Editors: Lawrence S. Chin William F. Regine

Lawrence S.
Chin
William F. Regine
Editors
Principles and
Practice of
Stereotactic
Radiosurgery
Second Edition
13
Principles and Practice of Stereotactic
Radiosurgery
Lawrence S. Chin • William F. Regine
Editors
Principles and Practice

of Stereotactic Radiosurgery
Second Edition
Editors
Lawrence S. Chin William F. Regine
Robert B. and Molly G. King Professor and Chair Isadore and Fannie Schneider Foxman Chair
Department of Neurosurgery and Professor of Radiation Oncology
SUNY Upstate Medical University Department of Radiation Oncology
Syracuse, NY, USA University of Maryland School of Medicine
Baltimore, MD, USA
ISBN 978-1-4614-8362-5 ISBN 978-1-4614-8363-2 (eBook)

DOI 10.1007/978-1-4614-8363-2
Springer New York Heidelberg Dordrecht London
Library of Congress Control Number: 2014952886
© Springer Science+Business Media New York 2015

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is
concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction
on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation,
computer software, or by similar or dissimilar methodology now known or hereafter developed. Exempted from this
legal reservation are brief excerpts in connection with reviews or scholarly analysis or material supplied specifically
for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work.
Duplication of this publication or parts thereof is permitted only under the provisions of the Copyright Law of the
Publisher’s location, in its current version, and permission for use must always be obtained from Springer. Permissions
for use may be obtained through RightsLink at the Copyright Clearance Center. Violations are liable to prosecution
under the respective Copyright Law.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not
imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and
regulations and therefore free for general use.
While the advice and information in this book are believed to be true and accurate at the date of publication, neither
the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may
be made. The publisher makes no warranty, express or implied, with respect to the material contained herein.
Printed on acid-free paper
Springer is part of Springer Science+Business Media (www.springer.com)

Contents
Part I Introduction
1 The History of Stereotactic Radiosurgery ............................................................. 3

Marina Kushnirsky, Vaibhav Patil, and Michael Schulder
2 Imaging Techniques in Stereotactic Radiosurgery ............................................... 11
Jessica Zhou, Yue Cao, James Balter, Neeraj Chaudhary, Aditya S. Pandey,
Greg Thompson, and Christina I. Tsien
3 Techniques of Stereotactic Localization ................................................................. 25
Victor C.K. Tse, M. Yashar S. Kalani, and John R. Adler
Part II Radiation Biology and Physics
4 The Physics of Stereotactic Radiosurgery ............................................................. 35

Siyong Kim and Jatinder Palta
5 Radiobiological Principles Underlying Stereotactic Radiation Therapy ............ 57
David J. Brenner and David J. Carlson
6 Treatment Planning for Stereotactic Radiosurgery .............................................. 73
David M. Shepard, Cedric Yu, Martin J. Murphy, Marc Bussière,
and Frank J. Bova
7 Designing, Building, and Installing a Stereotactic Radiosurgery Unit ............... 95
Lijun Ma, Andrew Hwang, and Arjun Sahgal
Part III Stereotactic Radiosurgery Techniques
8 Gamma Knife Radiosurgery ................................................................................... 111

Ajay Niranjan, Greg Bowden, John C. Flickinger, and L. Dade Lunsford
9 LINAC: Past, Present, and Future of Radiosurgery ............................................. 121
Maryam Rahman, Gregory J.A. Murad, Frank J. Bova,
and William A. Friedman
10 Charged Particles in Stereotactic Radiosurgery ................................................... 135
Shervin M. Shirvani and Joe Y. Chang
11 CyberKnife ............................................................................................................... 147
Carolina E. Fasola, Lei Wang, John R. Adler, Scott G. Soltys, Iris C. Gibbs,
Albert C. Koong, and Daniel T. Chang
12 GammaPod ............................................................................................................... 163
Yildirim D. Mutaf, Cedric Yu, Steven J. Feigenberg, and William F. Regine
v
vi Contents
13 Stereotactic Body Radiation Therapy .................................................................... 177

Steven J. Feigenberg, Randi Cohen, Navesh K. Sharma, Zain Husain,
Shifeng Chen, and Laura A. Dawson
Part IV Treatment of Disease Types
14 Metastatic Brain Tumors......................................................................................... 211

Edward W. Jung, John H. Suh, Samuel T. Chao, Michael A. Vogelbaum,
and Gene H. Barnett
15 Metastatic Brain Tumors: Viewpoint—Surgery .................................................. 233
Raymond Sawaya, David M. Wildrick, and Fassil B. Mesfin
16 Metastatic Brain Tumors: Viewpoint: Whole Brain Radiation Therapy ........... 241
William F. Regine, Sarah F. Grabowski, and Roy A. Patchell
17 Stereotactic Radiosurgery and Radiotherapy in the Management
of High-Grade Gliomas ........................................................................................... 249
David Roberge and Luis Souhami
18 Management of Glial Tumors: Viewpoint—Surgery
and Intra-cavitary Radiopharmaceutical Therapy .............................................. 269
Kaisorn L. Chaichana, Linda Chen, Salvador Manrique-Guzman,
Lawrence Kleinberg, and Alfredo Quinones-Hinojosa
19 Malignant Glioma: Viewpoint—Chemotherapy ................................................... 279
Roger Stupp, Krisztian Homicsko, and J. Gregory Cairncross
20 Meningioma .............................................................................................................. 295
Alessandra Gorgulho, Carlos A. Mattozo, and Antonio A.F. De Salles
21 Meningioma: Viewpoint—Surgery......................................................................... 315
Lawrence S. Chin, David J. Padalino, Pulak Ray, and John M. Caridi
22 Meningioma—Viewpoint: Fractionated Radiotherapy ........................................ 323
Igor J. Barani, Arie Perry, and C. Leland Rogers
23 Radiosurgery of Acoustic Schwannomas ............................................................... 339
John C. Flickinger, Hideyuki Kano, and L. Dade Lunsford
24 Acoustic Neuroma: Viewpoint—Surgery .............................................................. 347
Gabriel Zada and Steven L. Giannotta
25 Acoustic Tumors: Viewpoint—Stereotactic Radiotherapy .................................. 355
Christopher J. Farrell and David W. Andrews
26 Stereotactic Radiosurgery for Pituitary Adenomas .............................................. 369
Jason P. Sheehan and Brian Williams
27 Pituitary Tumors: Viewpoint—Surgery................................................................. 379
Khaled M. Krisht, William T. Couldwell, and Martin H. Weiss
28 Pituitary and Pituitary Region Tumors: Viewpoint—Fractionated
Radiation Therapy ................................................................................................... 391
Jonathan P.S. Knisely and Paul W. Sperduto
29 Pituitary Tumors: Viewpoint— Medical Therapy ................................................ 403
Rachel L. Hopkins
30 Pediatric Radiosurgery ............................................................................................ 409
Arthur K. Liu
Contents vii
31 Pediatric Disorders: Viewpoint—Surgery ............................................................. 415

Stephanie L. Da Silva and Mark D. Krieger
32 Pediatric Disorders: Viewpoint—Fractionated Radiotherapy ............................ 427
Thomas E. Merchant and Erin S. Murphy
33 Pediatric Brain Tumors: Viewpoint—Chemotherapy .......................................... 439
Nathan J. Robison
34 Pineal Region Tumors .............................................................................................. 445
Gregory P. Lekovic and Andrew G. Shetter
35 Pineal Region Tumors: Viewpoint—Surgery ........................................................ 459
Adam M. Sonabend, Alfred Ogden, and Jeffrey N. Bruce
36 Pineal Tumors: Viewpoint—Fractionated Radiation Therapy ........................... 469
William G. Rule and Steven E. Schild
37 Pineal Region Tumors: Viewpoint—Chemotherapy ............................................ 477
Christopher Dardis and Roy A. Patchell
38 Skull Base Tumors ................................................................................................... 483
Reinhart A.J. Sweeney and Matthias Guckenberger
39 Skull Base Tumors: Viewpoint—Surgery .............................................................. 499
Richard F. Schmidt, Smruti K. Patel, Robert W. Jyung, Jean Anderson Eloy,
and James K. Liu
40 Skull Base Tumors: Viewpoint—Fractionated Radiotherapy
or Stereotactic Radiotherapy .................................................................................. 517
René-O. Mirimanoff and Laura Negretti
41 Head and Neck Tumors ........................................................................................... 529
Daniel T.T. Chua
42 Head and Neck Tumors: Viewpoint—Surgery...................................................... 543
Scharukh Jalisi
43 Head and Neck Tumors: Viewpoint—Fractionated Radiation Therapy
and Chemotherapy................................................................................................... 549
Michael Rutenberg and Mohan Suntharalingam
44 Spinal Tumors and Radiosurgery ........................................................................... 563
Evangelia Katsoulakis, Ilya Laufer, and Yoshiya Yamada
45 Spinal Tumors: Viewpoint—Surgery ..................................................................... 571
Faiz U. Ahmad, Gabriel Zada, and Michael Y. Wang
46 Spinal Metastases: Viewpoint—Fractionated Radiation Therapy ...................... 583
Quynh-nhu Nguyen, Almon S. Shiu, and Eric L. Chang
47 Arteriovenous Malformation Radiosurgery .......................................................... 589
Bruce E. Pollock
48 Arteriovenous Malformations: Viewpoint—Surgery ........................................... 605
Eric M. Deshaies and Surasak Komonchan
49 Cerebral Arteriovenous Malformations Viewpoint: Endovascular
Therapy Perspective ....................................................................................................... 617
M. Yashar S. Kalani, Richard W. Williamson, Felipe C. Albuquerque,
and Cameron G. McDougal
viii Contents
50 Radiosurgery for Cavernous Malformations and Other Vascular Diseases ....... 623
Ajay Niranjan, Greg Bowden, John C. Flickinger, and L. Dade Lunsford
51 Cerebral Cavernous Malformations: Viewpoint—Surgery ................................. 637
Robert L. Dodd and Gary K. Steinberg
52 Trigeminal Neuralgia ............................................................................................... 649
Lawrence S. Chin, Seung S. Hahn, Shilpen Patel, Thomas Mattingly,
and Young Kwok
53 Trigeminal Neuralgia: Viewpoint—Surgery ......................................................... 659
Craig Rabb and Ahmed Cheema
54 Trigeminal Neuralgia: Viewpoint—Medical Management .................................. 665
Neil C. Porter
55 Stereotactic Radiosurgery for Movement Disorders ............................................ 671
Yoshinori Higuchi
56 Perspective on Radiosurgery Versus Conventional Surgery for
Movement Disorders ................................................................................................ 681
Zion Zibly, Andrew B. Shaw, John Y.K. Lee, and Ali R. Rezai
57 Movement Disorders: Viewpoint—Medical Therapy........................................... 695
Abraham N. Lieberman and Sara S. Dhanani
58 Radiosurgery for Drug-Resistant Epilepsies: State of the Art, Results,
and Perspectives ....................................................................................................... 699
Jean Régis, Romain Carron, Fabrice Bartolomei, and Patrick Chauvel
59 Epilepsy: Viewpoint—Surgery ............................................................................... 711
David Carter and Zulma Tovar-Spinoza
60 Epilepsy: Viewpoint—Medical ............................................................................... 723
Gregory K. Bergey
61 Stereotactic Radiosurgery for Psychiatric and Pain Disorders ........................... 731
Chun-Po Yen and Jason P. Sheehan
62 Ocular and Orbital Lesions..................................................................................... 743
Yan Michael Li, Gabriela Šimonová, Roman Lisčák, Josef Novotný Jr.,
Amit Singla, and Lawrence S. Chin
63 Ocular and Orbital: Viewpoint—Surgery ............................................................. 765
Bryant P. Carruth, Robert H. Hill, and Thomas A. Bersani
64 Ocular and Orbital Tumors: Viewpoint-Fractionated Radiation ....................... 773
Nicolas Girard, Françoise Mornex, and Alain Vighetto
Part V Patient Care and Socio-Economic Issues
65 Complications and Management in Radiosurgery ............................................... 785

Igor J. Barani and Minesh P. Mehta
66 Building a Radiosurgery Practice........................................................................... 807
Ali Farooqui and N. Scott Litofsky
67 Patient Care in Stereotactic Radiosurgery ............................................................ 817
Terri F. Biggins
Index .................................................................................................................................. 829

Contributors
John R. Adler, M.D. Department of Neurosurgery, Stanford University, Stanford, CA, USA
Varian Medical Systems, Stanford, CA, USA
Faiz U. Ahmad, M.D., M.Ch. Department of Neurosurgery, Emory University, Atlanta,
GA, USA
Felipe C. Albuquerque, M.D. Division of Neurological Surgery, Barrow Neurological
Institute, St. Joseph’s Hospital and Medical Center, Phoenix, AZ, USA
David W. Andrews, M.D. Department of Neurological Surgery, Thomas Jefferson University,
Philadelphia, PA, USA
James Balter, Ph.D. Department of Radiation Oncology, University of Michigan Hospital
and Health Systems, Ann Arbor, MI, USA
Igor J. Barani, M.D. Department of Radiation Oncology, University of California, San
Francisco, San Francisco, CA, USA
Gene H. Barnett, M.D., M.B.A. Department of Neurosurgery, Cleveland Clinic Foundation,
Cleveland, OH, USA
Fabrice Bartolomei, M.D., Ph.D. Department of Neurophysiology, Aix-Marseille
University—Timone University Hospital, Marseille, France
Gregory K. Bergey, M.D. Department of Neurology, Johns Hopkins University School of
Medicine, Baltimore, MD, USA
Thomas A. Bersani, M.D. Department of Ophthalmology and Otolaryngology, SUNY
Upstate Medical University, Syracuse, NY, USA
Terri F. Biggins, R.N., B.S.N., C.N.R.N. Department of Nursing, University of Maryland
Medical Systems, Baltimore, MD, USA
Frank J. Bova, Ph.D. Department of Neurosurgery, University of Florida, Gainesville,
FL, USA
Greg Bowden, M.D., M.Sc. Department of Neurological Surgery, University of Pittsburgh,
Pittsburgh, PA, USA
David J. Brenner, Ph.D., D.Sc. Center for Radiological Research, Columbia University
Medical Center, New York, NY, USA
Jeffrey N. Bruce, M.D. Department of Neurological Surgery, Brain Tumor Center, Columbia
University Medical Center, New York Presbyterian Hospital, New York, NY, USA
Marc Bussière, M.Sc. Department of Radiation Oncology, Massachusetts General Hospital,
Boston, MA, USA
ix
x Contributors
J. Gregory Cairncross, M.D. Southern Alberta Cancer Research Institute, University of

Calgary, Calgary, AB, Canada
Yue Cao, M.D. Department of Radiation Oncology, Radiology, and Biomedical Engineering,
University of Michigan Hospital and Health Systems, Ann Arbor, MI, USA
John M. Caridi, M.D. Department of Neurosurgery, Ichan School of Medicine at Mount
Sinai, New York, NY, USA
David J. Carlson, Ph.D., D.A.B.R. Department of Therapeutic Radiology, Yale University
School of Medicine, New Haven, CT, USA
Romain Carron, M.D., Ph.D. Department of Functional Neurosurgery, Aix-Marseille
Bryant P. Carruth, M.D. Department of Ophthalmology, SUNY Upstate Medical University,
Syracuse, NY, USA
David Carter, M.D., Ph.D. Department of Neurosurgery, SUNY Upstate Medical University,
Syracuse, NY, USA
Kaisorn L. Chaichana, M.D. Department of Neurosurgery, Johns Hopkins University School
of Medicine, Baltimore, MD, USA
Daniel T. Chang, M.D. Department of Radiation Oncology, Stanford University, Stanford,
CA, USA
Eric L. Chang, M.D. Department of Radiation Oncology, Keck School of Medicine of USC,
USC Norris Cancer Hospital, Los Angeles, CA, USA
Joe Y. Chang, M.D., Ph.D., M.S. Department of Radiation Oncology, The University of
Texas M.D. Anderson Cancer Center, Houston, TX, USA
Samuel T. Chao, M.D. Department of Radiation Oncology, Cleveland Clinic, Cleveland,
OH, USA
Neeraj Chaudhary, M.D., M.R.C.S., F.R.C.R. Department of Neuro-radiology, University
of Michigan Hospital and Health Systems, Ann Arbor, MI, USA
Patrick Chauvel, M.D. Department of Neurophysiology, Aix-Marseille University—Timone
University Hospital, Marseille, France
Ahmed Cheema, M.D. Department of Neurosurgery, University of Oklahoma-Oklahoma
City, Oklahoma City, OK, USA
Linda Chen, M.D. Department of Radiation Oncology, Johns Hopkins University School of
Shifeng Chen, Ph.D. Department of Radiation Oncology, University of Maryland School of
Lawrence S. Chin, M.D., F.A.C.S., F.A.A.N.S. Department of Neurosurgery, SUNY Upstate
Medical University, Syracuse, NY, USA
Daniel T.T. Chua, M.D., F.R.C.R., F.H.K.C.R. Department of Radiotherapy, Hong Kong
Sanatorium & Hospital, Happy Valley, Hong Kong
Randi Cohen, M.D., M.S. Department of Radiation Oncology, University of Maryland,
Baltimore, MD, USA
William T. Couldwell, M.D., Ph.D. Department of Neurosurgery, Clinical Neurosciences
Center, University of Utah, Salt Lake City, UT, USA
Contributors xi
Christopher Dardis, M.D. Barrow Neurological Clinics, Phoenix, AZ, USA

Laura A. Dawson, M.D., F.R.C.P.C. Department of Radiation Oncology, Princess Margaret
Cancer Centre, University of Toronto, Toronto, ON, Canada
Eric M. Deshaies Department of Neurosurgery, SUNY Upstate Medical University, Syracuse,
NY, USA
Sara S. Dhanani, M.D. Department of Neurology, Banner Sunreach Research Institute, Sun
City, AZ, USA
Robert L. Dodd, M.D., Ph.D. Department of Neurosurgery, Stanford University School of
Medicine, Stanford, CA, USA
Jean Anderson Eloy, M.D. Endoscopic Skull Base Surgery Program, Department of
Otolaryngology—Head and Neck Surgery, Rutgers New Jersey Medical School, Newark,
NJ, USA
Ali Farooqui, M.D. Division of Neurological Surgery, Department of Surgery, University of
Missouri School of Medicine, Columbia, MO, USA
Christopher J. Farrell, M.D. Department of Neurological Surgery, Thomas Jefferson
University, Philadelphia, PA, USA
Carolina E. Fasola, M.D. Department of Radiation Oncology, Stanford University, Stanford,
CA, USA
Steven J. Feigenberg, M.D. Department of Radiation Oncology, The University of Maryland
School of Medicine, Baltimore, MD, USA
John C. Flickinger, M.D. Department of Radiation Oncology, University of Pittsburgh,
Pittsburgh, PA, USA
William A. Friedman, M.D. Department of Neurosurgery, University of Florida, Gainesville,
FL, USA
Steven L. Giannotta, M.D. Department of Neurological Surgery, Keck School of Medicine
of USC, Los Angeles, CA, USA
Iris C. Gibbs, M.D. Department of Radiation Oncology, Stanford University, Stanford,
CA, USA
Nicolas Girard, M.D., Ph.D. Department of Respiratory Medicine, Oncology, Hospices
Civils de Lyon, Lyon, France
Alessandra Gorgulho, M.D., M.Sc. Department of Neurosurgery, Hospital de Coração-
Associação do Sanatório Sírio, São Paulo, Brazil
Department of Neurosurgery, University of California at Los Angeles, Los Angeles, CA, USA
Sarah F. Grabowski, M.D. Department of Radiation Oncology, University of Maryland
Medical Center, Baltimore, MD, USA
Matthias Guckenberger, M.D. Department of Radiation Oncology, University Würzburg,
Würzburg, Germany
Seung S. Hahn, M.D. Department of Radiation Oncology, SUNY Upstate Medical University,
Syracuse, NY, USA
Yoshinori Higuchi, M.D., Ph.D. Department of Neurological Surgery, Chiba University
Graduate School of Medicine, Chiba, Japan
Robert H. Hill, M.D. Department of Ophthalmology, SUNY Upstate Medical University,
Syracuse, NY, USA
xii Contributors
Krisztian Homicsko, M.D., Ph.D. Department of Oncology, Centre Hospitalier Universitaire

Vaudois, Lausanne, Switzerland
Rachel L. Hopkins, M.D. Division of Endocrinolgy & Metabolism, SUNY Upstate Medical
University, Syracuse, NY, USA
Zain Husain, M.D. Department of Therapeutic Radiology, Yale University, New Haven, CT, USA
Andrew Hwang, Ph.D. Department of Radiation Oncology, Alta Bates Summit Medical
Center, Oakland, CA, USA
Scharukh Jalisi, M.D., M.A. Division of Head and Neck Surgical Oncology and Skullbase
Surgery, Department of Otolaryngology and Neurological Sciences, Boston University,
Boston, MA, USA
Edward W. Jung, M.D. Department of Radiation Oncology, Cleveland Clinic Foundation,
Cleveland, OH, USA
Robert W. Jyung Department of Otolaryngology-Head and Neck Surgery, Rutgers New
Jersey Medical School, Newark, NJ, USA
M. Yashar S. Kalani, M.D., Ph.D. Division of Neurological Surgery, Barrow Neurological
Hideyuki Kano, M.D., Ph.D. Department of Neurological Surgery, University of Pittsburgh
School of Medicine, Pittsburgh, PA, USA
Evangelia Katsoulakis, M.D. Department of Radiation Oncology, Memorial Sloan Kettering
Cancer Center, New York, NY, USA
Siyong Kim, Ph.D. Department of Radiation Oncology, Virginia Commonwealth University,
Richmond, VA, USA
Lawrence Kleinberg, M.D. Department of Radiation Oncology, Johns Hopkins University
Jonathan P.S. Knisely, M.D. Department of Radiation Medicine, North Shore LIJ Health
System, Hofstra North Shore LIJ School of Medicine, Manhasset, NY, USA
Surasak Komonchan, M.D. Department of Neurology, Prasat Neurological Institute,
Rajthevee, Thailand
Albert C. Koong, M.D., Ph.D. Department of Radiation Oncology, Stanford University
School of Medicine, Stanford, CA, USA
Mark D. Krieger, M.D. Division of Neurosurgery, Children’s Hospital Los Angeles, Los
Angeles, CA, USA
Khaled M. Krisht, M.D. Department of Neurosurgery, Clinical Neurosciences Center,
University of Utah, Salt Lake City, UT, USA
Marina Kushnirsky, B.A. Department of Neurosurgery, Hofstra North Shore LIJ School of
Medicine, Manhasset, NY, USA
Young Kwok, M.D. Department of Radiation Oncology, University of Maryland Medical
Center, Baltimore, MD, USA
Ilya Laufer, M.D. Department of Neurosurgery, Memorial Sloan Kettering Cancer Center,
New York, NY, USA
John Y.K. Lee, M.D. Department of Neurosurgery, University of Pennsylvania, Philadelphia,
PA, USA
Contributors xiii
Gregory P. Lekovic, M.D., Ph.D. Good Samaritan Gama Knife Radiosurgery Center,
Los Angeles, CA, USA
Yan Michael Li, M.D., Ph.D. Department of Neurosurgery, UT MD Anderson Cancer Center,
Houston, TX, USA
Abrahim N. Lieberman, M.D. Department of Neurology, Barrow Neurological Clinics,
Phoenix, AZ, USA
Roman Lisčák, M.D. Third Faculty of Medicine, Clinical Department of Neurosurgery,
Charles University, Prague, Czech Republic
Department of Stereotactic and Radiation Neurosurgery, Na Homolce Hospital, Prague, Czech
Republic
N. Scott Litofsky, M.D., F.A.A.N.S., F.A.C.S. Division of Neurological Surgery, Department
of Surgery, University of Missouri School of Medicine, Columbia, MO, USA
Arthur K. Liu, M.D., Ph.D. Department of Radiation Oncology, University of Colorado
Cancer Center, Aurora, CO, USA
James K. Liu, M.D., F.A.A.N.S. Center for Skull Base and Pituitary Surgery, Rutgers New
Jersey Medical School Neurological Institute of New Jersey, Newark, NJ, USA
Departments of Neurological Surgery and Otolaryngology—Head and Neck Surgery, Rutgers
New Jersey Medical School Neurological Institute of New Jersey, Newark, NJ, USA
L. Dade Lunsford, M.D. Department of Radiation Oncology, University of Pittsburgh,
Pittsburgh, PA, USA
Lijun Ma, Ph.D. Department of Radiation Oncology, University of California San Francisco,
San Francisco, CA, USA
Salvador Manrique-Guzman, M.D. Department of Neurosurgery, Johns Hopkins Hospital,
Baltimore, MD, USA
Thomas Mattingly, M.D. Department of Neurosurgery, University of Maryland, Baltimore,
MD, USA
Carlos A. Mattozo, M.D. Department of Surgery, Division of Neurosurgery, David Geffen
School of Medicine at UCLA, Los Angeles, CA, USA
Assistant Professor of Neurosurgery, Department of Neurosurgery,Pontificia Universidade
Catolica de Curitiba, Brazil
Cameron G. McDougall, M.D. Division of Neurological Surgery, Barrow Neurological
Minesh P. Mehta, M.B.Ch.B., F.A.S.T.R.O. Department of Radiation Oncology, University
of Maryland, Baltimore, MD, USA
Thomas E. Merchant, D.O., Ph.D. Department of Radiological Sciences, St. Jude Children’s
Research Hospital, Memphis, TN, USA
Fassil B. Mesfin, M.D., Ph.D. Department of Neurosurgery, SUNY Upstate Medical
René-O. Mirimanoff, M.D. Radiation Oncology Department, CHUV, University Hospital
Lausanne, Lausanne, Switzerland
Clinique de La Source, Switzerland
Françoise Mornex, M.D., Ph.D. Department of Radiation Oncology, Centre Hospitalier
Lyon Sud, Hospices Civils de Lyon, Lyon, France
xiv Contributors
Gregory J.A. Murad, M.D. Department of Neurosurgery, University of Florida, Gainesville,

FL, USA
Erin S. Murphy, M.D. Department of Radiation Oncology, Taussig Cancer Institute,
Cleveland, OH, USA
Martin J. Murphy, Ph.D. Department of Radiation Oncology, Virginia Commonwealth
University, Richmond, VA, USA
Yildirim D. Mutaf, Ph.D. Department of Radiation Oncology, University of Maryland
Laura Negretti, M.D. Radiation Oncology Department, CHUV, University Hospital
Lausanne, Clinica Luganese, Ticino, Switzerland
Quynh–nhu Nguyen, M.D. Department of Radiation Oncology, Keck School of Medicine of
USC, USC Norris Cancer Hospital, Los Angeles, CA, USA
Ajay Niranjan, M.D., M.B.A. Department of Neurological Surgery, University of Pittsburgh,
Pittsburgh, PA, USA
Josef Novotný Jr., M.Sc., Ph.D. Department of Radiation Oncology, University of Texas
Southwestern Medical Center, Dallas, TX, USA
Alfred Ogden, M.D. Department of Neurological Surgery, Columbia University Medical
Center, New York Presbyterian Hospital, New York, NY, USA
David J. Padalino, M.D. Department of Neurosurgery, SUNY Upstate Medical University,
Syracuse, NY, USA
Jatinder Palta, Ph.D. Department of Radiation Oncology, Virginia Commonwealth
Aditya S. Pandey, M.D. Department of Neurosurgery, University of Michigan Hospital and
Health Systems, Ann Arbor, MI, USA
Roy A. Patchell, M.D. National Brain Tumor Center, Capital Institute for Neurosciences,
Capital Health Medical Center-Hopewell, Pennington, NJ, USA
Shilpen Patel, M.D. Department of Radiation Oncology, University of Washington Medical
Center, Seattle, WA, USA
Smruti K. Patel, B.A. Department of Neurological Surgery, University of Medicine and
Dentistry of New Jersey, New Jersey Medical School, Newark, NJ, USA
Vaibhav Patil, M.D., M.S. Department of Information Technology, Commonwealth Care
Alliance, Boston, MA, USA
Arie Perry, M.D. Department of Pathology and Neurological Surgery, University of California
San Francisco, San Francisco, CA, USA
Bruce E. Pollock, M.D. Department of Neurological Surgery and Radiation Oncology, Mayo
Foundation, Rochester, MN, USA
Neil C. Porter, M.D. Department of Neurology, University of Maryland School of Medicine,
Baltimore, MD, USA
Alfredo Quinones-Hinojosa, M.D. Department of Neurosurgery and Oncology, Johns
Hopkins University School of Medicine, Baltimore, MD, USA
Craig Rabb, M.D. Department of Neurosurgery, University of Oklahoma College of
Medicine, Oklahoma City, OK, USA
Contributors xv
Maryam Rahman, M.D., M.S. Department of Neurosurgery, University of Florida,

Gainesville, FL, USA
Pulak Ray, M.D. Department of Neurosurgery, Delaware Neurosurgical Group, Newark,
DE, USA
William F. Regine, M.D. Department of Radiation Oncology, University of Maryland
Medical Center, Baltimore, MD, USA
Jean Régis, M.D. Department of Functional Neurosurgery, Aix-Marseille University—
Timone University Hospital, Marseille, France
Ali R. Rezai, M.D. Department of Neurological Surgery and Neuroscience, The Ohio State
University, Columbus, OH, USA
David Roberge, M.D. Department of Radiation Oncology, Centre Hospitalier de l’Université
de Montréal—Notre Dame, Montreal, QC, Canada
Nathan J. Robison, M.D. Children’s Center for Cancer and Blood Diseases, Children’s
Hospital Los Angeles, University of Southern California Keck School of Medicine, Los
Angeles, CA, USA
C. Leland Rogers, M.D. GammaWest Cancer Services, Salt Lake City, UT, USA
William G. Rule, M.D. Department of Radiation Oncology, Mayo Clinic College of Medicine,
Mayo Clinic Arizona, Phoenix, AZ, USA
Michael Rutenberg, M.D., Ph.D. Department of Radiation Oncology, University of
Maryland, Baltimore, MD, USA
Arjun Sahgal, M.D. Department of Radiation Oncology, University of Toronto, Toronto,
ON, Canada
Antonio A.F. De Salles, M.D., Ph.D. Clinical and Scientific Division, Hospital do Coração
Neuroscience, Hospital de Coração-Associação do Sanatório Sírio, São Paulo, Brazil
Department of Neurosurgery, University of California at Los Angeles, Los Angeles, CA, USA
Raymond Sawaya, M.D. Department of Neurosurgery, The University of Texas
M.D. Anderson Cancer Center, Houston, TX, USA
Steven E. Schild, M.D. Department of Radiation Oncology, Mayo Clinic College of Medicine,
Mayo Clinic Arizona, Phoenix, AZ, USA
Richard F. Schmidt, B.A. Department of Neurological Surgery, Rutgers New Jersey Medical
School, Newark, NJ, USA
Michael Schulder, M.D. Department of Neurosurgery, Hofstra North Shore-LIJ School of
Medicine, Manhasset, NY, USA
Navesh K. Sharma, D.O., Ph.D. Department of Radiation Oncology, University of Maryland
Andrew B. Shaw, M.D. Department of Neurological Surgery and Neuromodulation, Wexner
Medical Center at The Ohio State University, Columbus, OH, USA
Jason P. Sheehan, M.D., Ph.D. Department of Neurological Surgery, University of Virginia,
Charlottesville, VA, USA
David M. Shepard, Ph.D. Swedish Cancer Institute, Seattle, WA, USA
Andrew G. Shetter, M.D. Section of Functional Stereotactic Neurosurgery, Division of
Neurological Surgery, Barrow Neurological Institute, Phoenix, AZ, USA
xvi Contributors
Shervin M. Shirvani, M.D. Department of Radiation Oncology, Banner MD Anderson

Cancer Center, Gilbert, AZ, USA
Almon S. Shiu, Ph.D. Department of Radiation Oncology, Keck School of Medicine of USC,
USC Norris Cancer Hospital, Los Angeles, CA, USA
Stephanie L. Da Silva, B.A. Division of Neurosurgery, Children’s Hospital Los Angeles, Los
Angeles, CA, USA
Gabriela Šimonová, M.D., Ph.D. Department of Stereotactic Radioneurosurgery, Hospital
Na Homolce, Prague, Czech Republic
Amit Singla, M.D. Department of Neurosurgery, SUNY Upstate Medical University,
Syracuse, NY, USA
Scott G. Soltys, M.D. Department of Radiation Oncology, Stanford University, Stanford,
CA, USA
Adam M. Sonabend, M.D. Department of Neurological Surgery, Columbia University
Medical Center, New York Presbyterian Hospital, New York, NY, USA
Luis Souhami, M.D., F.A.S.T.R.O. Division of Radiation Oncology, Department of
Oncology, McGill University, Montreal, QC, Canada
Paul W. Sperduto, M.D., M.P.P. Department of Radiation, University of Minnesota Gamma
Knife Center, Minneapolis Radiation Oncology, Waconia, MN, USA
Gary K. Steinberg, M.D., Ph.D. Department of Neurosurgery, Stanford University School of
Medicine, Stanford, CA, USA
Roger Stupp, M.D. Department of Oncology and Cancer Center, University Hospital Zurich,
Zurich, Switzerland
John H. Suh, M.D. Department of Radiation Oncology, Cleveland Clinic, Cleveland, OH, USA
Mohan Suntharalingam, M.D., M.B.A. Department of Radiation Oncology, University of
Maryland, Baltimore, MD, USA
Reinhart A.J. Sweeney, M.D., B.Sc. Department of Radiation Oncology, MVZ Innmed,
Klinik Bad Trissi, Oberaudorf, Germany
Greg Thompson, M.D. Department of Neurosurgery, University of Michigan Hospital
Zulma Tovar-Spinoza, M.D. Department of Neurosurgery, SUNY Upstate Medical
Victor C.K. Tse, M.D., Ph.D. Department of Neurosurgery, Kaiser Permanente Medical
Group, Redwood City, CA, USA
Christina I. Tsien, M.D. Department of Radiation Oncology, University of Michigan Hospital
Alain Vighetto, M.D. Service de neurologie, Hospital neurologique, Bron, France
Michael A. Vogelbaum, M.D., Ph.D. Department of Neurosurgery, Cleveland Clinic
Foundation, Cleveland, OH, USA
Lei Wang, Ph.D. Department of Radiation Oncology, Stanford University, Stanford, CA, USA
Michael Y. Wang, M.D. Department of Neurosurgery, University of Miami, Miami, FL, USA
Martin H. Weiss, M.D. Department of Neurological Surgery, Keck School of Medicine,
University of Southern California, Los Angeles, CA, USA
Contributors xvii
David M. Wildrick, Ph.D. Department of Neurosurgery, The University of Texas

Brian Williams, M.D. Department of Neurosurgery, University of Virginia, Charlottesville,
VA, USA
Richard W. Williamson, M.D. Division of Neurological Surgery, Barrow Neurological
Yoshiya Yamada, M.D. Department of Radiation Oncology, Memorial Sloan Kettering
Cancer Center, New York, NY, USA
Chun-Po Yen, M.D. Department of Neurological Surgery, University of Virginia,
Charlottesville, VA, USA
Cedric Yu, D.Sc. Department of Radiation Oncology, University of Maryland School of
Gabriel Zada, M.D. Department of Neurosurgery, Keck School of Medicine of USC,
University of Southern California, Los Angeles, CA, USA
Jessica Zhou, M.D. Department of Radiation Oncology, University of Michigan Hospital and
Health Systems, Ann Arbor, MI, USA
Zion Zibly, M.D. Department of Neurological Surgery, Center of Neuromodulation, The
Ohio State University Medical Center, Columbus, OH, USA
Part I
Introduction
The History of Stereotactic
Radiosurgery 1
Marina Kushnirsky, Vaibhav Patil, and Michael Schulder
surgery and fractionated stereotactic radiotherapy (FSRT)

Definition has become blurred. Sixty-two years since Leksell first
treated a patient with trigeminal neuralgia using an ortho-
Since its introduction in 1951, stereotactic radiosurgery voltage dental X-ray tube [4], SRS has changed so radically
(SRS) has become one of the most important technological that its definition has been reexamined. The argument in the
innovations in neurosurgery. Hundreds of thousands of first decade of the twenty-first century was framed thusly:
patients have undergone the procedure, which is now at the the “purists” stated that the unique radiobiological advances
cutting edge of radiation therapy techniques. This modality of SRS were lost once fractionation was incorporated, while
originated as the minimally invasive stereotactic application the “fractionators” claimed that it was overly rigid to not take
of various energy sources—ultrasound, orthovoltage, X-rays, advantage of the ability to fractionate, given the possibilities
and accelerated subatomic particles—for the treatment of opened up by increased computer speed and power.
movement and pain disorders [1]. SRS has since evolved into At a 2005 meeting of the American Association of
a precise, cost-effective, and noninvasive therapy for a wide Neurological Surgeons Stereotactic Radiosurgery Task
range of intracranial and extracranial pathologies. Advances Force, the members attempted to definitively separate
in imaging enable highly accurate treatment planning, and fractionated stereotactic radiosurgery and FSRT. A related
radiobiological innovations have improved beam targeting. goal was to agree with organized radiation oncology (repre-
The evolution of SRS has allowed for the treatment of an sented by ASTRO, the American Society for Therapeutic
ever-increasing range of oncological, cerebrovascular, and Radiation Oncology) on a definition of SRS as opposed to
functional targets, while simultaneously improving patient SRT. They dismissed radiobiological differences between
comfort and diminishing harmful side effects. the two modalities, stating they are unsubstantiated and theo-
Stereotactic radiosurgery was first defined in 1951 by the retical at best [1]. Instead, the meeting chose to focus on the
pioneering Swedish neurosurgeon Lars Leksell, who envi- intent of treatments as the key distinction between hypofrac-
sioned it as an alternative treatment during a time when mor- tionated SRS and FSRT. SRS uses steep radiation dose gra-
tality from open neurosurgery was close to 40% [2]. Leksell dients to ensure that only the target lesion is destroyed while
saw great potential in a technique that could use a single surrounding tissue is preserved. FSRT employs the differen-
heavy dose of radiation to “destroy any deep brain structure tial responses of tumors and normal tissue to fractionated
without the risk of bleeding or infection” [3]. His invention ionizing radiation. Thus, dose homogeneity is much more
predated advances in technology and radiobiology, which important for FSRT, in which abnormal and sensitive normal
now enable frameless, fractionated stereotactic localization. tissue may be in the same treatment field.
As a result of these advances, the boundary between radio- Ultimately, the task force defined SRS as “a distinct disci-
pline that utilizes externally generated ionizing radiation in
certain cases to inactivate or eradicate (a) defined target(s) in
M. Kushnirsky, B.A. • M. Schulder, M.D. (*) the head or spine without the need to make an incision. The
Department of Neurosurgery, Hofstra North Shore LIJ target is defined by high resolution stereotactic imaging.
School of Medicine, Manhasset, NY, USA
e-mail: mschulder@nshs.edu Stereotactic Radiosurgery (SRS) typically is performed in a
single session, using a rigidly attached stereotactic guiding
V. Patil, M.D., M.S.
Department of Information Technology, device, other immobilization technology and/or a stereotac-
Commonwealth Care Alliance, Boston, MA, USA tic image-guidance system, but can be performed in a limited
L.S. Chin and W.F. Regine (eds.), Principles and Practice of Stereotactic Radiosurgery, 3
DOI 10.1007/978-1-4614-8363-2_1, © Springer Science+Business Media New York 2015
4 M. Kushnirsky et al.
number of sessions, up to a maximum of five” [1]. As FSRT’s needed an improved method of intracranial targeting, and by
conformity and dose falloff improve, it may become neces- this time X-ray technology had sufficiently advanced to be
sary to rework the definition of SRS, perhaps including new practical and quick enough for use in the operating room.
information about radiobiological tissue response. Spiegel and Wycis employed X-ray ventriculography to
identify where a target lesion was in relation to any cerebral
structure. Imaging techniques combined with advances in
Early Roots physiology allowed for recognition of select target pathways:
the dorsomedial nucleus for psychosurgery, the lemniscal
Lars Leksell’s 1951 development of stereotactic radiosur- system for pain, and the extrapyramidal pathways for move-
gery was preceded by a multitude of technological innova- ment disorders [7]. In spite of these developments, the
tions in both stereotactic surgical fixation and medical Spiegel–Wycis frame, which consisted of a Horsley–Clarke
physics. The first stereotactic apparatus was described in device attached to patients’ skulls with a ring secured by a
1908 by Sir Victor Horsley and Robert Clarke [5]. It utilized plaster of Paris cap [5, 8], was not without drawbacks. During
Cartesian coordinates to locate structures within the brain, pneumoencephalography, cerebrospinal fluid was drained so
using bony landmarks for reference. Their work also included that air would enter the cerebral ventricles and allow for tar-
the first stereotactic atlas, which contained illustrated cross geting of brain structures using X-rays. The side effects were
sections of the monkey brain registered to distinguishing fea- so severe that patients were unable to undergo surgery on the
tures on the skull. The major issue with Horsley and Clarke’s same day their imaging and measurements were taken.
system was that intracranial anatomical structures vary sig- However, when positive contrast became available, the pro-
nificantly between individuals in their relationship to bony cedure became more easily tolerated. The Spiegel–Wycis
landmarks of the skull, making accurate targeting impossi- frame spurred the invention of at least 40 other stereotactic
ble. Interestingly, one of the engineers who worked on the devices [5], including Leksell’s.
apparatus designed a frame for human use. Unable to per- Concurrent radiobiological innovations played an equally
suade his neurosurgical colleagues to use it, he stored it in a important role as stereotactic ones in the development of
closet until his family discovered it, almost 60 years later [6]. SRS. Ionizing radiation was used for therapeutic purposes
The first stereotactic device accurate enough for use in soon after it was discovered; three months after Wilhelm
humans was developed in 1947 by Ernest A. Spiegel and Konrad Roentgen published his report in 1895, X-rays were
Henry T. Wycis (Fig. 1.1). This new stereotactic frame used to treat skin and breast cancers [3, 9]. The discovery of
radioactivity by Becquerel in 1896, and of radium by the
Curies soon after, opened the door for the medical use of
radioisotopes. X-rays were used to treat patients with pitu-
itary tumors as early as 1906, and radium brachytherapy was
applied to treat similar conditions at about the same time [10].
Harvey Cushing, the father of American neurosurgery, had
extensive experience with both X-ray and brachytherapy
treatments, although he remained skeptical of the utility of
either [11]. Dosimetry was poorly understood, and treat-
ments were not standardized. Nonetheless, some neurosur-
geons continued to explore the uses of ionizing radiation,
and new methods continued to be developed through the
1950s.
Emergence of Radiosurgical Practice
Lars Leksell conducted his earliest work with SRS while

serving as head of the neurosurgery department at the
University of Lund [12]. His preliminary work with a 250
kVp X-ray unit attached to a single beam port resulted in the
successful management of pain control in trigeminal neural-
gia patients (Fig. 1.2) [2, 5]. A ceaseless innovator, Leksell
Fig. 1.1 Henry T. Wycis holding Spiegel–Wycis stereotactic frame
noticed that if he used higher energy radiation, depth dose
with Horsley-Clarke frame on the table. From Textbook of Stereotactic
and Functional Neurosurgery. Lozano, Andres M.; Gildenberg, Philip L.; and beam definition could be improved, especially in smaller
Tasker, Ronald R. (Eds.) 2nd ed. 2009; used with permission treatment fields [12]. He then began working with Borje
1 The History of Stereotactic Radiosurgery 5
Fig. 1.4 Lars Leksell treating the first case of acoustic neurinoma with
Fig. 1.2 First stereotactic instrument for radiosurgery using 280 kV first version of gamma knife, 1968. From Textbook of Stereotactic and
X-rays, early 1950s. From Textbook of Stereotactic and Functional Functional Neurosurgery. Lozano, Andres M.; Gildenberg, Philip L.;
Neurosurgery. Lozano, Andres M.; Gildenberg, Philip L.; Tasker, Tasker, Ronald R. (Eds.) 2nd ed. 2009; used with permission
Ronald R. (Eds.) 2nd ed. 2009; used with permission
gamma knife (GK), containing 179 cobalt sources in a hemi-

spheric array (Fig. 1.4). The first unit was operational in
1968; Leksell’s first patient was immobilized using a molded
plaster headpiece and treated for a craniopharyngioma [4].
The potential of the GK to treat neoplasms and vascular
anomalies, as opposed to functional targets such as lesioning
for control of pain and movement disorders, was recognized
by Leksell early on. In the precomputed tomography (CT)
era these treatments were mostly limited to patients with
arteriovenous malformations (AVMs) [13] and acoustic neu-
romas, which could be imaged either on angiography or by
polytomography, respectively [3].
At the same time, work was continuing elsewhere with
focused heavy particle irradiation. John Lawrence, whose
brother Ernest won the 1939 Nobel Prize for his invention of
the cyclotron, began experimenting with charged particle
radiation in 1954. He treated patients with pituitary and other
intracranial disorders (Fig. 1.5) [14, 15] using proton and
helium ion beams. He initiated the use of the Bragg peak
principle according to which proton beams deposit their
energy at a distinct point, with minimal exit dose. In practice,
heavy particle beams must be carefully shaped and spread in
Fig. 1.3 Lars Leksell and his physicist colleague, Borje Larsson, pre- order to treat patients with intracranial lesions.
paring a patient for SRS with a particle beam accelerator in 1958. After visiting Stockholm in 1959, Raymond Kjellberg, a
(Photo courtesy of L. Dade Lundsford, MD) neurosurgeon at the Massachusetts General Hospital, began
the use of Bragg peak proton beam treatments in the USA
Larsson, a radiobiologist at the Uppsala University cyclotron [16]. He amassed a large series of patients with arteriove-
unit, to investigate the use of converging proton beams for nous malformations and pituitary tumors. Similar efforts
neurosurgery (Fig. 1.3). When these beams proved too com- were carried out in California with helium ions [17]. For
plicated and difficult to use in a hospital setting, Larsson and decades, the expense of building and maintaining a cyclotron
Leksell began collaborating with Kurt Liden, a medical has limited the use of heavy particle SRS to a few centers. In
physicist at Lund. the last few years, new technology has decreased the size and
After experimenting with particle beams and linear accel- cost of proton beam irradiation devices, and the number of
erators, Leksell and his colleagues ultimately designed the such facilities has grown.
electrons to almost the speed of light, then directing the elec-

tron beams to a heavy metal alloy. The resulting X-ray is
collimated and focused onto a target. Before inventing the
GK, Lars Leksell and his colleagues at the Karolinska
Institute experimented with LINAC, but abandoned the sys-
tem due to low photon energies and mechanical inaccuracy
[4]. However, the ability to use CT imaging for lesion local-
ization and beam targeting greatly increased the accuracy of
LINACs. These systems were a cheaper alternative to GK or
heavy particle accelerators [19] and were quickly adapted.
Working independently, in Buenos Aires and Vicenza, Italy,
respectively, Betti and Colombo reported the successful use
of LINACs for SRS [20, 21] in 1982. Their systems allowed
for the rotation of the LINAC gantry in a single plane.
At about the same time, Winston and Lutz described the
use of a commercially available stereotactic frame for
LINAC radiosurgery [22]. Following in their footsteps,
Loeffler and Alexander demonstrated how a LINAC dedi-
cated to SRS could be a practical alternative to a GK [23]. In
the late 1980s Friedman and Bova elected not to install the
second American GK unit, preferring to develop a new
LINAC SRS system [24]. Other advantages of these LINAC
Fig. 1.5 Raymond Kjellberg with a frame for proton beam therapy of systems, besides ubiquity and lower cost, included the avail-
a patient with an AVM. (Photo courtesy of Richard Wilson, Mallinckrodt
ability of collimators in a much greater variety of diameters
Research Professor of Physics, Harvard University)
than provided with the GK. This allowed for the use of single
isocenters when treating patients whose targets were over
Acceptance 18 mm in diameter, the width of the largest GK collimator.
However, at around the same time several GKs were installed
When SRS was first invented, stereotactic localization relied in several sites around the world.
on atlases, ventriculography, and angiograms. The advent of As clinical experience increased, publications appeared,
the computerized tomography (CT) scanner in the mid- indications broadened, and vendors became increasingly
1970s, and MRI some 10 years later, opened up the possibil- interested, a debate emerged regarding the merits of the
ity of direct targeting of tumors and other “soft tissue” targets Gamma Knife versus LINAC-based SRS. By now, clinical
inside the skull. Image-guided stereotaxy became available and physics studies indicate that SRS can be delivered effec-
for brain biopsy, craniotomy, tumor resection, and functional tively and accurately with either method [19, 25]. Numerous
neurosurgery. Due to this, the 1980s saw the evolution of reports demonstrating the efficacy of SRS with few if any
SRS from an esoteric technique, available at the original GK short-term complications and lower costs led to the prolifera-
in Stockholm (and as fractionated treatments at the few tion of GK and LINAC units around the world.
heavy particle accelerators around the world), to an emerg-
ing technology of increasing utility.
In 1984, after several years of intense regulatory reviews Fractionation, Extracranial Radiosurgery,
and logistical battles, Dade Lunsford and colleagues com- and Other Advances
pleted the installation of the first American GK at the
University of Pittsburgh [18]. This group was instrumental, As the Gamma Knife and LINAC systems continued to
via an ongoing series of peer-reviewed publications, in plac- improve treatment outcomes, it became clear that both tech-
ing the technique and clinical indications for SRS on a sound nologies had promise for wider application. Rapid advances
scientific basis. Since then, several new models of GK units in neuroimaging, robotic technology, and beam targeting
have become available. The most current high-end model, soon followed. John Adler, a neurosurgeon who trained at
the GK Perfexion, enables automated collimator changes the Brigham and Women’s Hospital in Boston, spent a
and improves access to intracranial targets. fellowship year with Lars Leksell in 1985 (Adler JR, per-
As the potential horizons of SRS broadened, other inves- sonal communication). Excited by his exposure to the GK,
tigators were able to adapt linear accelerators (LINACs) for Adler saw the potential of SRS being extended to other areas
SRS (Fig. 1.3) LINAC units create X-rays by accelerating of the body. This required a method of delivering focused
These and other articles have fostered a useful debate regard-

ing the concept of hypofractionation in SRS and indeed if
such treatments are still “radiosurgical” [30, 31].
Extracranial SRS
With frameless LINAC-based SRS and SRT, it became

possible to further step outside the bounds of traditional
radiosurgery and treat targets outside the brain. The first
radiosurgical moves out of the intracranial compartment
were in the logical direction of the skull base and past that
into the paranasal sinuses, using either GK [32, 33] or
LINAC units [34]. Creative modifications of standard
Fig. 1.6 The first Cyberknife treatment, 1994. (Photo courtesy of John
stereotactic systems, such as the vacuum pillow stabilizer
R. Adler, MD) described by Lax in 1994 [35], were designed to allow for
treatment of targets throughout the body. Hamilton and col-
radiation without a stereotactic frame. Partnering with leagues described the first truly extracranial radiosurgical
engineers at Stanford University and with private financial unit in 1995. This prototypical system relied on a skeletal
backing, the Cyberknife ultimately came into being in 1994 fixation frame and was designed to provide spinal SRS [36].
(Fig. 1.6). The need to surgically place a clamp on a spinous process
The Cyberknife delivers SRS via an X-band LINAC with and to treat the patient in a prone position limited the appeal
an output of 6 MV. It is nonetheless small enough to be of this groundbreaking concept. Non-CNS targets were later
mounted on an industrial robot, allowing for a theoretically treated using newer stereotactic technologies such as the
infinite number of beams to be aimed at the target. For treat- Elekta Stereotactic Body frame. However, with the advent of
ment of intracranial targets, the patient is immobilized using new frameless techniques and improvements in beam target-
a customized, molded plastic mask system that offers a high ing, such external devices for rigid fixation and localization
degree of reproducibility between multiple fractions. are no longer necessary. Table 1.1 summarizes the historical
Treatments are fashioned using an inverse planning method; landmarks in the development of SRS.
to allow for practical computation times, the number of beam
origins (“nodes”) and robot angles are limited. Various other
LINAC-based systems have been developed to improve dose The Role of Industry and Future Advances
planning and delivery through beam shaping and intensity
modulation. These systems include XKnife (Radionics), The convergence of image-guidance and radiation delivery
Novalis (BrainLAB), and Peacock (NOMOS Corporation). technologies initially encouraged the entry of multiple ven-
Single fraction SRS technologies have also been evolving; a dors into the SRS marketplace. This reflected the undeniable
rotating Gamma Knife system (RGS), invented in the 1990s, advantages of stereotactic localization and the resulting abil-
changes beam diameters during treatment without inter- ity to focus radiation treatments on the smallest possible vol-
changing collimator helmets, which increases the system’s ume. While GK was historically the gold standard in SRS, a
flexibility and decreases treatment time. variety of frame-based LINAC systems were designed to
In addition, CT scanners are now being incorporated into provide single fraction SRS beginning in the 1980s. Vendors
LINAC-based systems. The Tomotherapy Hi-Art system, included Radionics (X-Knife), Zmed (the University of
invented in the 1990s, provides integrated treatment plan- Florida system), BrainLab, and Fischer-Leibinger. As radia-
ning, patient setup, and CT-guided treatment. The newest tion oncologists accepted the concept of SRS, and neurosur-
Novalis TX system includes an on-board imaging cone beam geons accepted fractionation, the industry’s focus shifted
CT scan to improve visualization of soft tissue targets during toward frameless systems. As a result, Radionics and
treatment and an ExacTrac image-guidance system to adjust BrainLab adapted their LINAC-based SRS devices for fra-
for minute patient movements. These developments allow meless use, and Elekta created a frameless system that can be
for real-time updated imaging during treatment to ensure used to fractionate GK treatments.
positioning accuracy. Perhaps more significantly, this Additional technological advances in stereotactic devices,
improves intrafraction targeting accuracy as well. Peer- imaging modalities such as PET scans, and diffusion tensor
reviewed publications have demonstrated the acceptance of imaging (DTI) are rapidly altering methods for tumor analy-
the Cyberknife and Novalis frameless systems [26–29]. sis and surgical planning. DTI helps visualize nerve tracts,
Table 1.1 SRS landmarks Table 1.2 Peer-reviewed publication on SRS

Year Author Device/event Year Papers Citations
1947 Spiegel/Wycis First human SRS frame developed 1951–1960 1 7
1951 Leksell Invention of SRS with rotating 1961–1970 0 0
orthovoltage unit 1971–1980 17 402
1954 Lawrence Heavy particle treatment of pituitary 1981–1990 122 4,707
for cancer pain 1991–2000 1,024 20,056
1962 Kjellberg Proton beam therapy of intracranial 2001–2010 1,563 19,051
lesions
2011–present 505 477
1967 Leksell Invention of gamma knife
1970 Steiner GK SRS of AVMs
1980 Fabrikant Helium ion treatment of AVMs of the technique was to touch the desired structure with a
1982 Betti/Colombo LINACs adapted for SRS probe or electrode. Hence, the group chose the word
1984 Bunge Installation of commercial GK “Stereotactic,” a mongrel of the Greek stereo meaning “three
1986 Winston/Lutz LINAC SRS based on common dimensional” and the Latin tact meaning “to touch.”
stereotactic frame Neurosurgeons’ interest in SRS was slow to develop but
1992 Loeffler/Alexander Dedicated LINAC for SRS developed increased exponentially over time. In 1987, the year that the
1993 Mackie First tomotherapy system developed first American GK was installed at the University of
1994 Adler First Cyberknife treatment
Pittsburgh and early work on LINAC SRS had been pub-
1997 DeSalles, Solberg First Novalis shaped beam RS unit
installed
lished, there were no SRS-related presentations at the annual
2011 Sheehan et al. First Gamma Knife extend case studies meeting of the American Association of Neurological
Surgeons (AANS). By 1998 there were 31 such abstracts in
addition to practical course and seminars devoted to the
while PET produces three-dimensional images of processes topic. SRS has remained a key item of interest at the major
in the body. Together with fMRI, these technologies will annual meetings of the AANS and of the Congress of
likely improve the precision of SRS and SRT and may be Neurological Surgeons. In addition, the meetings of the
incorporated into the imaging suites of current units. American and World Societies for Stereotactic and
Nanotechnology, which can be used to detect tumor margins Functional Neurosurgery feature SRS as one of the main
or as a tool for radiation guidance [37], is also a major source topics. The number of peer-reviewed publications on SRS
of excitement in the industry. The advancement of SRS and has grown tremendously as well. From 1981 to 1990, 122
related techniques has been integral in defining the modality such papers were published; in 2011 alone, that number was
as a new standard in patient care. 532 (Table 1.2) [39].
As the field grew more interdisciplinary, the International
Stereotactic Radiosurgery Society (ISRS) was founded in
Organized Radiosurgery 1993 and held its first biannual meeting that year in
Stockholm. At first the papers presented dealt entirely with
Stereotactic societies have been critical in disseminating the treatment of intracranial conditions. As SRS has moved
information about radiosurgical advances. While today’s below the skull base, studies regarding patients with such
organizations host international symposiums with extensive conditions as tumors of the spine, lung, pancreas, and pros-
membership and attendance, their origins are much more tate have been included in the ISRS program. Thus, the
humble. After Spiegel and Wycis created their stereotactic expertise of clinicians in fields completely unrelated to neu-
device, neurosurgeons who were interested in learning the rosurgery is being applied to the study of SRS. Neurosurgeons
technique visited the inventors at their lab at Temple Medical comprise the single biggest specialty group in the organiza-
School in Philadelphia [38]. To facilitate the exchange of this tion, followed by radiation oncologists and medical physi-
information, the International Society for Research in cists. As interest in extracranial and non-neurosurgical SRS
Stereoencephalotomy was created in 1961. inevitably increased, the membership of ISRS will evolve to
Interestingly, the word stereotactic is a relatively recent reflect this. The ISRS publishes a peer-reviewed collection
product of organized neurosurgery. Originally, Horsley and of selected manuscripts from each meeting, entitled The
Clarke labeled their technique “stereotaxic,” meaning “three- Journal of Radiosurgery and SBRT. The growth of SRS and
dimensional arrangement” in Greek. However, when the its widespread acceptance have been such that yet another
International Society for Research in Stereoencephalotomy organization, the Radiosurgery Society, has begun to hold
voted on a new name in 1973, it was decided that the object annual meetings.
10. Hirsch O. Uber methoden der operativen behandlung von hypophy-

Conclusion sistumoren auf endonasalem wege. Arc Laryngol Rhinol. 1910;
24:129–77.
11. Schulder M, Loeffler J, Howes A, Alexander III E, Black P. The
In the 60 years since it was invented, it is clear that stereotactic radium bomb: Harvey Cushing and the interstitial irradiation of
radiosurgery has become a mainstream treatment for a wide gliomas. J Neurosurg. 1996;84:530–2.
variety of conditions. This is not surprising; due to innova- 12. Benedict SH, Bova FJ, Clark B, Goetsch SJ, Hinson WH, Leavitt
DD, et al. Anniversary paper: the role of medical physicists in
tions in stereotactic immobilization, imaging, and radiobiol- developing stereotactic radiosurgery. Med Phys. 2008;35(9):
ogy, the technique is now extremely precise, noninvasive, 4262–77.
well tolerated, and effective. Invented more than 60 years 13. Steiner L, Leksell L, Greitz T. Stereotaxic radiosurgery for cere-
ago, SRS is over half the age of its main “parent” special- bral arteriovenous malformations. Acta Chir Scand. 1972;138:
459–64.
ties, neurosurgery and radiation oncology. It is no longer 14. Kirn TF. Proton radiotherapy: some perspectives. JAMA. 1988;
considered an experimental or new treatment, but rather a 259(6):787–8.
continuously growing and advancing field. Acceptance by 15. Skarsgard LD. Radiobiology with heavy charged particles: a historical
neurosurgeons, surgical specialists, and radiation oncolo- review. Phys Med. 1998;14 Suppl 1:1–19.
16. Kjellberg RN, Abe M. Stereotactic Bragg peak proton beam ther-
gists means that as SRS evolves it will not be a technique apy. In: Lunsford LD, editor. Modern stereotactic neurosurgery.
for “radiosurgeons,” but one of the methods available to Boston: Martinus Nijhoff; 1988. p. 463–70.
treat patients. 17. Fabrikant J, Lyman J, Frankel K. Heavy charged particle Bragg
What remains to be seen is how the field will evolve in peak radiosurgery for intracranial vascular disorders. Radiat Res
Suppl. 1985;104:8244–58.
response to developments in imaging and nanotechnology. 18. Lunsford LD, Flickinger JC, Linder G, Maitz A. Stereotactic radio-
SRS has already seen one major revolution: once the genie surgery of the brain using the first United States 210 cobalt-60
of hypofractionation was let out of the radiosurgical bottle, source gamma knife. Neurosurgery. 1989;24:151–9.
fractionation became the norm among LINAC users, and 19. Podgorsak E, Pike G, Olivier A, Pla M, Souhami L. Radiosurgery
with high energy photon beams: a comparison among techniques.
GK created a frameless immobilization device in order to Int J Radiat Oncol Biol Phys. 1989;16:857–65.
fractionate treatments. The impetus toward more focused 20. Betti O, Derechinsky V. Hyperselective encephalic irradiation with
and less invasive treatments is likely to spur other such a linear accelerator. Acta Neurochir Suppl. 1984;33:385–90.
changes in the future. Clinical and technological advances 21. Columbo F, Benedetti A, Pozza F, Avanzo RC, Marchetti C,
Chierego G, et al. External stereotactic irradiation by linear accel-
will continue, and SRS will play an increasingly important erator. Neurosurgery. 1985;16(2):154–60.
part in the management of patients with a range of tumors 22. Winston KR, Lutz W. Linear accelerator as a neurosurgical tool for
and disorders. stereotactic radiosurgery. Neurosurgery. 1988;22:454–64.
23. Loeffler J, Shrieve D, Wen P, Fine H, Kooy H, Adessa A, et al.
Radiosurgery for intracranial malignancies. Semin Radiat Oncol.
1995;5:225–34.
References 24. Friedman W, Bova F. The University of Florida radiosurgery sys-
tem. Surg Neurol. 1989;32:334–42.
1. Barnett GH, Linskey ME, Adler JR, Cozzens JW, Friedman WA, 25. Luxton G, Petrovich Z, Joszef G, Nedzi L, Apuzzo M. stereotactic
Heilbrun MP, et al. Stereotactic radiosurgery—an organized radiosurgery: principles and comparison of treatment methods.
neurosurgery-sanctioned definition. J Neurosurg. 2007;106(1):1–5. Neurosurgery. 1993;32:241–59.
Epub 2007/01/24. 26. Rahimian J, Chen JT, Girvigian M, Miller M, Rahimian R. Frame-
2. Alexander E, Loeffler JS, Lunsford LD. Stereotactic radiosurgery. based and frameless accuracy of novalis® radiosurgery. In: De
New York: McGraw-Hill; 1993. p. 254. Salles AAF, Gorgulho A, Agazaryan N, Slotman B, Selch M,
3. Leksell L. Stereotactic radiosurgery. J Neurol Neurosurg Psychiatry. Burwick AJ, editors. Shaped beam radiosurgery. Berlin: Springer;
1983;46(9):797–803. Epub 1983/09/01. 2011. p. 37–46.
4. Niranjan A, Lunsford LD. Radiosurgery: where we were, are, and 27. Chang SD, Murphy M, Geis P, Martin DP, Hancock SL, Doty JR,
may be in the third millennium. Neurosurgery. 2000;46(3):531. et al. Clinical experience with image-guided robotic radiosurgery
5. Gildenberg PL, Krauss JK. History of stereotactic surgery. In: (the Cyberknife) in the treatment of brain and spinal cord tumors.
Lozano A, Gildenberg P, Tasker R, editors. Textbook of stereotactic Neurol Med Chir (Tokyo). 1998;38(11):780–3.
and functional neurosurgery. Berlin: Springer; 2009. p. 1–33. 28. Mehta VK, Lee QT, Chang SD, Cherney S, Adler Jr JR. Image
6. Olivier A, Bertrand G, Picard C. Discovery of the first human ste- guided stereotactic radiosurgery for lesions in proximity to the
reotactic instrument. Stereotact Funct Neurosurg. 1983;46(1–4): anterior visual pathways: a preliminary report. Technol Cancer Res
84–91. Treat. 2002;1(3):173–80.
7. Gildenberg PL. Spiegel and wycis—the early years. Stereotact 29. Ishihara H, Saito K, Nishizaki T, Kajiwara K, Nomura S, Yoshikawa
Funct Neurosurg. 2001;77(1–4):11–6. K, et al. CyberKnife radiosurgery for vestibular schwannoma.
8. Schulder M. Overview of radiosurgery technology. In: Lozano A, Minim Invasive Neurosurg. 2004;47(5):290–3.
Gildenberg P, Tasker R, editors. Textbook of stereotactic and func- 30. Pollock BE, Lunsford LD. A call to define stereotactic radiosur-
tional neurosurgery. Berlin: Springer; 2009. p. 867–96. gery. Neurosurgery. 2004;55(6):1371–3.
9. Schulder M, Rosen J. Therapeutic radiation and the neurosurgeon. 31. Adler Jr JR, Colombo F, Heilbrun MP, Winston K. Toward an
Neurosurg Clin N Am. 2001;12(1):91–100. viii. Epub 2001/02/15. expanded view of radiosurgery. Neurosurgery. 2004;55(6):1374–6.
32. Kondziolka D, Lunsford LD. Stereotactic radiosurgery for squa- 36. Hamilton A, Lulu B, Fosmire H, Stea B, Cassady J. Preliminary
mous cell carcinoma of the nasopharynx. Laryngoscope. 1991; clinical experience with linear accelerator-based spinal stereotactic
101(5):519–22. radiosurgery. Neurosurgery. 1995;36:311–9.
33. Firlik KS, Kondziolka D, Lunsford LD, Janecka IP, Flickinger 37. De Salles AAF, Sedrak M, Lemaire JJ. The future of radiosurgery
JC. Radiosurgery for recurrent cranial base cancer arising from the and radiotherapy. In: De Salles AAF, Gorgulho A, Agazaryan N,
head and neck. Head Neck. 1996;18(2):160–5. discussion 6. Slotman B, Selch M, Burwick AJ, editors. Shaped beam radiosur-
34. Kaplan ID, Adler JR, Hicks Jr WL, Fee WE, Goffinet gery. Berlin: Springer; 2011. p. 305–9.
DR. Radiosurgery for palliation of base of skull recurrences from 38. Gildenberg PL, Krauss JK. History of the stereotactic societies. In:
head and neck cancers. Cancer. 1992;70(7):1980–4. Lozano A, Gildenberg P, Tasker R, editors. Textbook of stereotactic
35. Lax I, Blomgren H, Näslund I, Svanström R. Stereotactic radio- and functional neurosurgery. Berlin: Springer; 2009. p. 35–44.
therapy of malignancies in the abdomen: methodological aspects. 39. Harzing AW. Publish or Perish. 2007. Available from http://www.
Acta Oncol. 1994;33(6):677–83. harzing.com/pop/htm.
Imaging Techniques in Stereotactic
Radiosurgery 2
Jessica Zhou, Yue Cao, James Balter, Neeraj Chaudhary,
Aditya S. Pandey, Greg Thompson, and Christina I. Tsien
Introduction Imaging and Treatment Planning

Considerations
Stereotactic radiosurgery (SRS) involves the delivery of a
very precise, focal dose of radiation to a target. Radiologic Conventional contrast-enhanced computed tomography
imaging is crucial for accurate delineation of the treatment (CT) scanning provides anatomic as well as electron density
volume in order to achieve maximal dose to the target and information with sufficient geometric accuracy to support
optimal sparing of the normal surrounding tissue. stereotactic targeting and to calculate dose. MR contrast-
Conventional imaging reflects only anatomic rather than enhanced T1-weighted and T2-weighted imaging provides
functional properties of the tumor. In contrast, metabolic and superior soft tissue contrast relative to imaging and has also
physiologic imaging provides in vivo measures of tumor tis- been routinely used in target volume definition for SRS of
sue properties. In this chapter, physiologic MR imaging tech- intracranial tumors.
niques including proton spectroscopy, blood volume, blood In order to combine the advantages of MRI and CT, it has
flow, vascular permeability, diffusion, and diffusion tensor been historically necessary to spatially register these images
imaging, as well as metabolic PET imaging, will be described. within the treatment planning system (Fig. 2.1). Image regis-
Although SRS treatment has been established for many years, tration permits accurate mapping of imaging data from other
recent advances in modern imaging combined with improved modalities to a single coordinate system, typically the treat-
radiotherapy techniques can provide better target definition, ment planning CT scan [1].
reduce normal tissue toxicity, and provide better assessment
of treatment response. This chapter describes the role of novel
MR and PET imaging in the treatment planning and posttreat- MRI Alone for SRS Simulation and Planning
ment follow-up of both malignant and benign central nervous
system (CNS) intracranial tumors. More recently, MRI has been used as a stand-alone modality
for SRS treatment planning. Traditionally 1.5 T MR scanners
have been used for this purpose. There is increasing interest
now in the use of 3 T MR scanners, which have a higher sig-
J. Zhou, M.D. • Y. Cao, M.D. • J. Balter, Ph.D.
nal to noise ratio and consequently higher spatial resolution
Department of Radiation Oncology, University of Michigan and shorter image acquisition time. A recent study by Zhang
Hospital and Health Systems, Ann Arbor, MI, USA et al. showed that the geometric accuracy achieved with 3 T
N. Chaudhary, M.D., M.R.C.S., F.R.C.R. MRI is comparable to 1.5 T MRI for SRS treatment planning [2].
Departments of Radiology & Neurosurgery, University of The advantage of MRI-based planning is that it removes
Michigan, Neuro-interventional Radiology, Ann Arbor, MI, USA systematic errors that may occur due to CT-MRI registration and
A.S. Pandey, M.D. • G. Thompson, M.D. eliminates radiation exposure from the CT scan. However,
Department of Neurosurgery, University of Michigan this must be balanced against potential disadvantages of
Hospital and Health Systems, Ann Arbor, MI, USA
omitting the CT scan including possible spatial distortion of
C.I. Tsien, M.D. (*) the MRI scan due to gradient field nonlinearity and magnetic
Department of Radiation Oncology, Washington University in
St. Louis, 4921 Parkview Place, Campus Box 8224,
field inhomogeneities, the lack of electron density informa-
St. Louis, MO 63110, USA tion used for dose calculation, and the difficulty in patient
e-mail: ctsien@radonc.wustl.edu setup verification during treatment delivery [3]. MRI simulation
12 J. Zhou et al.
Fig. 2.1 Split window display of image registration of MR and CT for a right parietal brain metastasis for SRS treatment planning in the axial
(right image), sagittal (middle image), and coronal (left image) views
may also require a larger bore for patient setup to allow for the RTOG 9005 trial showed significantly higher CNS toxic-
appropriate MRI compatible immobilization devices. ity in patients with larger tumor diameters [7]. Tumors with
Solutions to these shortcomings of MRI-based plans are diameters between 2.1 and 3 cm have a 7.3-fold increased
being developed. The lack of an identifiable relationship risk of developing toxicity over tumors with diameters of
between MR image values and electron densities has been 2 cm or less. Tumors between 3.1 and 4 cm had a 16-fold
addressed in a variety of ways. Early treatment planning increased risk. In general, SRS is recommended for tumors
using MRI alone involved assignment of bulk densities (e.g., less than 4 cm in size [7].
to bone and the rest of the brain), which resulted in a smaller The safety of SRS and the rates of post–treatment com-
(1–1.5 %) calculated dose difference between the MRI-based plication can also vary based on the location of the tumor
and a CT-based plans when compared to plans without bulk relative to the surrounding cerebral structures. In a review
density assignment. More recent work has included use of by Flickinger et al., the rate of symptomatic radiation
electron density atlases and statistical modeling of tissue necrosis was highest (threefold increased risk) for arterio-
composition to assign attenuation properties. Recent devel- venous malformation (AVM) lesions in the basal ganglia
opments which may benefit MRI–PET systems as well as and the brainstem (pons, midbrain, medulla) [8]. In con-
radiation oncology include ultrashort TE (UTE) imaging and trast, the risk of radiation injury remained low (even with
even more sophisticated attenuation models [4, 5]. These increased volumes) for lesions irradiated in the frontal and
methods may reduce uncertainties of MR planning in most temporal lobes.
regions of the brain to a fairly insignificant level in the near Similarly, the proximity of critical neuroanatomic struc-
future. One issue of note, however, is the potential for local tures to the targeted area is important to consider during SRS
distortion effects, e.g., due to air–tissue interfaces, which planning. Relative contraindications for SRS include close
may cause displacements of 1–2 mm over very short dis- proximity or involvement of the optic tract including the optic
tances from the interface (<1 cm) [6]. nerves and chiasm. The optic apparatus dose is generally lim-
ited to 10 Gy to limit the risk of optic neuropathy [9, 10].
In contrast, the remaining cranial nerves have somewhat
Radiologic and Anatomic Considerations higher dose tolerances [11, 12].
During Treatment Planning
The goal of SRS is not only to give a maximal dose of focal Physiologic MR Imaging and Metabolic
radiation to the target but also to prevent injury to the sur- PET Imaging in SRS
rounding normal tissue. Both tumor size and proximity to
surrounding structures need to be taken into consideration Physiologic MR and metabolic PET imaging can add to
during treatment planning. information obtained by morphological imaging and serve as
Tumor size dictates the feasibility of radiosurgery. a surrogate marker for specific biologic processes. These
Increasing tumor size results in an exponential increase in imaging techniques permit in vivo analysis of tumor tissue
the margin of normal brain tissue irradiated and is potentially properties including chemical composition, tumor vasculature,
associated with a higher risk of complications. Results from perfusion, and tumor cellularity [13].
2 Imaging Techniques in Stereotactic Radiosurgery 13
Assessing tumor response after SRS can be difficult due

to occurrence of early nonspecific imaging changes that can MR Spectroscopy Imaging
represent either recurrent or progressive tumor, or treatment-
related inflammatory and/or necrotic changes such as areas MR spectroscopy is a technique to detect proton metabolites
of contrast enhancement and edema on T1-weighted imag- in tissue, in vivo. This imaging technique provides informa-
ing. Underlying processes of radiation injury cause a tempo- tion regarding tumor proliferation, cell membrane break-
rary increase in contrast enhancement on MRI, termed down, neuronal activity, and tumor necrosis [15, 23]. Most
“pseudoprogression,” making the differentiation between commonly detected metabolites include choline-containing
true progression and radiation effects extremely difficult. compounds, creatine, lactate, lipid, and N-acetylaspartate.
Mechanisms that may contribute to radiation-induced neuro- Spectroscopic images are obtained with either two-
toxicity include vascular injury, glial and white matter dimensional or three-dimensional means of acquisition. This
changes, and immunological mechanisms. is achieved by mapping the concentration of each of the
Advanced MR imaging techniques are now becoming compounds within voxel sizes of approximately 1 cm3.
routinely available and will continue to play an increasingly Malignant tumors are characterized by an elevated choline to
important role in aiding in the precise definition of at risk N-acetyl aspartate (NAA) ratio due to greater cell membrane
target volumes as well as the assessment of treatment phospholipid turnover from increased tumor proliferation as
response [14–16]. In this section, a brief overview of these well as decreased NAA compared to the normal brain [24,
imaging techniques will be described. 25]. Several studies have demonstrated the utility of MR
spectroscopy in pretreatment evaluation and radiotherapy
planning as well as in differentiating radiation necrosis from
Dynamic Susceptibility Contrast MRI recurrent tumor [26–28].
Dynamic susceptibility contrast (DSC) T2*-weighted imag-

ing is the preferred method to map whole-brain perfusion MR Diffusion and Diffusion Tensor Imaging
properties [17–19]. Dynamic acquisition of T2*-weighted
MR imaging during intravenous injection of Gd-DTPA Diffusion MRI measures the mobility of water within tissues
allows estimations of cerebral (tumor) blood volume (CBV), at the cellular level and therefore can detect microenviron-
cerebral (tumor) blood flow (CBF), and mean bolus transit ment changes in tumor tissue. Changes in diffusion can occur
time (MTT). These parameters can be measured in the due to cell swelling, necrotic/apoptotic cell death after ther-
imaged volume by mathematical deconvolution of the arte- apy, or extracellular water space changes as a result of edema
rial input function (from a cerebral artery) and the tissue [29]. Changes in tumor such as cytolytic cell death following
response signal. These perfusion parameters provide assess- successful therapy lead to transient increases in regional dif-
ment of tumor viability, tumor vascular properties, and transfusion, which may be an important early indicator of treat-
port kinetics following therapy [13]. Several studies have ment response. The extent of directional diffusion can be
shown a correlation between cerebral blood volume and the estimated with an apparent diffusion coefficient (ADC).
histological grade of the tumor [20, 21]. Diffusion has traditionally been analyzed using average
ADC values throughout an entire tumor, which can signifi-
cantly underestimate regional changes following therapy
Dynamic Contrast-Enhanced MRI [30]. The parametric response map (PRMADC) has been
developed as a voxel-wise approach for evaluating ADC
Dynamic contrast-enhanced (DCE), MR imaging using changes and has been found to be an independent, early pre-
T1-weighted imaging provides a method for quantitative dictor of overall survival [31]. ADC maps are largely inde-
assessment of tumor vascular permeability. DCE MRI is pendent of the MRI system, vendor, and field strength.
obtained following dynamic acquisition over several minutes Therefore, it provides an accurate and noninvasive method of
during intravenous injection of a bolus of Gd-GTPA. performing longitudinal studies [32].
Estimations of vascular permeability with DCE MRI require Diffusion tensor imaging (DTI) is a diffusion technique
the application of pharmacokinetic models with an arterial that allows visualization of white matter architecture and has
input function and therefore are more complex than CBV been increasingly investigated as an imaging biomarker for
estimation. DCE MRI has been shown to be important in detecting physiological changes prior to any changes on con-
assessment of treatment response following both anti-angio- ventional imaging [33]. Previous studies have suggested that
genic agents and radiation [22]. changes in diffusion index of normal appearing brain white
14 J. Zhou et al.
Fig. 2.2 (a) MR diffusion tensor imaging and tractography illus- cingulum and fornix show significant decreases compared to the
trating a right parietal–occipital mass and the bilateral corticospi- corpus callosum 1 month following radiation, suggesting increased
nal tracts. Peritumoral edema appears to infiltrate and displace the radiation sensitivity. Preliminary data showing a correlation
posterior and superior part of the right corticospinal tract compared between early diffusivity changes and late decline in verbal recall
to the left. (b) Quantitative DTI analyses were performed on spe- suggest that DTI imaging may be a useful biomarker for late CNS
cific white matter structures including the posterior cingulate, for- toxicity [34]. Means and standard errors of the percentage decrease
nix, and corpus callosum. Fractional anisotropy (FA) values in the in FA 1 month post-RT are plotted
matter structures following radiation may also be an indica- increased protein transport mediated by L-type amino acid
tor for delayed radiation-induced neurotoxicity [34]. These carriers at the blood–brain barrier level compared with
studies examined changes in diffusion indices including normal brain tissue [41]. It is able to better differentiate
fractional anisotropy (FA) as an index of fiber integrity, mean tumor from background brain signals than 18F-labeled
diffusivity (MD) as an index of overall diffusivity, radial dif- 2-fluoro-2-deoxy-D-glucose (FDG)-PET, which is more dif-
fusivity (RD) as an index of demyelination, and axial diffu- ficult to interpret due to the high level of intrinsic glucose
sivity (AD) as an index of fiber degradation [35]. Surgical uptake in the brain.
series have demonstrated the utility of DTI in reducing the 18F-fluorothymidine (FLT)-PET uses the alternative
risk of morbidity of brain tumor resections near critical tracer FLT, which is a thymidine analogue that is incorporated
structures. Similar incorporation of DTI tractography imag- exclusively into DNA. This allows measurement of the activ-
ing in SRS may be valuable with regards to delineation of ity of cellular thymidine kinase, which increases several folds
normal tissue avoidance structures in treatment planning for as cells enter the S phase and begin DNA synthesis. Therefore,
lesions near eloquent regions (Fig. 2.2) [36]. Incorporation increased FLT uptake is a direct measure of cellular prolifera-
of tractography data in SRS treatment planning has been tion rate and correlates with Ki-67 staining [42].
shown in studies to lead to decreased dose to critical struc-
tures including the optic tract and pyramidal tracts [37].
Imaging Consideration in the SRS
Treatment of Malignant Tumors
Metabolic PET Imaging
Brain Metastases
With a number of tracer compounds available, PET and
SPECT allow for noninvasive measurements of tumor Whole-brain radiotherapy has been the mainstay of treat-
hypoxia, proliferation index, and markers of apoptosis [38, ment of brain metastasis. However, there has been a recent
39]. The cellular and physiological information obtained shift toward greater use of stereotactic radiosurgery in
may be used in combination with MR imaging, which still patients with a limited number of brain metastases and with
has relatively superior spatial resolution, to improve target controlled systemic disease in order to increase dose to
volume definition in radiotherapy planning [40]. tumor and minimize neurocognitive morbidity [43].
11C Methionine (MET)-PET visualizes increased radio- SRS target volume definition for brain metastasis is typi-
tracer uptake by metabolically active tumor. It is able to detect cally accomplished using conventional thin-cut CT or
increased cellular metabolism in brain tumors involving T1-weighted MR imaging with contrast enhancement.
Studies have reported on improved outcomes with the and following SRS [49]. ADC values for the group showing
addition of a 1–2 mm margin to account for microscopic dis- radiation-induced central necrosis was significantly higher
ease in the SRS treatment of brain metastases [44, 45]. The than those demonstrating recurrent tumor growth. Similar
additional margin may be necessary to account for possible results were noted in a prospective trial of 38 patients treated
infiltrative tumor growth beyond the enhancing border as with gamma knife radiosurgery for brain metastases with
well as the limited accuracy of delineation of the enhancing significant increases in mean ADC values indicative of stable
border based on current imaging methods. A study by disease and unchanged ADC levels indicative of tumor recur-
Baumert et al. evaluated specimens from 76 metastatic brain rence [50]. The changes in ADC occurred before any defini-
lesions and found that 63 % of the specimens showed tumor tive change in volume were evident on further imaging
infiltration beyond the metastases boundary, with small-cell studies. In another study, posttreatment ADC patterns com-
lung cancer and melanoma showing a maximum depth of pared with initial diffusion imaging were used to predict ulti-
infiltration of >1 mm and other histologies <1 mm [45]. mate treatment outcomes on serial MR imaging in 25
Non-small cell lung cancer, melanoma, and sarcoma also metastatic tumors [51]. The authors found that the normal-
showed the most number of infiltration zones outside the ized ADC patterns outperformed initial posttreatment tumor
boundary. Another study by Noel et al. found that the addi- volume in predicting long-term response to treatment.
tion of a 1 mm margin on MRI to account for microscopic At the University of Michigan, alterations in vascular vol-
disease extension for brain metastases improved the 2-year ume and permeability in brain metastases, as measured using
local control rate from 51 to 90 % [44]. dynamic contrast-enhanced MRI hve been analyzed to pre-
SRS is also frequently used as adjuvant treatment to max- dict subsequent volumetric response following radiation. In
imize local control following surgical resection, with a recent prospective study that evaluated 43 lesions from 20
reported control rates of approximately 80 % [46]. In plan- patients, results showed that the percentage decrease in the
ning treatment in the adjuvant setting, one also needs to take high-CBV-defined subvolumes of the tumor was greater in
into consideration changes in the size of the surgical cavity the group of responsive tumors than in the group of stable
when delineating the target volume. In a study by Atalar and progressive tumors (P < 0.007). Perfusion changes were
et al. on 68 lesions, the postresection cavity volume was a significantly better predictor for post-RT response than
smaller than the preresection tumor volume by a median changes in the gross tumor volume observed during the same
percent volume change of 29 % [47]. The authors found that time interval (P = 0.012), suggesting that physiological
the greatest volume change occurred immediately after sur- changes occur prior to volumetric changes [52].
gery during postoperative days 0–3, with no significant PET imaging, including FDG PET, has also been evaluated
change occurring up to 33 days after surgery. Thus, there for differentiating between recurrence and posttreatment
may not be a benefit to an extended waiting period prior to changes. Studies have shown that areas of radiation injury have
proceeding with SRS. A planning margin for microscopic lower glucose metabolism than normal brain tissue due to
extension has also been shown to improve control in the lower cellular attenuation [53]. Varying degrees of accuracy of
adjuvant setting. A study by Soltys et al. analyzed 76 resec- FDG-PET in detecting tumor recurrence versus radiation
tion cavities that were treated with radiosurgery [48]. Less necrosis have been reported. Several retrospective analyses
conformal treatment plans with a 2.4 mm margin of brain have shown sensitivity in detection of tumor recurrence
tissue around the cavity had a 100 % local control rate, com- ranging from 80 to 100 % and specificity higher than 80 %;
pared with 78 % for a 1.7 mm margin, 52 % for 1.5 mm, and however, these studies lacked histopathological validation [54, 55].
43 % for 0.8 mm. Toxicity was minimal and therefore the In a study of 15 patients with a histopathologically confirmed
inclusion of a 2 mm margin to improve local control was diagnosis, Thompson et al. showed that FDG-PET was
recommended. sufficiently sensitive to differentiate recurrent tumor from
radiation effect in only 43 % of the cases and was least accurate
when the lesion volume was less than 6 cm3 [56].
Assessment of Treatment Response MET–PET may provide a better method for differentiating
radiation necrosis from recurrent tumor. In 21 patients with
Novel imaging techniques in addition to conventional suspected recurrent metastatic brain tumor or radiation
contrast-enhanced MRI are now more frequently used to injury, MET-PET scans showed a sensitivity of 78 % and
assess brain metastasis response following radiosurgery. specificity of 100 % in detecting tumor recurrence [57]. In
MR diffusion imaging has been used in the evaluation of another study of 77 patients, a mean SUV ratio of lesion
SRS treatment response (Fig. 2.3). Several studies have uptake to contralateral normal gray matter uptake of greater
noted that longitudinal ADC values may be important in than 1.41 provided the best sensitivity (79 %) and specificity
predicting early tumor response to SRS treatment. In a (75 %) for metastatic brain tumor and a value of greater than
prospective study by Huang et al., 21 patients with 32 brain 1.58 provided the best sensitivity (75 %) and specificity
metastases were analyzed. ADCs were measured both prior (75 %) for recurrent glioma [58].
16 J. Zhou et al.
Fig. 2.3 (a) T1-weighted images (right), ADC map (middle), and (green), a region of interest corresponding to a brain metastases at
perfusion map (left) are shown at baseline prior to SRS for a right baseline (blue) and 2 weeks after SRS (red ). Quantitative ADC his-
parietal brain metastases (top row) and 1 month post SRS treatment togram analysis demonstrate the fractional tumor subvolume with
(bottom row). Qualitatively, there are significant decreases noted in low ADC appeared to increase 2 weeks after SRS consistent with a
tumor volume, diffusion, and perfusion 1 month after treatment. (b) nonresponsive or progressive tumor confirmed on subsequent fol-
Mean ADC histogram analyses performed for the normal tissue low-up imaging at 4 months
Recurrent Gliomas shown to potentially improve target delineation. Areas of con-

trast enhancement on T1-weighted conventional MRI do not
SRS has been increasingly used as a potential treatment always correspond to the most aggressive areas as reflected by
option for recurrent gliomas in combination with anti- high choline concentration. The extent of elevated choline has
angiogenic agents following an initial course of standard been found beyond the tumor volume defined by Gd enhance-
fractionation of RT in combination [59]. ment or hyperintense lesions on T2-weighted or FLAIR
Novel imaging techniques can be utilized in treatment plan- images [60, 61]. Several studies have suggested that elevated
ning, especially in cases where there is no tissue confirmation choline peaks correlate with higher tumor grade. In a study of
of recurrent disease (Fig. 2.4). MR spectroscopy has been 247 tissue specimens from 31 untreated patients with low- or
Fig. 2.4 Multimodality MR imaging for target volume definition for a defined by T1 post-gad (third image from left). This shows the potential
left temporal glioblastoma. T1 post-gadolinium notes a contrast-enhancing for perfusion MRI to identify aggressive subvolumes in gliomas [11]. C
lesion (far left image) with significant peritumoral edema shown on MR MET PET (far right image) also provides additional information for tar-
FLAIR imaging (second image from left). Perfusion map shows areas of get volume definition compared with conventional MR regarding tumor
high CBV (red) extending beyond the contrast-enhancing tumor rim as extension along the white matter tract
high-grade glioma, choline peak height normalized to contra- had relative CBV values from 0.21 to 0.71, while the recur-
lateral creatine and choline or ipsilateral NAA significantly rent tumor group had CBV values from 0.55 to 4.64. A
correlated with tumor cell density [62]. threshold value of 0.71 therefore optimized differentiation
The degree of spatial overlap between tumor volumes with between the two groups with a sensitivity of 91.7 % and
elevated choline and the radiosurgical target volume has been specificity of 100 %.
correlated with clinical outcome. A retrospective study in
patients with recurrent GBMs treated with SRS found an
improvement in survival if the target volume overlapped with Imaging Considerations in the SRS
the pretreatment MR spectroscopy metabolic lesion by 50 % Treatment of Benign Lesions
or greater (median OS 15.7 versus 10.4 months, P < 0.01) [63].
Newer imaging techniques are also crucial in posttreat- Acoustic Neuroma (Vestibular Schwannomas)
ment follow-up of recurrent gliomas as posttreatment
contrast-enhancing lesions often consist of a mixture of Acoustic neuromas are benign neoplasms derived from
tumor cells and tissue with radiation injury and therefore are Schwann cells of the myelin sheath that show a predilection
difficult to accurately differentiate noninvasively. for involvement of sensory nerves. Acoustic neuromas
MR spectroscopy is one modality that is being increas- account for approximately 5–8 % of intracranial tumors and
ingly used to distinguish the two in the posttreatment set- 80–90 % of cerebellopontine angle tumors. Retrospective
ting. However, there is currently no consensus in the data have shown excellent tumor control rates with stereotac-
literature on which calculated ratios can best differentiate tic radiosurgery [66]. A University of Pittsburgh study of 216
tumor recurrence from radiation injury. One study of 100 patients with unilateral acoustic neuromas treated with
tissue biopsy samples from 44 patients with grade 2-4 glio- Gamma knife (marginal tumor dose of 12–13 Gy) showed a
mas, a choline/NAA ratio cutoff threshold of 2.5, was able 10 year resection-free control rate of 98.3 %, low toxicity
to differentiate recurrent tumor from normal, edematous, with good rates of hearing preservation (Robertson–Gardner
and necrotic brain tissue, with an approximate 90 % sensi- serviceable hearing 45 %), and high preservation rates of tri-
tivity and 86 % specificity [64]. Another study showed the geminal and facial nerve function [67].
ability to differentiate tumor recurrence from radiation Contrast-enhanced MRI is used to evaluate treatment
injury in 27 of 28 patients by using a cutoff value of 1.8 or response. Following SRS treatment, clinical data suggest that
greater for the choline/NAA or choline/creatine ratio repre- acoustic neuromas may temporarily expand in size initially and
senting tumor recurrence [28]. then stabilize. Thus, a longer period of follow-up is required
MR perfusion imaging has also shown its utility in post before determining disease progression. A retrospective review
SRS treatment evaluation. One study prospectively evalu- of 75 patients showed that 23 % had evidence of pseudopro-
ated 42 image-guided tissue specimens from 13 patients gression with onset of enlargement at 6 months, of which 9 %
with high-grade glioma using threshold relative CBV val- remained larger than initial volume at last follow-up. No fur-
ues to distinguish recurrent tumor from radiation changes ther progression was seen beyond 24 months for all patients
[65]. The authors found that the treatment-related group [68]. Another study involving 87 patients showed that peak
18 J. Zhou et al.
tumor volume expansion was observed at 8.6 months after SRS in those patients who did not initially present with radiologic
[69]. At 5 years follow-up, the mean reduction in tumor volume or neurologic changes on the initial posttreatment scan. Nor
was 31 %. However, there were still 10 % of tumors that is serial repeat imaging recommended for those patients who
remained larger than their initial volume. respond well clinically unless neurological deficits emerge.
Considering the potential for pseudoprogression, follow- Imaging may be used for the assessment of potential
up recommendations include posttreatment MR imaging to radiation-related injury to the brainstem.
establish a baseline and at 24 months, followed by yearly
imaging. In general, salvage therapy is not recommended
before 2 years unless there is clinical deterioration. Arteriovenous Malformations
Hydrocephalus can also be seen on follow-up imaging, a
potential complication following SRS treatment of acoustic Arteriovenous malformations (AVMs) are abnormal commu-
neuromas. In a retrospective review of 444 patients treated nications between arteries and veins, usually associated with
with radiosurgery, symptomatic communicating hydroceph- high blood flow that results in blood vessel remodeling. The
alus developed in 5.6 % of patients, with a median time to incidence of AVMs is approximately 0.01–0.1 %, and present-
symptom development of 7 months [70]. Patient monitoring ing symptoms include hemorrhage, seizure, and headaches.
for development of hydrocephalus is recommended follow- The average annual risk of bleeding is 2–3 %, with a slightly
ing SRS treatment for up to 3–4 years. higher risk if there is a prior history of bleed (2–6 %) and
slightly lower risk if no prior bleed (1–3 %) [74]. AVMs are
graded based on the Spetzler–Martin scale, which reflects the
Trigeminal Neuralgia size and venous drainage, as well as eloquence of the adjacent
brain. Current treatment options include surgery, with opera-
Trigeminal neuralgia is a painful syndrome resulting in bility depending on AVM location, embolization, and SRS.
intense and often severe episodes of lancinating pain occur- Following SRS, clinical studies show obliteration rates of
ring in the distribution of cranial nerve V. Although benign, 60–90 % [75, 76]. In one study, 351 patients were treated to
trigeminal neuralgia can significantly diminish a patients’ a median dose of 20 Gy, and the rate of obliteration was
quality of life. Common treatment options include antiepi- 73 % by angiography, 86 % by MR alone, and 75 % overall
leptics, neuroleptics, opioids, and antidepressant medications. [77]. A persistent out-of-field nidus was seen in 18 % of pre-
When medication fails to alleviate the effects of trigeminal viously embolized and 5 % of nonembolized pts. Data on
neuralgia, available treatments include microvascular surgi- larger AVMs show decreasing obliteration and higher com-
cal decompression, ablation, and SRS. SRS has been highly plication rates with increasing volume.
successful in pain control, with excellent response in Although SRS is noninvasive, one disadvantage of SRS
70–90 % of patients [71]. In a prospective clinical trial, 100 compared with surgery is the potential for continued risk of
patients were treated with Gamma Knife radiosurgery to a hemorrhage in the 1–3 years after treatment. A retrospective
median dose of 85 Gy [72]. Of the total, 83% of the patients observational study of 500 patients on the rates of hemorrhage
were pain free at a minimum follow-up time of 12 months showed that, compared with the period between diagnosis
and 70 % had stopped taking medications during the study. and radiosurgery, the risk of hemorrhage decreased by 54 %
Post treatment complications were minimal, with facial par- during a median latency period of 2 years from treatment to
esthesia in 6 % and hypesthesia in 4 %. angiographic evidence of obliteration and by 88 % after
MRI is of paramount importance for the accurate planning angiographic obliteration [78].
of SRS treatment of trigeminal neuralgia. Studies have
shown relatively limited use of imaging in predicting effec- Cerebral Angiography
tive treatment outcome because the evidence of treatment Cerebral angiography has been the gold standard in the initial
benefit or treatment failure is most often readily apparent diagnosis and evaluation of AVMs, as part of radiosurgery
before enhancement occurs. Posttreatment follow-up with treatment planning, and in posttreatment residual AVM moni-
contrast-enhanced MRI taken at 6 months can provide for toring [79, 80]. It is a dynamic, real-time study that shows not
target confirmation due to the high prevalence of contrast only the presence or absence of an AVM but also the vascular
enhancement. The posttreatment imaging characteristics of transit time. It provides superior spatial resolution without
trigeminal neuralgia were assessed by Massager et al., who overlap of adjacent vascular structures as would be seen on CT
reported that at 6 months post-treatment, contrast enhance- or MRI. The image acquisition on cerebral angiography is
ment on T1-weighted MRI was seen in 83 % of trigeminal rapid and enables delineation of the arterial and venous phases
nerve roots [73]. However, they found no correlation between of the circulation, which allows better visualization of the
the occurrence of focal enhancement of the nerve root and AVM nidus and its shunt components. The potential complica-
pain outcome, or development of trigeminal dysfunction. tions of the invasive procedure include stroke, arterial dissec-
Repeated serial imaging is generally not considered helpful tion, and reactions to the contrast agent.
Fig. 2.5 (a) University of

Michigan Hybrid CT and biplane
digital subtraction angiography
(DSA) suite; (b) Intra-arterial CT
angiography showing the nidus
of a right-sided basal ganglia
AVM in both the axial (left
image) and coronal plane (right
image). (c) Deep venous drainage
is noted by the white arrow in the
coronal plane (right image)
Even though angiography is useful as a diagnostic tool in placement of endovascular catheters and subsequent capture
defining the nidus, feeding arteries and draining veins, it is of cross-sectional vascular imaging with an intra-arterial
limited in its use for radiosurgery treatment planning pur- CTA technique [83]. Together, the high spatial and temporal
poses because of its 2D dataset. This can lead to both over- resolution of biplane angiography can be combined with the
and underestimation of the AVM nidus in the target volume relatively high-contrast soft tissue resolution of CT imaging.
[81]. Alternate approaches include obtaining either intrave- Small cerebral AVMs are often not readily seen on IV-CTA
nous CTA or MRI angiography to provide a 3D view of or MRA and instead may require high-resolution intra-
the nidus [82]. Most recently, hybrid biplane angiography arterial CTA images that may provide better target volume
and CT-on-rails suites have been developed for the safe delineation for SRS treatment planning (Fig. 2.5).
20 J. Zhou et al.
AVM Radiosurgery Treatment Planning

The primary goal of AVM radiosurgery is to develop a plan References
with a target volume that conforms closely to the surface of
the AVM nidus while maintaining a steep dose gradient to 1. Petti PL, Kessler ML, Fleming T, Pitluck S, et al. An automated
image-registration technique based on multiple structure matching.
minimize the radiation dose to the surrounding brain. More
Med Phys. 1994;21(9):1419–26.
recently, tractography has been used in the AVM setting to 2. Zhang B, MacFadden D, Damyanovich AZ, Rieker M, Stainsby J,
identify better accuracy areas at risk in the brain that can Bernstein M, Jaffray DA, Mikulis D, Menard C. Development of a
cause patient morbidity if compromised [84]. A prospective geometrically accurate imaging protocol at 3T MRI for SRS treat-
ment planning. Phys Med Biol. 2010;55(22):6601–15.
study by Koga et al. integrated tractography in 71 AVM
3. Pinnaduwage D. MRI-based simulation and treatment planning: are
radiosurgery treatments where the tumor was less than we there yet? [Internet]. 2012 [updated 2012 Mar 17, cited 2013 Jan
1 cm from fiber tracts including the pyramidal tract, optic 22]. Available from: http://www.ucsfcme.com/2012/slides/MRO120
tract, and the arcuate fasciculus. Morbidity as a result of 01/16PinnaduwageMRSimulationAndPlanningAreWeThereYet.pdf.
4. Johansson A, Karlsson M, Nyholm T. CT substitute derived from MRI
treatment was minimized, with only one patient experienc-
sequences with ultrashort echo time. Med Phys. 2011;38(5):2708–14.
ing permanent and one experiencing transient pyramidal 5. Keereman V, Fierens Y, Broux T, De Deene Y, Lonneux M,
tract dysfunction [85]. Vandenberghe S. MRI-based attenuation correction for PET/MRI using
ultrashort echo time sequences. J Nucl Med. 2010;51(5):812–8.
6. Wang H, Balter J, Cao Y. Patient-induced susceptibility effect on
Post AVM Follow-Up Imaging geometric distortion of clinical brain MRI for radiation treatment
MR perfusion imaging has been used to assess response fol- planning on a 3T scanner. Phys Med Biol. 2013;58(3):465–77.
lowing SRS. In a study by Guo et al., perfusion relative cere- 7. Shaw E, Scott C, Souhami L, Dinapoli R, Kline R, Loeffler J,
bral blood volume (rCBV) and cerebral blood flow (rCBF) Farnan N. Single dose radiosurgical treatment of recurrent previ-
ously irradiated primary brain tumors and brain metastases: final
maps were assessed in 19 patients before and after radiosur-
report of RTOG protocol 90–05. Int J Radiat Oncol Biol Phys.
gery [86]. Prior to treatment, the ratio of rCBV and rCBF of 2000;47(2):291–8.
the ipsilateral hemisphere to the unaffected hemisphere was 8. Flickinger JC, Kondiolka D, Lunsford LD, Kassam A, Phuong LK,
found to be significantly higher in patients with AVM than Liscak R, Pollock B. Development of a model to predict permanent
symptomatic postradiosurgery injury for arteriovenous malforma-
those of healthy patients, consistent with a cerebral steal
tion patients. Int J Radiat Oncol Biol Phys. 2000;46(5):1143–8.
phenomenon. After treatment, starting at 6 months follow- 9. Tishler RB, Loeffler JS, Lunsford LD, Duma C, Alexander 3rd E,
up, the ratio of rCBV and rCBF of the AVM nidus to normal Kooy HM, Flickinger JC. Tolerance of cranial nerves of the cavern-
brain gradually decreased to 1.0. Decrease was also seen in ous sinus to radiosurgery. Int J Radiat Oncol Biol Phys.
1993;27(2):215–21.
the hemispheric perfusion ratios, reflecting reversal of steal
10. Leber KA, Bergloff J, Pendl G. Dose-response tolerance of the
toward normal perfusion. visual pathways and cranial nerves of the cavernous sinus to stereo-
Currently, the standard for posttreatment follow-up of tactic radiosurgery. J Neurosurg. 1998;88(1):43–50.
AVMs is MRI every year with cerebral angiography for con- 11. Morita A, Coffey RJ, Foote RL, Schiff D, Gorman D. Risk of injury
to cranial nerves after gamma knife radiosurgery for skull base menin-
firmation of obliteration of the nidus. An angiographic cure
giomas: experience in 88 patients. J Neurosurg. 1999;90(1):42–9.
requires that no nidus or shunting is identified on the follow- 12. Foote KD, Friedman WA, Buatti JM, Meeks SL, Bova FJ, Kubilis
up imaging study. PS. Analysis of risk factors associated with radiosurgery for ves-
tibular schwannoma. J Neurosurg. 2001;95(3):440–9.
13. Cao Y, Sundgren PC, Tsien CI, Chenevert TT, Junck L. Physiologic
and metabolic magnetic resonance imaging in gliomas. J Clin
Conclusion Oncol. 2006;24(8):1228–35.
14. Zakian KL, Koutcher JA, Ballon D, Hricak H, Ling
Novel functional and metabolic MRI and PET imaging pro- CC. Developments in nuclear magnetic resonance imaging and
spectroscopy: application to radiation oncology. Semin Radiat
vides the ability to analyze tumor tissue properties including
Oncol. 2001;11(1):3–15.
chemical composition, vasculature and perfusion, water 15. Nelson SJ. Multivoxel magnetic resonance spectroscopy of brain
mobility, permeability, hypoxia, and tumor proliferation. tumors. Mol Cancer Ther. 2003;2(5):497–507.
These imaging modalities have the potential to improve 16. Pardo FS, Aronen HJ, Kennedy D, Moulton G, Paiva K, Okunieff P,
Schmidt EV, Hochberg FH, Harsh GR, Fischman AJ. Functional
radiosurgery treatment planning and posttreatment monitor-
cerebral imaging in the evaluation and radiotherapeutic treatment
ing, including evaluation of differences in treatment efficacy planning of patients with malignant glioma. Int J Radiat Oncol Biol
between patients and heterogeneity of response within Phys. 1994;30(3):663–9.
tumors. Further investigation in randomized studies is war- 17. Paulson ES, Schmainda KM. Comparison of dynamic susceptibility-
weighted contrast-enhanced MR methods: recommendations for
ranted to better integrate these imaging techniques into rou-
measuring relative cerebral blood volume in brain tumors.
tine clinical practice. Radiology. 2008;249(2):601–13.
18. Maeda M, Itoh S, Kimura H, Iwasaki T, Hayashi N, Yamamoto K, 31. Moffat BA, Chenevert TL, Lawrence TS, Meyer CR, Johnson TD,
Ishii Y, Kubota T. Tumor vascularity in the brain: evaluation with Dong Q, Tsien C, Mukherji S, Quint DJ, Gebarski SS, Robertson
dynamic susceptibility-contrast MR imaging. Radiology. PL, Junck LR, Rehemtulla A, Ross BD. Physiologic diffusion map:
1993;189(1):233–8. a noninvasive MRI biomarker for early stratification of clinical brain
19. Thomsen H, Steffensen E, Larsson EM. Perfusion MRI (dynamic tumor response. Proc Natl Acad Sci U S A. 2005;102(15):5524–9.
susceptibility contrast imaging) with different measurement 32. Malyarenko D, Galban CJ, Londy FJ, Meyer C, Johnson TD,
approaches for the evaluation of blood flow and blood volume in Rehemtulla A, Ross BD, Chenevert TL. Multi-system repeatability
human gliomas. Acta Radiol. 2012;53(1):95–101. and reproducibility of apparent diffusion coefficient measurement
20. Roberts HC, Roberts TP, Brasch RC, Dillon WP. Quantitative mea- using an ice-water phantom. J Magn Reson Imaging. 2013;37(5):
surement of microvascular permeability in human brain tumors 1238–46.
achieved using dynamic contrast-enhanced MR imaging: correlation 33. Nagesh V, Tsien CI, Chenevert TL, Ross BD, Lawrence TS, Junck
with Histologic grade. AJNR Am J Neuroradiol. 2000;21(5):891–9. L, Cao Y. Radiation-induced changes in normal-appearing white
21. Provenzale JM, Wang GR, Brenner T, Petrella JR, Sorensen matter in patients with cerebral tumors: a diffusion tensor imaging
AG. Comparison of permeability in high-grade and low-grade brain study. Int J Radiat Oncol Biol Phys. 2008;70(4):1002–2010.
tumors using dynamic susceptibility contrast MR imaging. AJR 34. Chapman CH, Nagesh V, Sundgren PC, Buchtel H, Chenevert TL,
Am J Roentgenol. 2002;178(3):711–6. Junck L, Lawrence TS, Tsien CI, Cao Y. Diffusion tensor imaging
22. Morgan B, Thomas AL, Drevs J, Hennig J, Buchert M, Jivan A, of normal-appearing white matter as biomarker for radiation-
Horsfield MA, Mross K, Ball HA, Lee L, Mietlowski W, Fuxuis S, induced late delayed cognitive decline. Int J Radiat Oncol Biol
Unger C, O’Byrne K, Henry A, Cherryman GR, Laurent D, Dugan Phys. 2012;82(5):2033–40.
M, Marme D, Steward WP. Dynamic contrast-enhanced magnetic 35. Nazem-Zadeh MR, Jafari-Khouzani K, Davoodi-Bojd E, Jiang Q,
resonance imaging as a biomarker for the pharmacological response Soltanian-Zadeh H. Clustering method for estimating principal dif-
of PTK787/ZK 222584, an inhibitor of the vascular endothelial fusion directions. Neuroimage. 2011;57(3):825–38.
growth factor receptor tyrosine kinases, in patients with advanced 36. Yen PS, Teo BT, Chiu CH, Chen SC, Chiu TL, Su CF. White Matter
colorectal cancer and liver metastases: results from two phase I tract involvement in brain tumors: a diffusion tensor imaging analy-
studies. J Clin Oncol. 2003;21(21):3955–64. sis. Surg Neurol. 2009;72(5):464–9.
23. Preul MC, Caramanos Z, Collins DL, Villemure JG, Leblanc R, 37. Pantelis E, Papadakis N, Verigos K, Stathochristopoulou I, Antypas
Olivier A, Pokrupa R, Arnold DL. Accurate, noninvasive diagnosis C, Lekas L, Tzouras A, Georgiou E, Salvaras N. Integration of
of human brain tumors by using proton magnetic resonance spec- functional MRI and white matter tractography in stereotactic radio-
troscopy. Nat Med. 1996;2(3):323–5. surgery clinical practice. Int J Radiat Oncol Biol Phys.
24. Wald LL, Nelson SJ, Day MR, Noworolski SE, Henry RG, Huhn 2010;78(1):257–67.
SL, Chang S, Prados MD, Sneed PK, Larson DA, Wara WM, 38. Chao ST, Suh JH, Raja S, Lee SY, Barnett G. The sensitivity and
McDermott M, Dillon WP, Gutin PH, Vigneron DB. Serial proton specificity of FDG PET in distinguishing recurrent brain tumor
magnetic resonance spectroscopy imaging of glioblastoma multi- from radionecrosis in patients treated with stereotactic radiosur-
forme after brachytherapy. J Neurosurg. 1997;87(4):525–34. gery. Int J Cancer. 2001;96(3):191–7.
25. Negendank WG, Sauter R, Brown TR, Evelhoch JL, Falini A, 39. Barthel H, Cleij MC, Collingridge DR, Hutchinson OC, Osman S,
Gotsis ED, Heerschap A, Kamada K, Lee BC, Mengeot MM, He Q, Luthra SK, Brady F, Price PM, Aboagye EO. 3’-deoxy-3’-
Moser E, Padavic-Schaller KA, Sanders JA, Spraggins TA, Stillman [18F]fluorothymidine as a new marker for monitoring tumor
AE, Terwey B, Vogl TJ, Wicklow K, Zimmerman RA. Proton mag- response to antiproliferative therapy in vivo with positron emission
netic resonance spectroscopy in patients with glial tumors: a multi- tomography. Cancer Res. 2003;63(13):3791–8.
center study. J Neurosurg. 1996;84(3):449–58. 40. Grosu AL, Feldmann H, Dick S, Dzewas B, Nieder C, Gumprecht
26. Einstein DB, Wessels B, Bangert B, Fu P, Nelson AD, Cohen M, H, Frank A, Schwaiger M, Molls M, Weber WA. Implications of
Sagar S, Lewin J, Sloan A, Zheng Y, Williams J, Colussi V, Vinkler IMT-SPECT for postoperative radiotherapy planning in patients
R, Maciunas R. Phase II trial of radiosurgery to magnetic resonance with gliomas. Int J Radiat Oncol Biol Phys. 2002;54(3):842–54.
spectroscopy-defined high-risk tumor volumes in patients with 41. Torii K, Tsuyuguchi N, Kawabe J, Sunada I, Hara M, Shiomi
glioblastoma multiforme. Int J Radiat Oncol Biol Phys. S. Correlation of amino-acid uptake using methionine PET and his-
2012;84(3):668–74. tological classifications in various gliomas. Ann Nucl Med.
27. Huang J, Wang AM, Shetty A, Maitz AH, Yan D, Doyle D, Richey 2005;19(8):677–83.
K, Park S, Pieper DR, Chen PY, Grills IS. Differentiation between 42. Sato N, Suzuki M, Kuwata N, Kuroda K, Wada T, Beppu T, Sera K,
intra-axial metastatic tumor progression and radiation injury fol- Sasaki T, Ogawa A. Evaluation of the malignancy of glioma using
lowing fractionated radiation therapy or stereotactic radiosurgery 11C-Methionine positron emission tomography and proliferating
using MR spectroscopy, perfusion MR imaging or volume progres- cell nuclear antigen staining. Neurosurg Rev. 1999;22(4):210–4.
sion modeling. Magn Reson Imaging. 2011;29(7):993–1001. 43. Kocher M, Soffietti R, Abacioglu U, Villa S, Fauchon F, Baumert
28. Weybright P, Sundgren PC, Maly PV, Hassan DG, Nan B, Rohrer S, BG, Fariselli L, Tzuk-Shina T, Kortmann RD, Carrie C, Ben Hassel
Junck L. Differentiation between brain tumor recurrence and radia- M, Kouri M, Valeinis E, van den Berge D, Collette S, Collette L,
tion injury using MR spectroscopy. AJR Am J Roentgenol. Mueller RP. Adjuvant whole-brain radiotherapy versus observation
2005;185(6):1471–6. after radiosurgery or surgical resection of one to three cerebral
29. Hamstra DA, Galban CJ, Meyer CR, Johnson TD, Sundgren PC, Tsien metastases: results of the EORTC 22952–26001 study. J Clin
C, Lawrence TS, Junck L, Ross DJ, Rehemtulla A, Ross BD, Chenevert Oncol. 2011;29(2):134–41.
TL. Functional diffusion map as an early imaging biomarker for high- 44. Baumert BG, Rutten I, Dehing-Oberije C, Twijnstra A, Dirx MJ,
grade glioma: correlation with conventional radiologic response and Debougnoux-Huppertz RM, Lambin P, Kubat B. A pathology-
overall survival. J Clin Oncol. 2008;26(20):3387–94. based substrate for target definition in radiosurgery of brain metas-
30. Galban CJ, Chenevert TL, Meyer CR, Tsien C, Lawrence TS, tases. Int J Radiat Oncol Biol Phys. 2006;66(1):187–94.
Hamstra DA, Junck L, Sundgren PC, Johnson TD, Ross DJ, 45. Noel G, Simon JM, Valery CA, Cornu P, Boisserie G, Hasboun D,
Rehemtulla A, Ross BD. The parametric response map is an imag- Ledu D, Tep B, Delattre JY, Marsault C, Baillet F, Mazeron
ing biomarker for early cancer treatment outcome. Nat Med. JJ. Radiosurgery for brain metastasis: impact of CTV on local con-
2009;15(5):572–6. trol. Radiother Oncol. 2003;68(1):15–21.
22 J. Zhou et al.
46. Robbins JR, Ryu S, Kalkanis S, Cogan C, Rock J, Movsas B, Kim 62. Croteau D, Scarpace L, Hearshen D, Gutierrez J, Fisher JL, Rock
JH, Rosenblum M. Radiosurgery to the surgical cavity as adjuvant JP, Mikkelsen T. Correlation between magnetic resonance spectros-
therapy for resected brain metastasis. Neurosurgery. 2012;71(5): copy imaging and image-guided biopsies: semiquantitative and
937–43. qualitative histopathological analyses of patients with untreated
47. Atalar B, Choi CY, Harsh 4th GR, Chang SD, Gibbs IC, Adler JR, glioma. Neurosurgery. 2001;49(4):823–9.
Soltys SG. Cavity volume dynamics after resection of brain metas- 63. Chan AA, Lau A, Pirzkall A, Chang SM, Verhey LJ, Larson D,
tases and timing of postresection cavity stereotactic radiosurgery. McDermott MW, Dillon WP, Nelson SJ. Proton magnetic reso-
Neurosurgery. 2013;72(2):180–5. nance spectroscopy imaging in the evaluation of patients undergo-
48. Soltys SG, Adler JR, Lipani JD, Jackson PS, Choi CY, Puataweepong ing gamma knife surgery for Grade IV glioma. J Neurosurg.
P, White S, Gibbs IC, Chang SD. Stereotactic radiosurgery of the 2004;101(3):467–75.
postoperative resection cavity for brain metastases. Int J Radiat 64. Burtscher IM, Skagerberg G, Geijer B, Burtscher IM, Skagerberg
Oncol Biol Phys. 2008;70(1):187–93. G, Geijer B, Englund E, Stahlberg F, Holtas S. Proton MR spectros-
49. Huang CF, Chou HH, Yang MS, Lee JK, Lin LY. Diffusion mag- copy and preoperative diagnostic accuracy: an evaluation of intra-
netic resonance imaging as an evaluation of the response of brain cranial mass lesions characterized by stereotactic biopsy findings.
metastases treated by stereotactic radiosurgery. Surg Neurol. AJNR Am J Neuroradiol. 2000;21(1):84–93.
2008;69(1):62–8. 65. Hu LS, Baxter LC, Smith KA, Feuerstein BG, Karis JP, Eschbacher
50. Huang CF, Chiou SY, Wu MF, Tu HT, Liu WS, Chuang JC. Apparent JM, Coons SW, Nakaji P, Yeh RF, Debbins J, Heiserman
diffusion coefficients for evaluation of the response of brain tumors JE. Relative cerebral blood volume values to differentiate high-
treated by Gamma Knife surgery. J Neurosurg. 2010;113(Suppl): grade glioma recurrence from post-treatment radiation effect:
97–104. direct correlation between image-guided tissue histopathology and
51. Goldman M, Boxerman JL, Rogg JM, Noren G. Utility of apparent localized dynamic susceptibility-weighted contrast-enhanced per-
diffusion coefficient in predicting the outcome of Gamma Knife- fusion MR imaging measurements. AJNR Am J Neuroradiol.
treated brain metastases prior to changes in tumor volume: a pre- 2009;30(3):552–8.
liminary study. J Neurosurg. 2006;105(Suppl):175–82. 66. Meijer OW, Vandertop WP, Baayen JC, Slotman BJ. Single-fraction
52. Farjam R, Tsien CI, Feng FY, Gomez-Hassan D, Hayman JA, vs. fractionated linac-based stereotactic radiosurgery for vestibular
Lawrence TS, Cao Y. Physiological imaging-defined response- schwannoma: a single-institution study. Int J Radiat Oncol Biol
driven subvolumes of a tumor. Int J Radiat Oncol Biol Phys. Phys. 2003;56(5):1390–6.
2013;85(5):1383–90. 67. Chopra R, Kondziolka D, Niranjan A, Lunsford LD, Flickinger
53. Di Chiro G, Oldfield E, Wright DC, De Michele D, Katz DA, JC. Long-term follow-up of acoustic schwannoma radiosurgery
Patronas NJ, Doppman JL, Larson SM, Ito M, Kufta CV. Cerebral with marginal tumor doses of 12 to 13 Gy. Int J Radiat Oncol Biol
necrosis after radiotherapy and/or intraarterial chemotherapy for Phys. 2007;68(3):845–51.
brain tumors: PET and neuropathologic studies. AJR Am J 68. Hayhurst C, Zadeh G. Tumor pseudoprogression following radio-
Roentgenol. 1988;150(1):189–97. surgery for vestibular schwannoma. Neuro Oncol.
54. Kim EE, Chung S-K, Haynie TP, Kim CG, Cho BJ, Podoloff DA, 2012;14(1):87–92.
Tilbury RS, Yang DJ, Yung WK, Moser Jr RP, et al. Differentiation 69. Nagano O, Serizawa T, Higuchi Y, Matsuda S, Sato M, Yamakami
of residual or recurrent tumors from post-treatment changes in F-18 I, Okiyama K, Ono J, Saeki N. Tumor shrinkage of vestibular
FDG-PET. Radiographics. 1992;12(2):269–79. schwannomas after Gamma Knife surgery: results after more than 5
55. Ogawa T, Kanno I, Shishido F, Inugami A, Higano S, Fujita H, years of follow-up. J Neurosurg. 2010;113(Suppl):122–7.
Murakami M, Uemura K, Yasui N, Mineura K. Clinical value of 70. Han JH, Kim DG, Chung HT, Paek SH, Park CK, Kim CY, Hwang
PET with [18F]fluorodeoxyglucose and L-methyl-11C-methionine SS, Park JH, Kim YH, Kim JW, Kim YH, Song SW, Kim IK, Jung
for diagnosis of recurrent brain tumor and radiation injury. Acta HW. The risk factors of symptomatic communicating hydrocepha-
Radiol. 1991;32(3):197–202. lus after stereotactic radiosurgery for unilateral vestibular schwan-
56. Thompson TP, Lunsford LD, Kondziolka D. Distinguishing recur- noma: the implication of brain atrophy. Int J Radiat Oncol Biol
rent tumor and radiation necrosis with positron emission tomogra- Phys. 2012;84(4):937–42.
phy versus stereotactic biopsy. Stereotact Funct Neurosurg. 71. Dhople AA, Adams JR, Maggio WW, Naqvi SA, Regine WF,
1999;73(1–4):9–14. Kwok Y. Long-term outcomes of Gamma Knife radiosurgery for
57. Tsuyuguchi N, Sunada I, Iwai Y, Yamanaka K, Tanaka K, Takami classic trigeminal neuralgia: implications of treatment and critical
T, Otsuka Y, Sakamoto S, Ohata K, Goto T, Hara M. Methionine review of the literature. J Neurosurg. 2009;111(2):351–8.
positron emission tomography of recurrent metastatic brain tumor 72. Regis J, Metellus P, Hayashi M, Roussel P, Donnet A, Bille-Turc
and radiation necrosis after stereotactic radiosurgery: is a differen- F. Prospective controlled trial of gamma knife surgery for essential
tial diagnosis possible? J Neurosurg. 2003;98(5):1056–64. trigeminal neuralgia. J Neurosurg. 2006;104(6):913–24.
58. Terakawa Y, Tsuyuguchi N, Iwai Y, Yamanaka K, Higashiyama S, 73. Massager N, Abeloos L, Devriendt D, Op de Beeck M, Levivier
Takami T, Ohata K. Diagnostic accuracy of 11C-methionine PET M. Clinical evaluation of targeting accuracy of gamma knife radio-
for differentiation of recurrent brain tumors from radiation necrosis surgery in trigeminal neuralgia. Int J Radiat Oncol Biol Phys.
after radiotherapy. J Nucl Med. 2008;49(5):694–9. 2007;69(5):1514–20.
59. Cuneo KC, Vredenburgh JJ, Sampson JH, Reardon DA, Desjardins 74. Pollock BE, Flickinger JC, Lunsford LD, Bissonette DJ, Kondziolka
A, Peters KB, Friedman HS, Willett CG, Kirkpatrick JP. Safety and D. Factors that predict the bleeding risk of cerebral arteriovenous
efficacy of stereotactic radiosurgery and adjuvant bevacizumab in malformations. Stroke. 1996;27(1):1–6.
patients with recurrent malignant gliomas. Int J Radiat Oncol Biol 75. Colombo F, Pozza F, Chierego G, Casentini L, De Luca G,
Phys. 2012;82(5):2018–24. Francescon P. Linear accelerator radiosurgery of cerebral arterio-
60. Li X, Lu Y, Pirzkall A, McKnight T, Nelson SJ. Analysis of the venous malformations: an update. Neurosurgery. 1994;34(1):
spatial characteristics of metabolic abnormalities in newly diag- 14–20.
nosed glioma patients. J Magn Reson Imaging. 2002;16(3):229–37. 76. Ellis TL, Friedman WA, Bova FJ, Kubilis PS, Buatti M. Analysis of
61. Pirzkall A, Li X, Oh J, Chang S, Berger MS, Larson DA, Verhey LJ, treatment failure after radiosurgery for arteriovenous malforma-
Dillon WP, Nelson SJ. 3D MRSI for resected high-grade gliomas tions. J Neurosurg. 1998;89(1):104–10.
before RT: tumor extent according to metabolic activity in relation 77. Flickinger JC, Kondziolka D, Maitz AH, Lunsford LD. An analysis
to MRI. Int J Radiat Oncol Biol Phys. 2004;59(1):126–37. of the dose-reponse for arteriovenous malformation radiosurgery
and other factors affecting obliteration. Radiother Oncol. 2002; 83. Gemmete JJ, Chaudhary N, Pandey AS, Oweis Y, Thompson BG,
63(3):347–54. Maher CO, Gandhi D, Ansari SA. Initial experience with a
78. Maruyama K, Kawahara N, Shin M, Tago M, Kishimoto J, Kurita combined multidetector CT and biplane digital subtraction angiog-
H, Kawamoto S, Morita A, Kirino T. The risk of hemorrhage after raphy suite with a single interactive table for the diagnosis and
radiosurgery for cerebral arteriovenous malformations. N Engl J treatment of neurovascular disease. J Neurointerv Surg. 2013;5(1):
Med. 2005;352(2):146–53. 73–80.
79. Kondziolka D, Lundsford LD, Flickinger JC. Gamma knife 84. Stancanello J, Cavedon C, Francescon P, Causin F, Avanzo M,
stereotactic radiosurgery for cerebral vascular malformations. In: Colombo F, Cerveri P, Ferrigno G, Uggeri F. BOLD fMRI integra-
Aleksander E, Loeffler JS, Lundsford LD, editors. Stereotactic tion into radiosurgery treatment planning of cerebral vascular mal-
radiosurgery. New York: McGraw-Hill; 1993. p. 136–46. formations. Med Phys. 2007;34(4):1176–84.
80. Colombo F, Benedetti A, Pozza F, Marchetti C, Chierego G. Linear 85. Koga T, Maruyama K, Kamada K, Ota T, Shin M, Itoh D, Kunii N,
accelerator radiosurgery of cerebral arteriovenous malformations. Ino K, Terahara A, Aoki S, Masutani Y, Saito N. Outcomes of dif-
Neurosurgery. 1989;24(6):833–40. fusion tensor tractography-integrated stereotactic radiosurgery. Int
81. Spiegelmann R, Friedman WA, Bova FJ. Limitations of angio- J Radiat Oncol Biol Phys. 2012;82(2):799–802.
graphic target localization in planning radiosurgical treatment. 86. Guo WY, Wu YY, Wu HM, Chung WY, Kao YH, Yeh TC, Shiau
Neurosurgery. 1992;30(4):619–23. CY, Pan DH, Chang YC, Hsieh JC. Toward normal perfusion after
82. Kondziolka D, Lunsford LD, Kanal E, Talagala L. Stereotactic radiosurgery: perfusion MR imaging with independent component
magnetic resonance angiography for targeting in arteriovenous analysis of brain arteriovenous malformations. AJNR Am J
malformation radiosurgery. Neurosurgery. 1994;35(4):585–90. Neuroradiol. 2004;25(10):1636–44.
Techniques of Stereotactic
Localization 3
Victor C.K. Tse, M. Yashar S. Kalani, and John R. Adler
The word stereotactic stems from the Greek words solid and The innovation of their device was having a separate base
arrangement or to order. It describes the ability to locate the frame and a headpiece. The base frame fixed onto the skull
target in a defined three-dimensional space using an estab- rigidly and established the horizontal planes along Reid’s
lished coordinate system. Its application to neurosurgery is line, thus allowing the coordinates to be read along the infra-
credited to Horsley and Clarke at the turn of the twentieth orbital border to the external auditory meatus. The headpiece
century [1], but historically, it has been noted that the idea of was then incorporated perpendicular onto the basal frame to
using anatomical landmarks as reference points to describe adjust the vertical plane. The headpiece carried a micro-
the locations of deep brain structures precedes Horsley and manipulator that was perpendicular to the frame. It was
Clarke (Fig. 3.1). through this micro-manipulator that the aspiration needle
and stimulating electrode were placed into the deep structure
of the brain. The Horsely–Clarke system was based primar-
History of Stereotaxis ily on a set of predictable measurements available to them
that address the location of deep brain structures in reference
The development of sterotactic localization parallels those of to calvarial landmarks.
brain imaging, but the earliest applications of stereotaxis In the early part of the 1900s, most stereotatic localization
predate the advent of imaging modalities. Dittmar in 1873 in clinical use was based on modifications of the Horsley–
describes a study of the medulla oblongata of cat using a Clarke frame with calvarial anatomical landmarks [4]. Later,
device to read off the distance from fixed dependable surface several key technical developments, including the introduc-
anatomical landmark. Zernov in Russia used a similar tion of X-rays, ventriculography, and cerebral angiography,
approach in 1890 to locate deep structures in humans [2]. changed the course of stereotactic localization.
The L’encephalometer was a relatively simple device The introduction of the pneumoventriculogram (first
anchored onto the meridian of the human skull; it used the described by Dandy) with air as the contrast agent allowed
polar coordinate and arc-based angulation system to mark Spiegel and Wycis to use intraventricular landmarks to cali-
the trajectory to the designated structures of the brain. brate the stereotactic frame [5]. They established the use of
In 1889, it was used to locate the central sulcus of the human the posterior commissure, the infra-aural plane, and the
brain for the aspiration of an abscess [3]. Ultimately, its con- Frankfurt line (a straight horizontal line from the top of the
tribution to neurosurgery was overshadowed by the work of external auditory meatus to the inferior border of the orbit
Horsely and Clarke. along either side of the human skull) as reference points.
Not unlike Zenov’s device, Horsely’s device was This approach led to several modifications and refinements
also anchored onto the skull by using a pair of ear bars. of the published atlas for stereotactic localization. By the
late 1950s, the inter-commissural line joining the anterior
V.C.K. Tse, M.D., Ph.D. (*) commissure to the posterior commissure (as introduced by
Kaiser Permanente Medical Group, Redwood City, CA, USA Talairach) was the de facto reference point in defining deep
e-mail: Tsevictor@gmail.com structures in the brain [6] (Fig. 3.2). The internal structures
M.Y.S. Kalani, M.D., Ph.D. of the brain especially those surrounding the ventricles
Division of Neurological Surgery, Barrow Neurological Institute, could be located with a high degree of certainty by reading
St. Joseph’s Hospital and Medical Center, Phoenix, AZ, USA
off the coordinates from a grid and its companion atlas and
J.R. Adler, M.D. adjusted for the magnification of the X-ray taken during
Neurosurgery, Stanford University and Chief of New Clinical
Applications at Varian Medical Systems, Stanford, CA, USA pneumoencephalography.
26 V.C.K. Tse et al.
Fig. 3.1 The progression in stereotactic localization from 1906 to the operating room where the reference fiducial system is attached to the
modern time. (a) A replicate of the horsely and clarke frame (open head-rest and surface anatomical landmarks are co-registered with the
source from Google images). (b) A modern leskell frame (image used preoperative CT and MRI fused images, thus allowing real-time navi-
with permission from Elekta). (c, d) A frameless system used in the gation and target localization
In the 1930s, visualization of intracranial structures by

X-ray and pneumoencephalography was limited in part by to Image-Guided Stereotactic Localization
poor discrimination of brain tissue density and the over-
shadow effect caused by superimposed structures. Ziedes In 1947, Lars Leksell introduced the arc-based stereotactic
Des Plantes suggested that one could eliminate superim- frame, which bears his name and continues to be used today.
posed shadows if a single plane was visualized at any one The Leksell frame consists of a base frame and a mobile
time. He proposed to move the X-ray source and the detec- arc-quadrant headpiece. This design enables the operator to
tion plate simultaneously at a fixed distance. This blurred all put the desired target at the arc center and allows the target to
images casted by tissue anterior and posterior to the plan of be accessed at various angles. The ability to use the target
interest. This proposal led to the principle of tomography center and triangulation fundamentally altered the approach
and provided a mean of validating stereotactic localization of to localize lesions in the intracranial cavity. Leskell coupled
deep structures. an X-ray tube to the sterotactic frame so that the trajectory of
3 Techniques of Stereotactic Localization 27
Fig. 3.2 The incorporation of

ventriculogram in stereotactic
functional neurosurgery. The
AC–PC remains the classical
landmark used by surgeons of
today as it was the de facto
reference point in defining deep
structures in the brain in the days
of Talairach. (Used by
permission by Mark Sedrak, MD)
a treatment probe could be more readily and safely identi- integrals along an X-ray beam path [9]. This laid the ground-
fied. Extrapolating from his early concept Leksell would work for the development of CT imaging, which was readily
eventually imagine the idea of using a collimated radiation incorporated into stereotactic localization.
source instead of a physical probe to allow isocentric target- In 1967 Hounsfield envisaged a computational solution to
centered radiotherapy. By observing this evolution in stereo- reconstructing images based on photon transmission when
taxy it is clear to see how development of the Leksell frame X-rays were taken in all possible directions around an object
eventually resulted in the development of the field of stereo- while it was translating along an axis. He deduced that if
tactic radiosurgery [7]. X-rays were taken in all directions around an object, when
In the early 1960s, Oldendorf elaborated on the tomogra- the data were represented as a series of slices, a mathemati-
phy technique by rotating a radiation source and detector cal model based on Fourier transformation could be used to
around an axis in a constant linear manner, to differentiate differentiate the radiodensity of tissue within each individual
tissue density by its radiation absorption. Theoretically, tis- slice. Furthermore, in knowing topographical map of the
sue at the point of intersection would have constant radiation radiodensity distribution he could regenerate images of
absorption, whereas all other tissues in that plane had vari- structures within that slice. From this basic concept com-
able radiation absorption [8]. Cormack in 1963, used Fourier puted tomography was born. With the support of EMI Inc.,
transformation and backprojection to provide a mathemati- Hounsfield was successful in building the first CT machine.
cal solution and reconstruct the cross-section of radiographic Stereotactic surgery in its early years mainly relied on
images by recapitulating the sum of the radiodensities as line ventriculocisternography to localize internal structures for
Fig. 3.3 The use of a N-localizer (a) to translate two-dimensional coordinates into three-dimensional reference points in the CT images (b).
(c) The localization of the target can then be checked against a phantom base
lesioning, but by the late 1970s, target localization in stereo- To facilitate the digitization of surgical space, a series of
tactic surgery was mainly done by CT imaging. A localizing externally visible fiducial markers would be attached to dis-
plate with vertical fiducial posts filled with radio-opaque crete skin and scalp locations prior to the acquisition of plan-
material was fixed to the base of the stereotactic frame. These ning imaging studies. Upon completing presurgical head
posts essentially formed a picket fence around the perimeter immobilization, most typically in a Mayfield headrest, these
of the stereotactic frame while the patient was being imaged. same fiducials could be spatially registered and even tracked
The positions of the posts were read as the x–y coordinates in using any number of three-dimensional digitizing devices.
the CT images and used as reference points. The location of The earliest digitizer used for frameless stereotactic localiza-
the target was then triangulated. By late 1970s, the accuracy tion, was a mechanical arm (wand) consisting of a series of
of target localization was greatly improved through the inno- optically encoded joints that could be read by a computer. By
vation of Russell Brown who added a diagonal bar across the computationally determining the three-dimensional position
two picket posts to generate an N-localizer. This ingenious of the scalp fiducials, the digitizer enabled the entire head to
design simplified the conversion of the two-dimensional be registered within three-dimensional operative space.
coordinates into three-dimensional reference points in the When during actual surgery, calibrated tools were attached to
CT images (Fig. 3.3). The N-localizer allows the colocaliza- the same mechanical arm surgeons were able for the first
tion of the desired target in both CT and MRI images, and time to “navigate” through a patient’s anatomy by using the
improves the accuracy of multimodal image fusion. In the planning CT (and eventually MRI) as a map, thereby
1980s the N-localizer was universally adapted and was enabling the emergence of “image-guided surgery.” Over
incorporated into the design of the Leskell frame for stereo- time further advances in three dimensional digitization tech-
tactic radiosurgery [10–12]. nology translated directly to improved tools for image guid-
ance. These new digitization technologies, each with their
own advantages and disadvantages, utilized acoustic, elec-
From Frame-Based to Frameless System tromagnetic, and finally optical tracking concepts [13].
As the speed and the computation power of modern com-

puters increased, stereotactic localization started to shift Digital Reconstruction Radiography
from a frame-based to “frameless” technology. The
Cartesian coordinate system of computerized volumetric Stereotactic frameless radiosurgery is made possible by the
imaging (initially CT and subsequently MRI) enabled advancement of digital reconstruction radiography and
ready determination of target position within the cranium. image fusion algorithms. The digitally reconstructed radio-
By using common points of reference, the position of this graphs (DRRs) are X-ray images mathematically synthe-
same head could be accurately determined within the three- sized from the planning CT. DRRs are essentially a series of
dimensional space of operative tools, thereby making it ray-line re-constructions through a three-dimension volu-
possible to perform precision procedures without using rig- metric CT data set. A planar transmission image is repro-
idly attached external frames. duced by ray tracing from the source through the relevant
Fig. 3.4 The use of DRR in stereotactic localization: (a, b) The stimu- translational and three rotational dimensions (6 degrees of freedom) to
lation CT with its selected alignment center is used to generate a pair of be compared. A and C illustrate skull tracking for treating a patient with
DRR; (c, d): The fusion of two orthogonal X-ray projection images trigeminal neuralgia; B and D illustrate spine tracking for treating a
taken by the tracking system in Cyberknife treatment room and co- lung lesion
registered with the DRRs. This allows the target position in the three
adjacent CT slices to generate an effective transmission

value at each pixel on the image. DRRs can then be used to Extracranial Stereotactic Localization
compare to the portal films or their equivalent portal verifica-
tion images. In most stereotactic radiosurgery systems, the The application of extracranial localization in the nervous
frameless tracking systems use a similar set-up. A simulation system is generally applied to pathologies of the spinal col-
CT is taken according to the constraints of the radiosurgery umn and spinal cord structures. These techniques are pre-
machine, an alignment center is selected by the operator dominantly based on those techniques used in targeting
from the planning CT, and a set of DRRs is then generated. intracranial lesions. It is based on the premise that the
In the treatment room, the target’s location (in reference to motion of the spinal cord is relatively small and one can use
the alignment center) is aligned to the imaging center of the the bony landmarks as surrogate marker in tracking the
tracking X-ray source, which has a fixed center in the treat- motion of the cord. Spinal tracking can also be done by
ment room. Two orthogonal X-ray projection images taken implanting internal fiducials onto the spinous processes,
by the tracking system are then registered with the DRRs. lamina, or pedicles. Extracranial radiosurgery is also known
The 2D X-ray images are fused with the DRRs, and an auto- as stereotactic body radiation therapy (SBRT), is being
mated gradient correlation algorithm is used to calculate the increasingly used for the treatment of prostate, intrinsic
2D/3D correction vector that optimizes the translational and lung, and metastatic tumors. In extracranial stereotactic
rotational co-registration. This allows the target position in localization, specific considerations for organ movement
the three translational and three rotational dimensions (6 and subtle movement caused by respiration must be made.
degree of freedom) to be compared (Fig. 3.4). Motion not only makes localization of the target difficult, it
introduces an internal marginal error (the probability in both Unlike conventional CT, CBCT uses a cone-shaped rather
spatial and temporal dimensions within which the target than a fan-shaped beam thus allowing the acquisition of a
may be located) and possibility of causing an error in dose three-dimensional image volume during a single rotation of
delivery. To overcome the geometrical uncertainty of target the X-ray source/detector tandem. These localization vol-
localization caused by motion, one needs dynamic updating umes can be registered directly with the planning CT for bet-
of the spatial location of the target, the motion of the ter visualization and matching of the target location.
patient’s body, and internal organs with respect to time [14]. Additionally, it allows one to use soft tissue anatomy and
In other words, one has to track the target in real time either landmark to track and correct target localization (Fig. 3.5).
by following the movement of the target or by tracking some To date, Megavoltage CBCT has poor soft tissue contrast
moving structures that correlates with the movement of the due to Compton scatter and detection efficiency is low result-
target. The other approach is to use movement information ing in high noise levels with high energy photon. In the
to predict the position of the target or its surrogate at a par- future, with advancement of megavoltage cone beam CT one
ticular point of a motion cycle such as the inspiration or may even be able to do real-time tracking and correct target
expiration motion cycle of the thorax (i.e., gating). In both localization in tandem with beam delivery during radiosur-
cases, this can be achieved by placing infrared reflective gery treatment.
markers onto the patient to allow real time tracking of the
target in reference to the position of the patient. Additionally
internal radio-opaque markers or bony structure can be used
as surrogate markers to tag the target in real time using The Future of Stereotactic Localization
X-Ray or stereoscopic CT scanning system. The use of
markers and DRR has been shown to reduce error [15]; Over the last generation the field of stereotaxy has evolved
however, using bony structures as a surrogate marker runs from a small subdiscipline of neurosurgery used only in the
the risk of “marker migration” if the target and its reference treatment of a handful of functional disorder, to a broad and
bony landmark can move independently to each other. growing category of tools and techniques that now impacts
At the time of treatment planning, both markers are iden- many areas of surgery outside the brain. Moreover, going
tified and registered with respect to the planning isocenter forward the ongoing revolution in minimally invasive sur-
for a chosen phase of motion (i.e., inspiration or expiration gery, which by its nature requires the ever more precise map-
phase of the respiratory cycle). Prior to treatment the patient ping of patient anatomic space into operative space, will only
is positioned using infrared positioning system in the room be fully realized by incorporating at increasingly deeper lev-
to align the treatment isocenter with the isocenter of the els all the principles of stereotaxy. Although the idea of using
machine. The motion of the patient is monitored and recorded real-time imaging as a substitute for stereotactic mapping in
to generate a model of the motion cycle. The technician then open surgical procedures is seductive, e.g., intraoperative
sets the acquisition time of the radiographic monitoring sys- MRI and CT, even here stereotaxy is useful because continu-
tem to correlate with the body motion and compare the real ous imaging and perfect patient immobilization prove impos-
time images to that of the DRR. sible. In the setting of noninvasive surgical procedures, such
as radiosurgery and high intensity focused ultrasound, the
need for stereotaxy is inescapable. As the importance of such
Immobilization and Target Tracking noninvasive procedures grows, we will see further utilization
of robotic tools to enable dynamic compensation for move-
Immobilization of the body helps to reduce motion and ment. Meanwhile, other imaging modalities besides X-rays
increase accuracy. In targeting lesions in the head and neck and MRI can be turned into spatial localization “anatomic”
region this is commonly done using a facial fixation mask. sensors such as ultrasound and perhaps near infrared. By not
In extracranial tracking, devices such as conformal alpha requiring ionizing radiation this approach promises truly
cradle, Vac-lock bag, and Memorial body cradle have been continuous “real-time” “anatomic localization”. It is likely
used to achieve the same effect. The patient is fitted with more specialized tools will be developed to enable specific
these devices during image acquisition in the planning CT clinical applications. As the frontiers of medicine advance
and during stereotactic radiosurgery treatment. The Memorial inexorably, and noninvasive procedures subsume nearly all
body cradle has its own built-in external fiducial coordinate of surgery, stereotaxy is likely to lose nearly all resemblance
system that can be visualized in the planning CT. However, a to the stereotactic head frames which introduced the notion
built-in fiducial system in the immobiliizing device does not of precision targeted surgery a century ago. Nevertheless the
necessarily improve accuracy. intrinsic stereotactic principles of organizing the operating
More recently, the use of real-time “on-board” kilo- environment into a precisely calibrated three-dimensional
voltage cone beam CT (CBCT) facilitates target localization. functional space will remain at the heart of such change.
Fig. 3.5 The use of cone beam

CT in extracranial stereotactic
localization. (a) The stimulation
CT with its selected alignment
center is used to generate the
three-dimensions/volumetric
DRR. (b) The difference in the
bladder filling volume as imaged
by the cone beam CT. (c) The
fusion image from the
positioning cone beam CT and
the DRR using the Trilogy
system by Varian
2. Zernov DN. Léncephalometrie. Rev Gen Clin Ther. 1890;19:302.

References 3. Zernov DN. Entsefalometr—ustroistvo dlya lokalizatsii mozgovych
structur u zhivogo cheloveka. Predvaritel’noe soobshenie.
1. Clarke RH, Horsley V. On a method of investigating the deep gan- Encephalometer—a device for localising cerebral structures in a liv-
glia and tracts of the central nervous system (cerebellum). Br Med ing man. A preliminary communication. Trudy Fiziko-meditsynskogo
J. 1906;2:1799–800. Obshestva Moscovskogo Universiteta 1889; 2:70–80. (Rus).
4. Kirschner M. Arch Klin Chir. 1923;176(581):581–620. 12. Leksell L. Stereotaxis and tomography, a technical note. Acts
5. Dandy WE. Ventriculography after injection of contrast into the Neurochirurgica. 1980;52:1–7.
ventricles. Ann Surg. 1918;68(1):5–11. 13. Reinhardt HF, Zweifel HJ. Interactive sonar-operated device for
6. Reichert T, Hassler R. Special method of stereotaxic brain opera- stereotactic and open surgery. Sterotact Funct Neurosurg.
tion. Proc Roy Soc Med. 1955;48:469–70. 1990;54:393–7.
7. Leksell L. The stereotaxic method and radiosurgery of the brain. 14. Verhey LJ. Immobilizing patients and positioning for radiotherapy.
Acta Chir Scand. 1951;102:316–9. Semin Radiat Oncol. 1995;5:100–14.
8. Oldendorf WH. The quest for an image of brain: a brief historical 15. Soete G, Van de Steene J, Verellen D, Vinh-Hung V, Van den Berge
and technical review of brain imaging techniques. Neurology. D, Michielsen D, Keuppens F, De Roover P, Storme G. Initial clini-
1978;28(6):517–33. cal experience with infrared-reflecting skin markers in the position-
9. Cormack AM. Representation of a function by its line integrals, ing of patients treated by conformal radiotherapy for prostate
with some radiological applications. J Appl Phys. cancer. Int J Radiat Oncol Biol Phys. 2002;52:694–8.
1963;34:2722–7.
10. Brown RA, Nelson JA. Invention of the N-localizer for stereotactic
neurosurgery and its use in the Brown-Roberts-Wells stereotactic
frame. Neurosurgery. 2012;70(2 Suppl Operative):173–6. Further Reading
11. Brown RA, Nelson J. The origin of the N-localizer for stereotactic
neurosurgery. Cureus. 2012;4(8):e55. doi:10.7759/cureus.55. Levy M. A Short history of stereotactic neurosurgery. http://www.
http://www.cureus.com/papers/the-origin-of-the-n-localizer-
for-stereotactic-neurosurgery/203. neurosurgery.org/cybermuseum/stereotactichall/stereoarticle.html
Part II
Radiation Biology and Physics
The Physics of Stereotactic
Radiosurgery 4
Siyong Kim and Jatinder Palta
Protons, defined as positively charged particles found in

Introduction the nucleus of an atom, are also used for treatment [15–19].
Proton beams can be obtained from particle accelerators
In radiosurgery, instead of using a surgical knife when treating such as cyclotrons and synchrotrons [20], which cost a tre-
a patient, high-energy ionizing radiation is the tool of choice. mendous amount of money; thus, very few proton facilities
To understand the tumoricidal effects of ionizing radiation, it exist worldwide, and there are currently only five in the USA.
is important to know how radiation interacts with matter. Because the majority of radiosurgery treatments are per-
This chapter describes general concepts and principles in formed with photon beams, emphasis is placed on the photon
radiation physics, including basic physics that are applicable beam.
to stereotactic radiosurgery (SRS). Commonly used delivery
systems are also briefly reviewed. This chapter is written for
non-physics professionals, especially neurosurgeons and Cobalt-60
radiation oncologists.
Cobalt-60 is the radioactive isotope used as the radiation
source in the Leksell Gamma Knife treatment machine
The Radiation Source for Radiosurgery (Elekta. Nor-cross, GA), which is described later. A radio-
active isotope is an atom with an unstable nucleus that
Ionizing radiation is any electromagnetic or particulate radia- tries to stabilize itself through a process called radioactive
tion capable of producing ions either directly or indirectly in decay by emitting ionizing radiation such as alpha, beta,
its passage through matter [1, 2]. Among the many radiations and/or gamma particles. When 60Co undergoes radioactive
we have, the photon, defined as a discrete packet of electro- decay, it emits beta particles and two strong gamma radia-
magnetic energy, is most commonly used for patient treat- tions, one with 1.17 MeV of energy and the other with
ment [3, 4]. A photon is called by several different names 1.33 MeV of energy (Fig. 4.2). MeV, a commonly used unit
depending on its energy level, such as radio waves, infrared, of energy for particles in the treatment energy range, means
visible light, ultraviolet, and X-rays in the form of energy as 1 million electron volts, where 1 electron volt is the energy
illustrated in Fig. 4.1. Only the X-ray range of photons is a gained by an electron that accelerates through a potential
form of ionizing radiation, and, though the energy range of difference of 1 V. A real treatment beam of Gamma Knife
X-rays is fairly wide, only a high-energy range is used for can include photons of energy different from 1.17 and
treatment. Photon treatment beams can be obtained either from 1.33 MeV because some photons experience interactions
radioisotope sources or from X-ray-generating machines. with 60Co itself and/or its capsulation material and lose a
Currently, cobalt-60 (60Co) is the most widely used radioisotope part of that energy. Therefore, the effective energy of
[5–9] for radiosurgery, and the medical linear accelerator Gamma Knife is slightly lower than 1.25 MeV (the average
is the most popular X-ray-generating machine [10–14]. of 1.17 and 1.33 MeV).
Cobalt-60 ultimately decays to nonradioactive nickel. The
half-life of 60Co, that is, the length of time needed for 60Co to
lose half of its radioactivity from decay, is 5.26 years. At the
S. Kim, Ph.D. (*) • J. Palta, Ph.D.
end of 1 half-life, only 50 % of the original radioactive mate-
Department of Radiation Oncology, Virginia Commonwealth
University, Richmond, VA, USA rial remains. This means that treatment time after 5 years is
e-mail: skim2@mcvh-vcu.edu double that at the time of initial installation, and that the
36 S. Kim and J. Palta
lower higher
Energy of -9 -8 -7 -6 -5 -4 -3 -2 -1
photon (eV) 10 10 10 10 10 10 10 10 10 1 101 102 103 104 105 106 107
Therapeutic
Radio waves Infrared Ultraviolet X-ray
Microwaves Diagnostic
Visible
X-ray
Fig. 4.1 The electromagnetic spectrum is illustrated. Therapeutic X-ray or ionizing radiation is in the energy range of 105 eV and higher. The
energy of photon beams commonly used in radiosurgery are higher than 106 eV
60
Co 5.26 y
potential differences in diagnostic X-ray tubes up to several
27
hundred kiloelectron volts (keV). The acceleration mecha-
nism in a linear accelerator, however, is somewhat different
b 1– (99.8%) 0.313 MeV because it uses microwave technology similar to that used
for radar to accelerate electrons in a linear tube, often called
an accelerator tube [21]. In linear accelerators, electrons
b 2– (0.12%) 1.489 MeV are usually accelerated to the energy range of 4–25 MeV.
g 1 (99.8%) 1.173 MeV
A 6-MeV electron is most often used to create an X-ray for
radiosurgery. When an electron creates a photon, theoreti-
cally, the photon can take any energy from 0 to the same as
the energy of the incident electron. In other words, the pho-
g 2 (100%) 1.332 MeV
ton beam produced by a 6-MeV electron beam can have an
energy spectrum of 0–6 MeV. The energy of an X-ray pro-
60 duced in a linear accelerator is denoted MV instead of MeV.
Ni
28 The major components and auxiliary systems of typical
medical linear accelerators are shown in a block diagram
Fig. 4.2 For 60Co, most of the disintegrations (99.8 %) are the emission
of a β1− with a maximum energy of 0.313 MeV. This leads to an excita- (Fig. 4.3). The role of each part is briefly explained below.
tion of 60Ni, which releases its energy quickly by emitting two gamma
rays (1.173 and 1.332 MeV) in cascade. In some disintegration Power Supply
(0.12 %), a β2− particle with a maximum energy of 1.486 MeV is
The power supply provides direct current (DC) power to the
released and leads to the lowest excited state of 60Ni
modulator.
Modulator
source of Gamma Knife needs to be replaced after a certain The modulator generates high-voltage pulses lasting a few
period of time to avoid long treatment times. microseconds that are simultaneously introduced to the
radiofrequency source and the electron gun.
X-Ray Production in a Linear Accelerator Electron Gun

The electron gun produces electrons by thermionic emission.
One of the interesting characteristics of a fast-moving elec- In a conducting material, the electrons exist at or below the
tron is that when it interacts with a material, it can produce baseline electron energy at low temperatures. As the tem-
an X-ray. A part or whole of the kinetic energy of the elec- perature is increased, some of these electrons have sufficient
tron is transformed into electromagnetic energy. This prin- energy to pass over the surface–potential barrier between the
ciple is used to obtain X-rays in most X-ray production material and the vacuum. This process of increasing the tem-
machines in which electrons are accelerated and induced to perature of a bulk material to increase the number of elec-
hit a target and then generate an X-ray. The amount of trons that can leave the material is called thermionic emission.
X-ray particles produced increases as the kinetic energy of These electrons are pulse-injected into the accelerator tube
the incident electrons increases. Electrons are accelerated by according to the signal from the modulator.
4 The Physics of Stereotactic Radiosurgery 37
Fig. 4.3 The power supply provides DC to the modulator, which tube. The electron and the microwave are timed exactly to be met in the
provides a high-voltage pulse to the RF generator and electron gun accelerator tube by the modulator. The accelerator tube accelerates the
simultaneously. The electron gun introduces the electron in pulse to the electrons via the resonance cavity. As the accelerated electron comes
accelerator tube. Electron production occurs through the thermionic out of the accelerator tube, the path of the electron is bent through a
process. The RF generator introduces a microwave to the accelerator bending magnet to hit the target and generates an X-ray
Radiofrequency Source basically microwave resonators. Resonance is a very common

Radiofrequency (RF) is a range of electromagnetic frequencies physical phenomenon that can be observed in many types of
above sound and below visible light, generally in the 30 kHz systems. Imagine a child on a swing, for example. If the child
to 300 GHz range. RF is used for all broadcast transmissions pushes the swing at just the right time, the push can make
including AM and FM radio, television, shortwave, micro- the swing go higher and higher with very little effort. The
wave, and satellite transmissions. In a linear accelerator used swing has a natural frequency of oscillation. Pushing “at just
for radiosurgery, a magnetron is often the RF source. The the right time” means that energy is put into the system at a
magnetron is a diode-type electron tube that functions as a frequency known as the fundamental frequency. Now the
high-power oscillator to generate microwave pulses. Pulse system is said to be in resonance condition.
duration is several microseconds long, and the repetition rate The amplitude of the motion can increase rapidly with
is several hundred pulses per second. Following the pulses resonance. An analogous phenomenon can occur in a linear
from the modulator, microwave pulses are introduced into accelerator with RF resonance, which is generally derived
the accelerator tube through the waveguide system. The typi- from an RF cavity that is typically in a right circular cylindri-
cal frequency of microwaves used in a linear accelerator is cal shape with connecting holes to allow a charged particle’s
about 3 GHz. beam to pass through for acceleration. Figure 4.4 shows a
In some designs, a klystron is used instead of a magnetron. typical cylindrically symmetric RF cavity. In a fundamental
The klystron is a microwave amplifier rather than a source; RF mode, the electric field is roughly parallel to the beam
therefore, it requires a low-power microwave oscillator called axis and radically decays to zero upon approach to the cavity
an RF driver as an RF source. In general, a klystron produces walls. The pulsed electron particle beam traverses the cavity
microwaves of higher power than the magnetron and is often through the centered hole, creating an accelerating force
used with high-energy linear accelerators (>10 MV). along the axis of the cavity due to the electric field.
Waveguide System Bending Magnet

The waveguide system is made up of a pressurized rectangular When an electron moves in a magnetic field, it experiences
tube through which microwave pulses generated by either a force. The direction of the force is at right angles to both the
magnetron or klystron are transmitted into the accelerator tube. direction of the magnetic field and the direction of the elec-
tron motion (Fig. 4.5). Using this principle, the path of accel-
Accelerator Tube erated electrons is controlled so that they can hit the target
An accelerator tube consists of many RF resonant cavities. normally (note the electron path, bending magnet, and target
An RF cavity allows power to be coupled into the particle beam. in Fig. 4.3). Magnets used for this purpose are called bending
In an electron linear accelerator, RF accelerating cavities are magnets. In certain cases, straight-ahead beam designs can
Fig. 4.4 A typical cylindrically symmetric cavity
Fig. 4.6 A schematic of the treatment head of a medical linear accel-

erator is illustrated. Initially, as the accelerated electrons hit an X-ray
target, an X-ray beam is generated via the bremsstrahlung process. This
X-ray beam is collimated by the primary collimator. As it exits the pri-
mary collimator, the X-ray beam profile is flattened out by the flattening
filter. Then, it goes through the ion chamber, and, finally, it gets colli-
mated by the secondary collimator. In radiosurgery, generally one more
collimator is added to make a very small field (e.g., circular collimator
or micro-multileaf collimator)
and lose its energy as bremsstrahlung radiation (i.e., an X-ray).

The X-rays are produced in all directions, but high-energy
X-rays are produced in the forward direction (Fig. 4.7).
Therefore, in a linear accelerator, higher photon intensity is
observed in the area close to the central line (i.e., the extended
Fig. 4.5 As the electron travels into a uniform magnetic field, it experi- line following the direction of the incident electron) below
ences a downward force. This force causes the electron to travel in a
circular path. The crosses in the figure indicate the uniform magnetic the X-ray target. The X-ray beam is first collimated by the
field pointing into the page primary collimator, which has a circular aperture through
which a photon beam can pass to make a radiation field. The
primary collimator is typically made of tungsten.
be applied without the bending magnet if the acceleration For patient treatment, a radiation beam with uniform
structure is short enough to be placed in a vertical direction. intensity across the field is desirable, and to create a uniform
A linear accelerator in straight-ahead beam design is used in photon beam, a conical filter called a flattening filter is used.
the CyberKnife unit, which will be better described later. The central area of the flattening filter is thicker to attenuate
more photons in the central area. In general, a high-Z-number
Treatment Head material is used for the flattening filter to effectively attenu-
The major parts of the treatment head are the target (often ate photons.
called the X-ray target), primary collimator, flattening filter, The amount of radiation produced by the linear accelera-
ion chamber, and secondary collimators (Fig. 4.6). tor is monitored by an ionization chamber, often called a
Accelerated electrons collide with the target to generate monitor chamber. The monitor chamber not only controls the
an X-ray beam. For a given energy of electrons, more X-rays amount of radiation delivered to the patient but also checks
can be produced if a target material with a higher atomic and controls the uniformity of the delivered radiation.
number (Z number) is used. Tungsten and gold are com- Whereas the primary collimator defines the maximum
monly used as target materials. aperture of the radiation field, a real treatment field is deter-
The production of an X-ray is based on a process called mined by the secondary collimator, which defines a rectan-
bremsstrahlung (meaning “braking radiation”), which is the gular field that fits to each patient treatment in routine
result of radiative collisions between a fast-moving electron radiation therapy. In radiosurgery, smaller fields are needed.
and a nucleus. When the electron passes near a nucleus, it can Thus, an additional collimator system, either a circular col-
be deflected from its path by Coulomb forces of attraction limator or micro-multileaf collimator (mMLC) [22], is used.
4×4 cm2 (a), 6×6 cm2 (b), and 8×8 cm2 (c), for both regular
flattened 6 MV beam and 6 MV FFF beam from a Varian
TrueBeam linear accelerator. As can be seen, dose rates in
FFF mode are larger than those in regular flattened beams by
about two times in 6 MV while profiles are not much differ-
ent. Dose rates in FFF mode become about four times larger
in 10 MV but more nonuniform dose profiles are expected.
A Photon’s Interaction with Matter
The effectiveness of radiation treatment is based on how

much radiation energy is deposited in the tumor compared
with the energy deposited in the surrounding normal tissues.
Energy deposition occurs through interactions between a
radiation beam and the human body at either the atomic or
Fig. 4.7 Schematic illustration of the angular distribution of emission
of bremsstrahlung X-rays around a target. At the kinetic energy of an nuclear level. When photons traverse a material, they reduce
electron beam less than 100 keV, X-rays are emitted relatively equally the intensity of the incident photons. The three most com-
in all directions. As the kinetic energy of the electron beam increases, mon ways for a photon to interact with matter for energies
the direction of the bremsstrahlung X-ray emission becomes signifi- above a few keV are photoelectric absorption, Compton scat-
cantly forward. In megavoltage X-ray machines, the transmission
X-rays are therefore used to treat a patient tering, and pair production. These three mechanisms account
for more than 99 % of the interactions between photons and
matter, and the probability of each depends on the energy of
Flattening Filter Free Mode the photon and the material with which it interacts. Because
As described before, photon beams produced in a conventional these interactions are quantum in nature, a specific interac-
linear accelerator by bombarding a high-Z target with accel- tion is never guaranteed. It is simply more or less probable
erated electrons are strongly forward peaked (see Fig. 4.7), than the others (depending on the energy and the material).
resulting in significant centrally peaked fluence distributions
incident to patients. In the past, such a nonuniform fluence
was considered not suitable for routine clinical applications Photoelectric Absorption
and intentionally flattened by a flattening filter (FF) as shown
in Fig. 4.6. However, the rationale that a FF is required to In photoelectric absorption, an X-ray photon is absorbed by
obtain uniform dose distributions is getting less strong in the interacting material and an electron from a shell of the
modern radiation therapy. An example is the so-called inten- atom is ejected, leaving the atom ionized, as illustrated in
sity modulated radiation therapy (IMRT), one of the most Fig. 4.9. The ionized atom returns to the neutral state with the
significant advancements in modern radiation therapy. In emission of an X-ray, called the characteristic X-ray of the
IMRT, highly inhomogeneous beam fluences are often atom. The kinetic energy transferred to the ejected electron
applied to achieve extremely conformal dose distributions during the interaction is the same as the incident photon
even to concave targets, implying no need of flattened beam energy minus the ejected electron’s binding energy to the
in principle. Another example is in case of relatively small atomic nucleus. Photoelectric absorption occurs predomi-
field treatments such as stereotactic radiosurgery (SRS), ste- nately at low energies of photons and for high-atomic-number
reotactic radiotherapy (SRT), and stereotactic body radio- materials. In water (one of the closest materials to tissue), the
therapy (SBRT) because there is no significant difference in probability of this interaction becomes negligible once pho-
beam fluence profile across the field between flattened and ton energy reaches above 100 keV. Because the effective
unflattened beams when field sizes are small. Thus, flatten- energies of photon beams from the Gamma Knife (i.e., 60Co
ing filter free (FFF) mode is available in several state-of-the- source) and linear accelerator are of the order MeV, the effect
art commercial linear accelerators nowadays. of photoelectric absorption is not significant in radiosurgery.
The major advantage of a FFF beam is significantly large
dose rate compared to conventional flattened beams [23].
With larger dose rate, treatment time can be reduced and Compton Scattering
uncertainty related to patient motion may become smaller.
This feature is very important especially for SRS, SRT, and In Compton scattering, also known as an incoherent scat-
SBRT where fractional dose is large. Figure 4.8 shows dose tering, the incident X-ray photon ejects an electron from
rate profiles at 10 cm depth of water in several field sizes, an atom, loses some of its energy to the ejected electron,
Fig. 4.8 Dose rate profiles at 10 cm depth of water in three field sizes, TrueBeam linear accelerator. Dose rate scale is arbitrary and as illus-
4 × 4 cm2 (a), 6 × 6 cm2 (b), and 8 × 8 cm2 (c), for both regular flattened trated, dose rates in FFF mode are larger than those in regular flattened
6 MV beam (denoted as 6 MV) and 6 MV FFF beam from a Varian beams by about two times while profiles are not much different
the interaction. The energy and momentum are conserved

in this process. Kinetic energy transferred to the electron is
the energy difference between the primary and scattered
photons. The energy of the scattered photon depends on
the direction (i.e., the angle with respect to the direction
of the primary photon). The scattered photon has minimal
energy when it is backscattered (180° from its direction of
travel). If it is scattered in the same direction as the pri-
mary, the scattered photon energy is the same as the primary
photon energy. Compton scattering is important for
low-atomic-number materials and photon energies of
100 keV to 10 MeV. Therefore, Compton scattering is the
Fig. 4.9 In a photoelectric effect, an incident photon striking a bounded most significant interaction for photons used in radiosurgery.
electron is absorbed, and the electron is ejected from the atom with a Figure 4.10 illustrates the Compton scattering interaction.
kinetic energy equal to the difference between the incident photon
energy and the binding energy of the electron. The probability of the
photoelectric effect increases as the atomic number of the target mate-
rial increases; however, it decreases as the energy of the incident photon Pair Production
increases. (E is the energy of the incident photon, T is the kinetic energy
of the ejected electron, and BE is the binding energy. Photon energy is Photons have a quantum of energy but zero mass. In pair
expressed as hν, where h is Plank’s constant and ν is frequency)
production, the incident photon energy is completely
absorbed in the medium, and an electron and its antiparticle,
and continues to move in a direction different from the initial a positive electron called a positron, are created. This inter-
direction. The resulting incident photon is called a scat- action is illustrated in Fig. 4.11. Pair production can only
tered photon, whereas it is called a primary photon before occur when the energy of the incident photon is greater than
Fig. 4.12 A photon enters a volume of a medium and transfers a por-

Fig. 4.10 In Compton scattering, the incident photon strikes an elec- tion of its energy, thus giving the electron a kinetic energy. The electron
tron and ejects it out of the atom to which it was bound. During this will travel a short distance and deposit its energy in a medium. The
process, the incident photon energy is transferred to the ejected elec- original photon will continue to travel within the medium as it transfers
tron. The least energy is transferred to the electron when the incident more energy to another electron
photon has no scattering (i.e., θ = 0º), and the most energy is transferred
to the electron when the incident photon is scattered backward (i.e.,
θ = 180º). Compton scattering becomes most significant at an energy
range of 100 keV to 10 MeV, and it is almost independent of the atomic
number (i.e., Z value) of the interacting material. Thus, for the radiosur-
Dose and Dosimetry
gery energy range, Compton scattering is the dominating interaction. (E
is the energy of the incident photon. T is the kinetic energy of the The physical quantity that is used in radiation therapy to kill
ejected electron, and E′ is the energy of the scattered photon) neoplastic cells is the kinetic energy of the incident radiation
beam. Once photons enter a medium, they start to interact with
the medium with the probability of interactions described ear-
lier. Any photon particle loses a part or all of its energy when
it initially interacts with a medium and most of the lost energy
is absorbed by the medium around the interaction point.
Energy absorbed in a unit mass of the interaction medium dur-
ing interactions with a radiation beam is called the dose or,
more specifically, the radiation dose. The commonly used unit
of dose is the gray (Gy), which is equivalent to J/kg (joules per
kilogram), where joule is a unit of energy (1 J is the approxi-
mate energy required to lift 1 kg by 0.1 m assuming gravity
accelerates at 10 m/s2). Depending on the amount of dose, cGy
(centigray, or one-hundredth of a Gy) is also often used.
Fig. 4.11 When a photon with energy equal to or higher than 1.02 MeV Dosimetry is either the measuring or calculating of dose.
comes very close to the nucleus of an atom, it interacts with the nuclear
Coulombic force field and can create a pair of an electron and a posi-
As you may have noted, the energy absorption mecha-
tron. The total kinetic energy of the electron and position is the energy nism during photon interaction is a two-step process: (1)
of the incident photon minus 1.02 MeV. The probability of pair produc- energy transfers from the photon to the ejected electron and
tion is proportional to the atomic number of the target material and the (2) energy is deposited from the ejected electron to the
incident photon energy
medium (Fig. 4.12). Thus, energy deposition occurs within a
volume around the spot of interaction rather than at a point of
interaction. If the photon still has kinetic energy after the first
1.02 MeV, which is the minimum energy needed to create an interaction (e.g., Compton scattering), it can travel and inter-
electron and a positron. Positrons are very short lived and act at another point, then cause another energy deposition.
disappear with the creation of two 0.51-MeV photons each, This second interaction is independent of the first interac-
a process called positron annihilation. The total kinetic tion, thus, third, fourth, and multiple scatterings are possible.
energy transferred to the electron and positron pair is equal The energy absorbed from scattered photons is called the
to the incident photon energy minus 1.02 MeV. Pair produc- scattered dose, contrary to the primary dose, which is the
tion is of particular importance when high-energy photons energy absorbed in the first interaction. The same thing hap-
pass through materials of a high atomic number but plays a pens when a patient is irradiated with a photon beam. Photons
small role in photon beams used for radiosurgery. interact with human tissue and deposit energy into the tissue.
Fig. 4.13 (a) In the thin line of beam and thin medium, the scattered broad beam and thick medium, the single or multiple scattering of
photons are out of the medium and never return back to the medium. photons can occur at any line of beam. The broad beam is considered
There is thus no contribution by the scattered photons to the dose. as a bundle of thin lines of beam. The scattered photons coming from
Percent depth dose (PDD) is mainly determined by the energy deposi- each beam line to the center line of interest contribute to the total
tion of the primary photons. (b) On the other hand, in the thick dose. PDD values beyond the dmax (depth of dose maximum) increase
medium with the thin line of photon beam, the scattered photons can with field size because there are more line beams for larger fields.
further interact with the medium and return to the center line of PDD Because most radiosurgery fields are small, radiosurgery beams are
measurement. In general, however, this effect is insignificant. (c) In the close to case (b)
Radiation therapy’s aim is to kill the target tissue (or cell)

through enough impact from the energy deposition while
minimizing the impact to the normal tissue. In other words,
local control of treatment depends on how much dose is
delivered to the target (and how accurately, too, of course),
and normal tissue complication is dependent on how much
unwanted dose is given to the normal structure.
There are two commonly used dose-calculation methods:
measurement based and model based. Measurement-based
approaches directly use dose distributions measured in phan-
toms. On the other hand, model-based approaches try to cal-
culate the dose from the first principles with limited use of
measured data.
Fig. 4.14 A typical PDD with a narrow beam is illustrated. The dose is
close to zero at the surface and starts to build up with depth until reach-
ing the maximum. This region is called the build-up region. After the
Measurement-Based Dose Calculation depth of dose maximum, it then begins to decrease exponentially. Note
that depth dose (DD) is displayed in the “log” scale whereas depth is in
Dose calculation is mainly based on three parameters: per- the “linear” scale (i.e., log–linear plot). In a log–linear plot, an expo-
cent depth dose (PDD), profile, and output. nential decrease is expressed as a line
Percent Depth Dose electrons rather than by photons themselves. Because most
When a photon beam is incident on a medium, it creates a of the ejected electrons lose energy continuously while
dose distribution within the medium. Relative dose at any traveling through the medium, there is a distance lag
depth is called the depth dose (DD) and when it is expressed between the photon interactions and the dose. This mecha-
as a percentile it is called the PDD. Let us assume a certain nism also creates a so-called build-up region in the shallow
number of photons in a hypothetical narrow beam are inci- depth of the medium. Although the number of interactions
dent on a thin medium (Fig. 4.13a). The number of photons is the highest at the surface of the medium, dose is almost
decreases exponentially because some photons interact zero (because most ejected electrons move deeper into the
with the medium and disappear from the path of the narrow medium). Thus, dose is close to zero at the surface, starts to
beam. Now think of energy deposition (i.e., dose) at depth. build up with depth, and reaches the maximum. Then, the
The absorbed dose is proportional to the amount of radia- dose begins to decrease exponentially. Figure 4.14 shows a
tion interactions. Energy, however, is deposited by ejected typical shape of DD (or PDD) for a narrow beam. In reality,
Fig. 4.17 (a) A typical dose profile is illustrated for a conventional

beam. The region of rapid dose change is called the penumbra. (b) A
typical dose profile for a radiosurgery beam
edge of the field and slowly approaches zero dose (<1 %)

away from the field edge. The region of rapid dose fall-off is
Fig. 4.15 Comparison of PDD curves for various field sizes. As the called the penumbra. When a field is large, the flat plateau in
depth gets deeper, PDD increases with field size because of the the central area is clearly observed. As the field size becomes
increased amount of scatter dose contribution. Note that it is intention- smaller, the plateau disappears as illustrated in Fig. 4.17b,
ally shown in “linear–linear” scale to illustrate that PDD is often drawn
in two ways, “linear–linear” and “log–linear” (see Fig. 4.13)
which is a profile for a typical radiosurgery beam. In radio-
surgery, most radiation fields (<50 mm in diameter) show
either no plateau region or a very narrow plateau region.
Because of the large dose variation within a small region, a
high-resolution measurement detector is required for dose
profile measurement [24–30]. The shape of the individual
radiation field is defined by the collimator system that is spe-
cifically based on the radiosurgery delivery system.
Output
Fig. 4.16 An ideal dose profile in a plane is illustrated The magnitude of radiation produced by the delivery system
is called the output. In general, this is a relative value that is
normalized to the output of a reference field size. Output var-
the target volume is finite in size (Fig. 4.13b). In this sce- ies with field size (or collimator size). Larger field sizes have
nario, some of the scattered photons can scatter back to the higher output.
center of the field and contribute to the dose at depth. This With the PDD, OAR, and output for a given geometry,
effect is minimal, however. relative dose at any point is simply a multiplication of these
The scatter effect is significantly enhanced with a broad three parameters. Absolute dose is obtained when this value
beam (Fig. 4.13c) because more scattered photons can con- is multiplied with the machine calibration factor and monitor
tribute to the dose. The chance of multiple scattered photons units (MU) in the linear accelerator (or calibration dose rate
coming to the central line increases with field size because and treatment time in Gamma Knife).
more scattered photons exist in a larger field, so PDD
increases with field size (Fig. 4.15). In radiosurgery, how-
ever, most fields are relatively small, so the scatter effect is Model-Based Dose Calculation
insignificant. Thus, most radiosurgery beams are similar to
the case shown in Fig. 4.13b, and PDD can be analytically Most model-based dose-calculation methods define a kernel,
expressed as a simple exponential function. which is a function that describes how the dose is distributed
within a medium (mostly water). Two commonly used ker-
Profile or Off-Axis Ratio nels are point kernel and pencil-beam kernel.
Figure 4.16 shows an ideal dose profile, often called the off-
axis ratio (OAR), in a plane where a certain amount of dose Point Kernel
is given uniformly inside the field and no dose is deposited When a photon particle interacts at a point within a medium,
outside the field. In reality, it is not possible to obtain such a a certain amount of energy is released per mass, which is
profile. A typical dose profile in a plane at a certain depth for called the total energy released per unit mass (TERMA).
a clinical radiation beam is shown in Fig. 4.17a. The dose is A part of this energy is absorbed at the point of interaction
uniform in the central area but starts to gradually decrease as (often called the primary interaction point), but the rest of it
it nears the field edge. It then drops rapidly at the geometric is dispersed around the point. The scattered photon can also
travel to any other points, interact, and deposit energy around In general, kernel values are preobtained by calculations,
the area of interaction. Therefore, a photon interaction can and sometimes the pencil-beam kernel can be extracted from
contribute dose to the whole volume. The point kernel measurement data.
describes how dose is distributed with respect to the interac-
tion point within the medium, as illustrated in Fig. 4.18.
In general, the distributed dose is expressed as a ratio Isodose Line and Dose–Volume Histogram
between the energy deposited at any point and the total
energy released during the interaction at the primary interac- Once the dose is calculated in the treatment planning system,
tion point. The dose at any point is a summation of kernels each dose distribution is usually displayed in an isodose line,
multiplied with the TERMA from all of the primary photon which is made up of all the points of the same dose connected
interactions. TERMA is easily calculated using analytical in a line. In radiosurgery, isodose lines are expressed as a per-
expression of interaction mechanisms. centile of the maximum dose. Dose distribution within the tar-
get is much more nonuniform in radiosurgery than in
Pencil-Beam Kernel conventional radiotherapy. The maximum dose is usually
The pencil-beam kernel includes dose contributions from all observed in the center of the target, and it is usual to have very
of the primary interactions along a ray instead of at one inter- stiff dose fall-offs near the target boundary. Prescription is
action point. Figure 4.19 illustrates a pencil-beam kernel. If assigned to isodose lines of 50–80 % depending on the situa-
photon energy fluence (the same as the number of photons tion. A prescription isodose line is higher in the case of a sin-
times the corresponding energy per area) is known, dose at gle isocenter and lower with multiple isocenters. Figure 4.20
any point is simply a summation of kernels multiplied with shows isodose lines displayed in absolute value on three (axial,
energy fluence. sagittal, and coronal) planes of tomographic image.
Fig. 4.18 A point kernel is illustrated, which is a distribution of energy

deposited at any point to total energy released at the primary interaction Fig. 4.19 A pencil-beam kernel is illustrated, which is a distribution of
point deposited energy from all primary interactions along a pencil-beam line
Fig. 4.20 An isodose line connects all points of the same dose. In this figure, absolute isodose lines are displayed in axial, sagittal, and coronal
planes of the CT image
Fig. 4.21 In a dose– Dose Volume Histogram

volume histogram (DVH), 1.0
accumulated volume (i.e.,
histogram) is described 0.9
according to dose for any
particular volume of 0.8
interest (e.g., the target).
In this case, 95 % of the 0.7
volume of the target will
Norm.Volume
get a dose equal to or 0.6
more than 2,000 cGy
0.5
0.4
0.3
0.2
0.1
0.0
0 500 1000 1500 2000 2500 3000
Dose (cGy)
Dose–volume histogram (DVH) describes accumulated

volume (i.e., histogram) according to dose for any particular
volume of interest (e.g., the target) as illustrated in Fig. 4.21.
In Fig. 4.21, 95 % of the volume of the target will get a dose
equal to or more than 2,000 cGy. Although the DVH does not
show spatial information, it is a useful tool for plan evalua-
tion [31, 32]. In radiosurgery, the DVH for the target usually
shows a gentle slope on the high-dose side because of a sig-
nificantly nonuniform dose profile of small fields. With the
lower isodose line selected for prescription, a gentler DVH
slope on the high-dose side is expected. In the case of mul-
tiple isocenters, field overlap causes a significant dose
increase in the central volume of the target, resulting in a
high dose tail in DVH.
Photon Beam Delivery Systems

Fig. 4.22 A photograph of a Gamma Knife unit
Gamma Knife
Knife mainly consists of the radiation unit, the operating
In the late 1960s, Leksell introduced a prototype of Gamma table and sliding couch, a set of four collimator helmets, and
Knife, which is a dedicated radiosurgery unit [5]. Leksell’s a control unit. Figures 4.22 and 4.23 show a photograph of a
prototype incorporated 179 60Co radioactive sources placed Gamma Knife unit and a simple schematic diagram,
over a 60° × 160° sphere. The modern Gamma Knife models respectively.
incorporate 201 60Co radioactive sources that converge and The 60Co radioactive source is in pellet form, and 20 pellets
focus on the treatment target at a source-to-target distance of are encapsulated in a steel capsule. Each steel capsule is 1 mm
about 40 cm [6, 7]. Each source contributes a clinically insig- in diameter and 2 mm in height. Sources are aligned with three
nificant dose following the beam line. However, when many collimator systems, the precollimator, the stationary collima-
beam lines are converged on the focus where the target is, the tor, and the final collimator of a helmet. Both the precollima-
therapeutic dose is delivered to the target while the surround- tor and the final collimator of the helmet are made of tungsten
ing normal tissue dose remains under the limit. Gamma alloy whereas the stationary collimator is made of lead.
Rear shield Middle shield

Front shield
Central body Upper shielding

Plug door
Treatment table
Couch
Helmet
Position
indicator
Trolley
. . Lower shielding
door
Fig. 4.24 A photograph of a collimator helmet. A properly chosen hel-

met, depending on the target size, is connected to the sliding couch
.. ..
. .
Fig. 4.23 A simple schematic diagram of a Gamma Knife unit
The total activity of the Gamma Knife unit with new 201
sources loaded is of the order 6,000 Ci (2.22 × 1014 Bq, where
Bq is the same as disintegration per second, and Ci is
3.7 × 1010 Bq). The activity of each source is about 30 Ci and
its variation is expected to be within ±5 %. With 6,000 Ci, a
typical dose rate is about 300 cGy/min at the center of the
16-cm-diameter, spherical water phantom. As described ear-
lier, this value drops down to 150 cGy/min (50 % of initial
dose rate) 1 half-life (5.26 years) later.
Four helmets of different collimator sizes, 4, 8, 14, and Fig. 4.25 Picture of a Leksell stereotactic frame. The frame is fixed to
18 mm in diameter, are provided. Figure 4.24 is a photograph the patient’s skull and is connected to the couch
of a collimator helmet. A proper helmet is chosen depending
on the target size and is connected to the sliding couch. The
Leksell stereotactic frame (Fig. 4.25) is fixed to the patient’s directly escapes the unit even with the doors open. When the
skull and connected to the couch. Once the target position is doors are closed, exposure rates by leakage radiation are sig-
aligned to the focus, the sliding couch is pushed into the nificantly reduced.
mechanically set treatment position. Its mechanical spatial Several auxiliary capabilities have been developed lately
accuracy is claimed to be in submillimeters (about 0.3 mm). and they provide software that tracks the patient’s position
In treatment mode, the shielding doors are open, resulting in ten times per second. A robotic automatic positioning system
an increase of scatter dose inside the treatment room. When repositions each isocenter with submillimeter precision
the sliding couch is withdrawn and the shielding doors are without operator intervention.
still open, the scatter dose becomes about 1.3 times higher. The current state-of-art Gamma Knife® unit is called
Because of the unit’s design, however, no primary beam Perfexion™ and it is significantly different from previous
Fig. 4.27 The Gamma Knife® PerfexionTM unit in Princess Margaret

Hospital, Canada. A cone beam CT system has been added for better
image-guidance. (Courtesy of Young-Bin Cho, PMH)
Fig. 4.26 An inner view of the collimator in PerfexionTM unit. There rotation axes (gantry, collimator, and table) often do not
are eight sectors and 72 holes (3 hole sizes × 24 sub-collimators) in each intersect at a common point. Instead, they only near each
sector. (Courtesy of Young-Bin Cho, PMH)
other within a sphere. The center of this sphere is considered
the nominal isocenter.
versions in its design. There exist a total of 192 sources and The size of the sphere is the key factor in the linear accel-
it has a unique collimator that is a permanent device divided erator’s spatial accuracy, and the smaller the better. Common
into eight moveable sectors, each having 24 sub-collimators consensus says that the sphere needs to be less than 1 mm in
arranged in five concentric rings [33]. The collimator holes diameter for radiosurgery based on the fact that the resolu-
are sized as 4, 8, and 16 mm and either one of them or none tion of modern imaging modalities is around 1 mm. Another
of them can be assigned to each sector. Therefore four field important factor for spatial accuracy is how precisely posi-
sizes (i.e., 0, 4, 8, and 16 mm) are available for each sector tioned the patient is. In a linear accelerator, the patient’s
and automatic change of field size can be made by adjusting positioning is usually determined using the treatment room
the position of the movable sector. Such design enables effi- laser system. Three lasers, one from the ceiling and two from
ciently delivering many diverse combinations of beams to side walls, are aligned to point to the nominal isocenter. By
obtain dose distributions conforming to even highly irregular matching the laser crosshairs and the desired marks on the
targets. Figure 4.26 shows an inner view of the collimator. It frame attached to the patient’s skull, the target is aligned to
can be seen there are eight sectors and 72 holes (3 hole the nominal isocenter. Therefore, laser system accuracy is
sizes × 24 sub-collimators) in each sector. significantly important when used for patient positioning.
In Princess Margaret Hospital, Canada, a modification As described earlier, linear accelerators are equipped with
was made on a Gamma Knife® Perfexion™ unit to add a a so-called secondary collimator system that provides a rect-
cone beam CT system as shown in Fig. 4.27. angular field for conventional radiotherapy, but rectangular
Radiosurgery using Gamma Knife has been performed fields are not considered appropriate for radiosurgery.
for more than 250,000 patients worldwide during the past 30 Therefore, additional circular collimators are commonly
years. Many clinical studies with long years of follow-up used as tertiary collimators [10–14]. Circular beams are
have been reported. superior to rectangular beams for reasons like the sharper
beam, easier dosimetry calculation, more precise beam
delivery, and better field definition for small fields. Because
Isocentric Linear Accelerator tertiary collimation happens closer to the patient, the beam is
more accurately aligned and the penumbra is reduced. More
Most clinical linear accelerators are isocentric, which means rapid dose fall-off outside the target can therefore be
that the gantry, the collimator, and the table can be rotated obtained. Collimator sizes typically range from 5 to 40 mm
with respect to one point in the space called the isocenter. In to treat a variety of lesion sizes and shapes. Most circular
reality, the isocenter may not be a point because three collimators are made of tungsten alloy or lead and are
Fig. 4.28 Several circular collimators in different sizes of diameter (5,

10, 12, 14.20, 24, and 30 mm) used at the University of Florida are
shown
5–10 cm thick. Figure 4.28 shows several circular collima-

tors with different diameters used at the University of Florida.
At the University of Florida, a special device called a
floor stand was developed [34, 35] that improves the accu-
racy of the beam’s focus to the isocenter by adding a set of
bearings to the stereotactic collimator system that accounts
for imperfections in the gantry rotation. A set of bearings is
also attached to the patient and the target area to bypass the
Fig. 4.29 A setup of the floor stand used at the University of Florida is
inaccuracies of the table rotation. It is reported that the floor
shown. The floor stand is mechanically fixed to the holes in the floor
stand can provide mechanical spatial accuracy within
0.2 ± 0.1 mm to define the isocenter [34, 35]. Figure 4.29
shows how the floor stand is set up.
With the introduction of a mMLC, which has many colli-
mator leaves that can be individually moved to conform to
the shape of the target, it is possible to do two-dimensional
conformal radiosurgery for irregularly shaped targets with a
single isocenter (Fig. 4.30). In addition, dynamic conformal
radiosurgery can be performed with the aid of the automatic
field shaping capability while the gantry rotates. It is also
possible to do intensity modulated stereotactic radiosurgery
(IMSR) if the treatment planning system supports the appro-
priate dose optimization [36]. The collimator leaf is usually
made of tungsten. The width of each leaf is an important
parameter related to the conformality and it ranges from 1.5
to 6 mm at the isocenter depending on the manufacturer.
State-of-the-art linear accelerators incorporate a volumet- Fig. 4.30 A photograph of a commercial mMLC. As can be seen, there
ric imaging system referred to as a cone beam computed are many collimator leaves, and each of them can be individually moved
tomography (CT) system. A diagnostic X-ray tube is added to conform to the shape of the target
on a separate gantry at a 90° angle from the treatment gantry,
and a flat panel detector is installed on the opposite side as This advancement opened the door to exploring the possibili-
shown in Fig. 4.31. Volumetric imaging can be obtained in a ties with image-guided radiosurgery using isocentric linear
couple of minutes by recording multiple cone beam geometry accelerators. Improved setup accuracy by virtue of image-
planar images while the gantry rotates. Currently, it is not guidance also encouraged many clinicians to explore frame-
clear whether this capability needs to be used for radiosurgery. less radiosurgery with isocentric linear accelerators [37].
Cyberknife 6-MV beam. It is much smaller and lighter than a conventional

S-band (2,856 MHz) linear accelerator. CyberKnife incorpo-
CyberKnife, developed by Accuray Inc. (Santa Clara, CA) in rates an image-guided system consisting of two diagnos-
collaboration with Stanford University (Stanford, CA), uses tic X-ray tubes mounted on the ceiling and two detectors
a miniature linear accelerator mounted on a robotic arm as placed on the floor. With this X-ray image-guided system,
shown in Fig. 4.32 [38, 39]. The miniature linear accelerator CyberKnife continuously and automatically tracks a tumor’s
operates in the X-band RF (9,300 MHz) and provides a precise location throughout the procedure, detects any tumor
or patient movement, and makes corrections as needed.
Because of its image-guidance capability, the radiosurgery
procedure is performed without an invasive head frame.
Even without the frame the use of intelligent robotics tech-
nology enables treatments with submillimeter accuracy.
The robot can position the linear accelerator at any point
within a spherical shell of 60–100 cm from the target with a
precision of ±0.5 mm. In principle, the frameless approach
gives more choices in beam-angle selection compared with
radiosurgery with a frame. The robotic arm must, however,
avoid a collision with the patient, table, and imaging system
as well as direct beam incidence into the imaging system.
Another concern is the long treatment time compared with
other delivery systems like the Gamma Knife and the isocen-
tric linear accelerator. A total of 12 collimators from 5 to
60 mm in diameter are provided. Installing the unit would
require a 10-ft ceiling over a 12 × 16 ft2 area.
A variable aperture collimator, called IRIS, has been
introduced recently to allow the field size to be varied during
treatment delivery [40]. The IRIS incorporates twelve 60 mm
tall prism-shaped tungsten–copper alloy segments. They are
arranged in two banks stacked in the direction of radiation
Fig. 4.31 In an advanced linear accelerator, a diagnostic X-ray tube is path. The banks are rotated by 30° with respect to each other
added on a separate gantry at a 90° angle to the treatment gantry, and a to reduce leakage through gaps between segments. With this
flat panel detector is installed on the opposite side. Using an X-ray tube
and cone beam geometry reconstruction software, volumetric images variable aperture collimator, treatment time is expected to
can be obtained in the treatment room significantly decrease.
Fig. 4.32 A CyberKnife

unit uses a miniature linear
accelerator mounted on a
robotic arm. (Courtesy of
Hee Jung Kim, Soon Chun
Hyang University Hospital,
Korea.)
Fig. 4.33 In cyclotrons, protons are injected at the center of two halves
(“dees”) of an electrically powered circular dipole magnet with a con-
stant magnetic field. The proton acceleration occurs within the narrow
gap between the two half-magnets employing a correctly phased alter-
nating electric field. The protons gain energy, twice for each revolution,
while moving in a constant circular path. Accelerated protons with a
fixed energy and a continuous beam current can be extracted
Fig. 4.34 In synchrotrons, the protons are accelerated in an orbit with
a constant radius. As the proton energy increases, the magnetic field
strength also increases to keep them in a stable orbit
Proton Therapy
Clinical proton beams are produced by injecting hydrogen However, synchrotrons do not require energy degraders, thus
atoms with their electron stripped in an accelerating struc- minimizing secondary radiation production in the beam line.
ture; electric fields accelerate the free protons to the desired An elaborate beam transport system (BTS) carries the
clinical energy. The proton accelerators fall into two broad protons exiting the accelerator to the patient site. The beam
categories [20]. Cyclotrons are the classic proton accelera- line consists of focusing, bending, and switching magnets
tors in which protons are injected at the center of two halves that steer a pencil beam of protons to different treatment
(called “dees”) of an electrically powered circular dipole rooms. The length of the beam line can easily exceed 30 m.
magnet with a constant magnetic field (Fig. 4.33). The pro- A stable and precise BTS is required for reproducible dosim-
ton acceleration occurs within the narrow gap between the etry and patient treatments. The stability of the centroid of
two half-magnets employing a correctly phased alternating the beam position in the beam line must be better than
electric field. The protons gain energy, twice for each revolu- 1.00 mm. Therefore, the bending and focusing magnets have
tion, while moving in a constant circular path. Accelerated to be very stable, mechanically and electrically, during the
protons with a fixed energy and a continuous beam current of operation of the proton machine.
up to 1 milliampere (mA) can be extracted and transported
into an evacuated tube (beam line). Proton energy is changed
by inserting an energy degrader in the path of the beam A Proton’s Interaction with Matter
extracted from the cyclotron.
The other type of proton accelerator is the synchrotron, Protons are positively charged elementary particles that con-
which allows preaccelerated protons (up to a few MeV) to be tinuously lose energy and go through small-angle deflections
injected into an accelerating chamber equipped with a num- as they travel through matter before they get absorbed. While
ber of electromagnets. The protons are accelerated in an orbit the radiation dose deposited by photons drops off exponen-
with a constant radius (Fig. 4.34). As the proton energy tially with penetration depth, the dose deposited by protons
increases, the magnetic field strength also increases to keep increases very slowly for about three-quarters of its range of
them in a stable orbit. The acceleration cycle in synchrotrons travel in the medium before increasing sharply, reaching a
is typically 2 s. Therefore, it is only possible to obtain proton maximum value before rapidly dropping off to 0. The depth
beams in pulses with a repetition frequency of 30 cycles per at which the maximum energy is deposited by protons is
minute and the energy of the protons can be varied from called the Bragg peak. Figure 4.35 illustrates a depth dose of
pulse-to-pulse. Operating the synchrotrons is complex a proton beam showing its Bragg peak. The typical peak to
because it requires rapid magnetic field variations to operate. entrance dose ratio of narrow proton beams is 4 or higher.
than five non-coplanar arcs are used. A recent study for

trigeminal neuralgia treatment reports that compared with
Gamma Knife, linear accelerator-based deliveries exhibit a
flatter top at the high isodose line level (>50 %) and faster
dose-off at the low isodose line levels (<50 %) [41]. The pen-
umbral widths (80 to 20 %) were 1 mm for Gamma Knife
with a 4-mm helmet and 2.1 mm for the 6-MV linear accel-
erator with a 5-mm collimator in another study [42].
Radiosurgery with a proton beam is different from both a
linear accelerator and the Gamma Knife. Proton radiosur-
gery is performed with shaped and fully compensated fields,
widely separated in the entrance angle, and with a single iso-
center based on the characteristics of the proton beam, that
is, the Bragg peak [43].
The proton beam, Gamma Knife, and linear accelerator
Fig. 4.35 The depth dose of a proton beam is almost constant to a have been compared in five clinical cases [3]. The optimal
certain depth and starts to increase, then it jumps significantly at the modality for SRS depends on target size, shape, and location;
farthest end of the path. This peak is called the Bragg peak
however, all modalities are equally good if the target is small
and regular. In the case of trigeminal neuralgia, linear accel-
erator based SRS is effective and can be comparable with
The peak’s position is proportional to the energy of the Gamma Knife [42]. Gamma Knife is superior to the linear
proton beam. The range of 230-MeV protons is approxi- accelerator with regard to the conformality. In contrast, the
mately 33.00 cm, which is sufficient to penetrate any part of linear accelerator is superior with regard to the dose fall-off
the body. The typical range of proton energies used for radio- outer region of the target volume [44]. The linear accelerator
surgery is 70–150 MeV (4–15 cm depth). with mMLC can be better than Gamma Knife when hearing
The clinical target volumes are usually thicker than the preservation is important during treatment [45].
width of the Bragg peaks; therefore, the range of the proton
beam must be modulated to spread out the Bragg peak. By
changing the energy of the incident proton beam with a vari- Quality Assurance
able range shifter, the Bragg peaks will “stack” at different
depths. The variable range shifter can be a binary filter, a In most conventional radiation therapy treatments, pre-
double-wedge variable absorber, or a circular wedge. scribed dose is delivered in many fractions, and dose per
Scattering foils or active scanning of a narrow beam with each fraction is relatively small; however, a significantly
scanning magnets will laterally spread the proton beams to large dose is delivered in a single treatment during
achieve uniform intensity across the target volume. radiosurgery. Therefore, the impact of the inaccurate local-
ization of the target or dose delivery in radiosurgery can be
disastrous. To minimize such disasters, SRS quality assur-
Dose Characteristics in Different Modalities ance (QA) should be stringent. Both an international QA
task group and Task Group 42 of the American Association
In 1951, Leksell introduced the concept of SRS, and the first of Physicists in Medicine (AAPM) have published general
patient was treated with the Gamma Knife prototype in 1967. recommendations [46, 47]. The Radiation Therapy Oncology
Since then, Gamma Knife has only seen improvements with Group (RTOG) has also published guidelines [48].
the establishment of many linear accelerator-based radiosur- In general, radiosurgery QA consists of common QA for
gery programs during the past 20 years. The use of proton the overall performance of all radiosurgery equipment, valid
beams for radiosurgery is also increasing. Some research for a relatively long term, and specific QA for the calibration
groups have investigated and compared the dosimetric char- and preparation of equipment on each treatment day.
acteristics of each modality.
The linear accelerator is thought to be able to provide
dosimetric characteristics similar to Gamma Knife when Common QA
multiple, noncoplanar arcs are used. In the case of more than
five noncoplanar arcs, the full width at the half maximum Common QA procedures need to be set to periodically verify
(FWHM) of a 5-mm circular collimator field is very close to equipment status and performance. It should include target
that of Gamma Knife with a 4-mm helmet. The dose fall-off localization QA, basic dosimetry QA, treatment planning
of the linear accelerator is also almost identical when more QA, and output calibration and delivery QA. A detailed QA
Fig. 4.36 A sample of a patient-specific QA checklist used for linear accelerator-based radiosurgery at the University of Florida. In this sheet,
checklist items are listed chronologically
schedule must be set based on frequency and consequence:

any item that has either high failure rate or severe conse-
quence must frequently be checked.
Overall target localization can be tested using a phantom
containing a target located in the known geometry. Accuracy
of within 1 mm is recommended.
The treatment planning system’s dosimetric and nondosi-
metric conditions should both be checked. Specific items are
Fig. 4.37 An example of verification film. It can be seen that the steel well described in the recommendations of the AAPM Task
ball in the calibration jig and radiation beams are well aligned Group 53 [49].
Dosimetry-related QA procedures are specific to the angle, gantry end angle, and monitor unit, are verified. In
delivery system. In the Gamma Knife unit, the following are subsequent isocenter treatments, the floor stand is set by a
periodically verified: dose rate at the center of a 16-cm- physicist and coordinates are independently double-checked
diameter tissue-equivalent sphere at the focus; shutter error: by two other staff members. Specific QA in the Gamma
connections of frame, collimator helmets, and sliding couch; Knife unit is relatively less intense. It is basically ready for
and leak test on collimator helmets. For linear accelerators, treatment in most cases, and only a few safety checks need to
output is checked daily. If the laser system is used for patient be performed.
alignment, its accuracy needs to be verified rigorously. The
most important QA item is the mechanical accuracy of the
nominal isocenter. Less than a 1-mm diameter around the Conclusion
isocenter sphere is recommended [14, 50]. If an independent
floor stand is used, like at the University of Florida, the accu- SRS is a special radiotherapy technique that holds promise
racy is directly dependent on the floor stand. Thus, no addi- not only radio-therapeutically but also as a noninvasive
tional QA tests on the laser and linear accelerator isocenter, surgical tool. The underlying physics principles for radiosur-
other than those for routine conventional radiotherapy, are gery are the same as those for conventional radiation therapy;
required. however, the accuracy requirements in dosimetry and patient
positioning for radiosurgery are at the millimeter level. There
are several dosimetric parameters that must be measured in
Specific QA tissue-equivalent phantoms such as water including dose
calibration, percentage depth doses, relative dose output fac-
The goal of specific quality assurance is to make sure that the tors, and cross-beam profiles. Radiosurgery treatment fields
machine is running adequately, that all auxiliary devices are are often very small; therefore, the spatial resolution require-
prepared and working properly, and that treatment parameters ments for radiation detectors are much more stringent in
are correctly set just before and during patient treatments. radiosurgery.
Before placing the frame into the patient, the treatment
streamline should be checked to minimize unnecessary treat-
ment delay or cancellation after the frame is on. Figure 4.36 Glossary
shows a sample of a patient-specific QA checklist used for
linear accelerator based radiosurgery at the University of μSv Microsievert, the SI unit for an equiva-
Florida. As you can see, check items are listed chronologically. lent radiation dose that accounts for bio-
Once the secondary collimator is set as needed, the collimator logical effect.
drives are physically disconnected to prevent an accidental AAPM American Association of Physicists in
change of the secondary collimator. To prevent the treatment Medicine.
table from accidentally dropping, a mechanical block is Accelerator A device that accelerates subatomic
applied to the column of the table. Because the patient’s head charged particles or nuclei to high
is connected to the floor stand while his or her body is on the energies.
table, an abrupt drop of the table could cause serious prob- Alpha particle Positively charged particle consisting of
lems as serious as a patient death. For the first isocenter treat- two protons and two neutrons.
ment, the accuracy of focusing the radiation fields to the Beta particle High-speed electron or positron.
isocenter is checked by a Winston–Lutz filming test. Binding energy Net energy required to remove an atomic
A staff member sets the floor stand to the treatment posi- electron from its orbit.
tion. A steel ball in the calibration jig is set to the isocenter Bq Becquerel, the Système International (SI)
coordinates independently by another staff member and unit of radioactivity equal to one disinte-
attached to the floor stand. Then, the verification film is irra- gration per second.
diated in four different beam angles with a 24-mm-diameter Bragg peak Peak can be seen at the end of the depth
circular collimator. If both the floor stand and calibration jig dose curve of heavy charged particles
are set to the isocenter correctly, and beam alignment is such as a proton.
appropriate in all beam angles, images of the ball must be BTS Beam transport system, a proton acceler-
centered within all the circular fields. Figure 4.37 shows an ator system that carries the protons exit-
example of the verification film. Once the film is confirmed ing the accelerator to the patient site.
satisfactory, the patient is moved to the floor. Before the Cavity A device through which power is coupled
treatment beam is on, the size of the circular collimator and to the accelerating particles in a resonant
the treatment parameters, such as the table angle, gantry start mode.
Ci Curie, unit of radioactivity equal to keV Kiloelectron volts, a unit of energy of

3.7 × 1010 disintegrations per second. particles.
Collimator A device capable of collimating Klystron A microwave amplifier.
radiation. Linear accelerator A particle accelerator in which the
CT Computed tomography. path of the particles is straight.
Cyclotron A particle accelerator in which mA Milliampere, a unit of electric current.
charged particles are accelerated in a Magnetron A diode-type electron tube that gen-
circular path by an alternating elec- erates microwave pulses.
tric field in a constant magnetic field. MeV Million electron volts, a unit of
DD Depth dose, relative dose level energy of particles.
according to the depth of the mate- mMLC Micro-multileaf collimator, a special
rial when the radiation beam is collimator consisting of many small
incident. collimator leaves that can be individ-
Dipole A pair of electric charges or mag- ually moved to conform to the shape
netic poles. of the target.
Dose Energy absorbed per unit mass by mR Milliroentgen, a unit of exposure.
radiation. MV Million volts, a unit of beam quality
DVH Dose–volume histogram, accumu- of radiation therapy beams produced
lated volume according to dose for a in linear accelerators.
volume of interest. OAR Off-axis ratio, relative dose distribu-
FF Flattening filter. tion following a line on a plane nor-
FFF Flattening filter free. mal to the direction of the radiation
Filter A device used to modify the inten- beam, same as profile.
sity of radiation. Output Magnitude of radiation produced.
FWHM Full width at half maximum. PDD Percent depth dose, relative dose level
Gamma particle Electromagnetic radiation particle in percentile according to the depth of
(photon) emitted by radioactive decay. the material when the radiation beam
Gy Gray, a unit of dose equivalent to is incident.
joules per kilogram. Photon Quantum of electromagnetic energy
Half-life Length of time needed for a radio- with no mass and charge considered
active substance to lose half of its to be both particle and wave.
radioactivity from decay. Positron Positively charged particle of the
Ionization chamber A chamber for determining the inten- same mass and magnitude of charge
sity of ionizing radiation by measur- as an electron.
ing ionization of the enclosed gas. Profile Relative dose distribution following a
Ionizing radiation Electromagnetic or particulate radia- line on a plane normal to the direction
tion capable of producing ions in its of radiation beam, same as the OAR.
passage through matter. Proton Stable and positively charged sub-
IMRT Intensity modulated radiation therapy. atomic particle.
IMSR Intensity modulated stereotactic QA Quality assurance.
radiosurgery, a treatment technique Radioactive decay Spontaneous disintegration of a
in which radiation intensity is modu- radionuclide followed by the emis-
lated within a field to obtain an opti- sion of ionizing radiation in the form
mized dose distribution. of alpha, beta, or gamma particles.
IRIS A variable aperture collimator in Radioactivity Capability of radioactive decay.
CyberKnife system. Radioisotope Naturally or artificially produced
Isocenter A point in the space that is the center radioactive isotope of an element.
of rotation of the gantry, collimator, Range Distance charged particles can travel.
and table of an isocentric medical Resonance Condition of being resonant in which
linear accelerator. an increase in the amplitude of a
Isodose line A line made by connecting all points physical quantity can occur.
of the same dose. RF Radiofrequency, a range of electro-
Kernel A function that describes how the magnetic frequencies above sound
dose is distributed within a medium. and below visible light.
SBRT Stereotactic body radiotherapy. 17. Kjellberg RN, Shintani A, Frantz AG, Kliman B. Proton-beam ther-
SRS Stereotactic radiosurgery. apy in acromegaly. N Engl J Med. 1968;278(13):689–95.
18. Larsson B, Leksell L, Rexed B, Sourander P, Mair W, Andersson
SRT Stereotactic radiotherapy. B. The high-energy proton beam as a neurosurgical tool. Nature.
Synchrotron A particle accelerator in which 1958;182(4644):1222–3.
charged particles are accelerated 19. Breuer H, Smit BJ. Proton therapy and radiosurgery. Berlin:
around a fixed circular path by an Springer; 2000.
20. Wieszczycka W. Proton radiotherapy accelerators. Singapore:
electric field and held to the path by World Scientific Publishing; 2001.
an increasing magnetic field. 21. Karzmark CJ, Nunan CS, Tanabe E. Medical electron accelerators.
TERMA Total energy released per unit mass, New York: McGraw-Hill, Health Professions Division; 1993.
amount of energy released from radi- 22. Shiu AS, Kooy HM, Ewton JR, Tung SS, Wong J, Antes K, et al.
Comparison of miniature multileaf collimation (MMLC) with cir-
ation to the unit mass of the interac- cular collimation for stereotactic treatment. Int J Radiat Oncol Biol
tion material. Phys. 1997;37(3):679–88.
Z Atomic number; that is, the number of 23. Georg D, Knoos T, McClean B. Current status and future perspec-
protons in an element. tive of flattening filter free photon beams. Med Phys.
2011;38(3):1280–93.
24. Arcovito G, Piermattei A, D’Abramo G, Bassi FA. Dose measure-
ments and calculations of small radiation fields for 9-mv x rays.
References Med Phys. 1985;12(6):779–84.
25. Bova FJ. Radiation physics. Neurosurg Clin N Am.
1. Evans RD. The atomic nucleus. Malabar, FL: Krieger; 1955. 1990;1(4):909–31.
2. Attix FH. Introduction to radiological physics and radiation dosim- 26. Houdek PV, VanBuren JM, Fayos JV. Dosimetry of small radiation
etry. New York: Wiley; 1986. fields for 10-mv x rays. Med Phys. 1983;10(3):333–6.
3. Khan FM. The physics of radiation therapy. 2nd ed. Philadelphia: 27. Rice RK, Hansen JL, Svensson GK, Siddon RL. Measurements of
Lippincott Williams & Wilkins; 1992. dose distributions in small beams of 6 mv x-rays. Phys Med Biol.
4. Johns HE. The physics of radiology. 4th ed. Springfield, IL: Charles 1987;32(9):1087–99.
C. Thomas; 1983. 28. Heydarian M, Hoban PW, Beddoe AH. A comparison of dosimetry
5. Leksell L. Cerebral radiosurgery. I. Gammathalanotomy in two techniques in stereotactic radiosurgery. Phys Med Biol.
cases of intractable pain. Acta Chir Scand. 1968;134(8):585–95. 1996;41(1):93–110.
6. Wu A, Lindner G, Maitz AH, Kalend AM, Lunsford LD, Flickinger 29. Rustgi SN. Evaluation of the dosimetric characteristics of a dia-
JC, et al. Physics of Gamma Knife approach on convergent beams mond detector for photon beam measurements. Med Phys.
in stereotactic radiosurgery. Int J Radiat Oncol Biol Phys. 1995;22(5):567–70.
1990;18(4):941–9. 30. Vatnitsky S, Jarvinen H. Application of a natural diamond detector
7. Simonova G, Novotny J, Novotny Jr J, Vladyka V, Liscak for the measurement of relative dose distributions in radiotherapy.
R. Fractionated stereotactic radiotherapy with the Leksell Gamma Phys Med Biol. 1993;38(1):173–84.
Knife: feasibility study. Radiother Oncol. 1995;37(2):108–16. 31. Schell MC, Smith V, Larson DA, Wu A, Flickinger JC. Evaluation
8. Poffenbarger B. Viability of an isocentric cobalt-60 teletherapy unit of radiosurgery techniques with cumulative dose volume histo-
for stereotactic radiosurgery. MSc Thesis. Montreal, Canada: grams in linac-based stereotactic external beam irradiation. Int J
McGill University; 1998. Radiat Oncol Biol Phys. 1991;20(6):1325–30.
9. Poffenbarger BA, Podgorsak EB. Viability of an isocentric 32. Serago CF, Houdek PV, Bauer-Kirpes B, Lewin AA, Abitbol AA,
cobalt-60 teletherapy unit for stereotactic radiosurgery. Med Phys. Gonzalez-Arias S, et al. Stereotactic radiosurgery: dose-volume anal-
1998;25(10):1935–43. ysis of linear accelerator techniques. Med Phys. 1992;19(1):181–5.
10. Bellerive MR, Kooy HM, Loeffler JS. Linac radiosurgery at the 33. Giller CA, Fiedler JA, Gagnon GJ, Paddick I. Radiosurgical plan-
joint center for radiation therapy. Med Dosim. 1998;23(3):187–99. ning: gamma tricks and cyber picks. Hoboken, NJ: Wiley; 2011.
11. Betti OO, Derechinsky VE. Hyperselective encelphalic irradiation 34. Friedman WA, Bova FJ. The University of Florida radiosurgery
with linear accelerator. Acta Neurochir Suppl. 1984;33:385–90. system. Surg Neurol. 1989;32(5):334–42.
12. Colombo F, Benedetti A, Pozza F, Avanzo RC, Marchetti C, 35. Meeks SL, Bova FJ, Friedman WA, Buatti JM, Mendenhall
Chierego G, et al. External stereotactic irradiation by linear accel- WM. Linac scalpel radiosurgery at the University of Florida. Med
erator. Neurosurgery. 1985;16(2):154–60. Dosim. 1998;23(3):177–85.
13. Hamilton AJ, Lulu BA, Fosmire H, Stea B, Cassady JR. Preliminary 36. Webb S. Optimization by simulated annealing of three-dimensional
clinical experience with linear accelerator-based spinal stereotactic conformal treatment planning for radiation fields defined by a mul-
radiosurgery. Neurosurgery. 1995;36(2):311–9. tileaf collimator. Phys Med Biol. 1991;36(9):1201–26.
14. Lutz W, Winston KR, Maleki N. A system for stereotactic radiosur- 37. Minniti G, Scaringi C, Clarke E, Valeriani M, Osti M, Enrici
gery with a linear accelerator. Int J Radiat Oncol Biol Phys. RM. Frameless linac-based stereotactic radiosurgery (SRS) for
1988;14(2):373–81. brain metastases: analysis of patient repositioning using a mask
15. Larsson B. Dosimetry and radiobiology of protons as applied to fixation system and clinical outcomes. Radiat Oncol. 2011;6:158.
cancer and neurosurgery. In: Thomas RH, Perez-Mendez V, editors. 38. Adler JR, Cox RS. Preliminary clinical experience with the
Advances in radiation protection and dosimetry in medicine. cyberknife: image guided stereotactic radiosurgery. In: Kondziolka
New York: Plenum; 1980. p. 367–94. D, editor. Radiosurgery. Basel: Karger; 1995. p. 317–26.
16. Frankel KA, Phillips MH. Charged particle method: protons and 39. Maciunas RJ, Fitzpatrick M, Galloway RL, Allen GS. Beyond
heavy charged particles. In: Phillips MH, editor. Physical aspects of stereotaxy: extreme levels of application accuracy are provided
stereotactic radiosurgery. New York: Plenum; 1993. by implantable fiducial markers for interactive image-guided
neurosurgery. In: Maciunas RJ, editor. Interactive image guided micromultileaf collimator linear accelerator) for acoustic neuroma–
neurosurgery. Washington, DC: AANS; 1994. p. 261–70. and potential clinical importance. Int J Radiat Oncol Biol Phys.
40. Echner GG, Kilby W, Lee M, Earnst E, Sayeh S, Schlaefer A, et al. 2003;57(5):1450–9.
The design, physical properties and clinical utility of an iris collima- 46. Hartmann GH. Quality assurance program on stereotactic radio-
tor for robotic radiosurgery. Phys Med Biol. 2009;54(18):5359–80. surgery: report from a quality assurance task group. Berlin:
41. Ma L, Kwok Y, Chin LS, Yu C, Regine WF. Comparative analyses Pringer; 1995.
of linac and Gamma Knife radiosurgery for trigeminal neuralgia 47. Schell MC, Bova FJ, Larson DA, Leavitt DD, Lutz WR, Podgorsak
treatments. Phys Med Biol. 2005;50(22):5217–27. EB, et al. AAPM report no. 54. Stereotactic radiosurgery: report of
42. Gerbi BJ, Higgins PD, Cho KH, Hall WA. Linac-based stereotactic task group 42 radiation therapy committee. Woodbury, NY:
radiosurgery for treatment of trigeminal neuralgia. J Appl Clin Med American Institute of Physics; 1995.
Phys. 2004;5(3):80–92. 48. Shaw E, Kline R, Gillin M, Souhami L, Hirschfeld A, Dinapoli R,
43. Verhey LJ, Smith V, Serago CF. Comparison of radiosurgery treat- et al. Radiation Therapy Oncology Group: radiosurgery quality assur-
ment modalities based on physical dose distributions. Int J Radiat ance guidelines. Int J Radiat Oncol Biol Phys. 1993;27(5):1231–9.
Oncol Biol Phys. 1998;40(2):497–505. 49. Fraass B, Doppke K, Hunt M, Kutcher G, Starkschall G, Stern R,
44. Plowman PN, Doughty D. Stereotactic radiosurgery, X: clinical iso- et al. American Association of Physicists in Medicine radiation
dosimetry of Gamma Knife versus linear accelerator X-knife for therapy committee task group 53: quality assurance for clinical
pituitary and acoustic tumours. Clin Oncol (R Coll Radiol). radiotherapy treatment planning. Med Phys. 1998;25(10):
1999;11(5):321–9. 1773–829.
45. Perks JR, St George EJ, El Hamri K, Blackburn P, Plowman 50. Falco T, Lachaine M, Poffenbarger B, Podgorsak EB, Fallone
PN. Stereotactic radiosurgery XVI: isodosimetric comparison of BG. Setup verification in linac-based radiosurgery. Med Phys.
photon stereotactic radiosurgery techniques (Gamma Knife vs. 1999;26(9):1972–8.
Radiobiological Principles Underlying
Stereotactic Radiation Therapy 5
David J. Brenner and David J. Carlson
Since the gamma knife was first conceived in 1968 primarily

for arterial and functional lesions [1], single-fraction stereo- The Three Rs of Radiobiology:
tactic radiation therapy (SRT)1 often called radiosurgery, has Reoxygenation, Repair, and Repopulation
been widely used to treat a variety of cerebral lesions. By
1985, an alternative modality was available for SRT, using a These three radiobiological phenomena relate ultimately to
linear accelerator (linac) and a stereotactic head frame [5, 6]. cell killing processes, which is clearly an important (though
More recently, image-guided linac-based radiosurgery is now not the only, see later in this chapter) mechanism by which
in routine clinical practice [7–9], as is the CyberKnife, a fra- radiotherapy produces tumor control and induces side
meless robotic system for SRT [10], while proton radiosur- effects. We will discuss later in the chapter other potential
gery is increasingly used [11, 12]. The initial radiosurgery mechanisms of radiotherapeutic response, particularly at
concept was developed for intra-cranial lesions, but it is now high doses.
increasingly used for other sites, particularly for pulmonary
lesions [13].
In its early use, SRT was always applied in single frac- Reoxygenation
tions (i.e., as radiosurgery). More recently, there has been
increasing use of fractionated stereotactic radiotherapy [14–23]. The first radiobiological principle of importance here is that
This has been stimulated both by potential biological advan- malignant tumors, even those of limited size, almost always
tages, as discussed below, and by the development first of contain a proportion of hypoxic cells which, because of their
noninvasive relocatable frame systems [24, 25] and subse- deficiency in oxygen, are highly resistant to killing by X- or
quently of frameless image-guided stereotactic localization γ-rays [28]. Examples are shown in Fig. 5.1, showing
methods [26, 27]. hypoxic regions in sections of small tumors derived from
We will review here the radiobiological principles under- human glioma xenograph lines [29].
lying SRT, and their applications, to single- or multi-fraction Figure 5.2 shows a dose–response curve, derived from
radiotherapy of the three main types of lesions that might be the classic studies of Powers and Tolmach [30], illustrating
treated with this modality, malignant tumors, benign tumors, the fraction of cells surviving in very small tumors in a
and vascular disorders. mouse irradiated in a single fraction in vivo and subse-
quently assayed by transplantation to other animals. The
cellular survival curve is characterized by two distinct com-
1 ponents; the slopes of the two components differing by a
(The term “stereotactic radiation therapy” will be used here to apply
both to single-fraction treatment (often called radiosurgery) and to factor of 2–3. Up to doses of several Gy, the response is
multi-fraction stereotactic radiotherapy (often called stereotactic body dominated by the killing of aerobic cells, while, for higher
radiation therapy when applied to extracranial sites). There is still doses, the killing of hypoxic cells dominates. It is apparent
debate about the most appropriate terminology [2–4]).
that irradiating such tumors with a single large dose of sev-
D.J. Brenner, Ph.D., D.Sc. (*) eral tens of Gy is a futile exercise if the goal is sterilization,
Center for Radiological Research, Columbia University
since the hypoxic cells will not be adequately depopulated
Medical Center, New York, NY, USA
e-mail: djb3@columbia.edu with a dose of this magnitude. Figure 5.3 gives rough esti-
mates [31], based on in vitro data, of the single fraction dose
D.J. Carlson, Ph.D., DABR
Department of Therapeutic Radiology, Yale University School to sterilize a 30 mm diameter tumor, with and without a
of Medicine, New Haven, CT, USA hypoxic component.
58 D.J. Brenner and D.J. Carlson
100
90
80
70
60
Dose (Gy)
50
40
30
20
10
Fig. 5.1 Hypoxic regions (black) indicated by pimonidazole staining are 0

present even in small tumors. Imaged here is a section of a small tumor
sarcoma
SCLC
melanoma
NSCLC
colo-rectal
medullo
breast
prostate
brain
renal
ovarian
head & neck

derived from a human glioma xenograft line. Adapted from Rijken PF,
Peters JP, Van der Kogel AJ. Quantitative analysis of varying profiles of
hypoxia in relation to functional vessels in different human glioma xeno-
graft lines. Radiat Res. 2002 Jun;157(6):626–32; used with permission
Fig. 5.3 Rough estimate, based on in vitro data, of the single-fraction

100 dose required to sterilize various 30 mm diameter tumors, which are
either fully oxygenated (lower) or contain 20 % hypoxic cells (upper).
response dominated
Adapted from Leith JT, Cook S, Chougule P, Calabresi P, Wahlberg L,
10 –1 by aerated cells
Lindquist C, et al. Intrinsic and extrinsic characteristics of human
tumors relevant to radiosurgery: comparative cellular radiosensitivity
and hypoxic percentages. Acta Neurochir Suppl. 1994;62:18–27; used
10 –2 response dominated with permission from Springer Science + Business Media
by 1–2% hypoxic cells
Surviving Fraction
10 –3
ciple is illustrated in Fig. 5.4 [32]. In a fractionated regime,
therefore, each X- or γ-ray dose predominantly kills aerobic
10 –4
cells, and the interval between treatments allows hypoxic
cells to reestablish their oxygenated state, thereby reducing
10 –5
the problem of hypoxic radioresistance. An analysis [33] of
survival data for cervix cancer patients undergoing fraction-
10 –6 ated external-beam radiotherapy and high dose-rate brachy-
therapy suggests the clinically relevant oxygen enhancement
10 –7 ratio (OER) for hypoxic tumor subvolumes may be closer to
0 5 10 15 20 25 1.5, substantially lower than the typically reported value of
Dose (Gy) ~3 for in vitro clonogenic survival under anoxic conditions.
An OER of 1.5 agrees with values reported in vitro for
Fig. 5.2 Fraction of surviving cells, after a single radiation dose, in oxygen concentrations of ~5 mmHg (0.66 %). However,
small mouse lymphosarcomas irradiated in vivo. The shallow initial
clinically derived OER values are averaged over cells under a
part of the curve is due to the presence of aerated oxic cells, whereas the
steeper slope at higher doses is caused by the presence of 1–2 % of range of oxygen concentrations, so cells irradiated in vivo
hypoxic cells. Reprinted from Hall EJ, Brenner DJ. The radiobiology under extreme levels of hypoxia, e.g., <0.5 mmHg, may still
of radiosurgery: rationale for different treatment regimes for AVMs have OER values higher than 1.5.
and malignancies. Int J Radiat Oncol Biol Phys 1993; 25:381–385.
While stereotactic techniques offer valuable physical advan-
Copyright 1993, with permission from Elsevier
tages over conventional radiotherapy, the potential for reoxy-
genation is reduced as the total dose is delivered in a fewer
However, malignant tumors exhibit a characteristic known number of fractions. Hypofractionation may result in decreased
as reoxygenation whereby, between fractionated doses of biological effectiveness for hypoxic tumors [34]. The clinical
x- or γ-rays, tumors tend to reestablish their original pattern significance of tumor hypoxia and reoxygenation within the
and proportion of oxygenated and hypoxic cells. This prin- context of SRT is still a subject of rigorous debate [35].
5 Radiobiological Principles Underlying Stereotactic Radiation Therapy 59
100
0.36
rad/min
Fraction of Initial Colony-Forming

10 –1
Units (3.5 cells/col) Surviving

0.86
rad/min
10 –2
16
rad/min
10 –3
30
10 –4 rad/min
107
rad/min
10 –5
500 1000 1500 2000 2500
Dose (rad)
Fig. 5.5 Illustrating that, as a given single dose is protracted, either by low-
ering the dose rate (as here) or by increasing the number of fractions, the
Fig. 5.4 Schematic illustration of the process of reoxygenation: almost all
biological effect (in this case, cell killing of Chinese hamster cells)
tumors initially contain a mixture of aerated and hypoxic cells. A single
decreases, due to intra-fraction double-strand break (DSB) repair. Based on
radiation dose will kill a larger fraction of aerated oxic cells because they
data from Bedford JS, Mitchell JB. Dose-rate effects in synchronous mam-
are more radiosensitive (see Fig. 5.2). Thus after irradiation, a larger
malian cells in culture. Radiat Res. 1973 May;54(2):316–27. Reprinted
fraction of cells are hypoxic. But after a short time, the tumor tends to
from Hall EJ, Brenner DJ. The dose-rate effect revisited: radiobiological
return to its original proportion of oxic and hypoxic cells, allowing many
considerations of importance in radiotherapy. Int J Radiat Oncol Biol Phys
of the previously hypoxic (but now aerated) cells to be killed by a sec-
1991; 21:1403–1414. Copyright 1991; with permission from Elsevier
ond, and subsequent, dose fractions. Reprinted from Hall EJ, Brenner
DJ. The radiobiology of radiosurgery: rationale for different treatment
regimes for AVMs and malignancies. Int J Radiat Oncol Biol Phys 1993;
25:381–385. Copyright 1993, with permission from Elsevier
Repair
The second radiobiological principle is repair of DNA damage.

It has been well established for many decades that protracting
or fractionating an acute exposure reduces the level of cell kill-
ing. An example is shown in Fig. 5.5. If all tissues were equally
affected by changes in protraction or fractionation, then there
would be no radiotherapeutic significance to fractionation,
beyond the need to increase the dose to compensate for the
increased cellular repair of double-strand breaks (DSBs).
However, there is a wealth of experimental evidence
which indicates that there is a difference in shape between Fig. 5.6 The gamma ray dose–response curve for late responding tissues
the dose–response relationship characteristic of early (e.g., AVM obliteration, cerebral necrosis) has more curvature, i.e., a
small α/β ratio; for early responding endpoints such as tumor control, the
responding tissues and tumors and late responding tissues. dose–response curve is straighter, i.e., the α/β ratio is larger. Consequently
The inference from experiments in animals, which is con- dose fractionation spares late responding tissues more than early respond-
firmed in clinical practice, is that the dose–response relation- ing tissues. Reprinted from Hall EJ, Brenner DJ. The radiobiology of
ship for late responding tissues has more curvature than that radiosurgery: rationale for different treatment regimes for AVMs and
malignancies. Int J Radiat Oncol Biol Phys 1993; 25:381–385. Copyright
for early responding tissues, as shown in Fig. 5.6. 1993, with permission from Elsevier
In mathematical terms, if the dose (D)–survival (S) rela-
tionship is expressed in terms of a linear–quadratic relation,
then the ratio α/β (the dose at which the cell killing by the
S = exp ( -a D - b D 2 ) , (5.1) linear and quadratic terms is equal) tends to be small for late
responding tissues (less than ~3 Gy) and larger (greater than repopulation—if research in predictive assays has taught us
~8 Gy) for early responding tissues [32, 36–40]. anything, it is that inter-tumor variation can be large; thus, in
The practical consequence of the difference in shape this context, some tumors will grow slowly and some will
between the dose–response curves for early and late respond- grow very quickly. For example, it has been suggested that
ing tissues is a marked difference in the response to fraction- the onset to repopulation may be as long as 60 days for
ation of these two types of tissues. Late-responding tissues slowly proliferating prostate tumors [44].
are more sensitive to changes in fractionation than early
responding tissues. The overall outcome is illustrated in
Fig. 5.6. A fractionated regime spares late responding nor- How Do These Radiotherapeutic Principles
mal tissues more than a single acute dose, for a given level of Apply to Stereotactic Radiation Therapy?
tumor damage.
It is pertinent to ask what exactly constitute “early” and In general these three phenomena control the dose delivery
“late” responding tissues. Probably early responding tissues for cancer radiotherapy as follows:
contain a large proportion of cycling cells, while late 1. We must fractionate treatment:
responding tissues contain large proportions of non-cycling • To overcome hypoxia
cells. • For differential response with late effects
2. We would like to prolong treatment:
• To limit early sequelae
Accelerated Repopulation 3. We would like to shorten treatment:
• To prevent accelerated repopulation
As a tumor shrinks during radiotherapy, the surviving clono- Prima facie, requirements 2 and 3 are mutually exclusive
gens are stimulated to proliferate at an accelerated rate [41, and so, in most radiotherapeutic situations, the overall treat-
42]. For T3 laryngeal cancer, this accelerated repopulation ment time represents a compromise between short treatment
does not commence for about 30 days after the beginning of times to minimize tumor repopulation and long treatment
times to prevent unacceptable early complications. In those
protocols that have attempted to utilize shorter overall treat-
ment times, it has been necessary to reduce the overall dose
in order to avoid excessive early complications [45].
By contrast, stereotactic radiotherapy represents a situa-
tion where the contradictory aspects of overall treatment
time can be resolved without the need for compromise.
Because of the excellent dose distributions which can be
obtained using stereotactic radiotherapy, the need for long
overall treatment times to reduce early normal tissue compli-
cations will rarely apply. Specifically, early skin or mucosal
reactions, which are often a limiting factor in radiotherapy,
will not be an issue for intra-cranial stereotactic radiother-
apy. Thus short overall treatment times become a practical
possibility, with their attendant advantages in terms of limit-
ing tumor repopulation.
Fig. 5.7 Tumor control probability for T3 laryngeal cancer, as a func- Inherently, then, SRT is advantageous in terms of repopu-
tion of overall time and overall dose. The data points, at the top of the lation. We discuss here how the remaining two Rs of radio-
dashed vertical lines, are those reported by Slevin et al. [43], and the
surface is the result of a linear–quadratic (with repopulation) fit [42] to therapy (reoxygenation and repair) affect stereotactic
the data, with repopulation effectively starting about 33 days after the radiotherapy for the three main target lesions, malignancies,
beginning of the treatment. Reprinted from Brenner DJ. Accelerated AVMs, and benign tumors.
repopulation during radiotherapy: quantitative evidence for delayed
onset. Radiat Oncol Invest 1993; 1:167–172. Used with permission
Malignancies
treatment as illustrated in Fig 5.7 [43]. Radiation treatments
of duration longer than about 30 days would therefore need As we have discussed, it is very unlikely that a single dose
to increase the total dose to compensate for this clonogen fraction could sterilize even a small tumor which contains
proliferation. Note that it is unrealistic to categorize almost hypoxic malignant cells. Thus there is a clear rationale,
any tumor type as one which is not significantly prone to based on the issue of hypoxia, for stating unequivocally that
single-fraction radiotherapy, radiosurgery, is a suboptimal

modality when the target is a malignant tumor [46].
In addition, because of the differential α/β ratio of the
tumor and the surrounding late responding normal tissue
sequelae (such as delayed cerebral necrosis, or optic neu-
ropathy), even in the rare situation when there is minimal
hypoxia, single-fraction treatment of a malignancy would be
expected to give a suboptimal therapeutic ratio between
tumor control and late complications.
Thus an optimal stereotactic radiotherapy protocol for a
large intra-cranial malignancy would involve large numbers of
fractions over a short time period, for example, 20 fractions in
2 weeks might be an appropriate target. For smaller tumors,
where the relative dose to normal tissue surrounding the malig-
nancy is smaller—though still, as always, dose limiting—a
regime consisting of five or ten fractions over 1 week would
Fig. 5.8 Linear–quadratic fit to combined long-term AVM obliteration
combine the clinical advantages of a very short overall time,
data from Paris [57] and Pittsburgh [58]. The estimated α/β ratio from
the radiobiological advantages of reoxygenation and (partial) this analysis is 5.2 Gy, consistent with a recent estimate of 3.5 Gy by
repair of late responding normal tissue damage, and the practi- Kocher et al. [59], based on most currently available data, which is in
cal advantages of a small number of treatments. turn consistent with a typical α/β value for a late responding tissue
It has been argued [47] that, even for malignancies, “when
the treatment volume is small and contains little functioning
brain tissue, the need for fractionation may not apply.”
However, the dose-limiting factor of any radiotherapeutic The issue here is that the dose-limiting side effect of the
treatment of a malignancy must always be the normal tissue treatment, generally delayed cerebral radionecrosis, is, of
tolerance, so fractionation should always allow a greater course, also a late response [52–56]. Thus, in the treatment
tumoricidal dose. of AVMs, the tissue which it is desired to damage and the
Of course one of the potential advantages of radiosurgery is surrounding normal tissue that it is desired to spare are both
that it may be able to overcome the radioresistance of tumors of the same radiobiological type, i.e., they are both late
such as metastases from melanomas or renal-cell sarcomas. responding tissues. Consequently one would expect that
However, by keeping the number of fractions low (perhaps to nothing is to be gained by a fractionated course relative to a
five or six), the radioresistance of such tumors will still be single dose. In other words a change in fractionation pattern
adequately addressed while maintaining the other biological will not preferentially produce more damage in the AVM
advantages of fractionation that have been outlined here. than in the surrounding normal tissues. The ratio of damage
The arguments presented here do not exclude the fact that between the AVM and the surrounding normal tissue would
good results may be obtained using radiosurgery for brain be the same whether the dose is delivered in a single or in
malignancies. Indeed there are reports of good results with multiple exposures. Given that the issue of hypoxic cells is
this technique. Rather, for any given situation, better not relevant here, this is a strong rationale for radiosurgery,
results—in terms of therapeutic ratio between tumor control i.e., a single, highly localized radiation treatment.
and complications—would always be expected by fraction- In fact, there have been few estimates of α/β ratios reported
ating than can be obtained with a single fraction—an impor- in the literature corresponding to the end point of AVM oblit-
tant consideration in the treatment of cerebral malignancies. eration, as few dose–response data have been established. An
early estimate [46] of 0.6 Gy (range 0.2–5 Gy) was reported
based on limited dose–response results for complete AVM
Arteriovenous Malformations obliteration after stereotactic irradiation with helium ions.
Using the more recent dose–response data from Touboul
In treating AVMs, the goal is to produce progressive luminal et al. [57] and Flickinger [58], we estimated values of 5.2 Gy
closure through an inflammatory reaction in the vessel walls from long-term obliteration data and 2.4 Gy [range 0.9–8 Gy]
of the malformation, by irradiating the constituent epithelial from short-term obliteration data (see Fig. 5.8). A recent
cells [48–51]. This is a classic “late” response, occurring analysis by Kocher et al. [59] of most available data resulted
weeks to months after the radiation treatment and thus, prima in an overall best estimate of 3.5 Gy and an estimate of 4.6–
facie, one would expect a relevant α/β ratio similar to that for 6.4 Gy for small AVMs. All these values are consistent with
other late responding tissue. those of a classic late response.
In summary, in the treatment of AVMs there is probably specifically attributed to cell killing. By contrast, descriptive
nothing to be gained by a fractionated course relative to a empirical models can go disastrously wrong if used outside
single dose. Thus, the optimal treatment for AVMs is the dose/fractionation range from which they were derived—
expected to be single-fraction stereotactic radiotherapy or as when NSD was applied to large doses per fraction [69, 70].
radiosurgery, i.e., a single, highly localized radiation treat-
ment at, of course, the appropriate dose.
Mechanistic Basis
of the Linear–Quadratic Model
Benign Tumors
The standard model is that both tumor control and complica-
Less is known about the sensitivity to fractionation of tions result primarily from cell killing [38, 71, 72]—and the
benign tumors. It is less likely that hypoxia will be a signifi- LQ is a mechanistic model of cell killing (though see below
cant role here [60]; one estimate of the α/β ratio for benign for discussions of possible alternate mechanisms). In Eq.
parasellar meningiomas is 3.3 Gy [61], similar to that for (5.1), αD can be interpreted as the quantity of lethal DNA
late responding normal tissues. For benign tumors as with lesions formed by a single radiation track, i.e., unrejoinable
AVMs, the most likely scenario is that both the target and DSBs, lethally misrepaired or fixed DSBs, and lethal intra-
the surrounding normal tissue respond in a similar way to track exchange-type chromosome aberrations, while βD2 can
fractionation, like that of a late responding normal tissue, be interpreted as the quantity of lethal DNA lesions formed
suggesting that there is little to be gained from fractionation. by two radiation tracks, i.e., lethal inter-track exchange-type
For example, if the dose-limiting late sequelae were chromosome aberrations formed through binary misrepair of
radiation-induced optic neuropathy, which is caused by two separate DSBs [66, 73]. Underlying the application of
damage to the arteries supplying the optic nerve, this almost the LQ to fractionation and protraction effects is the pairwise
certainly responds to changes in fractionation like a classic misrepair of primary lesions such as DSBs or base damage
late responding tissue [62, 63]. formed by separate radiation tracks (hereon in, we shall refer
In summary, in the treatment of benign tumors there is to the DSB as the primary lesion, but base damage sites may
probably nothing to be gained by a fractionated course, rela- well also be relevant here [74]). As schematized in Fig. 5.9,
tive to a single dose. However, we certainly do not have ade- cell killing can occur via chromosome aberrations such as
quate data for a variety of benign tumors to consider this dicentric aberrations [75], formed when pairs of nearby
conclusion definitive.
Mechanistic Models of Radiotherapeutic

Response
Why are we interested in mechanistic models to describe

radiotherapeutic response? There are really two reasons:
First, we need some model for calculating isoeffect doses
when alternate fractionation schemes are considered. Second,
mechanistic models are a useful tool for understanding
mechanisms—if a model based on given mechanisms is
inconsistent with the data, then we can conclude that the
assumed mechanism is either wrong or not the dominant one.
Fig. 5.9 Examples of binary misrepair: (a) shows two chromosomes;
The tool most commonly used for quantitative predictions each has one double strand break (DSB), shown as a gap. Centromeres,
of dose and fractionation dependencies is the linear–qua- which are needed for proper transmission of chromosomes to daughter
dratic (LQ) formalism [37, 38, 64–66]. In radiotherapeutic cells at mitosis, are shown as black constrictions. Most DSB are cor-
applications, the LQ formalism is now almost universally rectly restituted, but a few undergo binary misrepair. (b) Binary misre-
pair can result in a dicentric chromosome aberration, which generally
used for calculating isoeffect doses for different fraction- destroys the clonogenic viability of the cell. In about half the binary
ation/protraction schemes. misrepair events, the two DSBs shown in (a) lead to a translocation (c).
In contrast to earlier methodologies, such as CRE, NSD, Translocations involve large scale rearrangements and can cause poten-
and TDF [67, 68], which were essentially empirical descrip- tially precarcinogenic alterations in cellular phenotype but most do not
impair cellular survival. Reprinted from Brenner DJ. The linear–qua-
tions of past clinical data, the LQ formalism has become the dratic model is an appropriate methodology for determining isoeffec-
preferred tool largely because it has a somewhat more biologi- tive doses at large doses per fraction. Semin Radiat Oncol. 2008; 18:
cal basis, with tumor control and normal tissue complications 234-239. Used with permission
DSBs wrongly rejoin to one another [76]. Protracting the Y μ a D + Gb D2 . (5.2)

exposure time potentially allows the first DSB to be repaired
before the second is produced, and the LQ approach quanti- (Note that the biologically effective dose, biological effec-
fies this effect [66]. Nowadays, this binary DSB misrepair tive dose (BED), is defined as Y/α [37].) Then, assuming the
model is the most usual way to motivate the standard LQ lethal lesions are Poisson distributed from cell to cell, the
approach, but different biological rationales for the same surviving fraction will be
mathematical formalism have also been given, as we will
discuss. S = exp ( -Y ) , (5.3)
It is important to stress here that the standard LQ formal-
ism, as applied to time–dose relationships, is not merely a and thus
truncated power series in dose. Its key feature here is a spe-
cific mechanistically based functional form for the protrac- S = exp éë - (a D + G b D 2 ) ùû . (5.4)
tion factor, usually designated by G, which takes into account
dose protraction or fractionation. Expressions for special In Eqs. (5.2) and (5.4), G is the generalized Lea–Catcheside
cases of the time factor, G, were derived by Lea and time factor, which accounts quantitatively for fractionation/
Catcheside [65, 77]; a general form was subsequently derived protraction; it is important to note that G acts only on the
[78] and has since been rederived from several different quadratic component, as described in Point 3 above. The
points of view [70]. We refer to this general form of the time generalized time factor has the form [78]
factor, given explicitly in Eq. (5.5), as the generalized Lea–
¥ t
- l (t -t )
G = ( 2 / D 2 ) ò D ( t ) dt ò e D ( t ¢ ) dt ¢ .
Catcheside time factor, G. ¢
(5.5)
-¥ -¥
The LQ Formalism Here D ( t ) describes the variation in dose rate over the entire
course of the radiotherapy, and λ is a characteristic damage
The LQ model, in its most usual current version, describes repair rate. Generically, the term after the second integral sign
cell killing in terms of the following mechanisms: refers to the first of a pair of DSBs required to produce a lethal
1. Radiation produces many types of clustered DNA lesions, lesion—the exponential term describing the reduction in num-
including the DSBs, proportional to absorbed dose up to bers of such DSB through repair; similarly, the term after the
several hundred Gy for low- and high-LET radiation first integral sign refers to the second DSB, which can interact
[79–82]. Experiments such as these imply that single- with DSBs, produced earlier, that still remain after repair.
track mechanisms are responsible for the induction of The time factor, G, can be calculated for any fractionation/
DSBs and all other types of clustered DNA damage. protraction scheme and systematically accounts for the effects
2. These DSBs can be repaired, with first-order rate con- of protracting the dose delivery in any way. G can take values
stant λ(=ln 2/T½, where T½ is the repair halftime). from zero to one, with G = 1 for a single acute dose, leading to
In practice, there may be more than one class of DSB the simplest single-fraction LQ formalism shown in Eq. (5.1).
which may be repaired with different rate constants; the The interpretation of G < 1 is a reduction in cell killing due to
LQ formalism can be simply extended to take this into intra-fraction DNA repair which occurs during protracted
account. radiotherapy. Two special cases, which illustrate the main fea-
3. In competition with DSB repair, binary misrepair of pairs tures of the general expression for G, are: (1) for irradiation
of DSBs produced from different radiation tracks (i.e., with n short fractions, each separated by a time T [83]:
different photons) can produce lethal lesions (often iden-
tified as predominantly dicentric chromosomal aberra- G = éë 2q / (1 - q ) ùû éë n - (1 - q n ) / (1 - q ) ùû , (5.6)
tions), the yield being proportional to the square of the
dose (see the quadratic term in Eqs. 5.2 and 5.4). The two
where θ = exp(−λT). When the time between the short frac-
independent radiation tracks can occur at different times
tions is very long, then
during the overall regimen, allowing repair of the first
DSB to take place before it can undergo pairwise misre- G = 1/ n, (5.7)
pair with the second; it is this phenomenon which is the
heart of the fractionation/protraction dependence in the and the most simple LQ model for well-separated fraction-
LQ formalism. ated exposures becomes
4. In addition, single radiation tracks can produce lethal
lesions, possibly by a variety of mechanisms, the yield S = exp éë - D (a + b D / n ) ùû . (5.8)
being linearly proportional to the dose.
Overall, in the LQ formalism, the yield (Y) of lethal Different fractionation/protraction schemes have different
lesions and the corresponding survival (S) equation are time factors, G, any of which can be calculated [64, 84] from
Eq. (5.5). Even for high dose rates, dose protraction effects In Vitro
may have a significant impact on cell killing and, ultimately,
treatment effectiveness [85, 86]. It is not easy to generate precise, accurate measurements of
The LQ model can further be used to relate cell death to cell survival at high radiation doses. One recent approach to
clinical endpoints of interest. The tumor control probability this problem is to use DNA flow cytometry for counting cell
(TCP) model quantifies the response of a group of patients to numbers, rather than colonies. This approach can produce
a specific radiotherapy treatment and is essential in the anal- very high statistics data and establish precise survival curves.
ysis of clinical data for local control. The TCP model relates Figure 5.10 shows such survival data for irradiated CHO
the tumor volume (or number of tumor cells) and the radia- cells [94], where it can be seen that the standard errors can be
tion dose to the probability of tumor cure, i.e., the probability made very small with this approach. The single-fraction
that no tumor cells survive the radiation treatment. The most LQ-model fit (Eq. 5.1) to these data clearly fits very well,
commonly used TCP algorithm [87, 88] is based on Poisson indicating that it can predict the pattern of dose–responses
statistics: over the key 1.8–7 Gy dose range.
Similar results have recently been reported by Garcia
TCP = exp éë - N × S ( D ) × eg (T -Tk ) ùû , (5.9) et al. [95], using a more standard colony assay to measure
cell survival. In this case, surviving fractions in the dose
where N is the initial number of tumor clonogens, S(D) is the range from zero to 16 Gy were measured. In order to esti-
surviving fraction of cells predicted by the LQ model, T is mate dose regions where the single-fraction LQ model
the overall treatment time, and Tk is the lag time before (Eq. 5.1) did/did not fit the data, the authors fitted the data
accelerated tumor repopulation begins. The repopulation rate from 0 to 4 Gy, and then from 0 Gy to progressively larger
of tumor cells, γ, is conveniently expressed as γ = ln(2)/Td, doses, each time assessing the goodness of fit of the model to
where Td is the effective tumor doubling time. The TCP for- the data. Typical results are shown in Fig. 5.11, where it can
malism can also be modified to include interpatient variabil- be seen that the quality of fit to the LQ model does not
ity in radiosensitivity [89]. decline significantly until doses above 15 Gy are included.
In Vivo
How Well Does the LQ Model Work,
At Low and High Doses? There are a fairly wide range of quantitative in vivo end-
points for which it is possible to test concordance with the
The goal here is to be able to make equi-effect regimen LQ predictions in the 1.8–20 Gy range [96]. Figure 5.12, for
extrapolations from standard regimes, typically with
1.8–2 Gy per fraction, to hypofractionated regimes. Possible
hypofractionated doses per fraction are, for example, up to
about 7 Gy for prostate [90], and up to about 15–20 Gy for
non-small cell lung cancer (NSCLC) [91]. For fractionated
stereotactic radiotherapy, we are interested in extrapolating
in the opposite direction, from single-fraction doses as high
as 18 Gy to hypofractionated exposures with lower doses per
fraction [92].
So the general question is whether the LQ formalism
described in the previous section describes radiotherapeuti-
cally relevant dose–responses in the dose-per-fraction range
from, say, 1.8 to 20 Gy. For many situations, such as prostate
cancer hypofractionation, the dose-per-fraction range of
interest is probably about 1.8–7 Gy.
We have previously argued that the LQ model does pro-
vide an appropriate methodology for determining isoeffective
doses at large doses per fraction [93]. Below, we discuss Fig. 5.10 Survival of x-irradiated CHO cells, determined by flow
some in vitro and in vivo experimental data to address the cytometry population counting, 5 days after treatment [94]. The curve
is the corresponding LQ model fit. Reprinted from Brenner DJ. The
question, and then discuss some alternate mechanistic mod-
linear–quadratic model is an appropriate methodology for determining
els to the LQ equations, and their potential impact, relative to isoeffective doses at large doses per fraction. Semin Radiat Oncol.
LQ-based predictions. 2008; 18: 234-239. Used with permission
Fig. 5.11 Goodness of fit of LQ model to measured cell-survival data,

as a function of the dose range which was fitted [95]. The quantity plot-
ted is χ2 per degree of freedom; hence, smaller values represent better
fits to the LQ model. For example the left-most point represents a good
fit of the LQ model to cell-survival data in the dose range from 0 to
4 Gy and the right-most point represents a less good fit of the LQ model
to cell-survival data in the dose range from 0 to 16 Gy. Reprinted from
Fig. 5.12 Isoeffect data for late response from three (open square,
Brenner DJ. The linear–quadratic model is an appropriate methodology
open circle, open triangle) different regions of the rat spinal cord [97],
for determining isoeffective doses at large doses per fraction. Semin
for acute skin reactions (filled diamond) in mice [98], and for early
Radiat Oncol. 2008; 18: 234-239. Used with permission
(filled circle) and late (⊕) murine intestinal damage [99]. The data are
plotted in a “reciprocal-dose Fe” form [98] such that, if they follow a
linear–quadratic relationship, the points fall on a straight line. Reprinted
example, shows some isoeffect results from Van der Kogel from Brenner DJ. The linear–quadratic model is an appropriate meth-
odology for determining isoeffective doses at large doses per fraction.
[97] for late responding damage to the rat spinal cord, from Semin Radiat Oncol. 2008; 18: 234-239. Used with permission
Douglas and Fowler [98] for acute damage in mouse skin
and from Peck and Gibbs [99] for early and late damage to
the murine small intestine. The form of the plot, the so-called
reciprocal-dose Fe plot [96, 98], is such that, if the LQ for- Is the LQ Model Correct?
malism applies, the data would fall on a straight line.
Although the reciprocal-dose plot approach is not an optimal We have shown that the LQ model is reasonably predictive of
methodology for parameter estimation [100, 101], it does dose–response relations, both in vitro and in vivo, in the dose
provide a visual indication of how well in vivo data agrees per fraction range of 1.8–15 Gy. Of course it goes without
with the LQ model, in the dose range of interest. All the saying that no mechanistic model describing dose–time pat-
quantitative in vivo endpoints in Fig. 5.12 are consistent with terns can be fully complete or correct, so we discuss here
the LQ model, over a wide range of doses per fraction, some of the main perceived mechanistic uncertainties associ-
including those of interest in hypofractionation. Extensive Fe ated with the LQ model.
plot analyses by Barendsen [96], using 12 normal tissue
response endpoints, reached the same conclusion. While
more sophisticated methods are available for assessing Pairwise Production of Chromosome
agreement with the LQ model [102], given the inherent Aberrations Is Not the Only Mechanism
uncertainties in the data, it is clear that all these data, includ- Mediating Cell Killing
ing those for single fractions of ~20 Gy, are consistent with
the LQ model. Again it is argued that cell killing is the dominant process
Clinical outcome data for local tumor control can also be mediating radiotherapeutic response [38, 71, 72], both for
used to test the validity of the LQ model over different dose early and late effects, including vascular effects. But not
ranges and fractionations. The application of TCP models to all cell killing is mediated through chromosome aberra-
local control data for early-stage NSCLC and the validity of tions produced by pairwise misrepair. However, other
this LQ-based approach at high stereotactic doses are further cell killing mechanisms—such as apoptosis and induction
discussed below. of small mutations—are dose-rate independent [103].
The LQ incorporates both dose-rate-independent mecha- In fact there is no, as yet, strong evidence that saturable
nisms such as these (in the linear term) and dose-rate-depen- repair mechanisms are important at the doses and dose rates
dent mechanisms (in the quadratic term). Assuming LQ of relevance to radiotherapy [116]. However, since SR mod-
correctly models the mechanisms involved in the dose-rate- els appear, prima facie, to be mechanistically different from
dependent term, then with appropriate parameters (α/β actu- binary misrepair models, there is the possibility that they
ally is the ratio of the dose-rate-independent term to the could make significantly different predictions of fraction-
dose-rate-dependent term), LQ should be adequate predict- ation/protraction effects, casting doubt on the validity of
ing fractionation/protraction effects. LQ-based isoeffect dose calculations. In fact, the formalisms
describing most SR models also lead, in an appropriate
approximation, to the same time–dose relations as does the
There Are Other Mechanistic LQ formalism [104]. This comes about because these SR
Models Describing Pairwise Production formalisms reduce to the specific form of the generalized
of Chromosome Aberrations, as Well as LQ Lea–Catcheside time factor, G, which describes protraction
effects in the LQ approach.
It has been known for some time that, in addition to the LQ
model, various other binary misrepair models lead to the
same generalized Lea–Catcheside time factor, G, in an appro- Different Mechanisms at High Single Doses?
priate approximation, and thus predict virtually the same
time–dose relations as does the LQ approach. This result was One additional aspect of the use of large dose fractions is the
demonstrated by several authors [73, 83, 104–107] for the possibility that tumor cell death might be enhanced by addi-
repair–misrepair (RMR) model [108], lethal–potentially tional biological processes resulting in tumor cell killing
lethal (LPL) model [105], the repair–misrepair–fixation [117], such as rapid endothelial cell apoptosis in tumor ves-
(RMF) model [73], and more general binary misrepair sels or systemic abscopal effects. Garcia-Barros et al. [118]
models. In light of the conceptual similarities between the have reported that vascular endothelial cell apoptosis occurs
LQ model and other binary misrepair models, the fact that at doses greater than ~8 to 10 Gy. Recently published clinical
the corresponding formalisms make virtually equivalent data showing a high rate of short-term local control for
predictions for dose–time relationships is not, in retrospect, patients with metastatic spinal cord tumors treated with large
particularly surprising. single doses have been attributed to this phenomenon [119].
Provocative data on the stimulation of antitumor immunity
for melanoma treated with stereotactic body radiation ther-
Lethal Lesions Are Not Poisson Distributed apy (SBRT) have also been reported [120, 121]. There
from Cell to Cell remains a need, however, for more preclinical and clinical
studies before this novel “high dose” biology should be
This Poisson assumption, embodied in Eq. (5.3), is used in accepted as a new mechanism for the effect of high dose
all models which calculate the fraction of inactivated cells fractions [122]. In addition, evidence presented below still
from the average yield of lethal lesions. It is certainly the suggests that the LQ model provides a good description of
case that, based on microdosimetric considerations, this dis- local control data at high doses per fractions despite these
tribution is not exactly Poisson [109]. For densely ionizing possible additional biological mechanisms that are not
radiations such as heavy ions or neutrons, this point is well captured in the model.
taken and important. But for sparsely ionizing radiation such
as X- or γ-rays, below ~50 Gy per fraction, deviations from
a Poisson distribution are quite minor [66, 110]. The Success of Stereotactic Radiation
Therapy: Better Dose Distributions,
Not Better Biology
Repair Mechanisms Saturate at High Doses
While other radiobiological mechanisms at high doses per
Various high dose saturable repair (SR) models have been con- fraction have been shown in the laboratory, there is to date no
sidered [111–115], all having in common the notion that per- clinical evidence of their relevance to SRT response, and the
lesion repair rate decreases as the dose—and the production of LQ model appears to be appropriate for use at all the doses
initial damage—increases. This might arise, for example, if and fractionation schemes of relevance to SRT. Rather than
repair enzymes can be overloaded. Such a change in repair new biological mechanisms, recent analyses of clinical data
rates with increasing dose might in principle result in different suggest that the great success of SRT is primarily related to
dose–response relations at high doses, compared with the LQ its improved conformal dose delivery that permits significant
approach where repair rates are dose independent. escalation of the biologically effective doses [123, 124].
Prima facie, single-fraction radiotherapy appears to rep-

resent a major paradigm shift from the practice of radiother-
apy that has developed over the past 90 years. Specifically,
essentially all non-stereotactic treatment plans to date involve
significant volumes of normal tissue receiving doses close to
the prescribed tumor dose. It follows that the goal in such a
scenario is to maximize tumor response for an acceptable
level of normal tissue damage probability—and the primary
tool that radiotherapists have used to maximize this thera-
peutic ratio has been fractionation. The benefit of dose frac-
tionation is supported by an abundance of animal and human
data generated over many decades.
SRT can ignore this classic fractionation paradigm, yet
still be remarkably successful, because of technical advances
brought about by image-guided robotic technology or multi-
leaf collimated beams: these techniques can deliver large
tumor doses with very small margins, thereby minimizing Fig. 5.13 Tumor control probability (TCP) as a function of biological
effective dose (BED) for stage I NSCLC. Symbols show crude local
large doses to large volumes of normal tissue. For many control rates (≥2 years) from a pooled analysis reported by Mehta et al.
tumor sites, such as tumors in the periphery of the lung, this [125] with different symbols distinguishing 3D-CRT and SBRT regi-
has reduced the concern about unacceptable normal tissue mens. Solid line shows an LQ-based fit to the available data
injury and allowed considerable dose escalation.
It is hardly surprising that additional biological processes if there were none at these very high BEDs. However, these
(described above) would come into play when tumor doses potential new mechanisms seem unlikely to be contributing
are increased so markedly. The important question is whether to clinical response more than the standard biological mech-
any or all of these processes, colloquially termed a “new anisms that dominate in conventional radiotherapy.
biology” [122], need to be invoked to explain the high local
control rates obtained with SBRT. Since most successful
SBRT regimens deliver very high BEDs to the tumor, is the Conclusions
impressive efficacy of SBRT the result of “new biology” or
simply the high BEDs delivered to the tumor? For malignant tumors:
Mehta and colleagues [125] recently reviewed the avail- • Radiobiological and clinical arguments, based on the
able clinical TCP data for SBRT and 3D conformal radio- need to overcome hypoxia, and the potential for a differ-
therapy (3D-CRT) for early-stage NSCLC. Figure 5.13 is a ential response between the tumor and the surrounding
replot of the NSCLC TCP data as a function of BED. It is normal tissue, strongly suggest that fractionation will
clear there is a monotonic relationship between TCP and result in improved outcome.
BED for SBRT—as the delivered BED to the tumor increases, • For most malignancies, even when good dose distribu-
so does the probability of local control. This conclusion is tions can be obtained, single-fraction radiation therapy
consistent with the results from 3D-CRT studies of NSCLC, has not been in general use for more than 70 years.
which consistently use lower BEDs and consistently demon- • For most sites, only a few fractions (5–10) would be
strate lower TCPs—though the different dose distributions optimal for treating malignant tumors.
involved here make comparison harder to interpret. • Modern technology has made fractionated stereotactic
Based on these two observations that (1) TCP increases radiotherapy practical and often logistically advantageous.
monotonically with BED and (2) the TCP vs. BED relation For AVMs:
is the same whether single or multiple fractions are used, we • The arguments in favor of fractionation probably do
can say with some confidence that the great success of SBRT not hold, and single-fraction treatment is generally
is due to the fact that the new stereotactic radiotherapy tech- optimal.
nologies provide dose distributions that permit the clinician • There is no hypoxic cell issue to overcome, and both the
to prescribe BEDs of 100 or more Gy [126]—tumor BEDs AVM and the surrounding normal tissue probably respond
that are simply unreachable with conventional dose distribu- the same way to changes in fractionation.
tions. A “new biology” is therefore not necessary to explain For benign tumors:
the high cure rates for NSCLC. They are simply a result of • There is some evidence that benign tumors respond to
the much higher delivered tumor doses. This does not mean fractionation like a late responding normal tissue, in
that different radiobiological mechanisms do not exist at the which case fractionation would not improve outcome,
high doses associated with SBRT, and it would be surprising unless hypoxic cells were an issue.
• Overall, the radiobiological evidence is probably insuffi- 16. Simonova G, Novotny J, Novotny Jr J, Vladyka V, Liscak
R. Fractionated stereotactic radiotherapy with the Leksell Gamma
cient to make recommendations on appropriate fraction-
Knife: feasibility study. Radiother Oncol. 1995;37(2):108–16.
ation schemes for benign tumors. Epub 1995/11/01.
For calculations of biologically effective dose with the lin- 17. Regine WF, Patchell RA, Strottmann JM, Meigooni A, Sanders
ear–quadratic model: M, Young B. Combined stereotactic split-course fractionated
gamma knife radiosurgery and conventional radiation therapy for
• The mechanistically based LQ approach appears to be
unfavorable gliomas: a phase I study. J Neurosurg. 2000;93 Suppl
appropriate for use at all clinically relevant doses and 3:37–41. Epub 2001/01/06.
fractionation schemes. 18. Higuchi Y, Serizawa T, Nagano O, Matsuda S, Ono J, Sato M,
• While novel radiobiological mechanisms at high doses et al. Three-staged stereotactic radiotherapy without whole brain
irradiation for large metastatic brain tumors. Int J Radiat Oncol
have been shown in the laboratory, there is to date no evi-
Biol Phys. 2009;74(5):1543–8. Epub 2009/01/13.
dence from the clinic of their relevance. 19. Ernst-Stecken A, Ganslandt O, Lambrecht U, Sauer R,
Grabenbauer G. Phase II trial of hypofractionated stereotactic
radiotherapy for brain metastases: results and toxicity. Radiother
Oncol. 2006;81(1):18–24. Epub 2006/09/19.
References 20. Narayana A, Chang J, Yenice K, Chan K, Lymberis S, Brennan C,
et al. Hypofractionated stereotactic radiotherapy using intensity-
1. Leksell L. Cerebral radiosurgery. I. Gammathalanotomy in two modulated radiotherapy in patients with one or two brain metas-
cases of intractable pain. Acta Chir Scand. 1968;134(8):585–95. tases. Stereotact Funct Neurosurg. 2007;85(2–3):82–7. Epub
2. Lunsford LD, Flickinger JC, Larson D. Regarding: Rosenthal DI, 2006/12/15.
Glatstein E. “We’ve got a treatment, but what’s the disease?” 21. Mori Y, Hashizume C, Kobayashi T, Shibamoto Y, Kosaki K,
The Oncologist 1996;1. Oncologist. 1997;2(1):59–61. Nagai A. Stereotactic radiotherapy using Novalis for skull base
3. Adler Jr JR, Colombo F, Heilbrun MP, Winston K. Toward an metastases developing with cranial nerve symptoms. J Neurooncol.
expanded view of radiosurgery. Neurosurgery. 2004;55(6):1374–6. 2010;98(2):213–9. Epub 2010/04/21.
4. Loo BW, Chang JY, Dawson LA, Kavanagh BD, Koong AC, 22. Matsuyama T, Kogo K, Oya N. Clinical outcomes of biological
Senan S, et al. Stereotactic ablative radiotherapy: what’s in a effective dose-based fractionated stereotactic radiation therapy for
name? Pract Radiat Oncol. 2011;1(1):38–9. metastatic brain tumors from non-small cell lung cancer. Int J
5. Colombo F, Benedetti A, Pozza F, Avanzo RC, Marchetti C, Radiat Oncol Biol Phys. 2013;85:984–90. Epub 2012/10/25.
Chierego G, et al. External stereotactic irradiation by linear accel- 23. Hattangadi JA, Chapman PH, Bussiere MR, Niemierko A, Ogilvy
erator. Neurosurgery. 1985;16(2):154–60. CS, Rowell A, et al. Planned two-fraction proton beam stereotac-
6. Houdek PV, Fayos JV, Van Buren JM, Ginsberg MS. Stereotaxic tic radiosurgery for high-risk inoperable cerebral arteriovenous
radiotherapy technique for small intracranial lesions. Med Phys. malformations. Int J Radiat Oncol Biol Phys. 2012;83(2):533–41.
1985;12(4):469–72. Epub 2011/11/22.
7. Chen JC, Rahimian J, Girvigian MR, Miller MJ. Contemporary 24. Gill SS, Thomas DG, Warrington AP, Brada M. Relocatable frame
methods of radiosurgery treatment with the Novalis linear accel- for stereotactic external beam radiotherapy. Int J Radiat Oncol
erator system. Neurosurg Focus. 2007;23(6):E4. Epub 2007/12/18. Biol Phys. 1991;20(3):599–603.
8. Dhabaan A, Elder E, Schreibmann E, Crocker I, Curran WJ, Oyesiku 25. Kooy HM, Dunbar SF, Tarbell NJ, Mannarino E, Ferarro N,
NM, et al. Dosimetric performance of the new high-definition mul- Shusterman S, et al. Adaptation and verification of the relocatable
tileaf collimator for intracranial stereotactic radiosurgery. J Appl Gill-Thomas-Cosman frame in stereotactic radiotherapy. Int J
Clin Med Phys. 2010;11(3):3040. Epub 2010/08/19. Radiat Oncol Biol Phys. 1994;30(3):685–91.
9. Roa DE, Schiffner DC, Zhang J, Dietrich SN, Kuo JV, Wong J, 26. Kamath R, Ryken TC, Meeks SL, Pennington EC, Ritchie J,
et al. The use of RapidArc volumetric-modulated arc therapy to Buatti JM. Initial clinical experience with frameless radiosurgery
deliver stereotactic radiosurgery and stereotactic body radiother- for patients with intracranial metastases. Int J Radiat Oncol Biol
apy to intracranial and extracranial targets. Med Dosim. Phys. 2005;61(5):1467–72.
2012;37(3):257–64. Epub 2012/03/01. 27. Verellen D, Soete G, Linthout N, Van Acker S, De Roover P, Vinh-
10. Adler Jr JR, Chang SD, Murphy MJ, Doty J, Geis P, Hancock Hung V, et al. Quality assurance of a system for improved target
SL. The Cyberknife: a frameless robotic system for radiosurgery. localization and patient set-up that combines real-time infrared
Stereotact Funct Neurosurg. 1997;69(1–4 Pt 2):124–8. tracking and stereoscopic X-ray imaging. Radiother Oncol.
11. Chen CC, Chapman P, Petit J, Loeffler J. Proton radiosurgery in 2003;67(1):129–41. Epub 2003/05/22.
neurosurgery. Neurosurg Focus. 2007;23(6):E5. Epub 2007/12/18. 28. Overgaard J, Horsman MR. Modification of hypoxia-induced
12. Halasz LM, Bussiere MR, Dennis ER, Niemierko A, Chapman radioresistance in tumors by the use of oxygen and sensitizers.
PH, Loeffler JS, et al. Proton stereotactic radiosurgery for the Semin Radiat Oncol. 1996;6(1):10–21.
treatment of benign meningiomas. Int J Radiat Oncol Biol Phys. 29. Rijken PF, Peters JP, Van der Kogel AJ. Quantitative analysis of
2011;81(5):1428–35. Epub 2010/10/12. varying profiles of hypoxia in relation to functional vessels in dif-
13. Zimmermann F, Wulf J, Lax I, Nagata Y, Timmerman RD, ferent human glioma xenograft lines. Radiat Res. 2002;157(6):
Stojkovski I, et al. Stereotactic body radiation therapy for early 626–32.
non-small cell lung cancer. Front Radiat Ther Oncol. 2010;42: 30. Powers WE, Tolmach LJ. A multicomponent x-ray survival curve
94–114. Epub 2009/12/04. for mouse lymphosarcoma cells irradiated in vivo. Nature.
14. Schwade JG, Houdek PV, Landy HJ, Bujnoski JL, Lewin AA, 1963;197:710–1.
Abitol AA, et al. Small-field stereotactic external-beam radiation 31. Leith JT, Cook S, Chougule P, Calabresi P, Wahlberg L, Lindquist
therapy of intracranial lesions: fractionated treatment with a fixed- C, et al. Intrinsic and extrinsic characteristics of human tumors
halo immobilization device. Radiology. 1990;176(2):563–5. relevant to radiosurgery: comparative cellular radiosensitivity and
15. Souhami L, Olivier A, Podgorsak EB, Villemure JG, Pla M, hypoxic percentages. Acta Neurochir Suppl. 1994;62:18–27.
Sadikot AF. Fractionated stereotactic radiation therapy for intra- 32. Hall EJ. Radiobiology for the radiologist. 5th ed. Philadelphia:
cranial tumors. Cancer. 1991;68(10):2101–8. Lippincott, Williams & Wilkins; 2000.
33. Carlson DJ, Stewart RD, Semenenko VA. Effects of oxygen on 55. Sheline GE, Wara WM, Smith V. Therapeutic irradiation and brain
intrinsic radiation sensitivity: a test of the relationship between injury. Int J Radiat Oncol Biol Phys. 1980;6(9):1215–28.
aerobic and hypoxic linear-quadratic (LQ) model parameters. 56. Wowra B, Schmitt HP, Sturm V. Incidence of late radiation necro-
Med Phys. 2006;33(9):3105–15. Epub 2006/10/07. sis with transient mass effect after interstitial low dose rate radio-
34. Carlson DJ, Keall PJ, Loo Jr BW, Chen ZJ, Brown therapy for cerebral gliomas. Acta Neurochir (Wien). 1989;
JM. Hypofractionation results in reduced tumor cell kill compared 99(3–4):104–8.
to conventional fractionation for tumors with regions of hypoxia. Int 57. Touboul E, Al Halabi A, Buffat L, Merienne L, Huart J, Schlienger
J Radiat Oncol Biol Phys. 2011;79(4):1188–95. Epub 2010/12/25. M, et al. Single-fraction stereotactic radiotherapy: a dose-response
35. Carlson DJ, Yenice KM, Orton CG. Tumor hypoxia is an impor- analysis of arteriovenous malformation obliteration. Int J Radiat
tant mechanism of radioresistance in hypofractionated radiother- Oncol Biol Phys. 1998;41(4):855–61.
apy and must be considered in the treatment planning process. 58. Flickinger JC, Kondziolka D, Maitz AH, Lunsford LD. An analy-
Med Phys. 2011;38(12):6347–50. Epub 2011/12/14. sis of the dose-response for arteriovenous malformation radiosur-
36. Brenner DJ, Hall EJ. Conditions for the equivalence of continuous gery and other factors affecting obliteration. Radiother Oncol.
to pulsed low dose rate brachytherapy. Int J Radiat Oncol Biol 2002;63(3):347–54.
Phys. 1991;20(1):181–90. 59. Kocher M, Wilms M, Makoski HB, Hassler W, Maarouf M, Treuer
37. Fowler JF. The linear-quadratic formula and progress in fraction- H, et al. Alpha/beta ratio for arteriovenous malformations esti-
ated radiotherapy. Br J Radiol. 1989;62(740):679–94. mated from obliteration rates after fractionated and single-dose
38. Thames HD, Hendry JH. Fractionation in radiotherapy. London: irradiation. Radiother Oncol. 2004;71(1):109–14.
Taylor & Francis; 1987. p. ix, 297. 60. Mayer A, Hockel M, Wree A, Leo C, Horn LC, Vaupel P. Lack of
39. Thames HD, Bentzen SM, Turesson I, Overgaard M, van den hypoxic response in uterine leiomyomas despite severe tissue
Bogaert W. Fractionation parameters for human tissues and hypoxia. Cancer Res. 2008;68(12):4719–26. Epub 2008/06/19.
tumors. Int J Radiat Biol. 1989;56(5):701–10. 61. Shrieve DC, Hazard L, Boucher K, Jensen RL. Dose fractionation
40. Withers HR, Thames Jr HD, Flow BL, Mason KA, Hussey in stereotactic radiotherapy for parasellar meningiomas: radiobio-
DH. The relationship of acute to late skin injury in 2 and 5 frac- logical considerations of efficacy and optic nerve tolerance.
tion/week gamma-ray therapy. Int J Radiat Oncol Biol Phys. J Neurosurg. 2004;101 Suppl 3:390–5.
1978;4(7–8):595–601. 62. Goldsmith BJ, Rosenthal SA, Wara WM, Larson DA. Optic neu-
41. Withers HR, Taylor JM, Maciejewski B. The hazard of acceler- ropathy after irradiation of meningioma. Radiology. 1992;185(1):
ated tumor clonogen repopulation during radiotherapy. Acta 71–6.
Oncol. 1988;27(2):131–46. 63. Bhandare N, Monroe AT, Morris CG, Bhatti MT, Mendenhall
42. Brenner DJ. Accelerated repopulation during radiotherapy—evi- WM. Does altered fractionation influence the risk of radiation-
dence for delayed onset. Radiat Oncol Invest. 1993;1:167–72. induced optic neuropathy? Int J Radiat Oncol Biol Phys.
43. Slevin NJ, Hendry JH, Roberts SA, Agren-Cronqvist A. The 2005;62(4):1070–7.
effect of increasing the treatment time beyond three weeks on the 64. Dale RG. The application of the linear-quadratic dose-effect equa-
control of T2 and T3 laryngeal cancer using radiotherapy. tion to fractionated and protracted radiotherapy. Br J Radiol.
Radiother Oncol. 1992;24(4):215–20. 1985;58(690):515–28.
44. Dasu A, Fowler JF. Comments on “Comparison of in vitro and 65. Lea DE. Actions of radiation on living cells. Cambridge:
in vivo alpha/beta ratios for prostate cancer”. Phys Med Biol. University Press; 1946.
2005;50(6):L1–4; author reply L5–8. Epub 2006/09/28. 66. Sachs RK, Hahnfeld P, Brenner DJ. The link between low-LET
45. Saunders MI, Dische S, Grosch EJ, Fermont DC, Ashford RF, dose-response relations and the underlying kinetics of damage
Maher EJ, et al. Experience with CHART. Int J Radiat Oncol Biol production/repair/misrepair. Int J Radiat Biol. 1997;72(4):
Phys. 1991;21(3):871–8. 351–74.
46. Hall EJ, Brenner DJ. The radiobiology of radiosurgery: rationale 67. Ellis F. Is NSD-TDF useful to radiotherapy? Int J Radiat Oncol
for different treatment regimes for AVMs and malignancies. Int J Biol Phys. 1985;11(9):1685–97.
Radiat Oncol Biol Phys. 1993;25(2):381–5. 68. Turesson I, Notter G. Skin reaction as a biological parameter for
47. Loeffler JS, Kooy HM, Wen PY, Fine HA, Cheng CW, Mannarino prospective studies of different dose schedules with the CRE for-
EG, et al. The treatment of recurrent brain metastases with stereo- mula. Bull Cancer. 1976;63(1):11–26.
tactic radiosurgery. J Clin Oncol. 1990;8(4):576–82. 69. Bates TD, Peters LJ. Dangers of the clinical use of the NSD for-
48. Ogilvy CS. Radiation therapy for arteriovenous malformations: mula for small fraction numbers. Br J Radiol. 1975;48(573):773.
a review. Neurosurgery. 1990;26(5):725–35. 70. Peters LJ, Withers HR. Morbidity from large dose fractions in
49. Friedman WA, Bova FJ. Radiosurgery for arteriovenous malfor- radiotherapy. Br J Radiol. 1980;53(626):170–1.
mations. Neurol Res. 2011;33(8):803–19. Epub 2011/10/19. 71. Gerweck LE, Zaidi ST, Zietman A. Multivariate determinants of
50. Barr JC, Ogilvy CS. Selection of treatment modalities or observa- radiocurability. I: prediction of single fraction tumor control
tion of arteriovenous malformations. Neurosurg Clin N Am. doses. Int J Radiat Oncol Biol Phys. 1994;29(1):57–66.
2012;23(1):63–75. Epub 2011/11/24. 72. Brown M, Bristow R, Glazer P, Hill R, McBride W, McKenna G,
51. See AP, Raza S, Tamargo RJ, Lim M. Stereotactic radiosurgery of et al. Comment on “Tumor response to radiotherapy regulated by
cranial arteriovenous malformations and dural arteriovenous fistu- endothelial cell apoptosis” (II). Science. 2003;302(5652):1894;
las. Neurosurg Clin N Am. 2012;23(1):133–46. Epub 2011/11/24. author reply.
52. Kjellberg RN, Hanamura T, Davis KR, Lyons SL, Adams 73. Carlson DJ, Stewart RD, Semenenko VA, Sandison GA. Combined
RD. Bragg-peak proton-beam therapy for arteriovenous malfor- use of Monte Carlo DNA damage simulations and deterministic
mations of the brain. N Engl J Med. 1983;309(5):269–74. repair models to examine putative mechanisms of cell killing.
53. Marks LB, Spencer DP. The influence of volume on the tolerance Radiat Res. 2008;169(4):447–59. Epub 2008/03/28.
of the brain to radiosurgery. J Neurosurg. 1991;75(2):177–80. 74. Preston RJ. Mechanisms of induction of specific chromosomal
54. Nedzi LA, Kooy H, Alexander III E, Gelman RS, Loeffler alterations. Basic Life Sci. 1990;53:329–36.
JS. Variables associated with the development of complications 75. Wlodek D, Hittelman WN. The relationship of DNA and chromo-
from radiosurgery of intracranial tumors. Int J Radiat Oncol Biol some damage to survival of synchronized X-irradiated L5178Y
Phys. 1991;21(3):591–9. cells. II. Repair. Radiat Res. 1988;115(3):566–75.
76. Savage JR. A brief survey of aberration origin theories. Mutat Res. 96. Barendsen GW. Dose fractionation, dose rate and iso-effect
1998;404(1–2):139–47. relationships for normal tissue responses. Int J Radiat Oncol Biol
77. Lea DE, Catcheside DG. The mechanism of the induction by radi- Phys. 1982;8(11):1981–97.
ation of chromosome aberrations in tradescantia. J Genet. 97. van der Kogel AJ. Chronic effects of neutrons and charged parti-
1942;44:216–45. cles on spinal cord, lung, and rectum. Radiat Res Suppl.
78. Kellerer AM, Rossi HH. The theory of dual radiation action. Curr 1985;8:S208–16.
Top Radiat Res. 1972;8:85–158. 98. Douglas BG, Fowler JF. The effect of multiple small doses of x
79. Frankenberg D, Brede HJ, Schrewe UJ, Steinmetz C, Frankenberg- rays on skin reactions in the mouse and a basic interpretation.
Schwager M, Kasten G, et al. Induction of DNA double-strand Radiat Res. 1976;66(2):401–26.
breaks by 1H and 4He lons in primary human skin fibroblasts in 99. Peck JW, Gibbs FA. Mechanical assay of consequential and pri-
the LET range of 8 to 124 keV/microm. Radiat Res. mary late radiation effects in murine small intestine: alpha/beta
1999;151(5):540–9. analysis. Radiat Res. 1994;138(2):272–81.
80. Frankenberg-Schwager M. Induction, repair and biological rele- 100. Taylor JM, Kim DK. The poor statistical properties of the Fe-plot.
vance of radiation-induced DNA lesions in eukaryotic cells. Int J Radiat Biol. 1989;56(2):161–7.
Radiat Environ Biophys. 1990;29(4):273–92. 101. de Boer RW. The use of the D versus dD plot to estimate the alpha/
81. Rothkamm K, Lobrich M. Evidence for a lack of DNA double- beta ratio from iso-effect radiation damage data. Radiother Oncol.
strand break repair in human cells exposed to very low x-ray 1988;11(4):361–7.
doses. Proc Natl Acad Sci U S A. 2003;100(9):5057–62. 102. Tucker SL. Tests for the fit of the linear-quadratic model to radia-
82. Sutherland BM, Bennett PV, Schenk H, Sidorkina O, Laval J, tion isoeffect data. Int J Radiat Oncol Biol Phys. 1984;10(10):
Trunk J, et al. Clustered DNA damages induced by high and low 1933–9.
LET radiation, including heavy ions. Phys Med. 2001;17 Suppl 103. Ling CC, Chen CH, Fuks Z. An equation for the dose response of
1:202–4. radiation-induced apoptosis: possible incorporation with the LQ
83. Thames HD. An ‘incomplete-repair’ model for survival after frac- model. Radiother Oncol. 1994;33(1):17–22.
tionated and continuous irradiations. Int J Radiat Biol. 104. Brenner DJ, Hlatky LR, Hahnfeldt PJ, Huang Y, Sachs RK. The
1985;47(3):319–39. linear-quadratic model and most other common radiobiological
84. Brenner DJ, Huang Y, Hall EJ. Fractionated high dose-rate versus models result in similar predictions of time-dose relationships.
low dose-rate regimens for intracavitary brachytherapy of the cer- Radiat Res. 1998;150(1):83–91.
vix: equivalent regimens for combined brachytherapy and external 105. Curtis SB. Lethal and potentially lethal lesions induced by
irradiation. Int J Radiat Oncol Biol Phys. 1991;21(6):1415–23. radiation–a unified repair model. Radiat Res. 1986;106(2):
85. Wang JZ, Li XA, D’Souza WD, Stewart RD. Impact of prolonged 252–70.
fraction delivery times on tumor control: a note of caution for 106. Hawkins RB. A microdosimetric-kinetic model of cell death from
intensity-modulated radiation therapy (IMRT). Int J Radiat Oncol exposure to ionizing radiation of any LET, with experimental and
Biol Phys. 2003;57(2):543–52. Epub 2003/09/06. clinical applications. Int J Radiat Biol. 1996;69(6):739–55.
86. Carlson DJ, Stewart RD, Li XA, Jennings K, Wang JZ, Guerrero 107. Obaturov GM, Moiseenko VV, Filimonov AS. Model of mam-
M. Comparison of in vitro and in vivo alpha/beta ratios for prostate malian cell reproductive death. I. Basic assumptions and general
cancer. Phys Med Biol. 2004;49(19):4477–91. Epub 2004/11/24. equations. Radiat Environ Biophys. 1993;32(4):285–94.
87. Webb S, Nahum AE. A model for calculating tumour control 108. Tobias CA. The repair-misrepair model in radiobiology: compari-
probability in radiotherapy including the effects of inhomoge- son to other models. Radiat Res Suppl. 1985;8:S77–95.
neous distributions of dose and clonogenic cell density. Phys Med 109. Zaider M. There is no mechanistic basis for the use of the linear-
Biol. 1993;38(6):653–66. Epub 1993/06/01. quadratic expression in cellular survival analysis. Med Phys.
88. Tome WA, Fowler JF. On the inclusion of proliferation in tumour 1998;25(5):791–2.
control probability calculations for inhomogeneously irradiated 110. Sachs RK, Brenner DJ. The mechanistic basis of the linear-
tumours. Phys Med Biol. 2003;48(18):N261–8. Epub 2003/10/08. quadratic formalism. Med Phys. 1998;25(10):2071–3.
89. Keall PJ, Webb S. Optimum parameters in a model for tumour 111. Kiefer J. A repair fixation model. In: Kiefer J, editor. Quantitative
control probability, including interpatient heterogeneity: evalua- mathematical models in radiation biology. New York: Springer;
tion of the log-normal distribution. Phys Med Biol. 1988.
2007;52(1):291–302. Epub 2006/12/22. 112. Haynes RH. The interpretation of microbial inactivation and
90. Fowler JF, Ritter MA, Chappell RJ, Brenner DJ. What hypofrac- recovery phenomena. Radiat Res. 1966;(Suppl 6):1–29.
tionated protocols should be tested for prostate cancer? Int J 113. Laurie J, Orr JS, Foster CJ. Repair processes and cell survival. Br
Radiat Oncol Biol Phys. 2003;56(4):1093–104. J Radiol. 1972;45(533):362–8.
91. Prevost JB, Voet P, Hoogeman M, Praag J, Levendag P, Nuyttens 114. Reddy NM, Mayer PJ, Lange CS. The saturated repair kinetics of
JJ. Four-dimensional stereotactic radiotherapy for early stage non- Chinese hamster V79 cells suggests a damage accumulation–
small cell lung cancer: a comparative planning study. Technol interaction model of cell killing. Radiat Res. 1990;121(3):
Cancer Res Treat. 2008;7(1):27–34. 304–11.
92. Brenner DJ, Martel MK, Hall EJ. Fractionated regimens for ste- 115. Sontag W. A cell survival model with saturable repair after irradia-
reotactic radiotherapy of recurrent tumors in the brain. Int J Radiat tion. Radiat Environ Biophys. 1987;26(1):63–79.
Oncol Biol Phys. 1991;21(3):819–24. 116. Ward JF, Limoli CL, Calabro-Jones PM, Aguilera J. An examina-
93. Brenner DJ. The linear-quadratic model is an appropriate method- tion of the repair saturation hypothesis for describing shouldered
ology for determining isoeffective doses at large doses per frac- survival curves. Radiat Res. 1991;127(1):90–6.
tion. Semin Radiat Oncol. 2008;18(4):234–9. 117. Kirkpatrick JP, Meyer JJ, Marks LB. The linear-quadratic model
94. Bartkowiak D, Hogner S, Nothdurft W, Rottinger EM. Cell cycle is inappropriate to model high dose per fraction effects in
and growth response of CHO cells to X-irradiation: threshold-free radiosurgery. Semin Radiat Oncol. 2008;18(4):240–3. Epub
repair at low doses. Int J Radiat Oncol Biol Phys. 2001;50(1): 2008/08/30.
221–7. 118. Garcia-Barros M, Paris F, Cordon-Cardo C, Lyden D, Rafii S,
95. Garcia LM, Leblanc J, Wilkins D, Raaphorst GP. Fitting the Haimovitz-Friedman A, et al. Tumor response to radiotherapy
linear-quadratic model to detailed data sets for different dose regulated by endothelial cell apoptosis. Science. 2003;300(5622):
ranges. Phys Med Biol. 2006;51(11):2813–23. 1155–9. Epub 2003/05/17.
119. Yamada Y, Bilsky MH, Lovelock DM, Venkatraman ES, Toner S, 123. Brown JM, Brenner DJ, Carlson DJ. Dose escalation, not “new
Johnson J, et al. High-dose, single-fraction image-guided biology,” can account for the efficacy of stereotactic body radia-
intensity-modulated radiotherapy for metastatic spinal lesions. tion therapy with non-small cell lung cancer. Int J Radiat Oncol
Int J Radiat Oncol Biol Phys. 2008;71(2):484–90. Epub Biol Phys. 2013;85(5):1159–60.
2008/02/01. 124. Brown JM, Carlson DJ, Brenner DJ. The tumor radiobiology of
120. Postow MA, Callahan MK, Barker CA, Yamada Y, Yuan J, SRS and SBRT: are more than the 5 Rs involved? Int J Radiat
Kitano S, et al. Immunologic correlates of the abscopal effect in a Oncol Biol Phys. 2014;88(2):254–62.
patient with melanoma. N Engl J Med. 2012;366(10):925–31. 125. Mehta N, King CR, Agazaryan N, Steinberg M, Hua A, Lee
Epub 2012/03/09. P. Stereotactic body radiation therapy and 3-dimensional confor-
121. Hiniker SM, Chen DS, Knox SJ. Abscopal effect in a patient with mal radiotherapy for stage I non-small cell lung cancer: a pooled
melanoma. N Engl J Med. 2012;366(21):2035; author reply 2036. analysis of biological equivalent dose and local control. Pract
Epub 2012/05/25. Radiat Oncol. 2012;2:288–95.
122. Brown JM, Koong AC. High-dose single-fraction radiotherapy: 126. Lo SS, Fakiris AJ, Chang EL, Mayr NA, Wang JZ, Papiez L, et al.
exploiting a new biology? Int J Radiat Oncol Biol Phys. Stereotactic body radiation therapy: a novel treatment modality.
2008;71(2):324–5. Epub 2008/05/14. Nat Rev Clin Oncol. 2010;7(1):44–54. Epub 2009/12/10.
Treatment Planning for Stereotactic
Radiosurgery 6
David M. Shepard, Cedric Yu, Martin J. Murphy,
Marc Bussière, and Frank J. Bova
Introduction Background on SRS Planning
This chapter provides an introduction to treatment-planning Over the past three decades, radiosurgery has been trans-
procedures for stereotactic radiosurgery (SRS). A brief formed from a curiosity available at a single institution to a
history of SRS planning is provided in section “Background routinely used therapy available at medical facilities through-
on SRS Planning.” The basic steps followed in the develop- out the world. During this increase in utilization, radiosurgi-
ment of an SRS treatment plan are described in section cal planning and delivery techniques have undergone
“Basic Steps in Developing an SRS Treatment Planning” significant technical improvements including: (1) availabil-
including imaging, contouring, selection of plan parameters, ity of true three-dimensional (3D) image data, (2) improved
and evaluating treatment plan quality. Section “Gamma plan delivery techniques, and (3) increased sophistication of
Knife” focuses on treatment planning for the Gamma Knife. planning algorithms.
Sections “Linac-Based SRS with Circular Cones” and The most critical improvements have come about through
“Linac-Based SRS with Multileaf Collimators” discuss lin- the availability of true three-dimensional image data pro-
ear accelerator-based SRS, and the final two sections (sec- vided primarily through computed tomography (CT) and
tions “CyberKnife” and “Proton SRS”) cover planning for magnetic resonance imaging (MRI) scanning techniques.
the CyberKnife and proton radiosurgery. Today’s practicing radiosurgeon has the advantage of view-
ing target and nontarget tissues with increased contrast reso-
lution and with dramatically improved spatial resolution.
Clinician has also benefited from the development of tools
for registering or fusing single and multimodality datasets.
The second area that has undergone significantly improve-
ment is that of plan delivery. For several years, the only
radiosurgical device available was the Gamma Knife. The
Gamma Knife has now been joined by several designs that
D.M. Shepard, Ph.D. (*)
Department of Radiation Oncology, Swedish Cancer Institute, use linear accelerators as the radiation source. While the ini-
Seattle, WA, USA tial linear accelerator (linac)-based systems were less accu-
e-mail: david.shepart@swedish.org rate than the Gamma Knife, devices were developed in the
C. Yu, D.Sc. late 1980s that provided significant improvement in the
Department of Radiation Oncology, University of Maryland delivery accuracy of linac-based radiosurgical units. Recent
School of Medicine, Baltimore, MD, USA years have seen additional advancements in basic linac tech-
M.J. Murphy, Ph.D. nology resulting in highly accurate base linear accelerators.
Department of Radiation Oncology, Virginia Commonwealth These devices provide efficient radiosurgical delivery, and
have made it possible to extend the benefits of radiosurgery
M. Bussière, M.Sc. to targets outside the cranium.
Department of Radiation Oncology, Massachusetts General
Hospital, Boston, MA, USA The third area of technological advancement in radiosur-
gery has been the increased sophistication of the planning
F.J. Bova, Ph.D.
Department of Neurosurgery, University of Florida, algorithms that has been made possible by the tremendous
Gainesville, FL, USA increase in computer processing power over the past three
74 D.M. Shepard et al.
decades. The ability to process images, register images, and a large number of beam paths are required to concentrate the
develop plans has gone from a process that required many dose in the target and create steep dose gradients. It is impor-
hours to one that can be accomplished in a matter of seconds. tant to note that as new SRS delivery techniques emerge, cli-
This has allowed radiosurgical teams to not only develop and nicians must examine the impact on dose conformity as
evaluate their first best guess at a plan but to iterate through compared with historical SRS delivery techniques using the
many different plans in order to arrive at a more optimal Gamma Knife or linac-based delivery with circular collima-
plan. tors. A loss in dose conformity may force the user to adopt a
While these planning and delivery tools help provide treatment schedule that deviates from basic radiosurgical
improved treatment plans, the basics principles set forth by principles and begins to depend more upon radiological prin-
Lars Leksell in the 1950s remain the foundation for radiosur- ciples. While this may be appropriate in certain clinical situ-
gery [1–4]. Leksell’s basic idea was to place the target tis- ations, a cautious approach is warranted.
sues in the center of a large number of beams of radiation.
The beams arrangement was designed so that the beams only
intersected over the target issues and they quickly diverged Basic Steps in Developing an SRS Treatment
from the target in all directions. Planning
The majority of clinical data published on radiosurgery is
based upon a planning technique known as sphere packing. The basic steps followed in radiosurgery treatment planning
This technique was initially developed by Leksell and con- are illustrated in Fig. 6.1.
stitutes the basis of the Gamma Knife planning process. It
has also been used in conjunction with many of the linac-
based delivery systems. In this technique, a set of beams are Stereotactic Localization
aimed at a point in space, known as the isocenter. The beams
are selected so that they approach and leave the isocenter Because SRS involves the delivery of a very high dose of
through unique paths, providing both geometric concentra- radiation in either a single delivery or a small number of
tion and a high-dose gradient. Much of the accuracy dis- fractions, it is essential to achieve a high degree of dose
cussed in the radiosurgery literature is a measure of the conformity and accurate patient positioning. This critical
delivery systems’ ability to precisely target radiation beams
at this “point” in space. The resultant dose distribution is
approximately spherical and therefore provides a reasonable
plan for spherical target volumes. For more complex target
shapes, a technique known as sphere packing is employed. In
this technique, the planner places the initial dose sphere
inside the target volume. The initial dose sphere is typically
selected to be the largest sphere that the system can produce
and that can fit “inside” the target volume. The planner then
continues to pack the target volume with spheres of equal or
smaller diameter until adequate target coverage is achieved.
An advantage of linear accelerator-based system over the
Gamma Knife is the ability to produce larger spheres and
therefore pack a specific volume with fewer total spheres or
isocenters. These spheres of dose are created using beam sets
that deliver radiation from a large number of beam angles.
The Gamma Knife makes use of 192 (or 201) individual
beams while multiple arcs are used in the linear accelerator
systems. It has been shown that in order to produce the con-
centration and gradient, which is responsible for a vast major-
ity of all published radiosurgical clinical outcomes, at least 15
beams must be spread over approximately 2π space [5].
In recent years, SRS delivery techniques have expanded
beyond the Gamma Knife and linac-based delivery using cir-
cular collimators to include the CyberKnife, tomotherapy,
and the adaptation of linacs with mini multileaf collimators
(MLC). While each of these systems offers the radiosurgeon
special features, they are all bound by the same principle that Fig. 6.1 The basic steps in SRS treatment planning
6 Treatment Planning for Stereotactic Radiosurgery 75
Fig. 6.3 (a) The fiducial indicator box for CT imaging. (b) The patient
is imaged with a fiducial box attached to the frame (Photos courtesy of
Elekta)
Fig. 6.2 Gamma knife coordinate frame (Photo courtesy of Elekta)
dosimetric requirement is accomplished by stereotactically

focusing many convergent radiation beams on the target. By
cross-firing a large number of beams, a high dose is delivered
in the region of intersection with a rapid falloff in dose out-
side of the target.
One approach to accurate stereotactic localization is
achieved with fixation frames such as the frame shown in
Fig. 6.2. On the day of the procedure, this fixation frame is
fixed to the patient’s skull under local anesthetic. The frame
serves two important purposes. First, it provides a rigid fixa-
tion system that ensures that the patient cannot move during
the delivery. Secondly, it provides a frame of reference
whereby the tumor location can be determined relative to the
frame and relative to the delivery unit.
One approach for facilitating the registration of the patient
images is to attach a fiducial box (see Fig. 6.3) to the stereo- Fig. 6.4 The fiducial marks appear as white dots on this MRI image of
tactic head frame before patient imaging. The box includes a Gamma Knife patient (Photo courtesy of Elekta)
tubes filled with a high-contrast solution that provides easy
to delineate fiducial marks on each of the patient’s images invasive nature of fixation frames makes them a less viable
(see Fig. 6.4). option. Fractionation radiosurgery, known as stereotactic
After the imaging is complete, the patient images are radiotherapy (SRT), is commonly delivered to patients where
loaded into the planning system. The images are registered the likelihood of late toxicity makes it inadvisable to treat a
with respect to the fiducial marks. The markings on the fidu- large target volume in a single high-dose fraction [7, 12–14].
cial box provide a coordinate system that makes it possible to
precisely determine the position of the treatment volume
with respect to the stereotactic head frame. Patient Imaging
In addition to the fixation frames described above, frame-
less SRS techniques have been developed that maintain a The three primary imaging techniques for SRS are MRI, CT,
highly precise delivery [6–12]. Frameless SRS is typically the and planar angiography. MRI is an excellent technique for
preferred option in fractionated radiosurgery where the imaging soft tissue contrast. MRI, however, does not provide
patient images into the treatment-planning system via a

computer network or data storage media. On the planning
computer, these images are typically displayed in two-dimen-
sional slices along the axial, sagittal, or coronal planes [15].
A screen capture from the GammaPlan system is shown
in Fig. 6.4 that illustrates the process of aligning the fiducial
marks in the SRS planning system. The planning system
must be able to accurately identify the fiducial marks and
have internal mechanisms for validating their consistency.
Once the fiducial markers are identified, the patient anatomy
Fig. 6.5 One approach to image registration shown on the images is placed in the coordinate system
defined by the stereotactic localization device.
attenuation coefficients, and must be fused with CT images Some systems use fiducial markers that fully define the
if one wishes to correct for to make appropriate heterogene- stereotactic reference system while others require specific
ities in the dose calculation. It should be noted that in most information from the scanner. For example, a fully defined
MRI units the spatial uniformity of the images degrades as system does not require information on the axial slice posi-
the radius of the images increases. Because most fiducial tion. This is because it can derive this information from the
systems place the fiducial markers at the outermost extent of actual scan image. Consequently, this approach eliminates
the MRI image volume, the accuracy obtained in mapping the need for QA measurements to guarantee accurate scanner
the imaged voxels to stereotactic space can be compromised. information such as linearity of table movement or degree of
This is one of several reasons some radiosurgical practitio- gantry tilt.
ners have opted for image fusion technologies as opposed to Frameless systems also need to be able to map the three-
direct stereotactic MRI scanning (see Fig. 6.5). dimensional dataset into a rigid and definable coordinate sys-
CT imaging provides inferior soft tissue contrast as com- tem. To achieve this, stereotactic systems have been designed
pared with MRI scanning. However, the spatial resolution of that incorporate surrogate fiducial markers into the CT or
CT is extremely reliable. CT also provides attenuation coef- MRI scans. These markers can also be referenced at the time
ficients that make it possible to make accurate heterogeneity of therapy through mechanical, electrical, or optical means.
corrections in the dose calculation. A critical feature of frameless systems is the ability to demon-
Planar angiography is an imaging option that can be used strate that the reference can in fact be reliably and precisely
for imaging patients with arteriovenous malformations fixed to the patient. Many systems offer testing only at the
(AVMs). While plane film angiography was the gold stan- time of design and development. These tests are usually at the
dard for AVM nidus identification, cut films have almost hands of the developer, an expert with extensive experience.
been fully replaced by electronically derived images. For the It is critical for a repeat fixation system to also provide a
past few decades, the primary mode of electronic image methodology of testing its accuracy for each patient.
acquisition has been the image intensifier. While this device Shows the RadioCameras™ Treatment Guidance System
is known for its excellent image contrast, it is equally known (Varian, Palo Alto, CA). This system provides patient-specific
for its poor spatial accuracy. Most image intensifiers have statistics on the device’s ability to be precisely reapplied.
multiple levels of image distortion. More recently, solid-state Two basic approaches are used in the registration of
image acquisition systems have been developed. These sys- images. The first is to use direct stereotactic MRI and CT
tems have all but eliminated concerns regarding spatially scanning. In this procedure, fiducial markers are incorpo-
nonuniform images. rated into both image datasets and each dataset is indepen-
An additional problem with angiography is that while plane dently mapped into stereotactic space. In the second
film angiography can provide images in multiple planes, the approach, stereotactic CT scanning is performed using a
basic image set that the clinician has to work with is two fiducial system when acquiring the images. Non-stereotactic
dimensional. This two-dimensional image dataset can lead to MRI scans are also obtained (see Fig. 6.5). The MRI scans
inaccuracies in the target definition. It is for this reason that are then mapped into stereotactic space by registration of the
most targeting for vascular radiosurgery targets are obtained MRI dataset to the CT dataset. In most cases, rigid registra-
either through CT angiography, MRI angiography, or both. tion techniques are used. Typically, the clinician visually
inspects the results of the registration to verify the quality of
the fit. An advantage of this second approach is that it allows
Registering the Images the MRI scan to be obtained prior to the placement of any
stereotactic ring. Consequently, the clinician can exam the
The general planning procedure starts with the acquisition of actual scan that will be used for scanning prior to deciding
three-dimensional image information and the import of that radiosurgery is the appropriate mode for treatment.
Knife and linac-based SRS using cones, the use of multiple

isocenters means that multiple spherical isodose distribu-
tions have to be packed into the target volume. Overlaps of
such high-dose spheres are inevitable. As a result, the dose is
highly nonuniform in the target. The isodose surface that
encloses the target (which is often taken as the prescription
dose) is typically 50 % of the maximum dose in the target.
For linac-based SRS using an MLC, the target dose is
highly uniform when a single isocenter is used. The isodose
line that encompasses the target can be greater than 80 % of
the maximum dose in the target. In SRS, the prescription
dose is normally set to the dose level that conforms to the
target, or the minimum target dose. As compared with frac-
tionated radiation treatments, radiosurgery treatment plans
have significantly less normal tissue included in the prescrip-
tion isodose volume. For this reason, the typical restrictions
on dose uniformity (a foundation of fractionated radiation
therapy) do not apply in radiosurgery cases. The radiosur-
geon is more concerned with the dose to normal tissue out-
side the target volume. Consequently, the goal in treatment
planning is to achieve maximum normal tissue sparing rather
than maximum target dose uniformity.
The radiosurgeon can minimize the dose to normal tissue
Fig. 6.6 Skull-scaling device used with the Gamma Knife to provide by designing a plan where the isodose that just covers the
mapping of the skull (Photo courtesy of Elekta)
target surface is along the steepest portion of the dose gradi-
ent. It can be shown that for single-isocenter plans, this
Contouring Structures places the optimal dose prescription at the 80 % isodose sur-
face [16]. This is because the dose gradient falls off most
After the images have been imported into the planning sys- quickly, between the 80 and 40 % isodose surfaces. Therefore,
tem and the fiducial marks have been registered, the target to prescribing to the 80 % isodose line minimizes the integral
be treated is then identified. To aid target delineation, fusion dose to all normal tissues, the tissues outside the target vol-
of images from different imaging modalities is possible with ume. For multiple isocenter plans, this optimal surface shifts
many SRS planning systems. While some planning systems to the 70 % isodose surface and at time can be extended to
may not require the users to outline the target and critical isodose surfaces as low as the 50 % dose surface, as is often
structures, contouring is required if plan evaluation tools the case in Gamma Knife plans.
such dose-volume histograms (DVHs) are to be used.
Contouring is also required by systems that employ inverse
planning. Designing the Treatment Plan
The patient’s external contour can either be manually
contoured or identified automatically by the planning system For optimal planning, the neurosurgeon, radiation oncolo-
based on grayscale information from the images. For Gamma gist, and physicist or dosimetrist should perform the treat-
Knife planning, the skin contour or scalp of the patient’s ment planning as a team and bring all aspects of expertise to
head is measured with a skull-scaling device (see Fig. 6.6). bear on the problem. After the images have been imported
The location of the surface is important for computing the and registered and contouring of structures has been com-
penetration depth for each beam during the dose calculation. pleted, one can proceed with the task of formulating a treat-
The skull-scaling device used for Gamma Knife planning ment plan that meets the dosimetric requirements specified
makes it possible to know the penetration depth for each by the physician. The planning steps that are followed differ
photon beam even if the entire skull has not been imaged. from one delivery technique to the next. For the Gamma
Knife and for linac-based SRS with circular collimators, the
task of planning is to find a set of isocenter locations, the size
Defining the Prescription of the collimators to use for each location, and the weights of
the isocenters. Multiple isocenters are commonly used in a
The dosimetric characteristics of the treatment plan vary Gamma Knife treatment plan due to the limited number of
from one delivery approach to another. With the Gamma collimator sizes to choose from and the relative efficiency
with which each isocenter can be delivered [17–24]. For Forward Versus Inverse Planning
linac-based SRS with circular collimators, the planner strives
to minimize the number of isocenters in a plan to compen- Forward planning is the most common planning approach
sate for the relative inefficiency of the beam delivery [25, for SRS. With forward planning, the treatment plan is devel-
26]. Linac systems, however, have the advantage of a much oped through an iterative trial and error approach. From one
wider range of collimators usually extending from 5 to iteration to the next, the planner attempts to determine a set
50 mm in diameter. of parameters (such as beam angle, beam weights) that give
For linac-based SRS with MLCs, only one isocenter is an acceptable plan. The iterative process is usually stopped
used [25, 27–37], and the planning methods are similar to when the planner is no longer able to make noticeable
those used for external beam radiation therapy. The selection improvements in the plan quality.
of beam directions and field shapes is aided by the use of With inverse treatment planning, the user begins by out-
tools such as beam’s-eye-view (BEV) visualization and digi- lining the target and any sensitive structures. A series of
tally reconstructed radiographs (DRRs). treatment goals are then defined, and an optimization algo-
rithm determines the plan parameters that provide a plan that
best satisfies these goals. The quality of the plan is scored
Dose Calculation using an objective function and constraints. An objective
function reduces an entire treatment plan into a single
A variety of dose calculation techniques have been employed numerical value. The job of the optimizer is to either mini-
for SRS treatment planning [38–48]. Simple empirical dose mize or maximize the value of the objective function.
calculation methods were employed with many of the earlier Typically, an optimization will also include one or more con-
SRS planning systems [17]. For Gamma Knife radiosurgery, straints. A constraint is a condition that must be satisfied in
a standard set of beam data is included with the system, and order for a solution to be considered feasible. The most basic
the user verifies the dose profiles of an individual beams for constraint in any radiotherapy optimization is that the beam
each of the focusing helmets. The use of a standard dataset is weights must be nonnegative.
possible because all Gamma Knife systems use the same As compared with forward planning, inverse planning
design with only slight variations in the single beam dosim- provides two potential advantages. First, the time required
etry. The dose distribution in the patient is calculated by for planning can be reduced because much of the trial and
adding the dose distributions for all 192 beams [20]. For error is removed. Secondly, inverse planning should lead
linac-based SRS with circular collimators, the tissue- to improved plan quality due to the ability of the optimizer
maximum ratios (TMRs) and dose profiles can be measured to consider thousands of plan configurations in selecting
for each collimator size as a function of depths. For a given the optimal plan. Unfortunately, the quality of inverse
arc, the mean TMR for all beam directions is calculated and planning tools for SRS varies significantly from one deliv-
the total dose contributed by the arc can be calculated. Dose ery technique to the next and one planning system to the
distributions can be computed by approximating an arc as a next. Users of some systems will find that an experienced
series fixed beams and summing the dose distributions from planner using forward planning techniques produces the
all of the beams. highest quality plans.
For intracranial SRS, heterogeneity corrections are
generally not applied in the dose calculation. Corrections
to account for variations in the attenuating properties of Evaluation of Plan Quality
tissue, bone, and air are not needed because of the simple
geometry that is employed [44, 49]. The magnitude of the The most common tool for evaluating plan quality is the
resulting error can be estimated by assuming that the aver- visualization of isodose curves superimposed on the patient’s
age beam passes through 5 mm of skull with an average anatomical images. Figure 6.7a shows a typical isodose plot
density of 1.2 g/cc. Assuming an attenuation of 4 %/cm for for a Gamma Knife patient. In this case, the isodose curves
a 6 MV beam, the maximum error in absolute TMR that are plotted as a percentage of the maximum target dose.
would result from ignoring heterogeneity corrections is Isodose curves can also be plotted as a percentage of the pre-
less than 1 %. However, when SRS is extended to extracra- scribed dose or as absolute dose lines. When evaluating plan
nial applications, traditional empirical dose calculation quality, it is common practice to visualize isodose curves in
methods are no longer adequate for accurate dose calcula- the axial, sagittal, and coronal planes. Some systems can also
tions due to the presence of bone and air. Therefore, many display three-dimensional dose clouds that make it possible
treatment-planning systems designed for planning extra- to quickly assess the quality of the dose coverage.
cranial SRS employ 3D pencil beam dose calculation DVHs also serve as an important tool for analyzing plan
methods. quality. A DVH plot from a Gamma Knife patient is shown
Fig. 6.7 (a) A Gamma Knife isodose plot with the target outlined in red, the 50 % isodose line in yellow, and the 30 % isodose line in green. (b)
Dose-volume histogram for target in this case
in Fig. 6.7b. For each structure, the DVH plots the fraction of identical to the target volume). As a planning goal, Lomax
the volume covered by each dose level. Both the dose and the and Scheib suggest a conformity index of 0.6 or higher.
volume can be expressed as either absolute or relative values. The RTOG has also defined a homogeneity index that is
DVHs are particularly useful because they reduce a three- equal to the maximum dose in the treatment volume divided
dimensional treatment plan into an easy to read two- by the prescription dose. A case is considered per protocol if
dimensional plot. The comparison of multiple plans is also this ratio is less than or equal to 2.0. In terms of target cover-
simplified by overlying DVHs on the same plot. The disad- age, the RTOG considers a case to be per protocol if the iso-
vantage of DVHs is the lack of spatial information. A DVH dose line equal to 90 % of the prescribed dose completely
will indicate the presence of hot or cold spots, but it does not encompasses the target. Lomax and Scheib suggest an alter-
provide specify where in the structure of interest this under- native volumetric definition of coverage where target cover-
dosage or underdosage is located. Consequently, slice-by- age is defined as the percentage of the target volume covered
slice evaluation of a plan is critical in making a final decision by the prescription [55].
regarding a treatment plan.
Additional parameters have been defined to score both
dose conformity and target coverage for SRS treatment plans Gamma Knife
[50–58]. The most commonly used parameter for scoring
dose conformity is the conformity index defined in the radio- Gamma Knife Unit
surgery quality assurance guidelines of the Radiation Therapy
Oncology Group (RTOG) [52]. The RTOG defined the con- The first Gamma Knife was built in 1967 under the direction
formity index as the volume of the prescription isodose sur- of Lars Leksell of the Karolinska Institute in Stockholm,
face divided by the target volume. For a perfectly conformal Sweden. Currently, there are over 300 units worldwide with
plan where the prescription isodose line exactly matches the approximately 60,000 patients treated annually [59]. Over
target volume, the conformity index would equal 1. A case is the years, there have been a number of advances and updates
considered to be per protocol if this ratio falls between 1.0 to the Gamma Knife. Two models of the Gamma Knife are
and 2.0. A shortcoming of this index is that it does not concurrently offered by Elekta, the 4C (introduced in 2004) and
sider the degree of overlap between the prescription isodose the Perfexion (introduced in 2006).
curve and the target. A plan with a complete geometric miss
of the target could still give a perfect conformity index. As an
alternative, Lomax and Scheib have suggested a conformity Gamma Knife 4C
index defined as “the ratio of the volume within the target
irradiated to at least the prescription isodose over the total Inside of the shielded treatment unit of the Gamma Knife 4C
volume enclosed by the prescription isodose” [55]. (Fig. 6.8), the beams from 201 radioactive sources are
Consequently, this value ranges from 0 (no conformity) to 1.0 focused so that they intersect at a single location. The result
(for perfect conformation, where the prescription isodose is is an elliptical region of high dose with a rapid falloff in dose
Fig. 6.8 A patient positioned for treatment on a Gamma Knife model

4C (Photo courtesy of Elekta)
outside of the boundaries of the ellipse. Each exposure to an

elliptical region of high dose is referred to as a “shot” of
radiation.
For each exposure, the focusing helmet dictates the size
of the high-dose region. A focusing helmet incorporates a
separate collimator for each of the 201 Co-60 sources (see
Fig. 6.9 The four focusing helmets dictate the size of each shot of
Fig. 6.9a). The four focusing helmets provided with the radiation (Photo courtesy of Elekta)
Gamma Knife can be used to produce a shot of radiation that
is 4, 8, 14, or 18 mm in diameter (see Fig. 6.9b). Within each
helmet, individual collimators may be plugged to provide couch is advanced into the high radiation field. Next, the
further shaping of the dose distribution. The benefits of plug- treatment couch is docked, and the shot of radiation is deliv-
ging are illustrated in Fig. 6.9. ered. After the exposure is complete, the couch retracts and
Figure 6.10a shows the isodose curve for a single 4 mm the door closes. If more than one shot of radiation is to be
shot for a trigeminal neuralgia patient. In Fig. 6.10b, a plug- delivered, the process is repeated with the appropriate focus-
ging pattern has been used to lower the dose to the brainstem. ing helmet and x, y, and z positioning. The model 4C Gamma
The plugging pattern is shown in Fig. 6.10c. Knife provides an automated positioning system in which
After the imaging is complete, the patient images are the shot positions and exposure times are directly transferred
loaded into the Gamma Knife’s treatment-planning system to a record-and-verify system, and the machine sets the coor-
(GammaPlan). The images are registered with respect to the dinates in an automated fashion.
fiducial marks, and the treatment volume is outlined by a The treatment planning for the Gamma Knife 4C is rela-
physician. tively straightforward for small spherical targets. For exam-
At the time of treatment, the patient lies on the couch of ple, Fig. 6.11 shows a case where the treatment volume is
the treatment unit, and the appropriate focusing helmet is relatively spherical and approximately 6 mm in diameter. An
affixed to the table. The patient’s stereotactic head frame is 8 mm shot of radiation was placed so that it covers the entire
then attached to the focusing helmet, and the x, y, and z coor- tumor volume.
dinates are adjusted to place the center of the planned shot of The treatment-planning process becomes more complex
radiation at the focal point of the 201 beams of radiation. when the tumor volume is large or irregularly shaped. These
After, the personnel leave the room, the exposure time is pro- cases typically require several shots of radiation. Through an
grammed, the door of the shielding unit is opened, and the iterative process, the planner must determine the number of
Fig. 6.10 The isodose curves for a trigeminal neuralgia patient before (a) and after (b) plugging. (c) The plugging pattern. 35 of the 201 sources
have been blocked (blocked collimators are shown in black)
In the GammaPlan system (the Gamma Knife specific

treatment-planning system), users typically normalize to the
maximum dose and seek to cover the target with the 50 %
isodose line.
Gamma Knife Perfexion
In 2006, Elekta introduced the Gamma Knife Perfexion

which includes a significant redesign of the Gamma Knife.
The most critical change in the Perfexion is the new collima-
tor system. The new system replaces the multi-helmet colli-
mator setup with a single integrated permanent collimator
system that incorporates openings for 4, 8, and 16 mm treat-
ment beams.
The collimator is partitioned into eight independently
moveable sectors each delivering 24 beams of radiation
meaning there is a total of 192 Co-60 sources. The beam size
can be changed dynamically by sector, and individual sectors
can be blocked to provide further shaping of each shot of
Fig. 6.11 A single shot treatment plan for a Gamma Knife treatment. radiation. For each shot, an ellipsoidal shot of radiation can
The target outlined in blue, the 50 % isodose line in yellow, and the be delivered by utilizing the same sized beams from each of
30 % isodose line in green
the eight sectors. More sophisticated shaping of the dose dis-
tribution can be obtained by mixing beams of various sizes
and/or blocking individual sectors. Unlike with the model 4C
shots of radiation that are required along with the size, the where it is necessary to manually block individual beams to
location, and the weight that should be assigned to each. shape the dose distribution, the Perfexion makes it possible
In Fig. 6.12, a simple 2D bean-shaped target is used to to simply move individual sectors into the off position to
illustrate the general planning procedure [22]. Each frame block the beams that pass directly through a particular criti-
illustrates the dosimetric impact of an added shot of radia- cal structure.
tion. In this case, five shots of radiation provide a conformal In addition to the updated collimator design, the Perfexion
treatment plan. Note that three different shot sizes were used introduces a number of other changes to the Gamma Knife.
in creating this plan. One change is in the positioning system design. The
During the process of treatment planning, the plan quality Perfexion positions the patient by moving the couch rather
is evaluated through the use of isodose plots and DVHs. than moving patient’s head. Rapid transition times are
a b c
5 100 5 1 5 1
90 0.9 0.9
80 0.8 0.8
centimeters
70
centimeters
0.7
centimeters
0.7
60 0.6 0.6
2.5 50 2.5 0.5 2.5 0.5
40 0.4 0.4
30 0.3 0.3
20 0.2 0.2
10 0.1 0.1
0 0 0
2.5 5 2.5 5 2.5 5
centimeters centimeters centimeters
d e f
5 1.2 5 5
1.2 1.2
1.0
1.0 1.0
centimeters
centimeters
centimeters
0.8 0.8 0.8
2.5 0.6 2.5 2.5
0.6 0.6
0.4 0.4 0.4
0.2 0.2 0.2
0
Fig. 6.12 (a) A bean-shaped target. (b–f) The process of adding shots to create a treatment plan is illustrated
achieved moving from one shot of radiation to the next through an asymmetric arc arrangement, ellipsoidal instead
because there is no need to change focusing helmets and the of spherical isodose distributions can be created.
beams are simply moved into the off position during the tran- For tumors that are small and convex in shape, it is often
sition. This ability to quickly switch from one shot of radia- possible to treat with only one isocenter. When one isocenter
tion to the next facilitates the treatment of larger tumors as is used, the dose uniformity in the target is high and the dose
well as patients with multiple brain metastases. can be prescribed to a high isodose level. When the target
The Perfexion provides improved patient comfort due the volume is large or it deviates from a sphere or an ellipsoid,
increased space inside the collimator body. The larger colli- multiple isocenters are typically required to achieve adequate
mator size also eliminates the need for eccentric frame posi- target dose conformity. The spherical high-dose regions must
tioning, and makes it possible to treat most peripheral lesions. overlap in order to avoid leaving a cold spot in the target.
The Perfexion is also the first Gamma Knife model that Consequently, one must accept a decreased target dose uni-
offers a stereotactic frame with a vacuum-assisted bite block formity as compared to plans using a single isocenter. To
(the Extend system) that makes it possible to deliver frame- date, no negative clinical consequences have been reported
less fractionated treatments. as a result of the lack of dose uniformity seen in Gamma
Knife and linac-based circular collimator treatments.
When multiple isocenters are used, the treatment-planning
Linac-Based SRS with Circular Cones problem is similar to sphere packing problem of the Gamma
Knife [61], where spherical dose distributions are packed
Initial developments in the use of linear accelerator-based together to achieve a composite dose distribution conform-
SRS techniques were centered on the use of multiple con- ing to the target volume. Because the setup and delivery time
verging arc treatments delivered with circular collimators. per isocenter is typically longer for linac-based radiosurgery
For each treatment plan, the user must select the number of as compared with Gamma Knife-based radiosurgery, it is
arcs, the angular arrangement of each arc, the weight generally desirable to deliver a limited number of isocenters.
assigned to each arc, and the number of isocenters. When However, the greater selection of collimator sizes and the
one isocenter is used, the high isodose levels at the isocenter availability of larger collimator sizes as compared with the
are nearly spherical. A variety of arc arrangements have been Gamma Knife reduce the need for the use of a large number
reported [26, 60]. As compared with the Gamma Knife, a of isocenters. Overall, the Gamma Knife centers and linac-
greater number collimator sizes is available with circular based SRS centers have reported similar cure rates and com-
collimators. By adjusting the weightings of the arcs or plication levels.
isocenter can lead to plans that are less conformal with a less
Linac-Based SRS with Multileaf Collimators steep dose gradient at the target edge. The decreased dose
gradient is in part the result of the use of larger beams and
An MLC is a field-shaping device that uses movable leaves fewer unique beam paths. As mentioned previously, a routine
made out of a highly attenuating material such as tungsten in Gamma Knife plan will involve as many as 201 beams and a
order to generate arbitrary field shapes (see Fig. 6.13). The typical plan for a linac-based cone-beam treatment would
first MLCs designed for routine external beam delivery typi- include five arcs each covering approximately 100° of gantry
cally had leaf widths that projected to 1 cm at isocenter. rotation. By contrast, a typical micromultileaf plan may have
These MLCs lacked the geometric precision for shaping as few as six beams.
small irregularly shaped fields such as those in commonly Target size is a second issue that impacts one’s ability to
encountered in SRS. create a conformal treatment plan using an MLC. For small
Interest in the use of MLCs for radiosurgery increased targets, such as in the treatment of trigeminal neuralgia, high
when vendors introduced micromultileaf collimators precision in target localization and positioning is required.
(mMLCs), a type of MLC where each leaf projects to a width MLCs may not be suitable for target sizes significantly less
of between 2 and 5 mm at isocenter. mMLCs are suitable for than one centimeter due to the undulating field edges caused
SRS applications, because they are capable of shaping small by the finite leaf width. For larger targets, MLC can provide
irregular fields with acceptable geometric error. Typical efficient beam delivery and dose uniformity due to the use of
mMLCs have between 20 and 80 leaves, arranged in pairs. a single isocenter [50]. However, the dose to surrounding
The maximum field size of mMLCs generally varies from 8 structures increases as the size of the target increases [50].
to 20 cm, much greater than those available with traditional As a result, the dose to normal structures may exceed the
circular collimators. As a result, extracranial SRS and stereo- acceptable limits if one attempts to treat a large tumor with a
tactic body radiation therapy (SBRT) can be delivered. single fraction.
Additionally, vendors now offer MLCs designed for Generally multileaf collimators (including mMLCs) can
both conventional treatments and radiosurgery. For exam- be used for four delivery approaches: (1) fixed-field confor-
ple, Elekta’s Agility MLC has 160 leaves of projected mal delivery; (2) dynamic conformal arc delivery; (3) fixed-
width 0.5 cm at isocenter and offers field sizes up to field intensity-modulated radiation therapy (IMRT); and (4)
40 × 40 cm. Varian offers their HD 120™ MLC, a 120 leaf volumetric modulated arc therapy (VMAT).
MLC which can deliver field sizes up to 22 × 40 cm with
2.5 mm leaves (projected at isocenter) across the central
8 cm of the field. Fixed Fields: 3D Conformal Delivery
Due to their ease of use and wide range of possible appli-
cations, the advent of MLCs has led to a decline in popular- With MLCs, three-dimensional (3D) conformal therapy
ity of SRS delivery using circular collimators. treatment-planning techniques can be applied to SRS plan-
As compared with Gamma Knife and linac-based SRS ning. If there are a sufficient number of MLC-shaped beams,
with circular collimators, the use of an MLC and a single the dose distribution quality can rival that obtained using
with multiple arcs using circular collimators and multiple
isocenters. This is especially true when the targets are larger
and non-spherical in shape. Because only one isocenter is
needed, the dose uniformity created with MLCs is generally
better than that achieved with the use of circular collimators
and multiple isocenters. With the exception of the need for
stereotactic reconstruction from stereotactic frames and
markers, the planning method for SRS essentially matches
that for 3D conformal therapy. Each non-coplanar field is
shaped based on the BEV of the target with additional mar-
gins to account for the width of the beam penumbra, which is
wider than the penumbra from circular collimators and
increases with leaf width.
The use of fixed fields delivered with an MLC is best
suited for convex target shapes. With fixed fields, there is
Fig. 6.13 A photo of Varian’s high definition multileaf collimator only a limited ability spare normal tissues in the center of a
(HD-MLC) (Varian Medical System, Inc. All rights reserved.) concave volume.
Dynamic Conformal Arcs nation produce a highly conformal dose distribution. For both
external beam radiation therapy and SRS, IMRT improves
In dynamic conformal arc delivery, the shape of the field dose conformity as compared with the use of conventional
changes continuously to conform to the BEV of the target unmodulated beams [63]. It should be noted that for small
during arced beam delivery. Dynamic conformal arc delivery intracranial targets, IMRT delivered with an MLC may be
combines the dosimetric advantages of arcs (reduced hot- unable to achieve the dose conformality obtained using sphere
streaks of dose through the patient) with the dose conformity packing due the fact the width of the leaves of the MLC are
that is possible with MLC beam shaping. Additionally, this on the order of the size of the targets to be treated.
approach only requires a single isocenter. With the use of The IMRT planning problem is modeled by subdividing
three or four non-coplanar arcs, a uniform high-dose volume each beam into a grid of beamlets. The weight, or intensity,
can be created that conforms to the target. of each of the beamlets is then determined. Because of the
Treatment planning for dynamic conformal arcs requires complexity of determining the appropriate beamlet weights
the delineation of the target and critical structures. The planner inverse (automated) planning techniques are employed.
must determine the number of arcs, their length, and their IMRT planning for SRS and IMRT planning for external
arrangements. For planning and delivery control, each arc is beam radiation therapy share the same general procedures.
approximated as a series of fixed fields. The shapes are The user defines a series of treatment goals, and an optimiza-
typically set based on the beam’s eye view (BEV) of the target tion algorithm determines the plan parameters that lead to a
and critical structures. For example, one can set each field plan that best satisfies those goals.
shape contained within an arc to match the BEV of the target
plus a 3 mm margin. Additionally, one may choose to use the
BEV of a critical structure to design field shapes that block Volumetric Modulated Arc Therapy
that structure throughout the arc path. If the gantry of the
accelerator rotates with a constant speed and maintains a con- Volumetric modulated arc therapy (VMAT) also known as
stant dose rate during rotation, one must assign the same num- intensity-modulated arc therapy (IMAT) is a rotational
ber of monitor units to each beam angle within a given arc. approach to the delivery of IMRT making use of a conven-
The dose calculation for conformal arcs is more compli- tional linear accelerator and a conventional MLC. The gantry
cated than that for conventional arcs due to fact that the field speed, dose rate, and MLC leaf positions are all varied
shapes changes while the beam is on. The planning system dynamically while the gantry rotates about the patient. The
must calculate the dose contributions from a large number of degree of intensity modulation is determined by the number
irregularly shaped fields. In determining the angular spacing of arcs as well as the number of aperture shapes included
of the fixed fields, one must balance the need for accuracy within each arc. The ability to modulate the intensity of radi-
with the desire for a reasonably quick dose calculation time. ation can provide significant dosimetric advantages relative
When an arc is approximated with fields spaced more than 5° to dynamic conformal arc therapy.
apart from one another, the lack of sufficient field overlap In 2007, Varian and Elekta each introduced delivery con-
results in undulating features in the lower isodose lines away trol systems capable of delivering VMAT. Since that time,
from the focal region. These features are not reflected in the VMAT has become a widely adopted delivery technique. A
actual delivery. Finer spacing of the beams will reduce such key advantage of VMAT is the highly efficient nature of the
artifacts in the displayed isodose lines. delivery. This is particularly beneficial when delivering the
In some treatment-planning system, a long arc can also be large doses utilized in radiosurgery treatments. The effi-
broken into multiple sub-arcs. The weights of these sub-arcs ciency can be further increased when VMAT is combined
can be optimized based on the user-defined dose-volume con- with flattening filter-free delivery, a technique where the
straints [62]. During delivery, each of these sub-arcs is treated beam is delivered without a flattening filter thereby permit-
as a separate beam. Generally, this technique works best on ting very high-dose rates.
convex targets and less well on targets with concave surfaces.
CyberKnife
Fixed-Field Intensity-Modulated Fields
Overview
With fixed-field IMRT, a modulated intensity pattern is deliv-
ered from each beam direction. Consequently, radiation can The CyberKnife (see Fig. 6.14) is a fully integrated radio-
be delivered to the target through preferred locations within surgery system that uses a pair of kilovoltage imaging
each beam. From each beam direction, the dose delivered to devices to locate the treatment site and direct the external
the target is nonuniform. However, all of the beams in combi- treatment beam [64]. The treatment beam is provided by a
Fig. 6.14 A CyberKnife (Figure courtesy of Accuray, Inc.)
linear accelerator (linac) mounted to a robotic arm that is Beam Characteristics

capable of maneuvering and pointing the linac with nearly
complete freedom within the treatment workspace (with the The treatment beam is provided by a 6 MV X-band linear
exception of the angles blocked due to the interposing of the accelerator that does not employ a flattening filter. The beam
linac between the diagnostic X-ray source and the amor- is collimated to a circular cross section by one of a set of
phous silicon imaging panels located in the floor). twelve interchangeable collimators. At a source-to-surface
During treatment, the imaging system repeatedly acquires distance (SSD) of 80 cm, these collimators provide a beam
planar images using a pair of amorphous silicon imagers and diameter that ranges from 5 to 60 mm. The SSD of the beam
compares the coordinates to the structures to be tracked can be varied from 60 to 100 cm. A continuous range of
(either implanted fiducial markers or visible anatomical beam diameters can be achieved by varying the collimator
structures) with those in the reference DRRs that are size and SSD. In 2008, Accuray introduced the Iris variable
generated from the planning CT images. After a rapid image aperture collimator. The Iris incorporates two stacked banks
comparison process (taking a few milliseconds), the updated of six tungsten segments that in combination create a
coordinates of the beams are supplied to the robot and the 12-sided variable aperture. The variable aperture automati-
position of the robot is updated to compensate for the any cally replicates the sizes of the 12 fixed collimators. The key
movement of the target. This enables the system to maintain benefit of the Iris aperture is that it reduces treatment time by
alignment of the beam with the target even if the patient facilitating the efficient combination of varying beam sizes
moves [11, 65]. within a single delivery.
The maneuverability of the beam, unconstrained by an
isocenter, combined with the capacity to adjust alignment
during treatment, endows the CyberKnife with unique dose Treatment Sites, Planning Scenarios,
delivery capabilities. These capabilities enable (1) the deliv- and Imaging Requirements
ery of highly conformal dose distributions to irregular target
volumes, (2) the delivery of fractionated SRT treatments, and All treatment planning for the CyberKnife is based on a CT
(3) the treatment of extracranial sites that are not amenable to study used alone or in conjunction with supplemental diag-
localization and/or fixation using conventional stereotactic nostic imaging such as MRI. As with most radiosurgical and
frames. They are exploited in a treatment-planning system SRT systems, the CT study is utilized for both target delinea-
that has been designed specifically for the CyberKnife. tion and dose calculation. However, with the CyberKnife, the
Fig. 6.15 Optimized plan for a spinal lesion
CT also plays a critical role in the image guidance aspect of places additional demands on its spatial resolution. To
the delivery process. If MRI images are to be used to enhance achieve optimal targeting accuracy [66], it is recommended
anatomical delineation, then the MRI study must be fused to that the CT slice thickness not exceed 1.25 mm [67]. While
the CT study prior to beginning the planning process. this slice thickness is commonly used for intracranial lesions,
The integrated image guidance system allows the it is a higher spatial resolution than is typically used for rou-
CyberKnife to target any treatment site that can be located tine diagnostic studies of extracranial sites.
via radiographic landmarks. Targeting is achieved by regis- CNS lesions, including tumors and AVMs, are the most
tering the landmarks in the kilovoltage (kV) alignment commonly treated sites. As examples of CNS applications,
images with their counterparts in DRRs derived from the we take note of two specialized treatment-planning prob-
treatment-planning CT study. lems: trigeminal neuralgia and spinal lesions. The trigeminal
For treatment sites that maintain a rigid relationship with nerve is visualized for planning using CT cisternography
bony landmarks that can be easily visualized on planar [68]. The patient is scanned over the full head in the
images (such as intracranial lesions), target localization can Trendelenburg position. Typically, 64 cGy is prescribed to
be achieved by measuring the position of the skeletal fea- the 80 % isodose line, encompassing approximately 8 mm of
tures in the treatment room. For treatment sites in soft tissue, nerve, and is delivered in one fraction. Beam alignment and
a common localization strategy is to make use of fiducial tracking during treatment is based on the position of the cra-
markers implanted near the lesion in a manner similar to nium visualized in the treatment room radiographs.
those being employed by other extracranial systems. Accuray Treatments of lesions along the spine are delivered in 1–5
has also introduced the Xsight Spine tracking system that fractions. Targeting is based on typically based on automated
uses advanced image-registration tools to automatically tracking of skeletal structures. The planning CT study (see
track lesions or targets in or near spinal structures. Accuray’s Fig. 6.15) is set up to visualize the region in and around the
Xsight Lung tracking system automatically compensates for lesion with the patient lying supine in an alpha cradle. The
the tumor motion without the need for implanted fiducials. typical treatment dose is 1,200–2,000 cGy prescribed to the
These targeting capabilities enable the user of the 80 % isodose line. Dose to the cord is limited to 800 cGy.
CyberKnife to plan treatments for central nervous system Experience from numerous treatments has shown that
(CNS) lesions in the cranium and anywhere along the spine, patients resting supine in an alpha cradle move less than
as well as soft tissue sites in the thorax, abdomen, and pelvis. 3 mm over the course of a 30 min treatment fraction [9]. The
Although the general treatment-planning procedure is not image guidance system detects and corrects for intra-fraction
site specific, each site introduces some distinctive elements movement.
to the planning process. In addition to CNS sites, the CyberKnife has been used
The important role that the planning CT study plays in to treat pancreatic, lung, prostate, and other soft tissue
target localization and beam alignment during treatment tumors.
The Planning Process semispherical lesions, the non-isocentric technique is most

commonly employed due to its greater flexibility.
The CyberKnife treatment-planning process combines a The third step is to choose either forward or inverse plan-
variety of user-managed and automatic planning operations ning. Non-isocentric plans are difficult to design using for-
to arrive at satisfactory deliverable plans. The planning sys- ward planning and so inverse planning is typically used. It
tem uses a CT study as the primary planning resource. The will be assumed from this point on that the clinician is inter-
dose calculation process makes a coarse accommodation of ested in inverse planning.
anatomical inhomogeneities by distinguishing air, tissue, One or more collimators can be employed in a plan.
and bone based on the Hounsfield number. Typically, for isocentric plans, the collimator diameter is
The robotic manipulator has a great deal of flexibility in matched to the diameter of the lesion. In non-isocentric
positioning and aiming the treatment beam. This makes it delivery, the clinician has more freedom in selecting the col-
possible to deliver both isocentric and non-isocentric dose limator diameter(s). Typically, the collimator diameter is
distributions. To utilize the full benefit of this flexibility and chosen to be approximately 60 % of the lesion’s smallest
at the same time make planning calculations tractable, the projected cross section. This makes it possible to obtain
planning system works with a discrete set of linac positions acceptable dose conformality, dose homogeneity, and dose
(called nodes) and a discrete set of pointing directions at falloff at the boundary of the target without unduly lengthen-
each node analogous to a multitude of isocenter-gantry-table ing the treatment time. As mentioned above, the beam diam-
positions in routine radiosurgery planning. During the deliv- eter can be fine-tuned by adjusting the SSD away from its
ery process the robotic manipulator moves the linac from nominal distance of 80 cm. The Iris variable aperture colli-
one node to another. At each node, the robot stops, aims the mator has simplified the use of multiple collimator diameters
linac in the selected directions, and delivers dose increments within a single plan.
in a “stop and shoot” sequence. The beam is not on while the Inverse planning works within a set of constraints to find
robot is in motion. the best deliverable plan. The clinician specifies a minimum
A typical treatment plan has at its disposal a set of up to target dose and a maximum dose to each critical structure
110 nodes distributed approximately uniformly over about that can potentially be transited by a beam. Additionally the
one half of a sphere centered on the treatment site. For each planner can create artificial dose constraint structures to fur-
plan, these nodes are selected from a total of more than 300 ther influence the plan. These structures can be used to atten-
possible linac positions. For non-isocentric delivery, the uate or block beams so as to shield critical structures. They
planning system defines up to 12 discrete pointing directions can also be used to modulate the dose reaching the lesion.
(vectors) at each node. These pointing vectors are aligned to The effective use of artificial constraints is a valuable
designated points within the treatment volume to produce a acquired skill.
set of overlapping beams designed to optimally cover the tar- In the contouring process, the clinician outlines the target
get volume while avoiding critical structures. The combina- volume, all relevant critical structures, and any artificial
tion of nodes and pointing vectors provides altogether a set structures needed to help manage the dose distribution. Part
of 1,320 “beams” from which to construct a plan. The plan is of this step involves identifying up to 12 locations within the
developed by selecting beams (i.e., nodes and pointing vec- target lesion that serve as endpoints for the pointing vectors
tors) from among the 1,320 available and assigning them at each node.
“weights” corresponding to the amount of dose each beam is After the contouring is complete, the planning system has
to deliver. up to 1,320 beams to choose from when optimizing the plan.
The planning process consists of a sequence of planning The inverse calculation then selects from among the avail-
choices by the clinician combined with a set of constraints able beams to get optimal target coverage and critical struc-
on the dose actually delivered by the plan. The first step is to ture avoidance. The optimizer then assigns beam weights
identify the treatment site. Because each anatomical location that meet the dose constraints. If constraints applied in an
is optimally treated by a particular configuration of nodes, inverse planning formulation are too strict, the system may
the planning system has predetermined groups of nodes that be unable to find a plan that simultaneously satisfies all of the
are considered the best choices for each treatment site. These constraints. Experienced CyberKnife planners have found it
groups of nodes are called “paths.” There are 80–110 nodes most effective to begin by defining minimum target doses
in each path, chosen from among 300+ possible linac posi- combined with fairly generous maximum doses to the criti-
tions. When the clinician designates the treatment site, the cal structures. If the system can find a solution for this first
planning system selects the appropriate path. approximation, the clinician then reduces the maximum
In the second step, the clinician chooses to construct doses to the critical structures until either the necessary con-
either a single-isocenter, a multi-isocenter, or a non-isocentric straints are met or a solution is no longer possible. If the
plan. While the isocentric option is reasonable for compact obtainable solutions cannot quite meet the original critical
structure constraints, then the clinician must decide if the dose with depth. As the beam passes through tissue, the dose
best obtainable solution is good enough. After an optimized initially remains approximately constant. Near the end of the
plan has been calculated, the clinician has an opportunity to range, however, the dose quickly increases to its peaks value
fine-tune it in a pseudo-forward planning step by selectively then rapidly falls off to near zero dose. The region of high dose
turning individual beams on and off. at the end of the particle range is called the Bragg peak [70].
The final decision made in the planning process is whether By adjusting the incident energy of the beam, the position of
the original fractionation scheme should be modified in light the Bragg peak can be controlled to match the depth of the
of the achievable dose limits to the target volume and critical lesions being treated. This is, however, a case of too much of a
structures. This is an important consideration for spine treat- good thing. The Bragg peak is usually much narrower than the
ments because there is incomplete knowledge of the spinal span of the average target. This in turn requires that the peak
cord’s tolerance of single-fraction doses. be “spread out” or modulated. The effect of this modulation is
an increase the surface dose of the entrance beam.
Proton delivery systems are not compact like conven-
Proton SRS tional linacs or Gamma Knife units (see Fig. 6.16). Until
recently, proton facilities that were originally designed for
Introduction nuclear research used retrofitted equipment to bring the radi-
ation to the patients. This meant using fixed beamlines and
In recent years, there has been a steady growth in the number innovative methods to deliver fields from various directions.
of proton therapy centers worldwide [69]. The upfront costs Modern proton facilities designed and built for radiation
of a new proton therapy center (frequently in excess of therapy have incorporated gantry systems making their use
$100 M) as well as the high operational costs have been a similar to linacs [71, 72].
limiting factor in the routine clinical adoption of proton ther-
apy. Of those centers offering proton therapy, fewer than half
have an active radiosurgery program. By comparison there Dose Modeling
are approximately 300 Gamma Knife centers worldwide [59]
and a much larger number of linac-based SRS programs. Doses from proton beams are clinically described in Cobalt
The major advantage of high-energy protons and other Gray Equivalent (CGE). The CGE represents the equivalent
heavy charged particles is the characteristic distribution of biological dose to achieve the same cell-kill as 60Co gamma
Fig. 6.16 (a) The STAR (Stereotactic Alignment Radiosurgery) iso- the top cranial hemisphere. (b) A 110-Ton isocentric gantry and 6-axis
centric patient positioning device designed and built by Product Genesis robotic patient positioner designed and built as a collaborative effort
Inc. [82] and used by the MGH (Massachusetts General Hospital, between General Atomic Inc. [83] and Ion Beam Application Inc. [84]
Boston) group in conjunction with horizontal fixed proton beamlines. and used for proton SRT and SRS at the NPTC (Northeast Proton
The patient can be rotated enabling any portal combination aimed from Therapy Center, MGH, Boston)
rays. The ratio of the physical to biological dose is called the the most from the reduction of dose to normal brain from
radiobiological effectiveness (RBE). Clinical applications of protons as compared with other modalities. Protons also
protons assume an RBE of 1.1 relative to 60Co [73, 74]. This make it possible to deliver a uniform target dose even with
means that 10 % less physical dose is required when deliver- large lesions. This may be considered an advantage when
ing proton beams compared to X-rays. treating AVM where the target and normal brain are inter-
The Bragg peak measured for a single energy beam is twined. However, additional clinical results are needed to
referred to as a “pristine” Bragg peak. The width of a pristine demonstrate this conclusively.
Bragg peak as measured from the proximal to distal 90 % The immobilization hardware and alignment techniques
dose varies with energy. The width is narrower for shallow used for linac SRS can be applied to proton radiosurgery
fields and broader for deeper fields; however, a typical width particularly when a gantry-based beamline is used. In fixed
is around 0.5 cm. In order to treat larger lesions, multiple beamlines, the need to rotate the patient can lead to torque
pristine Bragg peaks are combined to create a spread out on the patient’s head that can result in slight shifts of the
Bragg peak (SOBP). The SOPB 90–90 % width (also known head relative to an external reference frame. In such cases, it
as the modulation width) can be customized to the thickness would be unwise to rely on conventional stereotactic exter-
of any lesion by varying the depth and weight of each pris- nal reference frames. The insertion of three 1/16″ diameter
tine peak. 316LVM grade stainless steel spheres in the outer table of
It is important to note that the relative weights and depth the patient’s skull used in conjunction with diagnostic qual-
pullbacks of the individual pristine peaks needed to generate ity X-ray treatment alignment system provides a reliable
a flat SOBP are dependent not only on the 90–90 % width reference coordinate system [78]. A benefit of using an
but also on their entire depth dose profile. The pristine peak internal marker system is the flexibility to image, plan, and
shape is also affected by the field diameter. Therefore, for a treat on different days. This is an important consideration
specific beamline, both the energy and the field diameter for proton facilities that have significant overhead due the
must be modeled in the treatment-planning algorithm to need to fabricate custom beam shaping devices, which
properly generate the desired SOBP. An SOPB can be gener- require pretreatment QA/QC and possibly dosimetric verifi-
ated by delivering the dose from each pristine Bragg peak cation. Patients also benefit from not having to wear a ste-
separately or by using beam spreading devices such as spin- reotactic frame while waiting for their treatments as is
ning absorber wheels or ridge filters [75]. necessary with same day CT-Plan-Treat schedules. Using
The lateral dose penumbra of proton beams depends on internal alignment markers does not necessarily preclude
the inherent beam source geometry as well as individual the use of stereotactic bony fixation for CT imaging and/or
treatment field settings. Beamlines designed to treat large the treatment. It does however provide the flexibility to use
fields generally use a double scattering system which can stable noninvasive immobilization devices as assessed on an
results in a 5 cm FWHM source size. When this source size individual patient basis.
is combined with a source-to-axis distance (SAD) of 220 cm Treatment planning for proton therapy requires CT
that is typical for a proton gantry system, the result is a imaging because the dose algorithms convert the pixel den-
80–20 % isocenter plane penumbra of 2.8 mm for a range of sities to proton stopping power. The stopping powers are
8 cm and 6.3 mm for a range of 16 cm. In a small-field single used to calculate the penetration of the protons inside the
scattering system with an SAD of 450 cm, the source size is patient. When required, secondary imaging studies such as
around 1 cm FWHM, and the penumbra is reduced to 2.3 mm PET, MRI and angiography can be fused or merged with
for a range of 8 cm and 5.3 mm for a range of 16 cm. the primary planning CT as is done with conventional treat-
Treatment-planning systems with integrated proton beam ments [79, 80]. Because of the proton penetration conver-
algorithms, such as Elekta’s XiO planning system, must be sion from CT densities and the sharp dose fall off beyond
able to model both the lateral and depth dose profiles based the Bragg peak special precautions are necessary to ensure
on as many clinical variables as possible. This is achieved by that the planning CT’s densities are not artificially altered.
using pencil beam algorithms which model multiple This means that if a CT requires contrast for target and nor-
Coulomb scattering within beam modifying devices as well mal structure delineation it may be necessary to first obtain
as the patient [76, 77]. a non-contrast CT to be used for treatment planning. A sec-
ondary scan performed with contrast is used for structure
delineation and fused to the non-contrast planning CT. This
Pre-planning Consideration is especially important when treating vascular lesions,
where using the contrast CT scan to calculate the proton
As compared with linac and Gamma Knife delivery, protons stopping powers would overestimate the required proton
result in a lower integral dose to the patient due to the sharp range and modulation and result in an increased treatment
dose fall off beyond the Bragg peak. Larger lesions benefit volume.
Fig. 6.17 A custom brass Photon & Proton Depth Doses

aperture and Lucite compensator 190 MeV Photons
used to ensure a conformal 100 170 MeV Photons
proton beam
150 MeV Photons
Relative Dose (%)

75 130 MeV Photons
6 MeV Photons
50
25
0
0 5 10 15 20 25
Depth in Water (cm)
Clinical Treatment Planning dose uniformity across targets, it has been the MGH practice,
for proton SRS, to avoid truncating a target so as to treat it
Proton arc therapy is not feasible nor is it necessary to gen- using abutting or patch fields [81]. These techniques are reg-
erate conformal plans. Three to five static ports are gener- ularly used when treating fractionated proton SRT to improve
ally sufficient to obtain satisfactory target dose conformity. dose conformality of very irregular targets; however they
Using BEV (beams-eye-view) projections, the user deter- create small local high-dose points within the treatment
mines the best combination of ports to optimize the dose volume.
conformality. When practical, the use of orthogonal beams Patient demographics are dependent on the practice refer-
minimizes overlap regions between fields. Brass apertures ral patterns as well as the facility’s ability to offer alternative
are custom milled to the shape of the lesion’s BEV projec- treatments. Proton facilities that do not have an alternative
tion (see Fig. 6.17). The expansion of the aperture openings modality option are likely to have a significant fraction
accounts for the penumbra of the specific portal. Blocking (~25 %) of patients with metastatic disease compared to very
of specific critical structures is made easy using the BEV few if an alternative such as linac SRS is available. At the
approach. Custom lucite range compensators are fabricated MGH, 47 % of PSRS patients are AVMs, 19 % pituitary or
to enable dose shaping to conform to the distal edge of the cavernous sinus lesions, 13 % acoustic neuromas, 11 %
target. meningiomas, 4 % extracranial, and 6 % are miscellaneous
Although the beam directions are determined on a case- other disease.
by-case basis, an effort is made to use standard beam arrange-
ments. Pituitary adenomas are located in and around sella
and have similar dose constraints. A standard approach at the Other Considerations
Massachusetts General Hospital (MGH) for pituitary lesions
has been to use four fields (RL, LL, ASO, PSO) (Fig 6.18). When considering the overall advantage of protons for SRS,
The lateral fields avoid the optic structures and brainstem one should compare the dose gradient relative to standard
while passing through the temporal lobes; the PSO avoids Gamma Knife or linac-based systems. As with any such
the optic structures and temporal lobes while passing through treatment device there exist unique planning situations at
the brainstem; and the reduced weight ASO avoids the tem- which it will excel. However when one considers that
poral lobes and brainstem while passing through the optic Gamma Knife or linac-based system routinely produce plans
structures. that expose as little as 5–10 % of normal nontarget tissues in
Another standardized approach is used to treat acoustic the prescription isodose shell while providing dose gradients
neuromas using 3–4 fields (L/RPO, L/RAO, L/RSO, that drop the dose from the prescription dose to half of the
L/RPSO). This field combination limits the dose to the brain- prescription dose in 2.5–4 mm, the margin for improvement
stem, cerebellum, and temporal lobe while avoiding bone is relatively small. The image guidance techniques available
heterogeneities as much as possible. In other cases such as with linac-based systems such as cone-beam CT could also
AVMs, the variability in location, shape, and size makes impact the degree of the advantage seen with proton therapy
standardization difficult (see Fig. 6.19). In order to maximize for SRS and SBRT.
Summary
The technological advances in recent years have created a new

array of options for SRS. The dosimetric advantage of SRS
has also been gradually extended to extracranial sites. The
introduction of mMLCs, IMRT, and inverse planning into SRS
gives us more control in shaping the high-dose volume to con-
form to the target. Online or real-time image guidance also
provides a new alternative to the use of traditional stereotactic
frames. With these advances, fractionated treatments can now
be reliably administered. Although the dosimetric goals and
basic principles of SRS planning have not changed, the means
of achieving the dosimetric goals has shifted more towards
automation and computer optimization.
Glossary
Conformity index Defined by the RTOG as the volume of

the prescription isodose divided by the target volume.
Constraint A condition that must be satisfied during inverse
treatment planning for a plan to be considered feasible.
The most basic constraint is that all beam weights must
be nonnegative.
Dose-volume histogram (DVH) A plot of the volume ver-
sus dose used to analyze the dose distribution on a struc-
ture by structure basis.
Forward treatment planning An iterative approach to
planning where the user manually changes each of the
plan parameters until an acceptable plan is obtained.
Homogeneity index Defined by the RTOG as the maximum
dose in the treatment volume divided by the prescription
dose.
Inverse treatment planning An automated approach to
planning where the user defines the treatment goals and
an optimization algorithm is run that determines the
parameters that best meet the goals.
Micromultileaf collimator (mMLC) A device attached to
or incorporated into the head of a linear accelerator used
to define field shapes. As compared with a conventional
multileaf collimator (MLC), the leaves of an mMLC
are less wide and project to 5 mm or less in width at the
Fig. 6.18 A standard 4-field pituitary beam combination used for isocenter.
PSRS. The first image shows a 3D view of the ports and the structures
of interest. Typical doses are 18–20 CGE @ 90 % to the target with Objective function A scoring function that reduces the
constraints of 8 and 12 CGE to the surface of the optic structures and entire treatment plan into a single numerical value that is
brainstem, respectively to be either minimized or maximized.
Fig. 6.19 6-field PSRS b

radiosurgery isodose plan for a
9.6 cc right thalamic AVM in a 100 Dose
17-year-old patient. The
treatment was delivered in two
sessions, separated by 2 weeks,
each delivering 8 CGE @ 90 % 80
for a total of 16 CGE @ 90 %.
Three fields (LL, RPO, ASO)
were delivered on the first session a
and three different fields (RL, 60
LAO, LSO) were delivered on 90
the second session
Relative Dose (%)
40
45
Ø 10– 48 mm
20 20
Depth in Water (cm) 13 16
13. Meeks SL, Bova FJ, Wagner TH, Buatti JM, Friedman WA, Foote
References KD. Image localization for frameless stereotactic radiotherapy. Int
J Radiat Oncol Biol Phys. 2000;46(5):1291–9.
1. Ganz JC. Gamma knife surgery. 2nd ed. New York: Springer; 1997. 14. Buatti JM, Bova FJ, Friedman WA, Meeks SL, Marcus Jr RB,
2. Leksell L. Stereotactic radiosurgery. J Neurol Neurosurg Psychiatry. Mickle JP, et al. Preliminary experience with frameless stereotactic
1983;46(9):797–803. radiotherapy. Int J Radiat Oncol Biol Phys. 1998;42(3):591–9.
3. Leksell L. Cerebral radiosurgery. I. Gammathalanotomy in two 15. Gehring MA, Mackie TR, Kubsad SS, Paliwal BR, Mehta MP,
cases of intractable pain. Acta Chir Scand. 1968;134(8):585–95. Kinsella TJ. A three-dimensional volume visualization package
4. Leksell L. The stereotaxic method and radiosurgery of the brain. applied to stereotactic radiosurgery treatment planning. Int J Radiat
Acta Chir Scand. 1951;102(4):316–9. Oncol Biol Phys. 1991;21(2):491–500.
5. Wagner TH. Optimal delivery techniques for intracranial 16. Friedman WA, Bova FJ. The University of Florida radiosurgery
stereotactic radiosurgery using circular and multileaf collimators. system. Surg Neurol. 1989;32(5):334–42.
Gainesville, FL: University of Florida; 2000. 17. Coy SR, Houdek PV. Radiosurgery treatment planning. Semin
6. Takeuchi H, Yoshida M, Kubota T, Ishii H, Sato K, Handa Y, et al. Radiat Oncol. 1995;5(3):213–9.
Frameless stereotactic radiosurgery with mobile CT, mask immobi- 18. Ferris MC, Lim J, Shepard DM. Radiosurgery treatment planning
lization and micro-multileaf collimators. Minim Invasive via nonlinear programming. Ann Oper Res. 2003;119:247–60.
Neurosurg. 2003;46(2):82–5. 19. Ferris MC, Shepard DM. Optimization of gamma knife radiosur-
7. Ryken TC, Meeks SL, Pennington EC, Hitchon P, Traynelis V, gery. In: Du DZ, Pardolas P, Wang J, editors. Discrete mathematical
Mayr NA, et al. Initial clinical experience with frameless stereotac- problems with medical applications. Providence, RI: AMS; 2000.
tic radiosurgery: analysis of accuracy and feasibility. Int J Radiat 20. Flickinger JC, Lunsford LD, Wu A, Maitz AH, Kalend
Oncol Biol Phys. 2001;51(4):1152–8. AM. Treatment planning for gamma knife radiosurgery with multi-
8. Murphy MJ, Cox RS. The accuracy of dose localization for an ple isocenters. Int J Radiat Oncol Biol Phys. 1990;18(6):1495–501.
image-guided frameless radiosurgery system. Med Phys. 1996; 21. Luo L, Shu H, Yu W, Yan Y, Bao X, Fu Y. Optimizing computerized
23(12):2043–9. treatment planning for the Gamma Knife by source culling. Int J
9. Murphy MJ, Chang SD, Gibbs IC, Le QT, Hai J, Kim D, et al. Radiat Oncol Biol Phys. 1999;45(5):1339–46.
Patterns of patient movement during frameless image-guided radio- 22. Shepard DM, Ferris MC, Ove R, Ma L. Inverse treatment planning
surgery. Int J Radiat Oncol Biol Phys. 2003;55(5):1400–8. for Gamma Knife radiosurgery. Med Phys. 2000;27(12):2748–56.
10. Kamath R, Ryken TC, Meeks SL, Pennington EC, Ritchie J, Buatti 23. Zhang P, Wu J, Dean D, Xing L, Xue J, Maciunas R, et al. Plug pat-
JM. Initial clinical experience with frameless radiosurgery for tern optimization for gamma knife radiosurgery treatment planning.
patients with intracranial metastases. Int J Radiat Oncol Biol Phys. Int J Radiat Oncol Biol Phys. 2003;55(2):420–7.
2005;61(5):1467–72. 24. Zhang P, Dean D, Metzger A, Sibata C. Optimization of Gamma
11. Gerszten PC, Ozhasoglu C, Burton SA, Vogel W, Atkins B, knife treatment planning via guided evolutionary simulated anneal-
Kalnicki S, et al. Evaluation of CyberKnife frameless real-time ing. Med Phys. 2001;28(8):1746–52.
image-guided stereotactic radiosurgery for spinal lesions. Stereotact 25. Kubo HD, Pappas CT, Wilder RB. A comparison of arc-based and
Funct Neurosurg. 2003;81(1–4):84–9. static mini-multileaf collimator-based radiosurgery treatment plans.
12. Bova FJ, Buatti JM, Friedman WA, Mendenhall WM, Yang CC, Liu Radiother Oncol. 1997;45(1):89–93.
C. The University of Florida frameless high-precision stereotactic 26. Podgorsak EB, Pike GB, Olivier A, Pla M, Souhami L. Radiosurgery
radiotherapy system. Int J Radiat Oncol Biol Phys. 1997;38(4): with high energy photon beams: a comparison among techniques.
875–82. Int J Radiat Oncol Biol Phys. 1989;16(3):857–65.
27. Alheit H, Dornfeld S, Dawel M, Alheit M, Henzel B, Steckler K, 45. Kubsad SS, Mackie TR, Gehring MA, Misisco DJ, Paliwal BR,
et al. Patient position reproducibility in fractionated stereotactically Mehta MP, et al. Monte Carlo and convolution dosimetry for stereo-
guided conformal radiotherapy using the BrainLab mask system. tactic radiosurgery. Int J Radiat Oncol Biol Phys. 1990;19(4):
Strahlenther Onkol. 2001;177(5):264–8. 1027–35.
28. Bourland JD, McCollough KP. Static field conformal stereotactic 46. Dong L, Shiu A, Tung S, Hogstrom K. A pencil-beam photon dose
radiosurgery: physical techniques. Int J Radiat Oncol Biol Phys. algorithm for stereotactic radiosurgery using a miniature multileaf
1994;28(2):471–9. collimator. Med Phys. 1998;25(6):841–50.
29. Cardinale RM, Benedict SH, Wu Q, Zwicker RD, Gaballa HE, 47. Bardash M, Amols HI, Kohn S, Martel MK, Wuu CS, Sisti M, et al.
Mohan R. A comparison of three stereotactic radiotherapy tech- Rapid dose calculations for stereotactic radiosurgery. Med Phys.
niques; ARCS vs. noncoplanar fixed fields vs. intensity modulation. 1992;19(4):965–70.
Int J Radiat Oncol Biol Phys. 1998;42(2):431–6. 48. Ayyangar KM, Jiang SB. Do we need Monte Carlo treatment plan-
30. Grebe G, Pfaender M, Roll M, Luedemann L, Wurm RE. Dynamic ning for linac based radiosurgery? A case study. Med Dosim.
arc radiosurgery and radiotherapy: commissioning and verification 1998;23(3):161–8.
of dose distributions. Int J Radiat Oncol Biol Phys. 2001;49(5): 49. Verellen D, Linthout N, Bel A, Soete G, van den Berge D, D’Haens
1451–60. J, et al. Assessment of the uncertainties in dose delivery of a com-
31. Hamilton RJ, Kuchnir FT, Sweeney P, Rubin SJ, Dujovny M, mercial system for linac-based stereotactic radiosurgery. Int J
Pelizzari CA, et al. Comparison of static conformal field with mul- Radiat Oncol Biol Phys. 1999;44(2):421–33.
tiple noncoplanar arc techniques for stereotactic radiosurgery or 50. Verhey LJ, Smith V, Serago CF. Comparison of radiosurgery treat-
stereotactic radiotherapy. Int J Radiat Oncol Biol Phys. ment modalities based on physical dose distributions. Int J Radiat
1995;33(5):1221–8. Oncol Biol Phys. 1998;40(2):497–505.
32. Laing RW, Bentley RE, Nahum AE, Warrington AP, Brada 51. Smith V, Verhey L, Serago CF. Comparison of radiosurgery treat-
M. Stereotactic radiotherapy of irregular targets: a comparison ment modalities based on complication and control probabilities.
between static conformal beams and non-coplanar arcs. Radiother Int J Radiat Oncol Biol Phys. 1998;40(2):507–13.
Oncol. 1993;28(3):241–6. 52. Shaw E, Kline R, Gillin M, Souhami L, Hirschfeld A, Dinapoli R,
33. Shiu AS, Kooy HM, Ewton JR, Tung SS, Wong J, Antes K, et al. et al. Radiation Therapy Oncology Group: radiosurgery quality
Comparison of miniature multileaf collimation (MMLC) with cir- assurance guidelines. Int J Radiat Oncol Biol Phys.
cular collimation for stereotactic treatment. Int J Radiat Oncol Biol 1993;27(5):1231–9.
Phys. 1997;37(3):679–88. 53. Paddick I. A simple scoring ratio to index the conformity of radio-
34. Soanes T, Hampshire A, Vaughan P, Brownett C, Rowe J, Radatz surgical treatment plans. Technical note. J Neurosurg. 2000;93
M, et al. The commissioning and quality assurance of the Automatic Suppl 3:219–22.
Positioning System on the Leksell gamma knife. J Neurosurg. 54. Nedzi LA, Kooy HM, Alexander III E, Svensson GK, Loeffler
2002;97(5 Suppl):574–8. JS. Dynamic field shaping for stereotactic radiosurgery: a modeling
35. Tsai JS, Engler MJ, Ling MN, Wu JK, Kramer B, Dipetrillo T, et al. study. Int J Radiat Oncol Biol Phys. 1993;25(5):859–69.
A non-invasive immobilization system and related quality assur- 55. Lomax NJ, Scheib SG. Quantifying the degree of conformity in
ance for dynamic intensity modulated radiation therapy of intracra- radiosurgery treatment planning. Int J Radiat Oncol Biol Phys.
nial and head and neck disease. Int J Radiat Oncol Biol Phys. 2003;55(5):1409–19.
1999;43(2):455–67. 56. Leung LH, Chua DT, Wu PM. A new tool for dose conformity eval-
36. Tome WA, Meeks SL, Buatti JM, Bova FJ, Friedman WA, Li Z. A uation of radiosurgery treatment plans. Int J Radiat Oncol Biol
high-precision system for conformal intracranial radiotherapy. Int J Phys. 1999;45(1):233–41.
Radiat Oncol Biol Phys. 2000;47(4):1137–43. 57. Knoos T, Kristensen I, Nilsson P. Volumetric and dosimetric evalu-
37. Urie MM, Lo YC, Litofsky S, FitzGerald TJ. Miniature multileaf ation of radiation treatment plans: radiation conformity index. Int J
collimator as an alternative to traditional circular collimators for Radiat Oncol Biol Phys. 1998;42(5):1169–76.
stereotactic radiosurgery and stereotactic radiotherapy. Stereotact 58. Borden JA, Mahajan A, Tsai JS. A quality factor to compare the
Funct Neurosurg. 2001;76(1):47–62. dosimetry of gamma knife radiosurgery and intensity-modulated
38. Haedinger U, Krieger T, Flentje M, Wulf J. Influence of calculation radiation therapy quantitatively as a function of target volume and
model on dose distribution in stereotactic radiotherapy for pulmo- shape. Technical note. J Neurosurg. 2000;93 Suppl 3:228–32.
nary targets. Int J Radiat Oncol Biol Phys. 2005;61(1):239–49. 59. Elekta AB.
39. Chaves A, Lopes MC, Alves CC, Oliveira C, Peralta L, Rodrigues 60. Lutz W, Winston KR, Maleki N. A system for stereotactic radiosur-
P, et al. A Monte Carlo multiple source model applied to radiosurgery with a linear accelerator. Int J Radiat Oncol Biol Phys.
gery narrow photon beams. Med Phys. 2004;31(8):2192–204. 1988;14(2):373–81.
40. Pike B, Peters TM, Podgorsak E, Pla C, Olivier A, de Lotbiniere 61. Cho PS, Kuterdem HG, Marks II RJ. A spherical dose model for
A. Stereotactic external beam calculations for radiosurgical treat- radiosurgery plan optimization. Phys Med Biol. 1998;43(10):
ment of brain lesions. Appl Neurophysiol. 1987;50(1–6):269–73. 3145–8.
41. Pike B, Podgorsak EB, Peters TM, Pla C. Dose distributions in 62. Ma L, Yu CX, Earl M, Holmes T, Sarfaraz M, Li XA, et al.
dynamic stereotactic radiosurgery. Med Phys. 1987;14(5):780–9. Optimized intensity-modulated arc therapy for prostate cancer
42. Wu A, Lindner G, Maitz AH, Kalend AM, Lunsford LD, Flickinger treatment. Int J Cancer. 2001;96(6):379–84.
JC, et al. Physics of gamma knife approach on convergent beams in 63. Kramer BA, Wazer DE, Engler MJ, Tsai JS, Ling MN. Dosimetric
stereotactic radiosurgery. Int J Radiat Oncol Biol Phys. 1990;18(4): comparison of stereotactic radiosurgery to intensity modulated
941–9. radiotherapy. Radiat Oncol Investig. 1998;6(1):18–25.
43. Wu X, Ting JY, Markoe AM, Landy HJ, Fiedler JA, Russell J. 64. Adler JR, Jr., Murphy MJ, Chang SD, Hancock SL. Image-guided
Stereotactic dose computation and plan optimization using the con- robotic radiosurgery. Neurosurgery. 1999;44(6):1299–306; discus-
volution theorem. I. Dose computation. Stereotact Funct Neurosurg. sion 1306–7.
1996;66 Suppl 1:302–8. 65. Chang SD, Murphy M, Geis P, Martin DP, Hancock SL, Doty JR,
44. Solberg TD, Holly FE, De Salles AA, Wallace RE, Smathers et al. Clinical experience with image-guided robotic radiosurgery
JB. Implications of tissue heterogeneity for radiosurgery in head (the Cyberknife) in the treatment of brain and spinal cord tumors.
and neck tumors. Int J Radiat Oncol Biol Phys. 1995;32(1):235–9. Neurol Med Chir (Tokyo). 1998;38(11):780–3.
66. Chang SD, Main W, Martin DP, Gibbs IC, Heilbrun MP. An analy- 75. Chu W, Ludewigt B, Renner T. Instrumentation for treatment of
sis of the accuracy of the CyberKnife: a robotic frameless cancer using proton and light-ion beams. Rev Sci Instrum.
stereotactic radiosurgical system. Neurosurgery. 2003;52(1): 1993;64(8):2055–122.
140–6; discussion 146–7. 76. Russell KR, Isacsson U, Saxner M, Ahnesjo A, Montelius A,
67. Heilbrun MP. Cyberknife radiosurgery: a practical guide. Grusell E, et al. Implementation of pencil kernel and depth penetra-
Sunnyvale, CA: The CyberKnife Society; 2003. tion algorithms for treatment planning of proton beams. Phys Med
68. Mawad ME, Silver AJ, Hilal SK, Ganti SR. Computed tomography Biol. 2000;45(1):9–27.
of the brain stem with intrathecal metrizamide. Part I: the normal 77. Hong L, Goitein M, Bucciolini M, Comiskey R, Gottschalk B,
brain stem. AJR Am J Roentgenol. 1983;140(3):553–63. Rosenthal S, et al. A pencil beam algorithm for proton dose calcula-
69. Sisterson J. Sponsored by the Particle Therapy Co-Operative Group tions. Phys Med Biol. 1996;41(8):1305–30.
(PTCOG). Particles Newsletter. 2005. p. 35. 78. Gall KP, Verhey LJ, Wagner M. Computer-assisted positioning of
70. Kahn FM. The physics of radiation therapy. 3rd ed. Philadelphia: radiotherapy patients using implanted radiopaque fiducials. Med
Lippincott Williams & Wilkins; 2003. Phys. 1993;20(4):1153–9.
71. Slater J, Miller D, Archambeau J. Development of a hospital based 79. Wells III WM, Viola P, Atsumi H, Nakajima S, Kikinis R. Multi-
Proton Therapy Center. Int J Radiat Oncol Biol Phys. 1988;14(4): modal volume registration by maximization of mutual information.
761–75. Med Image Anal. 1996;1(1):35–51.
72. Flanz J, Bailey J, Bradley S, Goitein M, Gottschalk B, Jongen Y, 80. van Herk M, Kooy HM. Automatic three-dimensional correlation
et al. Recent performance of the NPTC equipment compared with of CT-CT, CT-MRI, and CT-SPECT using chamfer matching. Med
the clinical specifications. In: Proceedings of 15th conference on Phys. 1994;21(7):1163–78.
the application of accelerators in research and industry. New York: 81. Bussiere MR, Adams JA. Treatment planning for conformal pro-
AIP Press; 1999. p. 971–74. ton radiation therapy. Technol Cancer Res Treat. 2003;2(5):
73. Raju MR. Proton radiobiology, radiosurgery and radiotherapy. Int J 389–99.
Radiat Biol. 1995;67(3):237–59. 82. Product Genesis LLC. [cited]. Available from: www.productgenesis.com
74. Paganetti H, Niemierko A, Ancukiewicz M, Gerweck LE, Goitein M, 83. General Atomics, Inc. [cited]. Available from: www.ga.com
Loeffler JS, et al. Relative biological effectiveness (RBE) values for pro- 84. Ion Beam Application S.A. [cited]. Available from: www.iba-
ton beam therapy. Int J Radiat Oncol Biol Phys. 2002;53(2):407–21. worldwide.com
Designing, Building, and Installing
a Stereotactic Radiosurgery Unit 7
Lijun Ma, Andrew Hwang, and Arjun Sahgal
In addition, basic aspects of shielding and radiation safety

Introduction are described, as well as installation, acceptance, and qual-
ity assurance testing.
Stereotactic radiosurgery (SRS) units are fundamentally
based on delivering a high dose of “ablative” ionizing radia-
tion to small targets within the brain localized in three dimen- Design Principles
sions. The treatments require an overall delivery precision in
the order of 1–2 mm and delivery of highly conformal radia- Common types of ionizing radiation used for SRS include
tion with steep dose gradients beyond the target edge. high-energy gamma rays (e.g., 60Co), high-energy X-rays,
The term brain stereotaxy simply refers to a method for and charged particles such as protons. Photon-based (X-rays
locating points within the brain using an external 3D frame and gamma rays) and ion-based (protons and heavy charged
of reference usually based on the Cartesian coordinate sys- particles) SRS units use fundamentally different design prin-
tem. Although the use of stereotactic instruments have been ciples. The basic principle of photon-based SRS units is to
reported in the literature as early as the 1800s [1], the first converge a large number of beams at the isocenter in order to
definitive brain SRS unit was designed and built by the produce rapid dose fall off beyond the focal area. The beams
Swedish neurosurgeon Dr. Lars Leksell in the 1950s [2]. are created using radioactive sources (60Co) or high-energy
Without modern 3D imaging like CT and MRI, Lars Leksell electron linear accelerators. In contrast, ion-based units tend
achieved stereotaxy by designing a rigid head fixation device to use a limited number of beams, or points, in order to
attached to the patient’s cranium. The frame incorporated the spread the Bragg peak across the target volume. The rapid
required Cartesian coordinate system and, therefore, the dose fall off is largely created at the distal end of the Bragg
ability to pinpoint any location within the frame space to peak [8–13]. Charged particles such as protons, helium ions,
guide the radiation beams [3–7]. and carbon ions are accelerated using high-energy cyclotrons
Since that time, a large number of different techniques or synchrocylotrons. In this chapter, we will focus on photon-
have been used to deliver SRS. These include dedicated based SRS units as these units are dominant in current clini-
SRS units such as the Gamma Knife and Cyberknife in cal practice for both intracranial and the current extrapolation
addition to linac-based systems. In this chapter, we provide to extracranial body sites. State of-the-art photon-based SRS
an overview of the design principles of various SRS units. units currently include the Gamma Knife Perfexion,
Tomotherapy, Cyberknife, and linac-based systems that
sprouted from the dedicated Novalis brain radiosurgery units
(for example, the Varian and Elekta brand systems).
L. Ma, Ph.D. (*) One of four basic design schemes for radiation delivery is
Department of Radiation Oncology, University of California generally applied in building a SRS unit:
San Francisco, San Francisco, CA, USA 1. Fix the source of radiation and fix the patient position
e-mail: lijunma@radonc.ucsf.edu
2. Move the source of radiation and fix the patient position
A. Hwang, Ph.D. 3. Move the source of radiation and move the patient position
Department of Radiation Oncology, Alta Bates Summit
Medical Center, Oakland, CA, USA 4. Fix the source of the radiation and move the patient
The first principle is applied in Gamma Knife and most
A. Sahgal, M.D.
Department of Radiation Oncology, University of Toronto, proton and heavy-ion units, the second principle is used in
Toronto, ON, Canada the majority of linac-based SRS units and the Cyberknife
96 L. Ma et al.
unit [14], the third principle is used in Tomotherapy [15] and

linac-based simultaneous gantry and couch rotating units
[16], and the fourth principle is uncommon but has been
reported in the rotating chair SRS unit [17].
When applying these principles, recent technological
developments have been intensely focused on designing not
only accurate but more efficient and safer SRS units in a
turnkey type of solution. For example, repeat patient posi-
tioning during beam-off periods can now be automated via
computer controls with minimal user interventions; a large
amount of radiation with three- to fourfolds of conventional
dose rate can be now delivered in a matter of minutes; and
notably, full integration of online imaging capability that
facilitates patient setup in full six degrees of freedom. Some
of these features are still work in progress but highlight an
expansion in the design paradigm of building an SRS unit.
With respect to clinical developments, SRS has surpassed its
role in treating a small target with a single fraction to treating
multiple targets simultaneously (even up to 20–30), large tar-
gets and hypofractionation.
Components and Functions
Gamma Knife
The Gamma Knife (GK) is a teletherapy 60Co unit specifi-

cally designed for intracranial SRS treatments [18–21]. The
latest GK unit is named the Perfexion (PFX) and contains
192 60Co sources fixed on the exterior surface of a conical
collimator. Each source delivers gamma rays of 1.25 MeV in
mean energy. The center of the collimator is the “isocenter.”
The initial activity of the individual source is approximately
Fig. 7.1 The original model U GK unit as installed in the USA in the
3.0 Ci. Therefore the total initial activity in each unit exceeds late 1980s. (a) The side view of the unit and couch; (b) the helmet col-
5,000 Ci. This activity typically produces a dose rate of limator with a head phantom docked onto the helmet via the attached
approximately 300 cGy/min at the isocenter. Because radio- stereotactic frame
active 60Co decays with a half-life of 5.3 years, the dose rate
will be reduced by 50 % after 5.3 years thus making refuel- collimators, a GK unit weighs about 20 tons. A heavy and
ing a necessary consideration for the system. robust supporting structure, precise beam collimation via
The source-to-isocenter distance for all GK units is on high-Z materials such as tungsten, and a short source-to-
the order of 40 cm. This distance is significantly shorter than focal distance are major contributing factors for the sub-
the 100 cm source-to-axis distance (SAD) of the standard millimeter mechanical accuracy of the GK unit.
linear accelerator. Short source-to-focal distance is a hall- Historically, there have been several models of GK
mark design characteristic of the GK unit. This not only [Models U, B (B1 & B2), and C, and now the PFX]. In the
improves the required total activity of the sources (inversely USA, the original GK Model U was installed at the University
square law), the geometric penumbra (linear divergence) but of Pittsburgh in 1987. An illustration of the Model U unit is
also facilitates high mechanical precision near the isocenter. given in Fig. 7.1. After the FDA approval of later models
For example, if a minor milling error occurs in the primary (Model B, C, and PFX), the Model U has been gradually
collimator, the projected penumbra would be significantly phased out of service. Major differences between the origi-
narrower at a short source-to-isocenter distance than at a nal U and the later B and C models include the couch driving
large source-to-isocenter distance. mechanism (hydraulic vs. electrical), patient positioning
The high level of activity of the GK unit demands heavy assembly (manual vs. automatic), collimator design, and
density materials for proper shielding. With shielding and source arrangement pattern. However, the key components
7 Designing, Building, and Installing a Stereotactic Radiosurgery Unit 97
Besides elimination of the tertiary helmet collimators with

a single universal collimator, the new PFX design also trans-
forms the automatic patient positioning system (APS) via the
introduction of a single integrated translational couch.
Therefore, independent patient positioning devices such as
the tertiary trunnion bars and add-on APS units have been
eliminated. The translational couch enables the patient’s head
frame to be docked onto the end of couch and then entire
patient body plus the head frame is shifted automatically to
align with the treatment delivery. The maximum tolerance
level for the accuracy of the translational couch is less than
0.3 mm, which is comparable to the previous models.
The elimination of the tertiary collimator has also signifi-
cantly extended the maximum traveling distance achievable
Fig. 7.2 The redesigned PFX unit as released by Elekta in 2006. Note for the PFX as compared to the previous models. For exam-
the absence of helmet collimator and the universal collimator design as
compared with the original Model U of Fig. 7.1
ple, the overall reachable volume via the automatic couch
has been extended more than 300 %. Now the majority of
intracranial lesions can be readily treated with the stereotac-
remained nearly identical for these GK models (models U, tic frame placed in the central position, this greatly eases the
B, and C), which include a shielded source head ball housing effort of frame placement.
201 60Co sources, a movable couch, four tertiary helmet col- This new design of the Gamma Knife PFX has resulted in a
limators (18, 14, 8, and 4 mm in beam aperture), removable simple turn-key solution for most treatment indications
plug collimators (201 for each helmet), a separate console including complex multiple metastatic brain lesion treatment
control unit, patient positioning devices, audio/video moni- with significant improvements in the range of locations in the
toring devices, and a dedicated treatment planning system. brain that can be treated, radiation safety, and patient comfort.
The patient positioning device consists of the Leksell frame In addition to the above hardware design improvements,
(pinned to the patient skull) and its attachments to the helmet the treatment planning system for the PFX has also been
of the GK. The frame is attached manually via either a pair updated from the previous models. The salient features of the
of bars (i.e., trunnions) or the automatic positioning system system as well as the 3D dose calculation algorithms are cov-
(APS) device. ered in another chapter.
In 2006, Elekta introduced a new GK unit called Perfexion
(PFX). The PFX (Fig. 7.2) transforms and significantly
improves upon the design of previous GK models. One dis- Linear Accelerator-Based SRS Units
tinct feature is the use of a single collimator where three
sizes of collimation apertures (16, 8, and 4 mm) are drilled Historically, linac-based SRS units were built upon standard
directly into a single piece of tungsten shaped like a speaker linear accelerators with add-on collimation systems and couch
cone. The radioactive 60Co sources are placed inside metal stabilization devices [16, 22–32]. The majority of these SRS
housings called sectors that slide linearly on the outer sur- units employed tertiary cones, or add-on micro- or mini-mul-
face of the collimator to align themselves with these pre- tileaf collimators, to deliver noncoplanar converging arc
drilled apertures, to create confocal radiation beams of 16, 8, beams at the isocenter. The arrangement of these noncoplanar
and 4 mm of approximate full width half maximum (FWHM). arc beams aims to deliver an elliptical dose distribution similar
Because the sectors are driven by independent stepper to that of a single shot in Gamma Knife delivery. In general,
motors, beams of different aperture sizes can be mixed from the tertiary cones are made from high-Z materials and are
different directions at the focus point. The total number of intended to extend the source to diaphragm distance to sharpen
configuration for variable beam apertures at the isocenter is the beam penumbra at the isocenter. Note that the source to
given as 48 − 1 = 65,535. This is because at each isocenter, a isocenter distance is typically 100 cm, which is more than two
total of four beam aperture sizes (0 mm or blocked, 4, 8, and times of the source-to-isocenter distance of the GK unit.
16 mm) can be directed from eight equally spaced directions, Unlike GK units where the isocenter is mechanically
i.e., 45° separation between two neighboring sectors sur- aligned with the beam focus, traditional linac-based SRS
rounding the universal collimator. The overall beam preci- units predominantly rely on wall-mount lasers and special
sion is dependent upon the precise drilling of >500 apertures localizer boxes attached to the stereotatic frame for defining
and the precise alignment of the sliding sectors. Therefore, the target coordinates. Therefore, the accuracy of the laser
manufacturing the universal collimator and sector housing system is critical in ensuring the treatment accuracy for
comprises a major cost in building the PFX unit. linac-based SRS units. As a general practice, a pretreatment
98 L. Ma et al.
Fig. 7.3 Example of isocenter test tools: (a) an automatic diode scanning tool; (b) a film holder with central prick pin for marking the film before
exposure with the radiation
isocenter alignment check must be carried out for each six degrees of freedom (x, y, z translation, and pitch, yaw,
patient delivery. and roll) to correct for non-flat couch top and minor setup
Figure 7.3 illustrates the isocenter alignment devices for errors. The coordinates on the localizer is commonly set in
the PFX unit. In using the isocenter alignment check tool, the positive numbers in order to minimize misreading in setting
beam profile is first scanned via small-volume detector such up the stereotactic coordinates.
as diode (Fig. 7.3a) across axis directions. A film is pricked With the recent advent of onboard imaging devices, the
with a pin at the isocenter set by the mechanical device isocentric process has been gradually replaced by automatic
(Fig. 7.3b). Then the film is exposed and pinprick mark is image-based setups. An example of such a new generation
compared with the center of the beam profile to determine linear accelerator is shown in Fig. 7.6. Some of the features
agreement between the mechanical and radiation isocenters of this type of linear accelerator include digital control
[5, 25, 30, 33, 34]. boards, an onboard imager (OBI), high dose rate delivery
For linac-based SRS units, the most commonly used iso- without flattening filter attenuation as used for conventional
center alignment check is the Winston–Lutz test method radiotherapy. With an improved digital control system, such
[28, 32]. For the test, a metallic spherical ball is first aligned a linear accelerator is able to deliver arc beams with a high
with the isocenter using wall-mount lasers or a front-pointer. dose rate while being able to adjust the speed of gantry rota-
A series of port films are then exposed downstream for a tion and MLC leaf motions for beam intensity modulation,
“dry-run” of actual beam deliveries. Typically, the film is as well as collecting 2D and 3D onboard kV images during
placed on a special holder attached to the gantry port extend- the process. As a result, patient treatment setup does not
ing underneath the ball. If the linac is equipped with high- need to be referenced to the isocenter setup and verifications,
resolution electronic portal imager (EPID) [35], the image but can solely rely on direct online image studies via OBI for
can be taken with that device in lieu of the film. Once direct planar or volumetric match of the patient position with
exposed, the projected ball image is compared with the edge that defined in the reference CT image studies during the
field to detect any deviations in the isocenter. The Winston– treatment planning process. However, the imaging device
Lutz test is especially useful for linac-based SRS because it must be carefully calibrated so that its reference center coin-
not only checks the isocenter alignment but also detects any cides with the linear accelerator isocenter, so that any shift in
collisions among the gantry, couch, and tertiary collimators patient position detected from the online imaging process
during the patient delivery. can be reliably compensated using changes in couch position
Example isocenter measurement results for a Gamma or beam alignment (i.e., MLC shifts).
Knife unit and a linac-based SRS unit are given in Fig. 7.4. Once the patient is aligned, linac-based SRS commonly
Once the alignment between the laser and mechanical iso- uses either noncoplanar converging fixed beams or dynamic
centers is verified, the target coordinates for the linac-based arc beams. In dynamic rotation delivery, the gantry rotates
SRS treatments are set via a laser-based localizer. A picture while the couch can be either fixed or rotates at the same
of the setup localizer is shown in Fig. 7.5. To facilitate the time. The advantage of the dynamic arc-beam delivery is that
alignment of the isocenter, the localizer can be adjusted in only a single treatment setup is required and parallel-opposed
Fig. 7.4 Results of radiation and mechanical alignment tests for (a) agreement between the pinpricked center and full width half maximum
GK unit and (b) linac-based SRS unit. (a) Used the smallest circular (FWHM) field edge was found to be within 0.5 mm for both cases
field of 4-mm in diameter and (b) used a square field of 4 × 4 cm2. The
Fig. 7.5 A typical linac-based SRS treatment setup and illustrated beam arrangements
beams can be avoided during the entire delivery to produce a weights and intensity modulated arc beam delivery (IMAT or
better entrance and exit dose for a treatment plan. VMAT), MLC-based SRS may also be able to deliver even
With the ubiquitous adoption of high-definition MLC sys- better dose distributions for complex targets as compared to
tems, fixed-cone collimator-based SRS delivery is fast being the conventional fixed-field approach [31]. A picture of a
replaced by MLC-based SRS delivery. From the beginning, dedicated MLC-based SRS linear accelerator is shown in
these high-resolution MLCs were able to shape the beam Fig. 7.7. The unit is considered as a stereotactic unit because
aperture conformally in the beam’s eye view during the gan- it possesses a high-definition MLC with a slightly restricted
try rotation for linac-based SRS delivery; therefore, the tech- maximum field size of 22 × 40 cm2, as compared to the
nique is termed shaped-beam SRS delivery in contrast to the 40 × 40 cm2 maximum field size typical of conventional lin-
traditional fixed-beam cone SRS delivery [27, 29, 31, ear accelerators. In addition, this particular unit also pos-
36–38]. Recent advances in MLC technology have resulted sesses the capacity to perform online near real-time image
in leaf size on the order of even 2.5 mm which allows for guidance via a ceiling and floor-mounted kV X-ray imaging
even small targets <1 cm to be treated. system. The mechanical accuracy of the unit has been
For the majority of treatments, MLC-based SRS is capa- reported to be on the order of 0.5 mm [39], and notably
ble of delivering the full treatment regardless of a single or treated small targets such as the trigeminal nerve root entry
multiple targets with a single isocenter. With optimized beam zone for the treatment of trigeminal neuralgia [38].
100 L. Ma et al.
Fig. 7.8 Illustration of the Cyberknife SRS unit with integrated kV

Fig. 7.6 A digitally controlled linear accelerator capable of high dose imaging system
rate stereotatic radiosurgery, hypofractionated body radiotherapy, and
conventional radiotherapy delivery as released by Varian Medical
Systems in 2009 Considering the principles of the design, Cyberknife is
a unique SRS system in that the majority of the treatment
beams are non-isocentric (Fig. 7.8). The largest circular
field size for the Cyberknife is 6 cm in diameter so the
system is by design a dedicated SRS unit. However like
dedicated linac-based systems, cross firing and fast over-
lapping of multiple beams of variable field sizes allow
the Cyberknife to treat larger target volumes such as the
prostate.
Unlike those isocentric systems previously discussed,
several unique factors contribute to the successful operation
of the Cyberknife [1]. The linear accelerator operates in the
X-band microwave frequency which allows for a signifi-
cantly shorter waveguide length and a compact design of the
linear accelerator head such that it can be mounted onto a
robotic arm. This facilitates non-isocentric beam delivery
and, more importantly, precision handling of the linear
accelerator to achieve the necessary accuracy required for
SRS delivery [2]. The stereoscopic ceiling mount kV X-ray
Fig. 7.7 A dedicated linear accelerator unit Novalis (Brainlab tubes provide sufficient fast 2D imaging to monitor the
Company) with a built-in high-definition 2.5 mm leaf width at isocenter patient positioning, not only during the setup but also
MLC, for shaped-beam SRS delivery throughout the treatment, to constantly guide or correct the
individual beam direction from the linear accelerator head
beam to irradiate a target [3]. In-room kV X-ray images are
Cyberknife SRS Unit able to match accurately (<0.1 mm) with digitally recon-
structed images (DRRs) from a conventional CT scan to
The Cyberknife consists of a compact linear accelerator generate a virtual image center, which the robotic arm uses
mounted onto a robotic arm. The accelerator is collimated for guiding the individual beam directions. In essence, the
using fixed cones that range from 5 to 60 mm in nominal image center is the heart and soul of Cyberknife delivery,
diameter. In newer units, an optional variable diameter col- and it can be placed anywhere in space as long as the robotic
limator (Iris) can be used. Treatment is primarily guided arm knows where it is. This is fundamentally different from
using a pair of orthogonally mounted planar X-ray imagers, all other SRS units, where the image center must reference
although the system also has an optional optical tracking sys- to the mechanical isocenter of the linear accelerator and not
tem that can be used in conjunction with the X-ray imager. vice versa.
Tomotherapy Unit been immobilization of the patient to correctly match that

of reference position during the CT simulation process.
Tomotherapy is an integrated unit that is capable of deliver- Although patient roll angles can be corrected by changing
ing SRS using intensity modulated radiation therapy treat- the delivery angles, any significant rotational error in pitch
ments based on a helical fan beam geometry analogous to and yaw in the patient setup must be corrected manually
modern CT scanners [15, 40]. In design, one can think of the before the treatment. In addition, the existing Tomotherapy
Tomotherapy unit as a CT unit that also delivers therapeutic system does not have the ability to detect patient motion
high-energy X-rays. The unit has a compact Siemens 6 MeV during delivery. Longer treatments can therefore be broken
S-band linac mounted on a rotating slip ring gantry with a down into smaller sub-fractions to allow imaging to detect
source–axis distance of 85 cm. The linac has neither a bend- potential motion.
ing magnet nor a flattening filter and operates at a nominal
dose rate of approximately 900 MU/min. The bore of the
unit is >80 cm, and there are a series of individual binary Overall Accuracy
(either open or shut) collimators that produce a fan-beam
X-ray across the gantry. The binary MLC leaves open and Treatment accuracy has been the key characteristic of all
close to modulate the beam direction in the transverse SRS units. The mechanical accuracy is reported to be less
direction. than 0.3 mm for Gamma Knife, 0.5–1.0 mm for dedicated
The size of the field in the longitudinal direction (parallel SRS linac units, 0.5–1.0 mm for Cyberknife and Tomotherapy
to the bore) is defined by a pair of jaws. The field size cur- units, and approximately 1.0 mm for linac-based non-
rently in the longitudinal direction is 1, 2.5, or 5 cm. During dedicated SRS units. To evaluate the overall procedural
delivery, the patient is transported through the bore of the accuracy of a SRS treatment (unless for a simple treatment
gantry as the fan-beam rotates around to form a spiral trajec- geometry such as trigeminal neuralgia where mechanical
tory around the patient. The Tomotherapy unit is also accuracy can dominate), other sources of uncertainties must
equipped with 738 xenon detectors that enable online CT be carefully considered. These include (1) uncertainties in
scanning of the patient using MV photons using the treat- the diagnostic imaging studies and image fusion when appli-
ment beam. The unit is shown in Fig. 7.9. With binary colli- cable, (2) dose calculation uncertainties, and (3) target
mation and online imaging capabilities, Tomotherapy motion and setup errors. In general, the imaging resolutions
delivers SRS treatments similar to the shaped-beam linac- of CT and MR studies and the spatial distortions in the imag-
based SRS treatments except the target volume is irradiated ing acquisition process are the major source of uncertainties
in a slice-by-slice fashion. The physical accuracy of the unit intrinsic to all SRS units. For example, irregularities in the
is on the order of 0.5–1.0 mm. gradient field and the magnetic susceptibility from the ste-
For the specific application of high dose intracranial reotactic frame could cause significant spatial errors for the
SRS treatments, one challenge for the Tomotherapy unit has MR studies in SRS treatments. For dose calculations, the
acceptance level of a SRS dose algorithm is <2 % or <1 mm
in the high dose gradient regions.
Table 7.1 summarizes the estimated error levels for SRS
deliveries.
Overall, the imaging studies and setup variations can be
the major sources of uncertainties in most SRS deliveries
[41] but, nonetheless, the overall procedural accuracy can be
conservatively estimated to be on the order of 1–2 mm or
better.
Table 7.1 Estimated uncertainties in intracranial SRS deliveries in

terms of a peripheral isodose surface conforming to a clinically defined
target surface
Source of errors Standard deviation (mm)
Imaging studies (resolution and distortions) 0.5–1.5
Mechanical alignment and setup errors 0.3–1.0
Tissue/target motion 0.5–1.0
Treatment planning 0.5–1.0
Fig. 7.9 Tomotherapy unit with built-in fan-beam megavoltage X-rays
and online megavoltage CT capability for image guidance Overall (quadrature) 1.3−2.2
102 L. Ma et al.
C
Potential Issues and Challenges storage
EARTH BELOW GRADE

Over the past decade, image-guided or image-based SRS B Computer
Room Treatment
delivery such as Cyberknife and linear accelerator systems a Room
have created a significant driving force to deliver more pre-
CONSOLE
cise treatments with greater efficiency and patient through-
put. One common challenge for existing image guidance
solutions is to ensure that the imaging process is sufficiently A
fast. This is important not only for the purposes of treatment SRS
efficiency and patient comfort but also to ensure that any Unit
observed patient shifts can be corrected in a timely fashion
D Alley E
before and during the beam-on delivery. However, frequent
imaging via the current kV or MV X-ray beams has raised
concerns with respect to excessive imaging dose and risks of Fig. 7.10 A sample floor plan for installing a dedicated SRS unit. The
designated points (A, B, C, D, E) are selected calculation points for
secondary malignancies, particularly for young patients. One barriers transmissions
potential development is in the area of imaging modalities
such as online MR-guided teletherapy treatment devices.
Future studies will confirm whether such new developments For room shielding, the bunker design depends on numer-
would improve the overall accuracy and more importantly ous factors that include the machine load, how frequently a
clinical outcome of such devices over the traditional SRS particular wall is irradiated, and the distance between the
deliveries. wall and the isocenter. To ameliorate requirements for pri-
mary irradiations, it is often advantageous to construct a
bunker in the basement and direct the primary beam toward
Shielding Design the underground earth. Figure 7.10 shows a sample floor
plan hosting a SRS unit. As an illustration, we present here
One of major tasks in building and installing a SRS unit is to an example calculation for shielding design of the floor plan
design a bunker room and to enforce radiation protection in Fig. 7.10.
rules. There are two aspects of source shielding and radiation A simple equation for shielding design calculations is
protection for SRS delivery: (1) to reduce extrafocal radia- given as follows:
tion that contributes to dose deposition outside of the target
R = WUT ´ ( d 0 / d ) ´ 10( - t /TVL ) ´ S ,
2
volume and (2) to reduce radiation exposure to the operators
and the general public outside of the room.
The radiation to be shielded is generated from the primary where R is the exposure per week, W is the work load (i.e.,
beam directly hitting outside of the target, scattered radia- maximum delivered dose per week), U is the use factor (i.e.,
tions, and leakage radiation from the head of the radiation fraction of the time beam is directed or scattered toward the
source. For extrafocal radiation irradiating the patient barrier, U = 1 if scattered radiation is considered), T is the
internally, the concern is secondary malignancy risk, particu- occupancy factor (i.e., fraction of the time a person is present
larly in the treatment of young patients and patients with at the point), d0 is the reference dose calibration distance
benign diseases or functional conditions to which radiation (e.g., 100 cm for linac-based SRS unit), t is the thickness of
is being applied. Measurements have been carried out to the barrier materials, TVL is the tenth-value layer of the bar-
document such doses for both GK and linac-based SRS units rier material, and S is the scatter factor (e.g., S = 1 if primary
[5, 42]. As a percentage of the target dose, GK and linac- is used, S = 0.001 if scatter radiation is considered).
based units result in ~0.4 % to blood forming organs, ~0.5 % If we need to calculate the required wall thickness near
to the thyroid, ~0.04 % to the gonads, and ~0.05 % to the the console operator at point A, the following parameters
breast or thorax region. One difference between the GK and are first determined as follows: W = maximum 5 patients
the linac-based units is the dose to the lens. Since the dmax for per day × 5 days per week × 5,000 cGy per treatment =
60
Co is 0.5 cm while for 4–6 MV is in the range of 1.0– 125,000 cGy/week; TVL = 13.5 in. for the concrete of the
1.5 cm, the lens dose is estimated to ~2.5 % of the target wall; d0 = 1 m and d = 4 m; U = 1; T = 1 and S = 1 for the case;
dose. Such a percentage could be important for treatment and the exposure limit is 0.002 rad/week. We have
involving a large dose of delivery such as that for trigeminal
R = 125, 000 ´ 1´ 1´ (1 / 4 ) ´ 10( - t /13.5) ´ 1 £ 0.002.
2
neuralgia where the target dose ranges from 70 to 90 Gy.
Solving for t from the above equation, we have t ≥ 89 in. some states, the RSO is required to be a qualified operator of
This means that the minimum concrete wall thickness at point the SRS unit.
A should be about 7.5 ft in order to meet the exposure limit. In case of pregnancy, the RSO needs to determine whether
It is clear that either increasing source to barrier distance or it is appropriate for the worker to continue to operate. In gen-
adding high-density shielding materials such as lead in the eral, a declared worker may work as long as the total expo-
wall (reducing TVL value) will significantly reduce the wall sure to the embryo or fetus is maintained within the ALARA
thickness and increase room size. If necessary, the workload dose limits.
can also be restricted to maintain the exposure level to a sat-
isfactory level.
Note that the above calculations represented the simplest Installation and Acceptance
form of open-field single beam shielding calculations. When
considering specialized units such as intensity modulated Installation and acceptance a SRS unit requires extensive
shaped-beam SRS, Cyberknife, and Tomotherapy treat- tests of the functionality of the unit. Common to all SRS
ments, additional factors such as the so-called IMRT factor units, three categories of tests should be performed: (1) dose
should be considered. This is because traditional linear rate calibration and radiation survey; (2) radiation and
accelerators are normally calibrated to deliver 1 cGy per mechanical isocenter alignment; and (3) imaging acquisition
monitor unit at approximately 100 cm source-to-isocenter and treatment planning process.
distance, but units such as Cyberknife and Tomotherapy are The phantoms in Fig. 7.11 are sometimes also used for
calibrated differently. To deliver 1 cGy generally requires a regular quality assurance measurements or serve as the scat-
significantly more monitor units and, in general, a conserva- tering medium for radiation survey tasks. As an example, the
tive multiplicative factor such as 15 is currently assumed for absolute dose rate for a newly installed GK SRS unit should
the purpose of shielding calculations. This in effect increases be measured at least at 300 cGy/min using the 16- or 18-mm
the R value of the above equation for the leakage radiation helmet. This is important because this not only affects the
shielding by a factor of 15 assuming 100 % treatments are delivery efficiency but also ensures adequate source life
IMRT type. The actual percentage of IMRT cases can be before the next reloading is needed. The dose calibration
adjusted according to the individual clinical practices. geometry should be measured using tissue-equivalent phan-
Shielding design is covered in detail in National Council on toms of standard shapes as illustrated in Fig. 7.11.
Radiation Protection and Measurements (NCRP) report Another important specification that needs to be tested is
number 151 (http://www.ncrppublications.org/Reports/151). the radiation and mechanical isocenter alignment. This is
In addition, a number of publications describe shielding for performed using the special mechanical alignment tool for
specific modalities [43–45]. the GK unit and the Winston–Lutz test methods for linac-
based units as previously described. In general, the smallest
field size should be measured to ensure maximum accuracy.
Radiation Safety Program The most important test in installing a SRS unit is to
ensure the delivered dose profiles match the prescribed ones.
Before building or installing a SRS unit, an institutional radi- As a general practice, the beam profiles along all major axes
ation safety committee should be formed to implement a of a single-isocenter delivery should be compared with the
radiation safety program. The goal of the program is to main- calculated profiles. The agreement should satisfy the mini-
tain the radiation exposures to all employees and individual mum manufacturer specifications such as 2 % at the central
members of the public to be “as low as reasonably achiev- axis and 2 mm in the dose gradient region. For accurate
able” (ALARA). This is commonly referred to as ALARA delivery using small collimators (e.g., 4–5 mm in diameter),
principle. the requirements can be more stringent such as 1 % or 1 mm
To implement the program, a general rule is to monitor in the dose profiles. The dose profiles can be measured using
the radiation level that is significantly below the regulated radiochromic films with the insert pieces fit into the standard
legal dose limits (e.g., a factor of 10). In general, film badges calibration phantoms as illustrated in Fig. 7.11. The exposed
are used to monitor the radiation exposure to all SRS opera- radiochromic films should be scanned using the laser densi-
tors and visitors. A radiation safety officer (RSO) should be tometer with a resolution of 100 μm or better. An example
designated for the SRS treatments. The key responsibilities result for the measured dose profile is given in Fig. 7.12.
of RSO are to oversee the radiation exposure reports (e.g., Besides the major tests described above, additional accep-
monthly) and to provide refresher radiation safety training to tance tests may be carried out to meet federal, state regula-
the users (e.g., annually). The RSO needs to be responsive to tions, and institutional protocols for installing a SRS unit.
emergency situations and serve as a person of contact Most of these tests are also part of quality assurance program
between the institution and federal and state agencies. For in using the SRS unit. Detailed tasks are recommended in the
104 L. Ma et al.
Fig. 7.11 Example of customized calibration phantoms for SRS units: (a) an acrylic for linac-based SRS units; (b) the ABS plastic phantom used
for GK, (c) the solid-water phantom for the PFX released in 2009, and (d) the solid-water phantom for the Tomotherapy unit
110 audio–video monitoring and communication circuits. The

100 Measurements report by the AAPM task group TG-148 provides recom-
Reference mended QA tests for the Tomotherapy unit [46], and AAPM
90
task group TG-135 provides recommendations for robotic
RELATIVE DOSE (%)
80
70 radiosurgery (Cyberknife) [47].
60 One important but often neglected test in accepting a SRS
50 unit is to check imaging quality and accuracy. To carry out
40
such task, a standard MR/CT phantom with precalibrated
grid pattern should be scanned (preferably with fiducial
30
localizer box) and imported into the treatment planning sys-
20
tem to validate image resolution and detect any spatial dis-
10
tortion for the studies. One test phantom is shown in
0
85 90 95 100 105 110 115 Fig. 7.13. In general, it is recommended that an anthropo-
Z-AXIS morphic head phantom should be scanned for all imaging
modalities including CT, MR, and angiogram during the
Fig. 7.12 Measured dose profiles as compared with the calculated pro- acceptance of a SRS unit.
file in accepting a GK SRS unit. The agreement was found to be within
1 % and 1 mm requirements for the 4-mm helmet shown here
Service and Maintenance

AAPM task group TG42 report [5]. Additional basic tests
include timer, or monitor chamber accuracy, and linearity Proper functioning of a SRS unit is only as good as its weak-
tests; radiation survey; wipe tests (for Gamma Knife units); est link. To maintain high performance quality, routine ser-
emergency beam-off switch; door interlocks; and patient vice and maintenance is important. Before building and
Fig. 7.13 The grid phantom for detecting MR versus CT spatial distor-
tion of the SRS imaging studies
Fig. 7.14 Source loading unit with opened unit head for handling a
Table 7.2 The servicing period for major components of the SRS units high activity of 60Co sources
Recommended
check and
Major components servicing frequency Most of the service and maintenance jobs listed in
Couch motion mechanism (all units) Annually
Table 7.2 are performed in the treatment room except for the
APS (GK) Annually
source loading and reloading task specific to the GK. Since
Microswiches (GK) Monthly
the unit requires manipulation of >1,200 Ci of 60Co sources,
Source activity (GK) Every 7–10 years
the radiation exposure is significantly higher than the design
Stereotactic frames and fiducial boxes (all units) Weekly
Alarm and interlocks (all) Daily
limit of a treatment room. Typically, a special hot cell bunker
Audio/video system (all) Annually is constructed for carrying out such tasks.
Laser systems (LB, TU, CK) Daily The hot cell should be large enough to accommodate the
Gantry rotation (LB, TU) Annually source head and the source loading unit. An illustration of
X-ray camera systems (LB, CK) Weekly the source manipulator with the open unit head is shown in
Robotic arms (CK) Quarterly Fig. 7.14. The door of the hot cell should be also shielded to
GK Gamma Knife, LB linac-based units, CK Cyberknife, TU tomo- allow operation of the robotic arms by an operator from out-
therapy unit side. Lead bricks are typically used to shield the doors.
installing a SRS unit, it is recommended that a comprehen- Personnel Training and Qualifications
sive probable risk analysis (PRA) should be carried out. The
goal of PRA is to assign priorities or possible risk factors to Operation of the SRS units require qualified users. The quali-
the individual components and steps or sequences of the SRS fications and the training of the personnel should follow the
unit operations. This aims to eliminate potential cause lead- practice guideline from the American College of Radiology
ing to unrecoverable or disastrous events. (ACR). The minimum requirements for the key members of
For example, a relative low priority can be assigned to team are summarized in Table 7.3 [48].
the functionality of the skull surface measurement device, All core team members should receive basic operation
while a high priority should be given to the functionality of instructions and radiation safety training (physicians, physi-
the isocenter alignment tool such as the laser systems or cists, therapists, dosimetrists). Other ancillary SRS team
patient positioning device. Since the skull surface measure- members may include nursing staff, neuroradiologists, and
ment can be visually checked versus the patient imaging anesthesiologists in cases where pediatric patients are treated.
studies, the chance of operation error and malfunction is
small. Potential errors can be reversed and corrected with-
out significant risk of injuring the patient. However, mal- Quality Control, Quality Assurance,
functions in the laser system and patient positioning device and Quality Management
would directly affect the shot coordinates and may produce
unrecoverable gross delivery errors. Therefore, regular test- Quality control is the process to validate whether a SRS unit
ing and servicing of such major components of the SRS units is compliant to the design requirements. Once a SRS unit is
is required (Table 7.2). built and installed, a series of tests needs to be performed to
106 L. Ma et al.
Table 7.3 Minimum requirements of a qualified user for performing directive without deviation. In case of deviations, the reporting
SRS treatments level for misadministration and variance is narrow for SRS
Member of the team Qualifications deliveries. Typically, the misadministration of single fraction
Radiation oncologist • Certification (e.g., American Board SRS delivery is defined as when the delivered dose differs
of Radiology or equivalent) from the prescribed dose by 10 % or higher. Because of strin-
• SRS training either in residency or gent criteria, small deviations from the standard protocol are
in attending
likely to result in misadministrations to be reported.
Neurosurgeon • Certification (e.g., American Board
of Neurological Surgery) Therefore, most SRS programs require qualified physicists
• SRS training either in residency or to supervise all QA tasks of the procedure. In-treatment
in attending checks, such as validations of target coordinates matching
Qualified physicist • Certification (e.g., American Board the treatment plan, need also be carried out. Redundant QA
of Radiology or equivalent) checks by another team member are also recommended.
• SRS physics training and unit
All qualified SRS users should receive regular refresher
operations and quality assurance
Radiation therapist • State Licensing and Certification
training. The training should cover radiation safety instruc-
(e.g., American Registry of tions and reporting requirements for radiation workers as
Radiological Technologists (ARRT) well as quality assurance chart reviews. One important train-
• Unit operation training ing aspect is reviewing emergency procedures so that users
can respond quickly and appropriately to major equipment
failures or medical emergencies.
ensure initial quality of the unit. Quality control processes
governs all the manufacturing tests involved in testing the
quality of the unit. Cost and Budget
Because SRS units are medical devices, FDA regulates
the manufacturing quality control process. For the unit to be Budgeting for a SRS unit is a complex issue. Historically, the
used for human treatments in the USA, the manufacturers add-on option to an existing linear accelerator has been
need to submit proper documentations showing good manu- appealing because the accelerator is already in place for con-
facturing practice (GMP) in designing and manufacturing ventional radiation therapy treatments. However, it is a gross
process of the device. misconception that any conventional linear accelerator can
Similar to a teletherapy unit, GK and linac-based SRS be used for SRS delivery. Physically, the requirement in the
units involve moving components such as gantry, couch, and isocenter alignment is a factor of two or three higher for SRS
patient positioning devices. A strict quality assurance pro- deliveries as compared with standard fractionated treat-
gram should be enforced. The quality assurance program can ments. In addition, SRS demands high dose rate with a large
be separate into routine quality assurance and patient-specific number of monitor units being delivered regularly. Wear and
quality assurance procedures. The routine quality assurance tear on the expensive linear accelerator components such as
procedures include daily, monthly, and semiannual/annual the klystron (or the magnetron) and the microwave guide
quality assurance procedures, user refreshing training, and could also be substantial for these treatments.
emergency procedures. Patient-specific quality assurance For linac-based SRS delivery, the gantry speed in MU/
procedures include quality management program (QMP) degree should be at least 5 MU/degree and preferably
checks, pretreatment checks such as dry-run [49] and the 10 MU/degree or higher for efficient deliveries. Extra con-
Winston–Lutz tests. The routine quality assurance for SRS siderations need to be paid to pretreatment QA effort,
deliveries generally includes unit functionality, safety inter- machine scheduling, patient flow logistics, and room design.
locks, radiation monitor, and functionality and accuracy of Today, many centers opt to purchase a dedicated SRS unit in
imaging studies. order to maximize the capabilities of SRS treatments.
In addition to general procedures, qualified operators The estimated startup costs for different SRS units are
must carry out unit-specific quality assurance tests. For summarized in Table 7.4.
example, the GK uses radioactive 60Co sources, and the In general, the acquisition and the ownership cost are
Nuclear Regulatory Commission (NRC) regulations (or significantly higher for dedicated SRS units as compared
agreement state regulations) mandate additional tests for with standard radiation therapy equipment. Aside from
operating GK units. For example, annual wipe tests of all the operation logistics, patient care and patient quality of life
helmets are typically required to ensure no source leakage. should be the driving factor for purchasing a SRS unit. It is
A QMP is also required for most SRS deliveries. Major clear that the patients themselves are most likely to go for a
components of such a program include documentation of SRS option when it provides better results compared to
patient identification by at least two methods (name, photo, those standard radiotherapy practices of the past, such as
etc.), and the treatment plan and delivery follows a written whole brain radiotherapy.
Table 7.4 Total estimated acquisition cost of different SRS units

Units Estimated cost in million dollars (M)
Gamma Knife 3.0–5.0 M (include source units, planning software, bunker, service contracts, and
training); additional 0.5–1.0 M for source reloading with 1–2 weeks down time
Dedicated or combo linac-based unit 3.0–5.0 M (include dedicated accelerator, treatment planning system, bunker,
add-on collimators, service contract, and training)
Tomotherapy unit 3.0–4.0 M (including compact S-band accelerator units, online-CT, treatment
planning system, service, and training)
Cyberknife 3.0–4.0 M (including X-band accelerator units, bunker, imaging devices, service
contract, and training)
12. Levy RP, Fabrikant JI, Frankel KA, Phillips MH, Lyman JT,
Conclusions Lawrence JH, et al. Heavy-charged-particle radiosurgery of the
pituitary gland: clinical results of 840 patients. Stereotact Funct
Neurosurg. 1991;57(1–2):22–35.
Specialized equipment is needed to successfully deliver 13. Weber DC, Chan AW, Bussiere MR, Harsh GR, Ancukiewicz M,
SRS. This is accomplished using a number of different Barker II FG. Proton beam radiosurgery for vestibular schwan-
methods. However, all radiosurgery equipment must noma: tumor control and cranial nerve toxicity. Neurosurgery.
2003;53(3):577–86; discussion 586–8.
deliver highly conformal doses of radiation with steep
14. Adler Jr JR, Murphy MJ, Chang SD, Hancock SL. Image-guided
dose gradients to small targets precisely localized within robotic radiosurgery. Neurosurgery. 1999;44(6):1299–306; discus-
3D space. This equipment must be properly installed, sion 306–7.
serviced, used, and tested to ensure safe and effective 15. Mackie TR, Holmes T, Swerdloff S, Reckwerdt P, Deasy JO, Yang
J, et al. Tomotherapy: a new concept for the delivery of dynamic
patient treatments.
conformal radiotherapy. Med Phys. 1993;20(6):1709–19.
16. Podgorsak EB, Olivier A, Pla M, Lefebvre PY, Hazel J. Dynamic
stereotactic radiosurgery. Int J Radiat Oncol Biol Phys.
References 1988;14(1):115–26.
17. McGinley PH, Butker EK, Crocker IR, Landry JC. A patient rotator
1. Spielgel EA. History of human stereotaxy (stereoencephalotomy). for stereotactic radiosurgery. Phys Med Biol. 1990;35(5):649–57.
1st ed. New York: Georg Thieme; 1982. 18. Leksell DG. Stereotactic radiosurgery. Present status and future
2. Leksell L. The stereotaxic method and radiosurgery of the brain. trends. Neurol Res. 1987;9(2):60–8.
Acta Chir Scand. 1951;102(4):316–9. 19. Lindquist C. Gamma Knife radiosurgery. Semin Radiat Oncol.
3. Andrews DW, Scott CB, Sperduto PW, Flanders AE, Gaspar LE, 1995;5:197–202.
Schell MC, et al. Whole brain radiation therapy with or without 20. Maitz AH, Wu A, Lunsford LD, Flickinger JC, Kondziolka D,
stereotactic radiosurgery boost for patients with one to three brain Bloomer WD. Quality assurance for gamma knife stereotactic
metastases: phase III results of the RTOG 9508 randomised trial. radiosurgery. Int J Radiat Oncol Biol Phys. 1995;32(5):1465–71.
Lancet. 2004;363(9422):1665–72. 21. Wu A, Lindner G, Maitz H, Kalend M, Lunsford LD, Flickinger JC,
4. Corn BW, Curran Jr WJ, Shrieve DC, Loeffler JS. Stereotactic et al. Physics of gamma knife approach on convergent beams in ste-
radiosurgery and radiotherapy: new developments and new direc- reotactic radiosurgery. Int J Radiat Oncol Biol Phys. 1990;18:941–9.
tions. Semin Oncol. 1997;24(6):707–14. 22. Bourland JD, McCollough KP. Static field conformal stereotactic
5. Schell MC, Bova FJ, Larson DA, Leavitt DD, Lutz W, Podgorsak radiosurgery: physical techniques. Int J Radiat Oncol Biol Phys.
EB, et al. Stereotactic radiosurgery, report of the American 1994;28(2):471–9.
Association of Physicists in Medicine Task Group No. 42. College 23. Bova FJ, Friedman WA, Mendenhall WM. Stereotactic radiosur-
Park: American Institute of Physics; 1995. gery. Med Prog Technol. 1992;18(4):239–51.
6. Phillips MH, Stelzer KJ, Griffin TW, Mayberg MR, Winn 24. Colombo F, Benedetti A, Pozza F, Zanardo A, Avanzo RC, Chierego
HR. Stereotactic radiosurgery: a review and comparison of meth- G, et al. Stereotactic radiosurgery utilizing a linear accelerator.
ods. J Clin Oncol. 1994;12(5):1085–99. Appl Neurophysiol. 1985;48(1–6):133–45.
7. Loeffler JS, Shrieve DC, Wen PY, Fine HA, Kooy HM, Addesa AE, 25. Falco T, Lachaine M, Poffenbarger B, Podgorsak EB, Fallone BG.
et al. Radiosurgery for intracranial malignancies. Semin Radiat Setup verification in linac-based radiosurgery. Med Phys. 1999;
Oncol. 1995;5(3):225–34. 26(9):1972–8.
8. Harsh GR, Thornton AF, Chapman PH, Bussiere MR, Rabinov JD, 26. Friedman WA, Bova FJ, Spiegelmann R. Linear accelerator radio-
Loeffler JS. Proton beam stereotactic radiosurgery of vestibular surgery at the University of Florida. Neurosurg Clin N Am.
schwannomas. Int J Radiat Oncol Biol Phys. 2002;54(1):35–44. 1992;3(1):141–66.
9. Larsson B, Leksell L, Rexed B, Sourander P, Mair W, Andersson 27. Leavitt DD, Watson G, Tobler M, Williams G, Gaffney DK, Shrieve
B. The high-energy proton beam as a neurosurgical tool. Nature. DC. Intensity-modulated radiosurgery/radiotherapy using a micro-
1958;182(4644):1222–3. multileaf collimator. Med Dosim. 2001;26(2):143–50.
10. Larsson B, Sarby B. Equipment for radiation surgery using narrow 28. Lutz W, Winston KR, Maleki N. A system for stereotactic radiosur-
185 MeV proton beams. Dosimetry and design. Acta Oncol. gery with a linear accelerator. Int J Radiat Oncol Biol Phys. 1988;
1987;26(2):143–58. 14(2):373–81.
11. Lawrence JH, Tobias CA, Linfoot JA, Born JL, Manougian E, 29. Nedzi LA, Kooy HM, Alexander III E, Svensson GK, Loeffler
Lyman J. Heavy particles and the Bragg peak in therapy. Ann Intern JS. Dynamic field shaping for stereotactic radiosurgery: a modeling
Med. 1965;62:400–7. study. Int J Radiat Oncol Biol Phys. 1993;25(5):859–69.
108 L. Ma et al.
30. Podgorsak EB, Olivier A, Pla M, Hazel J, de Lotbiniere A, Pike B. radiosurgery system for trigeminal neuralgia. J Neurosurg. 2004;
Physical aspects of dynamic stereotactic radiosurgery. Appl 101 Suppl 3:351–5.
Neurophysiol. 1987;50(1–6):263–8. 40. Mackie TR, Balog J, Ruchala K, Shepard D, Aldridge S, Fitchard
31. Solberg TD, Boedeker KL, Fogg R, Selch MT, DeSalles AA. E, et al. Tomotherapy. Semin Radiat Oncol. 1999;9(1):108–17.
Dynamic arc radiosurgery field shaping: a comparison with static 41. Scheib SG, Gianolini S, Lomax NJ, Mack A. High precision radio-
field conformal and noncoplanar circular arcs. Int J Radiat Oncol surgery and technical standards. Acta Neurochir Suppl. 2004;91:
Biol Phys. 2001;49(5):1481–91. 9–23.
32. Winston KR, Lutz W. Linear accelerator as a neurosurgical tool for 42. Berk HW, Larner JM, Spaulding C, Agarwal SK, Scott MR, Steiner
stereotactic radiosurgery. Neurosurgery. 1988;22(3):454–64. L. Extracranial absorbed doses with Gamma Knife radiosurgery.
33. Colombo F, Francescon P, Cora S, Cavedon C, Terrin G. A simple Stereotact Funct Neurosurg. 1993;61 Suppl 1:164–72.
method to verify in vivo the accuracy of target coordinates in linear 43. Wu C, Guo F, Purdy JA. Helical tomotherapy shielding calculation
accelerator radiosurgery. Int J Radiat Oncol Biol Phys. for an existing LINAC treatment room: sample calculation and cau-
1998;41(4):951–4. tions. Phys Med Biol. 2006;51(21):N389–92.
34. Gibbs Jr FA, Buechler D, Leavitt DD, Moeller JH. Measurement of 44. Baechler S, Bochud FO, Verellen D, Moeckli R. Shielding require-
mechanical accuracy of isocenter in conventional linear-accelerator- ments in helical tomotherapy. Phys Med Biol. 2007;52(16):5057–67.
based radiosurgery. Int J Radiat Oncol Biol Phys. 1993;25(1): 45. McDermott PN. Radiation shielding for gamma stereotactic radio-
117–22. surgery units. J Appl Clin Med Phys. 2007;8(3):2355.
35. Boyer AL, Antonuk L, Fenster A, Van Herk M, Meertens H, Munro 46. Langen KM, Papanikolaou N, Balog J, Crilly R, Followill D,
P, et al. A review of electronic portal imaging devices (EPIDs). Med Goddu SM, et al. QA for helical tomotherapy: report of the AAPM
Phys. 1992;19(1):1–16. Task Group 148. Med Phys. 2010;37(9):4817–53.
36. Leavitt DD, Gibbs Jr FA, Heilbrun MP, Moeller JH, Takach Jr 47. Dieterich S, Cavedon C, Chuang CF, Cohen AB, Garrett JA, Lee
GA. Dynamic field shaping to optimize stereotactic radiosurgery. CL, et al. Report of AAPM TG 135: quality assurance for robotic
Int J Radiat Oncol Biol Phys. 1991;21(5):1247–55. radiosurgery. Med Phys. 2011;38(6):2914–36.
37. Pedroso AG, De Salles AA, Tajik K, Golish R, Smith Z, Frighetto 48. Potters L, Kavanagh B, Galvin JM, Hevezi JM, Janjan NA, Larson
L, et al. Novalis shaped beam radiosurgery of arteriovenous malfor- DA, et al. American Society for Therapeutic Radiology and
mations. J Neurosurg. 2004;101 Suppl 3:425–34. Oncology (ASTRO) and American College of Radiology (ACR)
38. Smith ZA, De Salles AA, Frighetto L, Goss B, Lee SP, Selch M, practice guideline for the performance of stereotactic body radia-
et al. Dedicated linear accelerator radiosurgery for the treatment of tion therapy. Int J Radiat Oncol Biol Phys. 2010;76(2):326–32.
trigeminal neuralgia. J Neurosurg. 2003;99(3):511–6. 49. Klein EE, Hanley J, Bayouth J, Yin FF, Simon W, Dresser S, et al.
39. Rahimian J, Chen JC, Rao AA, Girvigian MR, Miller MJ, Task Group 142 report: quality assurance of medical accelerators.
Greathouse HE. Geometrical accuracy of the Novalis stereotactic Med Phys. 2009;36(9):4197–212.
Part III
Stereotactic Radiosurgery Techniques
Gamma Knife Radiosurgery
8
Ajay Niranjan, Greg Bowden, John C. Flickinger,
and L. Dade Lunsford
In the USA, based on the available published literature,

Historical Review arteriovenous malformations (AVMs) and skull base tumors
that failed other treatments were considered the initial indi-
Professor Lars Leksell first coupled an orthovoltage X-ray cations for radiosurgery [3]. A cautious approach was
tube with his first-generation guiding device to focus radia- adopted while waiting for increased scientific documenta-
tion on the Gasserian ganglion to treat facial pain. He subse- tion. The encouraging results of radiosurgery for benign
quently investigated cross-fired protons as well as X-rays tumors and vascular malformations led to an exponential rise
from an early-generation linear accelerator (linac) for radio- of radiosurgery cases and sales of radiosurgical units
surgery [1]. In the 1960s, he became dissatisfied with the (Table 8.1). The research rate continues to escalate with hun-
cumbersome nature of cross-fired proton beams and the poor dreds of papers being published on this technology annually
reliability and wobble of then-existing linear accelerators. in the last several years [5]. Currently, metastatic brain
Leksell and Larsson finally selected cobalt-60 as the ideal tumors have become the most common indication for radio-
photon radiation source and developed the Gamma Knife [2, 3]. surgery, comprising 30–50 % of radiosurgery cases at busy
They placed 179 60Co sources in a hemispherical array so centers (Table 8.2). A wide variety of brain and body disor-
that all gamma rays (radiation from the decay of 60Co) were ders are now treated with radiosurgery [6]. Gamma Knife
focused on a single point thereby creating cumulative radia- radiosurgery is an example of a disruptive innovation in neu-
tion isocenters of variable volume depending on the beam rosurgery as a wide variety of brain disorders are now treated
diameter. The first Gamma Knife created a discoid-shaped with radiosurgery [7].
lesion suitable for neurosurgical treatment of movement dis-
orders and intractable pain management.
Clinical work with the Gamma Knife began in 1967 in The Evolution of Gamma Knife: Models A, B, C,
Sweden [4]. By 1975, a series of surgical pioneers at the and PERFEXION
Karolinska Hospital, Stockholm, began to use a reengineered
Gamma Knife (spheroidal lesion) for the treatment of intra- The Gamma Knife has evolved steadily since 1967. In the
cranial tumors and vascular malformations. The third and first models (model U or A), 201 cobalt sources were
fourth units created were placed in Buenos Aires and arranged in a hemispherical array. These units presented
Sheffield England in the early 1980s. Lunsford et al. intro- challenging 60Co loading and reloading issues. To eliminate
duced the first clinical 201 source Gamma Knife unit to this problem, the unit was redesigned so that sources were
North America (the fifth gamma unit worldwide). Dr. arranged in a circular (O-ring) configuration (models B, C,
Lunsford performed the first Gamma Knife radiosurgery in and 4C) (Fig. 8.1).
August 1987 at University of Pittsburgh Medical Center. Gamma Knife radiosurgery usually involves multiple iso-
centers of different beam diameters to achieve a treatment
plan that conforms to the irregular three-dimensional vol-
A. Niranjan, M.D., M.B.A. (*) • G. Bowden, M.D., M.Sc. umes of most lesions. The total number of isocenters may
Department of Neurological Surgery, University of Pittsburgh,
vary depending upon the size, shape, and location of the tar-
Pittsburgh, PA, USA
e-mail: niranjana@upmc.edu get. Each isocenter has a set of three x, y, z stereotactic coor-
dinates corresponding with its location in three-dimensional
J.C. Flickinger, M.D. • L.D. Lunsford, M.D.
Department of Radiation Oncology, University of Pittsburgh, space as defined using a rigidly fixed skull stereotactic frame.
Pittsburgh, PA, USA In terms of actual dose delivery, this means several changes in
112 A. Niranjan et al.
Table 8.1 Number and location of active Table 8.2 Brain disorders treated worldwide by gamma knife
Gamma Knife sites worldwide [6] radiosurgery by December 2011 [6]
Continent Country Units Category Indication Total
Africa Egypt 2 Malignant tumors Chondrosarcoma 957
Morocco 1 Nasopharyngeal carcinoma 1,770
Australia Australia 1 Other malignant tumor 10,400
Asia China 16 Malignant glial tumor (grade III and IV) 30,874
India 7 Metastatic tumor 252,372
Indonesia 1 Total 296,373
Iran 1 Benign tumors Hemangiopericytoma 1,533
Korea 15 Glomus tumor 2,335
Japan 48 Chordoma 2,395
Jordan 1 Other schwannoma 2,645
Pakistan 1 Hemangioblastoma 2,652
Philippines 1 Trigeminal schwannoma 4,139
Singapore 1 Pineal region tumor 4,420
Taiwan 7 Craniopharyngioma 5,107
Thailand 1 Benign glial tumors (grade I and II) 5,990
Vietnam 1 Other benign tumor 6,865
Europe Austria 2 Pituitary adenoma (nonsecreting) 19,575
Belgium 1 Pituitary adenoma (secreting) 31,226
Croatia 1 Vestibular schwannoma 63,797
Czech Rep. 1 Meningioma 90,761
France 3 Total 243,440
Germany 4 Vascular disorders Aneurysm 348
Greece 1 Other vascular disorder 5,325
Italy 5 Cavernous angiomas 6,128
Netherlands 1 AVM 71,566
Norway 1 Total 83,367
Poland 1 Functional OCD 193
Portugal 1 disorders Intractable pain 707
Romania 1 Other functional disorders 1,444
Russia 2 Parkinson’s disease 1,715
Spain 1 Epilepsy 2,731
Sweden 1 Trigeminal neuralgia 43,402
Switzerland 1 Total 50,192
Turkey 5 Ocular disorders Other ocular disorders 238
UK 5 Glaucoma 314
South America Argentina 1 Uveal melanoma 2,084
Brazil 2 Total 2,636
Chile 1 Total 676,008
Columbia 2
Venezuela 1
North America Canada 3 Center in March 2000. This technology combined advances
Dominican Rep. 1
in dose planning with robotic engineering and uses a submil-
Mexico 2
limeter accuracy automatic positioning system (APS)
USA 111
(Fig. 8.2). This technology obviates the need to manually
Puerto Rico 2
Total 269
adjust each set of coordinates in a multiple isocenter plan.
The robotic positioning system moves the patient’s head to
the target coordinates defined in the treatment plan. The robot
eliminates the time spent removing the patient from the hel-
the patient’s head position within the helmet. In 1999, the met, setting the new coordinates for each isocenter, and repo-
model C Gamma Knife was introduced. The first model C in sitioning the patient in the helmet. This has significantly
the USA was installed at the University of Pittsburgh Medical reduced the total time spent to complete the treatment and
8 Gamma Knife Radiosurgery 113
Fig. 8.1 Schematic diagram of model 4C Gamma Knife unit (Courtesy of Elekta AB, Stockholm, Sweden.)
(Fig. 8.3). The first unit was installed at the University of

Pittsburgh in January 2005. The model 4-C is equipped
with enhancements designed to improve workflow, increase
accuracy, and provide integrated imaging capabilities. The
integrated imaging, powered by Leksell GammaPlan, offers
the ability to fuse images from multiple sources. The plan-
ning information can be viewed on both sides of the treatment
couch. The helmet changer and robotic APS are faster and
reduce total treatment time.
The Gamma Knife PERFEXION unit is the latest inception
of this technology (Fig. 8.4). This system was developed in
response to input from expert neurosurgeons, radiation oncol-
ogists, and medical physicists. This resulted in a list of critical
features that included unlimited cranial reach, full automation
of the treatment protocol, improved patient and staff comfort,
radiation protection, and best dosimetry performance. The
first unit became active in 2006 and one was installed at the
University of Pittsburgh in 2007. The PERFEXION unit has
192 60Co sources arranged in a cylindrical configuration in five
concentric rings (Fig. 8.5). This differs substantially from the
Fig. 8.2 Gamma Knife 4C automatic positioning system, docking and previous hemispherical arrangements of the other units. The
4 mm collimator (Courtesy of the University of Pittsburgh Gamma tungsten collimator array ring eliminates the need for collima-
Knife Facility, 2013.) tor helmets with PERFEXION [9, 11]. Three collimators are
available for the PERFEXION system: 4, 8, and 16 mm. The
also increases accuracy and safety [8–13]. Because the treat- tungsten collimator array is subdivided into eight identical but
ment time is shortened, a precise three-dimensional (3D) plan independent sectors, each containing 72 collimators (24 colli-
can be generated using multiple smaller beams achieving mators for each of the three collimators). Each sector with 24
volumetric conformality. Such an approach results in a sources can be moved independently into five different posi-
steeper dose fall-off extending beyond the target (higher tions: (1) sector in home position when system is standby, (2)
selectivity). The other features of the model C unit include an 4 mm collimator, (3) 8 mm collimator, (4) 16 mm collimator,
integral helmet changer, dedicated helmet installation trol- and (5) sector off position (blocking). This revolutionary
leys, and color-coded collimators. In 2005, the fourth- change allows the ability to form composite shots using differ-
generation Leksell Gamma Knife, model 4-C, was introduced ent beam diameters.
Fig. 8.3 Gamma Knife 4C unit

(Courtesy of the University
of Pittsburgh Gamma Knife
Facility, 2013.)
Fig. 8.4 Gamma Knife

PERFEXION unit
(Courtesy of the University
of Pittsburgh Gamma Knife
Facility, 2013.)
The radiation cavity has been increased by more than reotactic coordinates (Fig. 8.6). This provides better patient
300 % compared to previous models. The increase in the vol- comfort and allows complete of the majority of radiosurger-
ume of the radiation cavity allows for a greater mechanical ies in one single run, and has repeatability better than
treatment range in X/Y/Z. It is (160/180/220 mm) for the 0.05 mm. Docking of the patient into the PPS is done
PERFEXION system compared to (100/120/165 mm) for by means of an adaptor that attaches to the standard stereo-
other Gamma Knife models. This provides virtually unlim- tactic Leksell G frame. The redesigned hardware of the
ited cranial reach, so crucial in the care of patients with mul- PERFEXION unit has had significant impact on the planning
tiple brain metastases [4]. The APS used in the C units was software Leksell GammaPlan PFX (LGP PFX), a new
replaced by the patient positioning system (PPS). Rather version of the LGP running on a PC platform with the
than just the head, the whole bed moves into preselected ste- Linux operating system. The PERFEXION unit provides the
Fig. 8.5 Diagrams of Gamma Knife PERFEXION collimator system. (a) Cross section of PERFEXION radiation unit. (b) Independently moving
sectors can be modified for varied shot compositions. (c–e) Sector position in 8, 4, and 16 mm collimator (Courtesy of Elekta AB, Stockholm, Sweden.)
The Radiosurgery Procedure
Following are the basic steps of Gamma Knife radiosurgery:

1. Daily quality assurance of the radiosurgery system
2. Application of the stereotactic guiding device to the
patient’s head
3. Stereotactic brain imaging using magnetic resonance
imaging (MRI), computed tomography (CT), and/or an
angiogram
4. Quality assurance of images
5. Determination of target volume(s)
6. Conformal radiosurgery dose planning by the radiosur-
gery team
7. Stereotactic delivery of radiation to the target volume by
positioning the patient’s head inside a collimator helmet
or docking system
8. Removal of stereotactic guiding device
Fig. 8.6 Docking mechanism of PERFEXION patient positioning system Daily Quality Assurance of the Radiosurgery
(Courtesy of the University of Pittsburgh Gamma Knife Facility, 2013.) System
potential to increase the spectrum of treatable indications Gamma Knife quality assurance testing is performed by an
including multiple brain metastasis, access to the upper authorized medical physicist every morning. The medical
cervical spine, and other pathologies of the head and neck. physicist ensures that the system tests required by US Nuclear
Regulatory Commission (NRC) regulations are performed is needed to bring the lesion closest to the center of the frame.
and functioning properly. These tests include the perma- For anterior target, short posterior posts are preferred to
nently mounted radiation monitor and its remote indicator, avoid collision of the posterior post/pin assembly with the
handheld radiation monitor, patient viewing and communica- collimator helmet. To target posterior lesions (occipital lobe
tion systems, door interlock, timer termination of exposure, tumors, transverse sinus tumors, cerebellar lesions), the
emergency stops, beam status indicators, availability of the frame is shifted backward by positioning the earpieces for-
release rods for the emergency removal of a patient, test run ward. The posterior edge of the earpiece is kept at 110–
of the automatic patient positioning unit, microswitch test, 125 mm instead of 95–100 mm. The anterior posts are
and function of the helmet hoist. Assessing the PERFEXION positioned as low as possible on the supraorbital region to
unit is very similar but includes a sector positioning check to avoid collision of the frontal post/pin assembly with the col-
verify proper functioning of the automatic collimator setup. limator helmet. For radiosurgery planning, a higher gamma
Additionally, the patient docking and positioning system is angle (120–140°) is used if a potential collision is detected at
assessed using a diode test tool and focus precision check. the default angle of 90°. To reach lateral targets (lateral
Apart from daily quality assurance tests, monthly and annual metastases, convexity tumors, far lateral tumors), the frame
quality assurance is also performed in addition to preventa- is shifted laterally (right or left) toward the lesion. While
tive maintenance of the Gamma Knife unit. shifting the frame laterally, it is important to make sure that
there is enough space on the contralateral side to allow posi-
tioning of the fiducial box on the base ring of the frame. The
Application of the Stereotactic Guiding Device MRI or CT fiducial should be tried on the frame prior to
sending the patient to the MRI unit. If the fiducial box does
For Gamma Knife radiosurgery, appropriate stereotactic not fit on to the frame due to excessive shifting of the frame,
frame placement is the initial critical aspects of the proce- the frame will have to be repositioned.
dure. Subtle changes to the positioning of the frame can be The frame adaptor (which attaches the frame to the table)
the difference between a simple treatment plan, complicated is checked for fit. If the frame is shifted too anteriorly and the
patient positioning, or even inability to proceed without back of head ring is too close to the neck, adaptor may not fit
changing the frame. These issues are greatly reduced in the and consequently treatment cannot be carried out. Tight fit-
PERFEXION system but the principles remain imperative to ting of the adaptor may cause neck discomfort to patient espe-
best patient care. Prior to frame placement, the radiosurgery cially during a long treatment. The frame cap check provides
team should review the preoperative images and discuss information about geometry of all stereotactic frame parts
optimal frame placement strategy. The preoperative images including posts and screws and also information about patient
should be kept in plain view while applying the head frame. head geometry to the treatment planning system. This infor-
An effort should be made to keep the lesion as close to the mation is needed for the prediction of potential collisions or
center of the frame as possible. The possibility of collision close contact with the gamma knife unit collimator system. If
by the frame base ring, the posts/pins assembly, or the frame cap does not fit then the exact post and screw measure-
patient’s head with the collimator helmet during treatment ments must be given to the treatment planning system.
should also be considered prior to the frame application.
Furthermore, significant angulation should be avoided as it
may affect collisions, planning integration, and patient com- Stereotactic Brain Imaging Techniques
fort when fixed in place for treatment. The authors use the
following strategies to avoid collisions and obtain optimal Aside from the frame application, the next most important
frame placement. To target lower lesions (skull base tumors, aspect of radiosurgery is accurate imaging of the target. MRI
acoustic tumors, cerebellar lesions), the frame is positioned is the preferred imaging modality. CT is used when MRI is
lower by placing the ear bars in the top holes of the earpieces not possible. Angiograms are used in conjunction with MRI
on the Leksell G frame. For higher lesions (sagittal sinus for AVM radiosurgery.
meningioma, metastases high in the frontal or parietal lobe),
the frame is positioned higher by placing the ear bars in the Stereotactic MRI
bottom holes of the earpieces. For anterior targets (anterior The critical aspects of stereotactic imaging include optimal
frontal lobe tumors, cavernous sinus tumors, sellar lesion contrast between normal and abnormal tissues in addition to
and lesion anterior to sella, and anterior temporal lesions), high spatial resolution, short scan time, and thin slices so that
the frame is shifted forward by placing earpieces posteriorly accurate target localization can be achieved. The use of MRI
on the base ring of the frame. The posterior edge of the ear- in stereotactic planning has enhanced accurate targeting of
piece is kept at 90 to 75 mm on the y-dimension of the head lesions that are usually not adequately defined by any other
frame (instead of 95–100 mm) depending upon the shift that imaging modality. Some physicians prefer the fusion of
magnetic resonance and CT images for stereotactic guidance, With modern CT scanners, 1- or 2-mm-thick slices (depending
as they believe that in certain types of scanners, distortion may upon the size of the lesion) without any gap can be obtained
affect the accuracy of target localization in MRI. For the initial in 4–5 min. Before removing the patients from the CT
2 years, the authors used both MRI and CT for stereotactic table, accuracy checks are performed to make sure that
planning. Significant target coordinate differences were not images would be accepted by the planning system and the
observed using the Leksell stereotactic system. Since 1993, lesions have coordinates that are achievable in the Gamma
MRI has been used for stereotactic radiosurgery planning in Knife unit.
almost all eligible cases using a 1.5-T unit.
At our institution, high-resolution, gadolinium-enhanced Stereotactic Angiography
3D localizer (T2* images) image sequence is used first to Angiography is the gold standard for AVM radiosurgery
localize the tumor in axial, sagittal, and coronal images [14]. planning. It should be used in conjunction with MR or CT
This sequence (3-mm-thick slices 2 mm apart) only takes imaging to provide the third imaging dimension. The tech-
45 s for 11 axial, 11 sagittal, and 11 coronal slices. Using the nique of angiography differs slightly from the conventional
axial images, the fiducials can be measured and compared digital angiography as the stereotactic angiographic images
with the opposite side to exclude the possibility of magnetic are used not only for AVM nidus definition but also to guide
resonance (MR) artifacts and to confirm that there is no radiation to the target. The orthogonal images (instead of
angulation or head tilt. Alternatively, T1-weighted sagittal oblique or rotated) are preferred but are not necessary. For
scout images (3-mm-thick slices with 1 mm) using spin echo AVM nidi that are not properly visualized in orthogonal
pulse sequence can be obtained for lesion localization. The planes, a rotation of up to 10° in two dimensions or aspects
average time for this sequence is approximately 1.5 min. For can be used without compromising the accuracy of radiation
stereotactic imaging of most lesions, a 3D volume acquisi- delivery. Before removing the angiography catheter, the
tion using fast spoiled-GRASS (gradient recalled acquisition images should be reviewed to make sure that all the fiducials
in steady state) sequence at 512 × 256 matrix and two NEX are seen on the images. Digital subtraction techniques, despite
(number of excitations) covering the entire lesion and sur- a potential radial distortion error, have proved satisfactory.
rounding critical structures are preferred. To define the radio-
surgery target, this volume is displayed as 1- or 1.5-mm-thick
axial slices. The field of view (FOV) is kept at 25 cm × 25 cm Quality Assurance of Images
in order to visualize all fiducials. The approximate imaging
time for this sequence is 6–8 min. Regular quality control checks of the MR unit are performed
We generally prefer 3D spoiled-GRASS sequence for in order to maintain accuracy of images. With a properly
most lesions. Additional sequences are performed when more shimmed magnet, regular servicing, and strict quality assur-
information is needed. Pituitary lesions are particularly diffi- ance on the unit as well as on the images, MRI provides high-
cult to image especially if there has been prior surgery. A half resolution images with accurate target localization. A special
dose of paramagnetic contrast is usually given to image pitu- frame holder is used in order to avoid head movement during
itary adenomas. For residual pituitary tumors, after trans- MRI. Accuracy of images is checked for each image sequence
sphenoid resection, a fat-suppression SPGR sequence is by comparing the known frame measurement with image
recommended in order to differentiate tumor from the fat measurements in addition to the distance from the posterior
packed in the resection cavity. For cavernous malformations, fiducials to the middle fiducials. The images are exported
an additional variable echo multi planar (VEMP) imaging is from Imaging Suite via the Ethernet. Images are defined
obtained to define the hemosiderin rim. For thalamotomy using Leksell GammaPlan software (LGP) after the images
planning, an additional fast inversion recovery sequence is are transferred to the LGP computer. The measurements are
performed to differentiate basal ganglia from white-matter again checked and compared with the known frame measure-
tracts. Brain metastases patients receive a double dose of con- ment and also the distance to the middle fiducial in order to
trast agent, and the entire brain is imaged by 2 mm slices to check for any distortion during image transfer.
identify all of the lesions. Before removing the patients from
the MR scanner, the images must be checked for accuracy.
Determination of Target Volume(s)
Stereotactic CT Imaging
When using CT imaging instead of MRI, it is advisable to use Target determination is an important step in order to make a
short posterior posts to avoid artifacts from the posts and conformal plan. Target volume can be outlined using the LGP
pins. Care should also be taken in deciding the optimal place software (manual or semiautomatic mode). Experienced sur-
for the pins because they cause artifacts on CT. An effort geons can create conformal dose plans without outlining the
should be made to keep the lesion away from the pin artifacts. target; however, for new centers especially where physicists
assume the initial responsibility for planning, target definition (multiple tumors) can be treated using different isodose pre-
and outlining by surgeon or oncologist becomes an important scription and different central doses with the use of multiple
step. The surgeon’s input is required to define radiosurgery matrices.
targets for patients with AVMs, tumors, and functional neuro- PERFEXION offers additional options to provide a con-
surgery as used by some centers. Defining the target volume formal dose plan. Three different approaches are available:
also helps in calculating conformality index. • The use of a classical combination of shots including 4, 8,
and 16 mm collimators to cover the entire target volume.
• The automatic dynamic shaping ability, which is a new
Techniques of Conformal Dose Planning feature in the treatment planning, introduced for the
PEREFXION system. This procedure will provide solu-
This area will focus primarily on the 4C and PERFEXION tions to block selected sectors to protect volumes defined
units. In the process of treatment planning, several strategies as critical structures.
can be used. The model C allows treatment using robotic • The use of single isocenters composed of different beam
automatic patient positioning system (APS mode), manual diameters or blocked sectors. Any pattern of sectors
positioning (trunnion mode), or mixed treatment (some iso- including 4, 8, 16 mm collimators and blocks can be gen-
centers in APS mode and some in trunnion mode). A differ- erated. This can aid significantly for shaping dose distri-
ent approach would be used if only a trunnion treatment plan bution especially for irregular volumes.
was possible versus an APS treatment plan. Universally, in
LGP one can start planning from the middle of the target and
then move to the top and bottom of the target. Another Techniques of Stereotactic Radiation
approach is to start at the bottom or top and build from the Delivery Using Gamma Knife
starting point. Some surgeons prefer to outline the target vol-
ume before planning the treatment volume. Beginners can The model C Gamma Knife allows radiation delivery using
also use the inverse dose-planning algorithm to create a plan trunnion mode (manual patient positioning) or APS mode
and then optimize it manually. When planning a treatment (robotic positioning) or a combination of the two (mixed
using trunnions only, one might tend to use larger collima- treatment). In trunnion treatment, the x, y, and z of each iso-
tors (especially for larger lesions) to reduce the time and center are set manually and double-checked to avoid errors.
maximize coverage of the target. For example, for a medium- The same coordinates and the time obtained from LGP are
size acoustic tumor, in trunnion mode, one might use a few entered into the control console, and radiation is adminis-
14 mm collimators for the majority of the tumor and a few tered. The APS plan is transferred directly from planning
4 mm collimators for the intracanalicular portion of the computer to control computer. The operator selects the run (a
tumor. In APS mode, however, one would most likely use combination of isocenters of same beam diameter) that
multiple 8-mm isocenters for the majority of the tumor and matches the collimator helmet on the Gamma Knife unit.
4 mm isocenters for the intracanalicular portion because the The APS is moved to the dock position, and the patient’s
total time spent would be less. There would be no need to go head frame is fixed into the APS. The accuracy of the dock-
into the treatment room to set each isocenter. As long as the ing position is checked. The system prompts the user to per-
isocenters are in close proximity to one another, the software form clearance checks first for all those planned isocenters in
would automatically put them into the same treatment run which the pins, posts, frame, or patient’s head would be less
and the patient would move from one set of coordinates to than 12 mm away from the inner surface of the collimator
the next until all isocenters of one collimator size were helmet (even though they may not match with the collimator
treated. The conformal dose planning is enhanced by the use size that is being used for first run). The clearance check is
of multiple small collimators. performed by moving the patient to those positions under
There are other tricks to use in planning; for example, APS manual control and by checking the risk of collision
using a steep (125°) gamma angle for posterior lesions (cer- with the collimator helmet. After the clearance check, the
ebellar or occipital) to avoid frame collisions. Another tech- system prompts the surgeon to carry out position checks. In
nique available for single isocenter lesions is to match the the position checks, all the isocenters using the same helmet
gamma angle to the angle of the target. In APS mode, during are checked, one by one, by moving the patient’s head to
the set up phase of planning, the idea is to try to group as these positions using APS manual control.
many isocenters in the same run as possible, even if it means The team moves out after the position checks are com-
changing all of the isocenters to high docking or a different pleted, and the radiosurgical dose is administered. The APS
gamma angle than the default 90°. If a different gamma angle moves the patient to all the planned positions, one by one,
is used, the plan must be rechecked for accuracy and adjusted until the isocenters using that size of collimator helmet are
if necessary. In the current version of LGP, multiple targets completed. The team monitors the patient and the coordinates
of different isocenters on the control computer. If other runs ability to conform to the target. The APS has also had a
using a different gamma angle but using the same helmet are significant impact in reducing the risk of human error and
planned, then the patient is taken out, the next run is selected, increasing the rate of treatment. In the future as these
APS is moved to the dock position, and the patient’s head is and many other modifications become more utilized,
again fixed in the APS using the planned angle (72, 90, 110, it should provide better treatments, resulting in better
or 125°). Position checks are performed, and procedure is patient outcomes.
begun. Similarly, if additional runs using different helmets
are planned, the helmet is changed, the patient’s head is posi-
tioned in the APS, and the position checks for all the isocen- References
ters for that helmet and gamma angle are performed.
Radiosurgery with the LGK PERFEXION is a fully auto- 1. Leksell L. The stereotaxic method and radiosurgery of the brain.
Acta Chir Scand. 1951;102:316–9.
mated process for all aspects of the procedure including set
2. Ammar A. Lars Leksell’s vision—radiosurgery. Acta Neurochir
up of the stereotactic coordinates, setup of different sector Suppl. 1994;62:1–4.
positions defining collimator size or blocked beams, and 3. Lunsford LD, Flickinger J, Lindner G, Maitz A. Stereotactic radio-
setup of exposure times. All treatment data are exported to surgery of the brain using the first United States 201 cobalt-60
source gamma knife. Neurosurgery. 1989;24:151–9.
the operating console. The only manual part of the procedure
4. Leksell L. Cerebral radiosurgery. I. Gammathalanotomy in two
is the positioning of the patient’s head in the docking cases of intractable pain. Acta Chir Scand. 1968;134:585–95.
device and adjustment of the couch height for optimal 5. Niranjan A, Lunsford LD. Radiosurgery: where we were, are, and
comfort. After confirmation of the patients’ identity, most may be in the third millennium. Neurosurgery. 2000;46:531–43.
6. International Leksell Gamma Knife Society. Indications treated
PERFEXION radiosurgeries are administered in one single
1991–2011. 1996. http://www.lgksociety.com.
run (95 %). Rarely a clearance check is needed. For this, a 7. Niranjan A, Madhavan R, Gerszten PC, Lunsford LD. Intracranial
special test tool simulating the shape and dimensions of the radiosurgery: an effective and disruptive innovation in neurosur-
inner collimator is attached and rotated around patient’s gery. Stereotact Funct Neurosurg. 2012;90:1–7.
8. Goetsch SJ. Risk analysis of Leksell Gamma Knife Model C with
head. Once radiosurgery begins, the team monitors the
automatic positioning system. Int J Radiat Oncol Biol Phys.
patient and the setup of coordinates, exposure times, and sec- 2002;52:869–77.
tor setup of different isocenters on the control computer of 9. Lindquist C, Paddick I. The Leksell Gamma Knife Perfexion and
operating console. The system allows audiovisual communi- comparisons with its predecessors. Neurosurgery. 2007;61:130–40;
discussion 140–1.
cation with the patient during irradiation and the process can
10. Niranjan A, Novotny Jr J, Bhatnagar J, Flickinger JC, Kondziolka
be interrupted at any time if needed. D, Lunsford LD. Efficiency and dose planning comparisons
between the Perfexion and 4C Leksell Gamma Knife units.
Stereotact Funct Neurosurg. 2009;87:191–8.
11. Novotny J, Bhatnagar JP, Niranjan A, Quader MA, Huq MS,
Conclusion Bednarz G, et al. Dosimetric comparison of the Leksell Gamma
Knife Perfexion and 4C. J Neurosurg. 2008;109(Suppl):8–14.
In the past 25 years, we have witnessed dramatic improve- 12. Regis J, Hayashi M, Porcheron D, Delsanti C, Muracciole X,
ments in the stereotactic radiosurgery technologies. Gamma Peragut JC. Impact of the model C and Automatic Positioning
System on gamma knife radiosurgery: an evaluation in vestibular
Knife radiosurgery now offers better image handling, faster
schwannomas. J Neurosurg. 2002;97:588–91.
and more compact platforms that make the calculations 13. Yu C, Jozsef G, Apuzzo ML, MacPherson DM, Petrovich Z. Fetal
almost real time. The advent of PERFEXION allows for radiation doses for model C gamma knife radiosurgery.
expanded indications and accessibility to lesions that were Neurosurgery. 2003;52:687–93; discussion 693.
14. Niranjan A, Maitz AH, Lunsford A, Gerszten PC, Flickinger JC,
challenging with previous systems. Furthermore, the ability
Kondziolka D, et al. Radiosurgery techniques and current devices.
to mix collimators within a single shot adds to the surgeons’ Prog Neurol Surg. 2007;20:50–67.
LINAC: Past, Present, and Future
of Radiosurgery 9
Maryam Rahman, Gregory J.A. Murad, Frank J. Bova,
and William A. Friedman
of knowledge about both targeting the lesion and radia-

Introduction tion delivery. This review covers the history of the develop-
ment of LINACs, the modifications necessary to deliver
Modern neurosurgery has witnessed a surge of new technol- radiosurgery, and current and future applications of LINAC
ogies and minimally invasive techniques that attempt to min- radiosurgery.
imize tissue damage and patient recovery times from
neurosurgical interventions. With improving accuracy of
imaging techniques, neurosurgeons are capable of delivering Radiosurgery
targeted treatment that causes little damage to surrounding
tissues without compromising efficacy. The search for mini- Regardless of the source of radiation, radiosurgical funda-
mally invasive neurosurgical treatment has led to the devel- mental concepts include the following [3]:
opment of the operating microscope, endovascular treatment, 1. A very high dose of radiation is delivered (usually in one
and endoscopic surgery. One of the most exciting discoveries treatment).
made during this search is the use of targeted, high-dose 2. A steep dose gradient is achieved with minimal dose to
radiation for neurosurgical disorders. surrounding structures.
Radiosurgery is any method for stereotactically focusing 3. The target is localized stereotactically.
multiple beams of radiation on a target [1]. Initially described 4. Computerized dosimetry planning is employed.
with the Gamma Knife, stereotactic radiosurgery (SRS) is 5. The radiation delivery system is very accurate.
now commonly delivered by linear accelerators (LINACs). Radiosurgery refers to a single session surgical procedure
To deliver SRS, modifications were made to LINACs for that delivers ionizing radiation to a target volume with accu-
radiosurgical use [2]. The LINAC is now the most frequently rate preplanning of three-dimensional (3D) isodose surface
used device for delivery of conventional radiotherapy contours [4, 5]. These concepts require precise knowledge of
and SRS. both the target volume and the behavior of the therapeutic
Radiosurgery is truly minimally invasive, delivering ther- energy beam [6].
apeutic energy to an accurately defined target without an The basis for SRS was conceived over 60 years ago by Lars
incision. It has been used to treat a wide variety of pathologic Leksell [1]. His team’s implementation of these concepts cul-
conditions including benign and malignant brain tumors, minated in the development of the Gamma Knife. The modern
vascular lesions such as arteriovenous malformations Gamma Knife employs cobalt radiation sources in a fixed
(AVMs), and pain syndromes such as trigeminal neuralgia. hemispherical array, such that all photon beams are focused
The last 50 years has produced a tremendous amount on a single point. The patient is stereotactically positioned in
the Gamma Knife so that the intracranial target coincides with
M. Rahman, M.D., M.S. (*) the isocenter of radiation. Using variable collimation, beam
Assistant Professor, Department of Neurosurgery, Preston A. Wells blocking, and multiple isocenters, the radiation target volume
Jr Center for Brain Tumor Therapy, University of Florida, is shaped to conform to the intracranial target.
PO Box 100265, Gainesville, FL, USA
An alternate radiosurgical solution is use of the linear
e-mail: mrahman@ufl.edu
accelerator (LINAC). Most LINAC radiosurgical systems
G.J.A. Murad, M.D. • F.J. Bova, Ph.D. W.A. Friedman, M.D. (*)
rely on the same basic paradigm: a collimated X-ray beam is
Department of Neurosurgery, University of Florida,
Gainesville, FL, USA focused on an intracranial target. The development and mod-
e-mail: friedman@neurosurgery.ufl.edu ification of LINAC for radiosurgery will be discussed.
122 M. Rahman et al.
collimated by a set of secondary and tertiary collimators [9].

The Early LINAC The photon then transfers energy and ionizes atoms within
tissue. These ionizing events lead to molecular changes.
Photon beam radiation was proposed by Swedish physicist Photon beam radiotherapy differs from particle beam radia-
Gustav Ising and subsequently developed by Rolf Wideroe, a tion which propagates particles such as protons, neutrons,
Norwegian physicist working in Switzerland in 1928 pions, and heavy charged particles through tissue. These
(Table 9.1) [7, 8]. He described a series of co-linear tubes charged particles directly disrupt the atomic structure of the
connected to a high-frequency generator. This apparatus material they are traversing and thereby cause biological
became widely used in experimental physics for heavy par- change [10]. The expense of particle beam units has partly
ticles, where increase in velocity is modest. However, elec- driven the development of LINACs for radiosurgery.
tron velocity increased so rapidly due to the small size of In 1938, William Hansen at Stanford described the con-
electrons, the tubes could not be long enough and the system cept of accelerating electrons by passing them repeatedly
was impractical for medical applications [8]. through a resonant microwave cavity, gaining velocity on
World War II drove the need for microwave technology each pass. He called this a “Rhumbatron” [11]. It was only
for military radar equipment. Innovation in this field led to improvements of microwave generators, necessitated by
the development of the modern LINAC. LINACs produce World War II, that enabled this concept to become a reality.
photon beams very differently from the Gamma Knife. These discoveries, along with Harry Boot and John Randall’s
Instead of decay of cobalt, LINACs use a microwave genera- creation of the “magnetron” in the UK and the Varian broth-
tor to accelerate electrons within a waveguide. The wave- ers’ development of the “klystron” in the USA, both in 1939
guide bunches the electrons onto a portion of the wave, [8], led to the development of microwave LINACs in both
where they can be efficiently accelerated up to 99.9 % of the the UK and the USA by 1945 [12]. All of these devices
speed of light. Once the electrons reach their full accelerat- worked similarly to accelerate electrons. The groups led by
ing potential, they collide with a heavy metal target. The Don Fry at the Atomic Energy Research Establishment and
energy generated from this collision is mostly lost as heat. Hansen at Stanford both described clinically workable
However, a small percentage of the electrons pass near the LINACs. Interestingly, the groups had little knowledge of
large nuclei of the metal target, are deflected, and undergo a each other’s work until the late 1940s [8].
change in acceleration. This interaction results in the emis-
sion of a photon from the electron, i.e., electromagnetic radi-
ation [9]. “Bremsstrahlung,” German for “braking radiation,” LINAC and Radiosurgery
is used to describe the production of radiation from deceler-
ating or “braking” electrons. Shortly after World War II, Leksell first used the term radio-
Once the photon beam is produced, it is limited by a pri- surgery [1]. This concept sprouted from an idea Leksell had
mary collimator, passes through a flattening filter to improve discussed with Sir Hugh Cairns at the first Scandinavian neu-
the spatial uniformity of the beam, passes through two inde- rosurgical meeting in Oslo after the war. He discussed his
pendent monitoring ionization chambers, and can then be concerns about then-available neurosurgical techniques and
Table 9.1 Early history of LINAC radiosurgery

Date Primary author Location Event
1928 R. Wideroe Baden, Switzerland Developed photon beam radiation with high-frequency generator
1938 W. Hansen Stanford, USA Description of “rhumbatron” using a resonant microwave unit to accelerate electrons
1939 H. Boot & J. Randall Birmingham, UK Creation of “magnetron” using resonant microwave for radar during WWII
1939 R. & S. Varian Stanford, USA Creation of “klystron” using resonant microwave for radar during WWII
1948 W. Hansen Stanford, USA Development of microwave LINAC
1953 C.W. Miller, D.W. Fry Great Malvern, UK Development of micowave LINAC
1957 L. Leksell Stockholm, Sweden Treatment of patients with Gamma Knife
1983 O. Betti & V. Derechinsky Buenos Aires, Argentina LINAC radiosurgery with Talairach stereotactic localization
1984 M.D. Heifetz Los Angeles, USA Description of high-dose radiation with LINAC to small targets in the brain
1984 F. Colombo Vicenza, Italy Description of stereotactic LINAC radiosurgery system
1985 G.H. Hartmann Heidelberg, Germany Modified stereotactic localization to deliver multiple arc radiosurgery treatments
1985 M.S. Ginsberg Miami, USA First description of LINAC radiosurgery in the USA
1988 K. Winston & W. Lutz Boston, USA LINAC radiosurgery with phantom target device to check accuracy
1989 W.A. Friedman & F. Bova Gainesville, USA High-precision bearings to control patient and gantry movements to improve beam
accuracy to 0.2 ± 0.1 mm
9 LINAC: Past, Present, and Future of Radiosurgery 123
his plans to mechanically direct a probe or narrow beam of of secondary circular collimators of varying size are used to
X-ray or ultrasound into the brain to ablate pathways for pain conform the beam [20].
alleviation. Cairns gave him a positive response and Leksell Modifications of LINACs were necessary to perform
began systematically investigating ionizing beams [13]. He radiosurgery including a system to rotate the couch in syn-
and his colleagues tested an orthovoltage X-ray tube and pro- chrony with the gantry, collimator development, and stereo-
ton beam produced by the cyclotron in Uppsala, calling pro- tactic localization [21]. Several factors made LINAC
ton radiosurgery “stralkniven” (ray knives) [14]. They also desirable for radiosurgery delivery. LINACs were used
considered LINAC as a potential radiation source. Ultimately, widely in the USA for conventional radiotherapy, and modi-
they constructed the “Gamma Knife” using cobalt radiation fying the LINAC for radiosurgery was much more cost-
[15]. Leksell began treating patients in 1957 [3, 14]. Other effective than purchasing a Gamma Knife. Gamma Knife
investigators worked with particle beam radiosurgery sys- units at the time were only capable of delivering radiosurgery
tems in Berkeley and Boston [16, 17]. and could not deliver conventional radiotherapy when not
Early radiosurgery researchers were aware of the poten- being used for radiosurgery. Additionally, extracranial radio-
tial for LINAC in delivering targeted radiation. Larsson, the surgery treatment was believed to be more feasible using
head physicist in the collaborative group at Uppsala LINACs. The theoretical benefits of LINAC discussed by
University with Leksell, wrote in 1974, “The choice between radiosurgery leaders in the past have now become a reality.
the two alternatives, e.g., roentgen or gamma radiation, Within 10 years of LINAC development, reports appeared
should be based on technical, clinical and economical rather on the use of external beam radiation for radiosurgery [12].
than physical considerations. If radiation surgery will reach In 1983 Oswaldo Betti and Victor Derechinsky reported the
a position as a standard procedure, improved electron accel- development of a multibeam LINAC coupled with a Talairach
erators for roentgen production, adapted for the purpose, stereotactic localization system in Buenos Aires [12, 22].
would seem a most attractive alternative” [18]. They used circular collimators that could be oriented in mul-
The initial LINACs developed in the UK and the USA tiple coronal planes of a patient sitting in a moveable chair
used traveling-wave technology [8]. They were limited by while attached to a rotating head frame [6, 22, 23]. Their
relatively low beam energy, low radiation output due to inef- system uniquely had the patient in a sitting position [3].
ficient waveguide design and limited microwave power, and In 1984, a standard LINAC with small modifications was
restricted range of movement. These systems were also used by Heifetz et al. (with Marilyn Wexler’s physics contri-
large, with the top of the gantry being approximately 4 m off butions) to deliver high-dose radiation to small targets spar-
the floor, to accommodate the accelerating waveguide. The ing normal brain, similar to Leksell’s Gamma Knife [24].
next generation of devices achieved full isocentric rotation Simultaneously, a neurosurgeon, Federico Colombo, and a
by mounting the accelerating structure to a gantry, allowing group of physicists led by Renzo Avanzo in Vicenza, Italy,
the radiation source to be rotated in a vertical plane about a reported their stereotactic LINAC radiosurgery system [25].
single point. Also, by mounting the waveguide horizontally They wrote about radiosurgical dose schemes of 40–50 Gy
with magnetic beam deflection (beam bending), the next over two fractions separated by 8–10 days for various intra-
generation of LINACs had more manageable heights. The cranial targets of 2–4 cm in diameter [26, 27]. The dose gra-
next major improvement came in 1968 when Knapp et al. dient achieved compared well with Gamma Knife data [27, 28].
developed the side-coupled standing wave structure which Hartmann et al. in Heidelberg, Germany, followed these
improved shunt impedances so that the total length of the achievements with the description of a modified stereotactic
accelerating structure was greatly reduced [19]. This devel- localization and positioning system to deliver multiple arc
opment did away with the need for beam bending (which had radiosurgery treatments [29]. They modified a Riechert–
its own problems) and still allowed for full isocentric rota- Mundinger stereotactic device, using laser lights to position
tion at a practical height [8]. Currently, the US and Japanese the frame within the isocenter.
manufacturers use standing-wave technology, while the UK The first published work on LINAC radiosurgery in the
manufacturers continue to use traveling-wave technology. USA came from the University of Miami in 1985 [30].
All LINAC radiosurgery systems focus a collimated X-ray However, their system relied on the jaws of the treatment
beam on a stereotactically identified target. The gantry of the machine for beam collimation instead of the secondary col-
LINAC rotates over the patient, producing an arc of radiation limation system described by Larsson et al. [18]. Their tech-
focused on the target. The patient couch is rotated in the hori- nique was regarded as fractionated, rather than single fraction.
zontal plane and another arc is performed. In this manner, One of the first solutions for the requirement to spread out the
multiple noncoplanar intersecting arcs of radiation are pro- radiation entrance path and minimizing treatment delivery
duced. Like the Gamma Knife, the intersecting arcs produce time was described by Ervin Podgorsak at McGill University
a high target dose, with minimal radiation to the surrounding [31]. He and his colleagues modified a LINAC using extra
brain [3]. Along with the LINAC rectangular collimator, a set collimators to define small circular fields and simultaneous
Fig. 9.2 Sphere packing technique. Dose planning has traditionally

used the sphere packing technique originally developed by Lars Leksell.
In this technique, sets of beams of radiation are aimed at the isocenter.
Fig. 9.1 Conformal dose. Nonspherical lesions can be treated with The beams are selected to reach the isocenter via unique paths. The
minimal dose to surrounding structures. The colored lines are as fol- resultant dose distribution is spherical. To cover the entire target vol-
lows: red = 70 % isodose, green = 50 % isodose, yellow = 20 % isodose ume, the initial dose sphere is the largest sphere that fits inside the target
volume. The target volume is then “packed” with equal or smaller
diameter until adequate target coverage is achieved
gantry and couch rotations. Additionally, the couch and
gantry were monitored from the control area, eliminating the
need to enter the room during treatment [32]. Due to increased This and other discoveries eventually led to the development
error rates with simultaneous gantry and couch rotation, most of a micromultileaf collimator from Varian and BrainLAB
institutions have not adopted this system. [36]. These collimators have changed treatment delivery
Concerned with error and quality control, Winston and from multiple noncoplanar arcs to fixed static fields and
Lutz published their work on multiple arc LINAC SRS in dynamic arcs and allow for treatment of larger and more geo-
1988 [33]. Their system included a phantom target device metrically complex lesions.
that could easily be used to check the accuracy of each patient LINACs are continuously being modified to improve
treatment as well as to evaluate sources of error. They found radiosurgery delivery. For example, John Adler and Richard
a mechanical accuracy of their system of 0.5 ± 0.2 mm. They Cox at Stanford University reported the development of an
suggested that the major error in any radiosurgery system industrial robot combined to a LINAC, called the Cyberknife®
was the error of localization and not mechanical error [3]. In (Accuray Inc., Sunnyvale, CA) [37]. This system can posi-
1989, Friedman and Bova reported on LINAC radiosurgery tion a circularly collimated beam of X-rays to a target from a
system at the University of Florida [34]. A portable add-on range of positions and angles. Moreover, the Cyberknife®
stereotactic devise was coupled to the LINAC, and high-pre- uses room-mounted imaging to localize the treatment isocen-
cision bearings in the device controlled all patient and gantry ter before treatment [38]. Other systems such as the Trilogy™
movements. As a result, the radiation beam accuracy was (Varian Medical Systems, Inc., Palo Alto, CA) also localize
improved to 0.2 ± 0.1 mm. The accuracy of treatment deliv- the isocenter prior to treatment (Fig. 9.3) [39]. Real-time
ery was further increased with imaging software improve- imaging of patients can be used to readjust beam coordinates
ments and the ability to fuse CT and MRI images [34]. for the target, making treatment of extracranial sites such as
As LINAC radiosurgery became more prevalent, increas- spine and abdomen possible [12, 20]. Spinal radiosurgery is
ingly larger and more complex lesions were being treated being implemented for benign and malignant tumors, over-
(Fig. 9.1). The circular collimators were still useful for these coming the tremendous difficulty of target localization in an
types of lesions using a sphere packing technique (Fig. 9.2). area with movement of multiple joints [40–43]. In addition to
An alternative approach to sphere packing was first described modifications to collimation and target localization, LINACs
by Dennis Leavitt who built the first dynamic field shap- have been modified to deliver SRS with intensity modulated
ing collimator for radiosurgery in 1989 [35]. The circu- radiation therapy (IMRT) [12]. LINACs provide tremendous
lar collimators were supplemented with independent possibilities for unique approaches to treatment. These sys-
rectangular vanes that “trimmed” the circular radiation field. tems demonstrate the versatility of LINAC for radiosurgery.
is typically well demarcated from surrounding tissues on

neuroimaging studies. The sharp borders of this noninvasive
tumor make it a convenient match for the characteristically
steep dose gradient produced at the boundary of a radiosurgi-
cal target. This allows the radiosurgeon to minimize radia-
tion of normal tissue. Excellent spatial resolution on
gadolinium-enhanced MRI facilitates radiosurgical dose
planning. These tumors typically occur in an older popula-
tion that may be less fit for microsurgical resection under
general anesthesia. Finally, the location of these tumors at
the skull base in close proximity to multiple critical neuro-
logic structures (i.e., cranial nerves, brain stem) leads to
appreciable surgical morbidity and rare mortality even in
expert hands. This makes the concept of an effective, less
invasive, less morbid alternative treatment that can be per-
formed in a single day under local anesthesia quite attractive.
Whether or not radiosurgery fits this description has been
extensively debated.
Certainly, the role of radiosurgery is limited by its inabil-
ity to expeditiously relieve mass effect in patients for whom
this is necessary. The radiobiology of SRS also requires
Fig. 9.3 The Trilogy™ system. This system includes inline CT
for target localization prior to treatment allowing for extracranial lower, potentially less effective doses for higher target vol-
radiosurgery targeting umes in order to avoid complications. This limits the use of
SRS to the treatment of smaller tumors. Despite these limita-
tions, there is substantial literature demonstrating radiosur-
gery is a safe and effective alternative therapy for acoustic
Radiosurgery for Benign Tumors schwannomas.
Spiegelmann et al. have reported their experience in 44
SRS has proved useful for the treatment of a variety of VS patients treated with LINAC SRS from 1993 to 1997
benign intracranial neoplasms. These tumors commonly [48]. CT scanning was selected as the stereotactic imaging
arise from the skull base, where their dramatic impact on modality for target definition. A single, conformally shaped
quality of life belies their benign histology and small size. isocenter was used in the treatment of 40 patients; two or
Despite progressive improvement in microsurgical tech- three isocenters were used in four patients who harbored
niques, outcomes for patients with these difficult tumors very irregular tumors. The radiation dose directed to the
continue to be less than optimal [44–46]. A significant tumor border was the only parameter that changed during
amount of experience has been accumulated using SRS in the study period: In the first 24 patients who were treated the
the treatment of schwannomas and meningiomas. We will dose was 15–20 Gy, whereas in the last 20 patients the dose
focus on each of these tumor types in turn. was reduced to 11–14 Gy. After a mean follow-up period of
32 months (range, 12–60 months), 98 % of the tumors were
controlled. The actuarial hearing preservation rate was 71 %.
Vestibular Schwannomas New transient facial neuropathy developed in 24 % of the
patients and persisted to a mild degree in 8 %. Radiation
Among benign intracranial tumors, vestibular schwannoma dose correlated significantly with the incidence of cranial
(acoustic neuroma) has been one of the most frequent targets neuropathy, particularly in large tumors (≥4 cm3).
for SRS. This common tumor (representing approximately Fractionated stereotactic radiation therapy (FSRT) has
10 % of all primary brain tumors) is a benign proliferation of been used as an alternative management for vestibular
Schwann cells arising from the myelin sheath of the vestibu- schwannomas. This method is proposed as a way of exploit-
lar branches of the eighth cranial nerve. These tumors are ing the precision of stereotactic radiation delivery to mini-
slightly more common in women, present at an average age mize dose to normal brain while employing lower fractionated
of 50 years, and occur bilaterally in patients with neurofibro- doses in an effort to minimize complications. Litre et al.
matosis (NF) type 2. recently reported on their experience with 155 VS patients
Leksell first used SRS to treat a vestibular schwannoma treated with fractionated LINAC SRS [49]. The patients
in 1969 [47]. SRS is a logical alternative treatment modality received five fractions of 1.8 Gy weekly for a total central
for this tumor for several reasons. A vestibular schwannoma dose of 55 Gy. Local tumor control rates were 99.3 % at
Fig. 9.4 Pretreatment MRI scan shows left-sided vestibular schwannoma

Fig. 9.5 Four years after treatment, the MRI scan shows the schwan-
noma of Fig. 9.4 to be much smaller
3 years, 97.5 % at 5 years, and 95.2 % at >7 year follow-up.
Tinnitus (70 %), vertigo (59 %), imbalance (46 %), and ear
mastoid pain (43 %) greatly improved post SRS. Median follow-up was 40 months and 65 % patients demon-
Complications included facial numbness (3.2 %), facial strated tumor shrinkage, 22 % had stable VS size, and 13 %
weakness (2.5 %), worsened tinnitus (2.1 %), and need for had growth. In 54 % of all patients, transient tumor swelling
ventriculoperitoneal shunt (2.5 %). Thus far, most radiosur- was observed.
geons feel that optimal results can be achieved with highly LINAC SRS is a well-described effective treatment for
conformal single-fraction radiosurgery while sparing the smaller VS tumors and has a lower complication rate com-
patient the inconvenience of a prolonged treatment course. pared to surgery.
Friedman et al. performed an analysis of 390 VS patients
treated with LINAC SRS at University of Florida (UF) from
July 1988 to August 2005 [50]. With a median follow-up of Meningiomas
32 months for the entire group, most tumors were unchanged
or smaller (Figs. 9.4 and 9.5), and only 11 (4 %) tumors were Meningiomas are the most common benign primary brain
larger. The 1- and 2-year actuarial control rates were both tumor, with an incidence of approximately 7/100,000 in the
98 %, and the 5-year actuarial control rate was 90 %. Four general population. Surgery has long been thought to be the
patients (1 %) required surgery for tumor growth. Seventeen treatment of choice for symptomatic lesions and is often
patients (4.4 %) reported facial weakness and 14 patients curative. Many meningiomas, however, occur in locations
(3.6 %) reported facial numbness after SRS. The risk of these where attempted surgical cure may be associated with mor-
complications rose with increasing tumor volume or radio- bidity or mortality, such as the cavernous sinus or petroclival
surgical dose to tumor periphery. Since 1994, when doses region [52, 53]. In addition, many of these tumors occur in
were deliberately reduced to 1,250 cGy, only 2 patients the elderly, where the risks of general anesthesia and surgery
(0.7 %) experienced facial weakness and only 2 patients are known to be increased. Hence, there is interest in alterna-
(0.7 %) developed facial numbness. Based on this and previ- tive treatments, including radiation therapy and SRS, either
ous studies, the authors currently recommend a peripheral as a primary or as an adjuvant approach.
dose of 12.5 Gy for almost all acoustics as that dose most Simpson, in a classic paper, described the relationship
likely to yield long-term tumor control without causing cra- between completeness of surgical resection and tumor recur-
nial neuropathy. rence [54]. A grade I resection, which is complete tumor
Recently, van de Langenberg et al. described 37 removal with excision of the tumor’s dural attachment and
VS patients treated with LINAC SRS with a 4-year probabil- involved bone, has a 10 % recurrence rate. A grade II resec-
ity of no additional intervention of 96.4 % ± 0.03 [51]. tion, complete resection of the tumor and coagulation of its
dural attachment, has up to a 20 % recurrence rate. Grade III

resection is complete tumor removal without dural resection
or coagulation. Grade IV resection is subtotal, and grade V
resection is simple decompression. Recurrence rates in
grades IV and V groups basically reflect the natural history
of the tumor, with high rates of recurrence over time.
Unfortunately, some common meningioma locations, such
as the cavernous sinus or petroclival region, are not readily
amenable to a complete dural resection or coagulation strat-
egy because of location and the proximity of vital neural and
vascular structures. In addition, relatively high complication
rates have been described for meningioma surgery in some
locations and in the elderly.
Pollock and colleagues recently analyzed 198 patients
with meningiomas less than 35 mm in diameter treated with
either surgical resection or Gamma Knife radiosurgery [55].
Tumor recurrence was more frequent in the surgical
resection group (12 % vs. 2 %). No statistically significant
difference was detected in the 3- and 7-year actuarial pro-
gression-free survival rate between patients with Simpson
grade 1 resections and those who underwent radiosurgery. Fig. 9.6 A patient with known breast carcinoma presented with symp-
Progression-free survival rates with radiosurgery were supe- tomatic pontine lesions. She was treated with radiosurgery (16 Gy to
the 80 % isodose line)
rior to Simpson grades 2, 3, and 4 resections. Complications
were lower in the radiosurgery group.
Multiple LINAC SRS series have been published [56–59].
Hakim and colleagues described one of the largest such
series, and the only one to report actuarial statistics [60]. One
hundred twenty-seven patients with 155 meningiomas were
treated. Actuarial tumor control for patients with benign
tumors was 89.3 % at 5 years. Six (4.7 %) patients had per-
manent radiation-induced complications.
The University of Florida report on LINAC SRS treat-
ment of meningiomas is one of the largest yet published [61].
Two hundred and ten patients were treated from May 1989 to
December 2001. All patients had follow-up for a minimum
of 2 years, and no patients were lost to follow-up. Actuarial
local control for benign tumors was 100 % at 1 and 2 years
and 96 % at 5 years (Figs. 9.6 and 9.7). Actuarial local con-
trol for atypical tumors was 100 % at 1 year, 92 % at 2 years,
and 77 % at 5 years. Actual control for malignant tumors was
100 % at 1 and 2 years but only 19 % at 5 years. Permanent
radiation-induced complications occurred in 3.8 %, all of
which involved malignant tumors. These tumor control and
treatment morbidity rates compare well with all other pub-
lished series. Fig. 9.7 Three years later, the site of the lesion of Fig. 9.6 is barely visible
We found that reliance on imaging characteristics rather
than surgical pathology did not yield a high incidence of Han et al. recently compared LINAC SRS to fractionated
missed diagnoses. During the time interval of this study, only radiation in the treatment of skull base meningiomas. A total
two patients were treated as presumed meningiomas and of 220 skull base meningiomas were treated using SRS
later found to have other diagnoses. One had a dural-based (n = 55), hypofractionated stereotactic radiation therapy
metastasis that was surgically excised when it enlarged. The (hFSRT) (n = 22), and FSRT (n = 143). Median follow-up
other had a hemangiopericytoma of the lateral cavernous was 32 months, and the median tumor volumes were 2.8 cm
sinus that was surgically excised when it enlarged. for SRS, 4.8 cm for hFSRT, and 11.1 cm for FSRT.
The median treatment doses were 1,250 cGy in one fraction aggressive treatment of metastatic disease has been shown to
for SRS, 2,500 cGy in five fractions for hFSRT, and 5,040 cGy restore neurologic function and prevent further neurologic
in 28 fractions for FSRT. Radiographic control was achieved manifestations. Debate exists concerning the optimum treat-
in 91 % of SRS patients, 94 % of hFSRT patients, and 95 % ment for metastatic brain disease.
of FSRT patients. The authors concluded that there was no In autopsy series, brain metastases occur in up to 50 % of
differences in clinical or radiologic response to the different cancer patients [65]. Approximately 30–40 % of patients
radiation strategies, although these data are limited due to present with a solitary metastasis. Brain metastases fre-
variability of patients within each group and the slow growing quently cause debilitating symptoms that can seriously
natural history of skull base meningiomas [62]. impact the patient’s quality of life. With no treatment or ste-
One of the advantages of LINAC SRS is the flexibility of roid therapy alone, survival is limited (1–2 months). Whole-
the system to treat systemic lesions. Increasingly, radiosur- brain radiotherapy (WBRT) extends median survival, but the
geons are reporting on their experiences with treating spinal duration of survival is typically low (3–4 months). Several
meningiomas with SRS. Recently, Gerszten et al. described randomized trials have suggested that, when possible, sur-
the treatment of 45 benign spine tumors with LINAC SRS gery followed by WBRT is superior to WBRT alone. Patchell
and CT for target localization [63]. They treated 14 cervical, et al. reported a randomized clinical trial involving 46
12 thoracic, 14 lumbar, and 5 sacral tumors. The majority of patients with a single metastasis and well-controlled sys-
the tumors (91 %) were intradural. The mean maximum dose temic disease [66]. They found a significant improvement in
to the tumor volume was 16 Gy (12–24 Gy) given in a single survival (40 vs. 15 weeks) and local recurrences in the CNS
fraction in 39 cases. Median follow-up was 32 months. Of (20 % vs. 52 %) for patients in the surgery plus WBRT arm
the 19 (42 %) patients who had pretreatment pain, 15 had of the study. Likewise, Noordijk et al. randomized 66 patients
significant improvement after SRS. Of the 15 patients with a and found a significant survival advantage (10 vs. 6 months)
neurologic deficit pretreatment, four motor deficits were for the combination therapy arm [67]. In contrast, Mintz
stable, one motor deficit worsened in patients with NF 1, 10 et al. studied a group of 84 patients and did not show an
sensory deficits improved, and one sensory deficit worsened advantage of surgery plus radiotherapy over radiotherapy
in a patient with NF 1. alone [68]. It has been suggested that the inclusion of a
Longer follow-up in these patients will be necessary for higher percentage of patients with active systemic disease
an accurate assessment of the efficacy of SRS for spinal and lower performance scores did not allow the benefit of
benign tumors. improved local control to affect survival in this series.
Haines points out that survival and quality of life are the
most important outcome measures in evaluating a clinical
Radiosurgery for Malignant Tumors treatment for cancer [69]. Surrogate end points, like local
control, are inherently unreliable, especially when the defini-
Malignant tumors are radiobiologically more amenable to tion of local control is changed. This applies to a comparison
fractionated radiotherapy than benign lesions. Malignancies of SRS with surgery for brain metastasis. In surgical series,
tend to infiltrate surrounding brain, resulting in poorly defin- local control means no visible tumor on follow-up scans. In
able tumor margins. A priori, these two traits of cerebral SRS series, local control means no growth (or sometime
malignancies would seem to make SRS an unattractive treat- minimal growth) on follow-up scans. These end points are
ment option. Nevertheless, SRS has proved to be a useful unlikely to be equivalent.
weapon in the armamentarium against malignant brain tumors. In addition, comparison of current results to historical con-
The most common applications of SRS to malignant tumors trols is fraught with hazard to selection bias. This issue led to
are the treatment of cerebral metastases and the delivery of an erroneous conclusions about the efficacy of brachytherapy for
adjuvant focal radiation “boost” to malignant gliomas. malignant gliomas and to overly optimistic reports regarding
the efficacy of intraarterial chemotherapy. Of equal import is
the difficulty and variability of reporting standards for local
Cerebral Metastases control. Few series provide actuarial local control. They simply
provide a “raw” number at an arbitrary point in time. Less
Metastatic brain tumors are up to ten times more common commonly appreciated is the difficulty in documenting local
than primary brain tumors with an annual incidence of control. Many of these patients die away from the medical
between 80,000 and 150,000 new cases each year [64]. center where radiosurgery was performed. It is frequently
Fifteen percent to 40 % of cancer patients will be diagnosed impossible to determine from family or local physician tele-
with a brain metastasis during the course of their illness. phone interview whether the proximate cause of death was
Once a brain metastasis has been diagnosed, the median life loss of local control, new intracranial disease (loss of regional
expectancy is less than 1 year; however, in many patients, control), or systemic disease. Most radiosurgical series have
assumed that, unless an MRI was performed documenting More than 80 % of recurrences are found within 2 cm of the
local loss of local control prior to death, local control was original tumor site. Many attempts have been made to
maintained. This assumption may lead to a systematic overes- improve long-term survival by improving local control. Such
timation of local control rates. therapies include aggressive surgical removal, brachyther-
Sturm [70–72], Black [73, 74], and Adler [75–77] pub- apy, chemotherapy wafers, and radiosurgery.
lished early reports on linear accelerator radiosurgery for A number of linear accelerator radiosurgery series have
brain metastases. Alexander reported on 248 patients [73]. been published. Shrieve and colleagues reported on 32
Median tumor volume was 3 cm3 and median tumor dose patients receiving interstitial brachytherapy and 86 patients
was 15 Gy. Median survival was 9.4 months. Actuarial local receiving radiosurgical boost [83]. They found similar sur-
control was 85 % at 1 year and 65 % at 2 years. Auchter et al. vival rates between the two groups and recommended radio-
reported a multi-institutional study of 122 patients [78]. surgery because of its outpatient, noninvasive nature. Hall
Actuarial 1- and 2-year survivals were 53 % and 30 %, and colleagues reported 35 patients and believed that radio-
respectively. Local control was 86 %. surgery did confer a survival advantage, with fewer compli-
As radiosurgery has emerged as a treatment option, clini- cations than brachytherapy [84]. Masciopinto and colleagues
cians have attempted to define prognostic factors, which may [69] reported on 31 patients so treated and found that the
help to define patient populations most likely to benefit from “curative value of radiosurgery is significantly limited by
radiosurgical treatment. Multiple factors have been discerned peripheral recurrence [85].”
from retrospective analysis and include Karnofsky perfor- A recurring theme in all retrospective studies of brain
mance scale score, status of systemic disease, histology, tumor therapies is the question of selection bias influencing
number of metastases, volume of metastases, time interval the results of therapy more than the therapy itself. In an
between the diagnosis of the primary lesion and the meta- attempt to control for selection bias in retrospective treat-
static lesion, pattern of enhancement [79, 80], the Radiation ment trials for malignant gliomas, Curran [74] developed the
Therapy Oncology Group (RTOG) recursive partitioning RPA categories [86], and Sarkaria and colleagues used this
categories [81], and radiation dose. methodology to analyze 115 patients from three institutions
At the University of Florida 619 patients treated with SRS treated with linear accelerator radiosurgery [87]. They found
for 1,569 brain metastases have been described [82]. Median that patients treated with radiosurgery had a significantly
dose to target periphery was 1,750 cGy (range 1,000– improved 2-year and median survival compared with RTOG
2,250 cGy). Median survival was 7.9 months. Actuarial local historical controls. The improvement was seen predomi-
control was 84.3 % (Figs. 9.6 and 9.7). The 1- and 2-year nately in the worst prognostic classes (3–6 classes).
actuarial control rates were 0.82 and 0.72. Melanoma histol- At the University of Florida, we have retrospectively
ogy predicted poorer survival. Recursive partitioning analy- reviewed 100 patients with WHO grade III and IV malignant
sis (RPA) class I or II were associated with improved gliomas who received SRS boost therapy for residual or
survival. Female sex, younger age, higher Karnofsky perfor- recurrent enhancing disease [88]. The patients in our study
mance status (KPS), controlled primary tumor, absence of were divided into RPA classifications for comparison with
systemic metastases, asynchronous presentation, fewer brain historical controls. Class III and IV patients had median sur-
metastases, smaller total tumor volume, surgery prior to vival times very similar to the historical controls. Class V
SRS, and multiple SRS treatments were also associated with patients demonstrated an increase in median survival (15.6
improved survival. vs. 8.9 months) and 2-year survival rate (12.5 % vs. 6 %)
compared with historical controls. Eloquent location corre-
lated with poorer survival. This may be due to the selection
Malignant Gliomas of less aggressive therapies for this group of patients.
Recurrence at time of radiosurgery was associated with lon-
Current conventional treatment for malignant gliomas ger survival. Very probably, this reflects the fact that patients
involves a combination of surgery, radiation, and, often, che- judged “eligible” for radiosurgery at time of recurrence are
motherapy. The prognosis in these patients remains poor. already selected for longer survival than the average patient
The majority of recurrences occur within 2 cm of the enhanc- treated up front. However, it remains possible that radiosur-
ing lesion as seen on initial imaging. Gross total excision gery at time of recurrence is truly more effective than upfront
may be associated with prolonged median survival in patients radiosurgery.
with malignant gliomas. Radiosurgery is an attempt at fore- What about drawbacks of the recursive partitioning tech-
stalling local recurrence by aggressive local therapy. nique? The RTOG classes used are broad and do not include
Malignant gliomas account for approximately 40 % of the all known prognostic variables, most notably tumor size. In
17,000 primary brain tumors diagnosed annually in the addition, important linear variables like age, mental status,
USA. The prognosis for long-term survival remains poor. and KPS are converted into binary ones. This approach,
therefore, is flawed, as are all attempts at retrospective analy- The problem lies with imaging. Although angiography very
sis. Irish and colleagues, in an analysis of 101 consecutive effectively defines blood flow (feeding arteries, nidus, and
malignant glioma patients, have shown that those “eligible” draining veins), it does so in only two dimensions. Using the
for radiosurgery have a median survival of 23.4 months, two-dimensional data from stereotactic angiography to rep-
compared with 8.6 months for “ineligible” patients [89]. resent the three-dimensional target results in significant
Likewise, Curran found a marked survival advantage in errors of both overestimation and underestimation of AVM
radiosurgery “eligible” vs. “ineligible” patients [90]. nidus dimensions. Underestimation of the nidus size may
The only complete solution to the issue of selection bias result in treatment failure, and overestimation results in the
affecting outcome is a prospective randomized study. Such a inclusion of normal brain within the treatment volume. This
study has been performed and the results published in 2004. can cause radiation damage to normal brain, which—when
RTOG Study 93-05 randomized patients with glioblastoma affecting an eloquent area—may result in a neurologic defi-
into two treatment arms [91]. One received postoperative cit. To avoid such targeting errors, a true three-dimensional
radiosurgery, followed by conventional radiotherapy and image database is required. Both contrast-enhanced CT and
BCNU chemotherapy. The other arm received radiotherapy MRI are commonly used for this purpose.
and chemotherapy without radiosurgery. At a median follow- We use contrast-enhanced, stereotactic CT as a targeting
up time of 61 months, the median survival in the radiosurgery image database for the vast majority of AVMs. Our CT tech-
group was 13.5 months compared with 13.6 months in the nique employs rapid infusion (1 mL/s) of contrast while
standard treatment arm. There were no significant differences scanning through the AVM nidus with 1-mm slices. The
in 2- or 3-year survival, patterns of failure, or quality of life head ring is bolted to a bracket at the head of the CT table,
between the two groups. Notably, RTOG 93-05 did not ensuring that the head/ring/localizer complex remains immo-
address the use of radiosurgery for recurrent malignant glio- bile during the scan. This technique yields a very clear three-
mas. The main limitation of this study is that SRS was deliv- dimensional picture of the nidus. Alternative approaches use
ered prior to systemic therapy, and since the completion of MRI/MRA as opposed to CT. Attention to optimal image
this trial, standard therapy for glioblastoma has become radia- sequences in both CT and MRI is essential for effective
tion and temozolamide due to the seminal paper in 2005 [92]. AVM radiosurgical targeting.
Dose Selection
Arteriovenous Malformations Various analyses of AVM radiosurgery outcomes have eluci-
dated an appropriate range of doses for the treatment of
Patient Selection AVMs. We prefer to deliver a dose of 20 Gy to the periphery
Open surgery is generally favored if an AVM is amenable to of the AVM nidus whenever possible. Larger AVMs, or those
low-risk resection (e.g., low Spetzler–Martin grade, young in critical locations, may require a lower dose—but this will
healthy patient) or is believed to be at high risk for hemor- reduce the chances of complete obliteration.
rhage during the latency period between radiosurgical treat-
ment and AVM obliteration (e.g., associated aneurysm, prior Follow-Up
hemorrhage, large AVM with diffuse morphology, venous Standard follow-up after AVM radiosurgery typically con-
outflow obstruction). Radiosurgery is favored when the AVM sists of annual clinic visits with MRI/MRA to evaluate the
nidus is small (<3 cm) and compact, when surgery is judged effect of the procedure and monitor for neurologic complica-
to carry a high risk or is refused by the patient, and when the tions. If the patient’s clinical status changes, he/she is fol-
risk of hemorrhage is not believed to be extraordinarily high. lowed more closely at clinically appropriate intervals.
Endovascular treatment, although rarely curative alone, Each patient is scheduled to undergo cerebral angiogra-
may be useful as a preoperative adjunct to either microsur- phy at 3 years after radiosurgery, and a definitive assessment
gery or radiosurgery. The history, physical examination, and of the success or failure of treatment is made based on the
diagnostic imaging of each patient are evaluated and the results of angiography. If no flow is observed through the
various factors outlined above are weighed in combination to AVM nidus, the patient is pronounced cured and is dis-
determine the best treatment approach for a given case. The charged from follow-up. If the AVM nidus is incompletely
decision about optimal AVM treatment is best made by a obliterated, appropriate further therapy (most commonly
multidisciplinary team composed of experts in operative, repeat radiosurgery on the day of angiography) is prescribed,
endovascular, and radiosurgical treatment. and the treatment/follow-up cycle is repeated.
Stereotactic Image Acquisition The University of Florida Experience

The most problematic aspect of AVM radiosurgery is target From May 18, 1988 to March 22, 2005, 544 patients with
identification. In some series, targeting error is listed as AVMs were treated at the University of Florida. The mean
the most frequent cause of radiosurgical failure [93, 94]. age was 40 (range, 4–78 years). The median treatment
volume was 7 cm3 (range, 2–45.3 cm3). Many patients early (documented persistence of AVM flow 3 years after retreat-
in the series were treated with single isocenters (259), but in ment), and death. The mean original lesion volume was
recent years an effort has been made to produce highly con- 13.8 cm3 and the mean volume at retreatment was 4.7 cm3,
formal plans by employing multiple isocenters. The median for an average volume reduction of 66 % after the initial
radiation dose to the periphery of the AVM was 1,750 cGy “failed” treatment. Only two (3.8 %) AVMs failed to demon-
and the mean follow-up duration was 31 months. strate size reduction after primary treatment. The median
Presenting symptoms included the following: headache/ doses on initial and repeat treatment were 12.5 and 15 Gy,
incidental (188), seizure (227), hemorrhage (179), progres- respectively. To date, 25 retreated patients have reached a
sive neurological deficit [23]. Spetzler–Martin scores were definitive end point. These include 15 (60 %) angiographi-
as follows: (1) 29; (2) 188; (3) 228; (4) 98. AVMs were fur- cally documented cures, nine (36 %) angiographically docu-
ther delineated into four nidus volume categories: (a) <1 cm3; mented failures, and one fatal hemorrhage. A single
(b) 1–4 cm3; (c) 4–10 cm3; (d) >10 cm3. Angiographic/MRI permanent radiation-induced complication occurred among
cure rates were as follows: (a) 92 %; (b) 79 %; (c) 64 %; and 52 (1.9 %) patients, and 1 patient experienced a transient
(d) 36 %. There was a dramatic increase in cure rates when deficit that resolved with steroid therapy. Two hemorrhages
the peripheral dose was raised to a least 15 Gy. There was a (one fatal) occurred during a total of 130 patient-years at
dramatic decrease in cure rate when AVM size exceeded risk, resulting in a 1.5 % annual incidence of posttreatment
10 cm3 (size D). hemorrhage. If one includes retreatments in the analysis of
In 2003, we determined which factors were statistically radiosurgical success, the results are as follows: (a) 100 %;
predictive of radiographic and clinical outcomes in the radio- (b) 92 %; (c) 85 %; (d) 82 %.
surgical treatment of AVMs [95]. The computerized dosim- Ten (1.8 %) patients sustained a permanent radiation-
etry and clinical data on 269 patients were reviewed. The induced complication. Seventeen (3.1 %) had a transient
AVM nidus was hand contoured on successive enhanced CT radiation-induced complication. These problems usually
slices through the nidus, to allow detailed determination of resolved within several months of steroid therapy. Most
nidus volume, target miss, normal brain treated, dose confor- importantly, 42 patients suffered hemorrhages after radiosur-
mality, and dose gradient. In addition, a number of patient gical treatment, and 8 were fatal. Hemorrhage during the
and treatment factors, including Spetzler–Martin score, pre- latent period after radiosurgery is the major drawback of this
senting symptoms, dose, number of isocenters, radiographic procedure. Only surgery at this point can immediately elimi-
outcome, and clinical outcome were subjected to multivari- nate the risk of hemorrhage in patients with AVMs.
ate analysis.
None of the analyzed factors were predictive of perma-
nent radiation-induced complications or of hemorrhage after Conclusion
radiosurgery in this study. Eloquent AVM location and 12 Gy
volume correlated with the occurrence of transient radiation- Modifications of LINACs for radiosurgery have led to close
induced complications. Better conformality correlated with collaboration between neurosurgeons, medical physicists,
a reduced incidence of transient complications. Lower and radiation oncologists. Such collaborations have increased
Spetzler–Martin scores, higher doses, and steeper dose gra- cross fertilization for each of these fields. Increasing use of
dients correlated with radiographic success. LINAC for radiosurgery has contributed to neurosurgeons
When AVMs are not cured, current practice frequently gaining a more in depth understanding of radiotherapy and
involves a “retreatment,” usually 3 years after the original the challenges of delivering radiation to patients, and SRS has
treatment. We reviewed the cases of 52 patients who under- now become an essential part of neurosurgical education.
went repeat radiosurgery for residual AVM at our institution The development of LINAC has led to widespread avail-
between December 1991 and June 1998 [96]. In each case, ability of radiotherapy for patients. LINACs are found in
residual arteriovenous shunting persisted beyond 36 months almost every major medical center in the USA With afford-
after the initial treatment. The mean interval between the first able modifications to LINACs, radiosurgery is also now
and second treatments was 41 months. Each AVM nidus was widely available to patients. LINACs are currently the most
measured at the time of original treatment and again at the common devices used to deliver radiotherapy and radiosur-
time of retreatment, and dosimetric parameters of the two gery. LINAC radiosurgery is being used to treat patients for
treatments were compared. After retreatment, patients were brain tumors, vascular malformations, pain syndromes, and
followed, and their outcomes evaluated, according to our functional indications. LINAC technology is continually
standard post-AVM radiosurgery protocol. Definitive end being improved and will continue to play a major role in
points included angiographic cure, radiosurgical failure clinical delivery of radiosurgery for patients.
24. Heifetz MD, Wexler M, Thompson R. Single-beam radiotherapy

References knife. A practical theoretical model. J Neurosurg. 1984;60(4):814–8.
25. Avanzo RC, Chierego G, Marchetti C, Pozza F, Colombo F,
Benedetti A, et al. [Stereotaxic irradiation with a linear accelerator].
1. Leksell L. The stereotaxic method and radiosurgery of the brain. Radiol Med. 1984;70(3):124–9.
Acta Chir Scand. 1951;102(4):316–9. 26. Colombo F, Benedetti A, Pozza F, Avanzo R, Chierego G, Marchetti
2. Schultz CJGM, Mueller WM. Modified linear accelerator C, et al. Radiosurgery using a 4MV linear accelerator. Technique and
radiosurgery: principles and techniques. In: IM G, editor. LINAC radiobiologic implications. Acta Radiol Suppl. 1986;369:603–7.
and Gamma Knife radiosurgery. Chicago: The American
27. Colombo F, Benedetti A, Pozza F, Avanzo RC, Marchetti C,
Association of Neurological Surgeons; 1999. p. 19–30.
Chierego G, et al. External stereotactic irradiation by linear accel-
3. Friedman WA. Linear accelerator radiosurgery. Clin Neurosurg.
erator. Neurosurgery. 1985;16(2):154–60.
1992;38:445–71.
28. Colombo F, Benedetti A, Zanardo A, Pozza F, Avanzo R, Chierego
4. Kondziolka D, Lunsford LD, Loeffler JS, Friedman
G, et al. New technique for three-dimensional linear accelerator
WA. Radiosurgery and radiotherapy: observations and clarifica-
radiosurgery. Acta Neurochir Suppl (Wien). 1987;39:38–40.
tions. J Neurosurg. 2004;101(4):585–9.
29. Hartmann GH, Schlegel W, Sturm V, Kober B, Pastyr O, Lorenz
5. Larson DA, Flickinger JC, Loeffler JS. The radiobiology of radio-
WJ. Cerebral radiation surgery using moving field irradiation at a
surgery. Int J Radiat Oncol Biol Phys. 1993;25(3):557–61.
linear accelerator facility. Int J Radiat Oncol Biol Phys. 1985;11(6):
6. Betti OO, Galmarini D, Derechinsky V. Radiosurgery with a linear
1185–92.
accelerator. Methodological aspects. Stereotact Funct Neurosurg.
30. Houdek PV, Fayos JV, Van Buren JM, Ginsberg MS. Stereotaxic
1991;57(1–2):87–98.
radiotherapy technique for small intracranial lesions. Med Phys.
7. Wideroe R. Uber ein neues Prinzip Zur Herstellung hoher
1985;12(4):469–72.
Spannungen. Archiv Elektrot. 1928;21:387.
31. Podgorsak EB, Olivier A, Pla M, Hazel J, de Lotbiniere A, Pike
8. Thwaites DI, Tuohy JB. Back to the future: the history and develop-
B. Physical aspects of dynamic stereotactic radiosurgery. Appl
ment of the clinical linear accelerator. Phys Med Biol.
Neurophysiol. 1987;50(1–6):263–8.
2006;51(13):R343–62.
9. Bova FJ. Radiation physics. Neurosurg Clin N Am. 1990;1(4): 32. Podgorsak EB, Olivier A, Pla M, Lefebvre PY, Hazel J. Dynamic
909–31. stereotactic radiosurgery. Int J Radiat Oncol Biol Phys. 1988;14(1):
10. Hoh DJ, Liu CY, Pagnini PG, Yu C, Wang MY, Apuzzo ML. Chained 115–26.
lightning, part I: exploitation of energy and radiobiological princi- 33. Winston KR, Lutz W. Linear accelerator as a neurosurgical tool for
ples for therapeutic purposes. Neurosurgery. 2007;61(1):14–27; stereotactic radiosurgery. Neurosurgery. 1988;22(3):454–64.
discussion 28. 34. Friedman WA, Bova FJ. The University of Florida radiosurgery
11. Hansen WW. A type of electrical resonator. J Appl Phys. system. Surg Neurol. 1989;32(5):334–42.
1938;9:654–63. 35. Leavitt DD, Gibbs Jr FA, Heilbrun MP, Moeller JH, Takach Jr
12. Benedict S, Bova F, Clark B, Goetsch S, Hinson W, Leavitt D, et al. GA. Dynamic field shaping to optimize stereotactic radiosurgery.
Anniversary paper: the role of medical physicists in developing ste- Int J Radiat Oncol Biol Phys. 1991;21(5):1247–55.
reotactic radiosurgery. Med Phys. 2008;35(9):4262–77. 36. Yin FF, Zhu J, Yan H, Gaun H, Hammoud R, Ryu S, et al. Dosimetric
13. Gildenberg P. The history of stereotactic neurosurgery. In: Friedman characteristics of Novalis shaped beam surgery unit. Med Phys.
WA, Winn HR, Mayberg MR, editors. Neurosurgery Clinics of 2002;29(8):1729–38.
North America: stereotactic neurosurgery. Philadelphia: 37. Adler Jr JR, Chang SD, Murphy MJ, Doty J, Geis P, Hancock
W.B. Saunders; 1990. p. 765–80. SL. The Cyberknife: a frameless robotic system for radiosurgery.
14. Larsson B, Leksell L, Rexed B, Sourander P, Mair W, Andersson Stereotact Funct Neurosurg. 1997;69(1–4 Pt 2):124–8.
B. The high-energy proton beam as a neurosurgical tool. Nature. 38. Murphy MJ. An automatic six-degree-of-freedom image registra-
1958;182(4644):1222–3. tion algorithm for image-guided frameless stereotaxic radiosurgery.
15. Leksell L. Cerebral radiosurgery. I. Gammathalanotomy in two Med Phys. 1997;24(6):857–66.
cases of intractable pain. Acta Chir Scand. 1968;134(8):585–95. 39. Huntzinger C, Friedman W, Bova F, Fox T, Bouchet L, Boeh
16. Kjellberg RN, Hanamura T, Davis KR, Lyons SL, Adams L. Trilogy image-guided stereotactic radiosurgery. Med Dosim.
RD. Bragg-peak proton-beam therapy for arteriovenous malforma- 2007;32(2):121–33.
tions of the brain. N Engl J Med. 1983;309(5):269–74. 40. Gerszten PC, Burton SA, Ozhasoglu C, McCue KJ, Quinn
17. Lyman JT, Phillips MH, Frankel KA, Fabrikant JI. Stereotactic AE. Radiosurgery for benign intradural spinal tumors. Neurosurgery.
frame for neuroradiology and charged particle Bragg peak radiosur- 2008;62(4):887–95. discussion 95–6.
gery of intracranial disorders. Int J Radiat Oncol Biol Phys. 41. Levine AM, Coleman C, Horasek S. Stereotactic radiosurgery for
1989;16(6):1615–21. the treatment of primary sarcomas and sarcoma metastases of the
18. Larsson B, Liden K, Sarby B. Irradiation of small structures through spine. Neurosurgery. 2009;64(2 Suppl):A54–9.
the intact skull. Acta Radiol Ther Phys Biol. 1974;13(6):512–34. 42. Thariat J, Castelli J, Chanalet S, Marcie S, Mammar H, Bondiau
19. Knapp EA, Knapp BC, Potter JM. Standing wave high energy lin- PY. CyberKnife stereotactic radiotherapy for spinal tumors: value
ear accelerator structures. Rev Sci Instrum. 1968;39:979–91. of computed tomographic myelography in spinal cord delineation.
20. Hoh DJ, Liu CY, Chen JC, Pagnini PG, Yu C, Wang MY, et al. Neurosurgery. 2009;64(2 Suppl):A60–6.
Chained lightning, part II: neurosurgical principles, radiosurgical 43. Tsai JT, Lin JW, Chiu WT, Chu WC. Assessment of image-guided
technology, and the manipulation of energy beam delivery. CyberKnife radiosurgery for metastatic spine tumors. J Neurooncol.
Neurosurgery. 2007;61(3):433–46. discussion 46. 2009;94:119–27.
21. Gillies BA, O’Brien PF, McVittie R, McParland C, Easton 44. Samii M, Matthies C. Management of 1000 vestibular schwanno-
H. Engineering modifications for dynamic stereotactically assisted mas (acoustic neuromas): surgical management and results with an
radiotherapy. Med Phys. 1993;20(5):1491–5. emphasis on complications and how to avoid them. Neurosurgery.
22. Betti O, Derechinsky V. [Multiple-beam stereotaxic irradiation]. 1997;40(1):11–21; discussion 23.
Neurochirurgie. 1983;29(4):295–8. 45. Samii M, Matthies C. Management of 1000 vestibular schwannomas
23. Betti OO. Treatment of arteriovenous malformations with the linear (acoustic neuromas): the facial nerve—preservation and restitution
accelerator. Appl Neurophysiol. 1987;50(1–6):262. of function. Neurosurgery. 1997;40(4):684–94. discussion 94–5.
46. Samii M, Matthies C. Management of 1000 vestibular schwannomas 67. Noordijk EM, Vecht CJ, Haaxma-Reiche H, Padberg GW,
(acoustic neuromas): hearing function in 1000 tumor resections. Voormolen JH, Hoekstra FH, et al. The choice of treatment of sin-
Neurosurgery. 1997;40(2):248–60. discussion 60–2. gle brain metastasis should be based on extracranial tumor activity
47. Leksell L. A note on the treatment of acoustic tumours. Acta Chir and age. Int J Radiat Oncol Biol Phys. 1994;29(4):711–7.
Scand. 1971;137(8):763–5. 68. Mintz AH, Kestle J, Rathbone MP, Gaspar L, Hugenholtz H, Fisher
48. Spiegelmann R, Lidar Z, Gofman J, Alezra D, Hadani M, Pfeffer B, et al. A randomized trial to assess the efficacy of surgery in addi-
R. Linear accelerator radiosurgery for vestibular schwannoma. tion to radiotherapy in patients with a single cerebral metastasis.
J Neurosurg. 2001;94(1):7–13. Cancer. 1996;78(7):1470–6.
49. Litre F, Rousseaux P, Jovenin N, Bazin A, Peruzzi P, Wdowczyk D, 69. Haines SJ. Moving targets and ghosts of the past: outcome mea-
et al. Fractionated stereotactic radiotherapy for acoustic neuromas: surement in brain tumour therapy. J Clin Neurosci. 2002;9(2):
a prospective monocenter study of about 158 cases. Radiother 109–12.
Oncol. 2013;106:169–74. 70. Sturm V, Kober B, Hover KH, Schlegel W, Boesecke R, Pastyr O,
50. Friedman WA, Bradshaw P, Myers A, Bova FJ. Linear accelerator et al. Stereotactic percutaneous single dose irradiation of brain
radiosurgery for vestibular schwannomas. J Neurosurg. metastases with a linear accelerator. Int J Radiat Oncol Biol Phys.
2006;105(5):657–61. 1987;13(2):279–82.
51. van de Langenberg R, Dohmen AJ, de Bondt BJ, Nelemans PJ, 71. Sturm V, Kimmig B, Engenhardt R, Schlegel W, Pastyr O, Treuer
Baumert BG, Stokroos RJ. Volume changes after stereotactic H, et al. Radiosurgical treatment of cerebral metastases. Method,
LINAC radiotherapy in vestibular schwannoma: control rate and indications and results. Stereotact Funct Neurosurg.
growth patterns. Int J Radiat Oncol Biol Phys. 2012;84(2):343–9. 1991;57(1–2):7–10.
52. Sekhar LN, Jannetta PJ, Burkhart LE, Janosky JE. Meningiomas 72. Voges J, Treuer H, Erdmann J, Schlegel W, Pastyr O, Muller RP,
involving the clivus: a six-year experience with 41 patients. et al. Linac radiosurgery in brain metastases. Acta Neurochir Suppl.
Neurosurgery. 1990;27(5):764–81. discussion 81. 1994;62:72–6.
53. Sekhar LN, Linskey ME, Sen CN, Altschuler EM. Surgical man- 73. Alexander 3rd E, Moriarty TM, Davis RB, Wen PY, Fine HA, Black
agement of lesions within the cavernous sinus. Clin Neurosurg. PM, et al. Stereotactic radiosurgery for the definitive, noninvasive
1991;37:440–89. treatment of brain metastases. J Natl Cancer Inst. 1995;87(1):34–40.
54. Simpson D. The recurrence of intracranial meningiomas after sur- 74. Black PM. Solitary brain metastases. Radiation, resection, or radio-
gical treatment. J Neurol Neurosurg Psychiatry. 1957;20(1):22–39. surgery? Chest. 1993;103(4 Suppl):367S–9.
55. Pollock BE, Stafford SL, Utter A, Giannini C, Schreiner 75. Joseph J, Adler JR, Cox RS, Hancock SL. Linear accelerator-based
SA. Stereotactic radiosurgery provides equivalent tumor control to stereotaxic radiosurgery for brain metastases: the influence of num-
Simpson Grade 1 resection for patients with small- to medium-size ber of lesions on survival. J Clin Oncol. 1996;14(4):1085–92.
meningiomas. Int J Radiat Oncol Biol Phys. 2003;55(4):1000–5. 76. Fuller BG, Kaplan ID, Adler J, Cox RS, Bagshaw MA. Stereotaxic
56. Valentino V, Schinaia G, Raimondi AJ. The results of radiosurgical radiosurgery for brain metastases: the importance of adjuvant
management of 72 middle fossa meningiomas. Acta Neurochir whole brain irradiation. Int J Radiat Oncol Biol Phys. 1992;23(2):
(Wien). 1993;122(1–2):60–70. 413–8.
57. Villavicencio AT, Black PM, Shrieve DC, Fallon MP, Alexander E, 77. Adler JR, Cox RS, Kaplan I, Martin DP. Stereotactic radiosurgical
Loeffler JS. Linac radiosurgery for skull base meningiomas. Acta treatment of brain metastases. J Neurosurg. 1992;76(3):444–9.
Neurochir (Wien). 2001;143(11):1141–52. 78. Auchter RM, Lamond JP, Alexander E, Buatti JM, Chappell R,
58. Engenhart R, Kimmig BN, Hover KH, Wowra B, Sturm V, van Friedman WA, et al. A multiinstitutional outcome and prognostic
Kaick G, et al. Stereotactic single high dose radiation therapy of factor analysis of radiosurgery for resectable single brain metasta-
benign intracranial meningiomas. Int J Radiat Oncol Biol Phys. sis. Int J Radiat Oncol Biol Phys. 1996;35(1):27–35.
1990;19(4):1021–6. 79. Goodman KA, Sneed PK, McDermott MW, Shiau CY, Lamborn
59. Spiegelmann R, Nissim O, Menhel J, Alezra D, Pfeffer MR. Linear KR, Chang S, et al. Relationship between pattern of enhancement
accelerator radiosurgery for meningiomas in and around the cav- and local control of brain metastases after radiosurgery. Int J Radiat
ernous sinus. Neurosurgery. 2002;51(6):1373–9. discussion 9–80. Oncol Biol Phys. 2001;50(1):139–46.
60. Hakim R, Alexander 3rd E, Loeffler JS, Shrieve DC, Wen P, Fallon 80. Shiau CY, Sneed PK, Shu HK, Lamborn KR, McDermott MW,
MP, et al. Results of linear accelerator-based radiosurgery for intra- Chang S, et al. Radiosurgery for brain metastases: relationship of
cranial meningiomas. Neurosurgery. 1998;42(3):446–53. discus- dose and pattern of enhancement to local control. Int J Radiat Oncol
sion 53–4. Biol Phys. 1997;37(2):375–83.
61. Friedman WA, Murad GJ, Bradshaw P, Amdur RJ, Mendenhall 81. Gaspar L, Scott C, Rotman M, Asbell S, Phillips T, Wasserman T,
WM, Foote KD, et al. Linear accelerator surgery for meningiomas. et al. Recursive partitioning analysis (RPA) of prognostic factors in
J Neurosurg. 2005;103(2):206–9. three Radiation Therapy Oncology Group (RTOG) brain metastases
62. Hashimoto N, Rabo CS, Okita Y, Kinoshita M, Kagawa N, Fujimoto trials. Int J Radiat Oncol Biol Phys. 1997;37(4):745–51.
Y, et al. Slower growth of skull base meningiomas compared with 82. Swinson BM, Friedman WA. Linear accelerator stereotactic radio-
non-skull base meningiomas based on volumetric and biological surgery for metastatic brain tumors: 17 years of experience at the
studies. J Neurosurg. 2012;116(3):574–80. University of Florida. Neurosurgery. 2008;62(5):1018–31. discus-
63. Gerszten PC, Chen S, Quader M, Xu Y, Novotny Jr J, Flickinger sion 31–2.
JC. Radiosurgery for benign tumors of the spine using the Synergy 83. Shrieve DC, Alexander 3rd E, Black PM, Wen PY, Fine HA, Kooy
S with cone-beam computed tomography image guidance. HM, et al. Treatment of patients with primary glioblastoma multi-
J Neurosurg. 2012;117(Suppl):197–202. forme with standard postoperative radiotherapy and radiosurgical
64. Lohr F, Pirzkall A, Hof H, Fleckenstein K, Debus J. Adjuvant treat- boost: prognostic factors and long-term outcome. J Neurosurg.
ment of brain metastases. Semin Surg Oncol. 2001;20(1):50–6. 1999;90(1):72–7.
65. DeAngelis LM. Brain tumors. N Engl J Med. 2001;344(2): 84. Hall WA, Djalilian HR, Sperduto PW, Cho KH, Gerbi BJ, Gibbons
114–23. JP, et al. Stereotactic radiosurgery for recurrent malignant gliomas.
66. Patchell RA, Tibbs PA, Walsh JW, Dempsey RJ, Maruyama Y, J Clin Oncol. 1995;13(7):1642–8.
Kryscio RJ, et al. A randomized trial of surgery in the treatment of 85. Masciopinto JE, Levin AB, Mehta MP, Rhode BS. Stereotactic
single metastases to the brain. N Engl J Med. 1990;322(8): radiosurgery for glioblastoma: a final report of 31 patients.
494–500. J Neurosurg. 1995;82(4):530–5.
86. Curran Jr WJ, Scott CB, Horton J, Nelson JS, Weinstein AS, radiosurgery followed by conventional radiotherapy with carmus-
Fischbach AJ, et al. Recursive partitioning analysis of prognostic tine to conventional radiotherapy with carmustine for patients with
factors in three Radiation Therapy Oncology Group malignant gli- glioblastoma multiforme: report of Radiation Therapy Oncology
oma trials. J Natl Cancer Inst. 1993;85(9):704–10. Group 93-05 protocol. Int J Radiat Oncol Biol Phys. 2004;60(3):
87. Sarkaria JN, Mehta MP, Loeffler JS, Buatti JM, Chappell RJ, Levin 853–60.
AB, et al. Radiosurgery in the initial management of malignant 92. Stupp R, Mason WP, van den Bent MJ, Weller M, Fisher B,
gliomas: survival comparison with the RTOG recursive partitioning Taphoorn MJ, et al. Radiotherapy plus concomitant and adjuvant
analysis. Radiation Therapy Oncology Group. Int J Radiat Oncol temozolomide for glioblastoma. N Engl J Med. 2005;352(10):
Biol Phys. 1995;32(4):931–41. 987–96.
88. Ulm AJ III, Friedman WA, Bradshaw P, Foote KD, Bova 93. Pollock BE, Flickinger JC, Lunsford LD, Maitz A, Kondziolka
FJ. Radiosurgery in the treatment of malignant gliomas: the D. Factors associated with successful arteriovenous malformation
University of Florida experience. Neurosurgery. 2005;57(3):512–7; radiosurgery. Neurosurgery. 1998;42(6):1239–44. discussion 44–7.
discussion 517. 94. Ellis TL, Friedman WA, Bova FJ, Kubilis PS, Buatti JM. Analysis
89. Irish WD, Macdonald DR, Cairncross JG. Measuring bias in uncon- of treatment failure after radiosurgery for arteriovenous malforma-
trolled brain tumor trials—to randomize or not to randomize? Can tions. J Neurosurg. 1998;89(1):104–10.
J Neurol Sci. 1997;24(4):307–12. 95. Friedman WA, Bova FJ, Bollampally S, Bradshaw P. Analysis of
90. Curran Jr WJ, Scott CB, Weinstein AS, Martin LA, Nelson JS, factors predictive of success or complications in arteriovenous mal-
Phillips TL, et al. Survival comparison of radiosurgery-eligible and formation radiosurgery. Neurosurgery. 2003;52(2):296–307; dis-
-ineligible malignant glioma patients treated with hyperfractionated cussion 308.
radiation therapy and carmustine: a report of Radiation Therapy 96. Foote KD, Friedman WA, Ellis TL, Bova FJ, Buatti JM, Meeks
Oncology Group 83–02. J Clin Oncol. 1993;11(5):857–62. SL. Salvage retreatment after failure of radiosurgery in patients
91. Souhami L, Seiferheld W, Brachman D, Podgorsak EB, Werner- with arteriovenous malformations. J Neurosurg. 2003;98(2):
Wasik M, Lustig R, et al. Randomized comparison of stereotactic 337–41.
Charged Particles in Stereotactic
Radiosurgery 10
Shervin M. Shirvani and Joe Y. Chang
are mostly limited to the tissue before or at the Bragg Peak.

Introduction: The Rationale for Charged Organs beyond this depth are spared the toxic effects of
Particle Radiation for SRS radiation. This difference between photons and charged
particle radiation is illustrated in Fig. 10.1.
The majority of radiotherapy delivery systems worldwide Stereotactic radiosurgery (SRS) and charged particle radio-
use photons generated from either a radioactive isotope such therapy are not mutually exclusive. SRS is conceptually
as Cobalt-60 or from a linear accelerator. Photon radiation defined as the delivery of large, hypofractionated doses of
has many advantages including its wide availability, cost- radiation to precise targets. This definition is not limited to any
effectiveness, and track record of controlling malignancies particular radiation modality, and charged particles can be
originating from a multitude of tissues. Nonetheless, photons used in place of photons for stereotactic treatment, provided
are not without their disadvantages, and over the last several that the essential processes of SRS—including accurate dose-
decades, radiation oncologists and medical physicists have modeling, image-guidance, and quality assurance—can be
investigated alternative radiotherapy modalities which cir- achieved. In fact, combing the radiobiological and dose esca-
cumvent the limitations of photons. Among these alterna- lation advantages of hypofractionation with the tissue-sparing
tives, charged particle radiotherapies such as proton and characteristics of charged particle radiation can be an elegant
carbon ion beams have received special attention because of way to maximize the therapeutic index.
their favorable physical characteristics. Until 2001, there were only two proton therapy centers in
Charged particles have an energy transfer pattern that is the USA. Since then, the attractive properties of charged par-
fundamentally different from photons. In the case of photon ticles have encouraged the construction of eight additional
radiation, dose deposition and resultant ionizations in mat- centers with four more to be completed soon. This sudden
ter are characterized by a shallow peak dose followed by a acceleration in proton center availability will likely fuel
smooth fall-off. Consequently, adjacent normal tissue active investigation into the use of charged particle radiation
located both shallow and deep to the target receives signifi- in a variety of clinical settings, including radiosurgery. This
cant radiation exposure. Unlike photons, charged particles chapter will review the physics and radiobiology of charged
deposit their energy sparingly until a certain depth is particle radiotherapy as well as published clinical outcomes
reached, at which point a very large deposition of energy to date in the hypofractionated, stereotactic setting.
occurs. Beyond this depth, called the “Bragg peak,” there
are no further ionization events attributable to the charged
particle (though a small number of ionizations due to nuclear Physical and Technical Considerations
scatter may still occur when larger nuclei like carbon ions
are used). The clinical consequence of the charged particle For therapeutic radiation, charged particles are first produced
depth–dose distribution is that the effects of radiotherapy by applying an electric field to a gas of the desired particle to
separate the nucleus from its electrons. Acceleration of the
S.M. Shirvani, M.D. nucleus-derived charged particle is then achieved via a
Department of Radiation Oncology, Banner MD Anderson Cancer cyclotron or a synchrotron. These machines operate by
Center, Gilbert, AZ, USA accelerating particles around a round cavity while subjecting
J.Y. Chang, M.D., Ph.D., M.S. (*) them to electric and magnetic fields. In a synchrotron, both
Department of Radiation Oncology, The University of Texas the electric field and the magnetic field vary in time to accel-
e-mail: jychang@mdanderson.org erate a particle while also keeping it along a defined circular
136 S.M. Shirvani and J.Y. Chang
SOBP (4cm) E=214.2MeV, Weight=8.7%

100
E=221.8MeV, Weight=47.4% E=211.6MeV, Weight=7.1%
E=219.3MeV, Weight=13.6% E=209 MeV, Weight=6.4%
E=216.7MeV, Weight=12.6% E=206.3MeV, Weight=5.0%
80
Dose (%)
60
40
20
0
0 5 10 15 20 25 30
Fig. 10.1 Comparison of dose-distribution curves of photons and pro- Depth (cm)
tons. The high entrance dose and tail of exiting dose are avoided by the
use of charged particles Fig. 10.2 Depth–dose curve for a spread-out Bragg peak (SOBP) with
its constituent pristine Bragg peaks also shown. The gray area repre-
senting the target is encompassed by the width of the SOBP plateau
track as it accelerates. In a cyclotron, the particle starts in the region
center of a cylindrical cavity and accelerates in a spiral, mov-
ing centripetally with time. In this case, the electric field shaping of the radiation field in passive scattering systems is
alternates with a constant frequency to accelerate the particle usually necessary. Either apertures or collimators are used to
during each revolution, but the magnetic field is kept uniform shape the perimeter of the beam, while customized compen-
to achieve the spiral movement. sators placed in the beam path after the aperture can be used
When the particle achieves the desired energy, it is ejected to finely modulate the field’s distal edge [1, 2]. Importantly,
from the synchrotron or cyclotron. In the case of protons, this the materials used for these devices must provide adequate
is usually 150–250 MeV which results in 25–30 cm of attenuation without producing harmful particles through
penetration into tissue. After ejection from the accelerator, scatter interactions. For example in the case of proton ther-
the particle is then guided by magnetic fields to the treatment apy, brass or tungsten apertures are often used.
gantry. A nozzle at the gantry directs the beam to the prop- Because passive scattering planning is similar to 3D con-
erly positioned and immobilized patient. Worldwide, proton formal radiation, complicated dose distributions and tight
and carbon ion cyclotrons are the most commonly used sys- conformation of dose around curved lesions are difficult to
tems for charged particle therapy. Other charged ions includ- achieve. However, with proper resources and skilled dosim-
ing helium, neon, argon, and silicon nuclei have been etry, these challenges are usually surmountable. Dosimetric
explored in research settings but are not currently in wide- studies of passive scattering proton therapy in the thorax
spread use. have shown that, with proper planning, their dose distribu-
A monoenergetic beam of charged particles results in a tions are superior to photon-based treatments (3D conformal
narrow Bragg peak, which has limited clinical utility for photon or photon IMRT) in most cases [1, 3].
treating typical tumor widths. Therefore, two different modes The second mode of delivery is a scanning beam system.
of delivery have been developed to adequately cover the In this mode, the charged particle is subjected to an electro-
dimensions of a typical target. The most common method of magnetic field prior to exiting the gantry. This electromag-
delivery is via passive scattering. In this system, the particle netic field modulates the energy and direction of the particle
beam is scattered and flattened to create a beam of desired beam so that the particles exiting the gantry deposit their
width and intensity. The monoenergetic beam can also be energy at a specific plane (“uniform scanning”) or voxel
modulated with a range shifter to generate a polyenergetic (“spot scanning” or “pencil beam scanning”) within the
beam; the resulting depth–dose curve of the polyenergetic tumor [4]. In these systems, all the planes or voxels that con-
beam is a superposition of the depth–dose curves of the constitute the target are treated in turn, usually starting at the
stituent Bragg peaks. It is characterized by a nonzero entrance most distal areas, until the entire volume of the tumor has
dose which rises to a plateau region before falling to zero at been irradiated. Among the voxel-based modes, spot scan-
the depth of the most distal constituent Bragg peak ning delivers dose to each voxel in a step-and-shoot fashion
(Fig. 10.2). The range shifter can customize the width of this while pencil beam scans and treats each voxel in a continuous
“spread-out Bragg peak” so that the tumor is adequately cov- fashion. In both cases, the dose that each voxel receives can
ered by the high-dose plateau. be independently modulated, and therefore treatment planning
To maximize conformality to the target’s three- for these techniques is typically done using inverse-planning
dimensional shape and to avoid normal tissues, additional algorithms. Because of this similarity with photon-based
10 Charged Particles in Stereotactic Radiosurgery 137
intensity-modulated radiotherapy (IMRT), these techniques RBE for a specific particle is high, then that particle exerts a
are often called intensity-modulated proton therapy (IMPT). significant amount of damage per absorbed unit of energy
Scanning beam delivery has two important advantages. and is said to have high “quality.”
First, compensators and apertures are not needed, and there- RBE is related to LET in a bell-shaped distribution: RBE
fore, contaminant scatter from these objects and the risk of rises until an LET of approximately 100–150 keV/μm and then
secondary malignancy is reduced. Second, scanning modes falls at higher LET values. The reason for the peak in RBE at
result in greater conformality and are especially useful for 100 keV/μm is that the likelihood of a single-particle track
eccentric tumors or tumors located near sensitive structures. causing a DNA double-strand break is maximized at this level.
A disadvantage, on the other hand, is that this approach is Because each charged particle has a different LET, we expect
less forgiving when intra- and interfraction tumor and organ RBE to be different among them. For instance, protons in the
motion occurs [5–9]. To address intrafractional motion, it is 250 MeV range have an RBE of approximately 1.1, whereas
imperative that tumor motion be accounted for during simu- carbon-ions have an RBE in the 2–3 range [11, 12]. This means
lation, treatment planning, and dose delivery. In the chest, for that protons are radiobiologically similar to photons, offering a
instance, tumors’ position can change significantly over the 10 % improvement in radiation quality, whereas heavy nuclei
course of respiration. Investigators at MD Anderson have cause considerably higher biological effect.
demonstrated that the creation of a four-dimensional, inter- Two other radiobiological phenomena deserve mention
nal gross tumor volume (“4D-iGTV”) based upon the maxi- with regard to the relative advantages of charged particle
mal intensity projection of the tumor on axial CT slices radiation. In photon radiobiology, the radiosensitivity of a
throughout the entire respiratory phase is an accurate method tumor cell can be diminished by the absence of oxygen or by
for minimizing the risk of a geometrical miss due to tumor the cell’s position in a radioresistant phase of the cell cycle
motion [10]. For interfraction changes due to an early tumor (such as the S phase). In both cases, the altered radiosensitiv-
response, adaptive planning can be considered [8, 9]. ity is attributable to the cell’s capacity for repair of
radiation-induced DNA damage, which is enhanced by
hypoxia and the presence of sister chromatids and DNA-
Radiobiologic Advantages of Charged repair enzymes in certain cell-cycle phases. However, the
Particle Therapy single-track, double-strand break described in the previous
paragraph is a subtype of DNA injury that is difficult to
Charged particle radiotherapy is directly ionizing because, repair even in the presence of these radiobiological modifi-
by virtue of their charge, the accelerated particles can inter- ers. Because high quality particles (such as heavy nuclei
act with atomic electrons through Coulombic forces. such as carbon ions) are more likely to induce this type of
Consequently, charged particles, unlike photons, are asso- injury, the oxygen and cell cycle effects become less impor-
ciated with high linear energy transfer (LET). Their inter- tant. This ability of heavy ion particle radiation to overcome
action with tissues is characterized by (1) direct damage to these intrinsic mechanisms of radioresistance is a major the-
cellular DNA and (2) a dense mobilization of secondary oretical advantage over photons. In fact, this difference in
electrons along the particle track, resulting in a high local radiobiology is the main distinguishing feature between
concentration of free radicals which also cause DNA dam- heavy nuclei ions and protons, whose radiological quality is
age. In clinical systems that utilize a SOBP, cellular dam- similar to photons.
age is considered to be nearly constant across the plateau
phase of the depth–dose curve, but in reality, there may be
variations over the course of the particle track. This is dis- Particle Radiosurgery: Early Clinical
cussed in more detail in section “The Challenges of Particle Outcomes
Therapy”.
Though different nuclei used in charged particle therapy As stated in section “Introduction: The Rationale for Charged
may have similarities with respect to the general shape of Particle Radiation for SRS”, there is no a priori reason that
their depth–dose curve, their ultimate radiobiological effects stereotactic hypofractionation cannot be combined with par-
is dependent upon the specific LET of the particle that is ticle radiotherapy so that patients can benefit from the advan-
used. Said differently, the shape of the SOBP is useful for tages of both. However, highly reproducible patient
determining the distribution of dose, but the extent of bio- immobilization devices (such as the Gamma Knife frame)
logical damage that is done to the target tumor is dependent and image-guidance technologies (such as cone-beam com-
on a different property: relative biological effectiveness puted tomography) have not been traditionally been avail-
(RBE). This quantity is defined as the ratio of the dose of able in proton or heavy-nuclei gantry systems. Fortunately,
photons to the dose of charged particles necessary to achieve newer systems have circumvented these limitations by
the same biological effect in a specified test system. If the incorporating customized vacuum molded masks and frames
for immobilization, four-dimensional simulation for treat- These anatomically challenging cases are one circum-
ment planning, and on-board imaging systems such as fluo- stance where particle-based radiotherapy may provide valu-
roscopy, CT imaging, and fiducials for image-guidance. able synergy to hypofractionated treatment. Provided that
Nonetheless, “particle radiosurgery,” as loosely defined by proper beam angles are used, particles such as protons and
the combination of charged particle radiotherapy with large carbon ions should significantly reduce the integral dose to
fraction sizes, is still a young concept whose preliminary evi- the lung and minimize incidental irradiation of nearby organs
dence is limited to a handful of single-institution series in the including the heart, esophagus, spinal cord, bronchial tree,
context of treating early-stage lung cancer. Still, despite their major vessels and brachial plexus (Fig. 10.3). Providing evi-
limitations, these experiences provide proof of principle of dence of this principle, dosimetric studies at the Mayo Clinic
the efficacy and organ-sparing characteristics of particle and MD Anderson Cancer Center have confirmed that the
radiosurgery. Bragg peak characteristics of proton therapy can be used to
deliver ablative doses of radiation to early stage tumors as
effectively as photon SRS while significantly reducing the
Non-small Cell Lung Cancer dose to surrounding normal tissues.
In the MD Anderson study, proton plans compared to pho-
Rationale for Particle Therapy in the Thorax: ton-based SBRT plans significantly reduced the mean total
Anatomy and Dose lung dose from 5.4 to 3.5 Gy (p < 0.001) and 2.8 Gy (p < 0.001)
SRS advances for non-small cell lung cancer in the last and reduced the total lung volume receiving 5, 10, and 20 Gy
decade have largely relied on photon techniques. These early (p < 0.001). Moreover, for cases where the PTV was within
experiences were successful in safely ablating peripheral 2 cm of the critical structures, the proton plans resulting in sig-
tumors but also resulted in considerable toxicity when used nificantly reduced mean maximal dose to the aorta, brachial
for centrally located tumors [13–17]. Life-threatening com- plexus, heart, pulmonary vessels, and spinal cord [21].
plications including radiation pneumonitis, great vessel rup- Similarly, the Mayo Clinic study found that proton-based tech-
ture, esophageal ulceration, and spinal cord myelopathy have niques resulted in significantly reduced total lung mean and V5
been reported when central tumors have been treated with dose; other normal tissues including heart, bronchial tree
ablative dose regimens [18, 19]. This difference in therapeu- (including trachea), esophagus, and spinal cord lower were also
tic ratio was especially stark in an early phase II study of 70 found to receive less dose with protons [22]. Interestingly, in
patients with both peripherally and centrally located tumors both studies, better dose distributions were achieved with scan-
treated to 60–66 Gy in three fractions [20]. In this cohort, ning beam methods than with passively scattered protons.
2-year freedom from severe toxicity was 83 % for those with A second rationale for particle radiotherapy is the con-
peripheral tumors compared with only 54 % for patients with verse of the first: While the goal described above is to mini-
perihilar or central tumors. mize normal tissue effects given a specified target dose, the
Fig. 10.3 Comparison of passively scattered proton therapy (right panel) and 3D conformal photon therapy (left panel)-based stereotactic ablation
of a lung tumor close to critical structure brachial plexus. The proton plan results in a diminished dose to the brachial plexus
other potential use of particle therapy is to escalate the dose utilized 50 or 60 CGE in ten fractions to treat a cohort of 21
delivered to the tumor given prespecified thresholds for adja- patients with stage I NSCLC [35]. All but one of the patients
cent organs. In radiosurgery for non-small cell lung cancer, (95 %) experienced durable local control following this treat-
this second rationale is particularly relevant because of the ment. Two-year overall and cause-specific survival rates were
observation that higher BED is more likely to achieve cure 74 and 86 %. As in the Bush study, the toxicity profile was
[13–16, 23–32]. This phenomenon was underlined in a retro- favorable with no grade 3 or greater reactions observed. Other
spective study by Onishi and colleagues who demonstrated hypofractionated proton regimens used in Japanese proton
that among different photon SRS regimens, patients who centers have demonstrated similar results [36].
received a BED greater than 100 Gy had much better 5-year The Massachusetts General Hospital reported outcomes
local control (92 % vs. 74 %) and overall survival (88 % vs. for 15 patients with 20 tumors treated with proton SBRT to
69 %) than those who received a lower BED [23]. In light of 42–50 CGE in 3–5 fractions between July 2008 and
these findings, particles may be a powerful tool for achieving September 2010. Only one grade 3 or higher toxicity was
100 Gy BED when adjacent organs-at-risk prevent dose observed (grade 3 pneumonitis in a patient with severe
escalation of photons to these levels. chronic obstructive pulmonary disease), and the 2-year over-
all survival and local control rates were 64 % and 100 %,
Outcomes for Highly Hypofractionated Regimens respectively, with a median follow-up of 24.1 months [37].
Thus far, seven experiences using hypofractionated particle The National Institute of Radiological Sciences (NIRS) in
therapy for non-small cell lung cancer have been reported, Chiba, Japan, has reported three trials of hypofractionated
four with proton therapy and three with carbon ion therapy radiotherapy using carbon ions. The first, a phase I/II study
(Table 10.1). Each of these studies has unique technical and of 50 patients treated to 72 CGE in nine fractions, resulted in
image-guidance procedures. The earliest study was per- a 5-year local control rate of 95 % with no grade 3 toxicities
formed by Bush et al. at Loma Linda. At this center, 68 observed [38]. A second study followed 76 patients with T1
patients were treated to 51–60 cobalt-60 Gray equivalents and T2 tumors treated to 52.8 CGE in four fractions and
(CGE) in ten fractions. Real-time fluoroscopy was used for 60 CGE in four fractions, respectively [39]. At 3-year fol-
verification of patient and tumor position, and notably, this low-up, local control for T1 tumors was 98 % and for T2
study did not include any compensatory actions for respira- tumors was 80 %. Again, no grade 3 toxicities were observed.
tory motion [33]. Nonetheless, 3-year local control and Finally, a dose escalation study in which patients receive
cause-specific survival rates were 74 % and 72 %, respec- single-fraction regimens from 28 to 44 CGE is ongoing at
tively, over a median follow-up of 30 months. With regard to NIRS. Preliminary results showed no grade 3 or greater tox-
toxicity, no cases of acute radiation pneumonitis were icities at any of the dose levels during a follow-up period of
observed nor were there any grade 3 early or late esophageal 16 months [40]. The local control and survival outcomes of
or cardiac toxicities. this cohort are yet to be reported.
Nihei and colleagues used three-dimensional CT simula-
tion, respiratory gating, and real-time digital radiographs for Outcomes for Modestly Hypofractionated
position verification. With these more modern techniques, 37 and Conventionally Fractionated Regimens
patients were treated to a total dose of 70–94 CGE using frac- In addition to the studies above, more modestly hypofrac-
tion sizes of 3.5–4.7 CGE, which resulted in a local control tionated regimens using proton therapy have been reported
rate of 80 % at 24 months [34]. Hata and colleagues in Japan for non-small cell lung cancer (Table 10.2). While strictly
Table 10.1 Highly hypofractionated proton or carbon ion for T1-T2 N0 non-small cell lung cancer
Study Dose (CGE)/no. of fractions Local control Overall survival
Proton therapy
Westover et al. [37] (2012) 42–50/3–5 100 % (2 years) 64 % (2 years)
Iwata et al. [36] (2010) 80/20 or 60/10 81 % (3 years) 73 % (3 years)
Hata et al. [35] (2007) 50–60/10 95 % (2 years) 74 % (2 years)
Nihei et al. [34] (2006) 70–94/20 80 % (2 years) 84 % (2 years)
Bush et al. [33] (2004) 60/10 74 % (3 years) 44 % (3 years)
Carbon ion therapy
Iwata et al. [36] (2010) 52.8/4 86 % (2 years) 87 % (2 years)
Miyamoto et al. [39] (2007) 52.8–60/4 98 % (3 years) 78 % (3 years)
Miyamoto et al. [38] (2007) 72/8 95 % (5 years) 50 % (5 years)
Table 10.2 Modestly hypofractionated proton radiotherapy for non-small cell lung cancer
Dose (CGE)/
Study no. of fractions Chemotherapy Local control Overall survival
Early stage
Chang et al. [41] (2010) 87.5/35 N/A 89 % (2 years) 54.5 % (2 years)
Shioyama et al. [42] (2003) 76a/22a N/A T1A—89 % (5 years) T1A—70 % (5 years)
T1B—39 % (5 years) T1B—16 % (5 years)
Locally advanced
Chang et al. [45] (2011) 74/37 Carboplatin/paclitaxel 80 % (1 year) 86 % (1 year)
Nakayama et al. [43] (2011) 67.1–91.3/33 None 66 % (2 years) 59 % (2 years)
a
Median values
speaking these fall outside the realm of radiosurgical abla- high radiation doses and/or concurrent chemotherapy in this
tion, they are nonetheless instructive with regard to the high setting. Nakayama et al. retrospectively reviewed 35 patients
doses that can be achieved with the depth–dose characteris- with stage II or III NSCLC who were medically inoperable or
tics of particle radiotherapy: Targeting the subset of patients who refused surgery and instead were treated with proton
with centrally or superiorly located Stage IA–II non-small therapy without concurrent chemotherapy [43]. The median
cell lung cancers, Chang et al. conducted a Phase I/II pro- proton dose delivered was 78.3 CGE (range, 67.1–91.3 CGE).
spective study of proton radiation with a total dose of The overall and local progression-free survival rates were
87.5 CGE at 2.5 CGE per fraction [41]. At initial reporting of 81.8 % and 93.3 %, respectively, at 1 year and 58.9 % and
outcomes with a median follow-up time of 16.3 months, the 65.9 %, respectively, at 2 years, and incredibly, no grade 3 or
most common adverse effects were Grade 2 dermatitis higher toxicity was observed. Similarly, Sejpal et al. pub-
(67 %), Grade 2 fatigue (44 %), Grade 2 pneumonitis (11 %), lished a retrospective analysis of 62 patients with locally
Grade 2 esophagitis (6 %), and Grade 2 chest wall pain advanced NSCLC who were treated with proton therapy and
(6 %). No grade 4 or 5 toxicities occurred, and the local con- concurrent platinum- or taxane-based chemotherapy and
trol rate was 88.9 %. Similarly, Shioyama and colleagues compared their outcomes with those of patients treated in ear-
reported 28 patients with Stage I NSCLC treated with proton lier eras with photons [44]. The median total radiation dose
therapy to a median total dose of 76 CGE in median 3 CGE was 74 CGE for the proton group and 63 Gy for the photon
fractions (range, 2–6 CGE) [42]. The 5-year local control group. The proton patients had lower rates of pneumonitis,
rate was 89 % for patients with Stage IA disease with one esophagitis, and hematologic toxicity than those observed in
case of Grade 3 acute toxicity observed. patients in the photon cohort despite the fact that the proton
Interestingly, the outcomes achieved in the Chang and dose was, on average, higher. Finally, Chang et al. completed
Shioyama studies with modest hypofractionation were com- a phase II study of 44 patients with stage III NSCLC who
parable to photon-based ablative approaches. This suggests received 74 CGE via conventional fractionation (2 CGE per
that, in certain settings, the radiobiological advantages of fraction) with weekly concurrent carboplatin and paclitaxel
high LET radiation can substitute for the radiobiological [45]. Despite the very high intensity of this treatment course,
advantages of hypofractionation, while still allowing for the grade 3 toxicities were minimal: 5 had grade 3 esophagitis, 5
interfractional recovery of normal tissues. In cases where had grade 3 dermititis, and only 1 patient had grade 3 pneu-
normal tissue toxicity is not a major concern, it stands to monitis. No grade 4 or 5 toxicities were observed. Eighty per-
reason that combining ablative hypofractionation with cent local control and 29 months overall survival were
charged particle radiotherapy would result in maximal achieved. The concept about combining stereotactic radio-
efficacy. therapy and hypofractionated radiotherapy using intensity-
Finally, in the locally advanced setting, several studies modulated proton-based integrated boost and/or dose painting
have confirmed the utility of using charged particles, specifi- has been proposed to further improve the therapeutic ratio
cally protons, to increase the radiation dose while sparing and improve cure rate in stage III NSCLC.
normal tissue (Table 10.2). Though these results also do not
reflect a stereotactic approach, they are nonetheless instruc- Summary
tive for demonstrating the advantages of proton therapy for While the studies above demonstrate promising tumor con-
treating large volumes in the thorax. Especially notable are trol after the use of charged particles in the setting of lung
the low rates of severe toxicity observed despite the use of cancer, an important caveat is that these experiences are
characterized by considerable heterogeneity with respect to Clinical Outcomes

doses, fractionation schemes, and techniques for positioning, Hypofractionated proton therapy for liver tumors has been
immobilization, respiratory compensation, and geometric studied extensively by the University of Tsukuba Proton
verification. Nonetheless, they uniformly demonstrate proof Medical Research Center. Among the first reported experi-
of principle that charged particles achieve high local control ences was a series of 162 patients for hepatocellular carci-
while limiting normal tissue effects in the thorax. Whether noma with regimens utilizing fractions in the 3.5–5 CGE
the costs of charged particle radiation are justified by these range for total doses between 50 CGE (ten fractions) and 84
advantages, however, is controversial until a prospective ran- CGE (24 fractions). Notably, 72 patients (44 %) had Child-
domized study can be completed. Fortunately, a trial at MD Pugh class B and C cirrhosis. At a median follow-up interval
Anderson Cancer Center is underway to compare stereotac- of 31.7 months, the 5-year local control rate was 86.9 %, and
tic proton and photon treatments in centrally located stage I, the overall survival rate was 23.5 %. The acute side effects
selective stage II, and recurrent NSCLC using a dose regime were limited primarily to elevation in liver enzymes (9.7 %)
of 50 Gy in four fractions. The results of this study will shed and only 5 patients (3 %) had grade 2 or higher late toxicity.
light on whether the financial cost of proton therapy, specifi- Several other studies by the Japanese group showed similar
cally, is justified by improved clinical outcomes. outcomes for patients with difficult to treat hepatic tumors
[48, 50–52]. A more recent prospective study employing a
regimen of 66 CGE in ten fractions using proton radiotherapy
Hepatocellular Carcinoma was reported by Fukumitsu [53]. In this trial of 51 patients,
overall survival and local control rates at 3 years were 49.2 %
Rationale for Particle Therapy for Hepatic Tumors and 94.5 %, respectively. Only 3 patients experienced late
The liver is a parallel organ and, as such, is particularly sen- sequelae of Grade 2 or higher.
sitive to the toxic effects as more functional subunits are irra- In the USA, Bush et al. conducted a phase II trial of hypo-
diated. Therefore, photons’ integral dose distribution fractionated proton therapy in which 34 patients with unre-
presents challenges when high doses are desired for a target sectable HCC were treated to 63 CGE in 15 fractions [54].
within the liver. In point of fact, hepatocellular carcinoma The 2-year overall survival was 55 % with local control rates
often arises in the background of chronic hepatic insuffi- of 75 %. Fewer than 10 % of these patients experienced a
ciency, making the dosimetric characteristics of charged par- gastrointestinal bleed, and only mild acute liver toxicity was
ticles even more appealing for sparing still-functioning observed. Of interest, 2 of these patients were observed to
hepatocytes. Additionally specific patient populations may have complete pathologic response upon examination of
garner additional benefit. For patients with liver transplants, their explanted livers (they were among 6 patients who had
the ability to reduce integral dose with charged particles may gone on to receive liver transplant after proton therapy).
have important ramifications for reducing secondary malig- Studies employing carbon ion therapy for hepatic tumors
nancy in a population characterized by an immunocompro- have mainly been performed at the National Institute for
mised state [46]. For those with portal venous thrombosis a Radiological Science (NIRS) in Chiba, Japan. The main eligi-
large volume of liver often requires treatment and, often, the bility criterion for carbon ion treatment at NIRS was that other
size and location of the carcinoma and thrombosis preclude therapies had to be considered ineffective. Dose-escalation
conventional photon radiotherapy at higher doses [47, 48]. studies were implemented with various fractionation schemes
In light of these clinical considerations, charged particles from 15-fraction to single-fraction regimens. At the more
present significant theoretical advantages for radiotherapy of extreme end of hypofractionation, 69 patients treated with
hepatic tumors. In the conventionally fractionated setting, 52.8 CGE in four fractions. Therapy-related impairment in
dosimetric comparison studies have shown that protons hepatic function was minimal and the 5-year local control and
result in better sparing of the fractional liver volume receiv- survival rates were 81 % and 33 %, respectively [55].
ing more than 30 Gy [49]. Improvements in the mean liver Interestingly, toxicity rates were no different whether the tumor
dose, V45 for the stomach and duodenum, V40 and V50 for was within 2 cm of the main portal vein or beyond [56]. An
the heart, and the maximal dose to the spinal cord were also even shorter regimen of two fractions in 2 days was employed
observed. As with lung tumors, treating hepatic malignan- in a dose-escalation study, in which a total of 117 patients were
cies with radiation involves addressing issues related to treated to doses ranging from 32.0 to 45.0 CGE. Thus far, there
tumor motion secondary to diaphragmatic movements. have been no therapy-related deaths or severe adverse events,
Multiple techniques including implantable fiducial markers, and the patients receiving a higher dose appear to have better
four-dimensional simulation, daily cone-beam CT, abdomi- local control and survival [57, 58].
nal compression, and a combination thereof can be used to The main difference between the proton regimens at
guide particle therapy directed at liver tumors [47]. Tsukuba and the carbon ion regimens at Chiba are the degree
of hypofractionation. In the carbon ion studies, dose regi- Arteriovenous Malformations

mens more closely resembled what would be considered Fifty nine patients at Massachusetts General Hospital with high
radiosurgery. Promising results from these studies should risk high-risk cerebral arteriovenous malformations (AVMS)
prompt increased interest around the world for employing received two-fraction proton radiosurgery [60]. The majority of
stereotactic particle therapy for hepatic malignancies. these patients had undergone prior therapies, and different
radiosurgery doses were used, with 16 CGE in two fractions
prescribed to the 90 % isodose line being the most common. At
Central Nervous System Tumors a median follow-up of 56.1 months, 9 patients (15 %) had total
and 20 patients (34 %) had partial obliteration. Five-year actu-
As in the thorax, dose to tumors in the central nervous sys- arial rate of hemorrhage was 22%, and 14 % (8 patients) even-
tem is often limited by their proximity to multiple radiosen- tually suffered fatal hemorrhage. One patient developed a
sitive structures including the optic chiasm, brainstem, Grade 2 generalized seizure disorder and two had mild neuro-
hypothalamus, and spinal cord. Therefore, considerable logic deficits. Although the total obliteration rate was low, the
investigation has focused on whether high doses of radiation low toxicity rate observed in this study suggests that higher
can be delivered through particle therapy to achieve better doses could be considered in future studies.
dose distributions than photon-based treatments. However,
although research into the role of charged particle radiation Meningioma
for CNS tumors has been ongoing for over 50 years, most Halasz et al. retrospectively reviewed 50 patients with 51
centers have focused on conventionally fractionated or mod- histologically proven or image-defined, presumed-benign
estly hypofractionated regimens [59]. With the construction meningiomas treated between 1996 and 2007 [64]. The
of multiple proton therapy centers in the USA, there has been series included patients undergoing primary treatment as
renewed interest in combining the dosimetric advantages of well as those with residual or recurrent tumor following sur-
particles with radiosurgery for the treatment of intracranial gery. The median dose delivered was 13 CGE (range, 10.0–
and spinal tumors. Phase I–II studies investigating this 15.5 CGE) prescribed to the 90 % isodose line. With a
approach are anticipated. For now, retrospective reviews of median follow-up of 32 months, the 3-year actuarial tumor
patients treated at the Harvard Cyclotron provide prelimi- control rate was 94 %. Symptoms were improved in 47 % of
nary insight into the role of particle radiosurgery—specifi- patients. Potentially permanent adverse effects after proton
cally proton radiosurgery—for treating CNS tumors in the radiosurgery occurred in 3 patients. These results compare
contemporary era. Patients with a variety of central nervous favorably with historical outcomes using conventional
system tumors have been treated stereotactically with proton fractionation.
therapy during the last decade. For image-guidance and
localization, the Harvard Cyclotron relied on implantable Pituitary Tumors
skull fiducials; patients were immobilized using either skel- Twenty-two patients with acromegaly who failed transsphe-
etal pin cranial fixation or a modified relocatable frame with noidal resection underwent proton radiosurgery to a median
dental fixation [60]. These experiences are reviewed here. dose of 20 CGE (range, 15–24 CGE) delivered in one frac-
tion [65]. With a follow-up of 6.3 years, complete or partial
Acoustic Neuroma response was observed in 21 of 22 patients (95 %). A com-
Excellent tumor control was reported at the Massachusetts plete response defined as sustained (≥3 months) normaliza-
General Hospital in their experience using proton radiosurgery tion of insulin-like growth factor-I without medical
for acoustic neuromas with a median marginal dose of 12 CGE suppression was attained in 13 patients (59 %). Therefore,
to the tumor was delivered to the tumor, and implanted skull the majority of patients were able to attain a durable bio-
fiducials were used for stereotactic positioning [61, 62]. Five- chemical response without medication. Eight patients (38 %)
year tumor control was 94 %, while the trigeminal and facial eventually had new pituitary deficits. A different series of 38
nerve preservation rates were both around 90 %, respectively. patients with corticotrophin adenomas were treated with
More recently, a South African group retrospectively reported single-fraction proton radiosurgery to a median dose of
on 51 patients with large acoustic neuromas treated with hypo- 20 CGE (range, 15–20) [66]. Complete response (defined as
fractioned proton therapy to a dose of 26 CGE (isocenter) in sustained (≥3 months) normalization of urinary free cortisol
three fractions [63]. A noninvasive stereophotogrammatic off medical therapy or normalization of plasma corticotro-
patient support and positioning system utilizing the markers phin) was achieved in 23 patients.
on the mask, video cameras, and a computerized adjustable In both of the above pituitary series, no visual complica-
treatment chair was used for stereotactic positioning. The tions, seizures, clinical evidence of brain injury, or secondary
5-year local control was 98 %. Hearing preservation was tumors were observed. However, 38 % of the patients in the
42 %. Facial and trigeminal nerve preservation rates were first study and 52 % of the patients in the second developed
90.5 % and 93 %, respectively. new pituitary deficits during follow-up.
Future Directions for CNS Tumors rently taken these RBE variations into account. However,
Currently, most centers employing charged particles for with increased use of particle therapy in the coming years, it
CNS tumors rely on conventionally fractionated or modestly may be necessary to account for this variability in specific
hypofractionated treatment regimens. Nonetheless, the expe- circumstances.
rience at the Harvard Cyclotron demonstrates proof of prin- Proton therapy can generate neutron contaminants via the
ciple that CNS tumors can be targeted for true radiosurgical collision of protons with the gantry system as well as the
treatment within a charged particle gantry. As more proton patient himself. These secondary neutrons have high LET
and carbon ion centers come online, integrated systems for and can travel beyond the proton’s Bragg peak to induce
achieving immobilization, positioning, and image-guidance DNA damage in normal tissue structures. This may result in
will be needed to enable stereotactic treatment and thus late toxic effects or the development of secondary malignan-
improve the therapeutic ratio for patients. cies [76]. Heavier nuclei can likewise generate neutrons and
other nuclear fragments through interactions with atomic
nuclei. These fragments, like neutrons, are capable of caus-
The Challenges of Particle Therapy ing cellular damage beyond the primary particle’s Bragg
peak. However, it is important to note that the incidental irra-
Three major challenges to charged particle therapy are cost, diation attributable to these fragments may ultimately be less
motion/modeling uncertainties, and potential nuclear frag- carcinogenic than the low-dose bath of radiation seen in con-
ment contamination. With regard to cost, charged particles formal photon radiotherapy. In fact, preliminary evidence in
entail considerably higher investment than linear accelera- pediatric patients suggests that the overall risk of secondary
tors used for generating photons and electrons [67]. Because malignancy is lower when particles are used in place of pho-
of their large mass, protons and heavy nuclei require large tons [77].
cyclotrons and high energy inputs to achieve acceleration of Though these uncertainties in charged particle therapy
the therapeutic particle to the necessary energy for medical should inspire caution, the best strategy is to test the assump-
use [68]. The total cost to build a proton center has histori- tions, approximations, and algorithms used in charged parti-
cally been between $120 and $200 million. A general rule of cle treatment through prospective trials. Similarly, the impact
thumb is that the required size of the cyclotron or synchro- of cost and nuclear contamination should not deter the use of
tron is directly related to the size of the charged particle. charged particle therapy unless it can be shown, prospec-
Therefore, heavy nuclei gantries entail even greater capital tively, that these limitations outweigh the beneficial effects
investment [69]. of particle radiation. Ultimately, randomized controlled trials
Newer technologies that generate high energy protons in are needed to fully assess both the comparative efficacy and
more compact systems are emerging and may provide a more cost-effectiveness of charged particle therapy. Though no
affordable avenue for introducing particle therapy to every- true randomized trials have been completed thus far, promis-
day practice [70]. Systems employing a single-room gantry ing trials comparing proton and photon radiation in several
may cost as little as $20 million. Still, whether charged par- cancer sites have fortunately been opened.
ticles are cost-effective when compared to photon-based
alternatives remains a controversial subject among radiation
oncologists [67, 68, 71]. Conclusion
Several planning uncertainties also present a challenge to
the widespread adoption of charged particle radiotherapy. Charged particle radiotherapy has distinct advantages over
First of all, proton dose distribution is very sensitive to photon radiation in terms of dose distribution and radiobiol-
motion/anatomy changes [9, 10]. In addition, the conversion ogy. The characteristic Bragg peak of charged particle radi-
of CT Hounsfield units to charged particle stopping party is ation results in minimal dose delivery beyond the treatment
not as straightforward as the case in photon radiation [72]. volume and, consequently, relative sparing of organs-at-
Particle therapy treatment modeling systems are variably risk. Recent advances in patient positioning, immobiliza-
effective in predicting the precise depth of the Bragg peak, tion, and image-guidance have allowed investigation into
and this uncertainty in spatial resolution can be exacerbated the combination of particle therapy with a hypofractionated,
when the particles traverse air cavities [73]. Finally, though stereotactic treatment approach. This approach, “particle
most treatment planning systems assign a constant RBE for radiosurgery,” has resulted in promising local control, sur-
converting photon doses to charged particle doses, the true vival, and toxicity outcomes in multiple tumor sites.
RBE is likely to be dynamic along the range of the charged Randomized controlled trials will determine whether the
particle. Additionally, the RBE can vary for different cancer degree of improved efficacy and reduced morbidity follow-
histologies and for different dose/fractionation regimens [74, ing particle radiosurgery justifies its use over photon tech-
75]. For practical reasons, treatment modeling has not cur- niques despite higher financial costs.
18. Nagata Y, Hiraoka M, Mizowaki T, Narita Y, Matsuo Y, Norihisa Y,

References et al. Survey of stereotactic body radiation therapy in Japan by the
Japan 3-D Conformal External Beam Radiotherapy Group. Int J
Radiat Oncol Biol Phys. 2009;75(2):343–7.
1. Chang JY, Zhang X, Wang X, Kang Y, Riley B, Bilton S, et al. 19. Joyner M, Salter BJ, Papanikolaou N, Fuss M. Stereotactic body
Significant reduction of normal tissue dose by proton radiotherapy radiation therapy for centrally located lung lesions. Acta Oncol.
compared with three-dimensional conformal or intensity-modulated 2006;45(7):802–7.
radiation therapy in Stage I or Stage III non-small-cell lung cancer. 20. Timmerman R, McGarry R, Yiannoutsos C, Papiez L, Tudor K, DeLuca
Int J Radiat Oncol Biol Phys. 2006;65(4):1087–96. J, et al. Excessive toxicity when treating central tumors in a phase II
2. Auberger T, Seydl K, Futschek T, Sztankay A, Sweeney RA, Lukas study of stereotactic body radiation therapy for medically inoperable
P. Photons or protons: precision radiotherapy of lung cancer. early-stage lung cancer. J Clin Oncol. 2006;24(30):4833–9.
Strahlenther Onkol. 2007;183(Spec No 2):3–6. 21. Register SP, Zhang X, Mohan R, Chang JY. Proton stereotactic
3. Georg D, Hillbrand M, Stock M, Dieckmann K, Potter R. Can pro- body radiation therapy for clinically challenging cases of centrally
tons improve SBRT for lung lesions? Dosimetric considerations. and superiorly located stage I non-small-cell lung cancer. Int J
Radiother Oncol. 2008;88(3):368–75. Radiat Oncol Biol Phys. 2011;80(4):1015–22.
4. Zhang X, Li Y, Pan X, Xiaoqiang L, Mohan R, Komaki R, et al. 22. Macdonald OK, Kruse JJ, Miller JM, Garces YI, Brown PD, Miller
Intensity-modulated proton therapy reduces the dose to normal tis- RC, et al. Proton beam radiotherapy versus three-dimensional con-
sue compared with intensity-modulated radiation therapy or passive formal stereotactic body radiotherapy in primary peripheral, early-
scattering proton therapy and enables individualized radical radio- stage non-small-cell lung carcinoma: a comparative dosimetric
therapy for extensive stage IIIB non-small-cell lung cancer: a vir- analysis. Int J Radiat Oncol Biol Phys. 2009;75(3):950–8.
tual clinical study. Int J Radiat Oncol Biol Phys. 2009;77(2): 23. Onishi H, Shirato H, Nagata Y, Hiraoka M, Fujino M, Gomi K,
357–66. et al. Hypofractionated stereotactic radiotherapy (HypoFXSRT) for
5. Albertini F, Bolsi A, Lomax AJ, Rutz HP, Timmerman B, Goitein stage I non-small cell lung cancer: updated results of 257 patients in
G. Sensitivity of intensity modulated proton therapy plans to a Japanese multi-institutional study. J Thorac Oncol. 2007;2(7
changes in patient weight. Radiother Oncol. 2008;86(2):187–94. Suppl 3):S94–100.
6. Engelsman M, Rietzel E, Kooy HM. Four-dimensional proton treat- 24. Fakiris AJ, McGarry RC, Yiannoutsos CT, Papiez L, Williams M,
ment planning for lung tumors. Int J Radiat Oncol Biol Phys. Henderson MA, et al. Stereotactic body radiation therapy for early-
2006;64(5):1589–95. stage non-small-cell lung carcinoma: four-year results of a prospective
7. Chang JY, Cox JD. Improving radiation conformality in the treat- phase II study. Int J Radiat Oncol Biol Phys. 2009;75(3):677–82.
ment of non-small cell lung cancer. Semin Radiat Oncol. 25. Baumann P, Nyman J, Lax I, Friesland S, Hoyer M, Rehn Ericsson
2010;20(3):171–7. S, et al. Factors important for efficacy of stereotactic body radio-
8. Zhao L, Sandison GA, Farr JB, Hsi WC, Li XA. Dosimetric impact therapy of medically inoperable stage I lung cancer. A retrospective
of intrafraction motion for compensator-based proton therapy of analysis of patients treated in the Nordic countries. Acta Oncol.
lung cancer. Phys Med Biol. 2008;53(12):3343–64. 2006;45(7):787–95.
9. Hui Z, Zhang X, Starkschall G, Li Y, Mohan R, Komaki R, et al. 26. Nagata Y, Negoro Y, Aoki T, Mizowaki T, Takayama K, Kokubo M,
Effects of interfractional motion and anatomic changes on proton et al. Clinical outcomes of 3D conformal hypofractionated single
therapy dose distribution in lung cancer. Int J Radiat Oncol Biol high-dose radiotherapy for one or two lung tumors using a stereotac-
Phys. 2008;72(5):1385–95. tic body frame. Int J Radiat Oncol Biol Phys. 2002;52(4):1041–6.
10. Kang Y, Zhang X, Chang JY, Wang H, Wei X, Liao Z, et al. 4D 27. Hara R, Itami J, Kondo T, Aruga T, Uno T, Sasano N, et al. Clinical
Proton treatment planning strategy for mobile lung tumors. Int J outcomes of single-fraction stereotactic radiation therapy of lung
Radiat Oncol Biol Phys. 2007;67(3):906–14. tumors. Cancer. 2006;106(6):1347–52.
11. Ando K, Kase Y. Biological characteristics of carbon-ion therapy. 28. Grills IS, Mangona VS, Welsh R, Chmielewski G, McInerney E,
Int J Radiat Biol. 2009;85(9):715–28. Martin S, et al. Outcomes after stereotactic lung radiotherapy or
12. Gerweck LE, Kozin SV. Relative biological effectiveness of proton wedge resection for stage I non-small-cell lung cancer. J Clin
beams in clinical therapy. Radiother Oncol. 1999;50(2):135–42. Oncol. 2010;28(6):928–35.
13. Timmerman R, Paulus R, Galvin J, Michalski J, Straube W, Bradley 29. Ricardi U, Filippi AR, Guarneri A, Giglioli FR, Ciammella P,
J, et al. Stereotactic body radiation therapy for inoperable early Franco P, et al. Stereotactic body radiation therapy for early stage
stage lung cancer. JAMA. 2010;303(11):1070–6. non-small cell lung cancer: results of a prospective trial. Lung
14. Nagata Y, Takayama K, Matsuo Y, Norihisa Y, Mizowaki T, Cancer. 2009;68(1):72–7.
Sakamoto T, et al. Clinical outcomes of a phase I/II study of 48 Gy 30. Lagerwaard FJ, Haasbeek CJ, Smit EF, Slotman BJ, Senan
of stereotactic body radiotherapy in 4 fractions for primary lung S. Outcomes of risk-adapted fractionated stereotactic radiotherapy
cancer using a stereotactic body frame. Int J Radiat Oncol Biol for stage I non-small-cell lung cancer. Int J Radiat Oncol Biol Phys.
Phys. 2005;63(5):1427–31. 2008;70(3):685–92.
15. Uematsu M, Shioda A, Suda A, Fukui T, Ozeki Y, Hama Y, et al. 31. Nagata Y, Hiraoka M, Shibata T, Onishi H. A phase II trial of
Computed tomography-guided frameless stereotactic radiotherapy Stereotactic body radiation therapy for operable TIN0M0 non-
for stage I non-small cell lung cancer: a 5-year experience. Int J small cell lung cancer: Japan Clinical Oncology Group (JCOG0403).
Radiat Oncol Biol Phys. 2001;51(3):666–70. Int J Radiat Oncol Biol Phys. 2010;78:S27.
16. Xia T, Li H, Sun Q, Wang Y, Fan N, Yu Y, et al. Promising clinical 32. Senan S, Verstegen N, Haasbeek C, Slotman B, Lagerwaard
outcome of stereotactic body radiation therapy for patients with FJ. Outcomes of stereotactic body radiotherapy in patients with
inoperable Stage I/II non-small-cell lung cancer. Int J Radiat Oncol clinical stage I non-small cell lung cancer who are fit to undergo
Biol Phys. 2006;66:117–25. surgery. J Clin Oncol. 2011;29:Suppl abst 7051.
17. Stauder MC, Macdonald OK, Olivier KR, Call JA, Lafata K, Mayo 33. Bush DA, Slater JD, Shin BB, Cheek G, Miller DW, Slater
CS, et al. Early pulmonary toxicity following lung stereotactic body JM. Hypofractionated proton beam radiotherapy for stage I lung
radiation therapy delivered in consecutive daily fractions. Radiother cancer. Chest. 2004;126(4):1198–203.
Oncol. 2011;99:166–71.
34. Nihei K, Ogino T, Ishikura S, Nishimura H. High-dose proton beam therapy for patients with hepatocellular carcinoma. Int J Radiat
therapy for Stage I non-small-cell lung cancer. Int J Radiat Oncol Oncol Biol Phys. 2009;74(3):831–6.
Biol Phys. 2006;65(1):107–11. 54. Bush DA, Hillebrand DJ, Slater JM, Slater JD. High-dose proton
35. Hata M, Tokuuye K, Kagei K, Sugahara S, Nakayama H, Fukumitsu beam radiotherapy of hepatocellular carcinoma: preliminary results
N, et al. Hypofractionated high-dose proton beam therapy for stage of a phase II trial. Gastroenterology. 2004;127(5 Suppl 1):S189–93.
I non-small-cell lung cancer: preliminary results of a phase I/II 55. Kato H, Yamada S, Yasuda S, Yamaguchi K, Ohno I, Ohto M, et al.
clinical study. Int J Radiat Oncol Biol Phys. 2007;68(3):786–93. Four-fraction carbon ion radiotherapy for hepatocellular carci-
36. Iwata H, Murakami M, Demizu Y, Miyawaki D, Terashima K, noma. J Clin Oncol. 2004;22:4090. ASCO Annual Meeting
Niwa Y, et al. High-dose proton therapy and carbon-ion therapy for Proceedings (Post-Meeting Edition) (14S).
stage I nonsmall cell lung cancer. Cancer. 2010;116(10):2476–85. 56. Imada H, Kato H, Yasuda S, Yamada S, Yanagi T, Kishimoto R, et al.
37. Westover KD, Seco J, Adams JA, Lanuti M, Choi NC, Engelsman Comparison of efficacy and toxicity of short-course carbon ion radio-
M, et al. Proton SBRT for medically inoperable stage I NSCLC. J therapy for hepatocellular carcinoma depending on their proximity to
Thorac Oncol. 2012;7(6):1021–5. the porta hepatis. Radiother Oncol. 2010;96(2):231–5.
38. Miyamoto T, Baba M, Yamamoto N, Koto M, Sugawara T, Yashiro 57. Imada H, Yasuda S, Shinoto M, et al. Carbon ion radiotherapy for
T, et al. Curative treatment of Stage I non-small-cell lung cancer liver cancer. In: Proceedings of NIRS-ETOILE second joint sympo-
with carbon ion beams using a hypofractionated regimen. Int J sium on carbon ion radiotherapy. Centre Etoile, Lyon: NIRS-M-243;
Radiat Oncol Biol Phys. 2007;67(3):750–8. 2011. p. 46–53.
39. Miyamoto T, Baba M, Sugane T, Nakajima M, Yashiro T, Kagei K, 58. Tsujii H, Kamada T. A review of update clinical results of carbon
et al. Carbon ion radiotherapy for stage I non-small cell lung cancer ion radiotherapy. Jpn J Clin Oncol. 2012;42(8):670–85.
using a regimen of four fractions during 1 week. J Thorac Oncol. 59. Levy RP, Schulte RW, Slater JD, Miller DW, Slater JM. Stereotactic
2007;2(10):916–26. radiosurgery—the role of charged particles. Acta Oncol.
40. Okada T, Kamada T, Tsuji H, Mizoe JE, Baba M, Kato S, et al. 1999;38(2):165–9.
Carbon ion radiotherapy: clinical experiences at National Institute 60. Hattangadi JA, Chapman PH, Bussiere MR, Niemierko A, Ogilvy
of Radiological Science (NIRS). J Radiat Res. 2010;51(4):355–64. CS, Rowell A, et al. Planned two-fraction proton beam stereotactic
41. Chang JY, Komaki R, Wen HY, De Gracia B, Bluett JB, McAleer radiosurgery for high-risk inoperable cerebral arteriovenous mal-
MF, et al. Toxicity and patterns of failure of adaptive/ablative pro- formations. Int J Radiat Oncol Biol Phys. 2012;83(2):533–41.
ton therapy for early-stage, medically inoperable non-small cell 61. Harsh GR, Thornton AF, Chapman PH, Bussiere MR, Rabinov JD,
lung cancer. Int J Radiat Oncol Biol Phys. 2011;80:1350–7. Loeffler JS. Proton beam stereotactic radiosurgery of vestibular
42. Shioyama Y, Tokuuye K, Okumura T, Kagei K, Sugahara S, Ohara schwannomas. Int J Radiat Oncol Biol Phys. 2002;54(1):35–44.
K, et al. Clinical evaluation of proton radiotherapy for non-small- 62. Murphy ES, Suh JH. Radiotherapy for vestibular schwannomas: a
cell lung cancer. Int J Radiat Oncol Biol Phys. 2003;56(1):7–13. critical review. Int J Radiat Oncol Biol Phys. 2011;79(4):985–97.
43. Nakayama H, Satoh H, Sugahara S, Kurishima K, Tsuboi K, 63. Vernimmen FJ, Mohamed Z, Slabbert JP, Wilson J. Long-term
Sakurai H, et al. Proton beam therapy of Stage II and III non-small- results of stereotactic proton beam radiotherapy for acoustic neuro-
cell lung cancer. Int J Radiat Oncol Biol Phys. 2011;81(4):979–84. mas. Radiother Oncol. 2009;90(2):208–12.
44. Sejpal S, Komaki R, Tsao A, Chang JY, Liao Z, Wei X, et al. Early 64. Halasz LM, Bussiere MR, Dennis ER, Niemierko A, Chapman PH,
findings on toxicity of proton beam therapy with concurrent chemo- Loeffler JS, et al. Proton stereotactic radiosurgery for the treatment
therapy for nonsmall cell lung cancer. Cancer. 2011;117(13): of benign meningiomas. Int J Radiat Oncol Biol Phys.
3004–13. 2011;81(5):1428–35.
45. Chang JY, Komaki R, Lu C, Wen HY, Allen PK, Tsao A, et al. 65. Petit JH, Biller BM, Coen JJ, Swearingen B, Ancukiewicz M,
Phase 2 study of high-dose proton therapy with concurrent chemo- Bussiere M, et al. Proton stereotactic radiosurgery in management
therapy for unresectable stage III nonsmall cell lung cancer. Cancer. of persistent acromegaly. Endocr Pract. 2007;13(7):726–34.
2011;117(20):4707–13. 66. Petit JH, Biller BM, Yock TI, Swearingen B, Coen JJ, Chapman P,
46. Taddei PJ, Howell RM, Krishnan S, Scarboro SB, Mirkovic D, et al. Proton stereotactic radiotherapy for persistent
Newhauser WD. Risk of second malignant neoplasm following adrenocorticotropin-producing adenomas. J Clin Endocrinol
proton versus intensity-modulated photon radiotherapies for hepa- Metab. 2008;93(2):393–9.
tocellular carcinoma. Phys Med Biol. 2010;55(23):7055–65. 67. Lodge M, Pijls-Johannesma M, Stirk L, Munro AJ, De Ruysscher
47. Skinner HD, Hong TS, Krishnan S. Charged-particle therapy for D, Jefferson T. A systematic literature review of the clinical and
hepatocellular carcinoma. Semin Radiat Oncol. 2011;21(4):278–86. cost-effectiveness of hadron therapy in cancer. Radiother Oncol.
48. Hata M, Tokuuye K, Sugahara S, Kagei K, Igaki H, Hashimoto T, 2007;83(2):110–22.
et al. Proton beam therapy for hepatocellular carcinoma with portal 68. Konski A, Speier W, Hanlon A, Beck JR, Pollack A. Is proton beam
vein tumor thrombus. Cancer. 2005;104(4):794–801. therapy cost effective in the treatment of adenocarcinoma of the
49. Wang X, Krishnan S, Zhang X, Dong L, Briere T, Crane CH, et al. prostate? J Clin Oncol. 2007;25(24):3603–8.
Proton radiotherapy for liver tumors: dosimetric advantages over 69. Peeters A, Grutters JP, Pijls-Johannesma M, Reimoser S, De
photon plans. Med Dosim. 2008;33(4):259–67. Ruysscher D, Severens JL, et al. How costly is particle therapy?
50. Mizumoto M, Tokuuye K, Sugahara S, Nakayama H, Fukumitsu N, Cost analysis of external beam radiotherapy with carbon-ions, pro-
Ohara K, et al. Proton beam therapy for hepatocellular carcinoma tons and photons. Radiother Oncol. 2010;95(1):45–53.
adjacent to the porta hepatis. Int J Radiat Oncol Biol Phys. 70. Hede K. Research groups promoting proton therapy “lite”. J Natl
2008;71(2):462–7. Cancer Inst. 2006;98(23):1682–4.
51. Hata M, Tokuuye K, Sugahara S, Fukumitsu N, Hashimoto T, 71. Jones B. The case for particle therapy. Br J Radiol. 2006;79(937):
Ohnishi K, et al. Proton beam therapy for hepatocellular carcinoma 24–31.
with limited treatment options. Cancer. 2006;107(3):591–8. 72. Jiang H, Seco J, Paganetti H. Effects of Hounsfield number conver-
52. Sugahara S, Oshiro Y, Nakayama H, Fukuda K, Mizumoto M, Abei sion on CT based proton Monte Carlo dose calculations. Med Phys.
M, et al. Proton beam therapy for large hepatocellular carcinoma. 2007;34(4):1439–49.
Int J Radiat Oncol Biol Phys. 2010;76(2):460–6. 73. Andreo P. On the clinical spatial resolution achievable with protons
53. Fukumitsu N, Sugahara S, Nakayama H, Fukuda K, Mizumoto M, and heavier charged particle radiotherapy beams. Phys Med Biol.
Abei M, et al. A prospective study of hypofractionated proton beam 2009;54(11):N205–15.
74. Carabe A, Moteabbed M, Depauw N, Schuemann J, Paganetti 76. Brenner DJ, Hall EJ. Secondary neutrons in clinical proton radio-
H. Range uncertainty in proton therapy due to variable biological therapy: a charged issue. Radiother Oncol. 2008;86(2):165–70.
effectiveness. Phys Med Biol. 2012;57(5):1159–72. 77. Miralbell R, Lomax A, Cella L, Schneider U. Potential reduction of
75. Paganetti H. Nuclear interactions in proton therapy: dose and rela- the incidence of radiation-induced second cancers by using proton
tive biological effect distributions originating from primary and beams in the treatment of pediatric tumors. Int J Radiat Oncol Biol
secondary particles. Phys Med Biol. 2002;47(5):747–64. Phys. 2002;54(3):824–9.
CyberKnife
11
Carolina E. Fasola, Lei Wang, John R. Adler,
Scott G. Soltys, Iris C. Gibbs, Albert C. Koong,
and Daniel T. Chang
vasive alternative to surgery and frame-based stereotactic

Introduction of CyberKnife Stereotactic techniques. Additionally, in contrast to the frame-based tech-
Radiosurgery nology, high dose treatments with CyberKnife could be
divided into more fractions to reduce normal tissue dose, thus
The CyberKnife (Accuray Inc., Sunnyvale, CA) was devel- minimizing toxicity from treatment. The concept of stereotac-
oped as a tool to deliver large and highly conformal doses of tic body radiotherapy was first described in 1994 by investiga-
ionizing radiation to well-defined targets with steep dose tors in Sweden [1, 2]. With its ability to deliver image-guided
gradients to spare surrounding tissue. It was initially used in treatments, the Cyberknife was uniquely suited to explore
the 1990s for the treatment of intracranial lesions, both some of the earliest applications of stereotactic treatments in
benign and malignant. Prior to the development of the body. Initially beginning with organ motion management
CyberKnife, radiosurgery was frame based and largely cen- by gated breath-hold, the CyberKnife’s efficiency of treating
tered on the Gamma Knife (Elekta AB, Stockholm, Sweden) lesions throughout the body improved dramatically with the
system as well as conventional linear accelerator (linac)- addition of the Synchrony™ respiratory tracking system to
based systems. In contrast to the frameless CyberKnife, both the platform, thus broadening the scope of clinical application
Gamma Knife and linac-based systems require the applica- to include a wide variety of tumors. In this chapter, we
tion of an invasive frame to the patient’s head to achieve the describe the CyberKnife and some of its more unique clinical
desired accuracy. indications.
With the advent of radiation devices capable of improved
accuracy and superior organ motion management over the
past decade, prolonged courses of conventional radiation CyberKnife Technology
therapy can be quickly shortened to span a few high dose
treatments. While frame-based stereotactic techniques were The CyberKnife system consists of a lightweight and com-
already able to provide high dose radiation in as few as one pact 6-MV linac mounted on a robotic arm (robotic manipu-
treatment, the necessity of skeletal fixation using a head frame lator) with six axes of movement, allowing the linac to point
was uncomfortable for patients and, at the same time, limited virtually any direction (Kuka GmbH, Augsburg, Germany),
treatment to intracranial sites of disease. The CyberKnife shown in Fig. 11.1. The robotic manipulator is guided with a
image-guided localization system emerged with advances in computerized localization system consisting of two paired
imaging and computer technology, providing precise target kilovoltage X-ray sources fixed on the ceiling, two silicon
delineation and replacing the need for a rigid stereotactic image detectors installed in the floor on each side of the
frame for immobilization. The CyberKnife became a nonin- patient with a live field size of 20 cm × 20 cm, and a couch
with six degrees of freedom. The high-resolution orthogonal
digital images acquired with the patient on the couch are reg-
C.E. Fasola, M.D. • L. Wang, Ph.D. • S.G. Soltys, M.D.
istered to digitally reconstructed radiographs (DRRs) derived
I.C. Gibbs, M.D. • A.C. Koong, M.D., Ph.D. • D.T. Chang, M.D. (*)
Department of Radiation Oncology, Stanford University, from the planning computed tomography (CT) scan
Stanford, CA, USA (Fig. 11.2), which is used for patient setup and patient motion
e-mail: dtchang@stanford.edu correction during delivery.
J.R. Adler, M.D. The system uses a large number of narrow circular beams
Neurosurgery, Stanford University, Stanford, CA, USA emanating from a wide array of directions that accumulate
Varian Medical Systems, Stanford, CA, USA within the predefined target volume, thus allowing more than
148 C.E. Fasola et al.
Fig. 11.1 The main components of the CyberKnife system. (Modified image used with permission from Accuray Incorporated)
Fig. 11.2 A library of digitally reconstructed radiographs (DRRs) is generated from the treatment planning CT. Each of these images, which
approximates an oblique projection, emulates a unique pose of the patient’s anatomy
100 non-isocentric, noncoplanar beams to be used during the Recent developments in hardware and software have sig-
process of designing a treatment plan and creating a dose nificantly enhanced the delivery efficiency to reduce patient
distribution that conforms to even irregularly shaped target treatment times. The newer CyberKnife G4 system is capa-
volumes (Fig. 11.3). In contrast, frame-based isocentric ble of a dose rate up to 1,000 MU per minute in up to 1,600
radiosurgery systems (e.g., Gamma Knife and linac-based beam directions. The newly designed RoboCouch™ aligns
systems) construct spherical dose distributions around a dis- patients quickly from the control room with a high degree of
crete isocenter, and irregular targets require multiple spheres accuracy in six degrees of freedom so that setup time can be
to ensure adequate dose coverage. reduced. The IRIS variable collimator system is capable of
11 CyberKnife 149
Fig. 11.3 (a) With standard

radiosurgery dispersed isocentric
beams, all beams intersect a
common region (light gray).
Because multiple spherical
volumes are needed to cover
irregular lesions, the resulting
dose distribution tends to be
inhomogeneous.
(b) Non-isocentric beams from
various directions. Beams do not
all cross in a single point and
dose need not be inhomogeneous
Fig. 11.4 The IRIS collimator

includes two banks of tungsten
segments capable of rapidly
manipulating beam geometry to
deliver up to 12 beam sizes from
each position
changing the collimator size during delivery (Fig. 11.4). In The CyberKnife treatment planning system is capable of
combination, these features have been able to significantly generating both isocentric and non-isocentric treatment
reduce treatment time. For brain and spine cases, the typical plans. Non-isocentric treatment delivery marks a stark con-
treatment time in the current G4 system ranges from 15 min trast to earlier technology and is particularly useful in treat-
to 1 h depending on the complexity of the treatment sites, ing irregularly shaped targets. The total system accuracy is
with an average of approximately 30 min. similar to previously published results from frame-based
radiotherapy and is less than 1 mm [3, 4]. The CyberKnife
uses 12 secondary collimators to shape cylindrical beams
CyberKnife Treatment Planning with diameters ranging from 5 to 60 mm. While other treat-
ment planning systems may use forward planning, in which
Treatment planning with CyberKnife is similar to frame-based beam angles and weights are adjusted manually by the phys-
technology in that it is developed on CT image data with other icists or dosimetrists, all the CyberKnife planning is done
imaging modalities such as MRI, fMRI, and PET scans coreg- using an inverse approach. For more complex and irregu-
istered for target volume delineation. For lesions of the brain, larly shaped tumors, robotic radiosurgery provides an enor-
head, and neck, a thermoplastic head mask is fabricated for mous range of possibilities for placing beams and beam
immobilization. For body lesions, a custom vacuum body arrays. In designing plans for these targets, the goal of the
mold may be used. In order to generate high-resolution DRRs CyberKnife planning algorithm is to find a beam arrange-
for image guidance during delivery, the planning CT is ment that returns a maximally conformal distribution for a
acquired at 1.0–1.5 mm slice thickness. given shape of target [5, 6].
The optimization process is based on a set of goals and either four-dimensional CT or by deforming an inhalation
dose constraints specified by the user. The treatment plan- image stack into an exhalation image stack), which is then
ning system determines the directions and activation dura- used for treatment planning. The linear programming
tions for each non-isocentric, noncoplanar beam (selected approach used by the CyberKnife can be extended to four-
from a universe of approximately 2,000 beam directions) dimensional inverse planning [14].
such that the user’s specifications are met. A typical plan uti- One of the major limitations of CyberKnife is the extended
lizes more than 100 beam directions in a single treatment. delivery time for patients. Options for reducing the treatment
The sequential optimization algorithm was introduced in time are available in the planning system by reducing the
newer planning systems to improve the efficiency of the number of beams, nodes (positions of the robotic arm), and
planning process [7]. Unlike the standard optimization pro- monitor units while preserving the integrity of the treatment
cess, which optimizes on a set of clinical objectives at the plan [15]. In addition, the optimized path traversal feature
same time, the sequential optimization option is composed now allows the robotic arm to travel the shortest safe dis-
of a series of individual optimization steps, which are per- tance between nodes to deliver the treatment, thereby mini-
formed in a defined sequence in the order of priority. Each mizing the amount of motion required by the robot and
step corresponds to a clinical objective. The solution of each reducing treatment time significantly [16].
prior step is maintained or relaxed by a user-defined value
when the next step is optimized. The process closely mimics
the clinical decision making of a physician in designing a CyberKnife Treatment Delivery
treatment plan. The steps and user-defined constraints com-
prise a script that can be saved so that the treatment planning The CyberKnife treatment delivery follows a step-and-shoot
process with similar targets can be standardized, thus reduc- manner. The robot travels along a set of predefined paths,
ing planning time for subsequent plans. stopping at treatment nodes to deliver the dose. The design
The use of multiple collimators can improve plan confor- advantage of the CyberKnife treatment delivery is its
mality and reduce the treatment time [8]. The introduction of automated near real-time image guidance. During treatment,
the IRIS variable collimator, which provides an aperture size X-rays are frequently acquired and compared to the DRRs
ranging from 5 to 60 mm, has made it practical and feasible obtained at setup to automatically compensate for patient
to use more collimator apertures in treatment planning. The movement within a range of 10 mm. The robotic arm of the
IRIS collimator is designed with two sets of tungsten colli- CyberKnife is realigned before delivery of each beam to
mator banks, each of which forms a hexagonal shape rotated account for translation and rotation of the target. If a signifi-
30° apart to form a semicircular shape (Fig. 11.4). In the cur- cant misalignment is present, radiation delivery is postponed
rent implementation, the IRIS beam apertures are incremen- until the misalignment is corrected by movement of the treat-
tal to match the fixed collimator set. ment couch [17]. While the system is capable of imaging
Previously, the raytracing algorithm (pencil-beam-based before every treatment beam, imaging every third to fifth
algorithm) was used exclusively in the CyberKnife treatment beam or every 20–60 s, depending on the compliance of the
planning system. Since 2008, the Monte Carlo (MC) calcula- patient, is generally sufficient. The accuracy of this system
tion option has been available for CyberKnife treatment design has been demonstrated to be submillimetric [18, 19]
planning. The CyberKnife MC calculation is a Monte Carlo and is comparable to frame-based systems [18–22].
simulation with modifications to improve calculation speed The CyberKnife is capable of real-time tracking of targets
[9, 10]. This calculation algorithm more accurately predicts that move with respiratory motion with the Synchrony™ sys-
dose in inhomogeneous areas such as the lung as evidenced tem. Synchrony™ synchronizes internal motion of the target
by a study by Wilcox et al., who compared the traditional by X-ray localization with the motion of the chest wall as
raytracing calculation with the MC calculation in lung assessed by continuous infrared light emitting diodes placed
tumors for the CyberKnife system and showed that the ray- on the chest wall near the diaphragm that are detected by an
tracing calculation consistently overestimated doses in low- optical camera in the treatment room (Fig. 11.1). As the
density structures (the maximum dose deviation up to a patient breathes, the movement of the chest wall is correlated
factor of 1.63 observed), and this effect was more pro- with internal movement by a predictive model updated dur-
nounced as the collimator size decreased [11]. ing treatment (Fig. 11.5). The most recent 15 data points cap-
CyberKnife planning has also been extended to the case tured from the X-ray imagers are used for model prediction
of moving targets [12, 13]. While the target may move dur- updated in a first-in first-out manner. With the feedback from
ing treatment, critical structures in the vicinity may not move Synchrony™, the radiation beams remain fixed on the target
with the same speed or direction. The relative motion of while the robotic arm moves to account for tumor motion
organs with respect to one another requires highly accurate with respiration. If the detected target position is significantly
planning. The temporal changes over the entire respiration different from the predicted position when the patient’s
cycle are represented in a motion model (obtained from breathing pattern changes, the treatment will be turned off
11 CyberKnife 151
misalignment is corrected with frequent imaging during the

treatment. Treatment of cranial lesions with the CyberKnife
compares favorably to more conventional frame-based
intracranial radiosurgery in terms of dose conformality and
targeting accuracy [18].
Fiducial tracking with or without Synchrony™ is often
used in stereotactic body radiotherapy with CyberKnife.
Historically, fiducial tracking has been implemented to ade-
quately localize spinal targets. This tracking system involves
percutaneous implantation of radio-opaque fiducials under
fluoroscopic guidance, typically three to four 4 × 2 mm stain-
less steel screws placed in the vertebrae adjacent to the target
lesion. For other thoracic and abdominal lesions, at least
three small 0.8 × 4 mm gold seeds are placed in close prox-
imity to the lesion. The Fiducial tracking algorithm com-
pares the positions of the fiducials on X-rays acquired near
real time to their location on the DRRs and adjusts the treat-
ment beam accordingly [23]. If the target motion is correlated
with respiration, such as lung, liver and pancreas lesions, the
Fig. 11.5 This schematic illustrates several aspects of dynamic respi- Synchrony™ motion tracking option is usually used during
ratory compensation with the Synchrony™ system. In this situation, the delivery. If the target movement is caused by patient move-
movement of a lung tumor is determined in real time by correlating the
location of internal gold seeds (not visible) with external markers (IR ment or internal bowel movement, such as spine and pros-
emitters). Tumor position is then fed back to the robot, which adjusts tate, frequent imaging between beams is applied.
accordingly for tumor motion The clinically relevant accuracy of Fiducial tracking in
spinal lesions was investigated by Yu et al. [19]. In this study,
until the patient resumes a normal breathing pattern. A new measurements were performed at three different CyberKnife
modeling procedure may be required if the patient moves facilities using head and torso phantoms loaded with packs
significantly or the breathing pattern changes significantly of radiochromic film. The absolute displacement of the dose
(e.g., when the patient is awake vs. when the patient is sleep- contours from that defined during treatment planning was
ing). The additional tracking with Synchrony™ typically reported to be submillimeter. The average offset was
prolongs the treatment time by 20 %. 0.3 ± 0.1 mm. The fiducial tracking error was below 0.3 mm
With the capability of image-guided motion tracking, for radial translations smaller than 14 mm and less than
CyberKnife stereotactic radiosurgery represents a departure 0.7 mm for rotations up to 4.5°.
from the previous notion of radiosurgery as solely an intracra- For treatments that require Synchrony™, the tracking
nial treatment. Such refinements in technology have broad- accuracy largely depends on the accuracy of the correlation
ened the clinical spectrum to include treatment to sites in the model of the internal motion and chest wall motion. In end-
spine, head and neck, lung, prostate, pancreas, and liver. to-end tests with a phantom (with perfect linear correlation
between external motion and internal motion), the
CyberKnife is found to be able to follow the lesion with
Image Guided Tracking System submillimeter accuracy [24]. Perfect correlations do not
exist on real patients due to the irregular breathing and
Image-guided tracking of lesions is one of the key compo- hysteresis in tumor motion, so small residual errors remain
nents in the CyberKnife radiosurgery. Different image track- during treatment. The residual error is dependent on the
ing algorithms are implemented for different anatomic sites, quality of the correlation model and is proportional to patient
which include Skull tracking, Fiducial tracking, XSight and tumor motion. Pepin et al. studied the correlation and
Spine tracking, and XSight Lung tracking. prediction uncertainties based on 100 CyberKnife
For intracranial lesions, the Skull tracking algorithm is Synchrony™ models for lung patients [25]. With 95 % mod-
applied. It uses the bony anatomy of the skull as a reference eler point coverage, the uncertainty was 1.2 mm in the lateral
to continually track intracranial targets and correct for trans- direction, 1.7 mm in the anterior–posterior direction, and
lational/rotational target shifts during radiation delivery. 3.5 mm in the superior–inferior direction. Appropriate mar-
A 2D–3D fast image registration algorithm [3] registers gin should be considered during PTV delineation to account
the images acquired during treatment with the treatment for this uncertainty.
planning CT (through a library of pregenerated DRRs at Fiducial-free tracking is now available for spine radiosur-
different rotation angles) using skull bony anatomy. Any gery, which increases patient convenience and reduces risks
Fig. 11.6 XSight Spine tracking

compares the (left) digital
reconstructed radiography (DRR)
with the (right) live image in a
region of interest (ROI) defined
by the user which includes the
maximum spine structures
located on or near the treatment
area. The DRR is synthesized
with the moving structures
excluded to increase the
robustness of the tracking method
Fig. 11.7 XSight Lung tracking

directly tracks the tumor volume
without implanted fiducials. The
displacement is determined with
maximum similarity measures
within the matching window
associated with fiducial placement. XSight Spine tracking shown to target spinal lesions with submillimeter accuracy
was initiated in 2004 and allows for tracking of spinal skel- and compares favorably to previously published precision of
etal anatomy based on anatomic landmarks [26, 27]. Similar fiducial-based localization [4].
to the skull-based algorithm, the XSight Spine tracking algo- Similarly, XSight Lung Respiratory tracking system has
rithm aligns the live X-ray images with the DRRs generated emerged as an option for fiducial-less tracking of lung lesions
from the setup CT. DRRs are often synthesized with the [28]. The XSight Lung algorithm directory tracks the tumor
spine region only to remove moving structures and are using the live images taken during the treatment (Fig. 11.7).
enhanced with imaging processing for more robust tracking. The GTV is contoured during the planning process to be used
Instead of using the whole image, the XSight Spine algo- as the tracking target, and a maximum similarity matching
rithm only uses a region of interest defined by the user that technique is applied near the GTV region for tracking. This
includes the spine bony structures near the treatment area tracking system only corrects for translational displacements,
(Fig. 11.6). Because the spine is not a rigid body, the ROI is and the rotations of the tumor are ignored. The patient using
divided into many small nodes (9 × 9 mesh), and the small XSight Lung is set up with XSight Spine tracking first to cor-
area near each node is treated as a rigid body. They are regis- rect for any patient rotations. This system works best for
tered to the DRRs separately using maximum similarity lesions with sufficient contrast from normal surrounding tis-
mapping. The detected displacement of each node constructs sue. Fu et al. found this approach can be used on peripheral,
a displacement vector field which is used to determine the radiodense tumors with diameters >15 mm [29]. The XSight
overall translations and rotations. This approach has been Lung tracking system uses Synchrony™ to correct the
11 CyberKnife 153
detected internal motion that is correlated with respiratory tion is readily extended to lesions that originate beneath or
chest wall movement [24]. extend through the skull base, such as tumors of the foramen
In summary, the CyberKnife tracking systems in place magnum and nasopharyngeal carcinoma.
can track lesions in any direction throughout the body, thus Shimamoto et al. reported results of 94 patients with 136
facilitating treatment of both intra-cranial and extracranial brain metastases treated with doses of 24–27 Gy using
sites. Submillimeter accuracy is achieved with CyberKnife CyberKnife, showing 89 % local progression free survival
systems for Skull, Fiducial, and XSight Spine tracking meth- [30]. In a series from Stanford University, Chang et al.
ods [17]. For lesions that move with respiration, adequate treated 53 patients with 149 brain metastases with a median
margin should be considered. dose of 19.6 Gy (range 14–30 Gy) with 91 % rate of local
control, defined as stabilization or decrease in lesion size.
Eight patients (5.4 %) had radiation brain necrosis [31].
Clinical Indications Another series of 71 patients with 148 metastatic brain
lesions with mean volume of 6.6 cm3 treated with median
CyberKnife is capable of targeting most lesions treated with dose 20.7 Gy over 1–3 fractions showed local control rate of
the more conventional frame-based radiosurgery techniques. 83 % at 44 weeks of follow-up [32]. Hara et al. reported a
With the advent of stereotactic body radiotherapy, the clini- local control of 87 % in 62 patients with 145 renal cell or
cal indications for use with the CyberKnife have been melanoma metastases. The median prescribed dose was
expanded from intracranial to extracranial, including bone, 20 Gy (range 14–24 Gy). Prior surgical resection did not
lung, pancreatic, hepatic, and prostate disease. Doses of radi- impact local control [33].
ation vary depending on location. The main clinical goals of Without adjuvant irradiation, local control after surgical
stereotactic radiosurgery are to improve local control, allevi- resection of a brain metastasis is no greater than 40 % in
ate pain, or prevent tumor progression which could result in prospective trials, leading investigators to examine adjuvant
symptoms. While a full review of each of these disease sites radiosurgery to the resection cavity [34, 35]. Using
is beyond the scope of this chapter and can be found in other CyberKnife, Choi et al. reported high local control rates and
chapters of this textbook, what follows is a brief discussion acceptable toxicity in an updated series of 112 patients with
of each site specifically with respect to how CyberKnife has 120 resection cavities treated with adjuvant stereotactic
been used and the reported clinical results. radiosurgery with a median dose of 18.6 Gy (range
15–30 Gy). The average cavity volume was 8.5 cm3 with 12-
and 24-month cumulative local failure rates of 9.5 % and
Intracranial 11.6 %, respectively [36]. Stanford is now leading a prospec-
tive trial investigating dose escalation for large brain tumors
Most conditions treated with standard frame-based radiosur- or large resection cavities of 2–5 cm in size [16].
gery are readily treatable with the frameless CyberKnife
using the same basic radiosurgical principles. However, the Glioblastoma Multiforme
combination of near real-time image guidance and robotic Investigators have reported on the use of CyberKnife for pri-
treatment delivery can offer distinct advantages over conven- mary brain tumors such as glioblastoma multiforme (GBM).
tional frame-based radiosurgery under some clinical circum- Villavicencio et al. reported a series of 20 patients treated
stances. Because of its robotic nature, the CyberKnife with CyberKnife at the time of initial diagnosis of GBM and
enables an increase in both the range of beam trajectories 26 patients treated at the time of disease recurrence or pro-
and, in some cases, the volume of space over which they are gression [37]. The group treated with CyberKnife at time of
distributed. Clinically speaking, this design provides greater recurrence or progression had a median survival rate of 21
homogeneity, when desirable, and enhanced dose conformal- months compared to 11.5 months in the group treated upfront
ity for irregularly shaped targets. Even more importantly, fra- with stereotactic radiosurgery. Patients with more extensive
meless stereotaxy allows for multi-fraction dosing schedules, surgical resection had improved survival rates.
which may be used if late radiation damage is of concern. Conti et al. recently investigated a series of 23 patients
The CyberKnife can be used to treat all the common intra- with unresected GBM treated with radiation alone using
cranial conditions currently treated with other radiosurgery CyberKnife stereotactic radiosurgery alone or combined
systems, such as intracranial tumors, arteriovenous malfor- with temozolomide. Median survival was 12 months for the
mations, and functional disorders such as trigeminal neural- group that received combination treatment compared to 7
gia. However, the use of hypofractionation makes it feasible months in the group that received stereotactic radiosurgery
to treat many larger tumors or lesions adherent to particu- alone. While surgery remains the standard of care for the
larly radiosensitive brain structures such as the anterior treatment of GBMs, the vast majority of failures (90 %)
visual pathways. Moreover, CyberKnife radiosurgical abla- occur within 2 cm of the original tumor [38]. CyberKnife
stereotactic radiosurgery is currently being investigated as with CyberKnife stereotactic radiosurgery to 18–21 Gy in
adjunct therapy following surgical resection in a prospective three fractions with 98 % local control, 74 % serviceable
trial at Stanford. hearing preservation, and no reports of trigeminal nerve
injury at 3 years of follow-up [45]. More recently Stanford
Perioptic Tumors updated their results of fractionated stereotactic radiosurgery
The use of hypofractionation enables treatment of lesions for 383 patients. The majority of patients received 18 Gy in
near radiosensitive tissue such as the optic structures, which three fractions. At a median follow-up of 3.6 years, local
are not well suited for single fraction treatment. High rates of control was 99 %. Serviceable hearing was preserved in
local control and visual preservation after hypofractionated 76 % of patients (among 200 patients) with serviceable hear-
radiosurgery have been reported [39–41]. Adler et al. updated ing prior to treatment. There were no cases of treatment-
a series of 49 patients with perioptic tumors within 2 mm of related facial nerve dysfunction, but 2 % of patients did
the optic apparatus treated with CyberKnife radiosurgery experience trigeminal nerve injury [46].
with excellent local control rates. Lesions included menin-
giomas, pituitary adenomas, craniopharyngiomas, and one
case of mixed germ cell tumor, most of which had been pre- Trigeminal Neuralgia
viously treated with surgical resections and/or conventional
radiation therapy. These lesions were treated with a mean Trigeminal neuralgia (TN) is characterized by chronic epi-
cumulative marginal dose of 20.3 Gy in 2–5 fractions. They sodic facial pain which is often refractory to medications.
reported local control >95 % and visual preservation rate of Frame-based radiosurgical rhizotomy, most commonly per-
94 % at 49 months of follow-up. There was one case of formed with the Gamma Knife, has been shown to be effec-
radiation-induced optic nerve injury in a patient previously tive in the management of TN. The standard Gamma Knife
treated with conventional radiation therapy and three courses rhizotomy was modified using the capacity of the CyberKnife
of perioptic radiosurgery [39]. to deliver non-isocentric plans. Rapid onset of pain relief
after CyberKnife treatment was first reported in a small
Vestibular Schwannomas series of patients by Romanelli et al. [47]. Lim et al. reported
Vestibular schwannomas are benign tumors that can cause a series of 29 patients with TN who had previously tried
significant morbidity including impaired hearing and bal- medical therapy and/or surgical procedures with persistent
ance from dysfunction of the vestibulocochlear nerve [42]. symptoms treated with CyberKnife stereotactic radiosurgery
While observation without initial treatment for these benign to a dose ranging between 60 and 70 Gy to a 6–8 mm length
lesions is a reasonable option in some cases, surgical resec- of the retrogasserian region of the trigeminal nerve. At 10
tion and stereotactic radiosurgery (single or fractionated ses- months of follow-up, 90 % of patients reported significant
sions) are standard treatment approaches. Compared to pain relief with a median time of 7 days. With regards to
surgery, stereotactic radiosurgery offers similar rates of local facial numbness, 14 % of patients reported worsening facial
control between 95 and 98 % with fewer side effects and numbness and 10 % reported new onset facial numbness fol-
complications and superior long-term hearing preservation lowing treatment [48]. Villavicencio et al. reported a multi-
rates [43]. Hearing preservation rates largely depend on pre- center experience of treatment of TN with CyberKnife
treatment hearing status. Andrews et al. presented a series of involving 95 patients. In this series, 67 % of patients reported
125 patients with vestibular schwannomas treated with significant pain relief at a median time of 14 days, but only
Gamma Knife stereotactic radiosurgery (n = 69) or linac- 50 % of patients had pain relief at 2 years of follow-up. Up
based fractionated stereotactic radiotherapy (n = 56). Local to 47 % of patients experienced some degree of facial numb-
control was >97 % for both groups, but functional hearing ness following treatment [49].
preservation was 2.5-fold higher in patients who were treated In targeting the segment of the retrogasserian cisternal
with the linac-based fractionated treatment compared to sin- portion of the trigeminal sensory root, pain relief can often be
gle session radiosurgery [44]. A recent meta-analysis on ste- achieved, but with a high incidence of facial numbness as
reotactic radiosurgery discussed 37 studies of treatment of shown in the above series. As a result, a gradual dose and
vestibular schwannomas with overall local control of 91.1 %, volume de-escalation approach was used by Adler et al. to
serviceable hearing preservation rate of 59.3 %, and posttreat a 6 mm segment of the nerve with a dose prescription of
treatment facial nerve dysfunction of 7.1 %. Of note, most of 60 and 75 Gy maximum dose. Among 46 patients, 85 %
these studies focused on single fraction stereotactic radiosur- experienced complete resolution of pain at a mean of 5 weeks.
gery in treatment of these lesions [42]. One of these patients had recurrence of symptoms at 7 months
In support of fractionated treatment of vestibular schwan- and was treated with a repeat session with resolution of
nomas, Chang et al. published a series of 61 patients treated symptoms. At a mean follow-up of 14 months, 96 % of
11 CyberKnife 155
patients reported >90 % improvement in symptoms and 15 % the Munich experience of single fraction radiosurgery using
reported significant facial numbness [50]. In an Italian series, XSight Spine (fiducial-free tracking). They reported a series
33 patients with TN were treated with CyberKnife to doses of 102 patients with 134 malignant spinal tumors, of which
between 55 and 75 Gy targeting a length of 4 mm of the ret- the majority were spinal metastases, treated 15–24 Gy in a
rogasserian portion of the trigeminal nerve, and 94 % of single fraction. In this cohort, 32 % of patients received prior
patients reported initial pain relief with a median time to pain spinal radiation. At a median follow-up of 15 months, local
relief of 30 days. There were no documented cases of facial control was 98 %. In an update, they reported the results of
numbness following treatment. However, 33 % of patients 287 patients with 348 lesions showing a local control rate of
experienced relapse of symptoms at a mean of 9 months [51]. 94 % at median follow-up of 9.6 months and only one patient
Currently, a dose of 60 Gy with maximum dose limited to developed myelopathy [55]. Heron et al. published the com-
75 Gy prescribed along 5 mm of the trigeminal nerve is used bined experience at University of Pittsburgh and Georgetown
at Stanford for TN. The target volume is kept 2–3 mm away University, which included 228 patients with 348 lesions
from the pons, and the brainstem maximum dose is limited to treated with CyberKnife spinal radiosurgery. Patients
<25 Gy and V12 < 0.2 cc in an effort to minimize the risk of received a mean dose of 16.3–24.5 Gy depending on frac-
brainstem necrosis. tionation (1–5 sessions). Pain control was excellent among
all patients and superior among patients treated in a single
session (100 %) compared to patients treated in multiple ses-
Spinal Indications sions (88 %). However, local control (96 % vs. 70 %), the
rate of retreatment (1 % vs. 13 %), and overall survival (63 %
With the degrees of freedom in beam arrangements available vs. 46 %) at 2 years were all superior among patients treated
with the robotic arm, spinal radiosurgery has emerged as a with multiple fractions compared to patients treated with a
natural application for CyberKnife with encouraging results. single fraction [58].
In the first reported use of image-guided robotics to perform These results demonstrate the safety and efficacy of using
spinal radiosurgery, Ryu and colleagues demonstrated the CyberKnife radiosurgery for treatment of spinal lesions. The
safety and short-term efficacy for a variety of neoplastic and local control and symptomatic relief rates are particularly
vascular lesions [52]. Similar to the principles of intracranial impressive given the high proportion of patients who had
radiosurgery, the goals remain controlling local disease to been treated with prior radiation therapy.
prevent or palliate bony pain, vertebral fracture, and/or nerve
and spinal cord damage. In an updated study, Gerszten et al.
reported on 125 patients treated with single fraction spinal Intrathoracic Tumors
radiosurgery to include 393 patients with 500 metastatic spi-
nal lesions [53, 54]. Lesions in the cervical spine were Many patients with non-small cell lung cancer present with
tracked using skull-based landmarks while tracking for lower underlying pulmonary disease precluding them from surgi-
spine lesions was based on implanted gold fiducials at the cal resection. Historically, conventionally fractionated radio-
vertebral bodies of interest. Of these patients, 69 % had therapy has been the treatment of choice for early stage
received prior radiation treatment. Patients received stereo- medically inoperable patients, but outcomes have been poor
tactic radiosurgery in a single fraction of 12.5–25 Gy. At a [59, 60]. SBRT was developed as a means to dose escalate
median follow-up period of 21 months, local control was radiation to improve local control and survival. Extracranial
88 % among all patients and 90 % in a group of patients stereotactic radiosurgery was first described by Lax and
without prior radiation. At 53 months of follow-up, long- Blomgren showing very high rates of tumor control in the
term pain relief was reported in 86 % of patients, and there lungs and abdomen [1, 2]. Since then, SBRT has emerged as
were no cases of radiation myelopathy [55, 56]. a standard treatment for early stage lung cancer with
In a report on the spinal CyberKnife experience at extremely good outcomes reported in prospective and retro-
Stanford, 74 patients with 102 spinal metastases were treated spective studies. While a full discussion of published data is
with 16–25 Gy in 1–5 fractions with the majority of patients beyond the scope of this chapter, in general, local control has
having received prior radiation. At a mean follow-up of 9 been reported to exceed 90 % with survival rates exceeding
months, 84 % of symptomatic patients experienced resolu- that of historical rates [61, 62].
tion or improvement of symptoms. There were three patients The feasibility of CyberKnife radiosurgical ablation
who developed treatment-related spinal injury. Upon further for pulmonary lesions was first investigated in 23 patients
investigation, there were no factors seen in dosimetry that with biopsy-proven lung tumors (15 primary and eight meta-
could account for the cases of spinal cord injury [57]. static lesions) by Whyte and coworkers [35]. After CT-guided
Other institutions have also reported their data on spinal percutaneous fiducial placement, each patient was treated
radiosurgery with the CyberKnife. Wowra et al. published with 15 Gy in a single fraction. Although radiosurgery itself
Table 11.1 Select studies reporting results of CyberKnife for lung tumors
N Tumors Dose Local control 1-Year survival
Brown et al. [70] (2009) 31 P 60–67.5 Gy/3–5 93 % 94 %
Le et al. [69] (2006) 32 P+M 15–30 Gy/1 78 %(P)/58 %(M) 85 %(P)/56 %(M)
van der Voort et al. [68] (2009) 70 P 45–60 Gy/3a 96 %(60 Gy)/78 %(45 Gy) 83 %
van der Voort et al. [67] (2010) 38 P 45–60 Gy/3–6 100 % 65 %
b
Pennathur et al. [66] (2009) 100 P+R+M 20 Gy or 60 Gy/3 67% (P) 44 % (P 2-year)
b
55% (R) 41 % (R 2-year)
b
42% (M) 84 % (M 2-year)
c
Vahdat et al. [65] (2010) 20 P 42–60 Gy/3 95 % 90%
c
Unger et al. [64] (2010) 20 P+M 30–40 Gy/5 63 % 54 %
Chen et al. [63] (2012) 40 P 42–60 Gy/3 91 % (3-year) 75 % (3-year)
Bibault et al. [28] (2012) 51 P 36–60 Gy/3 92 % 86 %
P primary non-small cell lung cancer, M metastatic tumors, R recurrent non-small cell lung cancer
a
1 patient received 36 Gy/3
b
Estimated from Kaplan–Meier curve
c
Larger central hilar tumors
was well tolerated, several patients experienced complica- the literature. Important to note is that the risk of grade 2 or
tions as a result of the implantation of fiducials. After limited higher gastrointestinal toxicity was 25 % at 1 year, highlight-
postsurgical follow-up, radiographic progression was found ing the importance of dose conformality and normal tissue
only in two patients; the follow-up period ranged from 1 to sparing.
26 months (mean, 7 months). Because several local failures Other investigators have also reported on CyberKnife
were subsequently observed, the radiosurgical dose for lung SBRT for pancreas tumors. Mahadevan et al. reported on 47
cancer has been gradually escalated. patients treated for locally advanced pancreas cancer, giving
Since then, there have been dozens of additional reports treatment after two cycles of gemcitabine [74]. Because of
showing the efficacy of CyberKnife for lung SBRT, some of the risk of duodenal toxicity, the investigators used a risk-
which are summarized in Table 11.1 [28, 63–70]. It is adapted dose based on tumor proximity to the duodenum.
expected that CyberKnife will continue to play an integral The dose ranged from 24 to 36 Gy in three fractions. Local
role in the treatment of early stage lung tumors and oligo- control was 85 % with a very encouraging median survival of
metastatic pulmonary disease, and the body of literature on 20 months. The rate of late grade 3 toxicity was 9 %. Studies
its clinical outcomes will grow significantly. using CyberKnife for pancreatic tumors are shown in
Table 11.2 [73–81].
Overall, rates of local control are encouraging and appear
Pancreatic Tumors superior to those reported for conventionally fractionated
external beam radiation. The duodenum, bowel, and stomach
The outcomes of unresectable pancreas cancer using frac- remain critical dose-limiting organs, and an improved under-
tionated radiotherapy is generally poor, with local recurrence standing of tolerance of these structures to hypofractionation
rates ranging from 20 to 60 % [71]. In an attempt to improve is needed to further improve the therapeutic ratio. In addi-
on these results, SBRT has been used as a method of dose tion, the high rate of distant metastases continues to be the
escalation. CyberKnife SBRT was first reported by Koong predominant mode of failure and death, and improvements in
et al. in a phase I dose escalation trial [72]. Fifteen patients systemic therapy may allow CyberKnife and SBRT to play a
were treated with 15, 20, or 25 Gy delivered as a single frac- potentially larger role in the treatment of this disease.
tion. No grade 3 or higher acute toxicity was observed, and
none of the 6 patients treated at 25 Gy in a single fraction
developed a local recurrence. Since then, the investigators Liver Tumors
have expanded on their experience with single fraction
CyberKnife treatment, with a local control rate of 84 % at 1 Liver SBRT continues to receive a great deal of interest, both
year and an isolated local failure rate of 5 % at 1 year [73]. for metastases and for primary tumors. Given the large
Although CyberKnife SBRT has proved highly effective in organ size, focal SBRT can potentially ablate tumors with
achieving local control, survival continues to be largely a high rate of success and a low risk of organ damage.
determined by the high rate of distant metastases, as the While resection remains the standard of care, tumors may be
median survival was 11.9 months, similar to that reported in unresectable due to tumor location or medical comorbidities
11 CyberKnife 157
Table 11.2 Studies reporting results of CyberKnife for pancreatic tumors

Series N Dose Local control (%) Overall survival (months) Toxicity
Chang et al. [73] (2009) 77 25 Gy/1 fraction 84 11.7 25 % Gr 3
Mahadevan et al. [74] (2011) 47 24–36 Gy/3 fractions 85 20 9 % Gr 3
Polistina et al. [81] (2010) 23 30 Gy/3 fractions 82a 10.6 0 % Gr 2+
Rwigema et al.b [80] (2012) 24c 20–24 Gy/1 fractiond 66e 26.7 0 % Gr 3+
Schellenberg et al. [79] (2008) 16 25 Gy/1 fraction 81 11.4 12.5 % Gr 3
Seo et al.f [78] (2009) 30 14–17 Gy/1 fraction 70 14 3 % Gr 4
Didolkar et al. [77] (2010) 85 15–30 Gy/3 fractions 92 18.6 22.3 % Gr 3+
Goyal [76] (2012) 20 22–30 Gy/3 fractions 81 14.4 16 % Gr 3
Lominska et al. [75] (2012) 28g 20–30 Gy/3–5 fractions 86 5.9 7 % Gr 3
a
At 6 months among 11 of 23 patients alive
b
Treatment given adjuvantly for close or positive margin
c
18 of 24 patients treated with CyberKnife
d
1 patient received 30 Gy in three fractions
e
At 1 year
f
Treatment given as boost after 40 Gy in 20 fractions
g
Treatment given for reirradiation
thus requiring nonsurgical ablative techniques such as radio- colorectal cancer, more patients will likely present with
frequency ablation, cryoablation, chemoembolization, and oligometastatic disease that requires locally ablative thera-
radioembolization. Several reports have been published on pies. Thus, SBRT and CyberKnife are poised to play an
SBRT showing encouraging local control rates [82–86]. As a increasingly important role in the management of these
result, SBRT has become an additional tool available for the patients.
treatment of unresectable liver tumors, and CyberKnife has
been successfully implemented for this clinical indication.
The Synchrony™ system allows for real-time respiratory Future Directions
tracking to compensate for breathing motion.
Investigators at Stanford have reported on single fraction Although preliminary, the above results with abdominal and
SBRT for liver tumors using CyberKnife, successfully esca- thoracic tumors have been encouraging. Nevertheless, much
lating the dose from 18 to 30 Gy with 2 patients experiencing more evidence is needed to establish a definite role for
grade 2 gastrointestinal toxicity among 26 patients [87]. The robotic radiosurgery in managing non-neurologic disorders.
cumulative local failure rate was 23 %. Lanciano et al. To this end, multiple clinical trials are under way worldwide
reported on 48 patients treated with liver tumors, showing that seek to discover if CyberKnife radiosurgery can sub-
that local control was dose dependent. Patients treated with a stantially impact the overall clinical outcome for a broad
biologically equivalent dose (BED) of >100 Gy had a 2-year range of extracranial indications.
local control of 75 % vs. 38 % for patients treated with Fiducial targeting of lesions within the major body cavi-
<100 Gy BED [88]. Goyal et al. reported 82 % local control ties is both robust and extremely accurate. However, the
in 17 patients treated for primary and metastatic liver tumors implantation process can be technically demanding, adds to
[89]. Others have reported results using CyberKnife specifi- the complexity of the overall radiosurgical procedure, and is
cally for hepatocellular carcinoma, with or without combina- modestly invasive. In the case of lung treatment, pneumotho-
tion treatment with transarterial chemoembolization, showing rax is not uncommon. Although the challenges are consider-
favorable response rates [90], suggesting that SBRT could be able, a technology for targeting intra-abdominal and thoracic
used as a bridge therapy for liver transplantation [91]. lesions without fiducials will be a clear improvement in
Much remains to be learned about the role of SBRT for robotic radiosurgery. Recent research suggests that motion
liver tumors including the optimal dose schedule, tolerance compensation without implanted gold markers may be clini-
of the liver and other abdominal organs using hypofraction- cally practical [39]. Such technology would considerably
ation, and how best to combine SBRT with systemic agents simplify treatment protocols and enhance patient comfort
and other ablative therapies. As our understanding of these and safety.
issues improves through continued clinical experience and Among the other stereotactic radiosurgery systems,
longer follow-up, the indications for SBRT will surely CyberKnife distinguishes itself with its automated image-
expand. As advancements in systemic therapies are made as guided tracking system and noncoplanar and non-isocentric
well, particularly for chemosensitive malignancies such as delivery fashion. The 2D–3D image registration is fast, the
dose accumulated from imaging is low, and the images are with systemic agents and other locally ablative therapies
excellent for localizing bony structures. The tracking process needs to be further investigated and likely represents the next
is highly automated, making the setup procedure quick and generation of studies. The CyberKnife is anticipated to play
easy. The noncoplanar and non-isocentric delivery method is an important role in this effort.
powerful in creating highly conformal plans without sacrific-
ing the uniformity of the resulting dose distribution.
However, when treating larger targets (e.g., >3 cm), the References
CyberKnife delivery time still faces challenges. The delivery
efficiency of the CyberKnife system is limited by the small 1. Blomgren H, Lax I, Naslund I, Svanstrom R. Stereotactic high dose
sizes of the collimators and the step-and-shoot delivery man- fraction radiation therapy of extracranial tumors using an accelera-
tor. Clinical experience of the first thirty-one patients. Acta Oncol.
ner. While the patient setup time is very quick (less than
1995;34(6):861–70. Epub 1995/01/01.
5 min), the delivery time for a 5–6 cm volume could be up to 2. Lax I, Blomgren H, Naslund I, Svanstrom R. Stereotactic radio-
45–60 min with the current G4 system. On the other hand, therapy of malignancies in the abdomen. Methodological aspects.
with a gated volumetric modulated arc therapy for a similar Acta Oncol. 1994;33(6):677–83. Epub 1994/01/01.
3. Fu D, Kuduvalli G. A fast, accurate, and automatic 2D-3D image
target, treatment can be delivered within 5 min once the
registration for image-guided cranial radiosurgery. Med Phys.
patient is set up. An MLC attached to the CyberKnife could 2008;35(5):2180–94. Epub 2008/06/20.
significantly reduce the treatment time. Further improve- 4. Ho AK, Fu D, Cotrutz C, Hancock SL, Chang SD, Gibbs IC, et al.
ments could be achieved with the implementation of contin- A study of the accuracy of cyberknife spinal radiosurgery using
skeletal structure tracking. Neurosurgery. 2007;60(2 Suppl
uous delivery method similar to the arc delivery on traditional
1):ONS147–56; discussion ONS56. Epub 2007/02/14.
linear accelerator systems. 5. Webb S. Conformal intensity-modulated radiotherapy (IMRT)
Alternative image-guided localization approaches could delivered by robotic linac—testing IMRT to the limit? Phys Med
be important to improve targeting. The CyberKnife system Biol. 1999;44(7):1639–54. Epub 1999/08/12.
6. Webb S. Conformal intensity-modulated radiotherapy (IMRT)
does not provide a volumetric imaging system, which, though
delivered by robotic linac—conformality versus efficiency of dose
perhaps less critical for intracranial lesions, could be very delivery. Phys Med Biol. 2000;45(7):1715–30. Epub 2000/08/16.
valuable for locating soft tissue structures in the thorax or 7. Schlaefer A, Schweikard A. Stepwise multi-criteria optimization
abdomen where appreciable deformation and movement for robotic radiosurgery. Med Phys. 2008;35(5):2094–103. Epub
2008/06/20.
occurs (e.g., liver, pancreas, lung, and prostate). Localization
8. Poll JJ, Hoogeman MS, Prevost JB, Nuyttens JJ, Levendag PC,
with non-ionizing methods, including optical, ultrasound, Heijmen BJ. Reducing monitor units for robotic radiosurgery by
and radiofrequency beacon systems could be interesting optimized use of multiple collimators. Med Phys. 2008;35(6):
areas to explore. Recently, XSight Spine tracking has been 2294–9. Epub 2008/07/25.
9. Deng J, Guerrero T, Ma CM, Nath R. Modelling 6 MV photon
made available for prone positioning to reduce the length of
beams of a stereotactic radiosurgery system for Monte Carlo treat-
the beam path to the target of posterior targets, allowing ment planning. Phys Med Biol. 2004;49(9):1689–704. Epub
improved dose distribution and reduced normal tissue 2004/05/22.
irradiation. 10. Deng J, Ma CM, Hai J, Nath R. Commissioning 6 MV photon
beams of a stereotactic radiosurgery system for Monte Carlo treat-
ment planning. Med Phys. 2003;30(12):3124–34. Epub 2004/01/10.
11. Wilcox EE, Daskalov GM, Lincoln H, Shumway RC, Kaplan BM,
Conclusion Colasanto JM. Comparison of planned dose distributions calculated
by Monte Carlo and Ray-Trace algorithms for the treatment of lung
tumors with cyberknife: a preliminary study in 33 patients. Int J
CyberKnife is a radiosurgical system well suited for stereo-
Radiat Oncol Biol Phys. 2010;77(1):277–84. Epub 2009/12/17.
tactic radiotherapy. The introduction of image guidance and 12. Li JG, Xing L. Inverse planning incorporating organ motion. Med
robotic delivery has dramatically expanded the scope of this Phys. 2000;27(7):1573–8. Epub 2000/08/18.
field. With the strong interest in SBRT and extracranial 13. Unkelbach J, Oelfke U. Incorporating organ movements in IMRT
treatment planning for prostate cancer: minimizing uncertainties
radiosurgery, indications for CyberKnife will continue to
in the inverse planning process. Med Phys. 2005;32(8):2471–83.
grow. As inevitable improvements in neuroimaging and Epub 2005/10/01.
computer technology emerge over the coming years, they 14. Schlaefer A, Fisseler J, Dieterich S, Shiomi H, Cleary K,
will serve as an impetus for further improvements in robotic Schweikard A. Feasibility of four-dimensional conformal planning
for robotic radiosurgery. Med Phys. 2005;32(12):3786–92. Epub
radiosurgical technology. There is much to be learned on the
2006/02/16.
role of SBRT for various disease sites, but with improve- 15. van de Water S, Hoogeman MS, Breedveld S, Heijmen
ments in technology and systemic therapy, the field of SBRT BJ. Shortening treatment time in robotic radiosurgery using a novel
will continue to expand. As our understanding of normal tis- node reduction technique. Med Phys. 2011;38(3):1397–405. Epub
2011/04/28.
sue tolerance with hypofractionation develops, we will be
16. Dieterich S, Gibbs IC. The CyberKnife in clinical use: current
able to further optimize dose fractionation schedules to max- roles, future expectations. Front Radiat Ther Oncol. 2011;43:
imize the therapeutic ratio. The role of SBRT in combination 181–94. Epub 2011/06/01.
11 CyberKnife 159
17. Antypas C, Pantelis E. Performance evaluation of a CyberKnife G4 whole-brain radiotherapy. Neurosurg Focus. 2009;27(6):E7. Epub
image-guided robotic stereotactic radiosurgery system. Phys Med 2009/12/03.
Biol. 2008;53(17):4697–718. Epub 2008/08/13. 35. Jensen CA, Chan MD, McCoy TP, Bourland JD, deGuzman AF,
18. Chang SD, Main W, Martin DP, Gibbs IC, Heilbrun MP. An analy- Ellis TL, et al. Cavity-directed radiosurgery as adjuvant therapy
sis of the accuracy of the CyberKnife: a robotic frameless stereotac- after resection of a brain metastasis. J Neurosurg. 2011;114(6):
tic radiosurgical system. Neurosurgery. 2003;52(1):140–6; 1585–91. Epub 2010/12/21.
discussion 146–7. Epub 2002/12/21. 36. Choi CY, Chang SD, Gibbs IC, Adler JR, Harsh GR, Lieberson RE,
19. Yu C, Main W, Taylor D, Kuduvalli G, Apuzzo ML, Adler Jr JR. et al. Stereotactic radiosurgery of the postoperative resection cavity
An anthropomorphic phantom study of the accuracy of for brain metastases: prospective evaluation of target margin on
Cyberknife spinal radiosurgery. Neurosurgery. 2004;55(5): tumor control. Int J Radiat Oncol Biol Phys. 2012;84(2):336–42.
1138–49. Epub 2004/10/29. Epub 2012/06/02.
20. Chung H-TK, DG. Modern radiosurgery equipment for treating 37. Villavicencio AT, Burneikiene S, Romanelli P, Fariselli L, McNeely
brain metastases. In: Kim DL, Lunsford LD, editors. Current and L, Lipani JD, et al. Survival following stereotactic radiosurgery
future management of brain metastasis Basel, Switzerland: Karger; for newly diagnosed and recurrent glioblastoma multiforme: a
2012. p. 236–47. multicenter experience. Neurosurg Rev. 2009;32(4):417–24. Epub
21. Muacevic A, Kufeld M, Wowra B, Kreth FW, Tonn JC. Feasibility, 2009/07/28.
safety, and outcome of frameless image-guided robotic radiosur- 38. Stupp R, Hegi ME, Mason WP, van den Bent MJ, Taphoorn MJ,
gery for brain metastases. J Neurooncol. 2010;97(2):267–74. Epub Janzer RC, et al. Effects of radiotherapy with concomitant and adju-
2009/10/06. vant temozolomide versus radiotherapy alone on survival in glio-
22. Takakura T, Mizowaki T, Nakata M, Yano S, Fujimoto T, Miyabe Y, blastoma in a randomised phase III study: 5-year analysis of the
et al. The geometric accuracy of frameless stereotactic radiosurgery EORTC-NCIC trial. Lancet Oncol. 2009;10(5):459–66. Epub
using a 6D robotic couch system. Phys Med Biol. 2010;55(1):1–10. 2009/03/10.
Epub 2009/12/02. 39. Adler Jr JR, Gibbs IC, Puataweepong P, Chang SD. Visual field
23. Mu Z, Fu D, Kuduvalli G. A probabilistic framework based on hid- preservation after multisession cyberknife radiosurgery for periop-
den Markov model for fiducial identification in image-guided radia- tic lesions. Neurosurgery. 2006;59(2):244–54; discussion 254.
tion treatments. IEEE Trans Med Imaging. 2008;27(9):1288–300. Epub 2006/08/03.
Epub 2008/08/30. 40. Mehta VK, Lee QT, Chang SD, Cherney S, Adler Jr JR. Image
24. Wong KH, Dieterich S, Tang J, Cleary K. Quantitative measure- guided stereotactic radiosurgery for lesions in proximity to the
ment of CyberKnife robotic arm steering. Technol Cancer Res anterior visual pathways: a preliminary report. Technol Cancer Res
Treat. 2007;6(6):589–94. Epub 2007/11/13. Treat. 2002;1(3):173–80. Epub 2003/03/08.
25. Pepin EW, Wu H, Zhang Y, Lord B. Correlation and prediction 41. Pham CJ, Chang SD, Gibbs IC, Jones P, Heilbrun MP, Adler Jr
uncertainties in the cyberknife synchrony respiratory tracking sys- JR. Preliminary visual field preservation after staged CyberKnife
tem. Med Phys. 2011;38(7):4036–44. Epub 2011/08/24. radiosurgery for perioptic lesions. Neurosurgery. 2004;54(4):
26. Fu D, Kuduvalli G, Maurer CR, Allision J, Adler J. 3D target local- 799–810; discussion 812. Epub 2004/03/30.
ization using 2D local displacements of skeletal structures in 42. Pannullo SC, Fraser JF, Moliterno J, Cobb W, Stieg PE. Stereotactic
orthogonal X-ray images for image-guided spinal radiosurgery. Int radiosurgery: a meta-analysis of current therapeutic applications in
J CARS. 2006;1(S1):198–200. neuro-oncologic disease. J Neurooncol. 2011;103(1):1–17. Epub
27. Thariat JS, Soltys S.G. Treatment of spinal tumors with Cyberknife 2010/12/15.
stereotactic radiotherapy. In: Hayat MA, editor. Tumors of the Central 43. Henderson F, Jean W, Nasr N, Gagnon G. CyberKnife: clinical
Nervous System, vol. 6, Netherlands: Springer; 2012. p. 165–79. aspects. In: Gildenberg PL, Lozano AM, Tasker RR, editors.
28. Bibault JE, Prevost B, Dansin E, Mirabel X, Lacornerie T, Lartigau Textbook of stereotactic and functional neurosurgery. Berlin:
E. Image-guided robotic stereotactic radiation therapy with fiducial- Springer; 2009. p. 1111–29.
free tumor tracking for lung cancer. Radiat Oncol. 2012;7:102. 44. Andrews DW, Suarez O, Goldman HW, Downes MB, Bednarz G,
Epub 2012/06/26. Corn BW, et al. Stereotactic radiosurgery and fractionated stereotac-
29. Fu DK, R., Wang, B. Xsight lung tracking system: a fiducial-less tic radiotherapy for the treatment of acoustic schwannomas: com-
method for respiratory motion tracking. In: Urschel HCK, J.J., parative observations of 125 patients treated at one institution. Int J
Luketich JD, Papiez L, Timmerman RD, editors. Treating tumors Radiat Oncol Biol Phys. 2001;50(5):1265–78. Epub 2001/08/03.
that move with respiration. Berlin: Springer; 2007. p. 265–82. 45. Chang SD, Gibbs IC, Sakamoto GT, Lee E, Oyelese A, Adler Jr JR.
30. Inoue T, Inoue T, Shiomi H, Shimamoto S. [Clinical experience of Staged stereotactic irradiation for acoustic neuroma. Neurosurgery.
new stereotactic radiotherapy system named CyberKnife]. Nihon 2005;56(6):1254–61; discussion 1261–3. Epub 2005/05/28.
Rinsho. 2001;59(8):1624–31. Epub 2001/08/25. 46. Hansasuta A, Choi CY, Gibbs IC, Soltys SG, Tse VC, Lieberson
31. Chang SD, Lee E, Sakamoto GT, Brown NP, Adler Jr JR. Stereotactic RE, et al. Multisession stereotactic radiosurgery for vestibular
radiosurgery in patients with multiple brain metastases. Neurosurg schwannomas: single-institution experience with 383 cases.
Focus. 2000;9(2):e3. Epub 2006/07/14. Neurosurgery. 2011;69(6):1200–9. Epub 2011/05/12.
32. Nishizaki T, Saito K, Jimi Y, Harada N, Kajiwara K, Nomura S, 47. Romanelli P, Heit G, Chang SD, Martin D, Pham C, Adler
et al. The role of cyberknife radiosurgery/radiotherapy for brain J. Cyberknife radiosurgery for trigeminal neuralgia. Stereotact
metastases of multiple or large-size tumors. Minim Invasive Funct Neurosurg. 2003;81(1–4):105–9. Epub 2004/01/27.
Neurosurg. 2006;49(4):203–9. Epub 2006/10/17. 48. Lim M, Cotrutz C, Romanelli P, Schaal D, Gibbs I, Chang SD, et al.
33. Hara W, Tran P, Li G, Su Z, Puataweepong P, Adler Jr JR, et al. Stereotactic radiosurgery using CT cisternography and non-
Cyberknife for brain metastases of malignant melanoma and renal isocentric planning for the treatment of trigeminal neuralgia.
cell carcinoma. Neurosurgery. 2009;64(2 Suppl):A26–32. Epub Comput Aided Surg. 2006;11(1):11–20. Epub 2006/03/15.
2009/01/30. 49. Villavicencio AT, Lim M, Burneikiene S, Romanelli P, Adler JR,
34. Karlovits BJ, Quigley MR, Karlovits SM, Miller L, Johnson M, McNeely L, et al. Cyberknife radiosurgery for trigeminal neuralgia
Gayou O, et al. Stereotactic radiosurgery boost to the resection bed treatment: a preliminary multicenter experience. Neurosurgery.
for oligometastatic brain disease: challenging the tradition of adjuvant 2008;62(3):647–55; discussion 655. Epub 2008/04/22.
50. Adler Jr JR, Bower R, Gupta G, Lim M, Efron A, Gibbs IC, et al. 68. van der Voort van Zyp NC, Prevost JB, Hoogeman MS, Praag J, van
Nonisocentric radiosurgical rhizotomy for trigeminal neuralgia. der Holt B, Levendag PC, et al. Stereotactic radiotherapy with real-
Neurosurgery. 2009;64(2 Suppl):A84–90. Epub 2009/01/30. time tumor tracking for non-small cell lung cancer: clinical out-
51. Fariselli L, Marras C, De Santis M, Marchetti M, Milanesi I, Broggi come. Radiother Oncol. 2009;91(3):296–300. Epub 2009/03/20.
G. CyberKnife radiosurgery as a first treatment for idiopathic tri- 69. Le QT, Loo BW, Ho A, Cotrutz C, Koong AC, Wakelee H, et al.
geminal neuralgia. Neurosurgery. 2009;64(2 Suppl):A96–101. Results of a phase I dose-escalation study using single-fraction ste-
Epub 2009/01/30. reotactic radiotherapy for lung tumors. J Thorac Oncol. 2006;1(8):
52. Ryu S, Fang Yin F, Rock J, Zhu J, Chu A, Kagan E, et al. Image- 802–9. Epub 2007/04/06.
guided and intensity-modulated radiosurgery for patients with spi- 70. Brown WT, Wu X, Fayad F, Fowler JF, Garcia S, Monterroso MI,
nal metastasis. Cancer. 2003;97(8):2013–8. Epub 2003/04/04. et al. Application of robotic stereotactic radiotherapy to peripheral
53. Gerszten PC, Burton SA, Ozhasoglu C, Welch WC. Radiosurgery stage I non-small cell lung cancer with curative intent. Clin Oncol
for spinal metastases: clinical experience in 500 cases from a single (R Coll Radiol). 2009;21(8):623–31. Epub 2009/08/18.
institution. Spine. 2007;32(2):193–9. Epub 2007/01/17. 71. Willett CG, Czito BG, Bendell JC, Ryan DP. Locally advanced
54. Gerszten PC, Ozhasoglu C, Burton SA, Vogel WJ, Atkins BA, pancreatic cancer. J Clin Oncol. 2005;23(20):4538–44. Epub
Kalnicki S, et al. CyberKnife frameless stereotactic radiosurgery 2005/07/09.
for spinal lesions: clinical experience in 125 cases. Neurosurgery. 72. Koong AC, Le QT, Ho A, Fong B, Fisher G, Cho C, et al. Phase I
2004;55(1):89–98; discussion 99. Epub 2004/06/25. study of stereotactic radiosurgery in patients with locally advanced
55. Martin A, Gaya A. Stereotactic body radiotherapy: a review. Clin pancreatic cancer. Int J Radiat Oncol Biol Phys. 2004;58(4):
Oncol (R Coll Radiol). 2010;22(3):157–72. Epub 2010/01/23. 1017–21. Epub 2004/03/06.
56. Hara W, Soltys SG, Gibbs IC. CyberKnife robotic radiosurgery sys- 73. Chang DT, Schellenberg D, Shen J, Kim J, Goodman KA, Fisher
tem for tumor treatment. Expert Rev Anticancer Ther. GA, et al. Stereotactic radiotherapy for unresectable adenocarcinoma
2007;7(11):1507–15. Epub 2007/11/21. of the pancreas. Cancer. 2009;115(3):665–72. Epub 2009/01/02.
57. Gibbs IC, Kamnerdsupaphon P, Ryu MR, Dodd R, Kiernan M, 74. Mahadevan A, Miksad R, Goldstein M, Sullivan R, Bullock A,
Chang SD, et al. Image-guided robotic radiosurgery for spinal Buchbinder E, et al. Induction gemcitabine and stereotactic
metastases. Radiother Oncol. 2007;82(2):185–90. Epub 2007/01/30. body radiotherapy for locally advanced nonmetastatic pancreas
58. Heron DE, Rajagopalan MS, Stone B, Burton S, Gerszten PC, cancer. Int J Radiat Oncol Biol Phys. 2011;81(4):e615–22. Epub
Dong X, et al. Single-session and multisession CyberKnife 2011/06/11.
radiosurgery for spine metastases-University of Pittsburgh and 75. Lominska CE, Unger K, Nasr NM, Haddad N, Gagnon
Georgetown University experience. J Neurosurg Spine. 2012;17(1): G. Stereotactic body radiation therapy for reirradiation of localized
11–8. Epub 2012/05/15. adenocarcinoma of the pancreas. Radiat Oncol. 2012;7:74. Epub
59. Sibley GS, Jamieson TA, Marks LB, Anscher MS, Prosnitz 2012/05/23.
LR. Radiotherapy alone for medically inoperable stage I non-small- 76. Goyal K, Einstein D, Ibarra RA, Yao M, Kunos C, Ellis R, et al.
cell lung cancer: the Duke experience. Int J Radiat Oncol Biol Phys. Stereotactic body radiation therapy for nonresectable tumors of the
1998;40(1):149–54. Epub 1998/01/09. pancreas. J Surg Res. 2012;174(2):319–25. Epub 2011/09/23.
60. Kaskowitz L, Graham MV, Emami B, Halverson KJ, Rush 77. Didolkar MS, Coleman CW, Brenner MJ, Chu KU, Olexa N,
C. Radiation therapy alone for stage I non-small cell lung cancer. Stanwyck E, et al. Image-guided stereotactic radiosurgery for
Int J Radiat Oncol Biol Phys. 1993;27(3):517–23. Epub 1993/10/20. locally advanced pancreatic adenocarcinoma results of first 85
61. Nagata Y, Takayama K, Matsuo Y, Norihisa Y, Mizowaki T, patients. J Gastrointest Surg. 2010;14(10):1547–59. Epub
Sakamoto T, et al. Clinical outcomes of a phase I/II study of 48 Gy 2010/09/15.
of stereotactic body radiotherapy in 4 fractions for primary lung 78. Seo Y, Kim MS, Yoo S, Cho C, Yang K, Yoo H, et al. Stereotactic
cancer using a stereotactic body frame. Int J Radiat Oncol Biol body radiation therapy boost in locally advanced pancreatic cancer.
Phys. 2005;63(5):1427–31. Epub 2005/09/20. Int J Radiat Oncol Biol Phys. 2009;75(5):1456–61. Epub
62. Timmerman R, Paulus R, Galvin J, Michalski J, Straube W, Bradley 2009/09/29.
J, et al. Stereotactic body radiation therapy for inoperable early 79. Schellenberg D, Goodman KA, Lee F, Chang S, Kuo T, Ford JM,
stage lung cancer. JAMA. 2010;303(11):1070–6. Epub 2010/03/18. et al. Gemcitabine chemotherapy and single-fraction stereotactic
63. Chen VJ, Oermann E, Vahdat S, Rabin J, Suy S, Yu X, et al. body radiotherapy for locally advanced pancreatic cancer. Int J
CyberKnife with tumor tracking: an effective treatment for high- Radiat Oncol Biol Phys. 2008;72(3):678–86. Epub 2008/04/09.
risk surgical patients with stage I non-small cell lung cancer. Front 80. Rwigema JC, Heron DE, Parikh SD, Zeh 3rd HJ, Moser JA, Bahary
Oncol. 2012;2:9. Epub 2012/06/02. N, et al. Adjuvant stereotactic body radiotherapy for resected
64. Unger K, Ju A, Oermann E, Suy S, Yu X, Vahdat S, et al. CyberKnife pancreatic adenocarcinoma with close or positive margins.
for hilar lung tumors: report of clinical response and toxicity. J Gastrointest Cancer. 2012;43(1):70–6. Epub 2010/09/03.
J Hematol Oncol. 2010;3:39. Epub 2010/10/26. 81. Polistina F, Costantin G, Casamassima F, Francescon P, Guglielmi
65. Vahdat S, Oermann EK, Collins SP, Yu X, Abedalthagafi M, R, Panizzoni G, et al. Unresectable locally advanced pancreatic
Debrito P, et al. CyberKnife radiosurgery for inoperable stage IA cancer: a multimodal treatment using neoadjuvant chemoradiother-
non-small cell lung cancer: 18F-fluorodeoxyglucose positron emis- apy (gemcitabine plus stereotactic radiosurgery) and subsequent
sion tomography/computed tomography serial tumor response surgical exploration. Ann Surg Oncol. 2010;17(8):2092–101. Epub
assessment. J Hematol Oncol. 2010;3:6. Epub 2010/02/06. 2010/03/13.
66. Pennathur A, Luketich JD, Heron DE, Schuchert MJ, Burton S, 82. Chang DT, Swaminath A, Kozak M, Weintraub J, Koong AC,
Abbas G, et al. Stereotactic radiosurgery for the treatment of lung Kim J, et al. Stereotactic body radiotherapy for colorectal liver
neoplasm: experience in 100 consecutive patients. Ann Thorac metastases: a pooled analysis. Cancer. 2011;117(17):4060–9. Epub
Surg. 2009;88(5):1594–600; discussion 1600. Epub 2009/10/27. 2011/03/25.
67. van der Voort van Zyp NC, van der Holt B, van Klaveren RJ, 83. Lee MT, Kim JJ, Dinniwell R, Brierley J, Lockwood G, Wong R,
Pattynama P, Maat A, Nuyttens JJ. Stereotactic body radiotherapy et al. Phase I study of individualized stereotactic body radiotherapy
using real-time tumor tracking in octogenarians with non-small cell of liver metastases. J Clin Oncol. 2009;27(10):1585–91. Epub
lung cancer. Lung Cancer. 2010;69(3):296–301. Epub 2010/01/12. 2009/03/04.
11 CyberKnife 161
84. Tse RV, Hawkins M, Lockwood G, Kim JJ, Cummings B, Knox J, 88. Lanciano R, Lamond J, Yang J, Feng J, Arrigo S, Good M, et al.
et al. Phase I study of individualized stereotactic body radiotherapy Stereotactic body radiation therapy for patients with heavily pretreated
for hepatocellular carcinoma and intrahepatic cholangiocarcinoma. liver metastases and liver tumors. Front Oncol. 2012;2:23. Epub
J Clin Oncol. 2008;26(4):657–64. Epub 2008/01/04. 2012/05/31.
85. Rusthoven KE, Kavanagh BD, Cardenes H, Stieber VW, Burri SH, 89. Goyal K, Einstein D, Yao M, Kunos C, Barton F, Singh D, et al.
Feigenberg SJ, et al. Multi-institutional phase I/II trial of stereotac- Cyberknife stereotactic body radiation therapy for nonresectable
tic body radiation therapy for liver metastases. J Clin Oncol. tumors of the liver: preliminary results. HPB Surg. 2010;2010.
2009;27(10):1572–8. Epub 2009/03/04. Epub 2010/08/07.
86. Andolino DL, Johnson CS, Maluccio M, Kwo P, Tector AJ, Zook J, 90. Choi BO, Choi IB, Jang HS, Kang YN, Jang JS, Bae SH, et al.
et al. Stereotactic body radiotherapy for primary hepatocellular car- Stereotactic body radiation therapy with or without transarterial che-
cinoma. Int J Radiat Oncol Biol Phys. 2011;81(4):e447–53. Epub moembolization for patients with primary hepatocellular carcinoma:
2011/06/08. preliminary analysis. BMC Cancer. 2008;8:351. Epub 2008/11/29.
87. Goodman KA, Wiegner EA, Maturen KE, Zhang Z, Mo Q, Yang G, 91. O’Connor JK, Trotter J, Davis GL, Dempster J, Klintmalm GB,
et al. Dose-escalation study of single-fraction stereotactic body Goldstein RM. Long-term outcomes of stereotactic body radiation
radiotherapy for liver malignancies. Int J Radiat Oncol Biol Phys. therapy in the treatment of hepatocellular cancer as a bridge to trans-
2010;78(2):486–93. Epub 2010/03/31. plantation. Liver Transpl. 2012;18(8):949–54. Epub 2012/04/03.
GammaPod
12
Yildirim D. Mutaf, Cedric Yu, Steven J. Feigenberg,
and William F. Regine
required to achieve clear surgical margins. Unfortunately, MR

Background imaging has not been shown to improve outcomes and may
actually increase the rate of mastectomy [5–7]. This is likely
Breast cancer is the most common malignancy among women because of the poor accuracy between breast imaging
in the USA, with more than 200,000 new cases per year [1]. (performed prone) and surgery (performed supine).
Awareness and regular screening have resulted in more Four to six weeks after surgical resection, if re-excision is
patients being diagnosed with early stage disease, for which not needed, radiation therapy is administered. In the USA, a
breast conservation therapy (BCT) is a standard treatment standard course of radiation encompasses the whole breast
option. Compared with mastectomy, BCT provides similar and requires an additional 5–7 weeks, with treatment deliv-
outcomes with superior cosmesis and reduced psychological ered daily, Monday–Friday. The overall duration of local
and emotional trauma [2, 3]. However, local treatments (sur- therapy/BCT (i.e., lumpectomy + radiation therapy) is a min-
gery and radiotherapy [RT]) have not taken advantage of sig- imum of 11 week. As many as 35 % of elderly patients who
nificant advances in imaging. BCT, as it is practiced today, live a long distance from a treatment facility decide against
has two key shortcomings: (1) local therapy lasts for at least RT after BC surgery—decisions that contribute directly to
11 weeks, is expensive, and significantly affects quality of decreases in overall survival [8].
life and (2) lack of precise image guidance results in poor
accuracy, which can negatively impact cure and cosmesis.
Breast-conserving surgery consists of lumpectomy, in Hypofractionated Whole Breast Irradiation
which the tumor plus a cuff of normal breast is resected. and Accelerated Partial Breast Irradiation
Because most diagnosed cancers are clinically occult, sur-
geons rely on image guidance for tumor localization. Using In order to reduce treatment costs and increase utility of
mammographic guidance, a needle is placed by the radiolo- radiotherapy, hypofractionated whole breast irradiation
gist into or through the tumor to guide the surgeon “blindly” (HWBI) and accelerated partial breast irradiation (APBI)
to the tumor. Because of the poor geometric link between were developed. HWBI decreases the duration of whole
imaging and surgery, the lumpectomy specimen is many breast radiation therapy to 3–4 weeks. However, patients are
times larger than the index lesion (Fig. 12.1). The large vol- still faced with quality-of-life difficulties associated with
ume loss is the primary reason for poor cosmetic outcomes. multiple repeat visits over a long time period. APBI is radi-
Compounding this poor geometric link, >20 % of patients cally different in that only the surgical bed, rather than the
have inadequate margins and require re-excision or mastec- whole breast, is targeted for irradiation. Treating smaller vol-
tomy to obtain optimal local control [4, 5]. Magnetic reso- umes allows safe delivery of higher doses of radiation per
nance (MR) imaging defines the breast disease more accurately treatment, obtaining the same biologic effectiveness in only
than mammography, theoretically reducing the excess normal 1–5 days [9–13]. Early results of several large prospective
breast removed at surgery and the number of operations phase III trials comparing APBI with whole breast irradia-
tion demonstrated similar outcomes, supporting the hypoth-
esis that for select patients [9, 14, 15], whole breast irradiation
Y.D. Mutaf, Ph.D. (*) • C. Yu, D.Sc. • S.J. Feigenberg, M.D. may not be necessary. In NSABP B-39, the largest phase III
W.F. Regine, M.D.
Department of Radiation Oncology, University of Maryland
trial comparing whole breast radiation and APBI, APBI
School of Medicine, Baltimore, MD, USA can be delivered using invasive approaches via various
e-mail: ymutaf@umm.edu brachytherapy applications or 3D conformal RT (3D-CRT).
164 Y.D. Mutaf et al.
brain metastases by improving local control, quality of life,

and survival [21]. Its success has hinged on minimizing the
volume of normal tissue receiving a “significant” dose of
radiation while maximizing dose to the tumor.
SRS outside the brain, called stereotactic body RT
(SBRT), has been more challenging [22]. Unlike intracranial
targets that do not move in reference to a rigid stereotactic
head frame, targets located outside the brain cannot be per-
fectly immobilized and the organ/tumor motion requires the
physician to add a margin to ensure appropriate dose cover-
age of the target volume. The planning target volume (PTV)
is defined by an additional margin to account for tumor
motion and setup uncertainties. This added margin increases
the volume of tissue irradiated, which theoretically increases
the risk of side effects [23]. Over the last 15 years, solutions
have allowed SBRT to treat tumors outside the brain [24, 25].
In fact, in lung cancer, the results [26, 27] have been so
impressive that investigators are currently challenging the
paradigm of surgery as the standard of care for patients with
small lung cancers [28].
SBRT of breast cancer has been addressed in only two
published reports [29, 30]. These studies have severe limita-
tions because of poor immobilization and lack of adequate
image guidance, allowing only subtherapeutic doses of radi-
ation and possible marginal misses because of set up uncer-
Fig. 12.1 (a) Preoperative CT scan of 1 cm breast cancer. Lumpectomy tainty. One of the major challenges of SBRT for breast cancer
demonstrated a 1 × 0.7 × 0.5 cm invasive cancer with 7 × 7 × 2 cm lumpec-
tomy specimen with margin of 2 mm. (b) Postoperative CT scan of sur-
is that the breast is pliable. These challenges have been
gical bed, which was 40 cm3, 100-fold greater than 0.4 cm3 shown in (a) solved in the brain, thorax, and abdomen with commercially
available immobilization devices. In addition in the breast,
surrounding structures have placed limitations on beam
On this study, more than 70 % of patients randomized to delivery. These technical obstacles have been solved through
APBI are treated using 3D-CRT as compared with brachy- an NIH-SBIR grant mechanism (4R44CA132254-02) at the
therapy, even though 3D-CRT treats a volume approximately University of Maryland by Yu et al. [31] with the inception
5× larger than the brachytherapy techniques [16]. This larger of GammaPod™ technology, commercialized by the Xcision
volume irradiated leads to a concern that 3D-CRT could Medical System, LLC (Xcision). A detailed description of
result in a higher rate of subcutaneous fibrosis [17] and unac- the GammaPod™ device and underlying technology will be
ceptable cosmesis [18, 19], as volume is one of the most provided in subsequent sections.
important predictors of morbidity [20]. Yet many patients The GammaPod™ device has several advantages over
choose to receive 3D-CRT instead of brachytherapy because current technologies capable of SBRT, including but not lim-
of the distinct advantage of being completely noninvasive. ited to (1) the vacuum assisted breast immobilization device
has improved accuracy and reproducibility. In addition, the
patient is treated in the prone position, similar to positioning
Role of Technologic Advances in Delivery of RT for stereotactic breast biopsy or MR imaging of the breast,
thus minimizing the effect of breathing motion. Subsequently,
Over the last 30 years, advances in immobilization, computer- the margin required for PTV can be decreased from 1.0 to
based treatment planning and delivery, and image guidance 0.3 cm, potentially reducing the volume radiated by three-
have revolutionized the ways in which small tumors are fold [32]. (2) This device uses 36 noncoplanar Co-60 sources
treated with RT. Stereotactic radiosurgery (SRS) was intro- that rotate during treatment to form 36 noncoplanar arcs,
duced in 1968 as a tool for noninvasive neurosurgery but producing a dose distribution similar to that of the
was not widely adopted until 3D imaging became available. GammaKnife™ for intracranial targets [33]. (3) This system
The pinpoint accuracy and high dose concentration achiev- allows differential dosing of target volumes [34] (dose paint-
able with SRS allow safe delivery of “ablative” doses of ing; similar to that in intensity modulated RT techniques),
radiation. SRS has been a success story for patients with which cannot be accomplished with any invasive APBI
12 GammaPod 165
approaches. This allows higher doses to the gross tumor with and implemented clinically in the future, select patients will
lower “subclinical doses” to ablate microscopic foci, which receive SBRT alone. The geometric accuracy will likely
can lie 1.0–1.5 cm beyond the index lesion [35, 36]. improve cosmesis by causing minimal-to-no volume loss,
GammaPod™ offers the opportunity to deliver curative eliminate surgical scarring, improve quality of life issues
doses of radiation using SBRT more precisely than current currently associated with prolonged treatment periods, and
surgical standards, thereby minimizing the amount of breast enhance overall outcomes.
tissue ablated and improving cosmesis. If this treatment is as
effective as it has proven for lung cancer, SBRT could lead to
a paradigm shift similar to that seen in lung cancer, with GammaPod™ Technology
completely noninvasive treatment of early breast cancer with
only a few high dose “ablative” treatments delivered over the The GammaPod™ device comprises three major components:
course of less than 1 week. the gamma irradiation unit housing the Cobalt-60 radioactive
sources and collimation structure, a breast immobilization unit
to facilitate stereotactic localization and prone patient support
Role of Preoperative RT in Early Stage and a continuous trajectory patient positioning assembly.
Breast Cancer
In preparation for developing a clinical trial using SBRT Irradiation Device

with ablative doses of RT, we investigated the role of preop-
erative 3D-CRT APBI as a method to reduce morbidity and The GammaPod™ irradiation unit comprises of two mechan-
allow greater utility of APBI. We have demonstrated that ical structures: a source carrier and a collimation unit as
using a preoperative target volume dramatically decreases shown in Fig. 12.2.
the PTV by a mean average value of 56 % when compared The hemispherical housing structure contains the radionu-
with the standard postoperative target [32]. In a subsequent clide source in the form of 36 Cobalt-60 pellets. These pellets
publication, we reported that preoperative 3D-CRT APBI are positioned equidistantly, 38 cm from the isocenter (source
showed significant clinical advantages by decreasing dose to to focal distance, SFD). With an SFD of 38 cm, achieving an
all surrounding normal tissues and substantially increasing initial dose rate greater than 5 Gy/min inside a large breast
(by 60 %) the number of patients eligible for APBI using requires an estimated total Cobalt-60 activity in the range of
3D-CRT [37]. Based on these dosimetric advantages, we are 4,000–5,000 Curies. The Cobalt-60 source is double-encap-
currently testing this hypothesis on a prospective trial at the sulated with the inner capsule packed tightly with smaller
University of Maryland. This trial uses the same treatment Cobalt-60 pellets. The source shell dimension is customized
guidelines (i.e., clinical treatment volume and PTV expan- to hold the required number of radioactive pellets. The cylin-
sions) and dose fractionation (38.5 Gy delivered in ten equal drically shaped inner capsule of the source shell measures
fractions over 5 days) as NSABP B-39, with the only differ- approximately 11 mm in height and 3.4 mm in diameter,
ence being the sequence of surgery and RT. Based on early reflecting the true radiation source size of a single source.
data from the first 25 patients treated on study, a complete Sources are aligned in six columns across the 2π surface
pathologic response was achieved. of the source carrier, with 60° longitudinal separation. Each
These pathologic findings have been quite exciting for source column is arranged in a spiral order, such that a source
two reasons. First, only a low “adjuvant” dose of radiation is positioned at every 10° longitudinally. The latitude of each
was administered, and, second and more important, a column is also designed such that all 36 sources span an
response was actually seen. These tumors are slow growing angular section between 18 and 53°, i.e., with 1° of latitudi-
and can take a long time to respond. Prostate cancer, which nal separation. This arrangement ensures that each of the 36
has similar growth patterns, takes years to show histologic sources occupies a unique longitude and latitude coordinate
resolution of tumor following radiotherapy. Therefore, we when the source carrier is rotated.
believe that small, early stage breast cancer, SBRT may Sources are collimated by the primary collimator, located
replace lumpectomy and postoperative RT. With the less at the exiting end of the source shell. The secondary collima-
aggressive, slower growing tumors in the breast, it is likely tor, with the same hemispherical shape, is nestled within the
that SBRT will approach or surpass the efficacy of 90 %+ primary collimator. Collimation is performed by pre-drilled
local control rates seen in SBRT for lung cancer. Although circular holes aligned with the sources. By altering the align-
clearly many studies will be necessary to prove this hypoth- ment of the source carrier and collimator, three different col-
esis, in order for SBRT to be realized as a standalone treat- limation modes can be chosen (two open, one blocked). The
ment, a noninvasive measure of success is key which will GammaPod™ is designed to have two collimator opening
require functional imaging such as MRI or PET. If successful sizes, 1.5 and 2.5 cm, in order to increase the dose shaping
Fig. 12.2 An artist rendering (left) and 3D model drawing (right) of GammaPod™ depicting prone placement of a patient. (Courtesy of Xcision
Medical Systems.)
Fig. 12.3 A schematic drawing for the two-layered breast immobilization cup is shown on left panel with embedded stereotactic frame imaging
fiducial wire. A photo of the prototype breast cup is also shown on right. (Courtesy of Xcision Medical Systems.)
capabilities of the device. In contrast to radiosurgery of intra- The shielding body and doors are designed to bring leakage
cranial lesions, in breast cancer larger collimator sizes are radiation levels below those set by the regulatory bodies
selected due to the typical tumor sizes and the need to steril- when the device is not in use. The irradiation unit and shield-
ize the tissues surrounding the tumor owing to its multicen- ing doors are equipped with an uninterrupted power supply
tricity nature. (UPS) unit to ensure that treatments can be terminated safely
When the selected collimator opening is aligned with the in the event of facility power failures. The source carrier and
source positions, the source carrier and the collimator are the shielding doors can also be controlled manually using a
rotated together to create 36 noncoplanar conical arcs all hand crank should the other safety measures fail.
converging at the isocenter. In addition, the 36 source block-
ing positions on the primary collimator are made of tungsten
rods, effectively reducing the radiation to acceptable leakage Breast Immobilization Apparatus
levels at the beginning and end phases of the fraction.
The GammaPod™ irradiation unit is encased in a hemi- The GammaPod™ system utilizes a two-layered breast cup
spherical shielding body made of cast iron. There are two in order to immobilize the breast (Fig. 12.3). Several inner
shielding doors on top of the irradiation unit which opens cup sizes accommodate patients with varying breast sizes.
during the irradiation and stays closed at other times. The outer layer is composed of a solid plastic material and
12 GammaPod 167
connected to a vacuum system working at a comfortable left or right breast while the contralateral hole is covered
level of negative air pressure. While the patient is in the with a lid. Once docking is accomplished, the patient is low-
prone position, the treated breast is positioned inside the ered to the treatment or the imaging unit (CT or MR device)
inner cup of the breast cup. The inner cup is perforated so via a cantilever action of the couch actuators. This automatic
that when a vacuum is applied to the space in between the patient positioning mechanism reduces the amount of motion
outer and inner cups, the breast volume fills the entire con- and effort required by the patient, which could otherwise
tour of the inner cup and the coordinate frame is transferred compromise breast immobilization.
to the patient’s anatomy. The inner and outer cups are tightly The treatment couch unit is also equipped with three-
coupled with a silicone flange, which is attached to the dimensional motor controls which move the patient in rela-
patient’s chest wall with a thin layer of silicone restricting tion to the radiation isocenter of the GammaPod™. This 3D
breast mobility with respect to the inner cup. In preliminary motion trajectory also enables continuous shaping of the
studies [38], the GammaPod™ immobilization system was dose distribution across the target volume. Therefore, instead
shown to reduce the setup uncertainty to 0.2 ± 0.1 (2σ) cm, of packing multiple spherical and static dose distributions,
and therefore the amount of margin necessary for PTV expan- generally referred to as shots, which result in large dose het-
sion may conservatively be set at 0.3 cm for radiosurgery erogeneities, the GammaPod™ attains a more uniform dose
applications. distribution. This dose shaping allows for the treatment of
The breast cup immobilization system also has the benefit different target volumes at different dose levels (such as pri-
of gently pulling of the breast volume away from the patient’s mary gross tumor volume (GTV) versus the surrounding rind
chest wall, therefore providing larger separation between the of subclinical disease extent receiving higher and lower dose
targeted area and the ribs, lung and the heart in the case of levels, respectively).
left breast treatments. The increased separation is an added The GammaPod™ treatment couch is also equipped with
advantage to the similar geometric advantages caused by the redundant safety features to prevent any collision of the
prone positioning of the patient’s breast especially for pen- device components with the patient. One of the safety fea-
dulous breasts. To further increase comfort and immobiliza- tures employs metal ports mounted on the shielding door
tion, a wide strap is affixed to the base of the cup, which which prevent the lowering of the treatment couch when the
wraps around the chest of the patient. shields are closed. The treatment couch also acts as a physi-
The outer cup also incorporates a stereotactic localization cal block, preventing the shielding doors from closing while
frame with the help of a spirally shaped CT and the patient is lowered into the treatment position. Knowing
MR-compatible fiducial as shown in Fig. 12.3. The frame the outer breast cup dimensions also allows for the imposi-
consists of a helical hollow plastic tube wrapped diagonally tion of trajectory limits for treatment plan optimization and
around the outer cup. The plastic tube is filled with either thus prevents inadvertent collisions with the inner surface of
MRI contrast solution or copper wires, and is solidly secured the hemispherical collimator structure.
to the machined groove on the outer cup. Once the fiducials
have been identified a shared coordinate frame system
between the immobilization device and the patient’s anat- GammaPod™ Treatment Planning System
omy is created.
The GammaPod™ stereotactic device is distributed with a
dedicated treatment planning system, called Ceres. In its
Continuous Trajectory Patient Positioning current implementation, Ceres provides capabilities for mul-
System timodality image import, stereotactic registration and local-
ization, manual contouring and automatic detection of the
The couch and the patient positioning system are critical breast/body surfaces, inverse optimization, dose calculation
components of GammaPod™ stereotactic radiotherapy and evaluation, treatment export and various quality assur-
device. The couch top has built-in pegs that are machined to ance (QA) utilities.
fit perfectly with the breast immobilization device, therefore After the breast CT/MRI images are imported, breast cup
creating a rigid registration system. Using this registration fiducials are identified in a single cross-sectional image,
protocol, the shared coordinates between the immobilization which initiates automatic detection of the entire spiral.
device and the patient anatomy become a global coordinate Refinements to the detected fiducial geometry are performed
frame between the treatment unit and the patient’s anatomy. to match the known helical geometry of the stereotactic
Two couch systems exist for GammaPod™ treatments: frame. After the registration of the stereotactic coordinate
The imaging couch top and the treatment couch. In both sys- frame of reference to the patient image set, the software
tems, the couch-tops are brought to an upright position using detects the actual breast contour within the breast cup and
a pivoted motor assembly that enables patient docking. Both creates the whole breast volume. At this point other organs
couch-tops consist of two holes to accommodate the treated of interest, such as chest wall and ribs, lungs, and/or heart
volume, can be manually drawn. The target volume creation files appearing as dips. The distances between the mechanical
starts with the delineation of the primary GTV and followed and radiation isocenters of the distributions were measured
by the generation of other concentric targets using margin as 0.2, 0.1, and 0.4 mm in the x, y, and z dimensions, respec-
expansions from the GTV, e.g., CTV. During the margin tively. This submillimeter level of isocenter localization
expansion process, the automatically generated contours are accuracy is deemed suitable for SRS units. Two-dimensional
restricted from expanding beyond the 0.5 cm distance to the isodose overlays also demonstrate the congruence of
breast skin surface. expected dose distributions, as shown by the obtained dosi-
After contouring is completed, the input of prescription metric measurements with radiochromic film in Fig. 12.6.
and other dosimetric parameters, such as critical organ dose
constraints, are performed on the inverse optimization
engine, where the continuous trajectory of the radiation iso- Absolute Dosimetry
center within the breast is optimized for two available colli-
mator sizes, 1.5 and 2.5 cm. Multiple targets and different Absolute dosimetry measurements with the GammaPod™
dose levels can be accommodated in the treatment planning unit can be performed with the vendor-supplied breast phan-
system. The planner also has the option to use treatment time tom, which is made of ultra-high molecular weight polyeth-
as an optimization constraint. After optimization, the final ylene (UHMW-PE) material with a physical density of
plan can be normalized (as a global scale) to achieve the 0.935 g/cm3. Because of the noncompliant geometry of the
desired dose–volume levels. treatment unit and measurement phantom, the contemporary
In Ceres, the dose calculation algorithm utilizes previously AAPM TG-51 photon calibration standard cannot be used
calculated dose kernels representing both the inner and outer for GammaPod™’s absolute dosimetry. Instead, the former
cup dimensions as well as multiple isocenter positions. Dose AAPM TG-21 photon beam calibration standard can be used
kernels for each collimator and isocenter positions are for dosimetric calibration.
calculated using the Monte Carlo (MC) simulation. The In a preclinical prototype unit installed at University of
GammaPod™ Co-60 source and collimation structures Maryland, the GammaPod™ device was loaded with 36
are modeled in detail using BEAMnrc code based on the Co-60 sources with an average activity of 120 Ci each for a
EGS4nrc [39] dedicated for radiation source simulation. total activity of 4,320 Ci. The absorbed dose in the polyeth-
The outcome phase space data modeling the energy deposi- ylene phantom at the isocenter was determined to be 5.61 Gy/
tion profile is used as the input for final dose kernel generation min for the static 2.5 cm collimator shot. Using the ratio of
using the DOSXYZnrc code [40]. An interpolation between the mass energy absorption coefficients, dose to water is
these dose kernels is also performed to describe the resultant determined to be 5.45 Gy/min at depth in the polyethylene
dose distribution due to continuous beam trajectory. phantom. This dose rate calibration was also cross checked
In the current version of Ceres, tissue heterogeneities of with OSLDs and determined to be 5.42 ± 0.09 (1σ) Gy/min,
the patient are not taken into account and the whole breast demonstrating a very good agreement between the two inde-
volume is regarded as having a uniform mass density of pendent measurements. The absorbed dose in water at a
0.935 g/cm3. water-equivalent depth of 7.5 cm at isocenter was also deter-
mined using a scale factor, as described in TG-21, to correct
for attenuation differences in polyethylene versus water
GammaPod™ Dosimetry using the ratio of linear attenuation coefficients (0.973).
Finally, given a total source activity of 4,320 Ci, the absorbed
Single collimator static dose profiles used as primary dose dose rate in water was established as 5.31 Gy/min for the
kernels in treatment planning systems for a medium-sized 2.5 cm collimator.
cup is shown in Fig. 12.4 for 1.5 and 2.5 cm collimators in
coronal and longitudinal cuts. In a preclinical prototype unit,
two-dimensional dose kernels are compared to dosimetric Treatment Plan Examples
measurements as performed with radio chromic films (only
for 2.5 cm collimation). Prior to making comparisons, the GammaPod™ can treat breast targets having a wide range of
radiation isocenter of the GammaPod™ was verified with shapes and sizes. In fact, the GammaPod™ device is capable
respect to the mechanical isocenter of the combined couch of treating target volumes up to 8 cm in its largest dimension
and phantom system. For isocenter verification, a small hole although since complications will be related to the surface
was punched on the film sheets marking the mechanical iso- receiving the same dose, the dose per fraction may be drasti-
center. As shown in Fig. 12.5, these holes, appearing as cally different between small and large targets. GammaPod™
peaks in the measured dose profiles, fall at the center of these is therefore technically capable of target volumes in breast
distributions, with the overlaid expected center of the pro- that are common both in pre- and post-lumpectomy settings.
12 GammaPod 169
Fig. 12.4 Dose kernels used for a static 1.5 and 2.5 cm shots (E-size breast cup) calculated using Monte Carlo. Apex preferential beam direction
is apparent in the low dose isodose lines with a characteristic cross shape in the YZ plane
Post-operative Treatment Planning Example 1.5 cm isotropic margins to include subclinical extension of the
disease and create a CTV (101.5 cm3). Next, a PTV contour was
The lumpectomy cavity (LC) of 4.1 cm3 on a patient’s left generated using a 1.0 cm expansion to account for patient setup
breast is identified as the target volume in the post-op CT uncertainties (256.4 cm3). In linac-based protocols, although
images of a prone patient with the help of surgical clips. In this PTV is used for beam aperture generation, it is trimmed
order to facilitate comparative analysis with conventional treat- from the breast skin and chest wall by 0.5 cm to establish dosi-
ment techniques (linac-based external beam, brachytherapy), metric goals and plan evaluation (PTV_EVAL). These volumes
we followed the target delineation methodology commonly are used to create three treatment plans for comparison: IMRT
prescribed in many accelerated partial breast protocols such as and 3D-CRT plans for linac-based external beam radiotherapy
NSABP B-39/RTOG 0413. Therefore, LC was expanded by (shown in Fig. 12.7a, b) and a GammaPod™ plan (Fig. 12.7c).
Fig. 12.5 One-dimensional dose profiles for 2.5 cm shot in cardinal directions and congruence of mechanical and radiation isocenters demonstrated
with a matching dip and peak pair in each direction
Fig. 12.6 Two-dimensional dose profile for 2.5 cm shot in XY direction is shown as measured with radiochromic film (left), calculated dose
distribution (middle), and a 2D overlay of measured and calculated isodose lines (right)
Fig. 12.7 Linac-based 3D-CRT

(a) and IMRT (b) plans for a
postoperative APBI treatment.
PTV_Eval is denoted with a
green contour. GammaPod™
plans for 1.0 cm (c) and 0.3 cm
(d) setup margin expansions for
the creation of planning target
volumes. The case with 1.0 cm
setup margin (upper margin) is
identical to the linac-based
treatments in the way PTV
structures are generated. PTV_
Eval volume is denoted by green
contour with white dots
12 GammaPod 171
Fig. 12.7 (continued)

Fig. 12.8 Dose–volume histogram comparisons of linac-based 3D-CRT, IMRT, and GammaPod™ plans for PTV_Eval, ipsilateral normal breast
tissue, and lung and heart volumes corresponding to the postoperative scenario
Although the same volumes are used for both linac-based istics, with GammaPod™ plans demonstrating a clear advan-
and GammaPod™ plans for one-to-one comparisons, the tage in sparing ipsilateral normal breast as compared to
requirement of a 1.0 cm expansion for PTV is clearly an linac-based treatment plans. Among the GammaPod™ plans,
overestimation for GammaPod™ treatments. This expansion the advantage of reducing the PTV margin from 1.0 to 0.3 cm
is prescribed mainly because of the setup uncertainties stem- is apparent as the ipsilateral breast volume receiving 50 % of
ming from APBI experience in supine setup linac-based the prescription dose drops by about 10 %. In terms of other
radiotherapy. The GammaPod™’s dedicated breast immobi- nearby critical organs, DVHs for heart and ipsilateral lung
lization device, stereotactic frame, vacuum suctioning of the are also displayed in Fig. 12.8. For the heart, all plans
breast, prone setup, and same day imaging and treatment resulted in low heart doses with the GammaPod™ 1 cm PTV
characteristics are believed to reduce treatment setup uncer- expansion plan slightly higher over the linac-based plans.
tainties to levels more common in stereotactic applications. Although the comparison is statistically underpowered due
As mentioned earlier, the GammaPod™ immobilization to single patient presentation, this could be expected from
system has indeed been shown to reduce the setup uncer- the beam topology associated with covering the larger
tainty to 0.2 ± 0.1 (2σ) cm, making it possible to reduce the PTV. On the other hand, GammaPod™ plans were effective
PTV expansion from 1.0 to 0.3 cm. In treatment plan com- in obtaining better lung DVH values (none of the plans were
parisons, we therefore included two GammaPod™ plans, corrected for tissue heterogeneities).
one with 1.0 cm PTV expansion and another with 0.3 cm
(Fig. 12.7d), all other margins being equal. The volumes
of two PTV structures were 256.4 and 145.8 cm3, a PTV Preoperative Treatment Planning Example
reduction of 57 %.
As shown in Fig. 12.7, both linac-based treatment plans On the same CT scan, a hypothetical preoperative plan was
yield larger irradiation of the ipsilateral breast volume to pre- created. Although the anatomy depicted in the images are
scription as well as lower level doses when compared to post-lumpectomy, we created a GTV volume corresponding
GammaPod™ plans. The dose–volume histogram compari- to a hypothetical spherical lesion located at the center of the
sons for ipsilateral normal breast volume and target volume cavity, with a volume matching the lesion volume reported in
are also provided in Fig. 12.8. All plans demonstrate similar other imaging studies (2.6 cm3). This volume was expanded
target dose coverage as well as dose heterogeneity character- with only setup margins used to create the final PTV contours
12 GammaPod 173
Fig. 12.9 GammaPod™ plan

isodose dose distribution
(upper panel) and dose volume
histograms (lower panel) in the
preoperative target scenario
discussed in the text. GTV is
displayed with the yellow
contour and white dots depicting
the pre-operative lesion, while
PTV is denoted with the green
contour and white dots
(0.3 cm margin expansion). The intermediate CTV volume disease. Therefore, two target volumes were created; a PTV_
and the treatment of subclinical disease extension will be high volume, receiving 100 % of the prescription dose, and a
discussed below as a separate scenario. The isodose distribu- PTV_low volume, receiving 50 % of the prescription dose
tions from these plans and corresponding DVHs are demon- (half-dose is chosen for demonstration purposes only). The
strated in Fig. 12.9. For such small targets, the dose outside PTV_high volume was the same volume as the previous pre-
the target volume falls to approximately 50 % of the target op scenario where GTV was expanded only with the setup
dose in the first centimeter and further drops to 10–15 % in margin (PTV_high = GTV + 0.3 cm). However, PTV_low
the second centimeter of tissue. volume corresponds to the GTV expanded by 1.5 cm to cre-
ate a CTV which is subsequently expanded by 0.3 cm setup
margins (PTV_low = GTV + 1.8 cm). Therefore, the two
Dose Painting with Pre-Op Target PTV volumes are concentric and the difference between the
and Subclinical Disease Extension two corresponds to the 1.5 cm CTV expansion. The isodose
distributions from these plans and corresponding DVHs are
In the last scenario, we evaluated the previous pre-op GTV demonstrated in Fig. 12.10. The DVH for PTV_low corre-
target within a dose-painting strategy in conjunction with a sponds only to the shell of 1.5 cm rind between the PTV_
PTV expansion of 1.5 cm for the subclinical extension of the high and PTV_low volumes.
Fig. 12.10 GammaPod™ plan

isodose dose distribution
(upper panel) and dose volume
histograms (lower panel) in the
dose painting scenario of high
and low dose levels discussed in
the text. GTV is displayed with
the yellow contour and white dots
depicting the pre-operative
lesion, while PTV_high is
denoted with the green contour
and white dots. The largest PTV
volume, PTV_low, is drawn with
the blue contour and white dots
angles surrounding the breast and can therefore shape the

Conclusions dose with better conformity around the target volume. The
GammaPod™ device also generates sharper dose profiles
The application of stereotactic principles to breast radio- due to small collimator openings and a short axis-to-
therapy has promising advantages over conventional linac- collimation distance, all of which contribute to the apparent
based applications. In Figs. 12.7 through 12.10, the dose dose conformity of GammaPod™ dosimetry.
distributions for treating the post- and preoperative target In addition to the advantages gained by its beam geometry,
volumes are demonstrated. The GammaPod™ is observed the GammaPod™ system also introduces a negative-pressure
to achieve good dose conformity. Additionally, linac-based breast immobilization cup with a built-in stereotactic frame,
external radiotherapy techniques are usually limited by the as well as a highly dynamic patient positioning system. The
degree of freedom available for the selection of beam angles combined effect of these dedicated breast stereotactic tech-
and are therefore typically restricted to tangentially oriented nologies is to reduce setup uncertainties with respect to the
angles. This beam topology results in undesirable dose spill- linac-based radiotherapy techniques. For a hypothetical clini-
age outside the target volume and poor dose conformity. cal target volume of 1.0 cm in radius, reducing the expansion
With the utilization of a breast-dedicated source/collimator margin from 1.0 to 0.3 cm, as previously discussed, brings
design, the GammaPod™ utilizes a larger range of beam about a volumetric reduction of nearly fourfold (34 vs. 9 cm3).
12 GammaPod 175
In summary, we present that in the application of stereo- 9. Vaidya JS, Joseph DJ, Tobias JS, Bulsara M, Wenz F, et al. Targeted
tactic concepts, GammaPod™ technology has the potential intraoperative radiotherapy versus whole breast radiotherapy for
breast cancer (TARGIT-A trial): an international, prospective,
to achieve specific benefits not available in conventional randomised, non-inferiority phase 3 trial. Lancet. 2010;376:91–102.
breast radiotherapy techniques: 10. Vaidya JS, Baum M, Tobias JS, Morgan S, D’Souza D. The novel
• A dose delivery system with rotating radiation sources technique of delivering targeted intraoperative radiotherapy (Targit)
positioned circumferentially around the breast, providing for early breast cancer. Eur J Surg Oncol. 2002;28:447–54.
11. NSABP B-39, RTOG 0413: A Randomized Phase III Study of con-
36 noncoplanar arcs, producing highly conformal dose ventional whole breast irradiation versus partial breast irradiation
distributions. for women with stage 0, I, or II breast cancer. Clin Adv Hematol
• A dedicated breast immobilization device with a built-in Oncol. 2006;4:719–21.
stereotactic frame to reduce localization uncertainty and 12. Arthur DW, Winter K, Kuske RR, Bolton J, Rabinovitch R, et al.
A Phase II trial of brachytherapy alone after lumpectomy for select
improve accuracy. breast cancer: tumor control and survival outcomes of RTOG
• The dynamic utilization and control of a patient position- 95–17. Int J Radiat Oncol Biol Phys. 2008;72:467–73.
ing system during dose delivery that enables delivery of 13. Vicini F, Winter K, Wong J, Pass H, Rabinovitch R, et al. Initial
uniform doses to the treated area. The same principle is efficacy results of RTOG 0319: three-dimensional conformal radia-
tion therapy (3D-CRT) confined to the region of the lumpectomy
also utilized for differential dose painting for multiple tar- cavity for stage I/II breast carcinoma. Int J Radiat Oncol Biol Phys.
gets with different levels of prescribed doses. 2010;77:1120–7.
• Optimal beam geometry to match geometric shape and 14. Polgár C, Fodor J, Major T, Németh G, Lövey K, et al. Breast-
orientation of a patient’s breast. conserving treatment with partial or whole breast irradiation for
low-risk invasive breast carcinoma—5-year results of a randomized
• Prone setup and irradiation geometry to maximize separa- trial. Int J Radiat Oncol Biol Phys. 2007;69:694–702.
tion of the targeted breast area from critical structures 15. Valachis A, Mauri D, Polyzos NP, Mavroudis D, Georgoulias V,
such as lungs, ribs, and heart. Casazza G. Partial breast irradiation or whole breast radiotherapy
• Automatic patient positioning system to transfer the for early breast cancer: a meta-analysis of randomized controlled
trials. Breast J. 2010;16:245–51.
patient from the imaging room to the treatment room 16. Scanderbeg D, Yashar C, White G, Rice R, Pawlicki T. Evaluation
without compromising the integrity of the stereotactic of three APBI techniques under NSABP B-39 guidelines. J Appl
coordinate frame. Clin Med Phys. 2009;11:3021.
• Simplified coordinate registration and treatment planning 17. Gupta S, Scanderbeg DJ, Kamrava M, Yashar CM. Unexpected tox-
icity in a patient treated with 3D conformal accelerated partial
with inverse planning optimization. breast radiotherapy. Brachytherapy. 2009;8:207–9.
18. Jagsi R, Ben-David MA, Moran JM, Marsh RB, Griffith KA, et al.
Unacceptable cosmesis in a protocol investigating intensity-
References modulated radiotherapy with active breathing control for acceler-
ated partial-breast irradiation. Int J Radiat Oncol Biol Phys. 2010;
76:71–8.
1. Jemal A, Siegel R, Ward E, Hao Y, Xu J, Thun MJ. Cancer statis- 19. Hepel JT, Tokita M, MacAusland SG, Evans SB, Hiatt JR, et al.
tics, 2009. CA Cancer J Clin. 2009;59:225–49. Toxicity of three-dimensional conformal radiotherapy for acceler-
2. Christian MC, McCabe MS, Korn EL, Abrams JS, Kaplan RS, ated partial breast irradiation. Int J Radiat Oncol Biol Phys.
Friedman MA. The National Cancer Institute audit of the National 2009;75:1290–6.
Surgical Adjuvant Breast and Bowel Project Protocol B-06. N Engl 20. Taylor ME, Perez CA, Halverson KJ, Kuske RR, Philpott GW, et al.
J Med. 1995;333:1469–74. Factors influencing cosmetic results after conservation therapy for
3. Knobf MT, Sun Y. A longitudinal study of symptoms and self-care breast cancer. Int J Radiat Oncol Biol Phys. 1995;31:753–64.
activities in women treated with primary radiotherapy for breast 21. Andrews DW, Scott CB, Sperduto PW, Flanders AE, Gaspar LE,
cancer. Cancer Nurs. 2005;28:210–8. et al. Whole brain radiation therapy with or without stereotactic
4. Turnbull LW, Brown SR, Olivier C, Harvey I, Brown J, et al. radiosurgery boost for patients with one to three brain metastases:
Multicentre randomised controlled trial examining the cost- phase III results of the RTOG 9508 randomised trial. Lancet.
effectiveness of contrast-enhanced high field magnetic resonance 2004;363:1665–72.
imaging in women with primary breast cancer scheduled for wide 22. Timmerman R, Galvin J, Michalski J, Straube W, Ibbott G, et al.
local excision (COMICE). Health Technol Assess. 2010;14:1–182. Accreditation and quality assurance for Radiation Therapy
5. Bleicher RJ, Ciocca RM, Egleston BL, Sesa L, Evers K, Sigurdson Oncology Group: multicenter clinical trials using stereotactic body
ER, Morrow M. Association of routine pretreatment magnetic reso- radiation therapy in lung cancer. Acta Oncol. 2006;45:779–86.
nance imaging with time to surgery, mastectomy rate, and margin 23. Nataf F, Schlienger M, Liu Z, Foulquier JN, Grès B, et al.
status. J Am Coll Surg. 2009;209:180–7; quiz 294–5. Radiosurgery with or without a 2-mm margin for 93 single brain
6. Kuhl C, Kuhn W, Braun M, Schild H. Pre-operative staging of metastases. Int J Radiat Oncol Biol Phys. 2008;70:766–72.
breast cancer with breast MRI: one step forward, two steps back? 24. Timmerman R, Abdulrahman R, Kavanagh BD, Meyer JL. Lung
Breast. 2007;16 Suppl 2:S34–44. cancer: a model for implementing stereotactic body radiation ther-
7. Morrow M. Should routine breast cancer staging include MRI? Nat apy into practice. Front Radiat Ther Oncol. 2007;40:368–85.
Clin Pract Oncol. 2009;6:72–3. 25. Hadziahmetovic M, Loo BW, Timmerman RD, Mayr NA, Wang
8. Clarke M, Collins R, Darby S, Davies C, Elphinstone P, et al. JZ, et al. Stereotactic body radiation therapy (stereotactic ablative
Effects of radiotherapy and of differences in the extent of surgery radiotherapy) for stage I non-small cell lung cancer—updates of
for early breast cancer on local recurrence and 15-year survival: an radiobiology, techniques, and clinical outcomes. Discov Med.
overview of the randomised trials. Lancet. 2005;366:2087–106. 2010;9:411–7.
26. Qiao X, Tullgren O, Lax I, Sirzén F, Lewensohn R. The role of 33. Li X, Zhang J, Wang C, Yu C. Stereotactic radiation therapy for
radiotherapy in treatment of stage I non-small cell lung cancer. breast cancer: dosimetry feasibility. Int J Radiat Oncol Biol Phys.
Lung Cancer. 2003;41:1–11. 2010;78:S139.
27. Timmerman R, Paulus R, Galvin J, Michalski J, Straube W, et al. 34. Mutaf YD, Yu C, Zhang J, Prado K, Feigenberg S, et al. Dosimetric
Stereotactic body radiation therapy for inoperable early stage lung characterization of a novel stereotactic radiotherapy device for
cancer. JAMA. 2010;303:1070–6. breast cancer. Radiother Oncol. 2011;99(S1):S160–1.
28. Timmerman RD. Surgery versus stereotactic body radiation therapy 35. Holland R, Veling SH, Mravunac M, Hendriks JH. Histologic mul-
for early-stage lung cancer: who’s down for the count? J Clin tifocality of Tis, T1–2 breast carcinomas. Implications for clinical
Oncol. 2010;28:907–9. trials of breast-conserving surgery. Cancer. 1985;56:979–90.
29. Bondiau PY, Lallement M, Bahadoran P, Birtwisle-Peyrottes I, 36. Faverly DR, Hendriks JH, Holland R. Breast carcinomas of limited
Chapellier C, et al. CyberKnife and neoadjuvant chemotherapy for extent: frequency, radiologic-pathologic characteristics, and surgi-
breast tumors: preliminary results. Cancer Radiother. 2009;13: cal margin requirements. Cancer. 2001;91:647–59.
79–84. 37. Nichols EM, Feigenberg SJ, Marter K, Lasio G, Cheston SB, et al.
30. Bondiau PY, Bahadoran P, Lallement M, Birtwisle-Peyrottes I, Pre-operative therapy increases patient eligibility for accelerated
Chapellier C, et al. Robotic stereotactic radioablation concomitant partial breast irradiation (APBI) and decreases normal tissue expo-
with neo-adjuvant chemotherapy for breast tumors. Int J Radiat sure. Int J Radiat Oncol Biol Phys. 2010;78:S251.
Oncol Biol Phys. 2009;75:1041–7. 38. Yu C, Feigenberg S. Radiat Oncol (London, England). Rad. Onc.
31. Yu XC. US National Cancer Institute 4R44CA132254-02; 2010. journal. 2011; 99:S305.
32. Nichols EM, Dhople AA, Mohiuddin MM, Flannery TW, Yu CX, 39. Nelson WR, Hirayama H, Rogers DW. Electron-photon transport
Regine WF. Comparative analysis of the post-lumpectomy target using the EGS4 Monte Carlo code. Trans Am Nucl Soc.
volume versus the use of pre-lumpectomy tumor volume for early- 1986;52:379–80.
stage breast cancer: implications for the future. Int J Radiat Oncol 40. Kawrakow I, Walters BR. Efficient photon beam dose calculations
Biol Phys. 2010;77:197–202. using DOSXYZnrc with BEAMnrc. Med Phys. 2006;33:3046–56.
Stereotactic Body Radiation Therapy
13
Steven J. Feigenberg, Randi Cohen, Navesh K. Sharma,
Zain Husain, Shifeng Chen, and Laura A. Dawson
to account for tumor motion and setup uncertainties. This

Introduction added margin increases the volume of tissue irradiated, which
theoretically increases the risk of side effects. Over the last 15
Over the last 30 years, advances in immobilization, computer- years, solutions have allowed SBRT/SABR to treat tumors
based treatment planning and delivery, and image guidance outside the brain.
have revolutionized the ways in which small tumors are SBRT refers to the use of a limited number of high-dose
treated with radiation therapy (RT). Stereotactic radiosur- fractions delivered conformally to targets with high accu-
gery (SRS) was introduced in 1968 as a tool for noninvasive racy, using biologic doses of radiation far higher than those
neurosurgery but was not widely adopted until three- used in standard fractionation. Stereotactic refers to the use
dimensional imaging became available. The pinpoint accu- of a reference system to aid in localizing the tumor. An exter-
racy and high-dose concentration achievable with SRS have nal reference system, such as a stereotactic head frame for
allowed high “tumor ablative” treatment of target tissues. cranial SRS or internal fiducial markers including the tumor
SRS has been a success story for patients with brain metasta- itself, may be used for localization and guidance. For tumors
ses by improving local control, quality of life, and survival. fixed to bony anatomy, such as base of skull cancers, image
Its success has hinged on minimizing the volume of normal guidance using bony anatomy may be appropriate; however,
tissue receiving a “significant” dose of radiation while maxi- often the position of the tumor is not highly correlated with
mizing dose to the tumor. the bones or an external reference system, and imaging and
SRS outside the brain, called stereotactic body RT (SBRT) targeting of the tumor itself (i.e., image-guided radiation
or more recently stereotactic radioablation (SABR), has been therapy [IGRT]) may be required.
more challenging. Treatment planning of intracranial targets SBRT is becoming more popular, as evidenced by increas-
involves delivery of a very conformal dose of radiation to the ing recognition at international meetings, a textbook devoted
radiographic target with no margin, because the brain does to the topic [1], and an American Society for Therapeutic
not move in reference to the rigid stereotactic head frame. Radiology and Oncology (ASTRO) consensus document on
When the target is located outside the brain, immobilization SBRT [2]. The ASTRO consensus document defines SBRT
is less accurate and organ/tumor motion requires the physi- as high-precision radiotherapy delivered in three or more
cian to add a margin to account for these differences to ensure fractions to very potent doses of highly conformal radiation
appropriate dose to the gross target volume (GTV). The plan- with steep dose gradients around the target. SBRT is “an
ning target volume (PTV) is defined by an additional margin evolving technology,” and definitions may change with time.
The philosophies and techniques of SBRT can be applied to
S.J. Feigenberg, M.D. (*) • R. Cohen, M.D., M.S. • N.K. Sharma, longer fractionation regimens (when the target is intimately
D.O., Ph.D. • S. Chen, Ph.D. associated with a critical serial functioning normal tissue, for
Department of Radiation Oncology, The University of Maryland example) or shorter fractionations, including SRS.
School of Medicine, Baltimore, MD, USA Similar to cranial SRS, multiple static or dynamic beams
e-mail: sfeigenberg@umm.edu
in a variety of beam arrangements, with or without segments
Z. Husain, M.D. or intensity modulation, can be used to produce a dose distri-
Department of Therapeutic Radiology, Yale University,
New Haven, CT, USA bution in which isodose lines tightly conform to the target
volume. Although most reports on SBRT refer to megavolt-
L.A. Dawson, M.D., F.R.C.P.C.
Department of Radiation Oncology, Princess Margaret Cancer age photon irradiation, protons are not excluded from the
Centre, University of Toronto, Toronto, ON, Canada SBRT definition. Inhomogeneity within the target volume is
178 S.J. Feigenberg et al.
permitted and hot spots have been encouraged within the stereotactic reference frame to aid in guidance of radiation.
PTV to increase the chance of tumor ablation, although with Blomgren et al. [5] reported their initial experience in 1995,
a larger rim of normal tissue found inside of the target vol- in which 31 patients with primarily solitary lung and liver
ume in comparison to cranial SRS that complications may be tumors were treated with one to four fractions. Reproducibility
increased. Dose is generally prescribed to an isodose line of this approach, based on patients who had repeat CT scans
covering the PTV similar to cranial SRS, with a very steep in the treatment position, was found to be within 5 (axial) to
dose gradient outside the PTV. 8 (cranial–caudal) mm for 90 % of setups with diaphragm
SBRT is a noninvasive, outpatient intervention, generally motion reduced to 5–10 mm with abdominal pressure.
completed within 1 or 2 weeks, allowing for little to no Experience in North America has been growing exponen-
delays in systemic therapy. The short radiation treatment tially following the pioneering work of Dr. Timmerman [6–9]
time and high dose per fraction in SBRT have potential who systematically evaluated SBRT by a series of prospec-
radiobiological therapeutic advantages compared with con- tive clinical studies. Based on this experience, he led the
ventional fractionation, due to less tumor repopulation and Radiation Oncology Therapy Group (RTOG) initiative with
repair. In addition, there is growing evidence that SBRT/ a multi-institutional phase II trial of SBRT in lung cancer
SABR may have an impact on disease outside of the treated which was published in 2010 in JAMA [10]. The results
volume decreasing regional and distant recurrences, the were so impressive that the paradigm of surgery as the stan-
“abscopal effect.” Furthermore, the shorter SBRT treatment dard of care for small tumors is currently being challenged in
times are more convenient for patients and have resource uti- RTOG 10–14 [11]. Similar North American studies have
lization advantages. been reported for liver metastases, vertebral metastases,
The rationale for SBRT is that there is a need for improved adrenal metastases, pancreatic carcinomas, recurrent head
local therapies for many primary cancers and also in the situ- and neck cancers as well as for tumors of the retroperito-
ation when there are “oligo” (i.e., isolated) metastases, spe- neum and pelvis.
cifically in sites where surgery has been shown previously to
be able to cure patients with “oligo” metastases (e.g.,
colorectal cancer (CRC), renal cell cancer, and sarcoma). Radiobiology
SBRT is less invasive than surgery, and once the safety of
SBRT has been established, SBRT has the potential to be Radiobiologic Models
used in place of surgery in situations when surgery may be
associated with high risk. Even in incurable situations, SBRT The linear quadratic model may not be valid for very short
may be of significant advantage to patients with metastatic fractionation schemes. In the early development of this
disease on systemic therapy where only a few tumors are model, one of the developers expressed that the model is not
increasing in size and the rest of their disease is stable. intended for doses higher than 8–10 Gy. The linear quadratic
Patients could be maintained on their current systemic ther- model may overpredict tumor cell kill and underpredict toxic-
apy, if they are tolerating, rather just changing to a new sys- ity at the increased dose per fraction. Fowler et al. [12–14]
temic therapy of unknown efficacy and morbidity potentially predicted that local control should be excellent for non-small
improving quality of life. cell lung cancer (NSCLC) with biological doses of the order
In this chapter, a historical perspective, rationale, techni- 100 Gy in 2-Gy equivalent dose or higher (α/β = 10), although
cal details, and literature on SBRT/SABR outcomes and based on the pioneering work at Michigan [15, 16] the local
complications will be reviewed. Specific applications of control was only 70 % at this dose level. Although there
SBRT will be reviewed in more detail where there is an seems to be a threshold of superior local control at 100
abundance of data including lung cancer and metastases, Gy (α/β = 10) when higher doses per fraction are used.
liver metastases and hepatocellular carcinoma (HCC), spinal Fractionation schemes with an estimated tumor control prob-
malignancies and pancreatic cancer. ability (TCP) of 95 % or greater include 45 to 60 Gy in three
fractions and 60 Gy in five fractions, although one must be
extremely careful to review how the dose was prescribed as
Historical Perspective investigators prescribe dose very differently. German investi-
gators [17] prescribe dose to the 65 % isodose shell while the
The first extracranial site to be investigated with SRS tech- Japanese routinely prescribe to the isocenter—which can lead
niques was the spine in 1993. A rigid skeletal fixation device to very different doses seen by the PTV. Due to the inadequa-
that immobilized the spine above and below the target was cies in the linear quadratic model in predicting outcomes,
used as a stereotactic reference frame to guide radiation [3]. investigators [18] have developed the universal survival
Around the same time, Blomgren and Lax [4, 5] from curve. This combines the linear quadratic model and the
Sweden began applying stereotactic techniques to body multi-target model which may be more accurate. Although as
targets such as lung and liver tumors. They developed a body the investigators correctly point out, “the complex biology
13 Stereotactic Body Radiation Therapy 179
that underlies cell ablation and loss of clonogenicity pre- associated with cytotoxic immune cell populations, is associ-
cludes an explanation by any simple mathematical model.” ated with improved tumor control [22]. Radiotherapy, in its
classic delivery with standard fractionation, has traditionally
been viewed as an immunosuppressive modality [23].
Mechanism of Action However, the actual effects on the immune system are
extremely complex with conflicting reports on whether they
It is clear that the mechanism of cell kill is different between promote or interfere with tumor reduction.
a conventionally fractionated course of radiotherapy as One of the leading hypotheses for this “abscopal” effect
opposed to the “ablative” doses used with SBRT. The classic with SBRT is its relationship to CD8+ T-cell-mediated
model of radiobiology is defined by the depletion of tumor immune response. The lower than expected regional failure
stem cells and undifferentiated progenitor clonogens for rate with SBRT may be due to radiation-induced systemic
radiotherapy to be curative. With conventionally fractionated stimulation of T-cell immune responses, leading to eradica-
radiotherapy where large volumes of normal tissue receive tion of occult regional micrometastases. Support for this
the prescribed treatment dose, normal tissue stem cells are theory derives primarily from experiments in animal models.
better at repairing radiation damage than tumor clonogens. A recent study examined the effect of ablative doses of radia-
The surrounding microenviroment was not thought to be an tion therapy in a mouse melanoma model [24]. In this study,
active contributor to this effect until recently. As a possible mice with B16 melanoma tumors received 20 Gy of radiation
explanation, Fuks et al. [19] initially suggested that contrary in a single fraction. Examination of the tumor microenviron-
to conventional (1.8–3 Gy) fractionation, high-dose radiation ment and lymphoid tissues 1–2 weeks posttreatment demon-
(>8–10 Gy per fraction) specifically impacts more robust strated tumor regression and an increase of infiltrating T
endothelial apoptosis and microvascular dysfunction, which cells. When the experiment was repeated with athymic nude
in turn leads to increased cell death. Furthermore, hypoxia mice, which lack T cells, no statistically significant decrease
resulting from standard fractionated radiotherapy regimens in tumor volume was observed, suggesting that the effects
results in a burst of pro-angiogenic activity in the tumor were T-cell mediated. Taken together, these studies suggest
microenvironment, generating HIF-1α, VEGF, and other that ablative doses of radiation alter the tumor microenviron-
vasculogenic factors which, in turn attenuate radiation- ment, causing T-cell infiltration resulting in vigorous prim-
induced apoptosis in endothelial cells. ing and expansion of effector T cells. The experiment was
repeated in wild-type mice with B16 tumors, which had been
subjected to antibody-mediated CD8+ T-cell depletion. The
Immune Response to Radiation therapeutic benefit to the ablative radiation was diminished,
and survival in this CD8+ depleted group was decreased by
SBRT appears to impact disease outside the radiated volume. 75 %. In a separate experiment using a transplantable mouse
This is best exemplified by a retrospective experience from 4T1 mammary carcinoma model (which mimics metastatic
William Beaumont Hospital [20] which compared patients ability of breast cancer cells to lung, bone, liver, etc.), abla-
who were treated with either a sublobar resection or SABR tive radiation therapy delivered to the primary tumor site led
during the same time period. SABR not only resulted in a to a complete elimination of lung metastasis. This observa-
drastically lower local failure rate (5 % versus 24 %, p = 0.05), tion strongly suggests the induction of a potent antitumor
but more unexpectedly had a lower regional lymph node fail- immune response. Separate experiments utilizing CD8 deple-
ure rate (18 % versus 0 %, p < 0.05) even though patients tion strategies demonstrate a relative increase in number of
were more likely to be surgically staged to be node negative distant metastases, again suggesting that CD8+ T cells are
in the resected cohort (71 % versus 20 %). Given that patients critical for mediating protection against tumors. Taken in
treated using SABR have a very small (tumor plus margin) combination, these studies suggest that CD8+ T cells play a
volume of tissue irradiated, and few if any lymph nodes critical role in radiation-induced antitumor immune responses
included in the treatment field, this was a surprising finding. (both locally and distantly), following ablative dose therapy.
One possible explanation of this phenomenon is the stimula- Clinically, similarly suggestive observations have been
tion of T-cell immunity by SABR, leading to the eradication reported in several tumor types. Mahmoud et al. [25, 26]
of occult regional micrometastases. recently demonstrated the influence of density and distribu-
The immune system has long been known to play a criti- tion of the cytotoxic CD8+ T cells in breast cancer patients.
cal role in tumor surveillance and control [21]. It is thought Using immunohistochemistry staining of tissue microarray
that tumors develop by escaping immune detection, either by cores, they established that infiltration of CD8+ T cells into
downregulating surface recognition antigens or by releasing the distant stroma correlated with improved overall survival.
immunomodulatory cytokines that dampen the immune In addition, the number of T regulatory cells, the ratio of
response, and are found in greater prevalence in immunosup- CD4+ to CD8+ T cells, and the presence of T lymphocytes in
pressed populations. Conversely, a robust immune response, the primary tumor correlated with grade, stage, and survival.
Fig. 13.1 Baseline CT of

isolated liver metastases and
follow-up imaging at 3 months
and 18 months revealing no
evidence of progressive disease
Furthermore, several groups have demonstrated that the the medically inoperable population, identifying the extent
presence of tumor-associated lymphocytes in breast cancer is of tumor response to initial SBRT is critical in determining
an independent predictor of response to anthracycline/ further treatment response, as is identifying dose–response
taxane-based chemotherapy [27]. Paulson et al. [28] have characteristics on current SBRT trials. Specifically, is there
demonstrated that intratumoral CD8+ lymphocyte infiltration an imaging technique that can be used as a biomarker or
was the most important predictor of outcomes in patients early surrogate marker of outcome? Currently, computerized
with Merkel cell carcinoma. Interestingly, in this population, tomography remains the imaging modality of choice to eval-
patients with tumors that demonstrated a robust CD8+ lym- uate response following SBRT. However, with the very high-
phocyte infiltration were observed to have the largest dose fractions utilized in SBRT, a significant percentage of
improvements in regional and distant control as compared to patients treated by SBRT will develop fibrosis within and in
local control. A recently published report [29] appears to close proximity to the treatment field. Takeda et al. [30, 31]
substantiate this combination of ipilimumab with SABR. In described solid consolidations on CT scans are seen as late
this case study, a patient with progressive metastatic mela- changes in the center or the periphery of the PTV in upwards
noma receiving maintenance ipilimumab therapy was treated of 73 % of patients. Radiographic findings, that occur as a
palliatively with SABR for a painful paraspinal mass. While result of SBRT, are more dramatic in appearance as opposed
there was no initial response, a single additional dose of ipili- to what is seen clinically (see Fig. 13.1). These consolida-
mumab not only significantly decreased the size of the para- tions can be quite complex making serial CT follow-up
spinal mass but also caused regression of additional sites of rather difficult, thereby making it difficult to determine
metastatic disease which were of considerable distance from whether these changes are consistent with recurrent carci-
the radiated paraspinal mass. Further analysis demonstrated noma or radiation changes.
that RT had increased antibodies to NY-ESO-1, an antigen In this regard, assessment of tumor response to therapy
associated with melanoma, by 30-fold. Since presence of using functional information could be very helpful in
antibodies to NY-ESO-1 had been associated with a more the absence of good anatomical imaging such as
robust response to ipilimumab, this may explain why this 18-fluorodeoxyglucose-positron emission tomography
patient had such an excellent response to disease within and (FDG-PET) scanning modality. PET scans have had a
outside the radiation portal. Furthermore, the RT regimen in significant impact on the management of patients with
this patient also led to a subsequent spike in inducible co- many cancers but it has been studied most extensively in
stimulator, a marker of activated CTLs, suggesting an addi- NSCLC. Their most important contribution has been in the
tional mechanism for the observed tumor control. improvement in the preoperative evaluation of mediastinal
lymph nodes over computed tomography (CT) scans alone
with increased sensitivity and specificity [32–34] and in
Imaging Response to SBRT staging improving patient selection for curative therapy.
FDG-PET has also been prognostic in the pretreatment
With these promising early results, two cooperative groups setting and following induction chemotherapy [35] or
are evaluating SBRT as an alternative to surgery. However, definitive chemoradiotherapy to determine tumor response
assessment of tumor response to SBRT in a time frame that as defined by FDG uptake for predicting outcomes [36, 37].
will allow those with failures the maximal curative benefit Following SBRT, PET findings have been mixed with only
from salvage surgery remains a challenge. Furthermore, in a few small series reported.
A small prospective series [38] from Fox Chase Cancer PET scanning has been recently challenged. Investigators
Center which included ten medically inoperable patients have looked at L-s-methyl-11C methionine (MET)
treated for NSCLC on a phase I dose escalation demon- PET. While MET has been shown to have decreased in
strated that a drop in the maximum standardized uptake val- inflammatory lesions as compared to FDG, a small series
ues (SUV) comparing the pre-SBRT and the 3-month from Japan in the SBRT setting showed very little difference
post-SBRT PET scan appeared to be a predictor of local con- between the two tracers [43]. 18F -3′-fluoro-3′-deoxy-L-
trol. A larger retrospective experience [39] from the same thymidine (FLT) [44], a nucleoside reverse transcriptase
institution including 26 patients addressed using FDG-PET inhibitor, has been suggested to be superior to that of FDG in
response (comparing pre-SBRT versus 3-month post-SBRT) detecting cellular and proliferative responses to chemother-
as a surrogate of local failure in greater detail. All 26 PET apy; the role of FLT-PET as a surrogate of response follow-
scans were reviewed and interpreted by a single physician ing stereotactic body radiotherapy (SBRT) for lung cancer
who was blinded to the clinical results. FDG-PET scans were remains unexplored. Similar data is available for patients
analyzed semi-quantitatively by comparing pre- and post- with liver metastases and lung metastases.
treatment changes in SUV. This study demonstrated two
major findings: (1) A lower initial SUV value was the only
significant factor associated with an increased risk of local Normal Tissue Response to SBRT
failure (3.4 versus 5.7; p = 0.045) with 4 out of the 5 local
failures having a low pre-SBRT SUV (<4). (2) Decreases in Normal tissues that function primarily as parallel normal tis-
post-SBRT SUV of >55 % of its pre-SBRT value signifi- sues are made up of predominately structurally defined func-
cantly decreased the risk of local recurrence. Similar find- tional subunits (FSUs) (e.g., peripheral lung, liver, kidney).
ings were seen when this data was combined with data from Serial functioning normal tissues, characterized by a chain of
the University of Maryland [40] which demonstrated a crude function, are composed of primarily undefined FSUs (e.g.,
local failure rate of 54.5 % failure rate (6/11) for patients gastrointestinal mucosa, trachea, spinal cord, bronchus).
with SUV <4, while only 11 % (4/36) of patients with SUV Parallel organ toxicity is mostly related to volume irradiated
≥4 failed locally. Hoopes et al. [41] have described with (e.g., mean dose, volume treated to a threshold dose such as
nearly 50 % of patients at 1 year having SUVmax changes V20), whereas serial organ toxicity is mostly related to the
>3.5 from their prospective study and 15 % in their retro- maximum dose delivered to that tissue. Although useful, this
spective experience (4 of 28 with a max SUV >2.5). distinction between serial and parallel functioning organs is
Presumably, though, all were lower than pretreatment. These likely too simplistic, and most normal tissues likely have
patients were too compromised to undergo further treatment both parallel and serial functionality. This has been demon-
and were thus followed. They remained without recurrence strated in elegant rat model experiments of spinal cord toler-
between 8 and 22 months post these PET scans. This illus- ance to radiation therapy by van der Kogel et al. [45]. The rat
trates that the use of a specific SUV value is not useful and spinal cord tolerance to radiation was found to be dependent
therefore trying to determine a specific cutoff SUV for a on the volume irradiated and the spatial distribution of dose.
population to predict recurrence is not appropriate. With The gray matter of the cord was found to be most resistant,
these inflammatory changes so prevalent, it is imperative for whereas the lateral white matter was most sensitive.
the reading nuclear medicine physician and treating physi- As the volume of normal tissues irradiated to high doses
cian to review the images to define the solid and inflamma- with SBRT is generally less than the volume of normal tissue
tory components and their corresponding SUV in relationship irradiated with conventional fractionation, for normal tissues
to the distribution of the treatment to minimize the possibil- that are primarily parallel functioning, if the volume irradiated
ity of false-positive readings. It is therefore to our advantage is low enough, the risk of toxicity may be extremely low
to have scans read by a single physician for consistency in despite delivery of very high doses to a small volume.
reporting. However, for serial organs that are in close proximity to
Further refinement of this data is vital as quantification tumors, even a small volume irradiated to a high dose may
with SUVmax oversimplifies the evaluation of tumor lead to irreversible toxicity. Thus, SBRT should be used cau-
response and sometimes results in inconclusive or inaccurate tiously for tumors adjacent to serial functioning organs such as
diagnosis. Currently investigators are trying to extract more the esophagus or spinal cord. An increase in the number of
detailed spatial–temporal PET/CT features [42] that compre- fractions may improve the therapeutic ratio and should be con-
hensively characterize the whole tumor’s intensities and dis- sidered in place of hypofractionated SBRT in this situation.
tributions, spatial variations (texture), and shape properties With a dramatic increase in the fraction size and total
which may become better predictors of treatment response. dose with SBRT, the repair mechanisms may not be initiated
In addition, novel radiotracers are in development. The use as they would at a lower dose per fraction, potentially lead-
of fluorodeoxyglucose (FDG) as the radiotracer of choice for ing to permanent damage to the normal tissue within the
high-dose volumes and/or unpredictable effects outside the Noninvasive stereotactic systems using a frame, with or
high-dose volume. Given the potential for normal tissue without implanted fiducial markers, have been used for para-
injury and that the volume tolerance of normal tissues to spinal tumor SBRT with accuracy within 2 mm [49]. Optical-
such high doses per fraction is unknown, most SBRT has guided three-dimensional ultrasound has also been used for
been applied to small tumors (<5 cm) in which the volume of spinal SBRT to ensure that the patient does not move during
normal tissue around the tumor is small. SBRT is being radiation [50].
investigated for larger tumor volumes [46], and it is hopeful
that clinical data will eventually be obtained to provide guid-
ance to what the dose–volume toxicity relationship is for Body Tumor Immobilization
organs irradiated with inhomogeneous doses from hypofrac-
tionated fractionation schemes. Nonrigid fixation has been performed for SBRT with
As very steep dose gradients are used with SBRT with specialized body frames that have fiducial systems attached
rapid falloff of dose in surrounding normal tissues, not only to the frame and a device to control respiration using abdom-
is there substantial variability of dose throughout normal inal compression [4]. Abdominal compression using such
tissues, but also there is substantial variability in the dose frames has been found to reduce diaphragm caudal–cranial
per fraction. Thus, when dose–volume characteristics of a motion to less than 5–10 mm in most patients. Reproducibility
normal tissue associated with toxicity risk are investigated, of target positioning using these frames has been reported to
correction for the variability in dose per fraction should be be better than conventional immobilization systems, with
considered. Unfortunately, accurate knowledge of the α/β positional deviations of lung cancer and liver cancer position
ratio for normal tissues is lacking. Clinical trials of novel less than 10 mm in 98 % of patients [4, 51, 52].
fractionation regimens are required to confirm outcomes and An option to these specialized body frames that facilitates
toxicities associated with the many different fractionation guidance is to not use the frames for guidance, but to image
schemes used in SBRT. Under each of the clinical sections, internal anatomic references, such as bones near the target or
additional data will be presented based on complications to the soft tissue tumor itself, to define the treatment coordi-
the bowel, chest wall, rib, liver, and brachial plexus. nates. Repositioning and repeat verification imaging are
required to ensure the patient was moved to the correct posi-
tion. A variety of immobilization devices can been used with
Immobilization this strategy, as long as they keep the patient immobilized
and preferably are (relatively) comfortable to minimize
Overview intrafraction patient motion.
Given the sensitivity of highly conformal SBRT plans to

setup uncertainty and organ motion, reduction of geometric Assessment of Breathing Motion
uncertainties and organ motion is important. The choice of
patient position and immobilization may impact setup error, Organ motion due to physiologic functions during a radia-
organ motion, as well as intrafraction motion secondary to tion treatment fraction can be substantial. For example, the
patient discomfort. For example, prostate motion due to liver can move up to 5 cm in the caudal–cranial direction
breathing is reduced in the supine position compared with during free breathing [53] causing motion of the upper
the prone position [47, 48]. The different immobilization abdominal and lower thoracic cavity. As this motion can
devices and strategies that have been used for SBRT are result in alternations in target and normal organ volume defi-
described below. nitions, PTV margins, and the entire dose distribution, inter-
ventions to reduce the impact of intrafraction organ motion
are required for many SBRT patients, to facilitate dose esca-
Spinal Tumor Immobilization lation and to reduce the volume of normal tissue irradiated.
Strategies to compensate for breathing motion include the
Similar to intracranial SRS with a cranial halo secured to the use of abdominal pressure, voluntary shallow breathing, vol-
skull with transcutaneous pins, rigid fixation systems have untary deep inspiration, voluntary breath holds at variable
been used for paraspinal and spinal SBRT. Hamilton in 1995 phases of the respiratory cycle, active breathing control
immobilized the spine for SBRT with rigid skeletal fixation (ABC), gated radiotherapy, and real-time tumor tracking.
above and below the involved segments. With this system, Although voluntary breath holds may be beneficial for some
excellent accuracy, with less than 2-mm offsets, was observed patients, there is potential for leaking air and patient error,
[3]. However, such an approach is invasive, and avoidance of particularly for patients with lung disease where usually
invasive fixation systems is desirable to minimize risks. less than 30 % of patients can even tolerate this approach.
ABC refers to organ immobilization with breath holds that situations, there may be rationale for breath hold, gating, or
are controlled, triggered, and monitored by a caregiver. In tracking during inhalation. For example, for lung tumors and/
approximately 60 % of patients with liver cancer, ABC was or tumors adjacent to the heart, inhalation will reduce the den-
used successfully, with excellent reproducibility of the liver sity of the lungs and/or may move critical structures away
relative to the vertebral bodies within the time period of one from the target volume.
radiation fraction (intrafraction reproducibility, σ, of the Other approaches to minimize respiratory and non-
liver relative to the vertebral bodies: 1.5–2.5 mm) [54, 55]. respiratory organ motion include maintaining the same pre-
However, with ABC, from day to day the position of the parative regimen prior to each treatment and ensuring
immobilized liver varies relative to the bones (interfraction comfortable immobilization and short overall treatment time
reproducibility, σ 3.4–4.4 mm), providing rationale for daily to reduce patient movement due to discomfort and physiologic
image guidance when ABC is used for liver SBRT although change in organs such as stomach filling. Another general
treatment time needs to be kept reasonable as patients with intervention is patient feedback, either auditory or visual.
poor performance status have increased treatment uncertain- Ideally, feedback would be from direct tumor imaging; how-
ties and all patients have increased uncertainties with treat- ever, feedback from imaging of adjacent organs, spirometry,
ment times >10 min due to fatigue, muscle relaxation, or nasal flow monitoring, external marker position, or optical
peristalsis. monitoring of body contour are often more practical options.
Gated radiotherapy, with the beam triggered to be on only If indirect measures of organ position are used, confirmation
during a predetermined phase of the respiratory cycle, usu- for an individual patient that the indirect measure is indeed
ally refers to the use of an external surrogate for tumor posi- directly related to organ position is mandatory.
tion (as opposed to direct tumor imaging) to gate the
radiation. This can be used to reduce the volume of normal
tissue irradiated. Changes in baseline organ position can Treatment Planning
occur from day to day [56], and thus image guidance is
important to avoid geographic misses, especially in the set- Overview
ting of SBRT. One concern regarding gated radiotherapy has
been brought to light by a French prospective phase III trial As dose gradients are steeper and doses are higher with
presented at ASTRO 2012 [57]. This trial was testing the SBRT than with conventional radiation therapy, the conse-
benefit of gating in locally advanced lung cancer setting and quences of error in tumor delineation, errors introduced by
surprisingly demonstrated an increase in radiation pneumo- dosimetry, and geometric uncertainties may be more dele-
nitis in patients treated with gating. The manuscript will terious. Thus, all aspects of treatment planning that are
hopefully shed more light on other variables that led to this important in conformal radiation planning are even more
paradoxical finding but more importantly should put caution crucial in SBRT, especially for tumors in close proximity to
in inexperienced users of this technology. critical normal tissues, where a systematic error could lead
Tumor tracking is another approach to reduce adverse to permanent serious toxicity if the normal tissue planned
effects of organ motion. An elegant real-time tumor tracking to be spared from radiation is irradiated to the high doses
system consisting of fluoroscopic X-ray tubes in the treat- planned for the tumor.
ment room allowing visualization of radiopaque markers in
tumors was first described by Shirato et al. The linear accel-
erator is triggered to irradiate only when the marker is located Imaging at Simulation
within the planned treatment region [58]. As an alternative to
turning the radiation beam off when the tumor moves outside At the time of simulation, patient positioning and the imaging
treatment region, multileaf collimators, the couch position, modality (CT, MRI, PET/CT), resolution (e.g., CT thick-
or the entire accelerator on a robotic arm may move with the ness), and phase of contrast (e.g., arterial IV contrast for HCC
tumor to ensure adequate tumor coverage at all times (e.g., or how to give IV contrast during a four-dimensional scan for
CyberKnife image-guided radiosurgery system; Accuray, a tumor near a major vessel) must be chosen carefully. The
Sunnyvale, CA). The latter system uses dual orthogonal fluo- patient treatment position should be kept the same as scan-
roscopy tubes to track radiopaque markers in or near the ning position. Due to the relatively long treatment time, a
tumor at a preset frequency. comfortable patient position is preferred to reduce the
There are advantages to gating, breath hold, and tracking in intrafraction motion. Since more beams are used at SBRT
exhalation phase of the respiratory breathing cycle versus planning and they need to be separated maximally in space,
inhalation. These include the fact that exhalation tends to be beam arrangement should be considered during patient simu-
more reproducible and is longer than inhalation, so that treat- lation. For example, for lung cancer patient, lifting both arms
ment during exhalation reduces duty time. However, in certain allows beam arrangements in 360°. The report of AAPM Task
Group 101 [59] has recommendations about the CT thickness may be as high as 60 Gy in six fractions at the periphery.
and scanning length. In this report, 1–3-mm CT slice thick- Careful radiologic–pathologic studies accounting for organ
ness was recommended and a typical CT scan for planning deformation and shrinkage are required to determine whether
purpose should cover the tumor site plus 5–10 cm in both the use of a CTV margin is required or not.
superior and inferior directions. If non-coplanar beams are Finally, appropriate PTV margins must be used to ensure
used, 15 cm is recommended to be extended in both superior that the actual planned doses are delivered to the tumor. The
and inferior directions. Motion must be considered at this PTV margins must consider setup uncertainty and internal
time, as breathing introduces artifacts in the tumor definition, organ motion. Individual institution setup uncertainty data
normal tissue definition, and resultant errors in TCP and nor- should be used if available. Individual patient internal organ
mal tissue complication probability (NTCP). Furthermore, if motion is preferable to using population-based respiratory
motion is not considered, there is potential for a systematic motion. Of note, the classic PTV margin papers on PTV mar-
error from the time of simulation to the time of treatment to gin determination, such as that published by Van Herk et al.
occur. Motion due to breathing is largest for tumors near the [62], do not apply to very tightly conformal plans delivered
diaphragm (i.e., the upper abdomen and the lower thorax). in a few fractions, such as those used in most SBRT cases.
One method to account for motion is to eliminate it, for exam- Thus, the margin recipes that are used for conventional radia-
ple, with a breath-hold scan. Diagnostic breath-hold scans are tion planning may be inappropriate for SBRT plans.
often obtained in the inhale position, which may not corre- Modeling has demonstrated that when PTV margins are too
spond with the treatment position (e.g., exhale breath hold). small, higher doses must be delivered for the same TCP [63].
An effort to conduct all imaging to be used for planning with
the patient in the same position, with the same phase of breath
hold, is required. If breath hold is not possible or if reproduc- Planning
ibility of this breath hold cannot be documented, reduction of
breathing motion may help reduce the negative impact of Often, many beams or dynamic conformal arcs are used to
breathing motion. However, even with a small range of develop a highly conformal SBRT dose distribution. In prin-
breathing motion, errors in tumor and normal tissue volumes ciple, a great number of beams maximumly separated in
may occur, resulting in a geographic miss or excessive toxic- space lead to more conformal plan; however, the beam opti-
ity. An option to breath-hold imaging is to obtain a four- mization should consider target size and irregularities, OAR
dimensional imaging data set. From this, any position could avoidance, the length of beam path, the treatment delivery
be used for planning and image guidance [56]. Planning on time, and patient safety. The energy of 6–10 MV is preferred
the exhale data set with asymmetric PTV margins is an option due to the neutron contamination and the larger penumbra of
[60], as is planning using the mean tumor position. The phase high-energy photon beam. Limited high-energy beam may
of the breathing cycle in which the patient is planned should be allowed if the beam path is larger than 10 cm. Sometimes,
correspond with the phase of breathing cycle used for image non-coplanar beams or arcs are used if required to reduce the
guidance and treatment. dose to normal tissues. For example, 8–12 beams may be
used for a typical lung SBRT plan. Non-coplanar beam setup
should be used carefully to avoid the patient, couch, and gan-
Target Volumes try collision. Although the more beams are preferred in
SBRT planning, the resultant longer treatment time has to be
A decision has to be made regarding whether a margin is considered in practice. One strategy to obtain highly steep
required for microscopic disease risk or the clinical target dose gradients at the edge of the PTV is to close the aperture
volume (CTV) margin. Although in many SBRT series, no of the beams to the PTV or less, not leaving a gap for penum-
extra margin for CTV has purposely been added but due to bra as usually done for conformal radiation therapy. When
the less steep falloff in dose when treating the body as com- enough beams are summed together, the PTV may be cov-
pared to the brain, there is always a dose gradient from the ered by a lower isodose such as 65 % that is often at the
prescription dose to a “microscopic dose” [61]. In addition, steepest part of the dose gradient. Choosing such a low
the additional margin added for motion and setup error likely prescription isodose line may have some theoretical disad-
overlaps with the needed margin extension. This added mar- vantages since a significant rind of normal tissue is in the
gin may also be significantly different based on histology PTV. Investigators at Fox Chase Cancer Center and at the
where subclinical disease distant from the index lesion is University of Maryland use a 3-mm distance between
more prevalent such as in breast cancer. At Princess Margaret the beam aperture and the PTV [64]. The prescription iso-
Hospital in Toronto, for our present SBRT liver cancer stud- dose line between 70 and 80 % is reasonable if the rind of nor-
ies, a 5-mm margin around the GTV within the liver is used mal tissue inside the PTV is small or can tolerate this high dose.
to define the CTV. The planned minimum dose to the CTV Resultant high doses/hot spots occur in the center of the
PTV is 27 Gy in six fractions, while the dose to the GTV PTV, perhaps giving the highest dose to the center hypoxic
volume (although the potential benefit of this is unproven). 2 and 7 cm, the conformality index was required to be less
Of note, the use of many beams for SBRT is not mandatory. than 1.2, and the plan was considered as minor deviation
For a peripheral liver tumor, three to five beams may be used with index between 1.2 and 1.4 but there is no good data to
to reduce the overall radiation path length through the liver. suggest complications are higher with a poor conformality
The use of segments within the beams can adjust the dose index. As with similar data with intracranial SRS, target vol-
distribution to ensure that the hot spot is within the GTV ume and proximity to critical structures usually trump the
while reducing the overall integral dose. effect of conformality.
As the adverse effects of dosimetric errors are most pro- A sharp dose gradient out of PTV is desirable for SBRT
nounced with SBRT, appropriate methods to consider cor- plans. The dose gradient can be evaluated by how much
rections for heterogeneities should also be used (see next %/mm or ratio of volume receiving 50 % of prescription
section). dose to the PTV that was used in RTOG protocols of lung
Typical prescription doses for SBRT range from 5 Gy in SBRT [10]. Usually a sharp dose-off isotropically is pre-
ten fractions to 20 Gy in three fractions. One to five fractions ferred although this can be limited to less degrees of freedom
are most often used, with a dose per fraction usually greater with non-coplanar fields, but the dose gradient can be larger
than 6 Gy. The common feature to most of the SBRT frac- in one direction if close to a critical structure.
tionation schemes is that they are biologically potent. The dose heterogeneity is probably not as important as
Multiple-fraction regimens have some radiobiologic advan- the above two parameters, although it needs attention. To
tages to single-fraction SBRT. Clinical data is not available increase the plan conformality index and dose gradient usu-
to provide guidance for the most appropriate fractionation ally increase the dose heterogeneity. The dose hot spot
for each clinical scenario. However, when the PTV is in very should be within the PTV, and the best place is the center of
close proximity to normal tissues that function serially, it is the PTV.
reasonable to consider prolonging fractionation to minimize The dose constraints to critical organs are different with
the risk of toxicity to the serial function normal tissue. what are used in the conventional therapy. So far the dose
It is a challenge to relate the prescribed dose to TCP, as tolerance is still immature and the dose to organ at risk
inhomogeneous doses are generally used and various deliv- should be evaluated carefully according to the peer-reviewed
ery and verification techniques are utilized. For a moving publications or protocols.
target not treated with image guidance, the actual delivered
minimal dose to the tumor may be lower than the prescribed Treatment Planning System Considerations
dose. Furthermore, there is heterogeneity in patterns of pre- Related to Heterogeneity Corrections
scribing dose. One method to account for inhomogeneous The dose calculation heterogeneity correction is desirable
dose distributions is to use equivalent uniform dose (EUD) to for SBRT planning, especially for lung or spine SBRT. The
tumor for reporting [65]. Of course, accounting for individ- treatment planning system (TPS) should be carefully evalu-
ual patient motion in the dose distribution or eliminating ated to see if it is suitable for SBRT planning or specific site.
motion would help ensure that the reported doses better Pencil beam algorithms have been discouraged for tumor
reflect delivered doses as well. Unfortunately, it is not pos- sites surrounded by low-density tissue [59]. It has been
sible to account for motion in current commercially available reported that there was significant dose difference in the tar-
planning systems. get periphery between pencil beam algorithm and superposi-
tion convolution algorithm for the lung SBRT [59]. Although
Plan Evaluations the Monte Carlo simulation is the most accurate dose calcu-
All conventional 3D plan evaluation methods apply to SBRT lation algorithm, it may not be widely available for commer-
plan, but three criteria are specially considered: dose confor- cial TPS and take more time for dose calculation. The
mality, dose gradient, and dose heterogeneity. convolution superposition algorithm is more suitable consid-
Since SBRT delivers high dose (often >5 Gy per fraction) ering the accuracy and calculation time is rather similar to
in a small number of fractions (≤5), high-dose conformality Monte Carlo-generated plans.
is required to minimize the dose to the normal tissue. A fre- Early in the North American experience with lung cancer,
quently used conformality index is defined as the ratio of heterogeneity corrections were not used because tissue den-
volume receiving prescription dose to the PTV. A perfect sity correction dose algorithms were not widely available. In
plan should have conformality index equal to 1, i.e., PTV is the absence of heterogeneity corrections, tumors completely
covered by 100 % of prescription dose and no normal tissue surrounded by air experience a dose build-up effect due to
receives the prescription dose or above. This is not realistic. the loss of electronic equilibrium, such that the periphery of
Usually the conformality index of 1.3 is achievable. a tumor is under-dosed, and a higher dose is required to
The index could be larger for smaller target or irregular tar- achieve optimal cell kill [66]. Tumors that lack this air–tissue
get. In current or recently closed RTOG protocols of lung interface, such as tumors abutting the chest wall or central
SBRT [10], for the target with maximum dimension between tumors, may have artificially high doses (“hot spot”) delivered
to the PTV, due to dose being delivered through low-density marks for patient position are routinely used for initial patient
air using very long path lengths. Xiao et al. [67] applied het- setup in practice. For some clinical situations, the use of skin
erogeneity corrections to the original treatment plans of 20 marks to align the patient can be done with high precision.
patients from RTOG 0236. They observed an increase in iso- However, in most body tumors, the internal structures cannot
centric dose in all patients following the application of het- be accurately localized with the use of skin marks. It has
erogeneity corrections. However, the mean volume of PTV been reported the random setup error is up to 6.6 mm for
receiving 60 Gy decreased from 95 to 85 %, and dose to pancreatic cancer patients receiving SBRT based only on
95 % of the tumor decreased by an average of 4.7 Gy [67]. body marker. The use of bony anatomy with electronic portal
Based on these findings, the use of heterogeneity corrections imaging is another standard practice in radiation therapy.
would be associated with larger hot spot doses but worsened However, for many clinical situations, the bony anatomy is
overall tumor coverage, potentially leading to increased not well correlated with the internal tumor position. Options
complications due to the high-dose areas and higher poten- for locating internal anatomy include the use of implanted
tial for failure due to underdosing parts of the tumor. This led radiopaque fiducial markers as surrogates for the target, tis-
the investigators to conclude that the RTOG standard dose sues adjacent to the tumor, or the tumor itself. Fiducial mark-
should in fact be 18 Gy per fraction delivered to the 95 % ers may also be used to measure organ motion and/or track/
isodose, for a total dose of 54 Gy. Notably, these patients all gate the beam.
had peripherally located tumors. Many of the published reports of SBRT have been on
Investigators at the University of Maryland repeated this patients treated with conventional linear accelerators.
work on patients with peripherally and centrally located lung However, more specialized treatment units are now available
tumors. This study shows that dose to the internal target vol- with the potential to allow soft tissue image guidance and
ume (ITV) increases with the application of heterogeneity reduced PTV margins. Examples of such systems include a
corrections and the effect was even greater with centrally lightweight and robotic linear accelerator (CyberKnife;
located tumors and larger. In centrally located tumors, where Accuray, Sunnyvale, CA) and modified linear accelerators to
more lung parenchyma is located between tumor and body allow image guidance including accelerators such as Novalis
surface, intervening tissue density is overestimated. When (BrainLAB, Inc., Westchester, IL), Synergy (Elekta
this is corrected, dose to the tumor increases, an effect Oncology, Stockholm, Sweden), TrueBeam and Trilogy
observed in this study. This too is true of larger tumors, where (Varian Medical Systems, Palo Alto, CA), Artiste (Siemens,
increased beam widths traverse more parenchyma prior to Concord, CA), and TomoTherapy (Madison, WI).
reaching the PTV. Given the data above, the inclusion of cen-
tral tumors in this sample may in part explain why the appli-
cation of heterogeneity corrections led to an increase in D95 Image-Guidance Strategies (Pretreatment)
in this study, versus a drop in D95 as observed in the paper by
Xiao et al. [67], in which central tumors were excluded. Imaging at the time of treatment can be used for localization
Without heterogeneity corrections applied to treatment for guidance, verification, and also as a quality assurance
plans, a higher dose must be prescribed to achieve adequate (QA) tool that is evaluated by the treating physician prior to
dose to the tumor, which also results in a higher dose to the sur- each treatment. Verification that the appropriate dose is actu-
rounding parenchyma. This increases the risk of adverse out- ally delivered is also important for SBRT. Soft tissue imag-
comes, which may help to explain the increased complications ing at the time of treatment can allow a measurement of the
seen in the initial Indiana experience. Based on these results, impact of geometric uncertainties (such as organ deforma-
the use of heterogeneity corrections allows the prescription tion) at the time of treatment. When image guidance and
of a lower total dose and dose per fraction, with a comparable repositioning are used, imaging after repositioning should be
dose to the tumor. This less aggressive dose prescription may used to ensure the positioning moves were made in the cor-
allow safer yet equally efficacious treatment of NSCLCs. rect direction.
When repositioning moves are required due to changes in
internal organ position, replanning is not routinely con-
Image Guidance ducted. However, when substantial changes in organ position
and or tumor size or breathing pattern occur, the dose deliv-
Overview ered may be altered due to changes in radiation path lengths
and position of organs with difference heterogeneities. The
Image guidance at the time of treatment can improve setup magnitude of dosimetric differences due to such changes
accuracy, reduce PTV margins, and irradiate the volume of should be studied to better define in which situations replan-
normal tissue, facilitating safe dose escalation. ning may be of benefit. In addition, when soft alignments are
Traditionally, surrogates for the target have been used in utilized, the treating physician should also ensure the volume
guiding the placement of treatment fields. For example, skin visualized on the volumetric imaging is similar to the ITV
generated on the 4D CT as differences may be related to has more priority during the treatment. The pitfall of this
different breathing patterns from day to day and may not strategy is that it may introduce the large setup uncertainty
improve alignments. for target which has respiratory-induced motion. Currently,
the CT scanner is fast enough to image the moving target on
one breathing phase. The CBCT is slow and the moving tar-
Two-Dimensional Image-Guidance Equipment get is more similar to internal moving target (ITV) on CBCT
images. The in-room CT is actually a CT scanner and the
Orthogonal megavoltage (MV) portal films and more moving target similar to one breathing phase. If the patient is
recently images from electronic portal imaging devices aligned based on the target between planning CT and CBCT/
(EPIDs) have traditionally been used for image guidance and in-room CT, the perfect target match may introduce the
may be appropriate for targets adherent to the bones. MV errors due to the difference in breathing phases between
images of large fields of view can be obtained with 2–8 mon- planning CT and CBCT/in-room CT. The correct strategy is
itor units (MUs) per image. These images not only can guide to align the patient based on the bony structure and make
therapy but also can verify the shape and orientations of the sure that the target acquired through CBCT/in-room CT is
treatment fields. If radiopaque fiducial markers are inserted within the PTV on the planning CT.
in or near the tumor, the fiducial markers themselves may be
used for guidance. Other alternatives for guidance include In-Room CT
using surrogates that are in close proximity to the tumor, for The placement of a diagnostic CT scanner in the treatment
example, the diaphragm as a surrogate for liver tumors room with a known geometric relationship to the linear
[68, 69]. An example of an anterior–posterior (AP) MV accelerator is one approach for volumetric imaging imme-
image for liver cancer guidance, in which the diaphragm is diately prior to treatment with the patient in their immobi-
used for cranial–caudal positioning. lization device. Uematsu et al. have used this approach to
Due to the low contrast of MV radiographs and the doses treat liver and lung cancers with SBRT. Uematsu et al.
delivered with repeat MV imaging, orthogonal kV radio- reported that with repeat CT for frameless head and neck
graphs and kilovoltage (kV) fluoroscopy have also been used cancer radiotherapy, there was a geometric vector error
for image guidance of tumors and/or fiducial markers, either ranging from 0.1 to 0.9 mm [70]. Multiple manufacturers
immediately prior to each radiation fraction [68, 69] or have developed products of this type, including Siemens’
throughout radiation delivery [58]; kV X-ray tubes may be Primatrom, Mitsubishi’s accelerator in combination with a
ceiling or wall mounted or attached to the linear accelerator. General Electric CT scanner, and Varian’s ExaCT target-
With both MV and kV orthogonal imaging, alignment ing system [71–73]. All systems place the CT scanner in
tools registering the images to digitally reconstructed radio- close proximity to the linear accelerator, allowing the
graphs (DRRs) can improve the accuracy and efficiency of couch to be moved from the imaging position to the treat-
image matching to determine the offsets in position. Such ment position. The CT scanner gantry is often translated
alignments tools include template matching tools based on during acquisition to minimize couch motion. Accuracy
therapists’ visualization of anatomy and/or automated image has been reported to be under 0.5 mm [72], and it has been
registration of the region of interest (e.g., mutual informa- reported to be improved with fiducial markers from 0.7 to
tion). Decision rules including tolerance levels for reposi- 0.4 mm [71].
tioning must be integrated with the overall system. Advantages of in-room CT include that state of the art,
diagnostic-quality CT can be used for optimal image qual-
ity and robustness. Disadvantages of this system are that
Three-Dimensional Volumetric Image the imaging and treatment isocenters are not coincident.
Guidance Accuracy of motion from the CT scanner gantry, the
accelerator couch, and the coincidence of the CT and lin-
Technological advances allowing volumetric imaging allow ear accelerator isocenters needs to be verified. Limiting
image guidance immediately prior to treatment using the patient movement between imaging and treatment (e.g.,
tumor or a soft tissue organ in close proximity to the tumor couch retraction <1 mm) should improve setup accuracy;
for guidance, rather than the bony anatomy. Advantages of however, organ motion between imaging and delivery
volumetric imaging systems include that adjacent normal may occur.
organs can also be visualized for more accurate avoidance of
critical structures. Some of these volumetric imaging tech- KV Cone Beam CT
niques can also measure tumor motion due to breathing. Jaffray et al. [74] first described the concept of cone beam CT
Since the soft tissue can be visualized with three- (CBCT) for image-guided radiation therapy in 1997. CBCT
dimensional volumetric images, it is possible to set up the refers to combined kV X-ray imaging and MV radiation deliv-
patients based on the tumor target or organ at risk if which ery in one integrated gantry-mounted system. Advancements
Fig. 13.2 Planning CT and kV

cone beam CT from each lung
cancer SBRT fraction. GTV from
the planning CT scan is shown in
orange and PTV in green
in large-area flat panel detector technology facilitated MV Cone Beam CT

volumetric imaging to be acquired in a single rotation of the CT imaging using MV beams has also been explored for
linear accelerator gantry. Planar kV images projections are more than 20 years [76, 77] and has also been made possible
obtained as the gantry rotates about the patient on the linear with advances in portal imaging technology. Advantages of
accelerator table, over 30 s to 4 min. CBCT three-dimensional MV cone beam CT are that the treatment MV beam is used
volume reconstruction images may then be obtained for posi- to obtain the imaging, requiring less modification to the lin-
tion verification or for image guidance (see Fig. 13.2). ear accelerator; the electron density estimates for treatment
Geometric calibration methodologies for CBCT systems [75] planning are accurate; and there is no high Z artifact that is
and quality assurance recommendations [74] have recently associated with kV imaging.
been proposed. MV cone beam CT has been used to aid in lung cancer
Doses delivered to obtain CBCT scans typically range SBRT, as described by Nakagawa et al. in 2000. MV CT-aided
from 0.5 to 2 cGy. Two vendors have developed kV CBCT lung SBRT was used for treatment of 22 lung tumors [78].
systems: Elekta Synergy and Varian Trilogy and TrueBeam. Pouliot et al. recently reported the feasibility of acquiring
In addition to providing volumetric imaging for verification low-exposure megavoltage CBCT. Phantom and pig cadaver
and guidance, these systems have the ability to be used for head and neck images were acquired using a linear accelera-
real-time kV tracking; the latter application has not been tor dose rate of 0.01–0.08 MU per image, for a set of 90–180
used clinically. projections, acquired in 1–2° increments over 45–60 s. MV
Similar to in-room CT, artifacts may occur with kV CBCT cone beam CT scans were obtained with doses of approxi-
reconstructions due to high Z structures such as surgical mately 5 cGy. Despite the low efficiency of this system, vis-
clips, hip prostheses, and dental fillings. Methods to reduce ibility of high-contrast structures, such as air and bone, was
these artifacts have been developed. reasonable [79].
MV TomoTherapy Disadvantages of this system include the need for fiducial

MV TomoTherapy combines tomographic scanning capabili- markers, long potential delivery times (up to 90 min), lack of
ties, from a conventional CT detector, with a linear accelera- suitability for large tumors with motion more than 10 mm,
tor mounted on a rotating gantry. Simpson described the and highly inhomogeneous dose distributions.
initial development of an MV CT scanner for radiation ther- Another system (Novalis: BrainLAB, Heimstetten,
apy in 1982 [76]. More recently, the TomoTherapy treatment Germany) also acquires kV orthogonal images and matches
platform has become available for image guidance and veri- the images to DRRs obtained from the planning CT. The
fication. The MV treatment beam is used to obtain imaging, imaging axes are not coincident with isocenter, and a transla-
with a lower energy, 3.5 MV instead of 6 MV. Computer- tion of patient position is required between imaging and
controlled multileaf collimators, also on the rotating gantry, treatment. Accuracy of this system has been reported to be
have two sets of leaves that open and close to modulate the within 3 % and 3-mm distance to agreement.
radiation beam while the couch advances the patient through Recently, wireless transponders and infrared cameras to
the gantry bore, for helical intensity-modulated radiation track tumors have also been proposed as an imageless local-
therapy (IMRT). ization system (Calypso Medical Technologies, Seattle, WA).
Similar to MV cone beam CT, there are no high Z artifacts
with MV TomoTherapy. SBRT Patient Safety
Since SBRT gives very high dose and uses much smaller
planning margin than conventional therapy, more caution
Image-Guidance Strategies (During Treatment) must be given to avoid catastrophic dose delivery. The patient
safety has got more and more attention in radiation therapy
Real-Time Tumor Tracking society. Recently, the ASTRO SRS/SBRT white paper was
Real-time tumor tracking while the radiation beam is on is published to address SRS/SBRT quality and safety consider-
another approach to reduce adverse effects of organ motion. ations. This chapter presented the personnel requirement and
An elegant highly integrated tracking system consisting of training for radiation oncologist, medical physicist, medical
four ceiling-mounted fluoroscopic X-ray tubes and four dosimetrist, radiation therapist, and administrator; presented
floor-mounted flat panel imagers in the treatment room the technology requirements of simulation, planning, and
allowing visualization of radiopaque markers in tumors was localization; and addressed the SRS/SBRT system accep-
first described by Shirato et al. [58]. This system has a tem- tance and commissioning.
poral resolution of 30 frames per second and a precision of The report of AAPM TG101 also has recommendations
1.5 mm. The linear accelerator is triggered to irradiate only for SBRT patient safety issue. It recommends that at least
when the fiducial marker is located within a predefined vol- one qualified medical physicist presents during first fraction
ume. This system has been used for image-guided radiation of SBRT treatment and the qualified medical physicist should
therapy of lung, liver, and paraspinal malignancies. be available for the following fractions. It also recommends
As an alternative to turning the radiation beam off when that the radiation oncologist should approve the image guid-
the tumor moves outside the treatment region, multileaf col- ance before each fraction treatment. The therapist should be
limators, the couch position, or the entire accelerator on a well trained for the SBRT procedure.
robotic arm may move with the tumor to ensure adequate
tumor coverage (e.g., CyberKnife image-guided radiosurgery Equipment Quality Assurance
system). The latter lightweight (330 lb) linear accelerator The equipment of SBRT mainly include the linac system and
mounted on a robotic arm uses 6 MV, 5- to 60-mm collima- imaging system. The QA frequency is daily, monthly, and
tors, and a dose rate of 300–400 MU per minute, combined annually. Usually the radiation therapist performs the daily
with dual orthogonal fluoroscopy tubes to track radiopaque QA and the qualified physicist reviews and approves it, and
markers in or near the tumor at a preset frequency. When the the qualified physicist performs the monthly and annual
beam is on, infrared external surrogates are continuously QA. Apparently SBRT linac has tighter tolerance than the
monitored, while the internal anatomy is monitored every linac only for conventional therapy. The report of AAPM
few seconds with kV imaging. The external surrogates are Task Group 142 [80] published recommendations for SRS/
used for determining the breathing model, and the model is SBRT linac QA. The main recommendations about accuracy
updated based on the X-ray data obtained every few seconds. tolerance are summarized here. It recommends that the laser,
The robotic linear accelerator responds to motion by moving the couch movement indictor, collimator size indicator, col-
to an appropriate position, within a range of ±10 mm x, y, z limator rotation isocenter, couch rotation isocenter, gantry
and ±1° pitch and roll, ±3° yaw. This system was found to rotation isocenter, and coincidence of radiation and mechan-
have a 0.3-mm accuracy when tested in phantom studies. ical isocenter have to be within 1 mm. The couch rotation
indictor has to be within 0.5°. The imaging system includes related to dose and dose per fraction. Stephans et al. [84]
kV/MV imaging and kV/MV cone beam CT. It recommends reported the experience at Cleveland Clinic, which clearly
that the imaging and treatment coordinate coincidence illustrates the effects of the higher biologic effective dose
should be within 1 mm. For cone beam CT, the geometric (BED) on morbidity. Their institutional practice of SBRT
distortion should be within 1 mm, and the contrast, spatial initially used 10 Gy × 5 fractions and subsequently changed
resolution, HU constancy, uniformity, and noise should agree to the RTOG standard (18–20 Gy × 3) when RTOG 0236’s
with the baseline. results were initially reported. The local control rates showed
no difference between the two schemas, although they did
experience a high rate of chest wall morbidity with the higher
Clinical Outcomes BED regimen (18 % versus 4 %, p = 0.028). In addition, fol-
low-up from the initial Indiana experience treating central
The next several sections will review the most common clini- tumors reported lower grade 3 and higher side effects when
cal sites where SBRT/SABR is delivered most commonly >3 fractions were used as opposed to three fraction (26.3 %
where pertinent issues related to patient selection, simula- versus 7.7 %) [85]. However, caution is still indicated.
tion, treatment planning, and morbidity specific to that clini- Washington University [86] has completed a phase I study
cal site. This data should be used with caution when demonstrating 60 Gy in five fractions is safe for treating cen-
extrapolating to other locations (i.e., primary liver versus tral tumors and is currently accruing to a phase II study.
metastatic liver since the background normal tissue is already Many retrospective series also suggest a five fraction is safe
partially injured) but will be good starting point prior to in treating centrally located tumors [87]. In addition, a multi-
treating a unique site. institutional phase I/II [88] RTOG study is currently nearing
completion using a similar fractionation schedule as the
Washington University. Led by Videtic et al. [88] from the
Stereotactic Ablative Radiotherapy for Lung Cleveland Clinic, the RTOG has currently completed testing
Tumors two less aggressive schemas (12 Gy × 4 fractions and
34 Gy × 1 fraction) in a randomized phase II trial, specifically
Lung cancer is the second most common cancer diagnosis looking at morbidity at 1 year as the primary end point as
and the most common cause of cancer death for both gen- local control is expected to be 90 % in both arms (RTOG
ders. In 2012, there was an estimated 226,160 new lung can- 0813). The less aggressive regimen will then be compared to
cer diagnoses in the United States and 160,340 deaths [81]. the current RTOG standard which is a subsequent phase III
Of the 15 % of patients diagnosed with early-stage lymph trial. There is no lower limit of lung function based on pul-
node-negative NSCLC, the 5-year relative survival is only monary function tests that predicts poor outcomes [89]. One
52 %. The current standard of care for these patients is sur- clinical entity that may have a higher risk of pulmonary
gery [82]. When a patient cannot tolerate or refuses surgery, injury is patients with interstitial pulmonary fibrosis. Onishi
radiation therapy is an excellent alternative with current lit- et al. [90] described 24 patients who died of fatal pneumoni-
erature suggesting outcomes following SBRT/SABR similar tis in the large multi-institutional experience. Seventeen of
to surgery. For patients with severe comorbidities, it is the 24 patient had IPF with 1 of the patients with usual inter-
important to understand what the natural history of early- stitial pneumonia. The absolute risk of fatal pneumonitis for
stage disease is. Approximately 50 % of patients require this cohort of patients in Japan is not known, but from
treatment for symptoms at 1 year and on average 50 % die of Washington University described a 20 % risk of pneumonitis
lung cancer at 2 years without treatment [83]. in patients with IPF. So a diagnosis of IPF is not a contrain-
dication for SBRT but does appear to be at a higher risk of
Patient Selection pneumonitis.
This approach is widely accepted as the current standard of
care for peripheral T1 or T2 tumors or certain T3 tumors Treatment Planning
(i.e., invading the chest wall) which measure less than During the CT simulation, 3-mm slice thickness or less is
5–7 cm without any nodal involvement. Central lesions may preferred. IV contrast may be helpful for tumors close to the
have a higher complication rate in comparison to peripheral brachial plexus or for central tumors adjacent to vasculature.
tumors based on the initial phase II experience published out The gross tumor volume (GTV) is contoured on pulmonary
of Indiana [8]. In their single institution experience, “exces- windows, although soft tissue windows are helpful for
sive toxicity” was reported for patients with tumors located tumors adjacent to the chest wall and central lesions.
within 2 cm of the proximal respiratory tree where the rate of Contouring all the spiculations on lung windows is contro-
grade 3 and above toxicity was significantly higher among versial and not uniformly performed. At the University of
patients with central tumors. This increase in toxicity may be Maryland, only the bulky component of the spiculation is
contoured (2-mm thickness) based on unpublished literature is primary based on our medically inoperable patients who are
suggesting that it is uncommon that these radiographic find- likely to have emphysema and poorly functioning diaphragms
ings correlate with clinically significant disease and is cov- [94]. Alignment using soft tissue can be performed if imaging
ered with subclinical doses in the area of dose falloff [61]. acquired on cone beam imaging is similar to that generated by
There are institutional preferences regarding the use of a the 4D CT-based ITV. If the volumes do not match, then the
margin for microscopic extension and creation of a CTV. In treating physician needs to consider that the patient’s breathing
RTOG 0236, it was specified that the GTV was equivalent to pattern may be different from the day of the initial 4D CT. This
the CTV. Assessment of respiratory motion and creation of difference may not be reflected accurately on the CBCT and
an ITV is preferred to limit the PTV margins similar to SRS alignment based on soft tissue may therefore be inaccurate
in the brain. The ITV can be assessed with a 4D CT scan or potentially leading to a geographic miss. In this situation, a
with fluoroscopy to track the tumor or a fiducial implanted bony alignment is preferable and treatment can be delivered if
into the tumor. A margin of 0.5 cm to create the PTV from the volume is safely within the PTV on the CBCT. If it is out-
the ITV is common. If an ITV is not created, population- side the volume, a repeat 4D CT is preferred.
based margins of 0.5 cm axially and 1.0 cm in the cranio- At the University of Maryland, consideration of the accu-
caudal directions were used historically in the Indiana racy of cone beam imaging is considered to be 2 mm such
experience and RTOG 0236. However, the use of that a shift will be performed only if the shift is 3 mm or more
patient-specific tumor motion is preferred. This can be per- based on the work from Princess Margaret Hospital [95].
formed by fusing maximum inspiration and expiration
breath-hold scans to a free breathing CT, obtain a “slow” CT Fractionation
simulation, if 4D CT is not available. 4D CT has been dem- There are two predominant fractionation schedules used in
onstrated to result in smaller PTV volumes than any of the the United States: the Japanese approach (12 Gy × 4) [96]
other described options [91]. If there is significant respira- and the original Indiana approach developed by Dr.
tory motion, some have advocated for managing this with Timmerman (18–20 Gy × 3) [6, 8]. Outcomes and dose–
abdominal compression or breath-hold maneuvers. The ben- response will be discussed under clinical outcomes. In the
efit of abdominal compression is likely to be greatest for original Indiana series that led to RTOG 0236, dose is calcu-
lower lobe tumors. However, abdominal compression may lated without heterogeneity corrections. When these are
increase tumor motion in some cases and should be consid- applied, the dose is on average 10 % lower for the subset of
ered on an individual basis and there are several studies patients treated on the study [22]. The prescription isodose
which demonstrate increased failure with the use of abdomi- line covering the PTV is usually 60–90 % depending on the
nal compression and gating [92]. At the University of block edge margin (2–3 mm, respectively) used on each
Maryland, a CTV margin of 3 mm is added based on the beam’s eye view. Typically a minimum of 7–9 non-coplanar,
unpublished work from Fox Chase Cancer Center which non-opposing beams are required to obtain adequate dose
evaluated 25 consecutive patients who underwent a lobec- falloff to minimize risk of morbidity (see Fig. 13.3). Plans
tomy or sublobar resection for a curative procedure for are assessed by dose conformality similar to strategies uti-
tumors <3 cm in maximum dimension and also had a diag- lized in the brain where the primary goal is to devise the most
nostic CT scan for review. Based on a comparison of CT and conformal plan to the PTV with the secondary goal to mini-
microscopic pathology review, the entire microscopic extent mize the volume receiving 50 % of the prescription dose.
of disease was covered when 3 mm was added to the maxi- The currently accruing RTOG protocols recommend that
mum extent of the disease contoured on standard lung win- 95 % of the PTV be covered by the prescription isodose and
dows without chasing the spiculations in 95 % of the cases. 99 % of the PTV receives a minimum of 90 % of the pre-
An additional 3 mm was added for setup error based on scription dose. A prescription isodose to PTV volume (PITV)
obtaining a daily CBCT prior to each treatment. ratio of 1.2 or less is recommended, except for tumors
Interfraction setup variation is reduced with image-guided smaller than 1.5 cm as this constraint is difficult to meet.
radiotherapy (IGRT). This can be performed by matching fidu- Additional RTOG recommendations are in place to limit tox-
cial location, which is vital to technologies which do not use icity to normal structures. This includes no more than 15 %
volumetric IGRT, and/or creating a CBCT with the patient of the PTV volume of tissues outside the PTV receiving
immobilized. These pretreatment images are compared to the more than 105 % of the prescription dose. The so-called
simulation images and adjusted as appropriate. Fusion of the intermediate dose spillage is limited by imposing criteria on
CBCT with the simulation CT using bony anatomy is recom- the ratios of the 50 % prescription isodose volume to the
mended, as there is minimal change in the target centroid posi- PTV volume and the maximal dose at 2 cm from the PTV,
tion relative to the skeletal frame [93]. The differences in which are dependent on PTV volume. This uniform dose
patient position based on matching with bony anatomy versus falloff may be important especially in the parenchyma of the
soft tissue can be off by upwards of 8 mm on average but this lung where subclinical disease may be present.
Fig. 13.3 SBRT dose distribution for lung cancer
Clinical Outcomes 500 patients, a BED of 105 or greater predicted was

SBRT/SABR has been most widely studied in primary and confirmed to predict for an excellent local control with no
metastatic lung cancer. A summary of clinical data is shown clear benefit from further dose escalation [99]. Regional
in Tables 13.1 and 13.2 and some of the major studies are nodal failures and distant metastases are more common for
discussed in greater detail below. central tumors and tumors >3 cm.
Initial reports from the Karolinska Institute and Japan, The North American experience was pioneered by
who first reported the use of extracranial stereotactic radio- Dr. Timmerman at Indiana University where a series of pro-
therapy included patients with lung cancer and lung metasta- spective protocols were conducted which led to the develop-
ses. In 2004, Onishi et al. reported a large multi-institutional ment of all of the RTOG SBRT trials. His initial phase I dose
experience from 13 institutions in Japan of nearly 250 escalation study included 37 medically inoperable patients
patients that had a major impact on clinicians taking care of with Stage I NSCLCs (<7 cm). The dose was escalated from
lung cancer patients. Using a variety of techniques and frac- 24 Gy in three fractions in 2 Gy/fraction intervals to 60 Gy
tionation scheme, local control was found to be 86 % at 5 and higher in three fractions. The maximum tolerated dose
years when adequate dose was delivered as compared to was not reached and 87 % of patients had tumor response,
50 %. In addition, the side effect profile was remarkably low. including 27 % with a complete response [6] and a dose–
Onishi suggested that the dose–response was based on deliv- response curve where a local control of >90 % was achieved
ering a BED greater than 100 Gy [25, 97]. This fraction may when 18 Gy or higher fraction was used. A phase II study
be inadequate for tumors >3 cm and for lung metastases was conducted using 60 Gy in three fractions for T1 tumors
where investigators advocate 12 Gy × 5 or 18–20 Gy × 3 and 66 Gy in three fractions for T2. Local control at 2 years
[98, 99]. A BED of greater than 100 was an important factor was 95 % and overall survival was 54.7 % at 2 years. Toxicity
for local control even for centrally treated lung tumors [100]. was significantly greater for those treated with central tumors
In a large pooled analysis from the Elekta consortium of over [8] and was subsequently excluded from RTOG 0236.
Table 13.1 Selected results of primary non-small cell lung cancer treated with SABR
Median Regional Distant
Reference (year Patient population Total dose (no. follow-up Local control metastasis Overall
published) No. patients (all N0 M0) fractions) (months) control rate (%) rate (%) survival
Uematsu (2001) [23] 50 18 pts received 50–60 Gy (5–6) 36 94 % 96 14 66 % at 3
40–60 Gy prior to years
SABR
Wulf (2004) [30] 20 10 % T1 30–37.5 Gy (3) 11 95 % NR 25 32 % at 3
50 % T2 years
40 % T3
Nagata (2005) [19] 45 71 % T1 48 Gy (4) 30 98 % 93 20 At 3 years:
29 % T2 83 % for T1
and 72 %
for T2
Timmerman (2006) [20] 70 50 % T1 60 Gy (3) 17.5 95 % at 2 100 10 55 % at 2
50 % T2 years years
Yoon (2006) [34] 21 62 % T1 30–48 Gy (3–4) 13 86 % 100 5 51 % at 2
38 % T2 years
Onishi (2007) [24] [25] 257 Multi-institutional 18–75 Gy (1–22) 38 86 % 89 20 47 % at 5
review years
63 % T1
37 % T2
Hof (2007) [32] 42 40 % T1 19–30 Gy (1) 15 95 % at 1 95 31 65 % at 2
50 % T2 year; 68 % years
10 % T2 at 2 years
Fakiris (2009) [35] 70 49 % T1 60–66 Gy (3) 50 88 % at 3 91 13 43 % at 3
51 % T2 years years
Baumann (2009) [36] 57 70 % T1 45 Gy (3) 35 93 % 95 16 60 %
30 % T2
Timmerman (2010) [29] 55 80 % T1 60 Gy (3) 34 91 % 87 22 48 % at 3
20 % T2 years
Ricardi (2010) [37] 62 69 % T1 45 Gy (3) 28 88 % at 3 94 24 57 % at 3
31 % T2 years years
Bral (2011) [28] 40 65 % T1 Central: 45 Gy (3) 16 92 % 95 15 52 % at 2
35 % T2 Peripheral: 60 Gy (3) years
Staging is AJCC sixth edition
NR not reported
Timmerman et al. reported the multi-institutional phase II Normal tissue constraints used in RTOG 0618 and 0915 are
experience including 55 medically inoperable, NSCLC shown in Table 13.3, although these values are more conser-
patients with T1 (80 %) and T2 (20 %) N0 M0 tumors less vative than the constraints employed by the Japanese. Dose–
than 5 cm treated to 60 Gy in three fractions. After a median volume analyses confirming and disputing these values are
follow-up of 34 months, the tumor control rate was 97.6 %. described below.
At 3 years, the involved lobe (local) control rate was 90.6 %, Pneumonitis is the dose-limiting complication after con-
the local regional control rate was 87.2 %, and the distant ventional radiation therapy for lung cancer. It is a clinical
metastasis rate was 22.1 %. Median survival was 4 years diagnosis of exclusion, consisting of cough, fever, pleuritic
[10]. This work has now led to subsequent studies at the pain, and dyspnea, with occasional respiratory failure and
RTOG (0618 and 1014) which are now challenging surgery general inflammation and can be extremely difficult to dif-
as the standard of care for early-stage lung cancer. ferentiate between a COPD exacerbation, an atypical pneu-
monia, and a viral infection. Characteristic radiographic
Complications findings are often present, typically in the distribution of the
Despite the majority of the patients on these studies being radiation fields. A prolonged course of prednisone is the
medically inoperable with significant comorbidities, there typical treatment using 50–60 mg/day initially followed by a
are relatively few grade 3–5 toxicities related to radiotherapy. slow taper. Based on the low incidence, arbitrary constraints
Table 13.2 Selected results of lung metastases treated with SABR

Reference (year No. patients Total dose (no. Median follow-up
published) (no. lesions) Patient population fractions) (months) Local control Overall survival
Wulf (2004) [30] 41 (51) Primary site: 30–37.5 Gy (3) 9 90 % 33 % at 2 years
45 % lung or 26 Gy (1)
10 % breast
8 % colorectal
8 % kidney
8 % sarcoma
Yoon (2006) [34] 53 (63) Primary site: 30–48 Gy (3–4) 14 82 % NR
26 % lung
23 % liver
19 % colorectal
8 % esophageal
Hof (2007) [38] 61 (71) Primary site: 12–30 Gy (1) 14 89 % at 1 year 65 % at 2 years
51 % lung
30 % other histologies
12 % colorectal
Milano (2008) [39] 121 (293) Primary site: Most 50 Gy (5) 41 67 % at 2 years 50 % at 2 years
32 % breast
26 % colorectal
Norihisa (2008) [40] 34 (43) Primary site: 48–60 Gy (4) 27 91 % (crude) 84 % at 2 years
44 % lung
26 % colorectal
Rusthoven (2009) [41] 38 (50) 1–3 lung mets, < 7 cm 48–60 Gy (3) 15 96 % at 2 years Median 19 months
Ricardi (2012) [42] 61; 74 % Primary site: 36–45 Gy (3–4) 20 89 % at 2 years 67 % at 2 years
single nodule 46 % lung 1 pt 26 Gy (1)
21 % colorectal
NR not reported
Table 13.3 SABR lung normal tissue maximal point dose constraints receive 11.6 Gy in four fractions and the maximum 1.0 L
for three and four fraction regimens based on RTOG 0618 and 0915, should receive 12.4 Gy. Barriger et al. [101] reported the
respectively
dose–volume analysis from the Indiana experience which
Three included 273 patients treated with SBRT. The overall grade 2
fractions Four fractions
(20 Gy × 3) (12 × 4 Gy)
and higher pneumonitis rate was seen in 9.4 % of their
Spinal cord 18 Gy (6 Gy 26 Gy (6.5 Gy
patients with the majority (7 %) graded as 2. The most
per fraction) per fraction) important predictor of grade 2 and higher pneumonitis was
Esophagus 27 Gy (9 Gy 30 Gy (7.5 Gy the mean lung dose (p = 0.02) and V20 (p = 0.03). When com-
per fraction) per fraction) paring the RTOG cut point in V20 of 10 %, the investigators
Brachial plexus 24 Gy (8 Gy 27.2 Gy (6.8 Gy found that 10 % did not predict an increase in side effects
per fraction) per fraction)
(p = 0.42), but when the median value of 4 % was used, this
Heart/pericardium 30 Gy (10 Gy 34 Gy (8.5 Gy
per fraction) per fraction)
increased the risk fourfold from 4 to 16 % (p = 0.03).
Trachea/ipsilateral 30 Gy (10 Gy 34.8 Gy (8.7 Gy Although their most recent update suggests a V20 cut point
bronchus per fraction) per fraction) of 6.5 % continues to have a very low risk of pneumonitis.
Skin 24 Gy (8 Gy 36 Gy (9 Gy per The majority of these patients were treated with a three-fraction
per fraction) fraction) approach. When using less aggressive schemas, no grade 3
or higher pulmonary events were reported by Nagata et al. in
treating patients with 48 Gy in four fractions [96]. Onishi
were developed. The RTOG used the percent of lung receiv- et al. reported a 5.4 % incidence of grade 2 and higher pneu-
ing 20-Gy total or more (V20) be less than 10 % in RTOG monitis with no grade 5 events [97, 102]. Other, potentially
0236 and 0618. RTOG 0915 which uses a less aggressive permanent pulmonary toxicities after SABR include bron-
regimen used absolute volumes as opposed to percentages of chial damage that can be associated with total obliteration
the total where more than 1.5 L of the lung volume should of the airway lumen and downstream atelectasis. Tracheal
stricture and/or necrosis with subsequent lung collapse or tripled in the last several decades from 3 % in 1975–1977 to
injury is another toxicity that is possible after highly potent 15 % in 2001–2007 (p < 0.05), most patients will die of their
SABR to medial lung cancers is rare. disease [81]. With approximately 748,300 new cases and
Chest wall pain and rib fracture have been reported in 695,900 cancer deaths globally in 2008, primary liver tumors
7–15 % of SBRT/SABR patients. Taremi et al. created a were the second most frequent cause of cancer death in men
nomogram using age, female gender, and dose to 0.5 cc of and the sixth in women [81]. The standard of care for HCC
the ribs to estimate the risk of bone injury following is surgical resection. For patients with cirrhosis, resection
SABR. In this study, 0.5-cc ribs receiving 60 Gy was associ- alone is problematic since the precipitating etiology is still
ated with a 50 % risk of fracture [95]. Dunlap et al. found present and subsequently orthotopic liver transplant (OLT)
that the median interval to onset of severe chest wall pain appears to have a superior outcome for select patients with
and/or fracture following lung SABR was 7 months and the reported 10-year survival rates of >70 % as it resects the
best correlate was the volume receiving 30 Gy. The authors tumor and corrects the underlying etiology [107, 108].
recommended limiting the chest wall volume receiving Unfortunately, only 30–40 % of HCC patients may be candi-
30 Gy in 3–5 fractions to less than 30 cc, but they similarly dates for OLT or surgical resection [109]. If untreated, median
saw a high risk for the volume receiving >60 Gy suggesting survival from HCC is 6–12 months [110]. Similarly, select
fractionation may play a critical role [103]. Stephans et al. patients with metastatic disease to the liver from CRC appear
[104] reported the dose–volume analysis from Cleveland to have better outcomes with complete surgical resection of
Clinic for their patients who received 60 Gy in three frac- all liver disease. These patients have a 5-year survival of
tions without the use of heterogeneity corrections that the 30–40 % [111]. Unfortunately, only 10–20 % of patients are
most important predictors of chest wall pain in the volume of resection candidates [112]. For individuals with unresected
chest all receiving 30 and 60 Gy should be kept below 25 and CRC, liver metastases have median survivals of 10–17
3 cc, respectively. Welsh et al. [105] reported the MD months which can improve to 20 months with systemic ther-
Anderson experience using a less aggressive schema (50 Gy apy [113–115]. For unresectable primary liver cancers, the
in four fractions) suggesting the volume receiving 30 Gy is use of radiofrequency ablation (RFA), radioembolization
also the most important predictor of grade 2 and higher pain, with yttrium-90 microspheres, cryotherapy, percutaneous
but the absolute volume cut point is higher (50–100 cc). ethanol or acetic acid injection, laser-induced thermal ther-
Brachial plexus and phrenic nerve injuries are extremely apy, and high-intensity focused ultrasound have been studied
rare. Indiana University experience described seven brachial [110, 116, 117]. SABR offers a completely noninvasive
plexus injuries, of which four were grade 2, one grade 3, and approach to treat primary and metastatic liver tumors not
one grade 4. When the dose was greater than the median amenable to resection.
dose (26 Gy over three fractions), the 2-year incidence of
brachial plexus injury was 46 % versus 8 % (p = 0.038) for Treatment Planning
lower doses [85]. Planning SABR for liver tumors is more difficult for several
reasons. IV contrast is vital for accurate tumor delineation
and timing of administration is vital. In addition, as these
Stereotactic Radiotherapy for Metastatic tumors near the diaphragm and in nonsmokers commonly
Tumors of the Spine motion tumor motion can be significant. Therefore, trying to
deliver contrast during a 4D CT can be extremely difficult to
For information on stereotactic radiotherapy for metastatic get an accurate tumor delineation and an accurate estimate of
spinal tumors, please see Chap. 45. motion. As the liver has a dual blood supply, tumors can be
fed either from the portal system or from the hepatic artery
and depending on the tumor type tumor delineation can vary.
SBRT/SABR for Liver Tumors HCC is typically hypervascular and fed from the arterial sys-
tem and subsequently the arterial phase imaging with CT or
SABR can be used to treat tumors originating from the liver, MRI is necessary for proper delineation as opposed to liver
such as HCC and cholangiocarcinoma, as well as those that metastases which are best visualized during the venous phase.
have metastasized from a different organ. Liver tumors from MRI or FDG-PET scans can be fused to the CT to improve
metastatic disease occur more frequently than primary liver visualization of the tumor, which radiation oncologists define
cancer, of which CRC is the most common [106]. In 2012, as gross tumor volume (GTV) although accuracy of the
there was an estimated 28,720 new cases of liver and intrahe- fusion is generally 1–2 mm. CT slice thicknesses of 2–4 mm
patic biliary duct cancer in the United States with an esti- are recommended to balance lesion detection rate and image
mated 20,550 deaths. Although the 5-year relative survival noise [81]. On MRI, CRC liver metastases tend to be hypo-
rate for patients with cancer of the liver and bile ducts has vascular and HCC demonstrates tumor arterial enhancement
with “washout” and a delayed enhancing pseudocapsule, liver tumors to a combination of both. Two of these studies
restricted diffusion, and T2 hyperintensity [118]. A CTV is incorporated patients treated with SABR to liver lesions and
created by adding margin to the GTV for microscopic tumor other anatomic locations within the same report [128, 129].
extension. In a clinicopathologic correlative study in HCC Radiation dose may be based on estimates of toxicity risk
patients, CT and MRI correlated equally with resected tumor [24, 130]. Additionally, some of the HCC patients underwent
specimens. The authors concluded that margins of 5 and liver transplant after SABR [131, 132]. Child–Pugh (CP)
10 mm would have encompassed 93 % and 100 %, respec- class B patients were sometimes permitted, while other stud-
tively, of the cases [119]. For CRC metastases, microscopic ies restricted to only class A patients. None included patients
extension can be found 0.2–10 mm from the main tumor, with CP class C. The fractionation schemes vary from a sin-
with 80 % within 3 mm and 90 % within 6 mm [120]. gle large fraction up to ten fractions.
Liver motion is predominant in the superior–inferior (SI) The ideal patient for liver SBRT has no extrahepatic
direction and can vary greatly among individuals and breath- disease with >700 cc of normal uninvolved liver, has ade-
ing patterns. SI liver motion ranges from 0.5 to 3.7 cm [121– quate liver function (CP B or better) and is located >1.5 cm
123]. Average anterior–posterior (AP) and left–right (LR) liver from a luminal gastrointestinal structure. The maximum
tumor motion has been reported to be 1.0 cm (range: 0.4– tumor size is less than 6 cm for a metastasis or less than
2.2 cm) and 0.8 cm (range: 0.4–1.5 cm), respectively [124]. 8 cm for a primary HCC. Vascular involvement is not a
Liver tumor motion can be measured in several different ways contraindication. Absolute contraindications are patients
by using (1) cine MRI, (2) fluoroscopy, or preferably (3) 4D with extrahepatic disease, >8 % of the liver involved with
CT scan. Fiducial placement is beneficial for several reasons cancer, have a small liver (<500 cc), CP C, <0.5 cm from a
when treating below the diaphragm with the only drawback luminal gastrointestinal structure or a tumor that is greater
being the rare, but real risk of tumor seeding along the inser- than 15 cm for a metastasis or greater than 20 cm for a
tion track. Fiducials, if placed in or near the tumor, can act as primary HCC [133, 134].
a surrogate for tumor location during the 4D CT scan or can be Wulf [135], Rule [136], and Chang [137] demonstrated
checked by fluoroscopy. In addition, fiducials can significantly improvements in local control with dose escalation. A recent
improve image guidance. Even with volumetric IGRT, the lack pooled analysis by Chang et al. of 65 patients with 102 meta-
of contrast difference in the liver makes it nearly impossible to static CRC lesions treated to a median dose of 41.7 Gy in
visualize the tumor unless it is at the edge of the liver or near six fractions demonstrated local control of 65 % at 1 year
the gall bladder. Finally, fiducial placement is the only way and 55 % at 2 years. They estimated that doses of 46–52 Gy
tumor tracking could be performed adequately. By measuring in three fractions were needed to have 1-year local control
organ motion, an appropriate margin can be added to the CTV of more than 90 % [137]. In comparison to lung SBRT,
to create an ITV. To reduce the size of the ITV margins, patients with metastatic disease to the liver appear to have
abdominal compression has been utilized by some to reduce the highest control rates with the most aggressive regimens.
diaphragmatic and liver motion. In a study by Heinzerling Rusthoven et al. [109] have demonstrated the best local
et al., abdominal compression almost halved tumor motion control using 60 delivered in three fractions where the local
from 13.6 to 7.2 mm [125]. Notably, abdominal compression control was 92 % at 2 years including 100 % for tumors less
can increase AP motion due to liver deformation [126, 127]. than 3 cm. Wulf et al. demonstrated an improved response in
An alternative to abdominal compression, which may not be patients treated with “high dose” which was defined as
tolerated by all patients, is tumor tracking or gating. 26 Gy in one fraction or 36 Gy in three fractions (n = 26)
A PTV is created from the ITV to account for setup errors versus “low dose” which was defined as 30 Gy in three frac-
between radiation treatments. Interfraction reproducibility is tions or 27 Gy in four fractions (n = 25) [135]. In a similar
improved by doing image-guided radiotherapy (IGRT) prior fashion Goodman et al. reported the single-fraction phase I
to each treatment. This is performed by matching fiducial from Stanford where better local control was seen with
location with kV imaging and/or creating a CBCT with the doses of 26 and 30 Gy as opposed to 18 and 22 Gy [138].
patient immobilized. These pretreatment images are com- Rule et al. demonstrated greater than 90 % local control
pared to the simulation images and adjusted as appropriate, when 50 or 60 Gy was delivered in five fractions as
reducing interfraction motion. opposed to lower doses based on their phase I experience
[136]. As for the patients with HCC, 90 % local control has
Clinical Outcomes and Patient Selection been more challenging due to the underlying liver disease
Local control rates of SABR for liver tumors range from 65 having a major impact on tolerance of dose escalation.
to 100 % at 1 year. Table 13.4 summarizes recently published Lower doses have been associated with higher response
articles for SABR of primary and metastatic liver tumors. rates compared to metastases; finding the appropriate dose
The small number of patients in these trials and variations in for tumor control that will not precipitate liver toxicity is a
eligibility criteria makes direct comparisons difficult. challenge for HCC, especially when underlying liver func-
Inclusion criteria ranged from only primary or metastatic tion is poor (e.g., Child–Pugh B or C).
13
Table 13.4 Studies of SABR for liver lesions
Study (year No. pts Tumor size Dose (no. Median
published) (lesions) Pt population (range) fractions) follow-up (range) Local control OS Toxicity
Herfarth (2001) [33] 37 (60) Inc: Surgical intervention 1–3 lesions. Max 14–26 Gy (1) 5.7 months 71 % at 1 year 72 % at 1 year No statistically significant
56 mets, 4 not possible each lesion 6 cm (1–26.1) (actuarial) (actuarial) changes in liver enzymes
primary liver Median 10 cc
tumors (1–132 cc)
Schefter (2005) [34] 18 mets, 1–3 lesions, < 6 cm Median 36–60 Gy (3) NR NR NR No grade 3–5 toxicities
including 6 >700 mL normal aggregate GTV
CRC mets liver receive < 15 Gy 18 cc (3–98)
Choi (2006) [35] 20 HCC (20) Single lesion Mean 3.8 cm 50 Gy (5–10) 23 months 80 % 70 % at 1 year No grade 3–5 toxicities
CP Class A-B (2–6.5 cm) (3–55) 43 % at 2 years
No extrahepatic
disease
Mendez Romero 25 (45) total HCC or metastatic Median 3.2 cm 30.0–37.5 Gy (3) if 12.9 months 94 % at 1 year 75 % One grade 5 liver–HCC,
(2006) [36] tumors, not eligible (0.5–7.2 cm) mets, HCC without (0.5–31) Child B
for other local cirrhosis, or HCC
treatment (RFA or <4 cm with
surgery) cirrhosis
15 (31) CRC 25 Gy/5 or 30 Gy/3 82 % at 2 years Two grade 3 liver—met
mets if HCC >4 cm and pts
8 (11) HCC cirrhosis For HCC, 75 % at 1 One grade 2 asthenia—
CP Class A-B and 2 years met pt
Wulf (2006) [37] 5 (5) primary Estimated life 14 cc primary Low-dose 15 months 92 % at 1 year 72 % at 1 year No grade 3 or higher
liver expectancy (max 516) group = 28–30 Gy (2–85) toxicities
>6 months (3–4)
39 (51) mets No more than 50 % 9 cc met (max High-dose 66 % at 2 years 32 % at 2 years
total functional liver 355) group = 12–12.5 Gy
receive > 5 Gy (3) or 26 Gy (1)
Not surgical
candidate
Extrahepatic disease
controlled or treated
Dawson (2006) [38] 79 total CP Class A; >800 cc Median 293 cc Median 36.6 (6) NR 74 % NR NR
33 HCC non-involved liver (2.9–3088 cc) Range: 24–57 Gy (95%CI = 61–83 %)
12 IHC
34 Mets
Hoyer (2006) [28] 65 total pts, of Inoperable CRC Max 6 cm 45 Gy (3) 4.3 years 86 % at 2 years 67 % at 1 year 1 death from liver failure
whom 44 mets, 1–6 lesions
treated to liver Included SBRT to 38 % at 2 years 1 colonic perforation; 2
non-liver sites duodenal ulcers
No chemo within 1 13 % at 5 years
month
(continued)
197
198
Table 13.4 (continued)

Study (year Dose (no. Median follow-up
published) No. pts (lesions) Pt population Tumor size (range) fractions) (range) Local control OS Toxicity
Tse (2008) [29] 31 HCC CP Class A; Median = 173 cc Median 36 Gy (6) 17.6 months 65 % at 1 year Median = 13.4 months 7 pts (23 %) progressed to
unresectable; (9–1,913 mL) (10.8–39.2) Child–Pugh class B
extrahepatic disease
okay if bulk of
disease in liver
10 IHC More than 800 cc Range: 24–54 Gy [11.7 months (HCC),
uninvolved liver 15.0 month (IHC)]
Every other day 1 year OS = 48 %
Lee (2009) [30] 68 met pts, Unsuitable or Median = 75.2 cc Median 41.8 Gy 10.8 months 71 % at 1 year Median = 17.6 months 7 grade 3–4 acute
including 40 refractory to (1.2–3,090 mL) (6) toxicities
CRC standard treatment
Life expectancy > 3 Range: 27.7–60 Gy 63 % 1 year for CRC 1 grade 5 late toxicity
months (6) mets (malignant bowel
>800 mL uninvolved 47 % at 18 months obstruction); 1 grade 4
liver (all histologies) late toxicity (small bowel
CP Class A obstruction)
Rusthoven (2009) 47 (63) mets, 1–3 lesions Max 6 cm 36–60 Gy 16 months 95 % at 1 year Median 20.5 months 1 grade 3 toxicity (skin/
[39] including 15 Extrahepatic disease Median 2.7 cm (6–54) 92 % at 2 years 30 % at 2 years soft tissue)
CRC pts permitted if treatable (0.4–5.8)
Cardenes (2010) 17 HCC (25) 1–3 lesions Cumulative 36–48 Gy (3); later 24 months 100 % at 1 year 58 % at 1 year and 2 2 pts radiation-induced
[31] diameter 6 cm or 40 Gy (5) for CTP B (10–42 months) years liver disease (RILD)
CP Class A or B; not less; median 1–2 fractions per
resection candidates; vol = 34 cc week
no progressive or (8–95)
untreated gross
extrahepatic disease
Goodman (2010) 26 (40) total 1–5 lesions Max 5 cm 18–30 Gy (1) 17 months 77 % at 1 year Median 28.6 months No acute or late grade 3
[40] Inc: 6 CRC Life expectancy > 6 (2–55) 64.3 % at 1 year. or higher toxicities
mets, 5 IHC, 2 months 50.4 % at 2 years
recurrent CP Class A
HCC Unresectable or
pt-refused resection
Seo 2010 [41] 38 Inoperable; all <10 cm 33–57 Gy (3–4) 79 % 68 % Gr 3 skin (1)
received TACE prior 2 years LC = 66 % 2 years = 25 %
to SBRT <42 Gy versus 81 %
>42 Gy
S.J. Feigenberg et al.
13
Andolino (2011) 60 HCC Liver-only disease Median 3.2 cm CTP A: median 27 months 90 % at 2 years 67 % at 2 years No grade 3 or higher
[32] 44 Gy (3) non-heme toxicities
36 CTP A CTP B: median Median TTP For transplant pts, OS 20 % progressed CTP
24 CTP B 40 Gy (5) 47.8 months 96 % at 2 years class within 3 months
Rule (2011) [42] 27 (36), 1–5 liver mets Median 2.5 cm Dose escalation 20 month (4–53) 56 % for 30 Gy, 50 % for 30 Gy, 67 % No grade 4–5 toxicities
almost half cohorts: 89 % for 50 Gy, for 50 Gy, and 56 %
CRC >700 mL liver Range: 30 Gy (3) 100 % for 60 Gy at for 60 Gy at 2 years 1 pt in 50 Gy cohort had
receiving <21 Gy 0.4–7.8 cm 50 Gy (5) 2 years grade 3 liver function
60 Gy (5) tests
Bae (2012) [27] 41 (50) CRC 1–4 lesions included Max 7 cm Median 48 Gy (3) 28 months 64 % at 3 years 60 % at 3 years 3 grade 3–4 toxicities (of
mets, of CRC mets to (6–65) whom only 1 treated with
whom 11 (15) non-liver sites SBRT to liver)
treated to liver Radical resection of Median 13 mL Range: 45–60 Gy 57 % at 5 years 38 % at 5 years
primary, inoperable, sum of (3)

or not amenable to individual GTVs
another local (2–123)
treatment
199
Complications keeping the volume of normal liver receiving less than from x
The use of radiotherapy to the liver has been limited in the Gy to be at least 700 cc, increasing the dose from 15 to 21 Gy
past due to radiation-induced liver disease (RILD). The exact based on delivering five fractions as opposed to three.
mechanism is unknown although it postulated that endothe- Similarly, Rule et al. did not observe dose-limiting toxicity.
lial injury or activation of hepatic stellate cells yields fibrin Cardenes et al. [131] reported a multi-institutional phase I
and collagen deposition, resulting in fibrous veno-occlusive experience for patients with primary HCC. They included
disease [139, 140]. Hepatitis B virus status is an independent patients with 1–3 tumors with a cumulative tumor size of 6 cm
predictor of RILD development [141] and is therefore antici- with Child–Pugh scores of A or B. The starting dose was 36 Gy
pated to be a greater factor in HCC patients than those with in three fractions with planned escalations of 2 Gy per fraction
metastatic disease to the liver. increments. The dose was escalated to 48 Gy in three fractions
Clinically, RILD usually develops 2 weeks to 4 months fol- without a dose-limiting toxicity for patients with a Child–Pugh
lowing radiation to the liver, although it has been reported as score of A. Two with Child–Pugh score B developed grade 3
late as 7 months after radiotherapy. The aspartate transami- hepatic toxicity at 14 Gy per fraction and the protocol was
nase (AST) and alanine transaminase (ALT) levels are moder- amended to a five-fraction approach to 40 Gy which has been
ately elevated with minimal or no increase in bilirubin, but tolerable. Mendez Romero et al. [120] described an 18 % inci-
there is an increase of alkaline phophatase to 3–10 times the dence of RILD although the patient population was very small.
upper limit of normal. In severe cases, anicteric ascites and The largest experience reported is from Princess Margaret
hepatomegaly develop. It is usually self-limited and is man- Hospital which utilized a six-fraction schema and included the
aged conservatively with medical management, though largest tumors as well. In their series, no patient developed
10–20 % of patients may die from liver failure, particularly in RILD but they exclude patients with Child–Pugh score B. They
patients with preexisting hepatic dysfunction such as cirrhosis did have 5 of 31 patients develop worsening of their Child–
[142]. In 1991, Emami et al. estimated normal tissue toler- Pugh score although 3 of the patients had progressive HCC
ances to radiotherapy based on pooled patient data. The 5 % and had Child–Pugh score A6.
(TD 5) and 50 % (TD 50) probability of liver failure within 5 Other potential toxicities include hematologic, fatigue,
years were estimated based on liver volume. In conventional and nausea that can occur shortly after radiation treatment.
fractionation, the TD5 at 5 years for 1/3, 2/3, and whole liver The greatest potential late toxicity other than liver is that of
was estimated to be 50, 35, and 30 Gy, respectively. At 5 years, the luminal GI organs, which include ulceration, bleeding,
the TD 50 for 1/3, 2/3, and whole liver was estimated to be 55, and obstruction [24, 129]. Using a maximum dose of 37 Gy
45, and 40 Gy, respectively [143]. More recent studies using in three fractions and 42 Gy in five fractions with additional
different modeling systems have demonstrated that the radia- criteria based on volume of the GI organ at risk, there were
tion tolerance of the liver was higher than previously thought. no grade 3 or higher acute toxicities and approximately 11 %
A median dose of 60.75 Gy delivered in twice-daily fractions of patients developed late grade 3 toxicities. One patient died
of 1.5 Gy resulted in 12 %, 9 %, and 1 patient (<1 %) develop- of complications from duodenal perforation 11 months fol-
ing grade 3, 4, and 5 toxicities, respectively [144]. Using the lowing SABR. Normal tissue constraints used in a Princess
Lyman NTCP model, there was an estimated 4 % increased Margaret six-fraction phase I study are shown in Table 13.5.
risk of RILD per 1-Gy mean liver dose above 30 Gy [69]. As
demonstrated by Dawson et al., the risk of liver injury is
related to whether the patient had primary or metastatic liver Pancreatic Cancer: The Rationale for SBRT
tumors. The TD 50 for whole liver irradiation was 45.8 Gy for
liver metastases versus 39.8 Gy for primary liver tumors [69]. Pancreatic ductal adenocarcinoma (PDAC) is the fourth
Although this may not be relevant in the patients undergoing leading cause of cancer-related death in the United States.
extreme hypofractionated regimens as used in SBRT.
Several prospective protocols have been reported and in Table 13.5 SBRT liver study normal tissue permitted tolerances, in
general have been minimal for patients with metastatic disease six fractions for HCC [38]
as opposed to those with primary disease related to the under-
Liver Iso-NTCP model, based on effective liver volume
lying liver dysfunction. Rusthoven et al. [109], which reported irradiated (for <10 % risk, suggest mean liver
the best outcomes, utilized a three-fraction approach. Patients dose <16 Gy)
included in their study include patients with 1–3 metastases Esophagus Maximum dose (to 0.5 cc) <30 Gy
with the largest tumor <6 cm. Their most important dosimetric Stomach Maximum dose (to 0.5 cc) <30 Gy
parameter was based on the surgical literature and required the Duodenum Maximum dose (to 0.5 cc) <30 Gy
volume of normal liver receiving less than 15 Gy to be at least Large bowel Maximum dose (to 0.5 cc) <34 Gy
700 cc. Following these parameters, no dose-limiting side Kidney D67% <15 Gy
effects were seen. Rule et al. [136] used a similar parameter Spinal cord Maximum dose <25 Gy
Due to the lack of an early detection method, only doses and treatment delivery tools have been used with
15–20 % of patients can undergo curative resection at the common result, very encouraging local control. In the
time of diagnosis [81]. Among resectable patients who absence of prospective studies, single institutional expe-
receive postoperative radiotherapy, the median OS is riences have played a major role in our understanding of
approximately 20 months but these patients do not reflect the role for SBRT in pancreatic cancer. These can be
the 25 % of patients with resectable disease who have divided into experiences in the palliative setting, where
radiographically occult metastases and do not undergo most of the prevalent data exists, and those in the neoad-
resection [145, 146]. For patients with resectable PDAC juvant setting.
who receive neoadjuvant therapy and do not undergo lap-
aroscopy prior to therapy, a median OS of 12–23 months Experience with SBRT as a Definitive (Including
could be expected [147, 148]. Unfortunately, even in the Palliative) Treatment
initially resectable population and despite curative sur- The initial reports with the use of SBRT in pancreatic cancer
gery followed by adjuvant chemotherapy and/or chemora- came from Koong and colleagues [149]. In a single institu-
diation, the risk of local and distant recurrence remains tion phase I dose escalation study from Stanford University,
high at 37 % and 75 %, respectively, and tumor recurrence 15 patients with locally advanced pancreatic cancer and
typically within 12 months after surgery [145, 146]. This ECOG performance status ≤2 were treated with a single
suggests the presence of disseminated cancer cells at the dose of 15 (3 patients), 20 (5 patients), or 25 Gy (7 patients)
time of surgery. For this reason, the role of conventionally to the primary tumor alone with no specific targeting of the
fractionated radiation therapy in addition to systemic che- elective lymph nodes. Treatment planning was performed on
motherapy alone has been questioned due to the predomi- thin-cut, contrast-enhanced CT images, and radiation was
nance of distant failures. Recently, improved systemic delivered using the CyberKnife® system (Accuray Inc.,
options have been developed which may again alter the Sunnyvale, CA). All patients had three to five fiducial mark-
patterns of failure such that local therapy may become ers implanted in the tumor to facilitate image guidance dur-
more important again. ing SBRT. Patients were treated with orthogonal X-rays prior
to treatment to ensure appropriate positioning of the fiducial
Early Experience in Pancreas markers, with the pancreas immobilized in voluntary breath
While the experience with SBRT for pancreatic malignan- hold. The treatment time ranged from 3 to 6 h. The proximal
cies is less than other sites, the rationale for using SBRT in duodenum was not allowed to receive more than 50 % of the
PDAC is quite similar to that in early lung cancer. The ben- prescribed dose though the prescription isodose line was
efits of hypofractionated RT allow for greater tumor kill due variable from 64 to 85 %. Of the 15 treated patients, only 2
to radiobiologic superiority and allow for less interruption had received any prior RT as part of their therapy. Despite
with more effective systemic chemotherapy regimens. Since treating patients with essentially advanced unresectable
hypofractionated courses of therapy can be delivered in 1–5 tumors, 84 % local control was seen at 1 year with no acute
treatments, generally in less than 1 week. This can make it grade 3 or higher toxicity despite use of a very large single
easy to incorporate into systemic therapy optimizing local dose [148]. In fact, local control in the patients receiving
control without interfering with distant disease control since 25 Gy was 100 % (6 patients with imaging follow-up) though
systemic therapy will not need to be altered as in more con- they all failed distantly. Median survival for the entire group
ventionally fractionated radiotherapy. Given its location, sur- of patients was 11 months. The most frequently seen adverse
rounded by extremely radiosensitive critical structures such events were grade 2 nausea, diarrhea, and abdominal pain in
as the small bowel, stomach, and liver, SBRT for PDAC rep- 5 patients. The same group subsequently performed a dose
resents a unique challenge. The small bowel has been known escalation study using SBRT as a planned boost following
to be susceptible to radiation-induced necrosis, bleeding, and conventionally fractionated RT (1.8 Gy/fraction) to the pri-
stenosis with conventional fractionation doses above 50 Gy mary and regional lymph nodes with 5-fluorouracil [151].
and the additional biologically equivalent dose can pose a Based on the safety demonstrated in their phase I results, a
significant threat to the integrity of these critical structures. single dose of 25 Gy was administered for the boost dose to
In addition, accounting for pancreatic motion often results in the tumor alone [152]. Overall, the regimen was well toler-
large ITV margins that may result in large volumes of small ated with grade 3 acute GI toxicity noted in 2 patients who
bowel being treated. subsequently went on to develop symptomatic duodenal
Compared to the available clinical data with SBRT for ulceration. Median survival was 33 weeks and median
malignancies of the lung and liver, only a few prospective progression-free survival was 17.5 weeks. All but one of the 16
studies have been published for PDAC similar to litera- enrolled patients were controlled locally; however, distant dis-
ture a decade ago for lung cancer [129, 149–151]. Similar ease manifested in all of them eventually. Additionally from
to the early experience in lung cancer, a wide range of the same group, Chang et al. [153] published a retrospective
experience of 77 patients treated with 25 Gy to the 95 % iso- SBRT for pancreas is associated with excellent local control,
dose line covering the PTV. In this rather heterogeneous with toxicity that seems similar when delivered with accu-
group of patients (73 % unresectable, 5 % medically inoper- rate precision and careful planning.
able, 3 % marginally resectable, and 19 % metastatic), Contrasting with the high level of safety and efficacy
median survival from date of diagnosis was 11.4 months. from these reports, a prospective phase II trial [129] was con-
Once again, grade 3 or higher acute toxicities were low with ducted in Denmark for patients with biopsy-proven adeno-
only one patient experiencing a gastric ulcer. Late toxicities carcinomas of the pancreas that were N0 and less than 6 cm
(≥grade 3) included gastric ulceration in 3, duodenal stric- in maximum dimension. They prescribed 45 Gy over three
ture in 1, and biliary stricture in 2 patients, respectively. Only fractions with the dose prescribed to the ICRU reference
one patient experienced a grade 4 toxicity consisting of small point such that 95 % of the prescribed dose was delivered to
bowel perforation. Murphy et al. [154] reported dosimetric the CTV and 67 % of the prescribed dose was delivered to
parameters that predicted grade 2 and higher duodenal toxic- the PTV. The CTV was defined as the GTV plus the sur-
ity which was seen in 11 % and 29 % of their patients at 6 rounding edema, while the PTV was defined as the CTV plus
and 12 months, respectively. Based on their data from deliv- 5 mm in the axial direction and 1 cm in the cranial–caudal
ering SBRT in a single fraction, the absolute volumes of the direction. Motion management was controlled by two differ-
duodenum receiving 10, 15, 20, and 25 Gy were all signifi- ent methods using abdominal compression or a “slow” CT
cant predictors of increased toxicity. The absolute volumes simulation. IGRT was performed using portal imaging. Their
for each parameter were 16, 9.1, 3.3, and 0.21 cc, respec- findings were concerning, with patients having one or more
tively. In addition, when the maximum dose was >23 Gy, the grade 2 side effects in 64 % of the patients by 2 weeks.
grade 2–4 duodenal toxicity rate was 45 % compared to Despite the use of prophylactic ondansetron and pantopra-
12 % when the dose was <23 Gy. zole for 4 weeks, acute toxicity at 14 days after therapy was
A recent retrospective publication of the University of pronounced, with worsening of nausea and pain that
Pittsburgh experience by Rwigema and coworkers [155] also resolved in 8 of 12 patients 3 months after SBRT. Progression
suggests a relatively safe profile of SBRT for unresectable to grade 2 or higher toxicity occurs in 79 % of patients and
pancreas. In their 71 patient experience of patients treated on the overall performance status is significantly deteriorating.
either the CyberKnife or the Trilogy linear accelerator system In addition some of the patients were lost to follow-up or
(Varian Medical, Palo Alto, CA), patients received a single deteriorated quickly with only 14 of the original 22 patients
fraction of between 18 and 25 Gy. Prescription was to the making it to 2 months after treatment and 4 patients with 6
80 % isodose shell when using the CyberKnife and to the 89 % months of follow-up. In 5 patients there was evidence of
isodose shell when using the linear accelerator-based system. severe mucositis (2), ulceration (2), and ulcus-perforation of
Median survival for the entire group was 10.3 months, and the stomach (1). The actuarial local failure rate was substan-
local failure was seen in 35 % of the patients. Dosimetric cor- tially higher than other experiences at 43 %. In addition,
relatives suggested that local control was improved with doses median survival and 1-year survival were also lower at
above 24 Gy being administered. Toxicity was once again low 5.7 months and 5 %, respectively. There were differences in
with only 4 % of patients reporting ≥ grade 3 GI toxicity. their treatment approach that may explain the dismal out-
Mahadevan and coworkers [151] adopted a slightly less comes. Treatment was administered using a linear accelera-
aggressive approach using three fractions of SBRT sand- tor-based approach using abdominal compression and a
wiched between cycles 3 and 4 of gemcitabine also using the stereotactic body frame, and was targeted to the primary
CyberKnife® platform. Local control was 85 % in their 47 tumor and surrounding edema. The CTV (created by expand-
patients with late grade 3 and higher toxicities related to ing the tumor plus edema by 1 cm in the sup-inf direction
radiotherapy seen in only 9 %. For disease abutting a large and 5 mm axially) was targeted to receive >95 % of the
area of the duodenal loop, the prescribed dose was 24 Gy in prescribed dose while the PTV (including the motion com-
three fractions while disease with minimal contact in this ponent) received at least 67 %. Therefore, these volumes
region was treated to 30 Gy in three fractions. PDAC with no were significantly larger with parts of the duodenum and
involvement of the duodenal c-loop was treated to 12 Gy in stomach (included in the PTV) receiving up to 67 % of the
three fractions for a total of 36 Gy on this study. Median prescribed dose. Abdominal compression may cause patient
progression-free survival for these patients was 15 months discomfort making longer treatments less reproducible
and the median overall survival was 20 months. In addition, which can increase the likelihood of marginal failures that
the sandwich approach with the initial systemic treatment could increase local failures and increase dose to critical
allowed patients with early metastases to be excluded similar surrounding normal tissues inadvertently increasing side
to the recommendations from the GERCOR study [156] and effects. Also of note, the prospective nature of this trial man-
thereby assuring local treatment only in appropriate patients. dated strict posttreatment biopsy protocols and may also
These and several other reports have demonstrated that account for the larger percentage of patients identified with
local failure when contrasted with those assessed by Future Directions

imaging-based RECIST criteria in other studies. Therefore, Recently, systemic therapy for pancreatic cancer has
this dosing scheme and planning methodology may not be improved significantly. Similar to breast cancer and lung
safely deliverable to the upper abdomen. Although, most cancer, improving systemic control may have a major impact
investigators believe lower dose SBRT is more reasonable on increasing local failures due to the predominant presence
for palliation. of residual disease subsequently improved local control will
become vital. The PRODIGE 4/ACCORD 11 study by
Experience in the Borderline Resectable Setting Conroy and coworkers [157, 158] demonstrated a significant
with a Goal to Achieve Resectability improvement in median survival in metastatic PDAC increas-
More recently, Schellenberg et al. [152] from Stanford ing from 6.9 to 10.5 months with 5-fluorouracil, leucovorin,
reported two studies, using 3 weekly doses of full dose oxaliplatin, and irinotecan or FOLFIRINOX. This regimen is
(1,000 mg/m2) of gemcitabine (days 1, 8, 15) followed 2 now being investigated in all settings of this disease.
weeks later by a single dose of 25 Gy administration to the Investigators are currently testing the combination of this
primary pancreas tumor on day 29. The first 16 patients regimen with SBRT to the involved disease.
underwent SBRT using the CyberKnife platform with the
dose prescribed to the 95 % isodose shell. Subsequently, Treatment Planning Considerations
patients went on to receive one further cycle of gemcitabine. The proximity of the pancreas to extremely radiosensitive
While none of the patients made it to resectability, all 16 critical structures such as the duodenum, liver, kidneys, and
enrolled patients were able to complete treatment, and 13/16 stomach, in addition to its association with respiratory and
exhibited local control with minimal acute toxicity. Median peristaltic motion, presents a unique challenge in choosing
and 1-year overall survival were 11.4 months and 50 %, appropriate modalities for immobilization and for motion
respectively. Despite the initial minimal acute toxicity pro- accounting. In addition, it is often difficult to delineate the
file, late toxicities included five grade 2, one grade 3, and one tumor based on CT imaging. In the absence of well-endorsed
grade 4 duodenal complications. Correlation of the toxicity guidelines, a commonsense approach combining the experi-
profile with treatment isodose lines did not seem to suggest ences of the currently available literature in this field suggests
correlation of duodenal toxicity with a particular minimal that the use of multiphasic thin (ideally 2.5 mm) CT slices in
level of dose received. The second phase 2 study used the addition to a four-dimensional CT scan should be used at a
sandwich approach once again using 25 Gy delivered follow- minimum. In experienced centers, the use of fiducial markers
ing 3 weekly doses of gemcitabine. However, treatment on can be an invaluable tool in daily setup verification or motion
this study was delivered using a linear accelerator and 4D CT tracking in the case of treatment with the CyberKnife. In all
and PET scan were used to delineate the target, and fiducials cases, the use of oral contrast to delineate the stomach and
were used to provide fluoroscopic target acquisition. A small bowel should be considered. In addition, intravenous
9-beam arrangement was optimized to deliver dose to the contrast should be used as long as renal function permits.
95 % isodose line. Median survival was 11.8 months and the
local failure rate was 5 %. Late grade 2 toxicity was seen in
20 % of the patients and late grade 3 toxicity was only seen Conclusion
in 1 (5 %) patient. Subsequently, gemcitabine was adminis-
tered till disease progression. Overall, this regimen did sup- SBRT is an exciting new field of radiation oncology that brings
port the limited use of SBRT in patients treated with together many of the technological advancements that have
gemcitabine and allowed for a sandwich approach which occurred recently in radiation oncology. SBRT is not specific
may be important as a strategy to address early treatment to one planning system or delivery method but requires utili-
with systemic therapy prior to local treatment. zation of high-quality imaging for target definition, immobili-
An Italian phase I study by Polistina and coworkers [150] zation, high-precision planning, and delivery and image
treated 23 patients with a 10 Gy × 3 fractionation scheme guidance. Using these technological advancements, potent
sandwiched between systemic dosing cycles of gemcitabine. radiobiological doses can be delivered in convenient fraction-
A complete (imaging based) response was seen in 9 % of ation schemes, generally ranging from one to five fractions.
patients, along with a partial response in 61 % of patients and Preliminary data suggest that SBRT can be delivered
no progression in 13 % of patients. Median survival was safely with a high likelihood of local control and an accept-
10.6 months and no reported ≥grade 2 acute or late toxicities able safety profile for most sites. Clinical outcomes are
were seen with local progression being seen in only 17 % of expected to be improved even further if SBRT can be
patients. Furthermore, selected patients getting gemcitabine combined with other treatments such as surgery and chemo-
and SBRT were still able to proceed to surgical resection, an therapy, and future research efforts should include determin-
important observation when considering SBRT as a compo- ing optimal combinations of SBRT with other therapeutic
nent of future neoadjuvant strategies. modalities.
The high doses and highly conformal SBRT plans are 13. Fowler JF. Linear quadratics is alive and well: in regard to Park
more prone to error introduced by geometric or dosimetry et al. (Int J Radiat Oncol Biol Phys 2008;70:847–852). Int J Radiat
Oncol Biol Phys. 2008;72(3):957; author reply 8.
uncertainties. Such errors could lead to permanent and seri- 14. Fowler JF. Review: total doses in fractionated radiotherapy—
ous late normal tissue toxicity, and thus caution is required implications of new radiobiological data. Int J Radiat Biol Relat
when SBRT is used clinically. Serious late toxicities have Stud Phys Chem Med. 1984;46(2):103–20.
been reported after SBRT, including radiation-induced liver 15. Hayman JA, Martel MK, Ten Haken RK, Normolle DP, Todd III
RF, Littles JF, et al. Dose escalation in non-small-cell lung cancer
injury and gastrointestinal bleeding. As we gain increased using three-dimensional conformal radiation therapy: update of a
experience with SBRT, the partial volume tolerances to phase I trial. J Clin Oncol. 2001;19(1):127–36.
hypofractionated heterogeneous radiation delivered to nor- 16. Narayan S, Henning GT, Ten Haken RK, Sullivan MA, Martel
mal tissues should be able to be better defined. Optimal frac- MK, Hayman JA. Results following treatment to doses of 92.4 or
102.9 Gy on a phase I dose escalation study for non-small cell
tionations, methods of guidance, and clinical applications of lung cancer. Lung Cancer. 2004;44(1):79–88.
SBRT are not well established, and clinical trials continue to 17. Wulf J, Hadinger U, Oppitz U, Thiele W, Ness-Dourdoumas R,
be required in this field, especially in North America. Flentje M. Stereotactic radiotherapy of targets in the lung and
liver. Strahlenther Onkol. 2001;177(12):645–55.
18. Park C, Papiez L, Zhang S, Story M, Timmerman RD. Universal
survival curve and single fraction equivalent dose: useful tools in
References understanding potency of ablative radiotherapy. Int J Radiat Oncol
Biol Phys. 2008;70(3):847–52.
1. Kavanagh B, Timmerman R. Stereotactic body radiation therapy. 19. Fuks Z, Kolesnick R. Engaging the vascular component of the
Philadelphia: Lippincott Williams & Wilkins; 2005. tumor response. Cancer Cell. 2005;8(2):89–91.
2. Potters L, Steinberg M, Rose C, Timmerman R, Ryu S, Hevezi 20. Grills IS, Mangona VS, Welsh R, Chmielewski G, McInerney E,
JM, et al. American Society for Therapeutic Radiology and Martin S, et al. Outcomes after stereotactic lung radiotherapy or
Oncology and American College of Radiology practice guideline wedge resection for stage I non-small-cell lung cancer. J Clin
for the performance of stereotactic body radiation therapy. Int J Oncol. 2010;28(6):928–35.
Radiat Oncol Biol Phys. 2004;60(4):1026–32. 21. Galon J, Costes A, Sanchez-Cabo F, Kirilovsky A, Mlecnik B,
3. Hamilton AJ, Lulu BA. A prototype device for linear accelerator- Lagorce-Pages C, et al. Type, density, and location of immune
based extracranial radiosurgery. Acta Neurochir Suppl. cells within human colorectal tumors predict clinical outcome.
1995;63:40–3. Science. 2006;313(5795):1960–4.
4. Lax I, Blomgren H, Naslund I, Svanstrom R. Stereotactic radio- 22. Franzke A, Buer J, Probst-Kepper M, Lindig C, Framzle M,
therapy of malignancies in the abdomen. Methodological aspects. Schrader AJ, et al. HLA phenotype and cytokine-induced tumor
Acta Oncol. 1994;33(6):677–83. control in advanced renal cell cancer. Cancer Biother Radiopharm.
5. Blomgren H, Lax I, Naslund I, Svanstrom R. Stereotactic high 2001;16(5):401–9.
dose fraction radiation therapy of extracranial tumors using an 23. Gray WC, Chretien PB, Suter CM, Revie DR, Tomazic VT,
accelerator. Clinical experience of the first thirty-one patients. Blanchard CL, et al. Effects of radiation therapy on T-lymphocyte
Acta Oncol. 1995;34(6):861–70. subpopulations in patients with head and neck cancer. Otolaryngol
6. Timmerman R, Papiez L, McGarry R, Likes L, DesRosiers C, Head Neck Surg. 1985;93(5):650–60.
Frost S, et al. Extracranial stereotactic radioablation: results of a 24. Lee Y, Auh SL, Wang Y, Burnette B, Wang Y, Meng Y, et al.
phase I study in medically inoperable stage I non-small cell lung Therapeutic effects of ablative radiation on local tumor require
cancer. Chest. 2003;124(5):1946–55. CD8+ T cells: changing strategies for cancer treatment. Blood.
7. McGarry RC, Papiez L, Williams M, Whitford T, Timmerman 2009;114(3):589–95.
RD. Stereotactic body radiation therapy of early-stage non-small- 25. Mahmoud SM, Lee AH, Paish EC, Macmillan RD, Ellis IO,
cell lung carcinoma: phase I study. Int J Radiat Oncol Biol Phys. Green AR. The prognostic significance of B lymphocytes in inva-
2005;63(4):1010–5. sive carcinoma of the breast. Breast Cancer Res Treat.
8. Timmerman R, McGarry R, Yiannoutsos C, Papiez L, Tudor K, 2012;132:545–53.
DeLuca J, et al. Excessive toxicity when treating central tumors in 26. Mahmoud SM, Paish EC, Powe DG, Macmillan RD, Grainge MJ,
a phase II study of stereotactic body radiation therapy for medi- Lee AH, et al. Tumor-infiltrating CD8+ lymphocytes predict clini-
cally inoperable early-stage lung cancer. J Clin Oncol. cal outcome in breast cancer. J Clin Oncol. 2011;29(15):1949–55.
2006;24(30):4833–9. 27. Denkert C, Loibl S, Noske A, Roller M, Muller BM, Komor M,
9. Fakiris AJ, McGarry RC, Yiannoutsos CT, Papiez L, Williams M, et al. Tumor-associated lymphocytes as an independent predictor
Henderson MA, et al. Stereotactic body radiation therapy for of response to neoadjuvant chemotherapy in breast cancer. J Clin
early-stage non-small-cell lung carcinoma: four-year results of a Oncol. 2010;28(1):105–13.
prospective phase II study. Int J Radiat Oncol Biol Phys. 28. Paulson KG, Iyer JG, Tegeder AR, Thibodeau R, Schelter J, Koba
2009;75(3):677–82. S, et al. Transcriptome-wide studies of merkel cell carcinoma and
10. Timmerman R, Paulus R, Galvin J, Michalski J, Straube W, validation of intratumoral CD8+ lymphocyte invasion as an inde-
Bradley J, et al. Stereotactic body radiation therapy for inoperable pendent predictor of survival. J Clin Oncol. 2011;29(12):1539–46.
early stage lung cancer. JAMA. 2010;303(11):1070–6. 29. Postow MA, Callahan MK, Barker CA, Yamada Y, Yuan J, Kitano
11. Timmerman RD. Surgery versus stereotactic body radiation ther- S, et al. Immunologic correlates of the abscopal effect in a patient
apy for early-stage lung cancer: who’s down for the count? J Clin with melanoma. N Engl J Med. 2012;366(10):925–31.
Oncol. 2010;28(6):907–9. 30. Takeda T, Takeda A, Kunieda E, Ishizaka A, Takemasa K, Shimada
12. Mehta M, Scrimger R, Mackie R, Paliwal B, Chappell R, Fowler K, et al. Radiation injury after hypofractionated stereotactic radio-
J. A new approach to dose escalation in non-small-cell lung therapy for peripheral small lung tumors: serial changes on
cancer. Int J Radiat Oncol Biol Phys. 2001;49(1):23–33. CT. AJR Am J Roentgenol. 2004;182(5):1123–8.
31. Kimura T, Matsuura K, Murakami Y, Hashimoto Y, Kenjo M, 46. Dawson LA, Eccles C, Craig T. Individualized image guided iso-
Kaneyasu Y, et al. CT appearance of radiation injury of the lung NTCP based liver cancer SBRT. Acta Oncol. 2006;45(7):856–64.
and clinical symptoms after stereotactic body radiation therapy 47. Dawson LA, Litzenberg DW, Brock KK, Sanda M, Sullivan M,
(SBRT) for lung cancers: are patients with pulmonary emphysema Sandler HM, et al. A comparison of ventilatory prostate move-
also candidates for SBRT for lung cancers? Int J Radiat Oncol ment in four treatment positions. Int J Radiat Oncol Biol Phys.
Biol Phys. 2006;66(2):483–91. 2000;48(2):319–23.
32. Scott WJ, Schwabe JL, Gupta NC, Dewan NA, Reeb SD, Sugimoto 48. Malone S, Crook JM, Kendal WS, Szanto J. Respiratory-induced
JT. Positron emission tomography of lung tumors and mediastinal prostate motion: quantification and characterization. Int J Radiat
lymph nodes using [18F]fluorodeoxyglucose. The members of the Oncol Biol Phys. 2000;48(1):105–9.
PET-Lung Tumor Study Group. Ann Thorac Surg. 49. Yenice KM, Lovelock DM, Hunt MA, Lutz WR, Fournier-Bidoz
1994;58(3):698–703. N, Hua CH, et al. CT image-guided intensity-modulated therapy
33. Shim SS, Lee KS, Kim BT, Chung MJ, Lee EJ, Han J, et al. Non- for paraspinal tumors using stereotactic immobilization. Int J
small cell lung cancer: prospective comparison of integrated FDG Radiat Oncol Biol Phys. 2003;55(3):583–93.
PET/CT and CT alone for preoperative staging. Radiology. 50. Meeks SL, Buatti JM, Bouchet LG, Bova FJ, Ryken TC,
2005;236(3):1011–9. Pennington EC, et al. Ultrasound-guided extracranial radiosur-
34. Magnani P, Carretta A, Rizzo G, Fazio F, Vanzulli A, Lucignani G, gery: technique and application. Int J Radiat Oncol Biol Phys.
et al. FDG/PET and spiral CT image fusion for mediastinal lymph 2003;55(4):1092–101.
node assessment of non-small cell lung cancer patients. 51. Wulf J, Hadinger U, Oppitz U, Olshausen B, Flentje M. Stereotactic
J Cardiovasc Surg (Torino). 1999;40(5):741–8. radiotherapy of extracranial targets: CT-simulation and accuracy
35. Decoster L, Schallier D, Everaert H, Nieboer K, Meysman M, of treatment in the stereotactic body frame. Radiother Oncol.
Neyns B, et al. Complete metabolic tumour response, assessed by 2000;57(2):225–36.
18-fluorodeoxyglucose positron emission tomography (18FDG- 52. Lohr F, Debus J, Frank C, Herfarth K, Pastyr O, Rhein B, et al.
PET), after induction chemotherapy predicts a favourable out- Noninvasive patient fixation for extracranial stereotactic radio-
come in patients with locally advanced non-small cell lung cancer therapy. Int J Radiat Oncol Biol Phys. 1999;45(2):521–7.
(NSCLC). Lung Cancer. 2008;62(1):55–61. 53. Booth JT, Zavgorodni SF. Set-up error & organ motion uncer-
36. Vansteenkiste JF, Mortelmans LA. FDG-PET in the locoregional tainty: a review. Australas Phys Eng Sci Med. 1999;22(2):29–47.
lymph node staging of non-small cell lung cancer. A comprehen- 54. Dawson LA, Brock KK, Kazanjian S, Fitch D, McGinn CJ,
sive review of the Leuven Lung Cancer Group experience. Clin Lawrence TS, et al. The reproducibility of organ position using
Positron Imaging. 1999;2(4):223–31. active breathing control (ABC) during liver radiotherapy. Int J
37. Higashi T, Saga T, Nakamoto Y, Ishimori T, Mamede MH, Wada Radiat Oncol Biol Phys. 2001;51(5):1410–21.
M, et al. Relationship between retention index in dual-phase (18) 55. Eccles C, Brock KK, Bissonnette JP, Hawkins M, Dawson
F-FDG PET, and hexokinase-II and glucose transporter-1 expres- LA. Reproducibility of liver position using active breathing coor-
sion in pancreatic cancer. J Nucl Med. 2002;43(2):173–80. dinator for liver cancer radiotherapy. Int J Radiat Oncol Biol Phys.
38. Cohen RJ, Sharma NK, Yu JQ, et al. A phase I radiation dose 2006;64(3):751–9.
escalation trial of stereotactic body radiotherapy for malignant 56. Sonke JJ, Zijp L, Remeijer P, van Herk M. Respiratory correlated
lung tumors. J Biomed Sci Eng. 2010;3(4):351–8. cone beam CT. Med Phys. 2005;32(4):1176–86.
39. Sharma NK, Ruth K, Konski AA, et al. Low morbidity and excel- 57. Claude L, Arpin D, Servois V, Ayadi M, Dussart S, Ferlay C, et al.
lent local control using image guided stereotactic body radiother- Acute radiation pneumonitis in non-small cell lung cancer: is
apy (IGSBRT) for lung tumors. Int J Radiat Oncol Biol Phys. respiratory-gated control useful? Results of a French Prospective
2008;72(1):S454. Randomized Study. Int J Radiat Oncol Biol Phys. 2012;84(3):S175.
40. Husain ZA, Sharma NK, Hanlon AL, Buyyounouski MK, 58. Shirato H, Shimizu S, Kitamura K, Nishioka T, Kagei K,
Mirmiran A, Dhople AA, Turaka A, Yu M, Chen W, Feigenberg Hashimoto S, et al. Four-dimensional treatment planning and fluo-
SJ. Low pretreatment PET SUV predicts for increased local fail- roscopic real-time tumor tracking radiotherapy for moving tumor.
ure following stereotactic body radiation therapy for lung cancer. Int J Radiat Oncol Biol Phys. 2000;48(2):435–42.
Int J Radiat Oncol Biol Phys. 2010;78(3):S525. 59. Benedict SH, Yenice KM, Followill D, Galvin JM, Hinson W,
41. Hoopes DJ, Tann M, Fletcher JW, Forquer JA, Lin PF, Lo SS, Kavanagh B, et al. Stereotactic body radiation therapy: the report
et al. FDG-PET and stereotactic body radiotherapy (SBRT) for of AAPM Task Group 101. Med Phys. 2010;37(8):4078–101.
stage I non-small-cell lung cancer. Lung Cancer. 60. Balter JM, Lam KL, McGinn CJ, Lawrence TS, Ten Haken
2007;56(2):229–34. RK. Improvement of CT-based treatment-planning models of
42. Sovik A, Malinen E, Skogmo HK, Bentzen SM, Bruland OS, abdominal targets using static exhale imaging. Int J Radiat Oncol
Olsen DR. Radiotherapy adapted to spatial and temporal variabil- Biol Phys. 1998;41(4):939–43.
ity in tumor hypoxia. Int J Radiat Oncol Biol Phys. 61. Arvidson NB, Mehta MP, Tome WA. Dose coverage beyond the
2007;68(5):1496–504. gross tumor volume for various stereotactic body radiotherapy
43. Ishimori T, Saga T, Nagata Y, Nakamoto Y, Higashi T, Mamede planning techniques reporting similar control rates for stage I non-
M, et al. 18F-FDG and 11C-methionine PET for evaluation of small-cell lung cancer. Int J Radiat Oncol Biol Phys.
treatment response of lung cancer after stereotactic radiotherapy. 2008;72(5):1597–603.
Ann Nucl Med. 2004;18(8):669–74. 62. van Herk M, Remeijer P, Rasch C, Lebesque JV. The probability
44. Direcks WG, Berndsen SC, Proost N, Peters GJ, Balzarini J, of correct target dosage: dose-population histograms for deriving
Spreeuwenberg MD, et al. [18F]FDG and [18F]FLT uptake in treatment margins in radiotherapy. Int J Radiat Oncol Biol Phys.
human breast cancer cells in relation to the effects of chemother- 2000;47(4):1121–35.
apy: an in vitro study. Br J Cancer. 2008;99(3):481–7. 63. van Herk M, Remeijer P, Lebesque JV. Inclusion of geometric
45. van Luijk P, Bijl HP, Konings AW, van der Kogel AJ, Schippers uncertainties in treatment plan evaluation. Int J Radiat Oncol Biol
JM. Data on dose-volume effects in the rat spinal cord do not sup- Phys. 2002;52(5):1407–22.
port existing NTCP models. Int J Radiat Oncol Biol Phys. 64. Jin L, Wang L, Li J, Luo W, Feigenberg SJ, Ma CM. Investigation of
2005;61(3):892–900. optimal beam margins for stereotactic radiotherapy of lung-cancer
using Monte Carlo dose calculations. Phys Med Biol. 84. Stephans KL, Djemil T, Reddy CA, Gajdos SM, Kolar M, Mason
2007;52(12):3549–61. D, et al. A comparison of two stereotactic body radiation fraction-
65. Kavanagh BD, Timmerman RD, Benedict SH, Wu Q, Schefter ation schedules for medically inoperable stage I non-small cell
TE, Stuhr K, et al. How should we describe the radiobiologic lung cancer: the Cleveland Clinic experience. J Thorac Oncol.
effect of extracranial stereotactic radiosurgery: equivalent uniform 2009;4(8):976–82.
dose or tumor control probability? Med Phys. 2003;30(3):321–4. 85. Shapiro R, Forquer JA, Henderson MA, Brabham JG, Barriger
66. DesRosiers PM, Moskvin VP, DesRosiers CM, Timmerman RD, RB, Andolino DL, Johnstone PAS, Fakiris AJ. Central Tumors in
Randall ME, Papiez LS. Lung cancer radiation therapy: Monte Node Negative Early-stage Non-small Cell Lung Cancer
Carlo investigation of “under dose” by high energy photons. (NSCLC): survival and toxicity outcomes following Stereotactic
Technol Cancer Res Treat. 2004;3(3):289–94. Body Radiation Therapy (SBRT). Int J Radiat Oncol Biol Phys.
67. Xiao Y, Papiez L, Paulus R, Timmerman R, Straube WL, Bosch 2009;75(3):S457.
WR, et al. Dosimetric evaluation of heterogeneity corrections for 86. Bradley JD, Robinson C, Parikh P, Bien-Willner L, DeWees T,
RTOG 0236: stereotactic body radiotherapy of inoperable stage Gao F. Prospective phase i dose escalation results of SBRT for
I-II non-small-cell lung cancer. Int J Radiat Oncol Biol Phys. centrally-located stage I NSCLC. Int J Radiat Oncol Biol Phys.
2009;73(4):1235–42. 2011;81(2):S79.
68. Dawson LA, Eccles C, Bissonnette JP, Brock KK. Accuracy of 87. Stephans KL, Djemil T, Reddy CA, Gajdos SM, Kolar M,
daily image guidance for hypofractionated liver radiotherapy with Machuzak M, et al. Comprehensive analysis of pulmonary func-
active breathing control. Int J Radiat Oncol Biol Phys. tion Test (PFT) changes after stereotactic body radiotherapy
2005;62(4):1247–52. (SBRT) for stage I lung cancer in medically inoperable patients.
69. Balter JM, Brock KK, Litzenberg DW, McShan DL, Lawrence J Thorac Oncol. 2009;4(7):838–44.
TS, Ten Haken R, et al. Daily targeting of intrahepatic tumors for 88. Buyyounouski MK, Balter P, Lewis B, D’Ambrosio DJ, Dilling TJ,
radiotherapy. Int J Radiat Oncol Biol Phys. 2002;52(1):266–71. Miller RC, et al. Stereotactic body radiotherapy for early-stage non-
70. Uematsu M, Sonderegger M, Shioda A, Tahara K, Fukui T, Hama small-cell lung cancer: report of the ASTRO Emerging Technology
Y, et al. Daily positioning accuracy of frameless stereotactic radia- Committee. Int J Radiat Oncol Biol Phys. 2010;78(1):3–10.
tion therapy with a fusion of computed tomography and linear 89. Guckenberger M, Kestin LL, Hope AJ, Belderbos J, Werner-
accelerator (focal) unit: evaluation of z-axis with a z-marker. Wasik M, Yan D, et al. Is there a lower limit of pretreatment pul-
Radiother Oncol. 1999;50(3):337–9. monary function for safe and effective stereotactic body
71. Court L, Rosen I, Mohan R, Dong L. Evaluation of mechanical radiotherapy for early-stage non-small cell lung cancer? J Thorac
precision and alignment uncertainties for an integrated CT/LINAC Oncol. 2012;7(3):542–51.
system. Med Phys. 2003;30(6):1198–210. 90. Onishi H, Marino K, Terahara A, Kokubo M, Onimaru R,
72. Kuriyama K, Onishi H, Sano N, Komiyama T, Aikawa Y, Tateda Shioyama Y, Matsuo Y, Kozuka T, Ishikura S, Hiraoka M. Case
Y, et al. A new irradiation unit constructed of self-moving gantry- series study of 26 patients who developed fatal radiation
CT and linac. Int J Radiat Oncol Biol Phys. 2003;55(2):428–35. pneumonitis (RP) after stereotactic body radiotherapy for lung
73. Onishi H, Kuriyama K, Komiyama T, Tanaka S, Sano N, Aikawa cancer. Int J Radiat Oncol Biol Phys. 2009;75(3):S62.
Y, et al. A new irradiation system for lung cancer combining linear 91. Hughes LL, Wang M, Page DL, Gray R, Solin LJ, Davidson NE,
accelerator, computed tomography, patient self-breath-holding, et al. Local excision alone without irradiation for ductal carci-
and patient-directed beam-control without respiratory monitoring noma in situ of the breast: a trial of the Eastern Cooperative
devices. Int J Radiat Oncol Biol Phys. 2003;56(1):14–20. Oncology Group. J Clin Oncol. 2009;27(32):5319–24.
74. Jaffray DA. Emergent technologies for 3-dimensional image- 92. Wambaka MA, Matsuo Y, Shibuya K, Ueki N, Nakamura M,
guided radiation delivery. Semin Radiat Oncol. Mukumoto N, Nakamura A, Sakanaka K, Mizowaki T, Hiraoka
2005;15(3):208–16. M. Abdominal compression and respiratory motion as predictors
75. Cho Y, Moseley DJ, Siewerdsen JH, Jaffray DA. Accurate tech- of local recurrence in patients treated with stereotactic body radia-
nique for complete geometric calibration of cone-beam computed tion therapy for primary lung cancer. Int J Radiat Oncol Biol Phys.
tomography systems. Med Phys. 2005;32(4):968–83. 2011;81(2):S608.
76. Simpson RG, Chen CT, Grubbs EA, Swindell W. A 4-MV CT 93. Wang L, Feigenberg S, Fan J, Jin L, Turaka A, Chen L, et al.
scanner for radiation therapy: the prototype system. Med Phys. Target repositional accuracy and PTV margin verification using
1982;9(4):574–9. three-dimensional cone-beam computed tomography (CBCT) in
77. Swindell W, Simpson RG, Oleson JR, Chen CT, Grubbs stereotactic body radiotherapy (SBRT) of lung cancers. J Appl
EA. Computed tomography with a linear accelerator with radio- Clin Med Phys. 2012;13(2):3708.
therapy applications. Med Phys. 1983;10(4):416–20. 94. Josipovic M, Persson GF, Logadottir A, Smulders B, Westmann
78. Nakagawa K, Aoki Y, Tago M, Terahara A, Ohtomo G, Bangsgaard JP. Translational and rotational intra- and inter-
K. Megavoltage CT-assisted stereotactic radiosurgery for thoracic fractional errors in patient and target position during a short course
tumors: original research in the treatment of thoracic neoplasms. of frameless stereotactic body radiotherapy. Acta Oncol.
Int J Radiat Oncol Biol Phys. 2000;48(2):449–57. 2012;51(5):610–7.
79. Pouliot J, Bani-Hashemi A, Chen J, Svatos M, Ghelmansarai F, 95. Li W, Purdie TG, Taremi M, Fung S, Brade A, Cho BC, et al.
Mitschke M, et al. Low-dose megavoltage cone-beam CT for radi- Effect of immobilization and performance status on intrafraction
ation therapy. Int J Radiat Oncol Biol Phys. 2005;61(2):552–60. motion for stereotactic lung radiotherapy: analysis of 133 patients.
80. Klein EE, Hanley J, Bayouth J, Yin FF, Simon W, Dresser S, et al. Int J Radiat Oncol Biol Phys. 2011;81(5):1568–75.
Task Group 142 report: quality assurance of medical accelerators. 96. Nagata Y, Takayama K, Matsuo Y, Norihisa Y, Mizowaki T,
Med Phys. 2009;36(9):4197–212. Sakamoto T, et al. Clinical outcomes of a phase I/II study of 48 Gy
81. Jemal A, Siegel R, Ward E, Hao Y, Xu J, Thun MJ. Cancer statis- of stereotactic body radiotherapy in 4 fractions for primary lung
tics, 2009. CA Cancer J Clin. 2009;59(4):225–49. cancer using a stereotactic body frame. Int J Radiat Oncol Biol
82. Mountain CF. Revisions in the International System for Staging Phys. 2005;63(5):1427–31.
Lung Cancer. Chest. 1997;111(6):1710–7. 97. Onishi H, Araki T, Shirato H, Nagata Y, Hiraoka M, Gomi K, et al.
83. McGarry RC, Song G, des Rosiers P, Timmerman R. Observation- Stereotactic hypofractionated high-dose irradiation for stage I non-
only management of early stage, medically inoperable lung can- small cell lung carcinoma: clinical outcomes in 245 subjects in a
cer: poor outcome. Chest. 2002;121(4):1155–8. Japanese multiinstitutional study. Cancer. 2004;101(7):1623–31.
98. Dunlap NE, Larner JM, Read PW, Kozower BD, Lau CL, Sheng 115. Kelly RJ, Kemeny NE, Leonard GD. Current strategies using
K, et al. Size matters: a comparison of T1 and T2 peripheral non- hepatic arterial infusion chemotherapy for the treatment of
small-cell lung cancers treated with stereotactic body radiation colorectal cancer. Clin Colorectal Cancer. 2005;5(3):166–74.
therapy (SBRT). J Thorac Cardiovasc Surg. 2010;140(3):583–9. 116. Salem R, Lewandowski RJ, Mulcahy MF, Riaz A, Ryu RK,
99. Grills IS, Hope AJ, Guckenberger M, Kestin LL, Werner-Wasik Ibrahim S, et al. Radioembolization for hepatocellular carcinoma
M, Yan D, et al. A collaborative analysis of stereotactic lung using Yttrium-90 microspheres: a comprehensive report of long-
radiotherapy outcomes for early-stage non-small-cell lung cancer term outcomes. Gastroenterology. 2010;138(1):52–64.
using daily online cone-beam computed tomography image- 117. Vogl TJ, Zangos S, Eichler K, Yakoub D, Nabil M. Colorectal
guided radiotherapy. J Thorac Oncol. 2012;7(9):1382–93. liver metastases: regional chemotherapy via transarterial chemo-
100. Rowe BP, Boffa DJ, Wilson LD, Kim AW, Detterbeck FC, Decker embolization (TACE) and hepatic chemoperfusion: an update. Eur
RH. Stereotactic body radiotherapy for central lung tumors. Radiol. 2007;17(4):1025–34.
J Thorac Oncol. 2012;7(9):1394–9. 118. Fowler KJ, Brown JJ, Narra VR. Magnetic resonance imaging of
101. Barriger RB, Forquer JA, Brabham JG, Andolino DL, Shapiro focal liver lesions: approach to imaging diagnosis. Hepatology.
RH, Henderson MA, et al. A dose-volume analysis of radiation 2011;54(6):2227–37.
pneumonitis in non-small cell lung cancer patients treated with 119. Kelsey CR, Schefter T, Nash SR, Russ P, Baron AE, Zeng C, et al.
stereotactic body radiation therapy. Int J Radiat Oncol Biol Phys. Retrospective clinicopathologic correlation of gross tumor size of
2012;82(1):457–62. hepatocellular carcinoma: implications for stereotactic body
102. Onishi H, Shirato H, Nagata Y, Hiraoka M, Fujino M, Gomi K, radiotherapy. Am J Clin Oncol. 2005;28(6):576–80.
et al. Hypofractionated stereotactic radiotherapy (HypoFXSRT) 120. Mendez Romero A, Wunderink W, Hussain SM, De Pooter JA,
for stage I non-small cell lung cancer: updated results of 257 Heijmen BJ, Nowak PC, et al. Stereotactic body radiation therapy
patients in a Japanese multi-institutional study. J Thorac Oncol. for primary and metastatic liver tumors: a single institution phase
2007;2(7 Suppl 3):S94–100. I-II study. Acta Oncol. 2006;45(7):831–7.
103. Dunlap NE, Cai J, Biedermann GB, Yang W, Benedict SH, Sheng 121. Balter JM, Dawson LA, Kazanjian S, McGinn C, Brock KK,
K, et al. Chest wall volume receiving >30 Gy predicts risk of Lawrence T, et al. Determination of ventilatory liver movement
severe pain and/or rib fracture after lung stereotactic body radio- via radiographic evaluation of diaphragm position. Int J Radiat
therapy. Int J Radiat Oncol Biol Phys. 2010;76(3):796–801. Oncol Biol Phys. 2001;51(1):267–70.
104. Stephans KL, Djemil T, Tendulkar RD, Robinson CG, Reddy CA, 122. Davies SC, Hill AL, Holmes RB, Halliwell M, Jackson
Videtic GM. Prediction of chest wall toxicity from lung stereotac- PC. Ultrasound quantitation of respiratory organ motion in the
tic body radiotherapy (SBRT). Int J Radiat Oncol Biol Phys. upper abdomen. Br J Radiol. 1994;67(803):1096–102.
2012;82(2):974–80. 123. Aruga T, Itami J, Aruga M, Nakajima K, Shibata K, Nojo T, et al.
105. Welsh J, Thomas J, Shah D, Allen PK, Wei X, Mitchell K, et al. Target volume definition for upper abdominal irradiation using CT
Obesity increases the risk of chest wall pain from thoracic stereo- scans obtained during inhale and exhale phases. Int J Radiat Oncol
tactic body radiation therapy. Int J Radiat Oncol Biol Phys. Biol Phys. 2000;48(2):465–9.
2011;81(1):91–6. 124. Kirilova A, Lockwood G, Choi P, Bana N, Haider MA, Brock KK,
106. Hess KR, Varadhachary GR, Taylor SH, Wei W, Raber MN, Lenzi et al. Three-dimensional motion of liver tumors using cine-
R, et al. Metastatic patterns in adenocarcinoma. Cancer. magnetic resonance imaging. Int J Radiat Oncol Biol Phys.
2006;106(7):1624–33. 2008;71(4):1189–95.
107. Prasad KR, Young RS, Burra P, Zheng SS, Mazzaferro V, Moon 125. Heinzerling JH, Anderson JF, Papiez L, Boike T, Chien S, Zhang
DB, et al. Summary of candidate selection and expanded criteria G, et al. Four-dimensional computed tomography scan analysis of
for liver transplantation for hepatocellular carcinoma: a review tumor and organ motion at varying levels of abdominal compres-
and consensus statement. Liver Transpl. 2011;17 Suppl 2:S81–9. sion during stereotactic treatment of lung and liver. Int J Radiat
108. Mazzaferro V, Bhoori S, Sposito C, Bongini M, Langer M, Miceli Oncol Biol Phys. 2008;70(5):1571–8.
R, et al. Milan criteria in liver transplantation for hepatocellular 126. Eccles CL, Patel R, Simeonov AK, Lockwood G, Haider M,
carcinoma: an evidence-based analysis of 15 years of experience. Dawson LA. Comparison of liver tumor motion with and without
Liver Transpl. 2011;17 Suppl 2:S44–57. abdominal compression using cine-magnetic resonance imaging.
109. Rusthoven KE, Kavanagh BD, Cardenes H, Stieber VW, Burri Int J Radiat Oncol Biol Phys. 2011;79(2):602–8.
SH, Feigenberg SJ, et al. Multi-institutional phase I/II trial of ste- 127. Eccles CL, Dawson LA, Moseley JL, Brock KK. Interfraction
reotactic body radiation therapy for liver metastases. J Clin Oncol. liver shape variability and impact on GTV position during liver
2009;27(10):1572–8. stereotactic radiotherapy using abdominal compression. Int J
110. Tiong L, Maddern GJ. Systematic review and meta-analysis of Radiat Oncol Biol Phys. 2011;80(3):938–46.
survival and disease recurrence after radiofrequency ablation for 128. Baek HM, Chen JH, Nie K, Yu HJ, Bahri S, Mehta RS, et al.
hepatocellular carcinoma. Br J Surg. 2011;98(9):1210–24. Predicting pathologic response to neoadjuvant chemotherapy in
111. Fong Y, Cohen AM, Fortner JG, Enker WE, Turnbull AD, Coit breast cancer by using MR imaging and quantitative 1H MR spec-
DG, et al. Liver resection for colorectal metastases. J Clin Oncol. troscopy. Radiology. 2009;251(3):653–62.
1997;15(3):938–46. 129. Hoyer M, Roed H, Sengelov L, Traberg A, Ohlhuis L, Pedersen
112. Berber E, Pelley R, Siperstein AE. Predictors of survival after J, et al. Phase-II study on stereotactic radiotherapy of locally
radiofrequency thermal ablation of colorectal cancer metastases to advanced pancreatic carcinoma. Radiother Oncol.
the liver: a prospective study. J Clin Oncol. 2005;23(7):1358–64. 2005;76(1):48–53.
113. Steele Jr G, Bleday R, Mayer RJ, Lindblad A, Petrelli N, Weaver 130. Tse RV, Hawkins M, Lockwood G, Kim JJ, Cummings B, Knox J,
D. A prospective evaluation of hepatic resection for colorectal car- et al. Phase I study of individualized stereotactic body radiother-
cinoma metastases to the liver: Gastrointestinal Tumor Study apy for hepatocellular carcinoma and intrahepatic cholangiocarci-
Group Protocol 6584. J Clin Oncol. 1991;9(7):1105–12. noma. J Clin Oncol. 2008;26(4):657–64.
114. Wagner JS, Adson MA, Van Heerden JA, Adson MH, Ilstrup 131. Cardenes HR, Price TR, Perkins SM, Maluccio M, Kwo P, Breen
DM. The natural history of hepatic metastases from colorectal TE, et al. Phase I feasibility trial of stereotactic body radiation
cancer. A comparison with resective treatment. Ann Surg. therapy for primary hepatocellular carcinoma. Clin Transl Oncol.
1984;199(5):502–8. 2010;12(3):218–25.
132. Andolino DL, Johnson CS, Maluccio M, Kwo P, Tector AJ, Zook 147. Pisters PW, Wolff RA, Janjan NA, Cleary KR, Charnsangavej
J, et al. Stereotactic body radiotherapy for primary hepatocellular C, Crane CN, et al. Preoperative paclitaxel and concurrent rapid-
carcinoma. Int J Radiat Oncol Biol Phys. 2011;81(4):e447–53. fractionation radiation for resectable pancreatic adenocarcinoma:
133. Swaminath A, Dawson LA. Emerging role of radiotherapy in the toxicities, histologic response rates, and event-free outcome.
management of liver metastases. Cancer J. 2010;16(2):150–5. J Clin Oncol. 2002;20(10):2537–44.
134. Klein J, Dawson LA. Hepatocellular carcinoma radiation therapy: 148. Varadhachary GR, Wolff RA, Crane CH, Sun CC, Lee JE, Pisters
review of evidence and future opportunities. Int J Radiat Oncol PW, et al. Preoperative gemcitabine and cisplatin followed by
Biol Phys. 2013;87:22–32. gemcitabine-based chemoradiation for resectable adenocarcinoma
135. Wulf J, Guckenberger M, Haedinger U, Oppitz U, Mueller G, of the pancreatic head. J Clin Oncol. 2008;26(21):3487–95.
Baier K, et al. Stereotactic radiotherapy of primary liver cancer 149. Koong AC, Christofferson E, Le QT, Goodman KA, Ho A, Kuo T,
and hepatic metastases. Acta Oncol. 2006;45(7):838–47. et al. Phase II study to assess the efficacy of conventionally frac-
136. Rule W, Timmerman R, Tong L, Abdulrahman R, Meyer J, Boike tionated radiotherapy followed by a stereotactic radiosurgery
T, et al. Phase I dose-escalation study of stereotactic body radio- boost in patients with locally advanced pancreatic cancer. Int J
therapy in patients with hepatic metastases. Ann Surg Oncol. Radiat Oncol Biol Phys. 2005;63(2):320–3.
2011;18(4):1081–7. 150. Polistina F, Costantin G, Casamassima F, Francescon P, Guglielmi
137. Chang DT, Swaminath A, Kozak M, Weintraub J, Koong AC, Kim R, Panizzoni G, et al. Unresectable locally advanced pancreatic
J, et al. Stereotactic body radiotherapy for colorectal liver metas- cancer: a multimodal treatment using neoadjuvant chemoradio-
tases: a pooled analysis. Cancer. 2011;117(17):4060–9. therapy (gemcitabine plus stereotactic radiosurgery) and subse-
138. Goodman KA, Wiegner EA, Maturen KE, Zhang Z, Mo Q, Yang quent surgical exploration. Ann Surg Oncol.
G, Gibbs IC, Fisher GA, Koong AC. Dose-escalation study of 2010;17(8):2092–101.
single-fraction stereotactic body radiotherapy for liver malignan- 151. Mahadevan A, Jain S, Goldstein M, Miksad R, Pleskow D,
cies. Int J Radiat Oncol Biol Phys. 2010;78(2):486–93. Sawhney M, et al. Stereotactic body radiotherapy and gemcitabine
doi:10.1016/j.ijrobp.2009.08.020. Epub 2010 Mar 28. for locally advanced pancreatic cancer. Int J Radiat Oncol Biol
139. Fajardo LF, Colby TV. Pathogenesis of veno-occlusive liver dis- Phys. 2010;78(3):735–42.
ease after radiation. Arch Pathol Lab Med. 1980;104(11):584–8. 152. Schellenberg D, Goodman KA, Lee F, Chang S, Kuo T, Ford JM,
140. Sempoux C, Horsmans Y, Geubel A, Fraikin J, Van Beers BE, et al. Gemcitabine chemotherapy and single-fraction stereotactic
Gigot JF, et al. Severe radiation-induced liver disease following body radiotherapy for locally advanced pancreatic cancer. Int J
localized radiation therapy for biliopancreatic carcinoma: activa- Radiat Oncol Biol Phys. 2008;72(3):678–86.
tion of hepatic stellate cells as an early event. Hepatology. 153. Chang DT, Schellenberg D, Shen J, Kim J, Goodman KA, Fisher
1997;26(1):128–34. GA, et al. Stereotactic radiotherapy for unresectable adenocarci-
141. Cheng AS, Chan HL, Leung WK, To KF, Go MY, Chan JY, et al. noma of the pancreas. Cancer. 2009;115(3):665–72.
Expression of HBx and COX-2 in chronic hepatitis B, cirrhosis 154. Murphy JD, Christman-Skieller C, Kim J, Dieterich S, Chang DT,
and hepatocellular carcinoma: implication of HBx in upregulation Koong AC. A dosimetric model of duodenal toxicity after stereo-
of COX-2. Mod Pathol. 2004;17(10):1169–79. tactic body radiotherapy for pancreatic cancer. Int J Radiat Oncol
142. Lawrence TS, Robertson JM, Anscher MS, Jirtle RL, Ensminger Biol Phys. 2010;78(5):1420–6.
WD, Fajardo LF. Hepatic toxicity resulting from cancer treatment. 155. Rwigema JC, Parikh SD, Heron DE, Howell M, Zeh H, Moser AJ,
Int J Radiat Oncol Biol Phys. 1995;31(5):1237–48. et al. Stereotactic body radiotherapy in the treatment of advanced
143. Emami B, Lyman J, Brown A, Coia L, Goitein M, Munzenrider adenocarcinoma of the pancreas. Am J Clin Oncol.
JE, et al. Tolerance of normal tissue to therapeutic irradiation. Int 2011;34(1):63–9.
J Radiat Oncol Biol Phys. 1991;21(1):109–22. 156. Van Laethem JL, Hammel P, Mornex F, Azria D, Van Tienhoven
144. Dawson LA, Ten Haken RK. Partial volume tolerance of the liver G, Vergauwe P, et al. Adjuvant gemcitabine alone versus
to radiation. Semin Radiat Oncol. 2005;15(4):279–83. gemcitabine-based chemoradiotherapy after curative resection for
145. Oettle H, Post S, Neuhaus P, Gellert K, Langrehr J, Ridwelski K, pancreatic cancer: a randomized EORTC-40013–22012/FFCD-
et al. Adjuvant chemotherapy with gemcitabine vs observation in 9203/GERCOR phase II study. J Clin Oncol.
patients undergoing curative-intent resection of pancreatic cancer: 2010;28(29):4450–6.
a randomized controlled trial. JAMA. 2007;297(3):267–77. 157. Conroy T, Gavoille C, Samalin E, Ychou M, Ducreux M. The role
146. Regine WF, Winter KA, Abrams RA, Safran H, Hoffman JP, of the FOLFIRINOX regimen for advanced pancreatic cancer.
Konski A, et al. Fluorouracil vs gemcitabine chemotherapy before Curr Oncol Rep. 2013;15:182–9.
and after fluorouracil-based chemoradiation following resection 158. Conroy T, Mitry E. [Chemotherapy of metastatic pancreatic ade-
of pancreatic adenocarcinoma: a randomized controlled trial. nocarcinoma: challenges and encouraging results]. Bull Cancer.
JAMA. 2008;299(9):1019–26. 2011;98(12):1439–46.
Part IV
Treatment of Disease Types
Metastatic Brain Tumors
14
Edward W. Jung, John H. Suh, Samuel T. Chao,
Michael A. Vogelbaum, and Gene H. Barnett
complication in cancer patients, yet long-term survival is pos-

Introduction sible [8]. As such, decisions regarding treatment options for
brain metastases present therapeutic dilemmas for both physi-
Brain metastases are a significant cause of morbidity and cians and patients. This chapter reviews prognostic factors,
mortality that affect 20–30 % of cancer patients [1]. The surgical results, WBRT, rationale for stereotactic radiosur-
incidence in the USA has been estimated to be 170,000— gery (SRS), surgery compared with SRS, results of SRS with
200,000 cases per year [2]. Brain metastases most commonly and without WBRT, SRS to the surgical resection cavity, use
originate from tumors of the lung, breast, kidney, and mela- of radiation sensitizers and protectors, complications of SRS,
noma. In some cases, the primary site is unknown. Melanoma and future directions in treatment of brain metastases.
and lung cancer often present with multiple lesions, whereas
single lesions are more common with breast, colon, and renal
cell cancers. Malignancies which can present with hemor- Prognostic Factors
rhagic lesions include melanoma, choriocarcinoma, testicu-
lar, thyroid, and renal cell cancers. Identifying factors that predict outcomes after treatment of
A single brain metastasis is defined as one lesion in the brain metastasis allows physicians to convey prognosis to
brain with at least one other region of extracranial metastasis patients and helps streamline treatment decisions and expecta-
or uncontrolled primary, while a solitary brain metastasis is tions. The most commonly used prognostic scale for patients
defined as one brain metastasis without any evidence of with brain metastasis has been the Radiation Therapy
extracranial metastasis. An estimated 30–40 % of patients Oncology Group (RTOG) recursive partitioning analysis
present with a solitary lesion [3]. Multiple metastases are (RPA) [9]. This scale divides patients into three classes based
defined as more than one intracranial lesion. on 1,200 consecutive patients enrolled in three RTOG trials
Despite improvements in imaging, surgical technique, and from 1979 to 1993. The vast majority of these patients had
radiation therapy, the prognosis for patients with brain metas- unresectable and/or multiple metastases and received standard
tases remains poor. Historically, patients who receive no doses of WBRT. The most important factors for predicting sur-
treatment have a median survival of only 1 month. Those who vival include extracranial metastases, patient age, Karnofsky
are treated with steroids survive a median of 2 months, and performance status, and control of primary tumor (Table 14.1).
those who undergo whole-brain radiation therapy (WBRT) A limitation of the RPA system is that it provides only three
survive a median of 4–7 months after treatment [4–7]. prognostic groups and has little power to discriminate between
Consequently, the development of brain metastasis is a feared patients since most fall into one category (Group II).
A study from Lagerwaald and colleagues reviewed 1,292
patients with brain metastases [10]. In this Dutch study, lung
cancer was the most common primary disease (56 %). Median
survival was 3.4 months, and the 1-year and 2-year survivals
E.W. Jung, M.D. • J.H. Suh, M.D. (*) • S.T. Chao, M.D. were 12 % and 4 %, respectively. The most important factors
Department of Radiation Oncology, Cleveland Clinic Foundation,
Cleveland, OH, USA
affecting survival were treatment modality, performance status,
e-mail: suhj@ccf.org extracranial disease burden, and response to steroid treatment.
M.A. Vogelbaum, M.D., Ph.D. • G.H. Barnett, M.D., M.B.A.
More recently, a prognostic scoring system was developed
Department of Neurosurgery, Cleveland Clinic Foundation, specifically for patients undergoing stereotactic radiosurgery,
Cleveland, OH, USA known as the score index for radiosurgery (SIR). This system
212 E.W. Jung et al.
Table 14.1 RTOG RPA classes for brain metastases Table 14.3 GPA scores for brain metastases
Factors Median survival (months) Graded prognostic assessment
Class I Age <65 and 7.1 Characteristic 0 0.5 1.0 Grade MS (mo)
KPS ≥70 and Age >60 50–59 <50 3.5–4 11.0
Controlled primary tumor KPS <70 70–80 90–100 3 6.9
and no extracranial metastases. # CNS mets >3 2–3 1 1.5–2.5 3.8
All 4 criteria have to be met. Extracranial mets Present – Absent 0–1 2.6
Class II All others not meeting criteria 4.2
From: Sperduto CM, Watanabe Y, Mullan J, Hood T, Dyste G, Watts C,
for class I and class III
et al. A validation study of a new prognostic index for patients with
Class III KPS <70 2.3 brain metastases: the Graded Prognostic Assessment. J Neurosurg.
RTOG Radiation Therapy Oncology Group, RPA recursive partitioning 2008 Dec;109 Suppl:87–9
analysis, KPS Karnofsky performance status
From Gaspar L, Scott C, Rotman M, Asbell S, Phillips T, Wasserman T,
et al. Recursive partitioning analysis (RPA) of prognostic factors in three extracranial metastases, and number of brain metastases. For
Radiation Therapy Oncology Group (RTOG) brain metastases trials. Int
melanoma and renal cell carcinoma, the significant prognos-
J Radiat Oncol Biol Phys. 1997;37(4):745-51; used with permission
tic factors are KPS and the number of brain metastases. For
breast and GI cancers, the KPS is the only prognostic factor.
Median survival by diagnosis is summarized in Table 14.4.
Table 14.2 Score index for radiosurgery in brain metastases (SIR) Breast cancer patients have the best overall survival based
Score on DS-GPA. This prompted a follow-up study to better strat-
Variable 0 1 2 ify these patients by analyzing a larger cohort [14]. Significant
Age (years) ≥60 51–59 ≤50 prognostic factors are Karnofsky performance status (KPS),
KPS ≤50 60–70 80–100 HER2, ER/PR status, and the interaction between ER/PR
Systemic disease status Progressive Stable disease or NED and HER2. The median survival time is 13.8 months.
disease partial remission Survival by GPA scores are as follows: 3.4 months for GPA
Lesions (n) ≥3 2 1 0–1.0, 7.7 months for 1.5–2.0, 15.1 months for 2.5–3.0, and
Largest lesion volume (cm3) >13 5–13 <5 25.3 months for 3.5–4.0. ER/PR positivity is a favorable
From: Lorenzoni J, Devriendt D, Massager N, David P, Ruíz S, prognostic factor, increasing median survival from 6.4 to
Vanderlinden B, et al. Radiosurgery for treatment of brain metastases:
9.7 months for HER2 negative patients and from 17.9 to
estimation of patient eligibility using three stratification systems. Int J
Radiat Oncol Biol Phys. 2004 Sep;60(1):218–24 20.7 months for HER2 positive patients.
KPS Karnofsky performance status, NED no evidence of disease/com- The DS-GPA will continue to evolve as larger cohorts of
plete clinical remission other tumor types are analyzed and our understanding of
disease-specific brain metastases is refined. Given the dispa-
takes into account five factors: age, KPS score, systemic dis- rate findings among different tumor types, diagnosis-specific
ease status, number of lesions, and largest lesion volume GPA provides a more accurate depiction of expected out-
(Table 14.2) [11]. The 0–10 point SIR scale is comparable to comes in patients by tumor type. It should be used to convey
RPA class in predicting outcome but needs to be validated in prognosis to patients and guide development of future clini-
larger studies. cal trials for patients with brain metastases.
The new and current RTOG standard for measuring prog-
nosis is the graded prognostic assessment [12]. This index is
based on 1,960 patients with data extracted from five con- Management of Brain Metastases
secutive RTOG brain metastases trials. The GPA is based on
four criteria: age, KPS score, number of brain metastases, WBRT
and the presence/absence of extracranial metastases. Each of
the four criteria is given a score of 0, 0.5, or 1.0, with the best Brain metastases originally were treated with whole-brain
prognosis in patients with a GPA score of 4.0 (Table 14.3). radiation therapy alone, which was first reported in the 1950s
As our understanding of brain metastases has evolved, it [15, 16]. Response to WBRT is best among patients with
has become increasingly important to distinguish prognosis small cell lung cancer, lymphoma, and germ cell tumors.
based on the primary tumor histopathology. A multi-institu- The RTOG conducted numerous trials from 1971 to 1993 to
tional analysis of 4,259 patients with newly diagnosed brain investigate various fractionation schemes and doses of
metastases from 11 institutions was performed to develop WBRT, which are listed in Table 14.5 [18–24]. In addition,
diagnosis-specific graded prognostic assessment (DS-GPA) the RTOG has investigated the use of hyperfractionation,
[13]. The significant prognostic factors vary by the type of radiation sensitizers such as bromodeoxyuridine, and radia-
primary cancer. For non-small cell lung cancer and small cell tion modifiers such as thalidomide [26] and found no
lung cancer, the significant factors are KPS, age, presence of improvement in overall survival. Based on these studies,
14 Metastatic Brain Tumors 213
Table 14.4 Diagnosis Specific Graded Prognostic Assessment (DS-GPA)

Median survival stratified by diagnosis and diagnosis-specific GPA score for patients with newly diagnosed brain metastases
DS-GPA
Diagnosis Survival in months 0–1.0 1.5–2.5 3.0 3.5–4.0 p (log-rank)
Non small cell lung cancer 7.00 (6.53–7.50) 3.02 (2.63–3.84) 6.53 (5.90–7.10) 11.33 (9.43–13.10) 14.78 (11.79–18.80) <0.0001
Small cell lung cancer 4.90 (4.30–6.20) 2.79 (2.04–3.12) 5.30 (4.63–6.83) 9.63 (7.50–14.95) 17.05 (6.10–27.43) <0.0001
Melanoma 6.74 (5.90–7.57) 3.38 (2.73–4.27) 4.70 (4.17–5.42) 8.77 (6.83–10.77) 13.23 (9.40–15.64) <0.0001
Renal cell carcinoma 9.63 (7.66–10.91) 3.27 (2.17–5.10) 7.29 (3.73–10.91) 11.27 (8.83–14.80) 14.77 (9.72–19.79) <0.0001
Breast cancer 11.93 (9.69–12.85) 6.11 (3.88–8.28) 9.37 (7.92–11.24) 16.89 (13.96–19.90) 18.74 (11.31–29.37) <0.0001
Gastrointestinal cancer 5.36 (4.30–6.30) 3.13 (2.40–4.57) 4.40 (3.37–6.53) 6.87 (5.03–11.63) 13.54 (9.92–27.12) <0.0001
Total 7.23 (6.90–7.60) 3.43 (3.02–3.84) 6.40 (5.78–6.90) 11.56 (10.47–12.78) 14.77 (12.85–17.05) <0.0001
Adapted from: Sperduto PW et al. Diagnosis-specific prognostic factors, indexes, and treatment outcomes for patients with newly diagnosed brain
metastases: a multi-institutional analysis of 4,259 patients. Int J Radiat Oncol Biol Phys. 2010;77(3):655–61
Data in parentheses are 95 % confidence intervals
Table 14.5 Prospective Radiation Therapy Oncology Group (RTOG) brain metastases studies (1971–1993)
Protocol (Reference) Year No. of patients Total dose/No. of fractions Median survival
RTOG 6901 [18] 1971–1973 233 30 Gy/10 Fx/2 weeks 21 Weeks
217 30 Gy/15 Fx/3 weeks 18 Weeks
RTOG 7361 [19] 1973–1976 447 20 Gy/5 Fx/1 week 15 Weeks
RTOG 6901 [20] 1971–1973 26 10 Gy/1 Fx/1 day 15 Weeks
RTOG 7361 [20] 1973–1976 33 12 Gy/2 Fx/2 days 13 Weeks
Ultra-Rapid
RTOG 7606 [21] 1976–1979 130 30 Gy/10 Fx/2 weeks 18 Weeks
Favorable pts. 125 50 Gy/20 Fx/4 weeks 17 Weeks
RTOG 8528 [10] 1986–1989 30 48 Gy/1.6 Gy bid 4.8 Months
53 54.5 Gy/1.6 Gy bid 5.4 Months
44 64 Gy/1.6 Gy bid 7.2 Months
36 70.4 Gy/1.6 Gy bid 8.2 Months
RTOG 9104 [22] 1991–1995 213 30 Gy/10 Fx 4.5 Months
216 54.4 Gy/1.6 bid 4.5 Months
RTOG 7916 [23] 1979–1983 193 30 Gy/10 Fx/2 weeks 4.5 Months
200 5 Gy/6 Fx/3 weeks 4.1 Months
Misonidazole 196 30 Gy/10 Fx + Miso 3.1 Months
190 5 Gy/6 Fx + Miso 3.9 Months
RTOG 8905 [24] 1989–1993 36 37.5 Gy/15 Fx/3 weeks 6.1 Months
BrdU 34 37.5 Gy/15 + BrdU 4.3 Months
Adapted from Sneed PK, Larson DA, Wara WM. Neurosurg Clin N Am 1996;7:505–515
Fx fraction, bid twice daily, Miso misonidazole, BrdU bromodeoxyuridine
30 Gy in 10 fractions became a popular fractionation scheme Surgery and WBRT

for patients with brain metastases. Although neurologic
symptoms improve in the majority of patients receiving Given the low local control rates associated with WBRT
WBRT, local control is low, resulting in neurologic death in alone, surgical removal of tumors—particularly for single or
25–54 % of patients [27]. symptomatic lesions—was explored in hopes of improving
local control and survival. Surgery can provide immediate
and effective palliation of symptomatic mass effect. Another
major advantage of surgery is establishing or confirming the
Table 14.6 Phase III trials of WBRT vs. surgery + WBRT

WBRT dose(cGy)/No. MS KPS >70 CNS Local
Author Year Surgery of fractions/No. of weeks N (weeks) (weeks) death (%) control
Patchell [28] 1990 Yes 3600/12/2.5 weeks 25 40 38 29 80 %
No 3600/12/2.5 weeks 23 15 8 52 48 %
Noordijk [30] 1994 Yes 4000/20/2 weeks 32 43 33 35 N/A
No 4000/20/2 weeks 31 26 15 33 N/A
Mintz [29] 1996 Yes 3000/10/2 weeks 41 22 N/Aa 14 N/A
No 3000/10/2 weeks 43 25 N/Aa 28 N/A
WBRT whole brain radiation therapy, N number of patients, MS median survival, N/A not applicable
N/Aa, the mean proportion of days that the patients had KPS >70 was not different
diagnosis, especially for patients without primary tumor with secondary endpoints of survival, cause of death, and
pathologic diagnosis. Although one would expect the functional independence. The addition of WBRT to surgery
probability to be low with advances in MR brain imaging, a decreased the chance for neurologic death (14 % vs. 44 %,
small percentage of patients undergoing surgical resection of p = 0.003), decreased local recurrence (10 % vs. 46 %,
a presumed single brain metastasis seen on imaging may not p < 0.01), and decreased tumor recurrence anywhere in the
actually have a brain metastasis, for which adjuvant WBRT brain (18 % vs. 70 %, p < 0.001). The majority of patients
would not be warranted. This percentage was as high as (61 %) received WBRT at time of recurrence resulting in a
11 % in the past [28]. large crossover of the observation arm to WBRT. Survival
Three randomized trials comparing WBRT alone vs. sur- was not different, although the study’s endpoint was local
gery followed by WBRT for patients with single metastasis control and was not powered to demonstrate a survival
have been completed and are summarized in Table 14.6 [28– advantage [32]. This second Patchell study established the
30]. These trials were based on the premise that improved value of adjuvant radiation therapy in management of a sin-
local control of a single brain metastasis would result in gle brain metastasis.
improved survival. Patchell et al. demonstrated that surgery The type of surgical resection may also influence local
and WBRT improved survival (40 vs. 15 weeks, p < 0.01), recurrence of brain metastases. A retrospective review of 570
local control (80 % vs. 48 %, p < 0.02), and functional inde- patients from MD Anderson Cancer Center treated with sur-
pendence (38 vs. 8 weeks, p < 0.005) compared with biopsy gical resection alone demonstrated that piecemeal resection
and WBRT [28]. This is the sentinel study that established a predicts for higher local recurrence compared with en bloc
clear advantage to surgically resecting a single lesion before resection, with a hazard ratio of 1.7 (p = 0.03) [33]. Size was
administering whole brain radiation therapy. Another study also predictive, with tumor volume exceeding 9.7 cm3 having
by Noordijk included 63 patients with CT-confirmed single a hazard ratio of 1.7 (p = 0.02) compared with smaller tumors.
metastasis who were randomized to surgery and WBRT vs. Histologic subtype did not predict for local recurrence in this
WBRT alone [30]. Surgery and WBRT improved survival cohort of patients.
and functional independence over WBRT alone (10 vs. 6
months (p = 0.04) and 7.5 vs. 3.5 months (p = 0.06), respec-
tively). Patients with active systemic disease, however, did Stereotactic Radiosurgery
not benefit from surgery. The third trial by Mintz did not
demonstrate a survival benefit for patients undergoing sur- Stereotactic radiosurgery is a technique that delivers a large
gery (6.3 vs. 5.6 months, p = 0.24) [29]. This was most likely dose of ionizing radiation in a single or few fractions using
a result of two factors: (1) the study consisted of a large per- highly focused narrow beams to small intracranial targets
centage of patients with active systemic disease and (2) while sparing surrounding brain tissue [34]. Interest devel-
patients with lower baseline KPS were included. oped in using SRS rather than surgery to improve local con-
Due to the improved outcomes with surgical resection of trol. Brain metastases have features that make them ideal
brain metastases, a second study by Patchell randomized targets for SRS. They are usually spheroid in shape, located
patients to surgery alone vs. surgery with WBRT to deter- in the gray-white junction, and have a maximum diameter of
mine if adjuvant radiation therapy is necessary in the man- less than 4 cm. Perhaps more importantly, brain metastases
agement of single brain metastasis [31]. In this phase III trial, are noninfiltrative, as opposed to primary gliomas. These
95 patients with a single metastasis status post gross total characteristics allow for accurate target delineation, plan-
resection were randomized to WBRT or no further treatment. ning, and treatment delivery. A single large fraction of radia-
The primary endpoint was tumor recurrence in the brain tion appears to have an equal effect in all tumor types, even
among radioresistant tumors such as renal cell carcinoma important prognostic factors. The median survival was
and melanoma [35–38]. Ever since Sturm’s initial report of 7.5 month for the SRS group and 16.4 months for the surgery
12 lesions treated on a modified linear accelerator, a number group. The local recurrence for SRS was 21 % vs. 8 % for
of papers have corroborated the benefit of SRS in newly surgery. They concluded surgery should be the preferred
diagnosed and recurrent brain metastases [39]. treatment for surgically accessible lesions. However, careful
The use of SRS alone or as an adjunct to whole brain analysis of the SRS dosing strategy in this study indicates
radiation therapy is based on the premise that improved local that relatively low tumor margin doses were used, well below
control will reduce morbidity, improve quality of life, and the doses used in RTOG studies. A study from the Mayo
prolong survival. Nieder et al. reviewed the CT scans of 332 Clinic compared outcomes of radiosurgery vs. surgery and
patients with brain metastases to evaluate local control and found no difference in 1-year survival between the groups
time to local progression based on dose [40]. A biologically [47]. This retrospective review included patients with tumors
effective dose (BED) using the linear quadratic model with who would have allowed for either type of treatment.
an alpha/beta of ten was derived for each patient. Partial A retrospective study from Muacevic compared sur-
response rates significantly improved for patients with higher gery + WBRT vs. Gamma Knife radiosurgery (GKRS) alone
BED suggesting that larger doses were needed to improve for patients with a single tumor ≤3.5 cm in diameter [48].
local control. The 1-year survival rates (53 % vs. 43 %, p = 0.19), 1-year
Brain tolerance to radiation therapy is related to total local control rates (75 % vs. 83 %, p = 0.49), and 1-year neu-
dose, dose per fraction, elapsed time of radiation delivery, rologic death rates (37 % vs. 39 %, p = 0.8) for the sur-
volume of normal brain irradiated, and the time interval from gery + WBRT group and SRS group, respectively, were not
prior radiation therapy [41]. From a radiobiological stand- statistically different.
point, brain metastases are considered category IV targets, Since all of the previously described studies are retro-
which are targets with early-responding tissue surrounded by spective, patient selection bias was probably a key contribu-
late-responding normal tissue [42]. Since the tumor cells are tor to the mixed results and controversy regarding the benefits
within the target, one would expect a high therapeutic index of surgery vs. SRS. The Joint Center for Radiation Therapy
for these targets given the steep dose gradient associated attempted to enroll patients onto a phase III trial comparing
with SRS. Unlike surgery, SRS also has the potential to ster- surgery to SRS in the 1990s. After 3 years, only six patients
ilize surrounding cells outside the high dose region. were enrolled due to patient or physician preference [49].
A prospective randomized trial from Germany compared
GKRS alone vs. surgery + WBRT for patients with a single
Surgery vs. SRS metastasis [50]. The study was closed early due to poor
accrual, with 33 patients in the surgery + WBRT arm and 31
Based on the results of surgical resection for a single metas- patients in the GKRS alone arm. Overall survival, neurologic
tasis, investigators explored the use of SRS for brain metas- death, and local recurrence were equivalent. Patients treated
tases. This interest was based on potential advantages of with GKRS developed more distant recurrences (p = 0.04);
SRS, including outpatient delivery, no general anesthesia use however, this difference was lost after adjusting for salvage
leading to shorter recovery time, minimal risk for bleeding or GKRS. GKRS patients had shorter hospital stay, less steroid
infection, and lower cost. In addition, SRS is not limited by treatment, and lower CNS Grade 1 and 2 toxicities.
location, number of lesions, comorbidity, or coagulopathy. M.D. Anderson performed a prospective randomized trial
For these various reasons, SRS has become a viable alterna- of surgery vs. radiosurgery for patients with a single brain
tive to surgery given the potential cost savings and quality- metastasis who are considered eligible for either treatment
of-life benefits [43, 44]. [51]. Fifty-nine patients were entered into the randomized arm
A multi-institutional retrospective review from the (30 underwent surgery and 29 SRS), and 155 patients were
University of Wisconsin, University of Florida, and Joint entered on the nonrandomized arm (89 chose surgery and 66
Center for Radiation Therapy included 122 patients with a chose SRS). Although the results demonstrating increased
newly diagnosed, potentially resectable single brain metasta- local recurrence with SRS alone have been presented in
sis treated by SRS with WBRT [45]. The local control rate abstract form, the statistical manipulations used in this study
was 86 %. The median survival was 56 weeks with 25 % of are controversial and the results are yet to be published.
deaths related to brain tumor progression. Based on results For patients with large resectable tumors or those without
of this retrospective review, the investigators concluded that histologic confirmation of a primary tumor, most clinicians
SRS could be delivered in place of surgery. would agree that surgical resection should be considered
Another study from M.D. Anderson Cancer Center com- rather than radiosurgery. Surgical removal of the tumor is
pared 13 patients treated with SRS and 62 patients treated by more effective in alleviating mass effect, offers faster
surgery [46]. These patients were retrospectively matched for improvement in neurologic function, and minimizes chronic
steroid therapy. Prolonged use of steroids can result in a

number of problems including diabetes, proximal muscle Radiosurgery with WBRT
weakness, peripheral edema, psychosis, susceptibility to
infections, and gastrointestinal perforation [52]. Conversely, Based on the early results of SRS for recurrent brain metas-
most would agree that small, deep lesions are best treated tasis, SRS with whole brain radiation therapy was investi-
with radiosurgery. To date, no randomized trial pitting radio- gated for newly diagnosed patients with the hypothesis that
surgery vs. surgery of comparable brain metastases has been the combination of WBRT with SRS would improve local
completed. and regional control of brain metastases. A number of studies
have demonstrated high local control rates with the addition
of WBRT to SRS [54, 58–66]. These local control rates
SRS for Recurrent Metastatic Disease ranged from 63 to 97 % with addition of WBRT vs. 42 to
87 % with SRS alone [36, 64, 66–68].
SRS for brain metastases was initially implemented in the RTOG 9005 is an important phase I/II trial which deter-
salvage setting. One of the first reports of SRS for brain mined the maximum tolerated radiosurgery dose for patients
metastases consisted of patients with recurrent brain meta- with recurrent primary brain tumors or brain metastases
stases after WBRT or surgery. Loeffler et al. reported the treated previously with fractionated radiation therapy [69]. In
Joint Center for Radiation Therapy retrospective results of this trial, 80 % of patients were treated on a linear accelerator
18 patients with recurrent metastases after previous surgery, (LINAC) and 20 % were treated with Gamma Knife radiosur-
WBRT, or both [53]. The majority of patients improved neu- gery. The maximum tolerated doses were inversely corre-
rologically after SRS with local control being achieved in all lated with the maximum tumor diameter. The doses were
patients. A follow-up report by Alexander et al. of patients 24 Gy for a tumor <20 mm in diameter, 18 Gy for 21–30 mm,
mostly with recurrent brain metastases reported tumor con- and 15 Gy for tumors 31–40 mm in diameter (Table 14.7). Of
trol rates of 85 % at 1 year and 65 % at 2 years, although the note, investigators were reluctant to escalate the dose for
control rates were lower for recurrent lesions [54]. Median tumors <20 mm in diameter even though the maximum toler-
survival was 9 months after SRS in that study. ated dose was not reached. The rates of radiation necrosis
The University of Cincinnati reported the results of 84 were 5 %, 8 %, 9 %, and 11 % at 6, 12, 18, and 24 months
patients with 1–6 lesions who developed recurrence after after SRS, respectively. Other factors that predicted for Grade
WBRT [55]. Median survival was 43 weeks after 3–5 neurotoxicity were tumor dose and KPS. Table 14.8 lists
SRS. Median time to local failure was 35 weeks for all the RTOG CNS toxicity criteria used for RTOG 9005. The
lesions and 52 weeks for lesions treated with ≥18 Gy. A study also reported on physics and quality control assess-
study by Noël et al. of 54 consecutively treated patients ments: MD/PD ratio, a measure of dose homogeneity, and
reported 1- and 2-year local control rates of 91.3 % and 84 % PIV/TV ratio, a measure of conformity of the treated volume
and 1- and 2-year brain control rates of 65 % and 57 %, relative to target volume [70]. The results of RTOG 9005 are
respectively [56]. Cleveland Clinic published their experi- the standard for determining dosing of SRS for brain metas-
ence of 111 patients who underwent initial WBRT followed
by SRS as salvage treatment [57]. The median overall sur- Table 14.7 RTOG 9005 dosing guidelines for SRS
vival from the initial diagnosis of brain metastasis was
Size (mm) Dose to tumor periphery (Gy)
17.7 months. Median survival after salvage SRS was
≤ 20 24
9.9 months. Median survival after salvage SRS was
21–30 18
12.3 months in patients who had their first recurrence
>30 15
>6 months after WBRT vs. 6.8 months for those who devel-
oped disease recurrence ≤6 months after (p = 0.0061).
Primary site did not affect survival. Twenty-eight patients Table 14.8 RTOG CNS toxicity criteria used for RTOG 9005 [76]
(25 %) developed local recurrence after salvage SRS with a Grade 1 Mild neurologic symptoms; no medication required
median time of 5.2 months. A dose <22 Gy and lesion size Grade 2 Moderate neurologic symptoms; outpatient medication
>2 cm were found to be predictive of local failure. In this required (e.g., steroids)
Grade 3 Severe neurologic symptoms; outpatient or inpatient
study, patients who recurred after WBRT and were treated
medication required
with salvage SRS were found to have good local control and Grade 4 Life-threatening neurologic symptoms (e.g., uncontrolled
survival after SRS. WBRT provided good initial control, as seizure, paralysis, coma); includes clinically and
45 % of these patients failed >6 months after WBRT. Those radiographically suspected radiation necrosis and
with a longer time to failure after WBRT had significantly histologically proven radiation necrosis at time of operation
longer survival after SRS. Grade 5 Death
tases based on size and have been incorporated into subse- WBRT + SRS, and WBRT arms, respectively). The local
quent studies. control rates were 87 %, 91 %, and 62 %, respectively. The
Shehata et al. evaluated the optimal SRS dose and influ- risk for developing new brain lesions was higher for the
ence of WBRT on tumor control among 160 patients with patients who did not receive WBRT (43 % SRS alone vs.
468 recurrent and newly diagnosed metastases <2 cm in 19 % WBRT + SRS, vs. 23 % WBRT alone). However, the
diameter [66]. On multivariate analysis, the most important study had major treatment bias given the unequal distribution
factor for local tumor control was the addition of WBRT to of patients undergoing surgical resection in the three arms.
SRS (97 % vs. 87 % for SRS alone; p = 0.001). For patients The RTOG performed a large prospective trial (RTOG
undergoing WBRT, doses >20 Gy resulted in higher Grade 3 9508) of 333 patients with 1–3 newly diagnosed brain metas-
and 4 neurotoxicity (5.9 % vs. 1.9 %) and did not result in tases on MRI, each <4 cm in diameter, randomized to WBRT
better local control. Thus, the authors recommended using (3,750 cGy in 15 fractions) and SRS vs. WBRT alone [75].
20 Gy for metastases <2 cm with combined WBRT. Patients had a minimum KPS of 70 and were excluded if a
The largest SRS dosing study is a single institution analy- metastasis was located in the brainstem or within 1 cm from
sis reported by Valery et al. on 377 patients with 760 lesions the optic nerves or chiasm. Survival significantly improved
treated with LINAC-based SRS. Seven patients had severe for patients with a single brain metastasis treated with
complications including nine patients who developed radia- SRS + WBRT vs. WBRT (6.5 months vs. 4.9 months,
tion necrosis. The median tumor volume was 4.9 cm3, and p = 0.04) despite the fact that 19 % of patients failed to
median prescribed tumor dose was 15.6 Gy. The only factor receive the intended treatment. On multivariate analysis,
that influenced the risk for radiation necrosis was the confor- RPA class (1 vs. 2) and tumor type (squamous or non-small
mality index [71]. Age, KPS, homogeneity index, tumor vol- cell vs. other) had a statistically significant effect on survival.
ume, and delivered dose were not significant. This study Overall survival did not improve for the patients with 2–3
highlights the importance of achieving a highly conformal lesions although KPS at 6 months had stabilized or improved
plan for LINAC-based radiosurgery. with the addition of SRS + WBRT vs. WBRT alone (43 % vs.
A multi-institutional retrospective study from ten institu- 27 %, respectively; p = 0.03). Steroid use was also lower in
tions analyzed outcomes for 502 patients with newly diag- the SRS + WBRT arm. Local control was improved at 1 year
nosed brain metastases treated by SRS and WBRT. Patients (82 % vs. 71 %; p = 0.01). The risk of developing local recur-
were stratified by RTOG RPA class [72]. For all RPA classes, rence was 43 % greater with WBRT alone (p = 0.0021). The
survival improved with the addition of SRS boost compared type of SRS unit did not influence results.
with WBRT alone. Median survival for RTOG RPA class I, A retrospective study from Cleveland Clinic of 202
II, and III were 16, 10, and 8 months for the SRS + WBRT patients that met study entry criteria of RTOG 9005 analyzed
patients vs. 7.1, 4.2, and 2.3 months for the WBRT alone local control rates based on differences in prescription dose
patients, respectively. This study suggests a survival benefit [76]. With a median follow-up duration of 10.7 months, a
to the addition of SRS for all patients. dose of 24 Gy to the tumor margin had significantly lower
Three prospective trials compared the use of SRS + WBRT local failure than 18 Gy or 15 Gy (hazard ratio 0.277,
vs. WBRT alone. The first trial from the University of p = 0.0005). The 18 and 15 Gy groups exhibited no statistical
Pittsburgh enrolled 27 patients with two to four brain metas- difference in local control. The 1-year local control rates by
tases measuring <2.5 cm in maximum diameter [73]. The dose were 85 % for 24 Gy, 49 % for 18 Gy, and 45 % for
study was stopped at an interim analysis at 60 % since a sig- 15 Gy. The improved control rates with 24 Gy suggests that
nificant difference in the 1-year local failure rate was noted small tumors ≤2 cm are better controlled than larger tumors,
(8 % in the SRS + WBRT arm vs. 100 % in the WBRT arm). as one would expect.
The median time to local failure was 6 months for WBRT vs.
36 months for the SRS + WBRT arm. Median survival was
not significantly different for the SRS + WBRT vs. WBRT Radiosurgery Alone
arm (11 months vs. 7.5 months, respectively). The small
sample size, as well as possible selection and statistical bias, Despite evidence supporting the use of WBRT for patients
may have affected the results. with single brain metastasis, SRS alone for treatment of
Brown University enrolled 96 patients onto a 3-armed, brain metastases has gained popularity and become a con-
prospective randomized trial of Gamma Knife radiosurgery, tentious issue. The use of SRS alone has been mostly driven
WBRT, or both [74]. Fifty-one patients underwent surgical by concerns regarding quality-of-life issues and the poten-
resection of large symptomatic lesions prior to randomiza- tial side effects of WBRT, especially cognitive effects for
tion, which was not evenly distributed among the treatment long-term survivors [77, 78]. One of the early and com-
arms. No difference in overall survival was noted among the monly referenced papers regarding cognitive decline in
three study arms (7, 5, and 9 months for the SRS alone, WBRT is by DeAngelis et al., which reported complications
in patients treated with fractionation schemes of occurred, but learning and memory skills declined. At 7–9
300–600 cGy/fraction to a total dose of 25–30 Gy [77]. Of months after SRS, there was a suggestion of global decline
these 12 patients identified, all developed progressive of neurocognitive function.
dementia, ataxia, and urinary incontinence, causing death in The RTOG reported on the feasibility of performing five
seven patients. Although the incidence of WBRT-induced neurocognitive measures and administering a quality-of-life
dementia was 1.9–5.1 % overall, it is likely that this is an instrument in patients with brain metastases [84]. The overall
underestimate of true incidence because only severely compliance rate for administration and completion of these
affected patients were identified from chart review. Another measures and instruments was >95 % prior to treatment,
argument to treating with SRS alone is that some physicians >84 % at the completion of WBRT, and >70 % at 1 month
believe that patients can be effectively salvaged by repeat after WBRT. Based on the encouraging results of this study
SRS rather than WBRT [79]. and others of similar design, future studies have more rou-
It is important to note that patients undergoing SRS alone tinely incorporate baseline and follow-up neuropsychometric
may be at higher risk for morbidity associated with brain testing as an inherent part of study design and patient out-
tumor recurrence. A retrospective study from the University come evaluation. It is becoming increasingly evident that
of Kentucky reviewed 36 patients treated with SRS alone neurologic/neurocognitive decline seen in cancer patients is
with newly diagnosed brain metastases of which 22 patients multifactorial and not exclusive to the effects of radiation
had a single lesion [80]. Seventeen of the 36 patients (44 %) therapy alone. This has been demonstrated in recent publica-
developed recurrent brain metastases, with 12 patients (71 %) tions implicating the potential significant impact of other
symptomatic and 10 (59 %) having neurologic deficits. treatment modalities such as chemotherapy and/or surgery
Several prospective trials have assessed whether patients on neurologic/neurocognitive function [85–88].
with brain metastases have neurocognitive deficits prior to
initiation of treatment. A recent prospective phase III trial
evaluated patients with brain metastasis with monthly neuro- Large Institutional/Multi-Institutional
cognitive testing consisting of memory, fine motor speed, Results of SRS Alone
executive function, and global neurocognitive testing [81].
Table 14.9 lists the neurocognitive tests that were performed A UCSF retrospective study by Sneed et al. reported
on this study. This trial demonstrated that 90.5 % of patients outcomes of patients with one to four metastases treated by
had impairment of one or more neurocognitive tests at base- SRS alone compared to WBRT plus SRS [68]. Physician
line and 42.4 % of patients had impairment in four or more preference and referral patterns influenced the use of
tests. Another study from the RTOG demonstrated that tumor upfront WBRT. The patients had similar median survivals
progression resulted in a significant decline in the mini- (11.1 months for SRS alone vs. 11.3 months for WBRT + SRS)
mental status exam (MMSE) after WBRT [82]. and local tumor control rates (71 vs. 79 %). The incidence
Chang et al. reported a prospective trial of neurocognitive of distant brain metastasis, however, was significantly higher
testing for patients with 1–3 newly diagnosed brain metasta- in the SRS only group vs. WBRT + SRS (72 vs. 31 %).
ses to determine if neurocognitive function was spared with Despite the higher rate of distant brain metastases, these
SRS alone [83]. Eighteen of the 27 enrolled patients had lesions were controlled with salvage therapies including
evaluable neurocognitive function data. At baseline, 66 % of WBRT, partial brain radiotherapy, SRS, and surgery.
patients had some degree of impairment that was related to The authors concluded that survival was not compromised
total brain metastases volume. After SRS, acute and subacute by the omission of WBRT.
improvements in executive function and processing speed Another retrospective paper from Hasegawa et al.
reviewed 172 patients with brain metastases (3.5 cm or less
in diameter) managed by SRS alone [89]. One hundred
twenty-one patients had follow-up imaging with 80 % of
Table 14.9 Neurocognitive testing used for phase III brain metastases patients having solitary lesions. The median survival was 8
trial [62] months, and the local tumor control rate was 87 %. At 2
Test Neurocognitive domain years, the local control rate, distant brain control, and total
Hopkins Verbal Learning Test (recall) Memory intracranial control rates were 75 %, 41 %, and 27 %, respec-
Hopkins Verbal Learning Test (recognition) Memory tively. Tumor volume significantly predicted for local con-
Hopkins Verbal Learning Test (delay) Memory trol (p = 0.02) with tumor volumes of at least 4 cm3 having
Trailmaking A Executive local control rates of 49 % at 1 and 2 years. Based on these
Trailmaking B Executive results, the authors advocated withholding WBRT for select
Controlled oral word association Executive patients with one or two tumors with good control of their
Grooved pegboard dominant hand Fine motor primary cancer, better KPS (90 or higher), and younger age
Grooved pegboard nondominant hand Fine motor (<60 years old).
One study prospectively evaluated the role of SRS alone The Japanese Radiation Oncology Study Group (JROSG)
for patients with up to three brain metastases measuring conducted a phase III study of SRS vs. SRS plus WBRT
<3 cm in maximum diameter [90]. The study consisted of (3,000 cGy/10 fractions) [93]. This trial randomized patients
101 patients with a total of 155 lesions. At 1 year, the local with one to four brain metastases <3 cm in diameter to
control, distant brain freedom from progression, and overall SRS + WBRT (65 patients) or SRS alone (67 patients). The
brain freedom from progression rates were 91 %, 53 %, and primary endpoint of the study was overall survival with the
51 %, respectively. The best predictor for improved overall secondary points including brain recurrence, salvage brain
brain freedom from progression was the presence of a single treatment, functional preservation, toxic effects of radiation,
lesion and a greater than 2-year interval from primary diag- and cause of death. At 12 months, WBRT decreased brain
nosis to diagnosis of brain metastases. The overall survivals tumor recurrence from 76.4 to 46.8 % (p < 0.001). Freedom
stratified by RTOG RPA class were 13.4, 9.3, and 1.5 months from new brain metastasis was also significantly better at 1
for class I, II, and III, respectively (p < 0.0001). Median year for the WBRT and SRS group (63.7 %) compared with
survival was 7.6 months. These results suggested that SRS the SRS alone group (41.5 %) (p = 0.003). Consequently, sal-
alone can provide good local control but that distant failure vage treatment was more frequently required in the SRS
can be expected. alone group. No significant difference in overall survival,
Another study from Pirzkall et al. reviewed an experience toxic effects of radiation, death due to neurologic causes, or
of 236 patients with 311 brain metastases who underwent differences in systemic or neurologic functional preserva-
SRS alone (158 patients) or SRS with WBRT (73 patients) tion were seen.
[67]. No difference was noted in terms of survival or local A prospective randomized trial for patients with 1–3
control. The SRS with WBRT group had a trend toward bet- lesions comparing SRS plus WBRT vs. SRS alone with the
ter survival compared with SRS alone (15.4 vs. 8.3 months; primary endpoint of neurocognitive function measured by
p = 0.08) for the patients with no evidence of extracranial dis- the Hopkins Verbal Learning Test-Revised (HVLT-R) was
ease. The local control and distant brain control rates were conducted by Chang et al. at MD Anderson [94]. After
worse with SRS alone. accrual of 58 patients, the trial was stopped early based on
A multi-institutional retrospective review from ten insti- the 96 % probability that the SRS + WBRT arm would show
tutions analyzed the results of 569 patients with newly diag- a statistically significant decline in learning and memory
nosed single or multiple brain metastases treated initially function (total recall) at 4 months. Congruous to the Japanese
with SRS alone vs. SRS and WBRT [91]. For all RPA classes, study, there were more CNS recurrences in the group receiv-
survival was comparable between the two treatment groups ing SRS alone; 73 % of patients in the SRS + WBRT group
(RPA class I: 14 vs. 15 months, RPA class II: 8 vs. 7 months, were free from CNS recurrence at 1 year, compared with
RPA class III: 5 vs. 6 months) suggesting that upfront WBRT 27 % of patients who received SRS alone (p = 0.0003). The
does not improve survival over SRS alone. authors conclude that SRS with close follow-up is the pre-
ferred strategy. However, this study did not provide informa-
tion regarding long-term impact of WBRT or distant tumor
Phase II/III Trials of SRS With progression on neurocognitive function.
or Without WBRT The EORTC completed a prospective phase III trial com-
paring the addition of WBRT to initial surgery or SRS for
The Eastern Cooperative Oncology Group (ECOG) phase II patients with stable systemic disease and up to three brain
trial of SRS alone for patients with renal cell carcinoma, sar- metastases [95]. The primary endpoint of this study was time
coma, and melanoma (radioresistant tumors) with one to to WHO performance status >2. There were no differences in
three brain metastases was conducted with a primary end- overall survival or time to performance status deterioration
point of intracranial progression [92]. The purpose of this in either group. WBRT reduced the 2-year relapse rate at the
trial was to evaluate the feasibility of SRS alone in manage- initial tumor site (surgery: 59 % vs. 27 %, p < 0.001; SRS:
ment of brain metastases before pursuing a phase III study. 31 % vs. 19 %, p = 0.04) and reduced new brain mets (sur-
This trial enrolled 36 patients; 31 were eligible by entry cri- gery: 42 % vs. 23 %, p = 0.008; SRS: 48 % vs. 33 %,
teria. The SRS dose was based on tumor size (24, 18, or p = 0.023) in both groups. Furthermore, neurologic deaths
15 Gy). The median survival was 8.3 months with a median were higher in the arms not receiving WBRT (44 % vs. 28 %,
follow-up of 32.7 months. The 3- and 6-month intracranial p < 0.002). As expected, salvage WBRT was more frequent in
failures with SRS alone were 25.8 % and 48.3 %, respectively. the arms not receiving WBRT initially. An update to this trial
Approximately 38 % of patients experienced death attribut- assessed health related quality-of-life (QOL) measures,
able to neurologic cause. The authors concluded that which was a secondary endpoint of the trial [96]. Patients in
although survival was equivalent to published reports of the observation arm reported higher QOL compared with the
SRS + WBRT, the high rate of intracranial failures suggests a arm receiving WBRT. However, compliance was poor with
judicial approach to omission of WBRT. 45 % of patients completing QOL assessment at 1 year.
The North Central Cancer Treatment Group (NCCTG) and clinical multiple brain cases treated with the Gamma
recently completed enrollment of a phase III trial (N0574) Knife Perfexion were used to derive the model. The total num-
comparing SRS vs. SRS followed by WBRT delivered within ber of targets ranged from 2 to 60. A significant increase in
14 days for patients with 1–3 brain metastases. The primary volume was noted for the 12 Gy peripheral isodose volume,
endpoint is neurocognitive function. which was strongly correlated with number of lesions but not
total target volumes. The 12 Gy target volume for multiple tar-
gets was increased by approximately 1 % per target when
SRS for More than Four Brain Metastases 15 Gy was used vs. 4 % per target increase when 20–24 Gy
was used. Based on these results, a reduction in dose of 1–2 Gy
With the success of SRS in treating patients with 1–4 brain is recommended when treating multiple lesions.
metastases and efforts to eliminate the effects of WBRT on
neurocognitive functioning, there has been increasing use of
SRS alone for patients with multiple (≥5) lesions. SRS for Brainstem Metastases
Previously, there had been concern regarding the cumula-
tive dose to the whole brain when treating many lesions with Metastases located in the brainstem portend a poor prognosis
SRS. However, Yang et al. reported that 50 % of the brain and frequently present a challenge in clinical management.
received less than 5 Gy when a maximum tumor dose of Surgical resection is rarely a viable option due to high risk of
40 Gy was used when treating 25 metastatic intracranial developing new or worsening neurological deficits. Patients
tumors [97]. A more recent study by Yamamoto et al. with brainstem metastases often present with focal neuro-
reviewed the median cumulative dose to the whole brain for logic signs and symptoms due to anatomic location and local
patients with at least ten lesions who were treated by GKRS mass effect, for which stereotactic radiosurgery would pro-
[98]. The median number of lesions treated with SRS was 17 vide better symptomatic outcomes than whole-brain radia-
(range: 10–43). The median volume for all tumors was tion therapy (WBRT) alone. However, stereotactic
8.02 cm3 (range: 0.46–81.41). The median prescribed dose radiosurgery to the brainstem presents unique consider-
was 20 Gy (range: 12–25 Gy). The median cumulative dose ations, mainly due to concerns of toxicity to critical struc-
to the whole brain was 4.71 Gy (range: 2.16–8.51 Gy). The tures and potential for brainstem necrosis. To date, there are
median brain volumes receiving >10 Gy, 15 Gy, and 20 Gy no prospective trials for brainstem metastasis SRS and no
were 64 cm3, 24 cm3, and 8 cm3, respectively. established guidelines regarding dosing. However, there are
With further assurance about the safety and efficacy of a number of retrospective single institution reviews in the
SRS to multiple lesions, there have been more recent studies. literature with favorable outcomes with limited toxicity
The local control and median survival of these studies are [108–119]. These results are summarized in Table 14.11.
summarized in Table 14.10 [99–106]. The North American Based on these studies, tumor margin dose ranges from
Gamma Knife Centers is currently proposing a trial of SRS 13 to 25 Gy, with Grade 3/4 toxicity of 0–11.5 %. Median
alone for patients with five or more metastases. survival for patients with treated brainstem metastases is
Since existing dose guidelines for SRS are based on 1–3 4–12 months, with a local control of 76—100 %. These stud-
lesions, a physical model was developed to correlate the ies demonstrate that SRS for brainstem metastases is an
peripheral isodose volume dependence on prescription dose effective strategy for treatment.
per target and increasing number of isocenters [107]. Simulated
Table 14.10 SRS for multiple brain metastases

Median
Number Median/mean Range (number of survival
Study Year of patients no. of brain mets brain metastases) Tumor type Local control (%) (months)
Suzuki [99] 2000 24 20 (mean) 10–47 All NR NR
Bhatnagar [100] 2006 205 5 (median) 4–18 All 71 % (1-year) 8
Kim [101] 2008 26 16.6 (mean) 10–37 All 79.5 (6 months) 7.85
Park [102] 2009 33 5.9 (median) 2–20 Lung NR 6.15
Yamamoto [103] 2009 456 6 (mean) 1–55 Non-lung NR 7
Nath [104] 2010 26 5 (median) 2–13 All 97 % (6-month) NR
Hunter [105] 2012 64 6 (median) 5 or more All NR 7.5
Grandhi [106] 2012 61 13.2 (median) 10–28 All 81 % 4
14
Metastatic Brain Tumors
Table 14.11 Review of the literature: stereotactic radiosurgery for brainstem metastases
Median Median 1-year
tumor tumor Median Median actuarial
No. of No. of SRS Median volume margin follow-up survival overall Symptomatic Local Grade 3–4
Study Year patients lesions WBRT (%) modality age (cm3) dose (Gy) (months) (months) survival (%) improvement control (%) toxicity (%)
Koyfman et al. [108] 2010 43 43 34 (79.1) GK 59 0.37 15 5.3 5.8 32.5 n.a. 85 0
Lorenzoni et al. [110] 2009 25 27 17 (68) GK 53 0.6 20 10.5 11.1 37 n.a. 95 0
Kased et al. [111] 2008 42 44 24 (57) GK 55 0.26 16 6.9 9 31 1/10 (10 %) 77 9.5
Hussain et al. [112] 2007 22 25 3 (13.6) GK 60 0.9 16 8.5 8.5 32 2/22 (9 %) 100 5
Fuentes et al. [113] 2006 28 28 6 (21.4) GK 57.5a 2.1 19 11 12 37 16/24 (66 %) 92 0
Yen et al. [114] 2006 53 53 21 (39.6) GK 57.3a 2.8 18 9.8 11 42 21/35 (60 %) 86.5 0
Shuto et al. [115] 2003 25 31 9 (36) GK 54 2.1 13 5.2 4.9 50 n.a. 77.4 n.a.
Huang et al. [116] 1999 26 27 24 (92.3) GK 62 1.1 16 9.4 9 21.7 10/20 (50 %) 95 11.5
Hatiboglu et al. [117] 2011 60 60 9 (15) LINAC 61 1 15 5.3 4 n.a. 49/60 (82 %) b 76 3.3
Valery et al. [118] 2011 30 30 4 (13.3) LINAC 57 2.82 13.4 10.4 10 40 8/14 (57 %) 90 0
Kelly et al. [119] 2011 24 24 23 (96) LINAC 57 0.2 13 6.6 5.3 29 2/4 (50 %) 78.6 8.3
Lin et al. [109] 2012 45 48 21 (47) LINAC 60 0.4 14 n.a. 11.6 n.a. n.a. 88 % 4.7
GK Gamma Knife radiosurgery, LINAC linear accelerator-based radiosurgery, n.a. not available, SRS stereotactic radiosurgery, WBRT whole brain radiation therapy
a
Mean age
b
Stable or improved
221
NCCTG N107C is an open phase III trial comparing SRS

Surgical Resection Cavity SRS to the surgical resection cavity vs. adjuvant WBRT in patients
with one to four brain metastases after surgical resection. The
There has been a growing interest in delivering SRS to the primary endpoints of the study are overall survival and neu-
resection cavity after surgical resection of a brain metastasis. rocognitive progression at 6 months. Tumor bed control is
This has been investigated due to the concern of cognitive defined as the absence of new nodular enhancement of the
decline with WBRT. The rationale is that by delivering radia- surgical bed. SRS treatment volume is defined as surgical
tion focally to the surgical resection cavity, one can take bed + 2 mm margin. Of note, unlike most SRS prescription
advantage of the local control benefits of radiation, while dose guidelines which are usually dictated by the maximum
minimizing normal brain toxicity. As well, surgical resection tumor diameter, the prescription dose for SRS to the surgical
may have benefit compared to SRS alone, especially for cavity is determined by the tumor volume, due to the frequent
patients with large tumors that cause mass effect, or for irregular shape of the surgical cavities (Table 14.13). The
patients who present with acute focal neurologic signs or results of this trial should give us further insight into future
symptoms, which would be resolved better with surgical directions in treatment management of brain metastases.
resection than SRS alone. There have been a number of sin-
gle institution studies looking at resection cavity SRS, with
results summarized in Table 14.12 [120–130]. SRS in Management of Large Brain
The initial study reported in 2008 from the University of Metastases
Pittsburgh examined 40 patients who underwent tumor bed
Gamma Knife radiosurgery after initial resection, 32 of Large brain metastases often cause neurologic symptoms
which had a gross total resection [120]. Median survival was due to peritumoral edema and mass effect. Surgical resection
13 months, with a local control rate of 73 %. GKRS was per- followed by WBRT has traditionally been the preferred man-
formed at a median time of 4 weeks after tumor resection. agement for large brain metastases. Traditionally, many cli-
The median tumor margin dose was 16 Gy. nicians use an upper limit of 3 cm in diameter for consideration
The largest and most recent study using LINAC-based of SRS for brain metastases. The role of SRS for larger
radiosurgery is from Stanford University, which retrospec- lesions is still an area of controversy and investigation.
tively examined 120 lesions in 112 patients [125]. A perti- A study from Pittsburgh examined the role of GKRS in
nent question to consider when treating a surgical resection patients with brain metastases >3 cm in diameter [131].
cavity with SRS is the appropriate margin to use after a gross A total of 70 patients without surgical resection were retro-
total resection. This study compared no margin vs. a 2 mm spectively reviewed, with 37 receiving GKRS alone, while
margin around the resection cavity to determine if there are 33 patients had prior WBRT. On the first follow-up MRI
differences in outcome. A total of 58 lesions were treated after treatment, 41 % of tumors had >50 % volume reduc-
with no margin vs. 62 lesions with a 2-mm margin. The study tion, 31 % had between 10 and 50 % reduction, and 28 % of
found that 12-month local failure was reduced from 16 to tumors were stable or larger. Tumor volume reduction was
3 % with addition of a margin (p = 0.042). The median OS associated with a single brain metastasis, no prior WBRT,
was 17 months with a 1-year OS of 62 %. and tumor volume <16 cm3. Improved peritumoral edema
Table 14.12 SRS to the resection cavity

Local Age
Median control (median Follow up
Lead author Institution Year Patients (n) survival (~1 year) Modality Tumor margin dose or mean) (months)
Mathieu [120] Pittsburgh 2008 40 13 months 73 % GKRS 16 Gy 59.5 13
Quigley [121] Allegheny 2008 32 16.4 months 93.75 % LINAC 14 Gy to 80 % IDL 60 14
Iwai [122] Japan(Osaka) 2008 21 20 months 82 % GKRS 17 Gy 61 21.6
Soltys [129] Stanford 2008 76 15.1 months 79 % LINAC 18.6 Gy to 79 % IDL 58 8.1
Jagannathan [130] Virginia 2009 47 11.5 months 94 % GKRS 19 Gy 61 14
Do [123] UC Irvine 2009 30 51 % at 12 months 82 % LINAC 15 Gy to 80–90 % IDL 61.5 NR
Hwang [124] Tufts 2010 43 15 months 72 % GKRS 15–20 Gy 56.68 NR
Choi [125] Stanford 2012 112 17 months 90.5 % LINAC 20 Gy to 79 % IDL 61 11
Robbins [126] Henry Ford 2012 85 12.1 months 81.4 % LINAC 16 Gy to 90 % IDL 58 11.2
Hartford [127] Dartmouth 2012 47 53 % at 12 months 85.5 % LINAC 10 Gy to 80–95 % IDL 64 9.3
Ogiwara [128] Northwestern 2012 56 20.5 91.1 % GKRS 17.1 Gy 58.5 24
Table 14.13 Dosing for surgical resection cavity stereotactic radiosurgery seizures, transient worsening of preexisting neurologic condi-
3
Resection cavity size (cm ) Dose (Gy) tions, and fatigue. The incidence of severe headaches after
<4.2 20 frame removal is low [65]. Nausea/vomiting can be mini-
≥4.2 to <8 18 mized if the dose to the area postrema is kept below 375 cGy
≥8 to <14.4 17 [52]. The risk for seizures has been reported to range from 2
≥14.4 to <20 15 to 6 % [54, 55]. Seizure risk is higher with cortical lesions and
≥20 to <30 14 for patients with history of seizures. For these patients, anti-
≥30 to <5 cm max diameter 12 convulsants should be considered, and steroid dose should be
Adapted from NCCTG N107C; http://www.cancer.gov/clinicaltrials/ increased.
search/view?cdrid=701474&version=HealthProfessional Subacute complications occur within the first 6 months
after SRS. Alopecia has been reported in 5.6 % of patients
was associated with a single lesion, no prior WBRT and [54]. These patients had superficial tumors that resulted in a
breast cancer histology. Because of persistent symptoms, dose of 4.4 Gy to the scalp. Neurological deterioration can
10 % [7] of patients required craniotomy for salvage therapy. occur in some patients, which is usually treated with steroids.
The authors conclude that surgical resection is the preferred A prospective study assessed the acute complications of
treatment if possible for large lesions. 76 patients undergoing GKRS using CTCAE v.3.0, the Brief
A more recent retrospective review from Stanford Pain Questionnaire (Short Form), Brief Fatigue Inventory,
University evaluated 97 patients with brain metastases >2 cm and the Tinnitus Handicap Inventory [134]. The majority of
who underwent surgical resection followed by cavity SRS patients (34 %) were treated for brain metastases, but also
[132]. The rationale for the study was based on high rates of included trigeminal neuralgia, schwannoma, meningioma,
local failure with single modality treatment of larger brain arteriovenous malformation, and pituitary adenoma. There
metastases. The 12-month local failure rate was 9.3 % with was no significant difference in face swelling, headache, eye
median time to local failure of 6 months. The 12-month inci- pain, vomiting, seizures, or passing out at any intervals com-
dence of distant brain failure was 53 %, and median survival pared to baseline. At 1 week, 24 % developed minimal scalp
time was 15.6 months. This study demonstrates that surgery numbness, which decreased to 2 % at 2 months. Pin site pain
with resection cavity SRS can achieve good local control of was present in 13 %, which declined to 3 % at 1 month with
large brain metastases, while sparing or delaying WBRT. a median intensity level of 3 out of 10. Four percent of
Due to potentially worse toxicity with SRS for large patients developed pin-site infection at 1 week and none at 1
lesions, institutions have looked at fractionated SRS. A group and 2 months. By 1 month, 10 % reported new local hair
from Germany compared differences in efficacy and toxicity loss. New or worsening fatigue was experienced by 23 %,
when comparing single fraction SRS vs. fractionated stereo- 16 %, and 15 % of patients at 1 week, 1 month, and 2 months,
tactic radiotherapy for brain metastases [133]. The study ret- respectively, but 40 % reported fatigue at baseline. After
rospectively examined 260 patients, with 138 patients treatment, 1 %, 6 %, and 3 % developed new tinnitus at 1
receiving single fraction SRS and 61 patients receiving frac- week, 1 month, and 2 months, respectively, which was sig-
tionated SRS. The fractionated SRS group included two nificant when comparing baseline to non-baseline
fractionation schemes: 5 Gy × 7 and 4 Gy × 10. They found (p = 0.0269). Thirty-two patients were employed prior to
no difference in local control, progression-free survival, or SRS. Three (9 %) patients did not return to work. Twenty-
overall survival. Single fraction SRS was associated with seven (84 %) patients returned to work a median of 4 days
higher Grade 1–3 toxicity. Of note, fractionated treatment after treatment. This prospective study demonstrates that
was used for large lesions or lesions near critical structures, GKRS is well tolerated with few patients developing signifi-
but local control did not show a difference. A hypofraction- cant acute effects. Many patients are able to return to work
ated treatment strategy may become a more prominent strat- shortly after GKRS.
egy for larger brain metastases, but local control rates and
survival need to be validated prospectively in larger studies.
Chronic Complications
Complications Associated with SRS Radiation necrosis is the most serious chronic complication
of SRS. In general, the risk of radiation necrosis increases
Acute Complications with higher doses, larger tumor volumes, and in patients
treated with prior radiation. The pathophysiology of radia-
Acute side effects during the first week after SRS are uncom- tion necrosis is a coagulative process that predominantly
mon. Some of the complications include headache after frame affects white matter. It is caused by small artery injury and
removal, infection of the pin sites, nausea and vomiting, thrombotic occlusion. On pathology, these small arteries
Table 14.14 Dose constraints for single fraction brain SRS

Organ Endpoint Dose (Gy) Rate (%)
Brain Symptomatic necrosis V12 < 5–10 cm3 <20
Brainstem Permanent cranial neuropathy or necrosis Dmax <12.5 <5
Optic nerve/chiasm Optic neuropathy Dmax <12 <10
Cochlea Sensory neural hearing loss Prescription dose ≤14 <25
Adapted from QUANTEC
From Marks LB, Yorke ED, Jackson A, Ten Haken RK, Constine LS, Eisbruch A, et al. Use of normal tissue complica-
tion probability models in the clinic. Int J Radiat Oncol Biol Phys. 2010;76(3 Suppl):S10–9; used with permission
demonstrate endothelial thickening, lymphocytic and macro- recommendations are based on a review of available data in
phagic infiltrates, presence of cytokines, hyalinization, fibri- the literature to summarize dose constraints for clinical use.
noid deposition, thrombosis, and finally occlusion. The With regard to tolerance dose to the optic nerves and chiasm,
primary mechanism of delayed injury in radiation associated consistent agreement has been reached on the low risk of
with necrosis is secondary to vascular endothelial injury or radiation-induced optic neuropathy for a Dmax of ≤10 Gy, and
direct damage to oligodendroglia. Depending on the tumor one major study indicated a low risk with a Dmax of ≤12 Gy
location and rate of growth, radiation necrosis can present for single fraction radiosurgery [139]. With regard to brain-
with seizure, elevated intracranial pressure, obstructive stem toxicity in single fraction SRS, Dmax <12.5 Gy is associ-
hydrocephalus, or herniation syndromes. ated with <5 % risk [140]. Higher doses of 15–20 Gy have
A number of MRI modalities can aid in diagnosis of brain been reported with low incidence of complications, but this
radionecrosis [135]. On MR spectroscopy, a lactate peak is may be an artifact of poor survival. With regard to overall
often manifested. On DWI, there can be increased signal due brain radionecrosis, toxicity increases rapidly once the vol-
to restricted diffusion in the region of necrosis. MR FLAIR ume of any part of the brain receiving >12 Gy is >5–10 cm3
may elicit regions of edema surrounding the necrosis. [137]. A modified QUANTEC summary of pertinent dose
Dynamic susceptibility-weighted contrast enhanced (DSC) constraints for single fraction brain SRS is provided in
perfusion MRI has also been used to delineate areas of Table 14.14 [141].
necrosis, which can be further characterized with pharmaco- The general strategy for treating radiation necrosis is to
kinetic modeling [136]. However, definitive diagnosis of decrease the edema and necrosis. Usually, high doses of ste-
radiation necrosis requires biopsy or surgical resection. roids are used to minimize neurological deterioration. If the
For single fraction radiosurgery, a clear correlation has patient becomes symptomatic despite steroids, surgical
been demonstrated between the target size and the risk of resection can be considered. Anticoagulation with coumadin
brain radionecrosis [137]. A study from Italy of 206 patients to prevent vessel thrombosis can also be used. In some cases,
treated with LINAC-based radiosurgery investigated factors hyperbaric oxygen has been used for patients who are poor
that predict for radionecrosis [138]. Median overall survival surgical candidates, have multiple areas of radiation necro-
and brain control were 14.1 months and 10 months, respec- sis, or have a surgically inaccessible lesion. In a retrospective
tively, with 1-year and 2-year local control rates of 92 % and review of 40 patients, 90 % reported subjective improvement
84 %, respectively. Severe neurological complications (RTOG and 80 % had objective neurologic improvement with hyper-
Grade 3 and 4) occurred in 5.8 % of patients. Brain radione- baric oxygen [142]. Steroids were discontinued or decreased
crosis occurred in 24 % of treated lesions, of which 10 % in two thirds of patients. A randomized trial is underway at
were symptomatic and 14 % asymptomatic. On multivariate the University of Cincinnati comparing surgery with hyper-
analysis, V10 through V16 Gy were independent risk factors baric oxygen.
for radionecrosis, with V10 Gy and V12 Gy being the most More recently, Bevacizumab (Avastin) treatment is being
predictive (p = 0.0001). For V10 Gy >12.6 cm3 and V12 Gy used in the management of radiation necrosis [143]. Since radi-
>10.9 cm3, the risk of radionecrosis was 47 %. The authors ation necrosis leads to hypoxia, elevated levels of VEGF have
suggest that since lesions with V12 Gy >8.5 cm3 confer a risk been seen which may lead to further edema. Therefore, block-
of radionecrosis >10 %, hypofractionated SRS should be con- ing VEGF from reaching capillary targets is a logical treatment
sidered for eloquently located lesions. As well, larger lesions option for radionecrosis [144, 145]. Another potentially prom-
would likely have larger volumes receiving high dose. ising treatment option is the use of laser interstitial thermal
A number of QUANTEC (quantitative analysis of normal therapy [146]. Laser interstitial thermal therapy was originally
tissue effects in the clinic) organ-specific papers published used for treatment for brain tumors in 1990. It delivers focal
in 2010 have reviewed late toxicity and radiation. These laser energy selectively to coagulate tissue with interstitial
hyperthermia and exhibits a sharp ablation boundary zone. were available for evaluation. New brain metastases were
This technique may be particularly useful for areas of radione- detected in 11 patients (24.4 %) at the time of the SRS treat-
crosis that are refractory to medical management as previously ment planning MRI. The median neurological progression-
described and are not amenable to surgical resection. free survival was 8 months and overall survival time was 9
months. The actuarial incidences of neurological progres-
sion at 6 months and 1 year were 17 and 20 %, respectively.
Costs Grade 3 or higher neurotoxicity was seen in two patients, and
the authors conclude that motexafin gadolinium does not
SRS for patients with multiple brain metastases, especially increase toxicity and should be further studied.
those with poor prognosis, will likely undergo further scru- A recent Phase I prospective study from Canada enrolled
tiny as the costs of various treatments are evaluated for cost- 26 previously irradiated patients with 1–4 progressive brain
effectiveness. Although the costs of SRS vary based on the metastases with a maximum diameter of 40 mm and KPS ≥60
type of technology used, number of lesions treated, and insti- [149]. Patients were enrolled to receive three different doses
tutional costs, estimated Medicare costs for SRS are of temozolomide (TMZ): 100, 150, and 200 mg/m2/day for
$10,000–$27,000 per procedure as compared to $2,300– days 1–5. SRS was administered on day 5. SRS dose was pre-
$7,650 for WBRT [53]. Unfortunately, no universally scribed as per RTOG 9005 guidelines. The primary endpoint
accepted threshold exists in the USA for determining whether was safety of increasing TMZ doses. Secondary endpoints
treatment provides good value for its cost. Traditionally, the included local control and survival. The median number of
quality-adjusted life years (QALY) has been used. Since the brain metastases was 1.5, with a median target diameter of
overuse of any technology can be a result of a number of fac- 21 mm. The most common Grade 1–2 adverse events were
tors including misaligned incentives under the fee-for- vomiting (23 %), nausea (23 %), edema (12 %), seizure (8 %),
service payment system and patient preference for psychosis (4 %), and thrombocytopenia (4 %). The frequency
cutting-edge technology, it is important that neurosurgeons of nausea and vomiting did not appear to be dose dependent.
and radiation oncologists demonstrate the value of radiosur- Grade 3–4 toxicities were not observed. Median overall sur-
gery for patients with multiple (>5) brain metastases. vival was 10.2 months. Crude local control was 87.5 %, with
Administering SRS with WBRT instead of SRS alone a radiological response seen in eight of 24 evaluable patients
may contribute to additional cost. A recent study by Lal et al. (33.3 %), and stable disease >6 months in 13 of 24 patients
compared SRS alone vs. SRS + WBRT for patients with three (54.2 %). The authors conclude that TMZ can be given at
or less brain metastases, using actual life years saved (LYS), doses up to 200 mg/m2/day prior to SRS with good local con-
quality-adjusted life years, and incremental cost-effectiveness trol. Further randomized studies with greater patient numbers
ratio (ICER) [147]. Compared to SRS with whole brain radi- will be needed to assess for the impact of addition of temo-
ation therapy, SRS alone has a higher average cost ($74,000 zolomide on local control and overall survival.
vs. $119,000). However, SRS alone has a higher average RTOG 0320 was a phase III trial which randomized
effectiveness of 1.64 LYS compared with combined SRS and NSCLC patients with 1–3 brain metastases to 3 arms: (1)
WBRT (0.60 LYS). The authors conclude that SRS alone SRS plus WBRT, (2) SRS plus WBRT with erlotinib, and (3)
without WBRT is more cost-effective, even with salvage SRS plus WBRT with temozolomide [150]. The rationale for
treatment delivered down the line. adding erlotinib or TMZ is its ability to cross the blood-brain
barrier. Of note, for patients randomized to receive chemo-
therapy, the treating physician could decide whether to give
Chemotherapy and Radiosensitizers erlotinib or TMZ. In the erlotinib and TMZ arms, chemo-
therapy was delivered with radiation therapy, and the investi-
Radiosensitizers are agents that can potentially increase the gator could decide whether to give adjuvant chemotherapy.
lethal effects of radiation therapy when given concurrently This trial closed early secondary to poor accrual. The addi-
with radiation. Recent studies have examined the role of che- tion of neither erlotinib nor TMZ resulted in an improvement
motherapy and radiosensitizers with WBRT and/or SRS in in OS or time to CNS progression compared to the control
the hopes of improving survival. arm: WBRT + SRS alone. Interestingly, WBRT + SRS alone
A phase II study, PCYC-0224, evaluated the use of resulted in an improved median survival time, which was not
motexafin gadolinium, a radiation sensitizer, given with attributable to excess toxicity in the chemo arms. The
WBRT (37.5 Gy in 15 fractions) and SRS (15–21 Gy) for WBRT + SRS arm had significantly less deterioration in per-
patients with ≤6 brain metastases [148]. The primary end- formance status at 6 months. Hence, there was no benefit to
point was to evaluate toxicity in preparation for a phase III the addition of chemotherapy and potentially a detriment to
study. A total of 65 patients were enrolled of which 45 (69 %) clinical outcome.
nitive decline (p = 0.01). The addition of memantine reduced

Future Directions decline in memory (p = 0.015), cognitive, executive, and global
function as well as processing speed (p < 0.01). The conclu-
Incorporation of newer imaging modalities such as MRS and sion is that memantine helps preserve cognitive function after
PET to aid in SRS planning and delivery of higher doses to WBRT in patients with brain metastases.
certain regions of the target may be the wave of the future Recognizing that hippocampus radiation damage may
[151, 152]. Another area of interest is in the development of induce cognitive decline in patients, a phase II study
improved hardware for SRS delivery. There have been recent (RTOG 0933) was opened to evaluate hippocampus sparing
developments in a frameless, surface image-guided radiosur- WBRT delivered by an IMRT technique. The primary end-
gery technique to treat brain metastases [153]. Originally, the point is neurocognitive decline measured by the HVLT-R
Gamma Knife uses a frame-based immobilization device, and administered 4 months after WBRT. This trial completed
the early LINAC-based radiosurgery techniques also used a accrual with results suggesting preservation of memory
frame-based system. However, patient discomfort and issues and QOL compared to historical series. (Gondi V et al.
with reproducibility in fractionated SRS are disadvantages ASTRO 2013).
with the frame system. More recently, frameless systems have
been developed for LINAC-based radiosurgery, by providing
a custom fit mask and position monitoring with X-rays or Conclusions
CT. However, these frameless systems subject the patient to
increased radiation exposure due to image guidance with radi- Historically, whole brain radiation therapy has been the stan-
ation. Further developments now allow for real-time monitor- dard treatment for patients with brain metastases. Over the
ing by externally attached infrared fiducials for intra-fraction past 20 years, stereotactic radiosurgery has developed an
monitoring. When first developed, these fiducials were placed increasingly important role in the management of brain
on a bite block, but recent developments now allow for sur- metastases. Although SRS has become an important thera-
face image guidance using camera pods that capture facial peutic option, the appropriate utilization of SRS is contro-
features for tracking. A recent study from UC San Diego of 44 versial. Studies have shown benefits of SRS as sole treatment,
consecutive patients demonstrates outcomes comparable to a as adjuvant therapy after surgery, as an adjunct to WBRT,
frame system, with a 6 and 12-month local control of 90 % and as salvage treatment after WBRT. Retrospective studies
and 76 %, respectively. Distant brain failure rate was 46 %. suggest that patients treated with SRS or surgery have com-
The Japan Clinical Oncology Group (JCOG) has a phase parable outcomes. RTOG 9508 provides level 1 evidence of
III randomized control trial (NCT00280475) of postoperative a survival benefit of SRS and WBRT compared with WBRT
WBRT vs. salvage SRS for patients with 1–4 brain metastasis alone for patients with a single lesion. For patients with more
from lung, breast, and colorectal cancers. The aim of this than one lesion, the use of SRS and WBRT should be consid-
study is to evaluate non-inferiority of salvage stereotactic ered based on performance status, extent and activity of
radiosurgery in the patients who received surgical resection extracranial disease, and steroid use. More recently, the
for brain metastases in comparison with postoperative whole omission of WBRT has been advocated by concerns of the
brain radiation therapy. The primary endpoint of this study is potential risks of cognitive decline with WBRT and apparent
overall survival. Secondary endpoints include performance lack of survival benefit of additional WBRT. Surgical resec-
status preservation, mini-mental status (MMSE) preservation followed by SRS boost exhibits good local control for
tion, and adverse events. This trial is still accruing patients. single brain metastasis. The potential cognitive side effects
As health care cost-effectiveness has become an increas- of WBRT must be balanced against the risk for increased
ingly important issue in our health care system, there have distant brain metastases and the cost of more frequent
been more recent RTOG trials examining ways to reduce cog- follow-up scans when treating brain metastases with SRS
nitive decline with WBRT without SRS. RTOG 0614 is a trial alone. A recent phase III trial has shown benefit in cognition
that assesses differences in cognitive decline in patients receiv- for patients undergoing WBRT who take memantine. Future
ing WBRT with or without the administration of memantine, studies may find other drugs which could mitigate the side
an NMDA receptor antagonist used to treat Alzheimer’s dis- effects of WBRT. Development and incorporation of radio-
ease [154]. With a median follow-up of 12.4 months, there sensitizers may help improve outcomes with radiation ther-
was no difference in OS or PFS. The primary endpoint was apy for brain metastases. As the prognosis improves for
difference in HVLT-R delayed recall. Memantine reduced cancer patients, the challenge of increasing progression-free
cognitive decline by 0.9 as measured by the HVLT-R delayed survival while limiting cognitive decline and other side
recall at 24 weeks, but the results were not statistically signifi- effects will continue to make the management of brain
cant (p = 0.059). However, this study may have been under- metastases controversial. We encourage the participation of
powered to detect a difference, as there were only 149 patients with brain metastases in clinical trials to improve
analyzable patients. Memantine reduced the relative risk of outcomes and to help answer many important questions
cognitive decline by 17 % (p = 0.01) and delayed time to cog- about management of this very common disease.
Case Presentation 14.1 to the 50 % local isodose line, which covered 100 % of the
History: This is a 73-year-old male with a history of renal target. The plan utilized 8 shots using an 8 mm helmet
cell carcinoma status post right nephrectomy 10 years without plugs. Tumor volume = 2.0 cm3. Maximum
ago. He presented with left-sided weakness. MRI was dose = 4,800 cGy. Maximum diameter = 1.9 cm. MD/
obtained, which revealed a 1.7 cm right frontal lesion PD = 2.00. PIV/TV = 1.70.
(Fig. 14.1a, b). Staging work-up was negative. Three-month post Gamma Knife radiosurgery MRI
He underwent Gamma Knife radiosurgery to the right (Fig. 14.1d, e). The patient’s left-sided weakness
frontal lesion (Fig. 14.1c). A dose of 24 Gy was prescribed improved.
Fig. 14.1 MRI of male with a history of renal cell carcinoma. See Case Presentation 14.1
Case Presentation 14.2 Knife radiosurgery within 1 week of completing WBRT

History: This is a 57-year-old female with history of a (Fig. 14.2c).
T3N1M0, Stage III adenocarcinoma of the esophagus sta- Treatment planning information, 1,800 cGy was pre-
tus post radiation therapy (3,000 cGy/20 fx, 150 cGy/fx scribed to the 50 % isodose line, which covered 100 % of
bid) to the distal esophagus with concurrent chemother- the target. The plan utilized 6 shots using an 18 mm hel-
apy status post esophageal resection followed by an addi- met without plugs. Tumor volume = 5.3 cm3. Maximum
tional 3,000 cGy/20 fx, 150 cGy/fx bid to the tumor bed dose = 36 Gy. Maximum diameter = 2.5 cm. MD/PD =
with concurrent chemotherapy. She did well for 10 months 2.00. PIV/TV = 1.64.
until she presented with a right-sided seizure. MRI Three-month post Gamma Knife radiosurgery film
revealed a 2-cm left frontoparietal lesion (Fig. 14.2a, b). (Fig. 14.2d, e). She had excellent response to WBRT and
She underwent whole brain radiation therapy Gamma Knife radiosurgery.
(3,750 cGy/15 fx using 250 cGy) followed by Gamma
Fig. 14.2 MRI of female with a history of adenocarcinoma of the esophagus. See Case Presentation 14.2
10. Lagerwaard FJ, Levendag PC, Nowak PJ, Eijkenboom WM,

References Hanssens PE, Schmitz PI. Identification of prognostic factors in
patients with brain metastases: a review of 1292 patients. Int J
1. Wen PY, Loeffler JS. Management of brain metastasis. Oncology Radiat Oncol Biol Phys. 1999;43(4):795–803.
(Huntingt). 1999;13(7):941–54. 11. Weltman E, Salvajoli JV, Brandt RA, de Morais HR, Prisco FE,
2. Greenberg H, Chandler WF, Sandler HM. Brain metastases. In: Cruz JC, et al. Radiosurgery for brain metastases: a score index for
Greenberg H, Chandler WF, Sandler HM, editors. Brain tumors. predicting prognosis. Int J Radiat Oncol Biol Phys. 2000;46(5):
Chapter 15. New York: Oxford University Press; 1999. 1155–61.
p. 299–317. 12. Sperduto CM, Watanabe Y, Mullan J, Hood T, Dyste G, Watts C,
3. Lohr F, Pirzkall A, Hof H, Fleckenstein K, Debus J. Adjuvant et al. A validation study of a new prognostic index for patients
treatment of brain metastases. Semin Surg Oncol. with brain metastases: the Graded Prognostic Assessment.
2001;20(1):50–6. J Neurosurg. 2008;109(Suppl):87–9.
4. Posner JB. Intracranial metastases. Neurologic complications of 13. Sperduto PW, Chao ST, Sneed PK, Luo X, Suh J, Roberge D, et al.
cancer. Vol 45. Philadelphia, PA: FA Davis Company; 1995. Diagnosis-specific prognostic factors, indexes, and treatment out-
p. 77–110. comes for patients with newly diagnosed brain metastases: a
5. Zimm S, Wampler GL, Stablein D, Hazra T, Young multi-institutional analysis of 4,259 patients. Int J Radiat Oncol
HF. Intracerebral metastases in solid-tumor patients: natural his- Biol Phys. 2010;77(3):655–61.
tory and results of treatment. Cancer. 1981;48(2):384–94. 14. Sperduto PW, Kased N, Roberge D, Xu Z, Shanley R, Luo X, et al.
6. Nussbaum ES, Djalilian HR, Cho KH, Hall WA. Brain metastases. Effect of tumor subtype on survival and the graded prognostic
Histology, multiplicity, surgery, and survival. Cancer. 1996;78(8): assessment for patients with breast cancer and brain metastases.
1781–8. Int J Radiat Oncol Biol Phys. 2012;82(5):2111–7.
7. Hazra T, Mullins GM, Lott S. Management of cerebral metastasis 15. CHAO JH, PHILLIPS R, NICKSON JJ. Roentgen-ray therapy of
from bronchogenic carcinoma. Johns Hopkins Med cerebral metastases. Cancer. 1954;7(4):682–9.
J. 1972;130(6):377–83. 16. CHU FC, HILARIS BB. Value of radiation theray in the manage-
8. Chao ST, Barnett GH, Liu SW, Reuther AM, Toms SA, Vogelbaum ment of intracranial metastases. Cancer. 1961;14:577–81.
MA, et al. Five-year survivors of brain metastases: a single- 17. Borgelt B, Gelber R, Kramer S, Brady LW, Chang CH, Davis LW,
institution report of 32 patients. Int J Radiat Oncol Biol Phys. et al. The palliation of brain metastases: final results of the first
2006;66(3):801–9. two studies by the Radiation Therapy Oncology Group. Int J
9. Gaspar L, Scott C, Rotman M, Asbell S, Phillips T, Wasserman T, Radiat Oncol Biol Phys. 1980;6(1):1–9.
et al. Recursive partitioning analysis (RPA) of prognostic factors 18. Gelber RD, Larson M, Borgelt BB, Kramer S. Equivalence of radi-
in three Radiation Therapy Oncology Group (RTOG) brain metas- ation schedules for the palliative treatment of brain metastases in
tases trials. Int J Radiat Oncol Biol Phys. 1997;37(4):745–51. patients with favorable prognosis. Cancer. 1981;48(8):1749–53.
19. Borgelt B, Gelber R, Larson M, Hendrickson F, Griffin T, Roth 35. Stinauer MA, Kavanagh BD, Schefter TE, Gonzalez R, Flaig T,
R. Ultra-rapid high dose irradiation schedules for the palliation of Lewis K, et al. Stereotactic body radiation therapy for melanoma
brain metastases: final results of the first two studies by the and renal cell carcinoma: impact of single fraction equivalent dose
Radiation Therapy Oncology Group. Int J Radiat Oncol Biol Phys. on local control. Radiat Oncol. 2011;6:34.
1981;7(12):1633–8. 36. Flickinger JC, Kondziolka D, Lunsford LD, Coffey RJ, Goodman
20. Kurtz JM, Gelber R, Brady LW, Carella RJ, Cooper JS. The pallia- ML, Shaw EG, et al. A multi-institutional experience with stereo-
tion of brain metastases in a favorable patient population: a ran- tactic radiosurgery for solitary brain metastasis. Int J Radiat Oncol
domized clinical trial by the Radiation Therapy Oncology Group. Biol Phys. 1994;28(4):797–802.
Int J Radiat Oncol Biol Phys. 1981;7(7):891–5. 37. Yu C, Chen JC, Apuzzo ML, O’Day S, Giannotta SL, Weber JS,
21. Sause WT, Scott C, Krisch R, Rotman M, Sneed PK, Janjan N, et al. Metastatic melanoma to the brain: prognostic factors after
et al. Phase I/II trial of accelerated fractionation in brain metasta- gamma knife radiosurgery. Int J Radiat Oncol Biol Phys.
ses RTOG 85-28. Int J Radiat Oncol Biol Phys. 1993;26(4): 2002;52(5):1277–87.
653–7. 38. Goyal LK, Suh JH, Reddy CA, Barnett GH. The role of whole
22. Murray KJ, Scott C, Greenberg HM, Emami B, Seider M, Vora brain radiotherapy and stereotactic radiosurgery on brain metasta-
NL, et al. A randomized phase III study of accelerated hyperfrac- ses from renal cell carcinoma. Int J Radiat Oncol Biol Phys.
tionation versus standard in patients with unresected brain metas- 2000;47(4):1007–12.
tases: a report of the Radiation Therapy Oncology Group (RTOG) 39. Sturm V, Kober B, Höver KH, Schlegel W, Boesecke R, Pastyr O,
9104. Int J Radiat Oncol Biol Phys. 1997;39(3):571–4. et al. Stereotactic percutaneous single dose irradiation of brain
23. Komarnicky LT, Phillips TL, Martz K, Asbell S, Isaacson S, metastases with a linear accelerator. Int J Radiat Oncol Biol Phys.
Urtasun R. A randomized phase III protocol for the evaluation of 1987;13(2):279–82.
misonidazole combined with radiation in the treatment of patients 40. Nieder C, Nestle U, Walter K, Niewald M, Schnabel K. Dose/
with brain metastases (RTOG-7916). Int J Radiat Oncol Biol effect relationships for brain metastases. J Cancer Res Clin Oncol.
Phys. 1991;20(1):53–8. 1998;124(6):346–50.
24. Phillips TL, Scott CB, Leibel SA, Rotman M, Weigensberg 41. Leibel SA, Sheline GE. Radiation therapy for neoplasms of the
IJ. Results of a randomized comparison of radiotherapy and bro- brain. J Neurosurg. 1987;66(1):1–22.
modeoxyuridine with radiotherapy alone for brain metastases: 42. Larson DA, Flickinger JC, Loeffler JS. The radiobiology of radio-
report of RTOG trial 89-05. Int J Radiat Oncol Biol Phys. surgery. Int J Radiat Oncol Biol Phys. 1993;25(3):557–61.
1995;33(2):339–48. 43. Mehta M, Noyes W, Craig B, Lamond J, Auchter R, French M,
25. Sneed PK, Larson DA, Wara WM. Radiotherapy for cerebral et al. A cost-effectiveness and cost-utility analysis of radiosurgery
metastases. Neurosurg Clin N Am. 1996;7(3):505–15. vs. resection for single-brain metastases. Int J Radiat Oncol Biol
26. Knisely JP, Berkey B, Chakravarti A, Yung AW, Curran WJ, Phys. 1997;39(2):445–54.
Robins HI, et al. A phase III study of conventional radiation ther- 44. Sperduto PW, Hall WA. Radiosurgery, cost-effectiveness, gold
apy plus thalidomide versus conventional radiation therapy for standards, the scientific method, cavalier cowboys, and the cost of
multiple brain metastases (RTOG 0118). Int J Radiat Oncol Biol hope. Int J Radiat Oncol Biol Phys. 1996;36(2):511–3.
Phys. 2008;71(1):79–86. 45. Auchter RM, Lamond JP, Alexander E, Buatti JM, Chappell R,
27. Borgelt B, Gelber R, Kramer S, et al. The palliation of brain Friedman WA, et al. A multiinstitutional outcome and prognostic
metastasis: final results of the first two studies by the Radiation factor analysis of radiosurgery for resectable single brain metasta-
Therapy Oncology Group. Int J Radiat Oncol Biol Phys. sis. Int J Radiat Oncol Biol Phys. 1996;35(1):27–35.
1980;6:1–9. 46. Bindal AK, Bindal RK, Hess KR, Shiu A, Hassenbusch SJ, Shi
28. Patchell RA, Tibbs PA, Walsh JW, Dempsey RJ, Maruyama Y, WM, et al. Surgery versus radiosurgery in the treatment of brain
Kryscio RJ, et al. A randomized trial of surgery in the treatment of metastasis. J Neurosurg. 1996;84(5):748–54.
single metastases to the brain. N Engl J Med. 47. O’Neill BP, Iturria NJ, Link MJ, Pollock BE, Ballman KV,
1990;322(8):494–500. O’Fallon JR. A comparison of surgical resection and stereotactic
29. Mintz AH, Kestle J, Rathbone MP, Gaspar L, Hugenholtz H, radiosurgery in the treatment of solitary brain metastases. Int J
Fisher B, et al. A randomized trial to assess the efficacy of surgery Radiat Oncol Biol Phys. 2003;55(5):1169–76.
in addition to radiotherapy in patients with a single cerebral 48. Muacevic A, Kreth FW, Horstmann GA, Schmid-Elsaesser R,
metastasis. Cancer. 1996;78(7):1470–6. Wowra B, Steiger HJ, et al. Surgery and radiotherapy compared
30. Noordijk EM, Vecht CJ, Haaxma-Reiche H, Padberg GW, with gamma knife radiosurgery in the treatment of solitary cerebral
Voormolen JH, Hoekstra FH, et al. The choice of treatment of metastases of small diameter. J Neurosurg. 1999;91(1):35–43.
single brain metastasis should be based on extracranial tumor 49. Loeffler JS, Shrieve DC. What is appropriate therapy for a patient
activity and age. Int J Radiat Oncol Biol Phys. 1994;29(4): with a single brain metastasis? Int J Radiat Oncol Biol Phys.
711–7. 1994;29(4):915–7. discussion 20.
31. Patchell RA, Tibbs PA, Regine WF, Dempsey RJ, Mohiuddin M, 50. Muacevic A, Wowra B, Siefert A, Tonn JC, Steiger HJ, Kreth
Kryscio RJ, et al. Postoperative radiotherapy in the treatment of FW. Microsurgery plus whole brain irradiation versus Gamma
single metastases to the brain: a randomized trial. JAMA. Knife surgery alone for treatment of single metastases to the brain:
1998;280(17):1485–9. a randomized controlled multicentre phase III trial. J Neurooncol.
32. Patchell RA, Regine WF. The rationale for adjuvant whole brain 2008;87(3):299–307.
radiation therapy with radiosurgery in the treatment of single brain 51. Lang FF, Suki D, Maor M, Chang E, Hess K, Graham S, et al.
metastases. Technol Cancer Res Treat. 2003;2(2):111–5. Conventional surgery versus stereotactic radiosurgery in the treat-
33. Patel AJ, Suki D, Hatiboglu MA, Abouassi H, Shi W, Wildrick ment of single brain metastases: a prospective study with both
DM, et al. Factors influencing the risk of local recurrence after randomized and nonrandomized arms. Presented in abstract form
resection of a single brain metastasis. J Neurosurg. 2010;113(2): at the American Association of Neurological Surgeons, 2008.
181–9. 52. Weissman DE. Glucocorticoid treatment for brain metastases and
34. Leksell L. The stereotaxic method and radiosurgery of the brain. epidural spinal cord compression: a review. J Clin Oncol.
Acta Chir Scand. 1951;102(4):316–9. 1988;6(3):543–51.
53. Loeffler JS, Kooy HM, Wen PY, Fine HA, Cheng CW, Mannarino 71. Valéry CA, Cornu P, Noël G, Duyme M, Boisserie G, Sakka LJ, et al.
EG, et al. The treatment of recurrent brain metastases with stereo- Predictive factors of radiation necrosis after radiosurgery for cerebral
tactic radiosurgery. J Clin Oncol. 1990;8(4):576–82. metastases. Stereotact Funct Neurosurg. 2003;81(1–4):115–9.
54. Alexander E, Moriarty TM, Davis RB, Wen PY, Fine HA, Black PM, 72. Sanghavi SN, Miranpuri SS, Chappell R, Buatti JM, Sneed PK,
et al. Stereotactic radiosurgery for the definitive, noninvasive treat- Suh JH, et al. Radiosurgery for patients with brain metastases: a
ment of brain metastases. J Natl Cancer Inst. 1995;87(1):34–40. multi-institutional analysis, stratified by the RTOG recursive par-
55. Breneman JC, Warnick RE, Albright RE, Kukiatinant N, Shaw J, titioning analysis method. Int J Radiat Oncol Biol Phys.
Armin D, et al. Stereotactic radiosurgery for the treatment of brain 2001;51(2):426–34.
metastases. Results of a single institution series. Cancer. 73. Kondziolka D, Patel A, Lunsford LD, Kassam A, Flickinger
1997;79(3):551–7. JC. Stereotactic radiosurgery plus whole brain radiotherapy versus
56. Noël G, Proudhom MA, Valery CA, Cornu P, Boisserie G, radiotherapy alone for patients with multiple brain metastases. Int
Hasboun D, et al. Radiosurgery for re-irradiation of brain metasta- J Radiat Oncol Biol Phys. 1999;45(2):427–34.
sis: results in 54 patients. Radiother Oncol. 2001;60(1):61–7. 74. Chougule P, Burton-Williams M, Saris S, et al. Randomized treat-
57. Chao ST, Barnett GH, Vogelbaum MA, Angelov L, Weil RJ, ment of brain metastasis with Gamma Knife radiosurgery, whole
Neyman G, et al. Salvage stereotactic radiosurgery effectively brain radiotherapy or both. Int J Radiat Oncol Biol Phys.
treats recurrences from whole-brain radiation therapy. Cancer. 2000;48(Suppl):114.
2008;113(8):2198–204. 75. Andrews DW, Scott CB, Sperduto PW, Flanders AE, Gaspar LE,
58. Kihlström L, Karlsson B, Lindquist C. Gamma Knife surgery for Schell MC, et al. Whole brain radiation therapy with or without
cerebral metastases. Implications for survival based on 16 years stereotactic radiosurgery boost for patients with one to three brain
experience. Stereotact Funct Neurosurg. 1993;61 Suppl 1:45–50. metastases: phase III results of the RTOG 9508 randomised trial.
59. Valentino V. The results of radiosurgical management of 139 sin- Lancet. 2004;363(9422):1665–72.
gle cerebral metastases. Acta Neurochir Suppl. 1995;63:95–100. 76. Vogelbaum MA, Angelov L, Lee SY, Li L, Barnett GH, Suh
60. Shiau CY, Sneed PK, Shu HK, Lamborn KR, McDermott MW, JH. Local control of brain metastases by stereotactic radiosurgery
Chang S, et al. Radiosurgery for brain metastases: relationship of in relation to dose to the tumor margin. J Neurosurg.
dose and pattern of enhancement to local control. Int J Radiat 2006;104(6):907–12.
Oncol Biol Phys. 1997;37(2):375–83. 77. DeAngelis LM, Delattre JY, Posner JB. Radiation-induced demen-
61. Matsuo T, Shibata S, Yasunaga A, Iwanaga M, Mori K, Shimizu T, tia in patients cured of brain metastases. Neurology.
et al. Dose optimization and indication of Linac radiosurgery for 1989;39(6):789–96.
brain metastases. Int J Radiat Oncol Biol Phys. 1999;45(4):931–9. 78. So NK, O’Neill BP, Frytak S, Eagan RT, Earnest F, Lee
62. Schoeggl A, Kitz K, Ertl A, Reddy M, Bavinzski G, Schneider RE. Delayed leukoencephalopathy in survivors with small cell
B. Prognostic factor analysis for multiple brain metastases after lung cancer. Neurology. 1987;37(7):1198–201.
gamma knife radiosurgery: results in 97 patients. J Neurooncol. 79. Bhatnagar A, Heron DE, Kondziolka D, Lunsford LD, Flickinger
1999;42(2):169–75. JC. Analysis of repeat stereotactic radiosurgery for progressive
63. Simonová G, Liscák R, Novotný J. Solitary brain metastases primary and metastatic CNS tumors. Int J Radiat Oncol Biol Phys.
treated with the Leksell gamma knife: prognostic factors for 2002;53(3):527–32.
patients. Radiother Oncol. 2000;57(2):207–13. 80. Regine WF, Huhn JL, Patchell RA, St Clair WH, Strottmann J,
64. Chidel MA, Suh JH, Reddy CA, Chao ST, Lundbeck MF, Barnett Meigooni A, et al. Risk of symptomatic brain tumor recurrence
GH. Application of recursive partitioning analysis and evaluation and neurologic deficit after radiosurgery alone in patients with
of the use of whole brain radiation among patients treated with newly diagnosed brain metastases: results and implications. Int J
stereotactic radiosurgery for newly diagnosed brain metastases. Radiat Oncol Biol Phys. 2002;52(2):333–8.
Int J Radiat Oncol Biol Phys. 2000;47(4):993–9. 81. Meyers CA, Smith JA, Bezjak A, Mehta MP, Liebmann J, Illidge
65. Joseph J, Adler JR, Cox RS, Hancock SL. Linear accelerator- T, et al. Neurocognitive function and progression in patients with
based stereotaxic radiosurgery for brain metastases: the influence brain metastases treated with whole-brain radiation and motexafin
of number of lesions on survival. J Clin Oncol. 1996;14(4): gadolinium: results of a randomized phase III trial. J Clin Oncol.
1085–92. 2004;22(1):157–65.
66. Shehata MK, Young B, Reid B, Patchell RA, St Clair W, Sims J, 82. Regine WF, Scott C, Murray K, Curran W. Neurocognitive out-
et al. Stereotatic radiosurgery of 468 brain metastases ≤2 cm: come in brain metastases patients treated with accelerated-
implications for SRS dose and whole brain radiation therapy. Int J fractionation vs. accelerated-hyperfractionated radiotherapy: an
Radiat Oncol Biol Phys. 2004;59(1):87–93. analysis from Radiation Therapy Oncology Group Study 91-04.
67. Pirzkall A, Debus J, Lohr F, Fuss M, Rhein B, Engenhart-Cabillic R, Int J Radiat Oncol Biol Phys. 2001;51(3):711–7.
et al. Radiosurgery alone or in combination with whole-brain radio- 83. Chang EL, Wefel JS, Meyers CA, et al. Prospective neurocogni-
therapy for brain metastases. J Clin Oncol. 1998;16(11):3563–9. tive evaluation of patients with 1 to 3 newly diagnosed brain
68. Sneed PK, Lamborn KR, Forstner JM, McDermott MW, Chang S, metastases treated with stereotactic radiosurgery alone. Int J
Park E, et al. Radiosurgery for brain metastases: is whole brain Radiat Oncol Biol Phys. 2004;60:S164–5.
radiotherapy necessary? Int J Radiat Oncol Biol Phys. 84. Regine WF, Schmitt FA, Scott CB, Dearth C, Patchell RA, Nichols
1999;43(3):549–58. RC, et al. Feasibility of neurocognitive outcome evaluations in
69. Shaw E, Scott C, Souhami L, Dinapoli R, Kline R, Loeffler J, et al. patients with brain metastases in a multi-institutional cooperative
Single dose radiosurgical treatment of recurrent previously irradi- group setting: results of Radiation Therapy Oncology Group trial
ated primary brain tumors and brain metastases: final report of BR-0018. Int J Radiat Oncol Biol Phys. 2004;58(5):1346–52.
RTOG protocol 90-05. Int J Radiat Oncol Biol Phys. 2000;47(2): 85. Sønderkaer S, Schmiegelow M, Carstensen H, Nielsen LB, Müller J,
291–8. Schmiegelow K. Long-term neurological outcome of childhood brain
70. Shaw E, Scott C, Souhami L, Dinapoli R, Bahary JP, Kline R, tumors treated by surgery only. J Clin Oncol. 2003;21(7):1347–51.
et al. Radiosurgery for the treatment of previously irradiated 86. Tchen N, Juffs HG, Downie FP, Yi QL, Hu H, Chemerynsky I,
recurrent primary brain tumors and brain metastases: initial report et al. Cognitive function, fatigue, and menopausal symptoms in
of radiation therapy oncology group protocol (90-05). Int J Radiat women receiving adjuvant chemotherapy for breast cancer. J Clin
Oncol Biol Phys. 1996;34(3):647–54. Oncol. 2003;21(22):4175–83.
87. van Dam FS, Schagen SB, Muller MJ, Boogerd W, vd Wall E, 104. Nath SK, Lawson JD, Simpson DR, Vanderspek L, Wang JZ,
Droogleever Fortuyn ME, et al. Impairment of cognitive function Alksne JF, et al. Single-isocenter frameless intensity-modulated
in women receiving adjuvant treatment for high-risk breast can- stereotactic radiosurgery for simultaneous treatment of multiple
cer: high-dose versus standard-dose chemotherapy. J Natl Cancer brain metastases: clinical experience. Int J Radiat Oncol Biol
Inst. 1998;90(3):210–8. Phys. 2010;78(1):91–7.
88. Brezden CB, Phillips KA, Abdolell M, Bunston T, Tannock 105. Hunter GK, Suh JH, Reuther AM, Vogelbaum MA, Barnett GH,
IF. Cognitive function in breast cancer patients receiving adjuvant Angelov L, et al. Treatment of five or more brain metastases with
chemotherapy. J Clin Oncol. 2000;18(14):2695–701. stereotactic radiosurgery. Int J Radiat Oncol Biol Phys.
89. Hasegawa T, Kondziolka D, Flickinger JC, Germanwala A, 2012;83(5):1394–8.
Lunsford LD. Brain metastases treated with radiosurgery alone: 106. Grandhi R, Kondziolka D, Panczykowski D, Monaco EA, Kano
an alternative to whole brain radiotherapy? Neurosurgery. H, Niranjan A, et al. Stereotactic radiosurgery using the Leksell
2003;52(6):1318–26. discussion 26. Gamma Knife Perfexion unit in the management of patients with
90. Lutterbach J, Cyron D, Henne K, Ostertag CB. Radiosurgery fol- 10 or more brain metastases. J Neurosurg. 2012;117(2):237–45.
lowed by planned observation in patients with one to three brain 107. Sahgal A, Barani IJ, Novotny J, Zhang B, Petti P, Larson DA, et al.
metastases. Neurosurgery. 2003;52(5):1066–73. discussion 73-4. Prescription dose guideline based on physical criterion for multi-
91. Sneed PK, Suh JH, Goetsch SJ, Sanghavi SN, Chappell R, Buatti ple metastatic brain tumors treated with stereotactic radiosurgery.
JM, et al. A multi-institutional review of radiosurgery alone vs. Int J Radiat Oncol Biol Phys. 2010;78(2):605–8.
radiosurgery with whole brain radiotherapy as the initial manage- 108. Koyfman SA, Tendulkar RD, Chao ST, Vogelbaum MA, Barnett
ment of brain metastases. Int J Radiat Oncol Biol Phys. GH, Angelov L, et al. Stereotactic radiosurgery for single brain-
2002;53(3):519–26. stem metastases: the Cleveland Clinic experience. Int J Radiat
92. Manon R, O’Neill A, Knisely J, Werner-Wasik M, Lazarus HM, Oncol Biol Phys. 2010;78(2):409–14.
Wagner H, et al. Phase II trial of radiosurgery for one to three 109. Lin CS, Selch MT, Lee SP, Wu JK, Xiao F, Hong DS, et al.
newly diagnosed brain metastases from renal cell carcinoma, mel- Accelerator-based stereotactic radiosurgery for brainstem metas-
anoma, and sarcoma: an Eastern Cooperative Oncology Group tases. Neurosurgery. 2012;70(4):953–8. discussion 8.
study (E 6397). J Clin Oncol. 2005;23(34):8870–6. 110. Lorenzoni JG, Devriendt D, Massager N, Desmedt F, Simon S,
93. Aoyama H, Shirato H, Tago M, Nakagawa K, Toyoda T, Hatano K, Van Houtte P, et al. Brain stem metastases treated with radiosur-
et al. Stereotactic radiosurgery plus whole-brain radiation therapy gery: prognostic factors of survival and life expectancy estima-
vs stereotactic radiosurgery alone for treatment of brain metasta- tion. Surg Neurol. 2009;71(2):188–95; discussion 195, 195–6.
ses: a randomized controlled trial. JAMA. 2006;295(21):2483–91. 111. Kased N, Huang K, Nakamura JL, Sahgal A, Larson DA, McDermott
94. Chang EL, Wefel JS, Hess KR, Allen PK, Lang FF, Kornguth DG, MW, et al. Gamma knife radiosurgery for brainstem metastases: the
et al. Neurocognition in patients with brain metastases treated UCSF experience. J Neurooncol. 2008;86(2):195–205.
with radiosurgery or radiosurgery plus whole-brain irradiation: a 112. Hussain A, Brown PD, Stafford SL, Pollock BE. Stereotactic
randomised controlled trial. Lancet Oncol. 2009;10(11):1037–44. radiosurgery for brainstem metastases: Survival, tumor control,
95. Kocher M, Soffietti R, Abacioglu U, Villà S, Fauchon F, Baumert and patient outcomes. Int J Radiat Oncol Biol Phys. 2007;67(2):
BG, et al. Adjuvant whole-brain radiotherapy versus observation 521–4.
after radiosurgery or surgical resection of one to three cerebral 113. Fuentes S, Delsanti C, Metellus P, Peragut JC, Grisoli F, Regis
metastases: results of the EORTC 22952-26001 study. J Clin J. Brainstem metastases: management using gamma knife radio-
Oncol. 2011;29(2):134–41. surgery. Neurosurgery. 2006;58(1):37–42. discussion 37-42.
96. Soffietti R, Kocher M, Abacioglu UM, Villa S, Fauchon F, 114. Yen CP, Sheehan J, Patterson G, Steiner L. Gamma knife surgery
Baumert BG, et al. A European organisation for research and for metastatic brainstem tumors. J Neurosurg. 2006;105(2):
treatment of cancer phase III trial of adjuvant whole-brain radio- 213–9.
therapy versus observation in patients with One to three brain 115. Shuto T, Fujino H, Asada H, Inomori S, Nagano H. Gamma knife
metastases from solid tumors after surgical resection or radiosur- radiosurgery for metastatic tumours in the brain stem. Acta
gery: quality-of-life results. J Clin Oncol. 2013;31(1):65–72. Neurochir (Wien). 2003;145(9):755–60.
97. Yang CC, Ting J, Wu X, Markoe A. Dose volume histogram anal- 116. Huang CF, Kondziolka D, Flickinger JC, Lunsford LD. Stereotactic
ysis of the gamma knife radiosurgery treating twenty-five meta- radiosurgery for brainstem metastases. J Neurosurg.
static intracranial tumors. Stereotact Funct Neurosurg. 1998;70 1999;91(4):563–8.
Suppl 1:41–9. 117. Hatiboglu MA, Chang EL, Suki D, Sawaya R, Wildrick DM,
98. Yamamoto M, Ide M, Nishio S, Urakawa Y. Gamma Knife radio- Weinberg JS. Outcomes and prognostic factors for patients with
surgery for numerous brain metastases: is this a safe treatment? Int brainstem metastases undergoing stereotactic radiosurgery.
J Radiat Oncol Biol Phys. 2002;53(5):1279–83. Neurosurgery. 2011;69(4):796–806.
99. Suzuki S, Omagari J, Nishio S, Nishiye E, Fukui M. Gamma knife 118. Valery CA, Boskos C, Boisserie G, Lamproglou I, Cornu P,
radiosurgery for simultaneous multiple metastatic brain tumors. J Mazeron JJ, et al. Minimized doses for linear accelerator radiosur-
Neurosurg. 2000;93 Suppl 3:30–1. gery of brainstem metastasis. Int J Radiat Oncol Biol Phys.
100. Bhatnagar AK, Flickinger JC, Kondziolka D, Lunsford 2011;80(2):362–8.
LD. Stereotactic radiosurgery for four or more intracranial metas- 119. Kelly PJ, Lin YB, Yu AY, Ropper AE, Nguyen PL, Marcus KJ,
tases. Int J Radiat Oncol Biol Phys. 2006;64(3):898–903. et al. Linear accelerator-based stereotactic radiosurgery for brain-
101. Kim CH, Im YS, Nam DH, Park K, Kim JH, Lee JI. Gamma knife stem metastases: the Dana002DFarber/Brigham and Women’s
radiosurgery for ten or more brain metastases. J Korean Neurosurg Cancer Center experience. J Neurooncol. 2011;104(2):553–7.
Soc. 2008;44(6):358–63. 120. Mathieu D, Kondziolka D, Flickinger JC, Fortin D, Kenny B,
102. Park SH, Hwang SK, Kang DH, Lee SH, Park J, Hwang JH, et al. Michaud K, et al. Tumor bed radiosurgery after resection of cerebral
Gamma knife radiosurgery for multiple brain metastases from metastases. Neurosurgery. 2008;62(4):817–23. discussion 23-4.
lung cancer. J Clin Neurosci. 2009;16(5):626–9. 121. Quigley MR, Fuhrer R, Karlovits S, Karlovits B, Johnson
103. Yamamoto M, Barfod BE, Urakawa Y. Gamma knife radiosurgery M. Single session stereotactic radiosurgery boost to the post-
for brain metastases of non-lung cancer origin: focusing on mul- operative site in lieu of whole brain radiation in metastatic brain
tiple brain lesions. Prog Neurol Surg. 2009;22:154–69. disease. J Neurooncol. 2008;87(3):327–32.
122. Iwai Y, Yamanaka K, Yasui T. Boost radiosurgery for treatment of 138. Minniti G, Clarke E, Lanzetta G, Osti MF, Trasimeni G, Bozzao
brain metastases after surgical resections. Surg Neurol. A, et al. Stereotactic radiosurgery for brain metastases: analysis of
2008;69(2):181–6. discussion 6. outcome and risk of brain radionecrosis. Radiat Oncol. 2011;6:48.
123. Do L, Pezner R, Radany E, Liu A, Staud C, Badie B. Resection 139. Mayo C, Martel MK, Marks LB, Flickinger J, Nam J, Kirkpatrick
followed by stereotactic radiosurgery to resection cavity for intra- J. Radiation dose-volume effects of optic nerves and chiasm. Int J
cranial metastases. Int J Radiat Oncol Biol Phys. 2009;73(2): Radiat Oncol Biol Phys. 2010;76(3 Suppl):S28–35.
486–91. 140. Mayo C, Yorke E, Merchant TE. Radiation associated brainstem
124. Hwang SW, Abozed MM, Hale A, Eisenberg RL, Dvorak T, Yao injury. Int J Radiat Oncol Biol Phys. 2010;76(3 Suppl):S36–41.
K, et al. Adjuvant Gamma Knife radiosurgery following surgical 141. Marks LB, Yorke ED, Jackson A, Ten Haken RK, Constine LS,
resection of brain metastases: a 9-year retrospective cohort study. Eisbruch A, et al. Use of normal tissue complication probability
J Neurooncol. 2010;98(1):77–82. models in the clinic. Int J Radiat Oncol Biol Phys. 2010;76(3
125. Choi CY, Chang SD, Gibbs IC, Adler JR, Harsh GR, Lieberson Suppl):S10–9.
RE, et al. Stereotactic radiosurgery of the postoperative resection 142. Warnick RE, Darakchiev BJ, Breneman JC. Stereotactic radiosur-
cavity for brain metastases: prospective evaluation of target mar- gery for patients with solid brain metastases: current status.
gin on tumor control. Int J Radiat Oncol Biol Phys. J Neurooncol. 2004;69(1–3):125–37.
2012;84(2):336–42. 143. Rahman M, Hoh BL. Avastin in the treatment for radiation necro-
126. Robbins JR, Ryu S, Kalkanis S, Cogan C, Rock J, Movsas B, et al. sis: exciting results from a recent randomized trial. World
Radiosurgery to the surgical cavity as adjuvant therapy for Neurosurg. 2011;75(1):4–5.
resected brain metastasis. Neurosurgery. 2012;71(5):937–43. 144. Gonzalez J, Kumar AJ, Conrad CA, Levin VA. Effect of bevaci-
127. Hartford AC, Paravati AJ, Spire WJ, Li Z, Jarvis LA, Fadul CE, zumab on radiation necrosis of the brain. Int J Radiat Oncol Biol
et al. Postoperative stereotactic radiosurgery without whole-brain Phys. 2007;67(2):323–6.
radiation therapy for brain metastases: potential role of preopera- 145. Wong ET, Huberman M, Lu XQ, Mahadevan A. Bevacizumab
tive tumor size. Int J Radiat Oncol Biol Phys. 2012;85(3):650–5. reverses cerebral radiation necrosis. J Clin Oncol. 2008;26(34):
128. Ogiwara H, Kalakota K, Rakhra SS, Helenowski IB, Marymont 5649–50.
MH, Kalapurakal JA, et al. Intracranial relapse rates and patterns, 146. Rahmathulla G, Recinos PF, Valerio JE, Chao S, Barnett GH. Laser
and survival trends following post-resection cavity radiosurgery interstitial thermal therapy for focal cerebral radiation necrosis: a
for patients with single intracranial metastases. J Neurooncol. case report and literature review. Stereotact Funct Neurosurg.
2012;108(1):141–6. 2012;90(3):192–200.
129. Soltys SG, Adler JR, Lipani JD, Jackson PS, Choi CY, 147. Lal LS, Byfield SD, Chang EL, Franzini L, Miller LA, Arbuckle
Puataweepong P, et al. Stereotactic radiosurgery of the postopera- R, et al. Cost-effectiveness analysis of a randomized study com-
tive resection cavity for brain metastases. Int J Radiat Oncol Biol paring radiosurgery with radiosurgery and whole brain radiation
Phys. 2008;70(1):187–93. therapy in patients with 1 to 3 brain metastases. Am J Clin Oncol.
130. Jagannathan J, Yen CP, Ray DK, Schlesinger D, Oskouian RJ, 2012;35(1):45–50.
Pouratian N, et al. Gamma Knife radiosurgery to the surgical cav- 148. McHaffie DR, Chabot P, Dagnault A, Suh JH, Fortin MA, Chang E,
ity following resection of brain metastases. J Neurosurg. et al. Safety and feasibility of motexafin gadolinium administration
2009;111(3):431–8. with whole brain radiation therapy and stereotactic radiosurgery
131. Yang HC, Kano H, Lunsford LD, Niranjan A, Flickinger JC, boost in the treatment of ≤6 brain metastases: a multi-institutional
Kondziolka D. What factors predict the response of larger brain phase II trial. J Neurooncol. 2011;105(2):301–8.
metastases to radiosurgery? Neurosurgery. 2011;68(3):682–90. 149. Roberge D, Souhami L, Fortin MA, Pouliot JF. Chemosensitized
discussion 90. radiosurgery for recurrent brain metastases. J Neurooncol.
132. Choi CY, Chang SD, Gibbs IC, Adler JR, Harsh GR, Atalar B, 2012;110(2):265–70.
et al. What is the optimal treatment of large brain metastases? An 150. Sperduto PW, Wang M, Robins HI, Schell MC, Werner-Wasik M,
argument for a multidisciplinary approach. Int J Radiat Oncol Biol Komaki RU, et al. RTOG 0320:A phase III trial comparing whole
Phys. 2012;84(3):688–93. brain radiation therapy (WBRT) and stereotactic radiosurgery
133. Fokas E, Henzel M, Surber G, Kleinert G, Hamm K, Engenhart- (SRS) alone versus WBRT with temozolomide (TMZ) or erlotinib
Cabillic R. Stereotactic radiosurgery and fractionated stereotactic for non-small cell lung cancer (NSCLC) and 1-3 brain metastases.
radiotherapy: comparison of efficacy and toxicity in 260 patients Presented at the AACR 103rd Annual Meeting 2012 in Chicago, IL.
with brain metastases. J Neurooncol. 2012;109(1):91–8. 151. Ma L, Chin LS, DiBiase SJ, Gullapalli R, Kennedy A, Simard JM,
134. Chao ST, Thakkar VV, Barnett GH, Vogelbaum MA, Angelov L, et al. Concomitant boost of stratified target area with gamma knife
Weil RJ, et al. Prospective study of the short-term adverse effects radiosurgery: a treatment planning study. Am J Clin Oncol.
of gamma knife radiosurgery. Technol Cancer Res Treat. 2003;26(4):e100–5.
2012;11(2):117–22. 152. Levivier M, Massager N, Wikler D, Lorenzoni J, Ruiz S, Devriendt
135. Kang TW, Kim ST, Byun HS, Jeon P, Kim K, Kim H, et al. D, et al. Use of stereotactic PET images in dosimetry planning of
Morphological and functional MRI, MRS, perfusion and diffusion radiosurgery for brain tumors: clinical experience and proposed
changes after radiosurgery of brain metastasis. Eur J Radiol. classification. J Nucl Med. 2004;45(7):1146–54.
2009;72(3):370–80. 153. Pan H, Cerviño LI, Pawlicki T, Jiang SB, Alksne J, Detorie N,
136. Barajas RF, Chang JS, Sneed PK, Segal MR, McDermott MW, et al. Frameless, real-time, surface imaging-guided radiosurgery:
Cha S. Distinguishing recurrent intra-axial metastatic tumor from clinical outcomes for brain metastases. Neurosurgery. 2012;71(4):
radiation necrosis following gamma knife radiosurgery using 844–52.
dynamic susceptibility-weighted contrast-enhanced perfusion MR 154. Brown PD, Pugh S, Laack NN, Wefel JS, Khuntia D, Meyers C,
imaging. AJNR Am J Neuroradiol. 2009;30(2):367–72. et al. Memantine for the prevention of cognitive dysfunction in
137. Lawrence YR, Li XA, el Naqa I, Hahn CA, Marks LB, Merchant patients receiving whole-brain radiotherapy: a randomized, dou-
TE, et al. Radiation dose-volume effects in the brain. Int J Radiat ble-blind, placebo-controlled trial. Neuro-oncol. 2013;15(10):
Oncol Biol Phys. 2010;76(3 Suppl):S20–7. 1429–37.
Metastatic Brain Tumors:
Viewpoint—Surgery 15
Raymond Sawaya, David M. Wildrick, and Fassil B. Mesfin
similar to those obtained by combining SRS and WBRT, or

Introduction potentially, even equivalent to the outcome with surgical
resection [16–20]. With this in mind and because of the ease
Metastatic brain tumors continue to occur more frequently of use of SRS and the perception that it costs less than con-
than primary intracranial neoplasms and are a serious com- ventional surgery, some have even suggested that most meta-
plication of systemic cancer [1]. Their annual incidence static brain tumors should be managed exclusively with
exceeds 100,000 [2, 3] and is on the rise as patients live lon- SRS. This controversial notion has already led to treatment
ger from improved treatments for extracranial cancer [4, 5]. of some series of brain metastasis patients with SRS alone
Thus, patients with brain metastases constitute a significant followed by observation [18, 19].
fraction of the case load in the neurosurgery services at many At The University of Texas M.D. Anderson Cancer Center
oncologic care centers. (“M.D. Anderson”), SRS is regarded as a specialized tool to
Early in the 1900s, the results of surgical treatment of sin- be used judiciously, as a given patient’s situation dictates,
gle brain metastases were disappointing [6], with patients sel- rather than as a generalized treatment modality for brain
dom showing a median survival time of longer than 6 months [7]. metastases. We continue to see surgical resection as playing
Although this was longer than the 4–6-week survival time the central role in the management of patients with a limited
typical of patients with untreated brain metastases, the surgi- number of brain metastases [4, 21–24]. In this chapter, we
cal morbidity was high (15–50 %) [8], and whole-brain radia- present our view of the relative roles of surgery and SRS in
tion therapy (WBRT) alone was frequently favored for terms of issues concerning patient selection, treatment out-
management of patients with brain metastases. Since then, come, cost-effectiveness of the treatment, and the patients’
the Radiation Therapy Oncology Group (RTOG) has demon- quality of life. These differences are summarized in
strated that treatment of brain metastasis patients with WBRT Table 15.1 [4, 13, 16, 23–27].
alone can extend their median survival time to 16 weeks [9–11].
In addition, two independent, prospective randomized studies
showed that surgery plus WBRT offered a survival advantage Patient Selection
superior to WBRT alone in such patients [12, 13].
After Leksell demonstrated the utility of focused beams When a patient presents with a brain metastasis and symp-
of high-energy X-rays in the ablation of intracranial tumors toms of mass effect, there is seldom a dispute that the lesion
[14, 15], the technique of stereotactic radiosurgery (SRS) should be removed surgically. Similarly, when a patient with
continued to evolve. During the 1990s, SRS was increasingly a brain metastasis presents in too poor a medical condition to
used in the management of brain metastases, with some be a surgical candidate (or declines surgery), SRS represents
investigators claiming that SRS alone produced results a logical treatment modality. Other patients with single brain
metastases can be sorted into three groups. The first group
has relatively large lesions (exceeding 3 cm in maximum
R. Sawaya, M.D. • D.M. Wildrick, Ph.D. (*) diameter) that can only be effectively removed by surgical
Department of Neurosurgery, The University of Texas
resection. Treatment of these tumors with SRS is not effec-
e-mail: dwildric@mdanderson.org tive because the radiation dose must be reduced as the tumor
size increases, to prevent damage to surrounding brain tis-
F.B. Mesfin, M.D., Ph.D.
Department of Neurosurgery, SUNY Upstate Medical University, sue. This relationship was clearly demonstrated by Mehta
Syracuse, NY, USA et al. [25], who showed that with SRS, the rate of complete
234 R. Sawaya et al.
Table 15.1 Differences between surgery and stereotactic radiosurgery in the treatment of brain metastasis
Surgical resection Stereotactic radiosurgery
Patient selection
Tissue diagnosis Confirms the lesion is a tumor Cannot confirm the lesion is a tumor
Lesion size Large, ≥1.5 cma; especially if there is mass effect Small, ≤2.0 cma; significant mass effect absent
Surgical candidate Yes No (or patient declines surgery)
Treatment outcome
Tumor status Removed (≥94 % on average) [4] Not removed
Local control 85 % for >40 months (up to 5 years) [23] 85 % at 12 months; 65 % at 24 months [25]
Local recurrence rate 8 % [23] to 12 % [4] 30 % [26] to 47 % [27]
Median survival time 10 [13] to 16.4 [23] months 7.5 [23] to 14 [16] months
Complications
Presentation Usually immediate Frequently delayed; necrosis may necessitate surgical resection
Major neurological 7 % in eloquent brain; 6 % overall [4] 25 % in eloquent brain (RTOG grade 3) [24]
Mortality (30-day)b <2 % 1.8 %
Cost-effectiveness Based on 3-day hospital stay Based on 1-day outpatient stay; includes no costs of
maintenance steroids, follow-up office visits, or follow-up MRI
Quality of life
Relief from mass effect Immediate Delayed
Steroid use Tapered off over 2–4 weeks Can last for months; may cause dependence
Follow-up visits Few Many
Patient debilitation Not debilitated at home Debilitated at home
a
Maximum diameter
b
Occurring within 30 days after the procedure
100
GTR Patients in the second group have very small lesions (less
than 10 mm in maximum diameter) that are not surgically
accessible and are located deep within the brain. In this situa-
80 tion, SRS provides an effective alternative to resection, superior
to WBRT, which was the only treatment previously available.
The third group of patients are those who have metastases
60
that are less than 3 cm in maximum diameter and are surgi-
Percent
cally accessible. Whether surgery or SRS is the best treat-

40 ment for these lesions is the subject of much current debate.
CR The 3-cm upper size limit referred to above is probably too
high for adequate SRS treatment. At M.D. Anderson, a rela-
20 tively recent study of 153 brain metastases from melanoma
treated with SRS [28] showed that the 1-year local control
rate of smaller tumors with a maximum diameter of no more
0 than 1.5 cm (volume = 2 cm3) was superior to that of larger
0 2 4 6 8 10 12 14 16 18 20
Tumor Size (cc)
lesions (75.2 % and 42.3 %, respectively; p < 0.05). Moreover,
another study at M.D. Anderson used SRS to treat brain
Fig. 15.1 Tumor size versus response to radiosurgery and surgery for metastases from different primary tumor types [26] and found
brain metastases. Comparison of a representative radiosurgery com- a 1-year actuarial local control rate of 86 % for lesions that
plete response (CR) curve [25] with a surgery gross-total resection
were 1 cm or less in maximum diameter (0.5 cm3) but only a
(GTR) curve (average percentage, 94 % from Table 15.2), showing that
CR rate with radiosurgery is tumor size dependent, whereas GTR rate 56 % local control rate for lesions larger than this (p = 0.0016).
with surgery is not (From Sawaya R. Surgical treatment of brain metas- Furthermore, an earlier study by a radiosurgery group in
tases. Clin Neurosurg 1999; 45:41–47. Used with permission) Pittsburgh [29] included only brain metastases that were no
larger than 2.5 cm in maximum diameter and found that treat-
response (CR; total disappearance of the magnetic resonance ment with SRS plus WBRT was superior to WBRT alone. It
[MR] image of the lesion after SRS) falls off dramatically is important to note that control rates similar to surgery for
with tumor volume such that if a tumor 2 cm3 in volume has lesions as large as 2 cm (volume ~5 cm3) can be achieved with
a 50 % CR rate, an 8–9-cm3 lesion shows about a 20 % CR the addition of fractionated WBRT [30]. Collectively, these
rate (Fig. 15.1). findings clearly indicate that for effective use of SRS alone
15 Metastatic Brain Tumors: Viewpoint—Surgery 235
(i.e., without the addition of WBRT), a cutoff point in brain recent retrospective studies have also labeled resection and
metastasis diameter that is lower than 3 cm is warranted. SRS as equally effective in the treatment of small to moder-
ately sized brain metastases [32].
At M.D. Anderson, Bindal and coworkers [23] also retro-
Treatment Outcome: Survival spectively compared surgical resection and SRS treatment of
brain metastases. Thirty-one patients treated with SRS were
The controversy remains as to whether surgery or SRS is followed prospectively, and 62 patients treated with
more appropriate and effective to treat patients in the group conventional surgery were retrospectively matched to those in
with single brain metastases that are surgically accessible the SRS group. Patients were matched according to primary
and range between 1 and 3 cm in maximum diameter. tumor histology, extent of systemic disease, preoperative KPS
Because no one has yet been able to complete a prospective score, time to brain metastasis, number of brain metastases,
randomized study comparing treatment of such patients with and patient age and sex. WBRT treatment was similar in both
surgical resection and SRS, the question of which of the two groups, and patient eligibility criteria for SRS were the same
is more effective lacks a definitive answer. as for surgery. Lesions treated by SRS were limited to those
In lieu of this, many retrospective studies have been per- <3 cm in maximum diameter, and 81 % of them were deemed
formed comparing surgical resection and SRS for brain surgically resectable. There was a statistically significant sur-
metastasis treatment. Among the more carefully constructed vival advantage in the surgically treated group (16.4 months
studies of this type are those of Auchter et al. [16], Bindal median survival) relative to the group treated with SRS
et al. [23], and Cho et al. [31]. (7.5 months median survival) (p = 0.0009 by multivariate anal-
Auchter and coworkers [16] performed a multi- ysis). In contrast to the conclusions of Auchter et al. [16] and
institutional retrospective outcome and prognostic factor Cho et al. [31], Bindal’s group [23] concluded that surgery
analysis of patients with single cerebral metastases who were was superior to SRS in clinically similar patients in terms of
treated with SRS plus WBRT. From their database of 533 survival, local recurrence, and morbidity. Thus, determination
patients with brain metastases treated with SRS and WBRT, of whether treatment of brain metastases with SRS is better
they selected 122 patients who fulfilled the criteria for surgi- than, equivalent to, or worse than surgical resection must await
cal resection established in the prospective randomized trial a phase III prospective trial comparing SRS with conventional
of Patchell et al. [13], including having a single brain metas- surgery for the treatment of single brain metastases.
tasis that was surgically resectable (but no urgent need for
surgery), no prior radiotherapy or surgical treatment, inde-
pendent functional status (a Karnofsky Performance Scale Tumor Local Control and Recurrence
[KPS] score >70), and a non-radiosensitive tumor. They then
compared the outcome of these 122 patients with that of the When brain metastases are removed surgically, local control
patients in the prospective randomized surgical series of is produced by total removal of the tumor and elimination of
Patchell et al. [13] and Noordijk et al. [12]. The actuarial surrounding edema (Fig. 15.2). At M.D. Anderson, gross-
median survival time after treatment with SRS plus WBRT total resection (equivalent to complete response in radiosur-
was 56 weeks compared with 40 weeks [13] and 43 weeks gery) of metastatic tumors is achieved in 90–96 % of
[12], respectively, after surgery plus WBRT. The median instances, in both eloquent and noneloquent brain regions
duration of functional independence was 44 weeks after SRS (Table 15.2). With surgery, local control (absence of recur-
and WBRT [16] compared with 38 weeks [13] and 33 weeks rent tumor growth after gross-total resection) typically per-
[12], respectively, after surgery and WBRT. These results sists at an 85 % level at 1 year and remains there for an
indicated to Auchter et al. [16] that SRS plus WBRT pro- interval ranging from 40 months to 5 years (Fig. 15.3).
duced outcomes for patients with single brain metastases that
were comparable with, if not better than, surgery plus WBRT. Table 15.2 Gross-total resections performed for metastatic tumors in
Cho and coworkers [31] retrospectively reviewed a series different brain regions (gross-total resection is equivalent to a complete
of 225 single brain metastases in patients treated with either response in radiosurgery)
WBRT alone, SRS plus WBRT, or surgery and WBRT. There Metastasis location
was a similar distribution of prognostic factors such as age, (tumor functional gradea) No. of patients GTR (%)
sex, KPS score, and metastasis location in these three groups Noneloquent (I) 79 75 (95)
except for extracranial cancer, which was 14 % higher in the Near-eloquent (II) 61 55 (90)
Eloquent (III) 54 52 (96)
group treated with SRS and WBRT. The actuarial median
Total 194 Average, 94
survival times for patients treated with SRS and WBRT
(9.8 months) and surgery plus WBRT (10.5 months) were GTR gross-total resection
a
Grade according to functional location
nearly identical, but both were significantly better than for Source: Adapted from Sawaya R. Surgical treatment of brain metasta-
those treated with WBRT alone (3.8 months). Other more ses. Clin Neurosurg 1999; 45:41–47. Used with permission
In contrast, when brain metastases are treated with SRS,

it is common for the tumor to remain visible on computed
tomography (CT) or MR images. Local tumor control with
SRS is less stringently defined and, in addition to tumor
regression, includes lesions remaining unchanged in size as
well as those undergoing no more than a 25 % increase in
size/volume above the baseline measurement. This makes it
difficult and misleading to compare values published for
local control of brain metastases by surgical resection and
SRS. In some patients who have a brain metastasis that is
stable on MR images or increases in size by less than 25 %,
surgery may be required to relieve neurologic deficits result-
ing from mass effect and persistent edema (Fig. 15.4) [33].
After surgical resection of brain metastases at
M.D. Anderson, we have observed local tumor recurrence at
a rate ranging from about 8–12 % [4, 23], which is consistent
with the observations of others. In patients treated for brain
metastases with SRS, similar 1-year local control failure
rates (6–15 %) are often reported [16, 18, 19, 34–38]. These
numbers reported for SRS are flawed not only because of the
less stringent definition of local control with SRS but also
because tumor recurrence in brain metastasis patients is a
function of survival time, and many radiosurgical series of
these patients report such short survival intervals (8 months
on average) that failure of local tumor control cannot be reli-
ably determined. Treatment of brain metastasis patients with
SRS at Brigham and Women’s Hospital [25] gave an 85 %
Fig. 15.2 (a) Preoperative and (b) postoperative gadolinium contrast- local control rate at 1 year, but this dropped steadily to 65 %
enhanced MR images of a brain metastasis in a patient with non-small after 2 years (Fig. 15.3). More recent series of brain metasta-
cell lung cancer. In each panel, the left and right images are T1- and
T2-weighted, respectively (From Sawaya R. Surgical treatment of brain sis patients treated with SRS have presented overall recur-
metastases. Clin Neurosurg 1999; 45:41–47. Used with permission) rence values (30–47 %) [26, 27] that are much higher than
those typical of surgically treated patients. These numbers
may be more realistic, reflecting a trend toward increased
Percent Free From Local Recurrence
100 follow-up monitoring of patients, with more routine neuro-

MDACC Surgery
imaging [28].
80
BRIGHAM Radiosurgery
60
Complications
40
MDACC Radiosurgery Complications arising from surgical resection of brain
metastases are usually evident immediately. Deleterious
20
effects of SRS treatment are frequently delayed and may be
0
missed in follow-up visits, leading to an underreporting of
0 10 20 30 40
complications. A study using SRS alone to treat brain metas-
Follow-up Time (months) tases showed a complication rate of only 8 % [19]; neverthe-
less, 70 % of the complications were acute and included
Fig. 15.3 Freedom from local recurrence (local control) of brain increases in seizures and worsening of preexisting neuro-
metastases with radiosurgery and with surgery. Curve showing time
logic symptoms. Similarly, in a study of recurrence of brain
from radiosurgical treatment to local failure (BRIGHAM Radiosurgery)
in 42 patients treated at Brigham and Women’s hospital [25] superim- metastasis after SRS, Regine et al. [27] observed that recur-
posed on curves showing time from surgical resection (MDACC rence was symptomatic in 71 % of patients and was associ-
Surgery) to local recurrence in 62 patients and time to local failure in 31 ated with a neurologic deficit in 59 %.
patients radiosurgically treated (MDACC Radiosurgery) at The
With the advent of modern neurosurgical techniques,
University of Texas M.D. Anderson Cancer Center [23] (From Sawaya
R. Surgical treatment of brain metastases. Clin Neurosurg 1999; 45: especially intraoperative stereotaxy and ultrasonography,
41–47. Used with permission) few metastatic brain tumors are surgically inaccessible.
Fig. 15.4 CT scan of a patient

with non-small cell lung cancer
who had a 3-cm brain metastasis
in the right anterior parietal lobe
(left) that was treated with
radiosurgery. Three months later,
the tumor shrank by 23 %, but
the edema and the patient’s
symptoms were unchanged
(right). The patient subsequently
underwent a craniotomy and total
resection of the mass. The results
of the surgery are shown in
Fig. 15.2 (From Sawaya
R. Surgical treatment of brain
metastases. Clin Neurosurg 1999;
45:41–47. Used with permission)
Table 15.3 Major neurologic complications of surgery for metastases mortality rates for SRS and surgical resection of brain metas-
in different brain regions
tases are similar.
Metastasis location No. patientsa No. complications (%) Another potential source of complications when treating
Noneloquent 79 1 (1) brain metastases with SRS stems from its inability to provide
Near-eloquent 61 6 (10) a tissue biopsy for pathologic diagnosis. Because 5–11 % of
Eloquent 54 4 (7) patients with a prior history of cancer and a brain lesion that
Average major neurologic (6) appears consistent with a metastasis may actually have non-
deficit rate
a
metastatic disease (primary brain tumor, abscess, multiple
Total = 194 patients
sclerosis plaque, or a hemorrhage) [13, 39], SRS treatment
Source: Adapted from Sawaya R. Surgical treatment of brain metasta-
ses. Clin Neurosurg 1999; 45:41–47. Used with permission of such lesions can lead to an adverse outcome. With surgical
resection of brain metastases, it is easy to histologically con-
firm that the lesion is metastatic cancer, and this situation
In a study of 194 patients at M.D. Anderson [4], the 30-day does not arise.
morbidity rate for any major neurologic deficit after brain
metastasis resection was 6 % (Table 15.3). Even for resec-
tions in eloquent brain regions, this value only rose to 7 %. Cost-Effectiveness
The common perception that SRS is safer to use than con-
ventional surgery to treat brain metastases within or near SRS is frequently promoted as a lower-cost alternative to
eloquent brain areas may not be warranted. In a retrospective conventional surgery in the management of brain metastases.
study of neurologic complications in patients treated with Two publications have emerged using retrospective analyses
SRS for brain metastases at M.D. Anderson, we found that to compare relative costs of SRS and surgery in an attempt to
such complications (including radiation necrosis, seizures, bolster this position [40, 41]. Unfortunately, these cost anal-
and neurologic symptom development or worsening) were yses have relied on surgical expenses and data on length of
highest for tumors in eloquent brain, reaching 25 % for com- hospital stay from publications on patient series that were
plications of RTOG grade 3 and above [24]. already more than a decade old, such as that of Patchell et al.
At M.D. Anderson, the current mortality rate for surgical [13], and have not taken into account the drastic cost-cutting
resection of brain metastases is less than 2 %. Although it measures adopted by most hospitals and surgeons around the
has been claimed that SRS poses little risk of permanent country during the past decade.
neurologic morbidity in brain metastasis treatment, mortali- Moreover, these cost-comparison studies [40, 41] make
ties have occurred. In the study by Bindal et al. [23], sponta- the assumption that SRS and surgery are “equiefficacious”
neous intratumoral hemorrhage occurred in 3 of 31 patients in the management of patients with cancer metastatic to the
treated with SRS, causing the death of one. Kondziolka and brain. They also assign inferior median survival values to
coworkers [29] compared brain metastasis treatment with studies employing surgical resection plus WBRT for brain
SRS plus WBRT versus WBRT alone and reported a 1.8 % metastasis treatment relative to SRS, thereby inflating
mortality rate within 4 weeks after SRS. Thus, it appears that the average cost of surgery. Because there have been no
100
A.Pre B.Post
80
60
% of Patients
40
20
0
28≤ 26 24 22 20 18 16 14 12 10 8 6 4 2 0 2 4 6 8 10 12 14 16 18 20 22 24 26 ≥28
Time (Weeks)
Fig. 15.5 Graphs depicting the duration (in weeks) of steroid treatment ing off of steroid use within 2–4 weeks (From Vecil GG, Suki D,
from the time of conventional resection of the index lesion (Time 0). Maldaun MV, et al. Resection of brain metastases previously treated
The duration of continuous treatment is shown both before (a) and after with stereotactic radiosurgery. J Neurosurg 2005; 102(2): 209–215.
(b) conventional resection. Note that surgical resection facilitated taper- Used with permission)
prospective randomized studies comparing SRS and surgery 7.1 months, 4.2 months, and 2.3 months for RPA classes I, II,
with respect to patient outcomes, there is no accurate and III, respectively [42]. RPA class I patients had a con-
statistical basis for the presumed “equiefficaciousness” or trolled primary tumor, a KPS score ≥ 70, were <65 years old,
the survival numbers put forth. and lacked extracranial metastases, whereas those in RPA
To accurately reflect the total costs of SRS and surgical class III had a KPS score < 70, and the intermediate RPA class
resection for brain metastases, the expenses involved in II consisted of all other patients. In RTOG trial 85-28, the
patient follow-up must be included. Typically, with SRS, only effect of higher doses of radiation was evaluated in patients
the 1-day cost of treatment is considered in cost-effectiveness who had a single brain metastasis [43]. In this trial, patients
studies, with follow-up being ignored. At M.D. Anderson, treated with a fractionated dose of 54.4 Gy or higher showed
follow-up costs after surgery for brain metastases have been improved survival and better neurologic function. When this
severely reduced because most patients are now discharged regimen (54.4 Gy) was compared with 30 Gy administered in
from the hospital after 3 days. Follow-up expenses after SRS ten fractions, no benefit to overall survival was noted [44].
are more significant than for surgery because many routine The neurotoxicity was similar in both trial arms. An
MR imaging studies and office visits are required to monitor M.D. Anderson trial showed that patients treated with SRS
the delayed effects of the radiation. These delayed effects for metastatic brain tumors have a significantly reduced risk
require extended administration of medications such as dexa- of neurotoxicity relative to patients treated with WBRT [45].
methasone, which becomes costly. In our study of patients Concerns about potential neurotoxicity with WBRT and
undergoing resection of brain metastases after failure of SRS the development of SRS have led to the introduction of newer
to control their lesions [33], after their SRS treatment and approaches for the treatment of metastatic brain tumors.
prior to surgery, 40 % of 53 patients required continuous ste- Currently, SRS is emerging as an initial management or as an
roid administration for 12 weeks, which is consistent with adjuvant treatment modality for brain metastases. The effec-
steroid dependency and is not a trivial expense. After surgery, tiveness of SRS is measured by local disease control without
more than 95 % of these patients were able to cease steroid the possible long-term neurotoxic or cognitive side effects of
usage within 2–4 weeks (Fig. 15.5). WBRT [46]. SRS boost improves survival in selected patients
in whom the predominant problem is brain disease rather
than extracranial disease, and it is also used as a salvage
Quality of Life treatment for progressive intracranial disease after surgery or
WBRT. Furthermore, SRS compromises the blood supply to
For many decades, WBRT has been the treatment of choice the tumor by indirect vascular injury and subsequent sclero-
for patients with intracranial metastasis. A recursive parti- sis of blood vessels [47]. This, in fact, supports the notion of
tioning analysis (RPA) performed on patients enrolled for local tumor control and possible tumor burden reductions
WBRT in RTOG trials showed median survival times of after SRS treatment of brain metastases.
Surgical resection of a brain metastasis usually provides

immediate relief of neurologic symptoms produced by mass References
effect of the tumor and the edema surrounding it. Thus, the
patient can be on his feet that much sooner, and few follow- 1. Posner JB, Chernik NL. Intracranial metastases from systemic
cancer. Adv Neurol. 1978;19:579–92.
up visits are required. With SRS treatment of a brain metas-
2. Tsukada Y, Fouad A, Pickren JW, Lane WW. Central nervous
tasis, radiation damage to the tumor frequently takes time to system metastasis from breast carcinoma. Autopsy study. Cancer.
develop, and the therapeutic effect with consequent resolu- 1983;52(12):2349–54.
tion of neurologic symptoms may be delayed for days or 3. Sawaya R, Bindal RK, Lang FF, Suki D. Metastatic brain tumors.
In: Kaye AH, Laws ER, editors. Brain tumors: an encyclopedic
weeks. Multiple follow-up visits are required to monitor the
approach. 3rd ed. London: Elsevier Saunders; 2011. p. 864–92.
tumor’s response to SRS with MR imaging or CT. For symp- 4. Sawaya R. Surgical treatment of brain metastases. Clin Neurosurg.
tom relief, the patient must receive continuous steroid 1999;45:41–7.
administration that can last for months, while remaining 5. Young RF. Radiosurgery for the treatment of brain metastases.
Semin Surg Oncol. 1998;14(1):70–8.
debilitated at home.
6. Grant FC. Concerning intracranial malignant metastases: their fre-
quency and the value of surgery in their treatment. Ann Surg.
1926;84:635–46.
Conclusion 7. Cairncross JG, Posner JB. The management of brain metastases. In:
Walker MD, editor. Oncology of the nervous system. Boston:
Martinus Nijhof; 1983. p. 341–77.
The debate continues as to whether surgical resection or SRS 8. Horwitz NH, Rizzoli HV. Postoperative complications of intracra-
is better in managing tumors metastatic to the brain. Settling nial neurological surgery. Baltimore: Williams & Wilkins; 1982.
the issue of which modality affords patients the longer 9. Borgelt B, Gelber R, Kramer S, Brady LW, Chang CH, Davis LW,
et al. The palliation of brain metastases: final results of the first two
median survival time awaits the outcome of a prospective
studies by the Radiation Therapy Oncology Group. Int J Radiat
randomized trial comparing the two. Both methods have Oncol Biol Phys. 1980;6(1):1–9.
clear benefits. Surgery has the advantages of immediate reso- 10. Diener-West M, Dobbins TW, Phillips TL, Nelson DF. Identification
lution of mass effect, procurement of tissue for pathologic of an optimal subgroup for treatment evaluation of patients with
brain metastases using RTOG study 7916. Int J Radiat Oncol Biol
diagnosis, and lack of the risk of radiation necrosis [48]. SRS
Phys. 1989;16(3):669–73.
carries a decreased risk of hemorrhage and infection, has no 11. Sneed PK, Larson DA, Wara WM. Radiotherapy for cerebral
risk of tumor seeding, and is less invasive, potentially less metastases. Neurosurg Clin N Am. 1996;7(3):505–15.
costly, and requires a shorter hospital stay than standard cra- 12. Noordijk EM, Vecht CJ, Haaxma-Reiche H, Padberg GW,
Voormolen JH, Hoekstra FH, et al. The choice of treatment of sin-
niotomy. Disadvantages of SRS include potential radiation
gle brain metastasis should be based on extracranial tumor activity
necrosis or exacerbation of peritumoral edema and a require- and age. Int J Radiat Oncol Biol Phys. 1994;29(4):711–7.
ment for long-term steroid administration [49, 50]. 13. Patchell RA, Tibbs PA, Walsh JW, Dempsey RJ, Maruyama Y,
In our experience at M.D. Anderson, we believe that the Kryscio RJ, et al. A randomized trial of surgery in the treatment of
single metastases to the brain. N Engl J Med. 1990;322(8):
size and location of the metastases along with the patient’s
494–500.
clinical presentation can be used to recommend one modality 14. Leksell L. The stereotaxic method and radiosurgery of the brain.
or the other. We almost always favor surgery in the case of Acta Chir Scand. 1951;102:316–9.
brain metastases larger than 3 cm in maximum diameter, 15. Leksell L. Stereotactic radiosurgery. J Neurol Neurosurg Psychiatry.
1983;46(9):797–803.
whereas deeply located lesions smaller than 1–1.5 cm in
16. Auchter RM, Lamond JP, Alexander E, Buatti JM, Chappell R,
maximum diameter are usually treated with SRS. If the Friedman WA, et al. A multiinstitutional outcome and prognostic
lesions lend themselves to treatment by either method, we let factor analysis of radiosurgery for resectable single brain metasta-
the patient’s symptoms be the guide; lesions that produce sis. Int J Radiat Oncol Biol Phys. 1996;35(1):27–35.
17. Flickinger JC, Kondziolka D. Radiosurgery instead of resection for
symptoms are more frequently treated surgically, and lesions
solitary brain metastasis: the gold standard redefined. Int J Radiat
that do not can be treated with SRS. Of course, this approach Oncol Biol Phys. 1996;35(1):185–6.
may be modified depending upon the patient’s systemic can- 18. Hasegawa T, Kondziolka D, Flickinger JC, Germanwala A,
cer status or medical condition. Patients who cannot tolerate Lunsford LD. Brain metastases treated with radiosurgery alone: an
alternative to whole brain radiotherapy? Neurosurgery. 2003;52(6):
surgery, who have progressive systemic disease, or who are
1318–26.
expected to live for less than 3 months are treated with 19. Lutterbach J, Cyron D, Henne K, Ostertag CB. Radiosurgery fol-
SRS. The best management strategy for patients with brain lowed by planned observation in patients with one to three brain
metastases should involve the complementary use of surgery, metastases. Neurosurgery. 2003;52(5):1066–73.
20. Sneed PK, Suh JH, Goetsch SJ, Sanghavi SN, Chappell R, Buatti
SRS, and WBRT.
JM, et al. A multi-institutional review of radiosurgery alone vs.
radiosurgery with whole brain radiotherapy as the initial manage-
Acknowledgment We thank the Barbara Falik Research Fund for ment of brain metastases. Int J Radiat Oncol Biol Phys. 2002;
helping make this work possible. 53(3):519–26.
21. Bindal RK, Sawaya R, Leavens ME, Lee JJ. Surgical treatment of 37. Muacevic A, Kreth FW, Horstmann GA, Schmid-Elsaesser R,
multiple brain metastases. J Neurosurg. 1993;79(2):210–6. Wowra B, Steiger HJ, et al. Surgery and radiotherapy compared
22. Bindal RK, Sawaya R, Leavens ME, Hess KR, Taylor with gamma knife radiosurgery in the treatment of solitary cerebral
SH. Reoperation for recurrent metastatic brain tumors. J Neurosurg. metastases of small diameter. J Neurosurg. 1999;91(1):35–43.
1995;83(4):600–4. 38. Petrovich Z, Yu C, Giannotta SL, O’Day S, Apuzzo ML. Survival
23. Bindal AK, Bindal RK, Hess KR, Shiu A, Hassenbusch SJ, Shi and pattern of failure in brain metastasis treated with stereotactic
WM, et al. Surgery versus radiosurgery in the treatment of brain gamma knife radiosurgery. J Neurosurg. 2002;97(5 Suppl):
metastasis. J Neurosurg. 1996;84(5):748–54. 499–506.
24. Dare AO, Sawaya R. Part II: Surgery versus radiosurgery for brain 39. Voorhies RM, Sundaresan N, Thaler HT. The single supratentorial
metastasis: surgical advantages and radiosurgical myths. Clin lesion. An evaluation of preoperative diagnostic tests. J Neurosurg.
Neurosurg. 2004;51:255–63. 1980;53(3):364–8.
25. Mehta MP, Rozental JM, Levin AB, Mackie TR, Kubsad SS, 40. Mehta M, Noyes W, Craig B, Lamond J, Auchter R, French M,
Gehring MA, et al. Defining the role of radiosurgery in the manage- et al. A cost-effectiveness and cost-utility analysis of radiosurgery
ment of brain metastases. Int J Radiat Oncol Biol Phys. vs. resection for single-brain metastases. Int J Radiat Oncol Biol
1992;24(4):619–25. Phys. 1997;39(2):445–54.
26. Chang EL, Hassenbusch 3rd SJ, Shiu AS, Lang FF, Allen PK, 41. Rutigliano MJ, Lunsford LD, Kondziolka D, Strauss MJ, Khanna
Sawaya R, et al. The role of tumor size in the radiosurgical manage- V, Green M. The cost effectiveness of stereotactic radiosurgery ver-
ment of patients with ambiguous brain metastases. Neurosurgery. sus surgical resection in the treatment of solitary metastatic brain
2003;53(2):272–80. tumors. Neurosurgery. 1995;37(3):445–53.
27. Regine WF, Huhn JL, Patchell RA, St Clair WH, Strottmann J, 42. Gaspar L, Scott C, Rotman M, Asbell S, Phillips T, Wasserman T,
Meigooni A, et al. Risk of symptomatic brain tumor recurrence and et al. Recursive partitioning analysis (RPA) of prognostic factors in
neurologic deficit after radiosurgery alone in patients with newly three Radiation Therapy Oncology Group (RTOG) brain metastases
diagnosed brain metastases: results and implications. Int J Radiat trials. Int J Radiat Oncol Biol Phys. 1997;37(4):745–51.
Oncol Biol Phys. 2002;52(2):333–8. 43. Epstein BE, Scott CB, Sause WT, Rotman M, Sneed PK, Janjan
28. Selek U, Chang EL, Hassenbusch 3rd SJ, Shiu AS, Lang FF, Allen NA, et al. Improved survival duration in patients with unresected
P, et al. Stereotactic radiosurgical treatment in 103 patients for 153 solitary brain metastasis using accelerated hyperfractionated radia-
cerebral melanoma metastases. Int J Radiat Oncol Biol Phys. tion therapy at total doses of 54.4 gray and greater. Results of
2004;59(4):1097–106. Radiation Therapy Oncology Group 85-28. Cancer. 1993;71(4):
29. Kondziolka D, Patel A, Lunsford LD, Kassam A, Flickinger 1362–7.
JC. Stereotactic radiosurgery plus whole brain radiotherapy versus 44. Murray KJ, Scott C, Greenberg HM, Emami B, Seider M, Vora NL,
radiotherapy alone for patients with multiple brain metastases. Int J et al. A randomized phase III study of accelerated hyperfraction-
Radiat Oncol Biol Phys. 1999;45(2):427–34. ation versus standard in patients with unresected brain metastases:
30. Shehata MK, Young B, Reid B, Patchell RA, St Clair W, Sims J, a report of the Radiation Therapy Oncology Group (RTOG) 9104.
et al. Stereotactic radiosurgery of 468 brain metastases < or =2 cm: Int J Radiat Oncol Biol Phys. 1997;39(3):571–4.
implications for SRS dose and whole brain radiation therapy. Int J 45. Chang EL, Wefel JS, Hess KR, Allen PK, Lang FF, Kornguth DG,
Radiat Oncol Biol Phys. 2004;59(1):87–93. et al. Neurocognition in patients with brain metastases treated with
31. Cho KH, Hall WA, Lee AK, Gerbi BJ, Higgins PD, Bohen M, et al. radiosurgery or radiosurgery plus whole-brain irradiation: a ran-
Stereotactic radiosurgery for patients with single brain metastasis. domised controlled trial. Lancet Oncol. 2009;10(11):1037–44.
J Radiosurg. 1998;1(2):79–85. 46. Chang EL, Wefel JS, Maor MH, Hassenbusch 3rd SJ, Mahajan A,
32. O’Neill BP, Iturria NJ, Link MJ, Pollock BE, Ballman KV, O’Fallon Lang FF, et al. A pilot study of neurocognitive function in patients
JR. A comparison of surgical resection and stereotactic radiosur- with one to three new brain metastases initially treated with stereo-
gery in the treatment of solitary brain metastases. Int J Radiat Oncol tactic radiosurgery alone. Neurosurgery. 2007;60(2):277–83.
Biol Phys. 2003;55(5):1169–76. 47. Szeifert GT, Massager N, DeVriendt D, David P, De Smedt F,
33. Vecil GG, Suki D, Maldaun MV, Lang FF, Sawaya R. Resection of Rorive S, et al. Observations of intracranial neoplasms treated with
brain metastases previously treated with stereotactic radiosurgery. gamma knife radiosurgery. J Neurosurg. 2002;97(5 Suppl):623–6.
J Neurosurg. 2005;102(2):209–15. 48. Loeffler JS, Alexander 3rd E. Radiosurgery for the treatment of
34. Alexander 3rd E, Moriarty TM, Davis RB, Wen PY, Fine HA, Black intracranial metastases. In: Alexander 3rd E, Loeffler JS, Lunsford
PM, et al. Stereotactic radiosurgery for the definitive, noninvasive LD, editors. Stereotactic radiosurgery. New York: McGraw-Hill;
treatment of brain metastases. J Natl Cancer Inst. 1995;87(1):34–40. 1993. p. 197–206.
35. Flickinger JC, Kondziolka D, Lunsford LD, Coffey RJ, Goodman 49. Sturm V, Kimmig B, Engenhardt R, Schlegel W, Pastyr O, Treuer
ML, Shaw EG, et al. A multi-institutional experience with stereo- H, et al. Radiosurgical treatment of cerebral metastases. Method,
tactic radiosurgery for solitary brain metastasis. Int J Radiat Oncol indications and results. Stereotact Funct Neurosurg. 1991;57(1–2):
Biol Phys. 1994;28(4):797–802. 7–10.
36. Kihlstrom L, Karlsson B, Lindquist C, Noren G, Rahn T. Gamma 50. Kondziolka D, Lunsford LD. Brain metastases. In: Apuzzo MLJ,
knife surgery for cerebral metastasis. Acta Neurochir Suppl (Wien). editor. Brain surgery: complication avoidance and management.
1991;52:87–9. New York: Churchill Livingstone Inc.; 1993. p. 615–41.
Metastatic Brain Tumors: Viewpoint:
Whole Brain Radiation Therapy 16
William F. Regine, Sarah F. Grabowski, and Roy A. Patchell
of progressive systemic cancer and not as a direct result of

Introduction brain metastases. This contrasts with the natural history
of untreated patients, which is that patients live for a median
Whole-brain radiation therapy (WBRT) is the mainstay of of 1–2 months and virtually all die as result of their brain
treatment for most patients with intracerebral metastases. It lesions [2, 5–7]. There is little doubt that WBRT does control
is useful for most newly diagnosed tumors and sometimes or improve symptoms, at least temporarily, in the majority of
also as salvage therapy for recurrent tumors. patients with brain metastases. Responses to radiation can be
seen early in the treatment course, with 50 % of patients
experiencing an improvement in neurologic symptoms by the
Efficacy of WBRT second week of treatment [8]. Overall, the clinical and radio-
graphic response to brain radiation ranges from 50 to 75 %
Given the general consensus that WBRT is an effective treat- [9]. Ultimately, the primary goals of the management of the
ment for brain metastases, it is surprising to note that there patient with brain metastases are to provide effective pallia-
have been limited randomized data comparing WBRT with tion, improve quality of life and, if at all possible, extend sur-
best supportive care or no treatment [1]. To our knowledge, vival. To these ends, stereotactic radiosurgery (SRS) and
only one randomized trial (performed in the pre-computed whole brain radiation therapy (WBRT) are effective treat-
tomography era) has evaluated the addition of WBRT to sup- ment modalities that often play a significant role in the man-
portive care [2]. In this trial, 48 patients with presumed brain agement of brain metastases. Although sometimes thought to
metastases were randomized to steroids with or without be competing modalities, they are, in fact, quite complemen-
40 Gy WBRT. Although p-values were not given, the addition tary; SRS delivers ablative radiation doses with a high degree
of WBRT led to improved duration of remission as well as of precision (typically in one fraction), whereas WBRT deliv-
median survival (14 vs. 10 weeks). Additional support for the ers lower amounts of radiation to the entire brain parenchyma
general efficacy of WBRT comes from comparing the out- (typically over the course of weeks). The former is more
comes of patients after WBRT treatment with the natural his- effective in treating clinically visible tumor deposits, whereas
tory of untreated brain metastases. Data from large the latter treats microscopic foci of disease and each has been
retrospective studies [1, 3, 4] have shown that WBRT shown to improve outcomes in specific settings.
increases the median survival time to 3–6 months and that
more than half of patients treated with WBRT die ultimately
Unsettled Issues
Although there is general agreement on efficacy, no consen-

W.F. Regine, M.D. (*) • S.F. Grabowski, M.D. sus exists on the optimum WBRT dose and schedule for the
Department of Radiation Oncology, University treatment of brain metastases. The best available data on the
of Maryland Medical Center, 22 South Greene Street, effect of dose and schedule for the treatment of brain metas-
Baltimore, MD 21201, USA
e-mail: wregine@umm.edu; sgrabowski@umm.edu
tases comes from several large-scale multi-institutional trials
conducted by the Radiation Therapy Oncology Group
R.A. Patchell
National Brain Tumor Center, Capital Institute
(RTOG) [1, 8, 10–12]. These studies have shown no signifi-
for Neurosciences, Capital Health Medical Center-Hopewell, cant difference in the frequency and duration of response for
Two Capital Way, Suite 456, Pennington, NJ 08534, USA total radiation doses ranging from 2,000 cGy over 1 week to
242 W.F. Regine et al.
5,000 cGy over 4 weeks. Regimens of 1,000 cGy in a single the radiation sensitizer, Motexafin Gadolinium. 401 patients
dose or 1,200 cGy in two doses were less effective and are no were enrolled and each patient underwent a battery of eight
longer in use. Typical radiation treatment schedules for brain neurocognitive function tests at baseline and following treat-
metastases consist of short courses (7–15 days) of WBRT ment. At baseline, 91 % of patients had a deficit, and 42 %
with relatively high doses per fraction (150–400 cGy per had deficits seen in four or more tests. Lesion volume was
day) and total doses in the range of 3,000–5,000 cGy. These predictive of initial neurocognitive deficit. This data is criti-
schedules minimize the duration of treatment while still cal as most studies do not describe baseline neurocognitive
delivering adequate amounts of radiation to the tumor. deficits. Not only does recurrence of intracranial disease
result in significant symptoms, it is also one of the most
important predictors of neurocognitive dysfunction. Meyers
Attempts to Improve WBRT et al. also noted a correlation between tumor progression and
worsening neurocognitive function [16]. Following treat-
Different fractionation and dosing schemes have been tried ment, patients with tumor shrinkage at 2 months greater than
in order to improve the effectiveness of WBRT with varying the population median had improved survival as well as a
results. Epstein et al. [13] reported a phase I/II dose-escalation longer time to neurocognitive dysfunction. Patients with a
study of hyperfractionated radiotherapy using total doses of partial response also had an improvement in execution func-
48, 54.4, 64, and 70.4 Gy. No increased toxicity was identi- tion and visual motor scanning from baseline. In compari-
fied and the three highest dose arms had a statistically signifi- son, patients with progressive disease had a significant
cant improvement in median survival over the lowest dose worsening of neurocognitive symptoms.
arm. This study suggested a dose-dependent effect with the One often quoted retrospective study by DeAngelis et al.
use of hyperfractionated radiation in unresected, single brain [17] suggests that as many as 11 % of long-term survivors
metastasis. Unfortunately, a large randomized RTOG trial, (>12 months) of brain metastases treated with WBRT develop
studying both single and multiple brain metastases, failed to dementia. However, virtually all of the patients in that sam-
show improvement in overall survival using a hyperfraction- ple who developed dementia had been treated with atypically
ated schedule of 54.5 Gy in 34 fractions compared with a large radiation fractionation schedules. The patients treated
standard regimen of 30 Gy in 10 fractions [14]. with fraction sizes of 3.0 Gy or less per day did not develop
Similarly, multiple randomized trials have failed to iden- clinically apparent dementia. Thus, the actual frequency of
tify a benefit to the addition of radiosensitizers or chemoradiation-related dementia when using convention fraction-
therapy to WBRT [9]. Investigators, nonetheless, continue to ation schedules is not known, but is certainly less than 11 %.
search for drugs that will improve the therapeutic ratio In any event, the frequency of long-term neuropsychological
between normal brain tissue and tumor cells. side effects of WBRT in adult brain metastases patients
appears to have been overestimated and seems to be within
the acceptable range when modern fractionation schemes are
Complications of WBRT and the Concern employed. As such, the question is not whether WBRT alone
for Neurotoxicity results in any neurotoxicity, but rather whether its overstated
risks are outweighed by the benefits of reduced intracranial
WBRT has complications. Almost all patients experience a failure with associated neurocognitive dysfunction. In our
temporary loss of hair; although hair growth usually resumes view it generally is. When discussing the pros and cons of
6–12 months after completing therapy. Also, in the short- WBRT with patients, it is important that the physician main-
term, patients may have a transient worsening of neurologic tains appropriate perspective regarding the toxicities associ-
symptoms while receiving therapy. Many physicians believe ated with WBRT. For instance, recently, there has been
that maintaining patients on steroids during radiotherapy accumulating data suggesting that many systemic chemo-
will minimize radiation complications, although conclusive therapies are associated with long-term neurotoxicities [18,
proof of this has not been forthcoming. The long-term side 19]. That is not to say that all chemotherapy need be aban-
effects of radiotherapy are usually not a significant issue in doned, but rather whenever prescribing treatment, one must
the treatment of patients with brain metastases because of the make sure that the benefits outweigh the risks.
relatively short survival time of these patients.
When discussing neurotoxicity, it is important to note that
nearly all patients with brain metastases present with neuro- WBRT With or Without SRS in the Treatment
cognitive deficits prior to beginning WBRT [15]. Meyers of Brain Metastases
et al. reported the most extensive study evaluating neurocog-
nitive function for patients with brain metastases. This Because the majority of patients have multiple metastases,
detailed work examined neurocognitive function in patients the efficacy data for WBRT comes in a large part from treat-
with brain metastases randomized to WBRT with or without ment of patients with multiple brain metastases. There is little
16 Metastatic Brain Tumors: Viewpoint: Whole Brain Radiation Therapy 243
doubt that WBRT is effective for multiple metastases, and it WBRT-alone is the treatment of choice for most patients
is used routinely. The main controversy regarding the treat- with multiple brain metastases (and the word “multiple” in
ment of multiple metastases involves the use of radiosurgery this context means more than one).
(either with or without WBRT).
To date, there have been two randomized trials [20, 21]
assessing the efficacy of radiosurgery in the treatment of Adjuvant WBRT in Conjunction with Focal
multiple metastases. The first randomized trial was reported Treatment of Brain Metastases
by Kondziolka et al. [20]. In that study, 27 patients with mul-
tiple brain metastases were randomized to treatment with For the treatment of single brain metastases, randomized trials
WBRT alone or WBRT plus a radiosurgery boost. The study have established the superiority of focal treatment (either con-
was stopped early because the authors claimed to have found ventional surgery or radiosurgery) plus WBRT over treatment
a large difference in the recurrence rates in favor of radiosur- with WBRT-alone. Therefore, good-prognosis patients with
gery. Unfortunately, the study used nonstandard end points single brain metastases should be treated with upfront surgery
to measure recurrence. The investigators used any change in or radiosurgery. However, with the establishment of the effi-
measurement of the lesion rather than the more usual 25 % cacy of focal treatments for single brain metastases, a new
increase in diameter. No attempt was made to control for ste- controversy has arisen as to whether adjuvant WBRT is really
roid use, radiation changes, or other factors that might pro- necessary after a “complete resection” or “successful” treat-
duce small fluctuations in lesion size on magnetic resonance ment with radiosurgery. Adjuvant WBRT is thought to be of
imaging (MRI). Also, as a study with only 27 patients, it benefit because there may be residual disease in the tumor bed
lacked the statistical power to support any meaningful con- or at distant microscopic sites in the brain. However, brain
clusion, regardless of p values. As a result, this study was metastases tend to be discrete masses that are theoretically
uninterpretable. capable of being removed totally or destroyed, and so WBRT
A second study was reported by Andrews et al. [21]. This may not be necessary after “successful” focal therapy.
study (RTOG 9508) contained 333 evaluable patients with There are several reasons for eliminating WBRT. First,
one to three brain metastases who were randomized to treat- WBRT has adverse, long-term neuropsychological side
ment with either WBRT (37.5 Gy) plus radiosurgery or effects. Second, there are also the costs and time commit-
WBRT (37.5 Gy) alone. The primary end point was survival. ment of the patient that must be considered. And finally,
Overall, there was no significant difference in survival there is the possibility that WBRT may simply not be needed
between the two treatment groups (median, 6.5 months for at all. It is theoretically possible to completely remove single
radiosurgery plus WBRT and 5.7 months for WBRT-alone, brain metastases by surgery, or to control them with radio-
p = 0.1356). There was no survival benefit from radiosurgery surgery. Furthermore, neuroimaging has improved, and it
in patients with multiple metastases (median, 5.8 months for may now be possible to detect reliably additional metastases
radiosurgery plus WBRT and 6.7 months for WBRT-alone, that may be present and treat these with additional focal ther-
p = 0.9776). (However, for patients with single metastases, apy. If these last two statements are true, then there would be
there was a significant survival advantage favoring radiosur- little justification for adjuvant WBRT.
gery, median 6.5 months vs. 4.9 months, p = 0.0393). Lower On the other hand, compelling reasons exist for giving
posttreatment Karnofsky scores and steroid dependence adjuvant WBRT. As a practical matter, it is likely impossible
were more common in the WBRT-alone group. Multiple sub- to completely remove all metastases with conventional sur-
group analyses were performed, and a benefit for radiosurgery, and radiosurgery does not completely control multiple
gery plus WBRT was found in several subgroups that tumors. In addition, neuroimaging may not have reached the
included patients with single and multiple metastases. These point yet where we can be absolutely certain that all metasta-
subgroups were RPA class 1 patients, patients with metastases are being detected, and therefore some type of additional
ses size equal to or larger than 2 cm, and lung cancer patients treatment may be needed. Also, although WBRT does have
with squamous cell histology. However, these subset analy- side effects, these side effects may not be as severe or as
ses were not prespecified, and the p values needed for signifi- common as was previously thought. Furthermore, most
cance should have been adjusted for inflation of type I error. patients with brain metastases have relatively limited overall
When this was done, none of these subgroup analyses survival times, and so the really serious long-term side
showed a positive benefit for radiosurgery [22]. So, for mul- effects are usually not an issue in their care.
tiple brain metastases, this was a completely negative trial Randomized trials [23–25] have addressed the question of
with regard to the major end points prevention of death due adjuvant WBRT in conjunction with focal treatment. A study
to neurologic causes and overall survival. published in 1998 by Patchell et al. [23] examined the effect
Radiosurgery has been put to the test in the treatment of WBRT in conjunction with conventional surgery. In that
of multiple metastases and has not been established as study, 95 patients who had single brain metastases that were
effective. Therefore, based on the best available evidence, completely surgically resected were randomized to treatment
with postoperative WBRT (50.4 Gy) or to observation with sons, the study could not be designed to have this large of a
no further treatment of the brain metastasis (until recur- sample size and, therefore, was not designed to detect mod-
rence). Recurrence of tumor anywhere in the brain was less erate differences in survival, even one as large as 11 %.
frequent in the radiotherapy group than in the observation There is an even stronger reason for discounting the
group (18 % vs. 70 %, p < 0.001). Postoperative radiotherapy apparent lack of efficacy of postoperative WBRT with regard
prevented brain recurrence at the site of the original metasta- to length of survival in the Patchell trial [23]. Recurrence of
sis (10 % vs. 46 %, p < 0.001) and at other sites in the brain tumor in the brain was the primary end point of that random-
(14 % vs. 37 %, p < 0.01). As a result, patients in the radio- ized trial, and this end point was the only truly direct mea-
therapy group were less likely to die of neurologic causes sure of the effects of adjuvant WBRT. Up until recurrence of
than patients in the observation group (6 of 43 who died tumor, the two treatment groups were distinct, and the
[14 %] vs. 17 of 39 [44 %]; p = 0.003). There was no signifi- patients in each had received the treatment assigned by ran-
cant difference between the two groups in overall length of domization. However, at recurrence, no specific treatment
survival or the length of time that patients remained func- was mandated by the study design, and as a result, patients
tionally independent. received a variety of additional treatments. There was an
The effect of WBRT in association with radiosurgery was extremely large crossover of the observation group to WBRT.
examined in a randomized trial conducted by the Japanese Of the 32 patients in the observation group who developed
Radiation Oncology Study Group and reported in abstract recurrent brain metastases, 28 patients got WBRT. Overall,
form by Aoyama et al. [24]. In that study, 132 patients with that means that 61 % (28 patients of 46 total) in the “no
one to four brain metastases were randomized to treatment WBRT” observation arm were, in fact, treated ultimately
WBRT plus radiosurgery or with radiosurgery-alone. The with WBRT. For the purposes of length of survival and func-
WBRT total dose was 30 Gy. The radiosurgery dose was tional independence, the study was virtually a comparison of
18–25 Gy at the periphery of the lesion in the radiosurgery- surgery plus immediate WBRT vs. surgery plus delayed
alone group and was reduced by 30 % in the WBRT plus WBRT. This substantially diluted the effect of WBRT given
radiosurgery group. The median survival time was 7.5 months immediately postoperatively because WBRT probably
in the WBRT plus radiosurgery group and 8.0 months in the improved the length of survival and functional independence
radiosurgery-alone group, p = 0.42. The 12-month brain in the observation group. Therefore, for all the reasons given
metastases recurrence rates were significantly (p < 0.001) above, the Patchell study did not “prove” that WBRT had no
different (47 % in the WBRT plus radiosurgery group and positive effect on survival.
76 % in the radiosurgery-alone group). Death due to neuro- The Aoyama trial [24] also failed to find any difference in
logic causes and neurologic functioning were not signifi- survival with the addition of WBRT. This study differed in
cantly different between the two groups. design from the Patchell study and was devised with survival
Despite the fact that both of the randomized trials [23, 24] as the primary end point; it was originally powered to be able
showed clearly that WBRT prevented recurrences, these to detect a 30 % or greater difference with 89 patients per
studies have actually provoked controversy rather than set- treatment arm. (Given that the Patchell study [23] and numer-
tling the issue. Results of these trials have been used as rea- ous retrospective studies had failed to find a difference in
sons both to give and not to give adjuvant WBRT. The survival even close to 30 %, the Aoyama study [24] should
justification for not giving WBRT holds that because no sur- have been powered as an equivalence trial [if survival was
vival difference was found in either of the trials, WBRT going to be used as the primary end point]). The study was
really adds nothing to the treatment. This argument fails on stopped after an interim analysis at 132 patients indicated
several counts. that the study was not going to show a statistically significant
The Patchell study [23] used tumor recurrence as the pri- difference in survival, even if all of the 178 patients origi-
mary end point and was not designed either to show a differ- nally projected to be in the trial had been randomized.
ence in survival or to rule one out. There was actually an Therefore, any negative conclusion based on the original
increase of 11 % in survival time in the WBRT group when power calculation, that is, that there was no more than a 30 %
compared with the observation group. The relative risk of difference in survival (or any other reasonable difference),
improved survival with WBRT was 1.1. However, this was cannot be supported. This study also had an 11 % crossover
not a statistically significant difference. Because there was a that further eroded the strength of any conclusions about sur-
statistically significant reduction in death due to neurologic vival. In addition, about half of the patients in both treatment
causes, ultimately adjuvant WBRT might have had some arms had multiple brain metastases. The study by Andrews
positive impact on overall survival time. The estimated sam- et al. [21] failed to establish the efficacy of radiosurgery in
ple size required to detect a significant difference of 11 % in the treatment of multiple metastases, and so the inclusion of
overall survival with adequate power would have been 1,005 these patients is problematic and further reduces the number
patients per group or 2010 patients total. For practical rea- of valid patients on which to base a conclusion about survival
benefit. Thus, like the Patchell trial [23] before it, this The trial, while commendable for its effort to prospectively
Aoyama study [24] was unable to show that there was no study neurocognitive function following WBRT, has several
difference in survival. Therefore, arguments based on the shortcomings which make its findings difficult to generalize
supposed lack of efficacy of (immediate) postoperative [28]. For one, the study used patient stratification variables not
WBRT on survival are based on a misunderstanding of the prognostic of the study’s primary neurocognitive endpoint. In
design and limits of the randomized trials. this study, patients were stratified by recursive partitioning
In addition, the Aoyama study demonstrated that omitting analysis class, number of brain metastases, and histology,
WBRT does not produce any difference in either gross neuro- which are factors linked to survival rather than neurocognitive
logic or neurocognitive functioning. From this information, function. The authors did not stratify by baseline neurocogni-
Aoyama et al. [24] and the Journal of the American Medical tive function, examination of which suggests at least a trend
Association (JAMA) editorial writer [26] concluded that the toward increased neurocognitive dysfunction in the WBRT
addition of WBRT is not necessary and can be safely omitted group. Also important is the fact that the authors failed to
in the treatment of most patients with brain metastases. account/stratify for many medications commonly prescribed
However, even if one takes the data presented in the paper at to cancer patients—including opioids, sedatives, anticonvul-
face value, it is possible to draw exactly the opposite conclu- sants, and steroids—that are known to cause neurocognitive
sion. As stated by the authors and the editorial writer, the dysfunction. Second, the authors chose an endpoint with
main reason for not giving WBRT is to avoid the long-term questionable clinical relevance. The primary endpoint of their
neurotoxic effects of WBRT. Yet, this study found no differ- study was an assessment of neurocognitive function using
ence in neurologic functioning, neurocognitive functioning, only the HVLT-R at a single time point of 4 months; however,
gross radiation-induced side effects, or survival times between given the complexity of cognitive function, a battery of tests
the two groups. In fact, deterioration in neurologic function over time is required to accurately depict the neurocognitive
attributable to progression of brain metastases was observed effects of RT [29]. In addition, this primary endpoint inadver-
in 59 % of patients in the WBRT group and 86 % in the SRS- tently biased the trial against WBRT as previous work had
alone group (p = 0.05), indicating a significantly higher rate of already shown that RT has a transient effect on memory, as
neurologic deterioration as a consequence of tumor progres- measured by verbal learning tests. In serial neuropsychologi-
sion in patients when WBRT is withheld. Interestingly, in cal studies, Armstrong and colleagues [30] assessed patients
subsequent analysis among patients with normal MMSE at with low-grade glioma who received RT and found a transient
baseline, WBRT resulted in an increased average duration decline in verbal memory, with the nadir after RT correspond-
until deterioration (17 vs. 8 months, p = 0.05) [27]. Thus, at ing exactly to the 4-month time interval used by Chang and
very least, WBRT appears to significantly reduce the recur- colleagues [25]. Finally, as mentioned above, patients receiv-
rence of brain metastases without demonstrable neurotoxic- ing WBRT inexplicably had a significantly shorter median
ity. Therefore, the trial by Aoyama et al. [24] seems to support survival than patients who received SRS alone, on the order of
strongly the use of WBRT as upfront treatment in the man- a threefold difference (5.7 vs. 15.2 months). This is particu-
agement of most patients with brain metastases. larly perplexing given the improved CNS disease control with
A more recent and widely publicized study evaluating the WBRT (27 % vs. 73 % at 1 year). Moreover, three previous
role of WBRT following SRS for patients with`brain metas- randomized studies reported equivalent survival with the addi-
tases was reported by Chang et al. from MD Anderson (level tion of WBRT to local therapy [23, 24, 31]. What specifically
II evidence) [25]. Patients in this study were randomized to led to such a survival difference is unclear, although close
15–24 Gy SRS with or without 30 Gy WBRT in 12 fractions. inspection of the baseline patient characteristics suggests at
The primary endpoint was a decline in Hopkins Verbal least a trend towards more favorable factors (RPA class I,
Learning Treatment-Revised at 4 months. The trial closed female gender, and lower median tumor volume) in the SRS
early after accrual of only 58 patients based on concern for alone group. The authors suggest that the survival difference
significantly worse neurocognitive outcomes secondary to might be due in part to more frequent surgical salvage in
the addition of WBRT. WBRT patients were found to have a patients who received SRS alone; however, the greater use of
52 % decline in HVLT-R at 4 months, whereas patients in the surgical salvage is more likely to be the result of increased
SRS alone arm had only a 24 % decline. Once again, the CNS disease recurrence rather than a disproportionate ration-
addition of WBRT resulted in decreased 1-year intracranial ing of salvage treatments. Also, although patients receiving
recurrence (27 % vs. 73 %, p = 0.0003), including both WBRT had a higher proportion of systemic deaths, this is due
improved local (100 % vs. 67 %, p = 0.012) and elsewhere to a delay in neurological mortality, rather than an increased
(73 % vs. 45 %, p = 0.02) brain tumor control. Paradoxically, systemic disease burden, a finding originally reported by
and unprecedented in similar randomized trials, patients in Patchell et al. [23]. Perhaps then the survival difference was
the WBRT arm had a significantly compromised median simply the result of statistical anomaly due to small patient
survival (5.7 vs. 15.2 months, p = 0.003). numbers. Regardless, this survival difference could in itself
explain the neurocognitive dysfunction found in patients patients with uncontrolled tumors (−2.42 points) than in
receiving WBRT. The primary endpoint of this study was an patients with controlled tumors (+0.076 points) (p = 0.0046).
assessment of neurocognitive function at 4 months (~1 month All of the patients in this study had received large total doses
prior to the median date of death of patients in the WBRT of conventional radiation therapy.
group). Many studies suggest that patients with terminal can- These studies all strongly suggest that uncontrolled brain
cer experience profound neurocognitive dysfunction [32, 33]. tumors result in a substantial decrease in mental performance
For example, Pereira and colleagues followed 348 cancer and that this reduction far outweighs any decrement seen
patients with serial mini-mental state examinations (MMSE) with cranial radiation therapy. Therefore, the side effects of
and found that patients who died had a trend toward decreased recurrent tumors are worse than the side effects of preventive
scores, with 68 % of patients having abnormal MMSE scores treatment. This is an extremely strong argument for the use
before death [32]. The decline in MMSE score was most proof adjuvant WBRT in association with focal therapy.
found 1 month before death. As such, the higher neurocogni-
tive dysfunction noted in the WBRT group cannot be
confidently ascribed to radiation given the large survival dif- WBRT for Recurrent Brain Metastases
ference noted between treatment groups, along with the tim-
ing of death in the WBRT group in relation to the single Brain metastases often recur, and CNS progression may be
4-month study endpoint (~1 month prior to median survival in accompanied by systemic tumor progression and a decline in
WBRT group). functional status. In general, the same types of treatment
The most forceful argument in favor of adjuvant WBRT used for newly diagnosed brain metastases are also available
involves an examination of the effects of not giving for recurrent tumors. However, the type of previous therapy
WBRT. Patients who do not receive adjuvant WBRT suffer may limit the therapeutic options available at recurrence, and
substantially more recurrent brain metastases than patients the development of radioresistance is not uncommon.
who are treated with WBRT. As previously noted, the harm- If patients have not already had WBRT, they should be
ful side effects of WBRT appear to have been overestimated treated with it on recurrence. However, often patients with
in the past and are probably in the acceptable range. recurrences have already been treated with WBRT, and this
Unfortunately, the same cannot be said of the side effects of limits the amount of subsequent radiation that can be given
recurrence of brain metastases. safely. The amount of additional radiation that can be offered
Several studies [34, 35] have demonstrated that the recur- is usually in the range of 1,500–2,500 cGy, a total dose usu-
rence of brain metastases has a negative effect on the neuro- ally inadequate to control tumor growth.
cognitive functioning of patients. A study by Regine et al. Several retrospective studies [37–41] have attempted to
[34] found that in 36 patients with brain metastases treated assess the efficacy of salvage WBRT. It is difficult to assess
with SRS alone, 47 % had recurrence of brain metastases and efficacy from these reports. Rates of improvement ranged
71 % of the recurrences were symptomatic. Significantly, from 27 to 70 %; however, the range of duration of response
59 % of the patients with recurrent tumors had associated was fairly uniform and was 2.5–3 months. The median sur-
neurologic deficits and 17 % were unable to undergo salvage vival ranged from 1.8 to 4.0 months. Relatively few long-
brain therapy because of their overall poor general status term complications were reported; however, because the
associated with brain tumor recurrence. These findings are median survival is quite short, most patients did not live long
now substantiated by the level 1 evidence provided by the enough to develop the long-term complications of radiation.
Aoyama phase III trial [27] where deterioration in neurologic The problem with the interpretation of these studies is that
function attributable to progression of brain metastases was they often used different end-point measurements for
observed in 59 % of patients in the WBRT group vs. 86 % in improvement and had heterogeneous patient populations.
the SRS-alone group (p = 0.05); indicating a significantly Some included patients with poor performance status and
higher rate of neurologic deterioration as a consequence of extensive disease, while others selected out favorable sub-
tumor progression in patients when WBRT is withheld [24]. groups for radiation. One recommendation based on a retro-
Another study by Regine et al. [35] showed that, at 3 months spective study [37] is to restrict reirradiation to patients who
after treatment, patients treated for brain metastases with showed an initial favorable response to radiotherapy, had a
WBRT had greater negative changes in their mini-mental sta- longer disease-free interval, and who remain in good general
tus examinations with uncontrolled brain tumors than they condition when the cerebral recurrence develops. However,
did with controlled brain tumors (−6.3 points vs. −0.5 points, even in this favorable subgroup, only 42 % of patients
p = 0.02). Also relevant (but perhaps somewhat farther afield) showed symptomatic improvement, and the median survival
was a study by Taylor et al. [36] showing that, in patients after reirradiation was 5 months. Despite such relatively poor
with primary brain tumors at 12 months after treatment, results, additional radiotherapy is frequently one of the few
changes in mini-mental status examinations were worse in treatment options for patients with recurrent disease.
resection of single brain metastases. Neurosurgery. 1989;24(6):

References 798–805.
18. Koppelmans V, Breteler MM, Boogerd W, Seynaeve C, Gundy C,
Schagen SB. Neuropsychological performance in survivors of
1. Berk L. An overview of radiotherapy trials for the treatment of breast cancer more than 20 years after adjuvant chemotherapy.
brain metastases. Oncology. 1995;9(11):1205–19. J Clin Oncol. 2012;30(10):1080–6.
2. Horton J, Baxter DH, Olson KB. The management of metastases to 19. Deprez S, Amant F, Smeets A, Peeters R, Leemans A, Van Hecke
the brain by irradiation and corticosteroids. Am J Roentgenol W, Verhoeven JS, Christiaens MR, Vandenberghe J, Vandenbulcke
Radium Ther Nucl Med. 1971;111(2):334–6. M, Sunaert S. Longitudinal assessment of chemotherapy-induced
3. Posner JB. Management of brain metastases. Rev Neurol. structural changes in cerebral white matter and its correlation with
1992;148(6–7):477–87. impaired cognitive functioning. J Clin Oncol. 2012;30(3):274–81.
4. Cairncross JG, Posner JB. The management of brain metastases. In: 20. Kondziolka D, Patel A, Lunsford LD, Kassam A, Flickinger
Walker MD, editor. Oncology of the nervous system. Boston: JC. Stereotactic radiosurgery plus whole brain radiotherapy versus
Martinus Nijhoff; 1983. radiotherapy alone for patients with multiple brain metastases. Int J
5. Markesbery WR, Brooks WH, Gupta GD, Young AB. Treatment for Radiat Oncol Biol Phys. 1999;45(2):427–34.
patients with cerebral metastases. Arch Neurol. 1978;35(11):754–6. 21. Andrews DW, Scott CB, Sperduto PW, Flanders AE, Gaspar LE,
6. Ruderman NB, Hall TC. Use of glucocorticoids in the palliative Schell MC, Werner-Wasik M, Demas W, Ryu J, Bahary JP, Souhami
treatment of metastatic brain tumors. Cancer. L, Rotman M, Mehta MP, Curran Jr WJ. Whole brain radiation
1965;18(3):298–306. therapy with and without stereotactic radiosurgery boost for
7. Chang DB, Yang PC, Luh KT, Kuo SH, Hong RL, Lee LN. Late patients with one to three brain metastases: phase III results of the
survival of non-small cell lung cancer patients with brain metasta- RTOG 9508 randomized trial. Lancet. 2004;363(9422):1665–72.
ses. Chest. 1992;101(5):1293–7. 22. Elveen T, Andrews DW. Summary of RTOG 95-01 phase III ran-
8. Borgelt B, Gelber R, Kramer S, Brady LW, Chang CH, Davis LW, domized trial of whole brain radiation with and without stereotactic
Perez CA, Hendrickson FR. The palliation of brain metastases: radiosurgery boost, including presentation of a clinical case study.
final results of the first two studies by the Radiation Therapy Am J Oncol Rev. 2004;3:592–600.
Oncology Group. Int J Radiat Oncol Biol Phys. 1980;6(1):1–9. 23. Patchell RA, Tibbs PA, Regine WF, Dempsey RJ, Mohiuddin M,
9. Kwok Y, Pathchell RA, Regine WF. Management of overt central Kryscio RJ, Markesbery WR, Foon KA, Young B. Postoperative
nervous system metastases: brain and spinal cord. In: Pass HI, edi- radiotherapy in the treatment of single metastases to the brain: a
tor. Principles and practice of lung cancer. 4th ed. Philadelphia: randomized trial. JAMA. 1998;280(17):1485–9.
Lippincott Williams & Wilkins; 2010. 24. Aoyama H, Shirato H, Tago M, Nakagawa K, Toyoda T, Hatano K,
10. Gelber RD, Larson M, Borgelt BB, Kramer S. Equivalence of radi- Kenjyo M, Oya N, Hirota S, Shioura H, Kunieda E, Inomata T,
ation schedules for the palliative treatment of brain metastases in Hayakawa K, Katoh N, Kobashi G. Stereotactic radiosurgery plus
patients with favorable prognosis. Cancer. 1981;48(8):1749–53. whole-brain radiation therapy vs stereotactic radiosurgery alone for
11. Kurtz JM, Gelber RD, Brady LW, Carella RJ, Cooper JS. The pal- the treatment of brain metastases: a randomized controlled trial.
liation of brain metastases in a favorable patient population: a ran- JAMA. 2006;295(21):2483–91.
domized clinical trial by the Radiation Therapy Oncology Group. 25. Chang EL, Wefel JS, Hess KR, Allen PK, Lang FF, Kornguth DG,
Int J Radiat Oncol Biol Phys. 1981;7(7):891–5. Arbuckle RB, Swint JM, Shiu AS, Maor MH, Meyers
12. Deiner-West M, Dobbins TW, Phillips TL, Nelson DF. Identification CA. Neurocognition in patients with brain metastases treated with
of an optimal subgroup for treatment evaluation of patients with radiosurgery or radiosurgery plus whole-brain irradiation: a ran-
brain metastases using RTOG study 7916. Int J Radiat Oncol Biol domised controlled trial. Lancet Oncol. 2009;10(11):1037–44.
Phys. 1989;16(3):669–73. 26. Raizer J. Radiosurgery and whole-brain radiation therapy for brain
13. Epstein BE, Scott CB, Sause WT, Rotman M, Sneed PK, Janjan metastases. JAMA. 2006;295(21):2535–6.
NA, Davis LW, Selim H, Mohiuddin M, Wasserman TH. Improved 27. Aoyama H, Tago M, Kato N, Toyoda T, Kenjyo M, Hirota S,
survival duration in patients with unresected solitary brain metasta- Shioura H, Inomata T, Kunieda E, Hayakawa K, Nakagawa K,
sis using accelerated hyperfractionated radiation therapy at total Kobashi G, Shirato H. Neurocognitive function of patients with
doses of 54.4 gray and greater. Results of the Radiation Therapy brain metastasis who received either whole brain radiotherapy plus
Oncology Group 85-28. Cancer. 1993;71(4):1362–7. stereotactic radiosurgery or radiosurgery alone. Int J Radiat Oncol
14. Murray KJ, Scott C, Greenberg HM, Emami B, Seider M, Vora NL, Biol Phys. 2007;68(5):1388–95.
Olson C, Whitton A, Movsas B, Curran W. A randomized phase III 28. Mahmood U, Kwok Y, Regine WF, Patchell RA. Whole-brain irra-
study of accelerated hyperfractionated versus standard radiation in diation for patients with brain metastases: still the standard of care.
patients with unresected brain metastases: a report of the Radiation Lancet Oncol. 2010;11(3):221–2.
Therapy Oncology Group (RTOG) 9104. Int J Radiat Oncol Biol 29. Regine WF, Schmitt FA, Scott CB, Dearth C, Patchell RA, Nichols
Phys. 1997;39(3):571–4. Jr RC, Gore EM, Franklin 3rd RL, Suh JH, Mehta MP. Feasibility
15. Meyers CA, Smith JA, Bezjak A, Mehta MP, Liebmann J, Illidge T, of neurocognitive outcome evaluations in patients with brain metas-
Kunkler I, Caudrelier JM, Eisenberg PD, Meerwaldt J, Siemers R, tases in a multi-institutional cooperative group setting: results of
Carrie C, Gaspar LE, Curran W, Phan SC, Miller RA, Renschler Radiation Therapy Oncology Group trial BR-0018. Int J Radiat
MF. Neurocognitive function and progression in patients with brain Oncol Biol Phys. 2004;58(5):1346–52.
metastases treated with whole-brain radiation and motexafin gado- 30. Armstrong CL, Corn BW, Ruffer JE, Pruitt AA, Mollman JE,
linium: results of a randomized phase III trial. J Clin Oncol. Phillips PC. Radiotherapeutic effects on brain function: double dis-
2004;22(1):157–65. sociation of memory systems. Neuropsychiatry Neuropsychol
16. Li J, Bentzen SM, Renschler M, Mehta MP. Regression after Behav Neurol. 2000;13(2):101–11.
whole-brain radiation therapy for brain metastases correlates with 31. Kocher M, Soffietti R, Abacioglu U, Villa S, Fauchon F, Baumert
survival and improved neurocognitive function. J Clin Oncol. BG, Fariselli L, Tzuk-Shina T, Kortmann RD, Carrie C, Ben Hassel
2007;25(10):1260–6. M, Kouri M, Valeinis E, van de Berge D, Collette S, Collette L,
17. DeAngelis LM, Mandell LR, Thaler HT, Kimmel DW, Galicich JH, Mueller RP. Adjuvant whole-brain radiotherapy versus observation
Fuks Z, Posner JB. The role of postoperative radiotherapy after after radiosurgery or surgical resection of one to three cerebral
metastases: results of the EORTC 22952-26001 study. J Clin Oncol. and implications. Int J Radiat Oncol Biol Phys. 2002;52(2):
2011;29(2):134–41. 333–8.
32. Pereira J, Hanson J, Bruera E. The frequency and clinical course of 36. Taylor BV, Buckner JC, Cascino TL, O’Fallon JR, Schaefer PL,
cognitive impairment in patients with terminal cancer. Cancer. Dinapoli RP, Schomberg P. Effects of radiation and chemotherapy
1997;79(4):835–42. on cognitive function in patients with high-grade gliomas. J Clin
33. Lawlor PG, Gagnon B, Mancini IL, Pereira JL, Hanson J, Suarez- Oncol. 1998;16(6):2195–201.
Almazor ME, Bruera ED. Occurrence, causes, and outcome of 37. Hocht S, Wiegel T, Hinkelbein W. Reirradiation for recurrent brain
delirium in patients with advanced cancer: a prospective study. metastases. Controversies in Neuro-Oncology. Front Radiat Oncol
Arch Intern Med. 2000;160(6):786–94. Basal. 1999;33:327–31.
34. Regine WF, Scott C, Murray K, Curran W. Neurocognitive outcome 38. Hazuka MB, Kinzie JJ. Brain metastases: results and effects of re-
in brain metastases patients treated with accelerated-fraction vs. irradiation. Int J Radiat Oncol Biol Phys. 1988;15(2):433–7.
accelerated-hyperfractionated radiotherapy: an analysis from 39. Kurup P, Reddy S, Hendrickson FR. Results of re-irradiation for
Radiation Therapy Oncology Group Study 91-04. Int J Radiat cerebral metastases. Cancer. 1980;46(12):2587–9.
Oncol Biol Phys. 2001;51(3):711–7. 40. Wong WW, Schild SE, Sawyer TE, Shaw EG. Analysis of outcome
35. Regine WF, Huhn JL, Patchell RA, St Clair WH, Strottmann J, in patients reirradiated for brain metastases. Int J Radiat Oncol Biol
Meigooni A, Sanders M, Young AB. Risk of symptomatic Phys. 1996;34(3):585–90.
brain tumor recurrence and neurologic deficit after radiosurgery 41. Cooper JS, Steinfeld AD, Lerch IA. Cerebral metastases: value of
al in patients with newly diagnosed brain metastases: results reirradiation in selected patients. Radiology. 1990;174(3 Pt 1):883–5.
Stereotactic Radiosurgery
and Radiotherapy in the Management 17
of High-Grade Gliomas
David Roberge and Luis Souhami
markers linked with malignant gliomas including chromo-

Introduction somal deletion, addition, mutation, and gene amplification
that may have important clinical implications [5]. Both GBM
Primary brain tumors are classified according to their pre- and AA can be of two types based on their clinical presenta-
dominant cell type. Glial neoplasms are the most common tion. Primary high-grade astrocytomas occur de novo and are
primary intracranial malignancies and are classified as astro- associated with short duration of symptoms and worse prog-
cytic tumors, oligodendroglial tumors, and ependymal tumors. nosis, while secondary high-grade gliomas often occur in
Astrocytomas are the most common variant of glioma and patients with a previous low-grade astrocytoma [6]. Unique
most adult astrocytomas are high grade. Of these, glioblas- sub-populations, such as secondary frontal tumors with an
toma multiforme (GBM or astrocytoma grade IV) represents isocitrate dehydrogenase 1 (IDH1) mutation, make the case
the most common subtype [1], making this the most common for groups of tumors having distinct pathogenesis [7].
malignant brain tumor in adults. Despite the best use of sur- High-grade glioma is a local disease; distant spread is
gery, radiation, and chemotherapy, long-term survival of uncommon and up to 90 % of initial recurrences are located
patients with GBM has been distinctly uncommon. In within 2 cm of the original enhancing lesion [8–11]. Overall,
unselected series, the 5-year-survival has been as low as 2 % the prognosis of HGG is related to tumor grade, performance
[2, 3]. Anaplastic astrocytoma (AA), or astrocytoma grade III, status, age, and treatment. Five other prognostic factors were
represents less than 20 % of malignant gliomas and are asso- incorporated in a 1993 prognostic scheme based on recursive
ciated with a more favorable outcome [1, 4]. For the scope of partitioning analysis (RPA) of the individual patient data
this chapter, only high-grade gliomas (HGG), which include from three Radiation Therapy Oncology Group (RTOG) tri-
both grade III and IV astrocytomas, will be discussed. als [4]. Using several prognostic variables, patients were
High-grade gliomas are infiltrating tumors that can enlarge divided into six classes with median survivals ranging from
rapidly, resulting in various signs and symptoms, such as 4.7 to 58.6 months. The RPA classification has been fre-
focal or generalized seizures, headache, visual disturbances, quently used in the comparison of treatment results and has
speech disturbances, changes in mental status, and motor or been updated to reflect improved outcomes following initial
sensory deficits. Malignant glial cells are thought to evolve combined treatment with temozolomide (Table 17.1).
from an accumulation of multiple genetic aberrations in nor- Despite intermittent waves of enthusiasm regarding vari-
mal precursor cells. In a stepwise fashion, this accumulation ous treatment modalities, the survival of patients with HGG
of deleterious genetic alterations may result in transforma- has only improved modestly from 1950 to 2000. In this chap-
tion of a glial cell into a low-grade glioma and subsequently ter, we describe and contextualize the use of stereotactic
into an aggressive phenotype associated with high-grade radiosurgery (SRS) and fractionated stereotactic radiother-
tumors. Several molecular studies have generated multiple apy (F-SRT) in the management of primary and recurrent
malignant gliomas.
D. Roberge, M.D. (*)
Department of Radiation Oncology, CHUM—Notre Dame,
1560 Sherbrooke E, Montreal, QC, Canada H2L 4M1
e-mail: david.roberge.chum@ssss.gouv.qc.ca
Historical Perspective
L. Souhami, M.D., F.A.S.T.R.O.
The poor prognosis of HGG is known for many years.
Division of Radiation Oncology, McGill University,
1650 Cedar Avenue, Montreal, QC, Canada H3G 1A4 In 1926, Bailey and Cushing, in reference to what they at
e-mail: Luis.souhami@muhc.mcgill.ca the time called “spongioblastoma multiforme,” made the
250 D. Roberge and L. Souhami
Table 17.1 RTOG recursive partitioning analysis classes [4]

Median survival (months) Two-year survival (%)
Class Original RTOG definition RTOG EORTC/NCIC RTOG EORTC/NCIC
I Age <50, AA, normal mental status 58.6 n/a 76 n/a
II Age ≥50, KPS 70–100, AA, ≥3 months from first symptoms to treatment 37.4 n/a 68 n/a
III Age <50, AA, abnormal mental status or 17.9 21.4 35 43.4
Age <50, GBM, KPS 90–100
IV Age <50, GBM, KPS <90 or 11.1 16.3 15 27.9
Age ≥50, KPS 70–100, AA, ≥3 months from first symptoms to treatment or
Age >50, GBM, surgical resection, good neurologic function
V Age ≥50, KPS 70–100, GBM, either surgical resection and poor neurologic 8.9 10.3 6 16.5
function or biopsy only followed by ≥54.4 Gy EBRT or
Age ≥50, KPS <70, normal mental status
VI Age ≤50, KPS <70, abnormal mental status 4.6 n/a 4 n/a
Age ≥50, KPS 70–100, GBM, biopsy only, <54.4 Gy EBRT
following observation: “Operative procedures, howsoever “While it is very difficult to judge of the effect of radiation in
radical [block extirpations repeated on signs of recurrence; these cases, it is our impression that this therapy is of definite
saturation with X-rays or radium emanations after wide value. A half-hearted attempt of giving every other patient
decompression with or without surgical interference with the Roentgen therapy was never carried through as we felt our-
tumor], have apparently done little more than to prolong life, selves unable to deprive the particular patient of what we felt
save vision, and alleviate headache for an average of a few was the best chance” [15].
months. Whether deep Roentgenization ever does more than It was only in the early 1980s that both the Brain Tumor
hold the growth temporarily in check is problematical” [12]. Cooperative Group (BTCG) and the Scandinavian
By the time Elvidge, Penfield, and Cone published their Glioblastoma Study Group (SGSG) published results of ran-
“McGill series” in 1937, the name “glioblastoma multi- domized trials of adjuvant irradiation [16, 17]. These trials
forme” had become widely accepted and neurosurgery was confirmed a large and significant increase in median survival
still faced with the same “difficult human problem” posed by following the administration of postoperative irradiation, as
these patients [13]. compared to surgery alone. In the Scandinavian trial, 45 Gy
Thirteen years after the publication of Elvidge and col- of whole-brain irradiation increased the median survival
leagues, the following case history was published as part of a from 5.2 months to 10.8 months. In BTSG study 6901,
paper reviewing 70 glioblastoma multiforme patients treated median survival increased from 3.2 to 8.1 months following
at the Montefiore Hospital [14]. 50–60 Gy of whole-brain radiotherapy.
Case 17.1 represented the longest surviving patient in the Looking at the role of chemotherapy, a 1993 meta-analysis
literature of the day [14]. Prognostic factors for the by Fine et al. collated data from 16 randomized trials involv-
Montefiore series were reviewed: “An analysis of factors in ing more than 3,000 patients [18]. Optimistic physicians saw
survival revealed that operation had no significant effect on a large (52 %) difference in 2-year survival; pessimistic phy-
longevity and irradiation a slight effect, if any. The longest sicians saw a relatively small difference in median survival.
survivals occurred in patients with onset at 24–42 years of A subsequent meta-analysis from the Glioma Meta-analysis
age. However, age at onset was not a constant prognostic fac- Trialists group confirmed a 1.5-month increase in median
tor. The sex of the patient, the location of the neoplasm, or survival and failed to demonstrate any benefit of multi-drug
the histologic appearance gave no indication as to longevity. regimens over single agent nitrosourea compounds [19].
A significant lengthening of survival was found in those Adjuvant BCNU chemotherapy was considered standard of
patients whose initial symptom was a motor seizure.” If care by most, although it was used variably in clinical prac-
patients dying within one week of surgery were excluded the tice. Subsequently, in a joint European Organization for
median overall survival was 13.1 months with postoperative Research and Treatment of Cancer (EORTC) and National
irradiation and 12 months without. It must, however, be Cancer Institute of Canada (NCIC) randomized trial [20],
noted that for all cases of prolonged survival reported in this concomitant and adjuvant temozolomide was shown to sig-
paper, the patient had received irradiation. nificantly increase the survival of patients diagnosed with
As demonstrated by the previous series, irradiation had glioblastoma multiforme. The median survival went from
been in common use long before randomized trials demon- 12.1 to 14.6 months and the 2-year survival more than
strating its efficacy. In 1947 Bush and Christensen wrote: doubled (26.5 % vs. 10.4 %). It is of interest to note that the
17 Stereotactic Radiosurgery and Radiotherapy in the Management of High-Grade Gliomas 251
control arm (radiotherapy alone) had the same median sur- survivals were 157 weeks and 88 weeks, respectively, for
vival as reported in the 1949 series from the Montefiore patients with grade III astrocytoma and GBM. Brachytherapy
Hospital, highlighting how little progress had been made in was associated with a high risk of radionecrosis and it was
the previous 50 years. felt that SRS might offer the focal dose escalation benefits
Based on this EORTC/NCIC trial, the current standard of of implants with lessened toxicity. Unfortunately concerns
care for GBM patients with a performance status of 0–1 (on about selection bias [29], as a possible explanation for
subgroup analysis, there was no benefit for patients with an these improved median survivals following brachytherapy,
ECOG performance status of 2) is external beam radiother- appear to be have been confirmed by two subsequent nega-
apy to 60 Gy and temozolomide. Although patients with AA tive phase III trials of brachytherapy by the BTSG [30] and
histology were not included in the trial, this regimen has the Princess Margaret Hospital [31] groups. In these ran-
become for many a de facto standard as we await data from domized studies, the added use of a focal brachytherapy
current prospective studies. boost did not lead to either a clinically or statistically sig-
nificant improvement in survival for patients harboring a
HGG. In the Toronto trial, the median survival was
Rationale for Stereotactic Radiosurgery 13.2 months on the standard arm and 13.8 on the brachy-
therapy arm, while in the BTSG study it was 58.8 weeks on
Infiltrating high-grade glial neoplasms would appear to be the standard arm and 64.1 on the brachytherapy arm.
poor candidates for the stereotactic application of single In 1985, Columbo et al. reported on the first patients
fraction irradiation. These tumors are hypoxic, acute- treated in Vinceza, Italy with a new technique for linear
responding [21] and admixed with normal tissue [22, 23]. accelerator-based radiosurgery using non-coplanar arcs [32].
Despite these biological roadblocks, stereotactic radiation Only 6/22 patients had adequate follow-up. Of these, one
has been pursued in the management of HGG. The use of patient was treated for a 3.5 cm grade III astrocytoma. Two
SRS (defined here as a radiotherapy technique characterized doses of 20 Gy were delivered during separate procedures.
by accurate delivery of high doses of radiation in a single This patient worsened within 2 months and underwent reop-
session to small intracranial targets in such a way that the eration. Following this early experience, several institutional
dose fall-off outside the target volume is very sharp) has reports have been published on the use of SRS boost in the
been based on the pattern of failure of this disease, the dose– management of patients with HGG either at presentation or
response data from external beam radiation, and early data time of recurrence. The non-invasive nature of SRS coupled
from interstitial brachytherapy trials [24, 25]. with the promising early results of interstitial brachytherapy
As mentioned previously, up to 90 % of patients recur led to the inflated enthusiasm for the technique in patients
within 2 cm of the contrast-enhancing lesions—this despite with HGG and a growing number of publications became
the fact that tumor cells can be found pathologically at available over the years (Fig. 17.1). Many of these reports
larger distances, often following the peri-tumoral edema. were published before the results of the randomized trials of
Furthermore, multicentric or metastatic disease is rare [11]. brachytherapy and SRS were available.
Moreover, in an analysis of the Brain Tumor Study Group
data, a dose–response relationship has been shown for doses
of 50 to 60 Gy [26]. In this data set, the median survival
increases from 28 weeks at 50 Gy to 42 weeks at the 60 Gy
level. Significant improvement in median survival was also
observed in the randomized British trial comparing 45–60 Gy
[27]. All of these characteristics made SRS an attractive option
to be used as a focal boost in selected patients with HGG.
In the early 1990s, phase I/II data suggested that inter-
stitial brachytherapy improved local control and survival in
selected primary and recurrent HGG patients. At University
of California at San Francisco (UCSF), the median survival
of the first 18 patients treated with interstitial brachyther-
apy for recurrent GBM was 52 weeks with two patients sur-
viving more than 5 years [28]. In a subsequent NCOG
study [25], 107 patients with HGG were enrolled in a pro-
gram of brachytherapy added to external beam radiation
therapy (EBRT) and adjuvant PCV chemotherapy. In the 63
evaluable patients who were actually implanted, the median Fig. 17.1 Publications on the radiosurgical treatment of HGG
Radiosurgery has predominantly been used in one of two Table 17.2 Comparison of common parameters for stereotactic
scenarios: in primary lesions as a supplement to wider volume irradiation modalities
EBRT and in recurrent lesions as a single modality. Initially Parameter SRS F-SRT
F-SRT was explored to increase the therapeutic ratio in previ- Tumor Size ≤40 mm ≤60 mm
ously irradiated patients; it has since also been explored as a Total Dose 15–24 Gy Variable (∼30 Gy in 5)
boost in the treatment of newly diagnosed HGG patients. Fractions Single Multiple
With the advent of non-invasive immobilization devices and Prescription isodose 50–80 % 70–90 %
sophisticated image-guidance, the use of F-SRT has become Margins 0–2 mm 2–5 mm
more common—it has also become more difficult to distin- SRS stereotactic radiosurgery, F-SRT fractionated stereotactic radiation
therapy
guish from what is now “conventional” radiotherapy. We have
thus chosen to exclude from our review protracted treatment
schedules—even if labeled “stereotactic.” for SRS. In these cases the GTV is the enhancing lesion on the
T1 MRI. On a case-by-case basis, a CTV of a few millimeters
(~2–5 mm) can be added. For rigid immobilization systems
Treatment Planning and systems with intra-fraction image-guidance, no additional
margin is added for the PTV.
Although each case must be approached on an individual Table 17.2 contrasts commonly used treatment parame-
basis, the following general guidelines represent a reason- ters for stereotactic irradiation in HGG.
able approach to the delivery of stereotactic therapy for HGG.
Fractionated Stereotactic Radiation Therapy

For both primary and recurrent lesions, larger volumes can
In the primary treatment of HGG, SRS is used as a boost to be considered for F-SRT than SRS. Depending on the
EBRT. The goal of radiosurgery is to inflict precise damage planned dose of F-SRT and the use (concurrent or prior) of
to tissue within the target volume, in this case proliferating EBRT, volumes of up to approximately 100 cm3 can be con-
glial cells. Thus, only tumors with a limited diameter (at pre- sidered for treatment. Margins will depend on the volume to
sentation or postoperatively) should be considered for be treated, proximity of critical structures, prior treatment,
SRS. The RTOG considered a diameter of 40 mm the maxi- and the image-guidance used. A reasonable approach for
mum diameter allowed for patient entry into the SRS trials. recurrent disease would have the GTV equal to the enhanc-
As a larger area of potential microscopic disease is targeted ing lesion on a gadolinium-enhanced T1 MRI with a com-
by the EBRT, the SRS target volume (PTV) is limited to the bined CTV/PTV margin of 3 mm (mainly accounting for the
tumor without margins. Individual programs have developed reproducibility of standard IGRT or a common three ply
expertise in PET or advanced MRI imaging. As long as the thermoplastic mask-based immobilization system [35, 36]).
technical and clinical limitations of these modalities are Table 17.3 [37–50] contains various published fractionation
understood, their use is reasonable—however, T1 contrast schemes. Depending on the technique, dose would be pre-
MRI remains the standard modality for GTV definition. scribed to the 50–90 % isodose surface after assessing the
Completely resected tumors are generally not considered for plan for volume irradiated, subjective isodose distribution,
SRS boosts but, in selected cases, reactive enhancement or a conformity, homogeneity, and dose to critical structures.
rim of normal-appearing parenchyma can form the radiosur-
gery target [33].
The maximum clinically tolerated doses for the treatment Linear Accelerator vs. Gamma Knife
of HGG have been derived from the RTOG 90–05 phase I
study and are tumor size-dependent [34]. Single doses of 24, HGG are large infiltrating tumors for which debates about
18, and 15 Gy were found to be the maximum tolerated doses fractions of a millimeter are somewhat irrelevant. Because of
for tumors diameters of ≤20 mm, 21–30, and 31–40, respec- the necrotic and presumed radio-resistant core of high-grade
tively. These doses are usually prescribed to an isodose sur- tumors, there might be a theoretical advantage to more inho-
face between 50 and 90 %. Caution should be exercised mogeneous plans. As Gamma Knife (GK) treatments com-
when treating brain stem lesions and lesions within 10 mm monly use multiple isocenters and are prescribed to lower
of the optic chiasm, as these lesions were not included in the isodose surfaces, they might offer a theoretical advantage
RTOG study and the maximum tolerated radiosurgery dose over more homogeneous single isocenter or dose painting
has not been clearly established for these structures. plans (if this were ever clinically verified, linear accelerator
For recurrent disease, similar principles apply. It is also a (LINAC) centers could then choose to prescribe treatments
general rule that only lesions smaller than 4 cm are considered to lower isodose surfaces). This hypothesis was tested by
Table 17.3 Results of F-SRT for primary and recurrent high-grade gliomas
Author Glass [37]* Shepherd [38] Hudes [39] Lederman [40] Regine [41, 42] Baumert [43] Cho [44] Vordermark [45] Cardinale [46] Schwer [47] Fokas [48] Gutin [49] Fields [50]
Institution Temple Royal Marsden Thomas Jefferson Staten Island Kentucky Zurich Minnesota Würzburg RTOG Colorado Marburg MSKCC Colorado
Date of publication 1997 1997 1999 2000 2000 2003 2004 2005 2006 2008 2009 2009 2012
Number of patients 20 recurrent 33 recurrent 20 recurrent 88 recurrent 8 primary 17 primary 10 primary 19 recurrent 76 primary 15 recurrent 53 recurrent 25 recurrent 10 recurrent
1 recurrent
Histology GBM 13 AA 29 GBM 19 GBM GBM 6 GBM 15 GBM 10 GBM 14 GBM GBM 11 GBM GBM 20 GBM 7
AA 7 AO 3 AA 1 AA 2 AA 2 AA 5 AA 4 GR III 5 AA 3
LGG 1
Median age (range) 44.5 (6–73) 37 (19–55)a 52 (26–77) 56 (21–82) ≥18 51 (25–64.8_ 57 (17–80) 50 (11–74) 57.5 (25–83) 47 (23–65) 53 (22–71) 56 (30–80) 40 (22–72)
KPS – 80 (60–100)a 80 (60–100) 70 (50–90) ≥60 (5/12 WHO 70 (60–90) 90 (60–90) Zubrod 0–1 80 (60–90) 70 ≥ 70 60 (60–80)
0/1, 7 WHO 2)
Time from – 29a 3.1 (0.8–45.5) 6.5 n/a n/a n/a 19 (3–116) n/a 12 (3–57) – 14.5 (6–135) 14.5 (7–123)
Diagnosis (for From EBRT from EBRT
recurrent lesions)
Median fu (months) – – 8.5 >12 15a 25 (9–50) – – 22 7 8 6.6 4
Median tumor 14.3 24 (3–93)a 12.66 32.7 7.4 (1.5–27.2)a – 34 (4–70) 15 (4–70) – 41 (8.4–151) – 34 (2–62) 54 (6.7–123)
volume (range) (1.76–122) (1.5–150.3)
Median peripheral 42/7 (–) 35/7 30/10 24/4 14–28/2–4 19 27.5/11 30/6 50 Gy EBRT + 18–36/3 30/5 30/5 36/3
dose (range) (20/4–50/10) (21/7–35/10) (18/4–36/4) 7 (20–35/8–14) (20–30/2–6) 20–28/4 (20–60/5–30)
Median survival 12.7 AA 11 10.5 7 11 20 15.9 9.3 12.5 10 9 GBM 12.5 6
(months)
1-year survival (%) – – 20 17 56 77 67 – – – – GBM 54 –
Median prescription ≥70 90 (80–90) 89 (80–95) 90 (80–90) 50 prescribed 88 (75–90) 80 (70–90) 80–90 90 (80–93) prescribed 94 92 (90–97.6)
isodose (range) at isocenter at isocenter
EBRT All prior All prior All prior EBRT 99 % 59.4/33 60 60 (50.4–60) All prior EBRT 50 All prior EBRT All prior All prior All prior
EBRT? EBRT EBRT EBRT EBRT
Toxicity 15 % 45 % at 24 no GR III – 4/15 GR IVa 6 % necrosis 10 % – 20 % GR IV No symptomatic No > GR II 6 GR IV no 3 DLT
necrosis monthsa necrosis (mainly necrosis nocrosis
hematological)
*With cis-platinum
a
Of a larger group including other histologies
100 [33, 52, 53, 56–66]. Ages ranged from 3 to 84 years and
Karnofsky performance status from 40 to 100. Overall the
Linac median survival of patients treated with SRS was quite
80
Gamma Knife encouraging, although some series reported no benefit in sur-
vival when comparing results to historical controls. However,
Survival Rate
60 in a disease where more than 95 % of patients will ultimately

die of their disease, the median survival is dictated more by
40 patient-related than therapy-related factors [4]. By the very
nature of SRS, patients are selected for small tumors, more
complete resection (size determination being based on post-
20 operative imaging), and a good response to initial therapy
(when SRS is administered after EBRT ± chemotherapy,
0 patients with progressive disease or decreasing performance
0 6 12 18 24 30 36 status are likely to be excluded). Other potential sources of
Months
bias include high patient motivation, more favorable tumor
Fig. 17.2 Survival by SRS treatment delivery system in RTOG 93–05
biology, and aggressive treatment of recurrences. Taking these
(From Souhami L, Seiferheld W, Brachman D, et al. Randomized com- selection factors into account, only one tenth to one quarter of
parison of stereotactic radiosurgery followed by conventional radio- GBM patients will be eligible for radiosurgical boost [67].
therapy with carmustine to conventional radiotherapy with carmustine The results of these radiosurgical series were viewed with
for patients with glioblastoma multiforme: report of Radiation Therapy
Oncology Group 93–05 protocol. Int J Radiat Oncol Biol Phys 2004;
skepticism. In an attempt to reduce bias, authors sought out
60:853–860. Used with permission from Elsevier.) retrospective control populations. One of the largest reviews
[68] attempted to retrospectively stratify 115 patients from
three institutions (Harvard, the University of Florida, and the
Larson and colleagues at UCSF in a Phase II trial [51]. In this University of Wisconsin) according to the prognostic classes
trial prescriptions were made to isodoses as low as 25 %. The of the RTOG recursive partitioning analysis of patients
results, however, appear no different than would be expected enrolled on RTOG 74–01, 79–18, and 82–02 [69–71]. This
with conventional treatment (15 weeks median survival for analysis concluded that there was a significant improvement
patients with recurrent grade IV tumors). in both 2-year and median survival favoring SRS-treated
A review of the single-institution series will reveal a mix patients. Unfortunately, along with other flaws, the RTOG
of GK and LINAC-based treatments. The two largest series classes are broad, do not include all known prognostic fac-
are still from Pittsburgh [52] and Boston [53]—historically tors (most notably—tumor size), and convert important linear
prototypical gamma-knife and LINAC-based programs. Not variable into binary ones (age, mental status and KPS). Thus,
unexpectedly, there is no difference in result between these this approach, along with all retrospective comparisons, is
two series—both reporting a median survival of 20 months inherently flawed. Irish et al. [67] published an elegant anal-
for the treatment of primary tumors. In the initial RTOG ysis of 101 consecutive patients seen at the London Regional
Phase I trial of radiosurgery for recurrent tumors, there Cancer Center. Of these 27 % were deemed eligible for
appeared to be a large difference in results favoring GK in radiosurgery. The median survival of these patients was
the treatment of a mixed bag of primary brain tumors [54]. 23.4 months compared to 8.6 for the radiosurgery-ineligible
This was not born out in the much larger and more homoge- patients. This study highlights the significant effects of
neous experience of the RTOG 93–05 randomized trial [55]. patient selection. It also demonstrates that a patient group
In this trial, in a subgroup analysis, patients treated with can be selected that performs better than the most favorable
LINAC systems had a 14-month survival, not statistically group of glioblastoma patients in the RTOG recursive parti-
different from the 12.1-month survival of GK patients tioning analyses—the RTOG class III patients (age > 50,
(Fig. 17.2). If patients are to be treated with radiosurgery, KPS ≥ 90) for whom the expected median survival was 18
delivery system should not be an issue. months in the pre-temozolomide era. The bottom line is that
only randomization can control efficiently for all confound-
ing factors.
SRS in the Primary Management of High- Having previously demonstrated the feasibility of a multi-
Grade Gliomas institutional radiosurgery trial [54] and encouraged by the
favorable survival of patients treated on phase I/II protocols,
The 1990s saw the publication of several small, retrospective the RTOG opened protocol 93–05 in 1994. This was a pro-
and prospective, series of patients treated with SRS boost in spective randomized trial evaluating upfront SRS followed
the primary treatment of malignant glioma (Table 17.4) by EBRT with BCNU (Arm 1) vs. EBRT and BCNU (Arm 2).
Table 17.4 SRS in the primary management of high-grade gliomas
Kondziolka
Authors Selch [56] Masciopinto [57] Gannett [58] Buatti [59] [52] Shenouda [60] Shrieve [53] Nwokedi [61] Hsieh [62] Smith [33] Biswas [63] Villavicenci [64] Pouratian [65] Einstein [66]
Institution UCLA Wisconsin Arizona Florida UPMC McGill Harvard Maryland Northwestern Phoenix Rochester Multi- Virginia Case Western
center
Year 1993 1995 1995 1995 1997 1997 1999 2002 2005 2008 2009 2009 2009 2012
Number of patients 18 31 30 11 65 14 78 31 25 27 15 20 22 35
Histology 12 GBM GBM 17 GBM 6 GBM 45 GBM GBM GBM GBM GBM GBM GBM GBM GBM GBM
6 AA 10 AA 5 AA 20 AA
KPS (%) 100 % >70 57 % >70 97 % >70 all >90 Mean KPS 90 79 % >70 median 90 61 % >70 70 >60 >70 Median 82 Median 40 Median 90
(50–100)a (50–100) (60–100) (40–100) (60–100)
Median 20 (8–46) 16 (2–60) 24 (2–115) 14 (6–23) 6.5 (1–31)a < 34 10 25 23.6 34.6 13.2 5.8 (0.7–47) 13.4 (4.4–56) –
volume (cc) (1.5–85)
Median MPD 30 (15–35) 12 (10–20) 10 (0.5–18) 13 (10–15) 16 (12–25)a 20 12 (6–24) 17 12 12 10.4 20 (12–25) 5.6 (4–10) 15–24
(range)
Sequence Post-EBRT Pre-EBRT 12 Within 8 12–109 GBM median Pre-EBRT Post-EBRT, Within Post-EBRT Pre-EBRT Post-EBRT Alone or 91 % Pre-EBRT
weeks days 6.2 months median 4 weeks post-EBRT Post-EBRT
Post-EBRT 17 (median 4) post EBRT post diagnosis 14.2 weeks post-EBRT
AA median from diagnosis
3.9 months (range 1–42)
No EBRT 2 (typically post diagnosis
2–3 weeks)
Median survival 9 9.5 GBM 13 17 GBM 20 10 19.9 25 10 11.5 10.3 11.5 15.1 15.8
(months) AA 28 AA 56
2-year GBM 33 – GBM 8 – GBM 41 – 35.9 – – 22 % – – – –
survival (%) AA 100 AA 53 AA 88
Median 62 mean Mean 67 70 80 (70–90) 50 (40–90)a – 85 (60–100) 50 (45–65) 50 50 80 74.9 (66–89) 70 (50–80)a 50
prescription (40–80) (50–100)
isodose (%)
Median age 56 (35–79) 57.7 (20–78) 54 (5–74) 42.1 GBM 51 67.5 (45–78) 51 (12–84) 68 % >50 60.3 61 (19–79)a 57.8 61.3 (27–81) 58.6 (12–76) 62 (12–84)
(range) (15–77) (3–72) (33–81)a
AA 45
(3–73)a
EBRT 78 %, 0–66 Gy Median Median GBM all, 60 Gy 73/78 All, 24/31 > All, median All 60 Gy All 75 % EBRT 96 %, All 60 Gy
45–60 Gy 59.4 Gy 60 Gy mean 60 Gy 59 Gy 60 Gy 50–64 Gy (40–60 Gy) 45–66.6 Gy
(44–62) (54–60)
Chemo 15 % – 47 % None From 1990 None 68 % 100 % 48 % TMZ 75 % 27 % 46 % TMZ
Comments Accelerated Prospective + 1–5 fractions Prospective
EBRT gliadel and MRS-defined
TMZ
a
For larger group, including recurrent tumors
TMZ temozolomide, EBRT external beam radiation therapy
Fig. 17.3 Patient HL, a 73-year-old gentleman treated for glioblas- using a three-isocenter plan using the McGill dynamic stereotactic
toma multiforme on the radiosurgery arm of RTOG 93–05. Prior to technique. 15 Gy was delivered to the 50 % isodose surface, as per pro-
60 Gy of EBRT, the residual tumor in the left temporal lobe was boosted tocol guidelines (a). The tumor failed locally 13 months later (b)
Eligibility criteria required patients to be at least 18 years 100

of age, have a KPS of at least 60 %, and to have a histologi-
RT
cally proved supratentorial, unifocal GBM. All lesions were 80
to be 4 cm or less in maximal diameter. Patients presenting SRS+RT
with tumors larger than 40 mm preoperatively were eligible
Survival Rate
60
only if the postoperative imaging studies showed residual
tumors of ≤40 mm in maximal diameters.
In the investigational arm, radiosurgery was to be given 40
up front, within 5 weeks of surgery. External beam radiation
was the same in both arms: a volume encompassing the 20
tumor, surrounding edema (for the first 46 Gy) and a margin
was treated to a total of 60 Gy with daily fractions of 2 Gy.
BCNU was administered at a dose of 80 mg/m2 on days 1, 2, 0
0 6 12 18 24 30 36
and 3 of EBRT and then repeated every 8 weeks for a total of
Months
6 cycles.
The first patient was treated in February 1994 and the Fig. 17.4 Overall survival by treatment arm on RTOG 9305 (From
study was closed in June 2000 after the 203rd patient was Souhami L, Seiferheld W, Brachman D, et al. Randomized comparison
enrolled (thus meeting the accrual target of 200 patients). of stereotactic radiosurgery followed by conventional radiotherapy with
carmustine to conventional radiotherapy with carmustine for patients
Results were published in 2004 [55]. Figure 17.3 illustrates with glioblastoma multiforme: report of Radiation Therapy Oncology
the case of a patient treated on the experimental arm. Group 93–05 protocol. Int J Radiat Oncol Biol Phys 2004; 60:853–860.
At a median follow-up of 61 months (Fig. 17.4), the Used with permission from Elsevier.)
median survival for Arm 1 was 13.5 months (95 % CI: 11.0–
14.9) and it was 13.6 months (95 % CI: 11.3–15.2) for Arm Both acute and late RTOG grade 3 toxicities were more
2 (p = 0.53). The 2-year actuarial survival rates were 21 % for frequent on Arm 1, although not significantly so. There were
Arm 1 and 19 % for Arm 2. Disappointingly, patterns of fail- no incidences of radiation-related grade 4 toxicity in either
ure were not influenced by the therapy with 90 % of patients arm. A total of seven patients did die of chemotherapy-
presenting with a component of local failure. There was also related toxicity.
no difference in the mini-mental or Spitzer [72] indices (a Questions may arise about the timing of the radiosurgery
validated quality of life index). Despite the exclusion from boost in the RTOG trial. Although both pre and post-EBRT
analysis of seven patients having progressed prior to SRS, no boosts have been used, a look at Table 17.4 will reveal that
subgroup was identified as benefiting from SRS. An post-EBRT boosts are more common in reported series. At
intention-to-treat analysis including all randomized patients the time of the study design in 1993, the choice of an upfront
produced nearly identical results. boost for the RTOG 93–05 trial was motivated primarily by
Table 17.5 Timing of SRS in the primary management of high-grade gliomas

SRS before EBRT SRS after EBRT
Pros Increases number of eligible patients Better definition of tumor margins
All patients will receive protocol program Target may be smaller
Decreases selection bias SRS may be planned far ahead
Patients fully recovered from surgery
Cons Patients not fully recovered from surgery Target may enlarge
Radiosurgery potentially delivered to a larger target Tumor may grow beyond 4 cm
Because of postoperative changes, imaging less satisfactory Performance status may deteriorate
Selection bias
SRS stereotactic radiosurgery, EBRT external beam radiation therapy
three factors: (1) to benefit from controlling potential- vs. broadly concluding that SRS (or F-SRT) is unproven as
accelerated tumor repopulation, (2) to avoid exclusion of any a beneficial treatment for any patient with a malignant
patient entering the radiosurgical arm because of tumor pro- glioma [75].
gression during the EBRT, and (3) to avoid the bias inherent
in selecting patients after fractionated radiotherapy. Two
reports on the use of a focal boost post-EBRT published at SRS for Recurrent High-Grade Gliomas
that time [67, 73] showed that a large proportion of patients
(10–47 %) developed tumor progression while undergoing Treatment of recurrent HGG is essentially palliative; cases of
the EBRT. Even with the strategy of an up front SRS boost, patients surviving more than 5 years are anecdotal. Strategies
7 % of the patients in the SRS arm had to be excluded available [76] include reoperation, chemotherapy, EBRT
because of tumor progression. Biologically early intensifica- reirradiation, SRS, F-SRT, interstitial brachytherapy, and
tion of the radiation seems favorable and there is no convinc- supportive care. Each of these modalities has its own toxici-
ing evidence that a post-EBRT SRS boost is radiobiologically ties and indications. Current published Phase III studies sup-
advantageous over a pre-EBRT SRS boost. port an improvement in median survival from the use of
Case series of post-EBRT boosts cannot report patient BCNU polymers [77] and an improvement in softer measures
outcomes on an intent to treat basis and have included of outcome from the use of oral temozolomide [76, 78, 79].
patients receiving radiosurgery up to a mean of 6.2 months Survival outcomes from prospective randomized trials are
following their diagnosis [52]. The additional bias intro- presented in Table 17.6 [80–83]. There are no phase III trials
duced by including patients who still have small non- of radiosurgery for recurrent tumors and only selected cases
progressing tumors and a good performance status several are appropriate for salvage SRS. Figure 17.5 illustrates the
months after their initial diagnosis is obvious. Of interest, case of a patient treated at our institution with apparent clini-
despite using a different temporal sequence of their boost cal benefit.
implants, in the two randomized trials of stereotactic brachy- An early 1994 series from the University of California
therapy boost [31, 74] both had similar results, failing to San Diego included 15 patients with recurrent HGG [84].
demonstrate a survival benefit for this technique when given Doses prescribed ranged from 12 to 15 Gy based on the tar-
either before or after the EBRT. A summary of the relative get volume. Of the total group of 20 patients, seven suffered
merits of pre and post-EBRT SRS is presented in Table 17.5. intra-cranial pressure-related acute toxicity (fatal in one
Although the Phase III data does not unequivocally disprove patient) and one suffered from a late somnolence syndrome.
any benefit to post-EBRT radiosurgical boosts, there is no Outcome was not reported by tumor type.
reason to believe that this approach is superior. Kondziolka et al. reported on the University of Pittsburgh
Thus, the best evidence currently available does not sup- experience treating 42 patients with recurrent HGG [52].
port an improvement in survival following the addition of a Median survival from radiosurgery was unusually favorable,
SRS to a standard treatment regimen for GBM. There is 30 months for GBM, 31 for AA. There were no cases of
also no evidence of an improvement in quality of life or a acute toxicity and three patients experience late radiosurgery-
change in pattern of failure. There is, however, an increase related morbidity. Overall 22/107 (21 %) of patients in this
in both acute and late radiation-related toxicity. These are series required craniotomy following SRS.
essentially the conclusions of 2005 ASTRO evidence-based With the passage of time, little has changed in the scope
guidelines–guidelines that can be interpreted narrowly as or results of series reporting SRS for recurrent HGG. More
stating that the benefit of SRS is only disproven for T1 recent reports do tend to be limited to GBM and include a
MRI-guided pre-EBRT boosts in the pre-temozolomide era greater proportion of chemotherapy-treated patients. As an
Table 17.6 Survival outcomes from prospective randomized trials of previously irradiated, lesions where there was a concern for
recurrent glioblastoma toxicity. Beginning in 1987, patients were treated at our
Therapy Overall survival (months) institution with a regimen of 42 Gy given in fractions of 7 Gy
Lumustinea [80] 9.8 on alternating days over 2 weeks [91]. For these treatments,
Bevacizumab [81] 9.2 rigid immobilization was performed with a halo type frame
Gliadel wafersa [82] 8.8 [92]. Since this early experience, the general trend for F-SRT
NovoTTF-100A [83] 6.6 has been towards smaller fractions using non-invasive
a
Control arm immobilization.
Authors from the Royal Marsden Hospital have reported
example, in a 2009 report, Villavicencio and colleagues a series of 33 patients treated with F-SRT for recurrent pri-
reported on 26 patients treated with salvage SRS [64]. In this mary brain tumors [38]. This was a dose escalation study
multi-institutional series, all patients had GBM and 96 % conducted from January 1989 to July 1994. All of these
received systemic chemotherapy. Survival from the time of patients had been previously irradiated. Doses ranged from 4
SRS was only 7 months. The longest surviving patient died to 10 daily fractions of 5 Gy. The median survival of the 21
34 months after SRS. patients with recurrent HGG (11 AA, 10 GBM) was
Additional institutional series can be found in Table 17.7 9.6 months. A matched-pair analysis was performed using
[44, 51, 52, 56, 62–65, 84–88]. Common themes from these patients treated for their recurrence with nitrosourea-based
series are: chemotherapy. There was a small, statistically significant,
• Despite median survivals of 7.5–30 months [52, 84], only difference in median survival favoring the RT cohort. Late
one patient is reported as surviving 5 years post- toxicity (not graded) was seen in 36 % of the patients.
radiosurgery (<0.5 %) [51]. Between April 1991 and January 1998, 71 patients with
• Toxicity is incompletely reported with up to 46 % of recurrent tumors (13 % anaplastic oligodendrogliomas, 87 %
patients experiencing treatment complications [84]. high-grade astrocytomas, 15 % of these were dedifferenti-
• Reported prognostic factors include histology, age, KPS, ated low-grade tumors) were treated at the University of
and tumor volume [52, 85, 89]. Minnesota Hospital. A total of 46 of the patients in this retro-
The most credible investigation of the potential benefits spective review were treated with SRS and 25 with F-SRT
of salvage radiosurgery is a case–control study from the [89]. For the SRS cases, the median minimum peripheral
M. D. Anderson Cancer Center [86]. In this report, 41 dose was 17 Gy. F-SRT was delivered to a median of 37.5 Gy
patients treated with salvage SRS were matched for KPS at in 15 fractions. The median survival from stereotactic irra-
presentation, age, year of diagnosis, and size of recurrence diation was 11 months for SRS and 12 for F-SRT (p = 0.3).
with 41 controls. Images of the control patients were Acute complications were seen in 40 % of both groups. Late
reviewed to confirm that the recurrences would have been complications (clinical/pathological necrosis or cranial nerve
eligible for SRS. For the cases, the median survival from palsy) were seen more commonly in the SRS group—30 %
SRS was 11 months vs. 10 months for the controls (calcu- vs. 8 % (p < 0.05). The authors concluded that as the F-SRT
lated from the time of the imaging study showing the SRS- group had worse prognostic factors, a similar median sur-
treatable recurrence). Although the duration of survival was vival, and less late toxicity, it might represent a better treat-
similar in both groups (p = 0.45), the SRS-treated patients ment option for recurrent high-grade gliomas.
had significantly fewer salvage craniotomies (mean: 0.9 vs. Table 17.3 summarizes some of the larger published sin-
1.4). The authors concluded that SRS was of benefit, in par- gle institution series of F-SRT for HGG. Fraction sizes range
ticular in avoiding a number of salvage craniotomies. from 2.5 to 7 Gy for what are mostly recurrent tumors.
Median survival for variable patient populations ranges from
7 to 20 months. Complication rates are incompletely reported
Fractionated Stereotactic Radiation Therapy and range from none to 27 % grade 4 [38, 39].
Authors from Marburg, Germany, reported on 53 patients
When introduced by Leksell more than 50 years ago [90], with recurrent GBM treated at their institution with F-SRT
stereotactic radiosurgery was intended as a means of creating [48]. Between 1998 and 2008, a wide variety of regiments
a “radiation injury.” The intent at that time was to destroy were used to deliver a median total dose of 30 Gy (20–60 Gy)
brain tissue, not selectively inactivate tumor cells intertwined in 2–5 Gy fractions. The reirradiation was well-tolerated
with normal brain tissue. In an attempt to decrease complica- without acute or late grade 3 toxicity. After re-treatment, the
tions through normal tissue repair and to potentially increase median survival was 9 months.
efficacy by allowing reoxygenation and cell cycle reassort- The same issues of bias that plagued the experience with
ment, several centers have treated small series of patients single fraction SRS apply to the experience utilizing
with F-SRT. Often these series are of patients with recurrent, F-SRT. Treatment options for recurrent lesions are numerous.
Fig. 17.5 Patient CB, a 55 year old, was treated for a left frontal GBM 15 Gy was prescribed to the 50 % isodose surface. He subsequently
on the radiotherapy only arm of the EORTC/NCIC protocol. His disease developed what was felt to be an area of asymptomatic radionecrosis (b),
progressed soon after the end of radiotherapy and he was given temo- managed conservatively. He was well until the tumor progressed a sec-
zolomide. The tumor progressed following the second cycle of temo- ond time, 3 years following SRS (c). This second failure was at an edge
zolomide. The patient was switched to BCNU chemotherapy and his of the tumor bed outside the radiosurgery target volume (arrow) and
disease remained stable. Ten months post-EBRT, there was evidence of repeat SRS was performed. Five months later he was reoperated for fur-
further disease progression and he was treated with SRS (a). A dose of ther progression and went on to live for 1 year following his second SRS
If stereotactic radiation is used, the choice of fractionation Motivated by the favorable survival (18–19 months) of
may represent a compromise between convenience and risk patients previously treated with the regimen of Cardinale and
of radionecrosis. colleagues [94], the RTOG conducted a Phase II trial (RTOG
A more definitive answer may not soon come for BR-0023) testing a concurrent F-SRT scheme in patients with
F-SRT. The EORTC and the Medical Research Council had primary supratentorial GBM [46]. This study was open from
jointly initiated a randomized trial of F-SRT in the primary March 2001 to June 2003 and was performed to assess the
management of high-grade gliomas. The patients eligible for feasibility, toxicity, and efficacy of this approach. On this
this study were those with ≤4 cm WHO grade III–IV glio- study, a relocatable immobilizer was used for the 4 weekly
mas, age <65 years with a good performance status. These fractions (5 Gy each for targets >40 mm, 7 Gy for smaller
patients were to be randomized to partial brain radiation targets) of F-SRT that were delivered using 3DCRT (GK and
(54–60 Gy) followed or not by F-SRT (4 × 5 Gy daily). This cone-based SRS systems were not permitted) during weeks
study closed in December 2001 because of poor accrual. 3–6 of EBRT (50 Gy). Radiation was followed by 6 cycles of
Median survival for the 14 patients randomized the F-SRT BCNU (80 mg/m2 for 3 days every 8 weeks). A total of 76
arm was a favorable 21.4 months, but outcomes were not patients were analyzed. Treatments were well tolerated with
reported for the control patients [93]. only one acute grade 4 (lethargy) and one late grade 3 (necro-
The use of hypofractionated F-SRT in combination with sis) toxicity observed. Although patients with gross total resec-
standard radiation therapy has been tested in Phase I/II trials, tions did fare better than expected, the overall median survival
in selected patients with primary GBM. Regine et al. [42] was 12.5 months and no difference was seen when the overall
treated 18 previously unirradiated patients with brain tumors results were compared to the RTOG historical database. Thus,
using a combination of conventional external beam radiation although feasible and well-tolerated, this dose-intense, accel-
therapy and a split-course F-SRT. Of these 18 patients, 11 erated regimen did not appear to improve survival. Although
were malignant gliomas. F-SRT dose schedule was based in the biological concept of F-SRT is interesting, attempting to
the residual tumor volume and varied from 7 Gy given 2–4 deliver a higher dose while shortening the overall treatment
times. F-SRT was delivered pre and post-external beam radi- time to limit accelerated repopulation, this prospective RTOG
ation therapy. Of 15 patients evaluated so far at a median trial suggests that a hypofractionated boost strategy may be of
follow-up of 15 months, excessive toxicity was seen only in limited value. In any case, with the merging of conventional
the group with larger tumor volumes. Of interest, 38 % of the and stereotactic technologies, further attempts at dose escala-
patients requiring reoperation (3/8) were found to have only tion are more likely to involve simultaneous-integrated boost
necrosis without evidence of tumor cells. strategies than intercalated “stereotactic” treatment sessions.
Table 17.7 Results of SRS for recurrent high-grade gliomas
Author Selch [56] Chamberlain [84] Shrieve [85] Kondziolka [52] Cho [44] Larson [51] Larson [51] Mahajan [86] Hsieh [62] Patel [87] Pouratian [65] Biswas [63] Villavicencio [64] Cuneo [88]
Institution UCLA UCSD Harvard University of University UCSF UCSF M.D. Anderson Northwestern Henry Ford Virginia Rochester Multicenter Duke
Pittsburgh of Minnesota
Date of 1993 1994 1995 1997 1999 2002 2002 2005 2005 2008 2008 2009 2009 2012
publication
Number 17 13 86 42 46 26* 54 41 26 26 26 18 26 63
of patients
Histology 12 GBM 5 GBM 86 GBM (14 19 GBM 27 GBM GBM 14, GBM 39, GBM GBM GBM GBM GBM GBM GBM 49,
5 AA 8 AA transformed 23 AA 15 AA GRIII 12 GRIII 15 GRIII 14
from LGG)
4 AO
Time from Median 8 Median 15 Median 10.3 GBM mean Median 10 GBM median GBM median Median 11 Median 6.7 Median 12.5 Median 8 Median Median 13 19.6
diagnosis (4–96) (2.3–115) 18.9 (1–166) 12 (3–50), 7 (2–70), (3–49) 12.1 (4–44) (5–89)
(months) AA mean 19.8 GRIII median GRIII median
43 (7–175) 35 (2–145)
Median KPS 100 % >70 70 (50–100) 90 (70–100) Mean KPS 90 70 (40–90) 90 (70–100) 70 (40–90) 90 (70–100) Mean 70 80 (50–100) 80 (40–100) ≥ 70 84 (50–100) 90
(50–100)† 80 for GBM
Median tumor 28.6 36 (3–50.6) 10.1 6.5 (0.9–31)† 30 (3–125) GBM 8 GBM 9.1 4.7 (0.15–16.3) Mean 21.6 10.4 21 (0.3–110) 12.1 7 (0.4–48.5) 4.8
volume cc (6–121) (2.2–83) (1.6–29.7), (0.3–29.1), (0.3–60) (4.1–44.4)
(range) GRIII 2.7 GRIII 6
(0.4–13.4) (0.3–20.3)
Median 24.4 (mean) 12.5 (12–15.7) 13 (6–20) 15.5 (12–25)† 17 (9–40) GBM 15 GBM 16 – 12 18 (12–20) 6 (3–15) 12 20 (8–25) 15 (12
peripheral (12–17.5), (10–20), patients had
dose Gy GRIII 16.5 GRIII 17 F-SRT)
(range) (15–18) (15–18.5)
Median 10 GBM 15 10.2 GBM 30 11 (1–36) GBM 8.7 GBM 10.1 11 10 8.4 9.4 5.3 7 GBM 9
survival AA 7.5 AA 31 GRIII 15.6 GRIII 13.6
(months)
1-year – GBM 40 – 45 42 – 42 39 – – – – –
survival (%) AA 12.5
Median 64 80 80 (50–100) 50 (40–90)† 50 (40–90) 25–30 50 (45–55) – 50 90 70†† (50–80) 80 (40–90) 77.7 (65–88)
prescription
isodose
(range)
Median age 49 (27–79) 36 (17–62) 46 (9–77) GBM 51 48 (16–75) GBM 53 GBM 50 54 (17–69) Mean 58.2 53 (25–70) 60.7 (12–76) 57.8†† 56.4 (36–82) 47 (19–76)
(range) (3–72), (22–74), (21–77), (33–81)
AA 45 (3–73)† GRIII 44 GRIII 35
(24–62) (24–58)
Prior Chemo – 100 % 33 % – 57 % – – 90 % 73 % All had 10/26 100 % 96 % 100 % mean
subsequent 3.6 prior
chemo regimens
Reoperations 5/17 (46 % 19/86 3/22† 22 % – – 28/41 – 11/26 – 2/18 – –
complication
rate)
Survivors None None None 1 (74 months) None None None None None None None None None None
>5 years
* Patients treated on a protocol with marimastat
†
Including patients treated for a primary glioma
††
For a larger group of patients
surgery, 7 (25 %) had only necrosis in the resection speci-

Treatment Toxicities men, as compared to 3 out of 31 patients (9.6 %) in the
standard arm.
Significant acute complications—i.e., those occurring within Clark et al. [95], from McGill University, have described
days to weeks of treatment—are unusual and generally self- a single number parameter for use during F-SRT that may be
limited. Occasionally an exacerbation of existing symptoms used to represent effective dosage to intracranial structures
may occur, particularly in patients with moderate brain that are irradiated with inhomogeneous dose distribution
edema and with larger lesions (3–4 cm). Most patients will having steep dose gradients. Evaluating biologically effec-
respond to increasing doses of corticosteroids. tive dose volume histograms, the authors obtained an inte-
Late complications attributable to SRS are usually defined gral biologically effective dose (IBED) for each case and
as necrosis within the treatment volume. In patients with demonstrated a threshold value for late damage to the brain
GBM, distinguishing radiation-induced necrosis from tumor stem consistent with similar thresholds that have been deter-
recurrence is often a very difficult problem and may lead to mined for external beam radiation therapy. Using α/β [alpha/
an overestimation of the rate of treatment-induced complica- beta] ratio of 2.5 as representative of the dose–response of
tions. The rate of reoperation for ranges from 0 [58] to 54 % brain stem tissue, they were able to establish a threshold
[53]. However, viable cells are identified in the majority of value in the region of 60 Gy2.5. Despite its apparent value,
such reoperation specimens. Pure necrosis without residual IBED has not been integrated into practice and plan evalua-
tumor cells is a rare event (Fig 17.6). tion typically continues to rely on maximum dose to normal
In the review involving three institutions and 115 patients structures. Hopefully, following future ICRU recommenda-
[68], late complications occurred in 16 % of the patients. tions for small field radiotherapy, dose to a small volume will
Radiation necrosis was diagnosed in the majority of the gradually supplant less reliable point dose reports.
patients (17/19) either by reoperation or by imaging evalua- When symptomatic radiation necrosis does occur, beva-
tion. Prolonged corticosteroid therapy was required in 47 % cizumab offers an alternative to surgery and steroids. Case
of the patients. In the RTOG randomized trial (RTOG 93–05), series have shown benefit but effects on oft intertwined
there were four cases (5 %) of grade III radiation toxicity in tumor and necrosis can be difficult to distinguish in HGG. A
the SRS arm and no cases in the standard arm. Of the 28 small randomized trial that excluded glioblastoma patients
patients in the SRS arm undergoing subsequent salvage was able to show the benefits of a 12-week course of
Fig. 17.6 Patient EM is a 29-year-old patient at the time of diagnosis (c) (potentiated by further chemotherapy) which required emergent cra-
of glioblastoma. From left to right, T1 contrast MRIs: after the initial niotomy (d) and resection. The patient is now without evidence of
surgery (a), at the time of 18 Gy SRS salvage (b) (following 60 Gy active disease more than 2 years from diagnosis (e)
EBRT, TMZ and reoperation), showing symptomatic radiation necrosis
bevacizumab as all bevacizumab-treated patients had both Results of the Temple University and Staten Island look-
clinical and radiological response vs. none of the control ing at chemotherapy (cis-platinum and Taxol) and F-SRS for
patients (p = 0.0013) [96]. recurrent glioma have been reported [37, 99]. Retrospectively,
these results are similar to patients treated without chemo-
therapy and further use of these agents with SRS is unlikely.
Future Directions in Stereotactic Radiation A phase I trial (NCI-T99–0041) investigating the sensiti-
for Glial Neoplasms zation of SRS with Gadolinium Texaphyrin was reported at
AACR-NCI-EORTC conference with emphasis on the 8 T
New strategies in the application of stereotactic irradiation MRI correlative studies [101]. In this trial GBM patients
for malignant gliomas include: with a KPS >60 and a tumor <40 mm were first treated with
• Changes in imaging. conventional EBRT. Following EBRT an IV infusion of
• Concurrent use of systemic chemotherapy. Gadolinium Texaphyrin was given 3 h prior to a 10 Gy SRS
• İntegration with bevacizumab. boost. In the first eight patients accrued, the median time to
There is persistent interest in functional imaging modali- progression was a disappointing 4.5 months.
ties for HGG. This might be of interest in target definition, The UCSF group has published its experience using an
especially for recurrent tumors treated with SRS where non- MMPI inhibitor (Marimastat starting on the day of SRS and
enhancing tumor may be missed by gadolinium-enhanced continued until progression, death, or toxicity) in combina-
T1 MRI images. The group from UCSF has compared SRS- tion with SRS for recurrent HGG [51]. In this phase I trial,
treated volumes with metabolically active tumor volumes patients’ outcomes were not clearly different than expected
defined by proton magnetic resonance imaging (1H-MRS) for patients with recurrent HGG.
[97]. A total of 26 patients were retrospectively divided into Investigators from Colorado have reported two Phase I
high-risk or low-risk groups according to the amount of trials of kinase inhibitors and radiation for recurrent HGG
overlap between the metabolic target and the SRS target. [47, 50]. In a trial with gefitinib, the drug dose was constant
The median survival was 15.7 months for patients in the (250 mg/day) and the F-SRT was escalated (from 18 Gy in 3
low-risk group and 10.4 months for those in the high-risk fractions to 36 Gy in 3 fractions). In the subsequent trial with
group. vandetanib, the F-SRT dose was constant (36 Gy in 3) and
In a trial for newly diagnosed GBM patients, MRS was the drug escalated. Drug and radiation toxicities were addi-
used to target a post-resection, pre-EBRT radiosurgery boost tive without a clear increase in radiation effects or median
[66]. As a novel means of managing the large 10 × 10 × 15 mm survival (6 and 10.3 months).
MRS voxels, an 8 mm Gamma Knife “shot” was used to tar- Bevacizumab, a humanized monoclonal antibody to
get each suspicious voxel. Dose was selected as per RTOG VEGF, is unique in that it promises to both increase the
90–05. Radiosurgery was followed by 60 Gy of EBRT. As anti-tumor effect of radiation at the same time as may lower
the results of the EORTC/NCIC trial became available, the risk of symptomatic necrosis—particularly appealing in
temozolomide was incorporated in the regimen. Survival for re-irradiation. Bevacizumab already appears to improve
temozolomide-treated patients was 20.8 months. progression-free survival in primary (Roche press release, Aug
Combining new strategies in HGG stereotactic radiother- 2012) and recurrent [81] glioblastoma and has pre-clinical and
apy, Neider et al have presented the results of a phase II study clinical support for synergy with radiation. Early outcomes
looking at adding chemotherapy and molecular targeting to from radiosurgery plus bevacizumab have been reported. A
F-SRT for recurrent HGG [98]. Forty-six patients were retrospective analysis from the Duke group compared recur-
treated with F-SRT (30 Gy in 6 fractions) to a target defined rent GBM patients having received SRS with (33 patients) or
using C11-methionine PET, CT, and MRI information. In 30 without (16 patients) bevacizumab [88]. Although the cases
patients, F-SRT was sandwiched in a 6-month course of and control were not matched, the groups were similar and the
temozolomide. Median survival was 11 months with temo- median survival of the patients who received bevacizumab was
zolomide, 6 months without. Toxicity was acceptable with significantly longer (11 vs. 4 months, p = 0.015). Bevacizumab
three reoperations for apparent radionecrosis, which, on patients also had lower rates of radionecrosis (5 % vs. 19 %).
pathology, was combined with recurrent tumor. Based on this experience, Duke investigators conducted a
Temozolomide is now in standard use for HGG, but there prospective study to evaluate SRS plus bevacizumab for
have been a number of small trials combining SRS or F-SRT recurrent high-grade gliomas [102]. Fifteen patients (9
with other alkylators (carmustine implants, cis-platinum) GBM, 6 grade III) were treated: SRS for lesions 0–2.9 cm,
[33, 41], other conventional chemotherapy agents (pacli- F-SRT lesions measuring 3–5 cm. Patients received 10 mg/
taxel) [99] kinase inhibitors (gefitinib, vandetanib) [47, 50], kg of bevacizumab on the day of radiosurgery and again 14
or dedicated radiation sensitizers (hyperbaric oxygen, days later. No grade 4 toxicity was seen and the median sur-
motexafin gadolinium) [100, 101]. vival was 13 months.
Gutin and colleagues at the Memorial Sloan-Kettering

Cancer Center also reported a prospective study of bevaci- glioma, probably a glioblastoma multiforme. Following
zumab and SRS for recurrent HGG [49]. Twenty-five patients the operation, the patient as given a course of roentgen
(20 GBM, 5 grade III) received bevacizumab 10 mg/kg every therapy for a total of 6,068 r … .
14 days along with 30 Gy in 5 fractions of F-SRT. No radia- In September, 1937, he was struck by an automo-
tion necrosis was seen. For the GBM patients, the median bile and was unconscious for 24 h. Sometimes thereaf-
overall survival was 12.5 months. ter his seizures returned. He gradually lost power on
Finally, Park and colleagues reported on 11 GBM patients the right side, and the aphasia became worse.
treated at recurrence with bevacizumab and SRS (median In October, 1939, he was seen by Dr. Leo Davidoff
16 Gy, range 13–18 Gy) [103]. These patients were compared … .An encephalogram was interpreted as showing
to SRS-treated 44 case-matched controls irradiated without left-sided cerebral atrophy without evidence of
bevacizumab. Overall survival was improved in the bevaci- regrowth. Nothing further was done.
zumab-treated patients (18 months vs. 12 months, p = 0.005) He entered Montefiore Hospital in August, 1939, at
and adverse radiation effects were less common (9 % vs. the age of 49 … .There were now astereognosis in the
46 %, p = 0.037). right hand, a partial motor aphasia, and some blurring of
The bottom line is that, despite past failures, as long as a the right disc margin … .During his stay in the hospital,
majority of HGG continue to fail locally, there will be inter- both focal motor and generalized seizures were noted.
est in dose intensification and local re-irradiation. As new Irradiation was not deemed advisable, and he was dis-
techniques to achieve this are tested [104], it will be impor- charged to be followed in the outpatient clinic. He had
tant to better report toxicity and remain wary of selection difficulties in expressing himself. Because of this, he
bias to which small cohorts are prone. insisted on being re-admitted in July 1941. At this time
it was felt that the paretic side was more spastic, and that
the motor aphasia now had a sensory component … .
Conclusion He was given a course of roentgen-therapy amounting
to 3,000 r. with slight improvement. Thereafter he con-
The best level 1 evidence currently available does not support tinued to have right-sided seizures until July, 1943,
an improvement in outcome from the use of SRS for primary when he rapidly became worse and died … .
HGG. Despite the modest benefits achieved by escalating Autopsy. A broncho-pneumonia was found to be
EBRT dose to match normal tissue tolerance, it is unlikely the immediate cause of death. The brain weighed
that variations in fractionation, total dose, or treatment deliv- 1,420 g … .There was a hemorrhagic tumor nodule,
ery will lead to a significant change in the outcome for measuring 3 × 2 cm., in the left 3rd frontal convolution
patients with HGG. For a select group of patients with recur- … .The tumor was a glioblastoma multiforme. There
rent disease, SRS or F-SRT is a safe, reasonable but palliative were some spongioblastic portions, but there were, in
treatment. For recurrent HGG other treatment modalities addition, extensive necrosis, pseudo-pallisading
have been tested in phase III trials and only prospective inves- around focal necrotic zones, thrombosis of vessels and
tigation can define the exact role of stereotactic irradiation. endothelial proliferation, perivascular lymphocytes,
numerous mitotic figures, and pleomorphism.”
Case Study 17.1

“Case 1. S.S. Autopsy #9588.
In 1929, at the age of 37, this right-handed white
male had a convulsive seizure lasting a few minutes … . References
In 1931, following an attack, he noticed loss of sen-
sation on the right side of the body and a right 1. Surawicz TS, McCarthy BJ, Kupelian V, Jukich PJ, Bruner JM,
Davis FG. Descriptive epidemiology of primary brain and CNS
hemiparesis. tumors: results from the Central Brain Tumor Registry of the
In 1933, a partial motor aphasia developed. Dr. Paul United States, 1990–1994. Neuro Oncol. 1999;1(1):14–25.
C Bucy operated on him in October, 1933. A firm area 2. Melendon R, EC H, editors. Long term survival after treatment of
measuring 2 cm. in diameter was found on the left side glioblstoma multiforme: is it overstated. 44th Annual Meeting of
the American Society for Therapeutic Radiology and Oncology
in the gyrus just anterior to the precentral gyrus. An (ASTRO); 2002.
attempt at complete removal was not made because of 3. Scott JN, Rewcastle NB, Brasher PM, Fulton D, MacKinnon JA,
the location. A small piece was taken for study. The tis- Hamilton M, et al. Which glioblastoma multiforme patient will
sue was diagnosed by Dr. Percival Bailey as a malignant become a long-term survivor? A population-based study. Ann
Neurol. 1999;46(2):183–8.
4. Curran Jr WJ, Scott CB, Horton J, Nelson JS, Weinstein AS, 24. Loeffler JS, Alexander 3rd E, Hochberg FH, Wen PY, Morris JH,
Fischbach AJ, et al. Recursive partitioning analysis of prognostic Schoene WC, et al. Clinical patterns of failure following stereo-
factors in three Radiation Therapy Oncology Group malignant tactic interstitial irradiation for malignant gliomas. Int J Radiat
glioma trials. J Natl Cancer Inst. 1993;85(9):704–10. Oncol Biol Phys. 1990;19(6):1455–62.
5. Rao RD, James CD. Altered molecular pathways in gliomas: an 25. Gutin PH, Prados MD, Phillips TL, Wara WM, Larson DA, Leibel
overview of clinically relevant issues. Semin Oncol. 2004;31(5): SA, et al. External irradiation followed by an interstitial high
595–604. activity iodine-125 implant “boost” in the initial treatment of
6. Kitange GJ, Templeton KL, Jenkins RB. Recent advances in the malignant gliomas: NCOG study 6G-82–2. Int J Radiat Oncol
molecular genetics of primary gliomas. Curr Opin Oncol. Biol Phys. 1991;21(3):601–6.
2003;15(3):197–203. 26. Walker MD, Strike TA, Sheline GE. An analysis of dose-effect
7. Lai A, Kharbanda S, Pope WB, Tran A, Solis OE, Peale F, et al. relationship in the radiotherapy of malignant gliomas. Int J Radiat
Evidence for sequenced molecular evolution of IDH1 mutant glio- Oncol Biol Phys. 1979;5(10):1725–31.
blastoma from a distinct cell of origin. J Clin Oncol. 27. Bleehen NM, Stenning SP. A Medical Research Council trial of
2011;29(34):4482–90. Epub 2011/10/26. two radiotherapy doses in the treatment of grades 3 and 4 astrocy-
8. Wallner KE, Galicich JH, Krol G, Arbit E, Malkin MG. Patterns toma. The Medical Research Council Brain Tumour Working
of failure following treatment for glioblastoma multiforme and Party. Br J Cancer. 1991;64(4):769–74.
anaplastic astrocytoma. Int J Radiat Oncol Biol Phys. 1989;16(6): 28. Gutin PH, Leibel SA, Wara WM, Choucair A, Levin VA, Philips
1405–9. TL, et al. Recurrent malignant gliomas: survival following inter-
9. Hochberg FH, Pruitt A. Assumptions in the radiotherapy of glio- stitial brachytherapy with high-activity iodine-125 sources.
blastoma. Neurology. 1980;30(9):907–11. J Neurosurg. 1987;67(6):864–73.
10. Cerame MA, Guthikonda M, Kohli CM. Extraneural metastases in 29. Florell RC, Macdonald DR, Irish WD, Bernstein M, Leibel SA,
gliosarcoma: a case report and review of the literature. Gutin PH, et al. Selection bias, survival, and brachytherapy for
Neurosurgery. 1985;17(3):413–8. glioma. J Neurosurg. 1992;76(2):179–83.
11. Erlich SS, Davis RL. Spinal subarachnoid metastasis from pri- 30. Shapiro WR, Green SB, Burger PC. A randomized trial of intersti-
mary intracranial glioblastoma multiforme. Cancer. tial radiotherapy (IRT) boost for the treatment of newly disgnosed
1978;42(6):2854–64. malignant glioma (glioblastoma multiforme, anaplastic astrocy-
12. Bailey P, Cushing H. Classification of the tumors of the glioma toma, anaplastic oligodendroglioma, malignant mixed glioma):
group on a histogenetic basis with a correlation study of progno- BTCG Study 8701. Neurology. 1994;44 Suppl 2:A263.
sis. Philadelphia: Lippincott; 1926. 175 p. 31. Laperriere NJ, Leung PM, McKenzie S, Milosevic M, Wong S,
13. Elvidge A, Wilder P, Cone W. The gliomas of the central nervous Glen J, et al. Randomized study of brachytherapy in the initial
system. Res Publ Ass Nerv Ment Dis. 1937;16:107–81. management of patients with malignant astrocytoma. Int J Radiat
14. Netsky MG, August B, Fowler W. The longevity of patients with Oncol Biol Phys. 1998;41(5):1005–11.
glioblastoma multiforme. J Neurosurg. 1950;7(3):261–9. 32. Colombo F, Benedetti A, Pozza F, Avanzo RC, Marchetti C,
15. Bush E, Christensen E. Three Types of Glioblastoma. J Neurosurg. Chierego G, et al. External stereotactic irradiation by linear accel-
1947;4(3):200–20. erator. Neurosurgery. 1985;16(2):154–60.
16. Kristiansen K, Hagen S, Kollevold T, Torvik A, Holme I, 33. Smith KA, Ashby LS, Gonzalez LF, Brachman DG, Thomas T, Coons
Nesbakken R, et al. Combined modality therapy of operated astro- SW, et al. Prospective trial of gross-total resection with Gliadel wafers
cytomas grade III and IV. Confirmation of the value of postopera- followed by early postoperative Gamma Knife radiosurgery and con-
tive irradiation and lack of potentiation of bleomycin on survival formal fractionated radiotherapy as the initial treatment for patients
time: a prospective multicenter trial of the Scandinavian with radiographically suspected, newly diagnosed glioblastoma multi-
Glioblastoma Study Group. Cancer. 1981;47(4):649–52. forme. J Neurosurg. 2008;109(Suppl):106–17. Epub 2009/01/10.
17. Walker MD, Alexander Jr E, Hunt WE, MacCarty CS, Mahaley Jr 34. Shaw E, Scott C, Souhami L, Dinapoli R, Bahary JP, Kline R,
MS, Mealey Jr J, et al. Evaluation of BCNU and/or radiotherapy et al. Radiosurgery for the treatment of previously irradiated
in the treatment of anaplastic gliomas. A cooperative clinical trial. recurrent primary brain tumors and brain metastases: initial report
J Neurosurg. 1978;49(3):333–43. of radiation therapy oncology group protocol (90–05). Int J Radiat
18. Fine HA, Dear KB, Loeffler JS, Black PM, Canellos GP. Meta- Oncol Biol Phys. 1996;34(3):647–54.
analysis of radiation therapy with and without adjuvant chemother- 35. Willner J, Flentje M, Bratengeier K. CT simulation in stereotactic
apy for malignant gliomas in adults. Cancer. 1993;71(8):2585–97. brain radiotherapy–analysis of isocenter reproducibility with
19. Glioma Meta-Analysis Trialists (GMT) Group. Chemotherapy for mask fixation. Radiother Oncol. 1997;45(1):83–8.
high-grade glioma. Cochrane Database Syst Rev. 2002;4, 36. Alheit H, Dornfeld S, Dawel M, Alheit M, Henzel B, Steckler K,
CD003913. et al. Patient position reproducibility in fractionated stereotacti-
20. Stupp R, Mason WP, van den Bent MJ, Weller M, Fisher B, cally guided conformal radiotherapy using the BrainLab mask
Taphoorn MJ, et al. Radiotherapy plus concomitant and adjuvant system. Strahlenther Onkol. 2001;177(5):264–8.
temozolomide for glioblastoma. N Engl J Med. 2005;352(10): 37. Glass J, Silverman CL, Axelrod R, Corn BW, Andrews
987–96. DW. Fractionated stereotactic radiotherapy with cis-platinum radio-
21. Zietman AL, Suit HD, Tomkinson KN, Thames HD, Sedlacek sensitization in the treatment of recurrent, progressive, or persistent
RS. The response of two human tumor xenografts to fractionated malignant astrocytoma. Am J Clin Oncol. 1997;20(3):226–9.
irradiation. The derivation of alpha/beta ratios from growth delay, 38. Shepherd SF, Laing RW, Cosgrove VP, Warrington AP, Hines F,
tumor control, and in vitro cell survival assays. Int J Radiat Oncol Ashley SE, et al. Hypofractionated stereotactic radiotherapy in the
Biol Phys. 1990;18(3):569–75. management of recurrent glioma. Int J Radiat Oncol Biol Phys.
22. Marin LA, Smith CE, Langston MY, Quashie D, Dillehay 1997;37(2):393–8.
LE. Response of glioblastoma cell lines to low dose rate irradia- 39. Hudes RS, Corn BW, Werner-Wasik M, Andrews D, Rosenstock
tion. Int J Radiat Oncol Biol Phys. 1991;21(2):397–402. J, Thoron L, et al. A phase I dose escalation study of hypofraction-
23. Raza SM, Lang FF, Aggarwal BB, Fuller GN, Wildrick DM, ated stereotactic radiotherapy as salvage therapy for persistent or
Sawaya R. Necrosis and glioblastoma: a friend or a foe? A review recurrent malignant glioma. Int J Radiat Oncol Biol Phys.
and a hypothesis. Neurosurgery. 2002;51(1):2–12; discussion12–3. 1999;43(2):293–8.
40. Lederman G, Wronski M, Arbit E, Odaimi M, Wertheim S, 55. Souhami L, Seiferheld W, Brachman D, Podgorsak EB,
Lombardi E, et al. Treatment of recurrent glioblastoma multi- Werner-Wasik M, Lustig R, et al. Randomized comparison of ste-
forme using fractionated stereotactic radiosurgery and concurrent reotactic radiosurgery followed by conventional radiotherapy with
paclitaxel. Am J Clin Oncol. 2000;23(2):155–9. carmustine to conventional radiotherapy with carmustine for
41. Regine WF, Patchell RA, Strottmann JM, Meigooni A, Sanders patients with glioblastoma multiforme: report of Radiation
M, Young AB. Preliminary report of a phase I study of combined Therapy Oncology Group 93–05 protocol. Int J Radiat Oncol Biol
fractionated stereotactic radiosurgery and conventional external Phys. 2004;60(3):853–60.
beam radiation therapy for unfavorable gliomas. Int J Radiat 56. Selch MT, Ciacci JD, De Salles AA, Goetsch S, Brekhus
Oncol Biol Phys. 2000;48(2):421–6. SD. Radiosurgery for primary malignant brain tumors. In: De
42. Regine WF, Patchell RA, Strottmann JM, Meigooni A, Sanders M, Salles AAF, Goetsch SJ, editors. Stereotactic surgery and radiosur-
Young B. Combined stereotactic split-course fractionated gamma gery. Madison, Wisconsin: Medical Physics Pub; 1993. p. 335–52.
knife radiosurgery and conventional radiation therapy for unfavor- 57. Masciopinto JE, Levin AB, Mehta MP, Rhode BS. Stereotactic
able gliomas: a phase I study. J Neurosurg. 2000;93 Suppl 3:37–41. radiosurgery for glioblastoma: a final report of 31 patients.
43. Baumert BG, Lutterbach J, Bernays R, Davis JB, Heppner J Neurosurg. 1995;82(4):530–5.
FL. Fractionated stereotactic radiotherapy boost after post- 58. Gannett D, Stea B, Lulu B, Adair T, Verdi C, Hamilton
operative radiotherapy in patients with high-grade gliomas. A. Stereotactic radiosurgery as an adjunct to surgery and external
Radiother Oncol. 2003;67(2):183–90. beam radiotherapy in the treatment of patients with malignant
44. Cho KH, Hall WA, Lo SS, Dusenbery KE. Stereotactic radiosur- gliomas. Int J Radiat Oncol Biol Phys. 1995;33(2):461–8. Epub
gery versus fractionated stereotactic radiotherapy boost for 1995/09/30.
patients with glioblastoma multiforme. Technol Cancer Res Treat. 59. Buatti JM, Friedman WA, Bova FJ, Mendenhall WM. Linac radio-
2004;3(1):41–9. surgery for high-grade gliomas: the University of Florida experi-
45. Vordermark D, Kolbl O, Ruprecht K, Vince GH, Bratengeier K, ence. Int J Radiat Oncol Biol Phys. 1995;32(1):205–10.
Flentje M. Hypofractionated stereotactic re-irradiation: treatment 60. Shenouda G, Souhami L, Podgorsak EB, Bahary JP, Villemure JG,
option in recurrent malignant glioma. BMC Cancer. 2005;5:55. Caron JL, et al. Radiosurgery and accelerated radiotherapy for
Epub 2005/06/01. patients with glioblastoma. Can J Neurol Sci. 1997;24(2):110–5.
46. Cardinale R, Won M, Choucair A, Gillin M, Chakravarti A, 61. Nwokedi EC, DiBiase SJ, Jabbour S, Herman J, Amin P, Chin
Schultz C, et al. A phase II trial of accelerated radiotherapy using LS. Gamma knife stereotactic radiosurgery for patients with glio-
weekly stereotactic conformal boost for supratentorial glioblas- blastoma multiforme. Neurosurgery. 2002;50(1):41–6; discussion
toma multiforme: RTOG 0023. Int J Radiat Oncol Biol Phys. 6–7; Epub 2002/02/15.
2006;65(5):1422–8. Epub 2006/06/06. 62. Hsieh PC, Chandler JP, Bhangoo S, Panagiotopoulos K,
47. Schwer AL, Damek DM, Kavanagh BD, Gaspar LE, Lillehei K, Kalapurakal JA, Marymont MH, et al. Adjuvant gamma knife ste-
Stuhr K, et al. A phase I dose-escalation study of fractionated stereotactic radiosurgery at the time of tumor progression potentially
reotactic radiosurgery in combination with gefitinib in patients improves survival for patients with glioblastoma multiforme.
with recurrent malignant gliomas. Int J Radiat Oncol Biol Phys. Neurosurgery. 2005;57(4):684–92; discussion 684–92. Epub
2008;70(4):993–1001. Epub 2007/10/31. 2005/10/22.
48. Fokas E, Wacker U, Gross MW, Henzel M, Encheva E, Engenhart- 63. Biswas T, Okunieff P, Schell MC, Smudzin T, Pilcher WH, Bakos
Cabillic R. Hypofractionated stereotactic reirradiation of recurrent RS, et al. Stereotactic radiosurgery for glioblastoma: retrospective
glioblastomas : a beneficial treatment option after high-dose analysis. Radiat Oncol. 2009;4:11. Epub 2009/03/19.
radiotherapy? Strahlenther Onkol. 2009;185(4):235–40. Epub 64. Villavicencio AT, Burneikiene S, Romanelli P, Fariselli L,
2009/04/17. McNeely L, Lipani JD, et al. Survival following stereotactic radio-
49. Gutin PH, Iwamoto FM, Beal K, Mohile NA, Karimi S, Hou BL, surgery for newly diagnosed and recurrent glioblastoma multi-
et al. Safety and efficacy of bevacizumab with hypofractionated forme: a multicenter experience. Neurosurg Rev. 2009;32(4):
stereotactic irradiation for recurrent malignant gliomas. Int J 417–24. Epub 2009/07/28.
Radiat Oncol Biol Phys. 2009;75(1):156–63. Epub 2009/01/27. 65. Pouratian N, Crowley RW, Sherman JH, Jagannathan J, Sheehan
50. Fields EC, Damek D, Gaspar LE, Liu AK, Kavanagh BD, Waziri JP. Gamma Knife radiosurgery after radiation therapy as an
A, et al. Phase I dose escalation trial of vandetanib with fraction- adjunctive treatment for glioblastoma. J Neurooncol.
ated radiosurgery in patients with recurrent malignant gliomas. Int 2009;94(3):409–18. Epub 2009/03/31.
J Radiat Oncol Biol Phys. 2012;82(1):51–7. Epub 2010/11/03. 66. Einstein DB, Wessels B, Bangert B, Fu P, Nelson AD, Cohen M,
51. Larson DA, Prados M, Lamborn KR, Smith V, Sneed PK, Chang et al. Phase II Trial of Radiosurgery to Magnetic Resonance
S, et al. Phase II study of high central dose Gamma Knife radio- Spectroscopy-Defined High-Risk Tumor Volumes in Patients with
surgery and marimastat in patients with recurrent malignant gli- Glioblastoma Multiforme. Int J Radiat Oncol Biol Phys.
oma. Int J Radiat Oncol Biol Phys. 2002;54(5):1397–404. 2012;84(3):668–74. Epub 2012/03/27.
52. Kondziolka D, Flickinger JC, Bissonette DJ, Bozik M, Lunsford 67. Irish WD, Macdonald DR, Cairncross JG. Measuring bias in
LD. Survival benefit of stereotactic radiosurgery for patients with uncontrolled brain tumor trials–to randomize or not to randomize?
malignant glial neoplasms. Neurosurgery. 1997;41(4):776–83; Can J Neurol Sci. 1997;24(4):307–12.
discussion 83–5. 68. Sarkaria JN, Mehta MP, Loeffler JS, Buatti JM, Chappell RJ,
53. Shrieve DC, Alexander 3rd E, Black PM, Wen PY, Fine HA, Kooy Levin AB, et al. Radiosurgery in the initial management of malig-
HM, et al. Treatment of patients with primary glioblastoma multi- nant gliomas: survival comparison with the RTOG recursive parti-
forme with standard postoperative radiotherapy and radiosurgical tioning analysis. Radiation Therapy Oncology Group. Int J Radiat
boost: prognostic factors and long-term outcome. J Neurosurg. Oncol Biol Phys. 1995;32(4):931–41.
1999;90(1):72–7. Epub 1999/07/21. 69. Werner-Wasik M, Scott CB, Nelson DF, Gaspar LE, Murray KJ,
54. Shaw E, Scott C, Souhami L, Dinapoli R, Kline R, Loeffler J, et al. Fischbach JA, et al. Final report of a phase I/II trial of hyperfrac-
Single dose radiosurgical treatment of recurrent previously irradi- tionated and accelerated hyperfractionated radiation therapy with
ated primary brain tumors and brain metastases: final report of carmustine for adults with supratentorial malignant gliomas.
RTOG protocol 90–05. Int J Radiat Oncol Biol Phys. Radiation Therapy Oncology Group Study 83–02. Cancer. 1996;
2000;47(2):291–8. 77(8):1535–43.
70. Nelson DF, Diener-West M, Weinstein AS, Schoenfeld D, Nelson 85. Shrieve DC, Alexander 3rd E, Wen PY, Fine HA, Kooy HM,
JS, Sause WT, et al. A randomized comparison of misonidazole Black PM, et al. Comparison of stereotactic radiosurgery and
sensitized radiotherapy plus BCNU and radiotherapy plus BCNU brachytherapy in the treatment of recurrent glioblastoma multi-
for treatment of malignant glioma after surgery: final report of an forme. Neurosurgery. 1995;36(2):275–82; discussion 82–4.
RTOG study. Int J Radiat Oncol Biol Phys. 1986;12(10): 86. Mahajan A, McCutcheon IE, Suki D, Chang EL, Hassenbusch SJ,
1793–800. Weinberg JS, et al. Case–control study of stereotactic radiosurgery
71. Nelson DF, Diener-West M, Horton J, Chang CH, Schoenfeld D, for recurrent glioblastoma multiforme. J Neurosurg. 2005;103(2):
Nelson JS. Combined modality approach to treatment of malig- 210–7. Epub 2005/09/24.
nant gliomas–re-evaluation of RTOG 7401/ECOG 1374 with 87. Patel M, Siddiqui F, Jin JY, Mikkelsen T, Rosenblum M, Movsas
long-term follow-up: a joint study of the Radiation Therapy B, et al. Salvage reirradiation for recurrent glioblastoma with
Oncology Group and the Eastern Cooperative Oncology Group. radiosurgery: radiographic response and improved survival.
NCI Monogr. 1988;6:279–84. J Neurooncol. 2009;92(2):185–91. Epub 2008/12/11.
72. Spitzer WO, Dobson AJ, Hall J, Chesterman E, Levi J, Shepherd 88. Cuneo KC, Vredenburgh JJ, Sampson JH, Reardon DA, Desjardins
R, et al. Measuring the quality of life of cancer patients: a concise A, Peters KB, et al. Safety and efficacy of stereotactic radiosur-
QL-index for use by physicians. J Chronic Dis. gery and adjuvant bevacizumab in patients with recurrent malig-
1981;34(12):585–97. nant gliomas. Int J Radiat Oncol Biol Phys. 2012;82(5):2018–24.
73. Loeffler JS, Alexander 3rd E, Shea WM, Wen PY, Fine HA, Kooy Epub 2011/04/15.
HM, et al. Radiosurgery as part of the initial management of 89. Cho KH, Hall WA, Gerbi BJ, Higgins PD, McGuire WA, Clark
patients with malignant gliomas. J Clin Oncol. 1992;10(9): HB. Single dose versus fractionated stereotactic radiotherapy for
1379–85. recurrent high-grade gliomas. Int J Radiat Oncol Biol Phys.
74. Selker RG, Shapiro WR, Burger P, Blackwood MS, Arena VC, 1999;45(5):1133–41.
Gilder JC, et al. The Brain Tumor Cooperative Group NIH Trial 90. Leksell L. The stereotaxic method and radiosurgery of the brain.
87–01: a randomized comparison of surgery, external radiother- Acta Chirurgica Scandinavia. 1951;102(4):316–9.
apy, and carmustine versus surgery, interstitial radiotherapy boost, 91. Souhami L, Olivier A, Podgorsak EB, Villemure JG, Pla M,
external radiation therapy, and carmustine. Neurosurgery. Sadikot AF. Fractionated stereotactic radiation therapy for intra-
2002;51(2):343–55; discussion 55–7. cranial tumors. Cancer. 1991;68(10):2101–8.
75. Tsao MN, Mehta MP, Whelan TJ, Morris DE, Hayman JA, 92. Clark BG, Podgorsak EB, Souhami L, Olivier A, Sixel KE, Caron
Flickinger JC, et al. The American Society for Therapeutic JL. A halo-ring technique for fractionated stereotactic radiother-
Radiology and Oncology (ASTRO) evidence-based review of the apy. Br J Radiol. 1993;66(786):522–7.
role of radiosurgery for malignant glioma. Int J Radiat Oncol Biol 93. Baumert BG, Brada M, Bernier J, Kortmann RD, Dehing-Oberije
Phys. 2005;63(1):47–55. Epub 2005/08/23. C, Collette L, et al. EORTC 22972–26991/MRC BR10 trial: frac-
76. Nieder C, Grosu AL, Molls M. A comparison of treatment results tionated stereotactic boost following conventional radiotherapy of
for recurrent malignant gliomas. Cancer Treat Rev. 2000;26(6): high grade gliomas. Clinical and quality-assurance results of the
397–409. stereotactic boost arm. Radiother Oncol. 2008;88(2):163–72.
77. Brem H, Piantadosi S, Burger PC, Walker M, Selker R, Vick NA, Epub 2008/05/06.
et al. Placebo-controlled trial of safety and efficacy of intraopera- 94. Cardinale RM, Broaddus WC, Benedict SH, Schmidt-Ullrich RK,
tive controlled delivery by biodegradable polymers of chemother- Kavanagh BD. Accelerated radiotherapy using weekly stereotactic
apy for recurrent gliomas. The Polymer-brain Tumor Treatment comformal boosts for glioblastoma multiforme. Int J Radiat Oncol
Group. Lancet. 1995;345(8956):1008–12. Biol Phys. 2000;48(3 Suppl 1):257.
78. Osoba D, Brada M, Yung WK, Prados M. Health-related quality of life 95. Clark BG, Souhami L, Pla C, Al-Amro AS, Bahary JP, Villemure
in patients treated with temozolomide versus procarbazine for recurrent JG, et al. The integral biologically effective dose to predict brain
glioblastoma multiforme. J Clin Oncol. 2000;18(7):1481–91. stem toxicity of hypofractionated stereotactic radiotherapy. Int J
79. Yung WK, Albright RE, Olson J, Fredericks R, Fink K, Prados Radiat Oncol Biol Phys. 1998;40(3):667–75.
MD, et al. A phase II study of temozolomide vs. procarbazine in 96. Levin VA, Bidaut L, Hou P, Kumar AJ, Wefel JS, Bekele BN, et al.
patients with glioblastoma multiforme at first relapse. Br J Cancer. Randomized double-blind placebo-controlled trial of bevacizumab
2000;83(5):588–93. therapy for radiation necrosis of the central nervous system. Int J
80. Batchelor T, Mulholland P, Neyns B, Nabors LB, Campone M, Radiat Oncol Biol Phys. 2011;79(5):1487–95. Epub 2010/04/20.
Wick A, et al. A phase III randomized study comparing the effi- 97. Chan AA, Lau A, Pirzkall A, Chang SM, Verhey LJ, Larson D,
cacy of Cediranib as monotherapy, and in combination with et al. Proton magnetic resonance spectroscopy imaging in the
lomustine, with lomustine alone in recurrent glioblastoma evaluation of patients undergoing gamma knife surgery for Grade
patients. J Clin Oncol. 2013;31(26):3212–8. IV glioma. J Neurosurg. 2004;101(3):467–75.
81. Friedman HS, Prados MD, Wen PY, Mikkelsen T, Schiff D, Abrey 98. Nieder C, Weidenmann S, Stark S, Grosu AL. Phase II trial of
LE, et al. Bevacizumab alone and in combination with irinotecan functional imaging-optimized stereotactic fractionated radiother-
in recurrent glioblastoma. J Clin Oncol. 2009;27(28):4733–40. apy plus temozolomide for recurrent high-grade glioma. Journal
Epub 2009/09/02. of Clinical Oncology, 2004 ASCO Annual Meeting Proceedings
82. Kunwar S, Chang S, Westphal M, Vogelbaum M, Sampson J, (Post-Meeting Edition). 2004;22(14S):1552.
Barnett G, et al. Phase III randomized trial of CED of IL13- 99. Lederman G, Arbit E, Odaimi M, Lombardi E, Wrzolek M,
PE38QQR vs Gliadel wafers for recurrent glioblastoma. Neuro Wronski M. Fractionated stereotactic radiosurgery and concurrent
Oncol. 2010;12(8):871–81. Epub 2010/06/01. taxol in recurrent glioblastoma multiforme: a preliminary report.
83. Stupp R, Wong ET, Kanner AA, Steinberg D, Engelhard H, Int J Radiat Oncol Biol Phys. 1998;40(3):661–6.
Heidecke V, et al. NovoTTF-100A versus physician’s choice che- 100. Kohshi K, Yamamoto H, Nakahara A, Katoh T, Takagi
motherapy in recurrent glioblastoma: A randomised phase III trial M. Fractionated stereotactic radiotherapy using gamma unit
of a novel treatment modality. Eur J Cancer. 2012;48(14):2192– after hyperbaric oxygenation on recurrent high-grade gliomas.
202. Epub 2012/05/23. J Neurooncol. 2007;82(3):297–303. Epub 2006/11/23.
84. Chamberlain MC, Barba D, Kormanik P, Shea WM. Stereotactic 101. Grecula JC, Grever M, Gupta N, Goofman J, Barth R,
radiosurgery for recurrent gliomas. Cancer. 1994;74(4):1342–7. Christoforidis G. Phase I trial of motexafin gadolinium (Gd-Tex)
during G-knife boost for glioblastoma multiforme. Clin Cancer by bevacizumab for recurrent glioblastoma multiforme: a
Res. 2001;7(S). case–control study. J Neurooncol. 2012;107(2):323–33. Epub
102. Cuneo KC, Cabrera AR, Sampson JH, Allen K, Duffy E, 2011/11/08.
Vredenburgh JJ, et al. Bevacizumab and Radiosurgery for 104. Tatter SB, Shaw EG, Rosenblum ML, Karvelis KC, Kleinberg L,
Recurrent Malignant Glioma: Toxicity Results from a Prospective Weingart J, et al. An inflatable balloon catheter and liquid 125I
Pilot Study. Int J Radiat Oncol Biol Phys. 2011;81(2):S185. radiation source (GliaSite Radiation Therapy System) for treat-
103. Park KJ, Kano H, Iyer A, Liu X, Niranjan A, Flickinger JC, ment of recurrent malignant glioma: multicenter safety and feasi-
et al. Salvage gamma knife stereotactic radiosurgery followed bility trial. J Neurosurg. 2003;99(2):297–303.
Management of Glial Tumors:
Viewpoint—Surgery and Intra- 18
cavitary Radiopharmaceutical
Therapy
Kaisorn L. Chaichana, Linda Chen,

Salvador Manrique-Guzman, Lawrence Kleinberg,
and Alfredo Quinones-Hinojosa
[2–6, 10, 11, 18–20]. High-grade gliomas are characterized

Introduction by their invasive and infiltrative nature, making curative
resection unlikely [20]. In fact, survival has changed little in
Gliomas are primary tumors of the central nervous system recent years despite advances in medical and surgical thera-
that can be subdivided into low- and high-grade gliomas. pies [21]. Walter Dandy in the 1920s performed hemispher-
Patients with low-grade gliomas are considered to have a ectomies for these lesions, and they would still recur on the
better prognosis than patients harboring high-grade tumors contralateral side [22]. Survival for these patients was still
[1–13]. The median survival for patients with low-grade gli- less than 2 years [22]. As a result, the efficacy of surgery
omas is between 5 and 10 years as compared to 1–3 years for alone for these lesions has been questioned and raises the
patients with high-grade tumors [1–13]. Regardless of tumor need to implement further treatment modalities to comple-
grade, the overwhelming majority of patients will eventually ment surgery.
die from progressive and/or recurrent disease [1–13]. Despite In recent years, there have been an increased number of
the poor prognosis for patients with gliomas, surgery has studies supporting a role for extensive surgery in patients
been a mainstay of treatment [1–13]. An emerging adjunct to with high-grade gliomas.
surgical therapy is stereotactic brachytherapy (SBT) [14, 15]. Lacroix et al. in 2001 analyzed 416 consecutive patients
This chapter will discuss the role that surgery and brachy- with GBM [23]. They found that resections greater than
therapy have for patients with gliomas. 98 % were needed to significantly improve survival, where
survival increased from 8.8 to 13 months [23]. This study,
however, did not include adjuvant therapy into their survival
Role of Surgery for Patients analysis [23]. Sanai et al. more recently studied 500 patients
with High-Grade Gliomas with GB in 2011 and found that 78 % resection was needed
to confer a survival advantage [24]. Keles et al. in 2006 eval-
High-grade gliomas include WHO grade III (anaplastic uated 102 patients with anaplastic astrocytomas [25]. They
astrocytomas) and IV (glioblastoma—GB). Patients with found that the amount of residual tumor correlated with time
high-grade gliomas are considered to have poor prognoses to progression and overall survival, where the most signifi-
[3–6, 12, 13, 16–19]. Patients with anaplastic tumors have a cant threshold was 10 cm3 [25]. This study, however, did not
median survival of approximately 3 years, while patients include patients with GB and did not account for extent of
with GB have a median survival of approximately 14 months resection, just residual disease [25]. Laws and colleagues
evaluated outcomes for 788 patients with malignant astrocy-
K.L. Chaichana, M.D. (*) • S. Manrique-Guzman, M.D. tomas from several institutions [26]. They found that patients
Department of Neurosurgery, Johns Hopkins University School who underwent surgical resection had significantly longer
of Medicine, Baltimore, MD, USA survival times than patients who underwent biopsy [26]. This
e-mail: Kaisorn@jhmi.edu remained significant even after eliminating patients who
L. Chen, M.D. • L. Kleinberg, M.D. were considered poor surgical candidates [26]. This study,
Department of Radiation Oncology, Johns Hopkins University however, did not evaluate the role of gross total, near total, or
subtotal resection [26].
A. Quinones-Hinojosa, M.D. (*) We recently evaluated the role of extent of resection on
Department of Neurosurgery and Oncology, Johns Hopkins
University School of Medicine, Baltimore, MD, USA survival for 949 patients with malignant astrocytomas [12].
e-mail: Aquinon2@jhmi.edu We found that there was a survival benefit for patients who
270 K.L. Chaichana et al.
underwent both gross total and near total resection [12]. This that gross total resection of gliomas improves overall survival
survival benefit was independent of age, degree of disability, in comparison to subtotal resection and biopsy [2–6, 10, 11,
or subsequent treatment modalities for both WHO grade III 30–32]. In addition to the possible benefit of prolonging sur-
and IV gliomas [12]. The median survival for patients with vival with safe, maximal surgical debulking, aggressive
newly diagnosed GB who underwent gross total, near total, resection provides a more representative sample for patho-
and subtotal resection was 13, 11, and 8 months, respectively logical diagnosis and may also provide more improvement in
[12]. Likewise, the median survival for patients with recur- mass effect with subsequent symptomatic relief [26].
rent GB who underwent gross total, near total, and subtotal Maximal resection should therefore be pursued when safe to
resection was 11, 9, and 5 months, respectively [12]. do so for high-grade lesions (Table 18.1) [19].
Similarly, for patients with WHO grade III astrocytomas,
patients who underwent gross total resection had improved
survival as compared to patients with subtotal resection, even Role of Surgery for Patients
after controlling for age, functional status, and revision with Low-Grade Gliomas
resection [12]. The median survival after primary resection
of WHO III astrocytomas for gross total, near total, and sub- Patients with low-grade gliomas often have a better progno-
total resection was 58, 46, and 34 months, respectively [12]. sis than patients harboring higher-grade tumors [3–10, 12,
These studies therefore suggest that achieving an increased 18–21]. In fact, the median survival for patients with low-
extent of resection may be associated with prolonging grade gliomas is typically between 5 and 10 years [3–10, 12,
survival for patients with malignant astrocytomas, but 18–21]. Despite this more favorable prognosis, 50–75 % of
randomized control studies do not exist. patients with low-grade gliomas eventually die from their
Despite this growing body of evidence supporting disease because of either tumor progression or degeneration
increased extent of resection for patients with high-grade to a higher malignant grade [1, 7, 8]. The optimal treatment
gliomas, not all studies support a positive correlation between for patients who harbor low-grade gliomas, however, remains
extent of resection and survival [27]. Levin et al. studied 90 unclear because many patients are followed radiographically
patients with WHO grade III gliomas enrolled into a phase II and intervention is only pursued when clinical or radio-
study for radiation and carboplatin [27]. They found that graphic progression becomes evident [35]. Moreover, the
extent of resection was not significantly associated with sur- role of surgery in delaying recurrence and/or malignant
vival for this cohort [27]. Totrosa et al. also found no associa- degeneration as well as prolonging survival remains contro-
tion between survival and extent of resection for a similar versial [7, 35].
group of patients with anaplastic tumors [28]. Additionally, This lack of understanding of the effectiveness of surgery
Pope et al. evaluated 153 patients with high-grade gliomas for patients with low-grade gliomas is because there are no
(43 with WHO grade III and 110 with WHO grade IV glio- randomized control trials or matched pair analyses. Keles
mas) [29]. They found no difference in survival between et al. performed a literature review of the role of surgery for
patients who underwent surgery versus biopsy [29]. patients with low-grade gliomas and found 30 studies evalu-
There have been five systematic reviews of the literature ating the role of surgery for patients with low-grade gliomas
to date evaluating the role of resection and survival [30–34]. [1]. In this review, all but five studies were excluded because
Sanai and colleagues identified 28 studies on high-grade they included patients with grade I gliomas, relied upon
gliomas with more than 75 patients [34]. They found of these intraoperative determination of extent of resection, and had
28 studies, 14 studies demonstrated an association between small sample sizes, among others [36–39]. These five studies
extent of resection and survival and ten studies found no sta- that met the inclusion criteria were also limited because
tistical association [34]. Grant et al. performed a Cochrane extent of resection was determined by intraoperative gross
review on biopsy versus resection for patients with high- visualization rather than neuro-imaging, and a distinction
grade gliomas [30]. They excluded all 2,100 identified stud- between gross total and near total resection was not made
ies because they lacked statistical power, patient information, [36–39]. This review and others have found data supporting
and/or outcome data [30]. Hess [31] and Nazzaro [32] also a role of surgery in delaying recurrence and prolonging sur-
independently performed similar reviews and also concluded vival controversial [1, 40].
that there was a lack of high-quality studies to conclude if Despite this, there are an increasing number of studies
surgical resection provided a benefit for patients with high- demonstrating a benefit for increasing extent of resection for
grade gliomas [31]. patients with low-grade gliomas [7]. Wisoff et al. evaluated
Although the survival benefit of extensive surgical resec- 518 pediatric patients with low-grade glioma, and found that
tion is not demonstrated in all retrospective studies, the pre- extent of resection was significantly associated with
vailing consensus in numerous studies have demonstrated improved progression-free survival [41]. The 5-year survival
18 Management of Glial Tumors: Viewpoint—Surgery and Intra-cavitary Radiopharmaceutical Therapy 271
Table 18.1 Surgical management of glioma: summary and level of existing evidence supporting management
Tumor characteristics Low-grade glioma–surgical management High-grade glioma–surgical management
and presentation and considerations and considerations
Large tumors, significant Early surgical debulking with extensive resection Extensive resection which minimizes neurologic
neurologic deficit, and/or while minimizing neurologic dysfunction dysfunction
medically intractable Decompression of tumor provides Gross total resection improves patient overall survival
seizures symptomatic relief and reduces tumor burden
Decreases risk for malignant transformation (Level III evidence)
(Level III evidence)
Use of image guided resection to achieve more
extensive resection (Level I evidence)
Decompression of tumor provides symptomatic relief
and reduces tumor burden
Small tumors without mass Consider watchful waiting and delayed Extensive resection which minimizes neurologic
effect found on MRI, intervention dysfunction
transient neurologic Monitor for radiologic progression Gross total resection improves patient overall survival
symptoms Monitor for clinical progression (Level III evidence)
of neurologic symptoms
Use of image-guided resection to achieve more
extensive resection (Level I evidence)
Decompression of tumor provides symptomatic relief
and reduces tumor burden
Tumors located in motor, Stereotactic biopsy only Extensive resection which minimizes neurologic
sensory, or language dysfunction
functional areas and/or Monitor for radiologic progression Gross total resection improves patient overall survival
deep seated tumors Monitor for clinical progression (Level III evidence)
of neurologic symptoms
Consider resection Intraoperative cortical mapping to identify functionally
critical areas (Level III evidence)
Intraoperative cortical mapping to identify Use of image-guided resection to achieve more
functionally critical areas (Level III evidence) extensive resection (Level I evidence)
Consider resection in conjunction
with stereotactic brachytherapy
(Level III evidence)
rate was 99 % with gross total resection, 95 % with 1.5 cm3 progression was 7.0, 4.0, and 3.5 years, respectively [11].
residual disease, 94 % with 1.5–2.9 cm3 residual, and 87 % The median time to malignant degeneration after gross total,
with >3 cm3 residual [41]. The 5-year progression-free sur- near total, and subtotal resection was 12.5, 5.8, and 7 years,
vival rate was 90 % with gross total resection and 45–65 % respectively [11]. In a separate study, we found that gross
with any volume (1.5–3 cm3) of residual disease [41]. More total resection was independently associated with delayed
recently, Smith et al. performed a retrospective volumetric malignant degeneration [7].
analysis of extent of hemispheric LGG resection in 216 There is a growing body of evidence that surgery is an
patients and found that patients with at least 90 % resection effective treatment for patients with low-grade gliomas,
had 5- and 8-year survival rates of 97 % and 91 %, respec- despite an absence of randomized control trials. For patients
tively, while patients with less than 90 % resection had 5- who present with large tumors, significant neurological defi-
and 8-year survival rates of 76 % and 60 %, respectively cits, and medically intractable seizures, immediate surgical
[42]. We recently evaluated 170 patients with low-grade gli- intervention is generally the preferred treatment option [7, 43].
omas and found that gross-total resection was associated Extensive resection without causing an iatrogenic deficit can
with increased overall survival as compared to patients who lead to improvements in overall survival, progression-free
underwent subtotal resection [11]. Five-year overall survival survival, and malignant transformation-free survival [7–9, 44].
after gross total, near total, and subtotal resection was 95 %, However, for incidentally discovered small tumors without
80 %, 70 %, respectively, and 10-year survival was 76 %, mass effect or symptoms, observation with surveillance
57 %, and 49 %, respectively [11]. Additionally, after gross imaging is a viable option given the slow growth and natural
total, near total, and subtotal resection, median time to tumor history of low-grade gliomas (Table 18.1) [45].
sparing surrounding tissue [15]. Additionally, SBT avoids the

Intra-Cavitary Radiopharmaceutical Therapy systemic side effects of chemotherapy and the potential neuro-
or Stereotactic Brachytherapy cognitive effects of external beam radiation therapy (EBRT)
[50]. However, while numerous retrospective studies have
The standard of care for patients with high-grade gliomas is demonstrated SBT to be safe and efficacious in gliomas [15,
extensive resection, followed by radiation and temozolo- 51–56], three small single institution prospective randomized
mide chemotherapy [46]. Regardless of treatment, tumor trials failed to demonstrate a survival advantage [57–59].
recurrence typically occurs close to the tumor margins [20]. Therefore, the efficacy of this treatment option requires further
At the tumor margin, there is an increased tumor cell density investigation as well as potential advances in imaging and
[20]. Beyond the periphery, there is a sharp fall-off in cell implant technology before it will be widely used.
numbers as the distance from the resection cavity increases
[20]. Extensive surgical resection of gliomas can prolong
progression-free and overall survival by decreasing tumor Procedure
burden and possibly increasing adjuvant therapy efficacy as
it has with other types of solid tumors [47]. Moreover, the SBT is based on the principle that radiation will induce
use of adjuvant therapies have been developed to help necrosis in a targeted tumor volume and that necrotic tissue
reduce the tumor burden at the tumor periphery. One of will then be removed by macrophages [60]. Stereotactic
these developments is SBT (Table 18.2). computed tomography and magnetic resonance images
It is generally considered beneficial to provide radiation (MRI) are co-registered for treatment planning and dosime-
to the gliomatous areas around the surgical cavity, especially try calculations [61]. The typical cumulative, prescribed
for high-grade gliomas [14, 48, 49]. The techniques to deliver dose ranges from 50–65 Gy over 3–5 days [14, 48, 49]. In
this radiation include stereotactic radiation therapy and SBT interstitial radiotherapy, 125I seeds are administered via ste-
[14, 48, 49]. SBT involves the delivery of radiation by stereotactically implanted catheters intraoperatively [15, 62].
reotactically implanting radioactive sources near or within Dose rates vary and individual institutions vary over whether
tumors [14, 48, 49]. SBT can be applied peri- or postopera- a low dose, which allows build up of damage in tumor cells,
tively by either permanent isotope seeds or by temporary or a high dose rate, which decreases total treatment time, is
catheters containing radioisotopes [14, 48, 49]. There are a administered [14]. A potential barrier to the success of radio-
variety of different isotopes that can be administered for active seed brachytherapy may be the inhomogeneous distri-
radiation therapy including Iodine-125, Iodine-131, and bution of doses [14, 48, 49]. Tumor areas adjacent to seeds
Iridium-192, among others [14, 48, 49]. The exposure times receive a very high dose that may contribute to toxicity,
of these seeds range from a few days to several months and whereas the areas in between seeds may have cold spots and
vary whether temporary implants or permanent seeds are the tumor is potentially under-dosed in these regions [63].
used [14, 48, 49]. In the central nervous system, the most Recently, the advent of GliaSite brachytherapy allows for
common radioactive source used for SBT are Iodine-125 administration of 125I on an outpatient basis [64–66]. With
seeds (125I) which emit low-energy radiation of 0.028– GliaSite, an expandable balloon catheter can be implanted
0.035 meV and have a steep falloff dose that reduces periph- intracranially near the tumor or within a resection cavity
eral normal tissue exposure [14, 15]. thereby providing a conformal fit (Fig. 18.1) [64–66]. Liquid
125
SBT has been used as adjuvant treatment following surgi- I is then used as the source of radiation and the balloon is
cal resection for both primary and recurrent low- and high- inflated through an accessible subcutaneous port; this admin-
grade gliomas and has been used for tumors in eloquent istration provides the opportunity for treatment to be admin-
regions that are considered unresectable [14, 48, 49]. The istered after several weeks of postoperative recovery [66, 67].
theoretical benefit of SBT is that it allows for precise surgical As the liquid radioisotope distributes uniformly throughout
implantation of radioisotopes within the tumor or tumor cav- the balloon, there is a uniform delivery of radiation to the
ity, thus providing local radiation to adjacent tissue while region surrounding the balloon [64–66]. This minimizes the
Table 18.2 Summary of benefits and risks of stereotactic brachytherapy use in gliomas
Stereotactic brachytherapy benefits Stereotactic brachytherapy risks
Minimizes surgical intervention in functionally critical tissue Prospective, randomized controlled trials have failed to demonstrate
a survival advantage
Avoids systemic toxicity of chemotherapy Invasive procedure when compared to external beam radiation therapy
Avoids neuro-cognitive effects in children who undergo EBRT Limited to tumors <5 cm in diameter and spherical in shape
Retrospective series demonstrate both safety and efficacy Complications: Radiation-induced necrosis and neurological dysfunction
Fig. 18.1 GliaSite radiation therapy system device. (a) Schematic shows device components: balloon, subcutaneous tube, and infusion port.
(b) Intraoperative photograph of device implantation after tumor resection
potential for “hot” or “cold” radiation regions and also allows tumor in several retrospective studies [56, 70, 71]. For high-
for a predictable distribution of radiation dose, which grade gliomas, SBT has been utilized for both operated and
decreases with distance from the balloon surface [64–66]. inoperable gliomas in combination with adjuvant therapies
Whereas stereotactic implantation of radiation seed catheters and has been shown to be safely tolerated [55, 72]. Several
can be placed in the midst of actual gross tumor, the GliaSite studies have retrospectively studied the safety and feasibility
technique is only useful when a surgical cavity has been cre- of SBT in high-grade gliomas using GliaSite in combination
ated by resection of all or most of the tumor [64–66]. There with resected brain tumors [66, 67].
also continues to be interest in developing targeted delivery
of radiation dose by using radioisotopes bound to molecules,
which preferentially bind to glioma cells [64–66]. In addi- Previous Studies
tion to 125I, I-131 labeled TM-601, a synthetic chlorotoxin,
has also been demonstrated to effectively deliver radiation Previous studies (Table 18.3) have demonstrated the safety
[68]. Efficacy has not been demonstrated, but this general and efficacy of brachytherapy for patients with gliomas.
approach remains promising and continues to be a subject of Gutin et al. studied 63 patients who were administered
investigation. brachytherapy (125I) in addition to EBRT, procarbazine,
lomustine, and vincristine from 1982 to 1990. The median
survival for patients with anaplastic gliomas was 157 weeks
Indications for those with implants and 165 weeks for those without
implants, while the median survival for patients with GB was
Therapeutic management of gliomas includes some combina- 88 weeks for those with implants and 67 weeks for those
tion of stereotactic biopsy and radiologic monitoring, surgical without implants [73]. Wen and colleagues treated 56
resection, and chemo-radiation therapy [3–5, 16]. Although patients with GB from 1987 to 1993 [74]. The patients in this
surgical resection is the gold standard, tumors located in elo- study underwent surgery, limited field EBRT, and brachy-
quent or critical areas of the brain are not always amenable to therapy with 125I [74]. The median survival for patients
extensive debulking [49]. In children with inoperable low- undergoing brachytherapy was 18 months compared with 11
grade gliomas, SBT is a treatment option that avoids the months for a matched brachytherapy control group with sim-
hematologic and otologic toxicity of chemotherapy and the ilar clinical and radiologic features [74]. Thirty six patients
cognitive and neuropsychological deficits associated with (64 %), however, underwent reoperation for symptomatic
EBRT [50, 69]. In these children, retrospective studies have radiation necrosis [74]. Welsh et al. studied 20 patients from
found SBT to be safe and have an efficacy that is comparable eight institutions who underwent Gliasite brachytherapy
to conventional chemo-radiation with 5 and 10-year survival boost [66]. These patients underwent extensive debulking
rates of 97 % and 92 %, respectively [51, 52, 69]. Additionally, followed by EBRT [66]. The average survival for this study
in adults with low-grade gliomas, the combination of SBT population was 11.4 months, which was 3 months longer
and microsurgery reduces the high complication rate of injury than historical controls [66]. Three patients (14 %) in this
to eloquent brain regions while effectively irradiating the series experienced central nervous system toxicity [66].
Table 18.3 Summary of previous studies on the use of brachytherapy

Number Brachytherapy EBRT Median survival
Author Years of patients Glioma type type dose (cGy) (months) Complications
Gobitti et al.a 2006–2008 15 Recurrent WHO IV Gliasite 4,500 13 Radiation necrosis
(20 %), death (13 %)
Suchorska et al. 1982–2006 95 Recurrent WHO Iodine-125 20 3.3 % permanent
grade II morbidity
Chen et al.a 1999–2002 18 WHO IV Iodine-125 4,000 26.3 Radiation necrosis (61 %)
Welsh et al. 2000–2004 20 WHO IV Gliasite 5,000 11.4 Radiation necrosis (14 %)
Reardon et al. 1996–2003 43 Recurrent WHO Iodine-131 4,600 14.8 Radiation necrosis
III-IV (12 %), hematologic
toxicity (23 %)
Akabani et al. 1996–2003 18 WHO III-IV Iodine-131 4,800 18.2 Radiation necrosis (39 %)
Selker et al.b 1987–1994 133 WHO III-IV Iodine-125 6,000 15.7 Radiation necrosis
(15.3 %)
Koot et al. 1998–1995 66 WHO IV Iodine-125 (45), 5,000 16 Radiation necrosis (5 %),
Iridium-192 (21) Vascular ischemia (14 %)
Johannsen et al. 1992–1996 44 WHO III-IV Iridium-192 6,600 11.7
Laperriere et al.b 1986–1996 71 WHO III-IV Iodine-125 6,000 13.8 Worsened neurological
function (10 %)
Patchell et al. <1997 33 WHO III-IV Californium-252 5,400 10.9 Scalp necrosis (30 %)
Micheletti et al. <1996 24 WHO III-IV Iridium-192 2,500 8 Acute complication
Bernstein et al. 1986–1992 46 Recurrent WHO Iodine-125 10.6 Radiation necrosis (26 %)
III-IV
Wen et al. 1987–1993 56 WHO IV Iodine-125 18 Radiation necrosis (64 %)
Riva et al. <1994 24 Recurrent WHO IV Iodine-131 3,650 16
Gutin et al. 1982–1990 63 WHO III-IV Iodine-125 26
a
Study stopped because of complications related to brachytherapy application
b
Randomized control trial
In addition to these studies, there have been two random- Limitations and Complications
ized control studies to date on the use of brachytherapy for
patients with gliomas (Table 18.3) [58, 59]. Selker and col- The risks and benefits of SBT are summarized in Table 18.2.
leagues randomized 270 patients with newly diagnosed Patients who are more amenable to SBT include those with
malignant gliomas to surgery, EBRT, and carmustine or sur- tumors that are smaller <5 cm in diameter to minimize
gery, EBRT, carmustine, and interstitial radiotherapy boost damage to surrounding brain parenchyma [62]. Moreover,
with 125I from 1987 to 1994 [59]. The median survival for patient selection is limited to those with spherical tumors, as
patients undergoing treatment with 125I was 68.1 weeks as ovoid or complex polycyclic shapes create challenges in
compared to 58.8 weeks [59]. The difference in survival was obtaining uniform dose distributions [62, 70]. Complications
not statistically significant [59]. Likewise, Laperriere et al. that are due to SBT radiation most commonly lead to blood-
randomized 140 patients from 1986 to 1996 to brachyther- brain barrier breakdown at high radiation doses and result in
apy (125I) plus EBRT versus EBRT for patients with biopsy- radiation-induced necrosis [75]. Necrotizing radiation
proven malignant astrocytomas [58]. The 71 patients changes can cause mass effect and lead to devastating side
randomized to the implant arm had a median survival of effects including both transient and long-term neurological
13.8 months as compared to 13.2 months for the 69 patients dysfunction, increased use of steroids to reduce edema, and
with EBRT [58]. There was no significant difference in sur- formation of cysts that can exert mass effect [62, 70].
vival between the two treatment arms [58]. Thus, while SBT The GliaSite balloon brachytherapy is designed for sur-
remains a treatment option for both low- and high-grade gical cavities up to 3–4 cm in size [64–66]. A substantial
gliomas, the data is conflicting and inconclusive. This is true radiation dose can be delivered up to a cm or more beyond
when SBT is used in combination with standard surgical the balloon surface such that minimal gross residual disease
resection and/or chemo-radiation. can be encompassed [64–66]. However, the higher doses
required to deliver a sufficient amount of radiation at 7. Chaichana KL, McGirt MJ, Laterra J, Olivi A, Quinones-Hinojosa
increasing distances from the balloon surface may result in A. Recurrence and malignant degeneration after resection of adult
hemispheric low-grade gliomas. J Neurosurg. 2010;112(1):10–7.
more toxicity [64–66]. More study of this technique are 8. Chaichana KL, McGirt MJ, Niranjan A, Olivi A, Burger PC,
therefore required. Quinones-Hinojosa A. Prognostic significance of contrast-
enhancing low-grade gliomas in adults and a review of the litera-
ture. Neurol Res. 2009;31(9):931–9.
9. Chaichana KL, McGirt MJ, Woodworth GF, Datoo G, Tamargo RJ,
Conclusion Weingart J, et al. Persistent outpatient hyperglycemia is indepen-
dently associated with survival, recurrence and malignant degener-
While there are no randomized control trials, there is a grow- ation following surgery for hemispheric low grade gliomas. Neurol
ing body of evidence supporting maximal resection without Res. 2010;32(4):442–8.
10. Chaichana KL, Zaidi H, Pendleton C, McGirt MJ, Grossman R,
causing an iatrogenic deficit for patients with high-grade Weingart JD, et al. The efficacy of carmustine wafers for older
gliomas, as well as low-grade glioma patients with symp- patients with glioblastoma multiforme: prolonging survival. Neurol
toms or enlarging tumors. SBT can be utilized as adjuvant Res. 2011;33(7):759–64.
treatment for primary and recurrent low- and high-grade 11. McGirt MJ, Chaichana KL, Attenello FJ, Weingart JD, Than K,
Burger PC, et al. Extent of surgical resection is independently asso-
gliomas in combination with surgical resection, chemother- ciated with survival in patients with hemispheric infiltrating low-
apy, and/or EBRT. The use of intracranial SBT can also be grade gliomas. Neurosurgery. 2008;63(4):700–7; author reply 7–8.
helpful for patients with tumors located in functionally criti- 12. McGirt MJ, Chaichana KL, Gathinji M, Attenello FJ, Than K, Olivi
cal areas not amenable to surgical resection. However, to A, et al. Independent association of extent of resection with survival
in patients with malignant brain astrocytoma. J Neurosurg. 2009;
date, randomized control trials have yet to show a survival 110(1):156–62.
benefit for patients with SBT. Thus, while SBT can be uti- 13. McGirt MJ, Than KD, Weingart JD, Chaichana KL, Attenello FJ,
lized in a select group of patients to minimize adverse side Olivi A, et al. Gliadel (BCNU) wafer plus concomitant temozolo-
effects of surgical and chemo-radiation treatment, more evi- mide therapy after primary resection of glioblastoma multiforme.
J Neurosurg. 2009;110(3):583–8 [Comparative Study].
dence is needed to determine whether patient outcomes are 14. Liu BL, Cheng JX, Zhang X, Zhang W. Controversies concerning
significantly improved with this therapy. The development the application of brachytherapy in central nervous system tumors.
of new approaches or new radiopharmaceutical agents to J Cancer Res Clin Oncol. 2010;136(2):173–85 [Review].
deliver intra-cavitary brachytherapy such as the GliaSite bal- 15. Ruge MI, Simon T, Suchorska B, Lehrke R, Hamisch C, Koerber F,
et al. Stereotactic brachytherapy with iodine-125 seeds for the treat-
loon catheter which delivers a predictable homogeneous ment of inoperable low-grade gliomas in children: long-term out-
dose distribution or the use of radiolabelled antibodies to tar- come. J Clin Oncol. 2011;29(31):4151–9.
get treatment at tumor cells has the potential to enhance the 16. Chaichana KL, Halthore AN, Parker SL, Olivi A, Weingart JD,
therapeutic ratio of this approach. Brem H, et al. Factors involved in maintaining prolonged functional
independence following supratentorial glioblastoma resection.
J Neurosurg. 2011;114(3):604–12. Clinical article.
17. Chaichana KL, Parker SL, Mukherjee D, Cheng JS, Gokaslan ZL,
References McGirt MJ. Assessment of the extent of surgical resection as a
predictor of survival in patients with primary osseous spinal neo-
1. Keles GE, Lamborn KR, Berger MS. Low-grade hemispheric glio- plasms. Clin Neurosurg. 2011;58:117–21.
mas in adults: a critical review of extent of resection as a factor 18. Chaichana KL, Parker SL, Olivi A, Quinones-Hinojosa A. Long-
influencing outcome. J Neurosurg. 2001;95(5):735–45. term seizure outcomes in adult patients undergoing primary resec-
2. Piepmeier J, Baehring JM. Surgical resection for patients with tion of malignant brain astrocytomas. J Neurosurg. 2009;111(2):
benign primary brain tumors and low grade gliomas. J Neurooncol. 282–92. Clinical article.
2004;69(1–3):55–65. 19. McGirt MJ, Mukherjee D, Chaichana KL, Than KD, Weingart JD,
3. Chaichana KL, Chaichana KK, Olivi A, Weingart JD, Bennett R, Quinones-Hinojosa A. Association of surgically acquired motor and
Brem H, et al. Surgical outcomes for older patients with glioblas- language deficits on overall survival after resection of glioblastoma
toma multiforme: preoperative factors associated with decreased multiforme. Neurosurgery. 2009;65(3):463–9. discussion 9-70.
survival. J Neurosurg. 2011;114(3):587–94. Clinical article. 20. Quinones-Hinojosa A, Chaichana K. The human subventricular
4. Chaichana K, Parker S, Olivi A, Quinones-Hinojosa A. A proposed zone: a source of new cells and a potential source of brain tumors.
classification system that projects outcomes based on preoperative Exp Neurol. 2007;205(2):313–24.
variables for adult patients with glioblastoma multiforme. J Neurosurg. 21. Tait MJ, Petrik V, Loosemore A, Bell BA, Papadopoulos
2010;112(5):997–1004 [Research Support, N.I.H., Extramural MC. Survival of patients with glioblastoma multiforme has not
Research Support, Non-U.S. Gov’t]. improved between 1993 and 2004: analysis of 625 cases. Br J
5. Chaichana KL, Garzon-Muvdi T, Parker S, Weingart JD, Olivi A, Neurosurg. 2007;21(5):496–500 [Randomized Controlled Trial
Bennett R, et al. Supratentorial glioblastoma multiforme: the role of Research Support, Non-U.S. Gov’t].
surgical resection versus biopsy among older patients. Ann Surg 22. Dandy WE. Removal of right cerebral hemisphere for certain
Oncol. 2011;18(1):239–45 [Clinical Trial]. tumors with hemiplegia. JAMA. 1928;90:823–5.
6. Chaichana KL, Kosztowski T, Niranjan A, Olivi A, Weingart JD, 23. Lacroix M, Abi-Said D, Fourney DR, Gokaslan ZL, Shi W,
Laterra J, et al. Prognostic significance of contrast-enhancing ana- DeMonte F, et al. A multivariate analysis of 416 patients with glio-
plastic astrocytomas in adults. J Neurosurg. 2010;113(2):286–92 blastoma multiforme: prognosis, extent of resection, and survival. J
[Research Support, Non-U.S. Gov’t]. Neurosurg. 2001;95(2):190–8.
24. Sanai N, Polley MY, McDermott MW, Parsa AT, Berger MS. An 43. Sanai N, Chang S, Berger MS. Low-grade gliomas in adults.
extent of resection threshold for newly diagnosed glioblastomas. J Neurosurg. 2011;115(5):948–65 [Review].
J Neurosurg. 2011;115(1):3–8. 44. Jung TY, Jung S, Moon JH, Kim IY, Moon KS, Jang WY. Early
25. Keles GE, Chang EF, Lamborn KR, Tihan T, Chang CJ, Chang SM, prognostic factors related to progression and malignant transforma-
et al. Volumetric extent of resection and residual contrast enhance- tion of low-grade gliomas. Clin Neurol Neurosurg. 2011;113(9):
ment on initial surgery as predictors of outcome in adult patients 752–7 [Randomized Controlled Trial].
with hemispheric anaplastic astrocytoma. J Neurosurg. 2006;105(1): 45. Recht LD, Lew R, Smith TW. Suspected low-grade glioma: is
34–40. deferring treatment safe? Ann Neurol. 1992;31(4):431–6.
26. Laws ER, Parney IF, Huang W, Anderson F, Morris AM, Asher A, 46. Stupp R, Mason WP, van den Bent MJ, Weller M, Fisher B,
et al. Survival following surgery and prognostic factors for recently Taphoorn MJ, et al. Radiotherapy plus concomitant and adjuvant
diagnosed malignant glioma: data from the Glioma Outcomes temozolomide for glioblastoma. N Engl J Med. 2005;352(10):987–
Project. J Neurosurg. 2003;99(3):467–73. 96 [Clinical Trial Clinical Trial, Phase III Multicenter Study
27. Levin VA, Yung WK, Bruner J, Kyritsis A, Leeds N, Gleason MJ, Randomized Controlled Trial Research Support, Non-U.S. Gov’t
et al. Phase II study of accelerated fractionation radiation therapy Research Support, U.S. Gov’t, P.H.S.].
with carboplatin followed by PCV chemotherapy for the treatment 47. House MG, Gonen M, Jarnagin WR, D’Angelica M, Dematteo RP,
of anaplastic gliomas. Int J Radiat Oncol Biol Phys. 2002;53(1): Fong Y, et al. Prognostic significance of pathologic nodal status in
58–66. patients with resected pancreatic cancer. J Gastrointest Surg. 2007;
28. Tortosa A, Vinolas N, Villa S, Verger E, Gil JM, Brell M, et al. 11(11):1549–55.
Prognostic implication of clinical, radiologic, and pathologic 48. Kreth FW, Thon N, Siefert A, Tonn JC. The place of interstitial
features in patients with anaplastic gliomas. Cancer. 2003;97(4): brachytherapy and radiosurgery for low-grade gliomas. Adv Tech
1063–71. Stand Neurosurg. 2010;35:183–212 [Review].
29. Pope WB, Sayre J, Perlina A, Villablanca JP, Mischel PS, Cloughesy 49. Schwarz SB, Thon N, Nikolajek K, Niyazi M, Tonn JC, Belka C,
TF. MR imaging correlates of survival in patients with high-grade et al. Iodine-125 brachytherapy for brain tumours—a review. Radiat
gliomas. AJNR Am J Neuroradiol. 2005;26(10):2466–74. Oncol. 2012;7:30 [Review].
30. Grant R, Metcalfe SE. Biopsy versus resection for malignant gli- 50. Jalali R, Mallick I, Dutta D, Goswami S, Gupta T, Munshi A, et al.
oma. Cochrane Database Syst Rev. 2005;2, CD002034. Factors influencing neurocognitive outcomes in young patients
31. Hess KR. Extent of resection as a prognostic variable in the with benign and low-grade brain tumors treated with stereotactic
treatment of gliomas. J Neurooncol. 1999;42(3):227–31. conformal radiotherapy. Int J Radiat Oncol Biol Phys. 2010;
32. Nazzaro JM, Neuwelt EA. The role of surgery in the management 77(4):974–9 [Research Support, Non-U.S. Gov’t].
of supratentorial intermediate and high-grade astrocytomas in 51. Korinthenberg R, Neuburger D, Trippel M, Ostertag C, Nikkhah
adults. J Neurosurg. 1990;73(3):331–44. G. Long-term results of brachytherapy with temporary iodine-125
33. Quigley MR, Maroon JC. The relationship between survival and the seeds in children with low-grade gliomas. Int J Radiat Oncol Biol
extent of the resection in patients with supratentorial malignant Phys. 2011;79(4):1131–8.
gliomas. Neurosurgery. 1991;29(3):385–8; discussion 8-9. 52. Voges J, Sturm V, Berthold F, Pastyr O, Schlegel W, Lorenz
34. Sanai N, Berger MS. Glioma extent of resection and its impact on WJ. Interstitial irradiation of cerebral gliomas in childhood by per-
patient outcome. Neurosurgery. 2008;62(4):753–64; discussion manently implanted 125-iodine–preliminary results. Klin Padiatr.
264-6. 1990;202(4):270–4.
35. Low-Grade Glioma Guidelines Team in association with the 53. Ostertag CB, Kreth FW. Iodine-125 interstitial irradiation for cere-
Guidelines and Outcomes Committee of the American Association bral gliomas. Acta Neurochir (Wien). 1992;119(1–4):53–61.
of Neurological Surgeons. Practice parameters in adults with sus- 54. Voges J, Sturm VV. Interstitial irradiation with stereotactically
pected or known supratentorial nonoptic pathway low-grade gli- implanted I-125 seeds for the treatment of cerebral glioma. Crit Rev
oma. Neurosurg Focus. 1998;4(6):e10. Neurosurg. 1999;9(4):223–33.
36. Philippon JH, Clemenceau SH, Fauchon FH, Foncin 55. Voges J, Treuer H, Schlegel W, Pastyr O, Sturm V. Interstitial
JF. Supratentorial low-grade astrocytomas in adults. Neurosurgery. irradiation of cerebral gliomas with stereotactically implanted
1993;32(4):554–9. iodine-125 seeds. Acta Neurochir Suppl (Wien). 1993;58:
37. Rajan B, Pickuth D, Ashley S, Traish D, Monro P, Elyan S, et al. 108–11.
The management of histologically unverified presumed cerebral 56. Kreth FW, Faist M, Grau S, Ostertag CB. Interstitial 125I radiosur-
gliomas with radiotherapy. Int J Radiat Oncol Biol Phys. 1994; gery of supratentorial de novo WHO Grade 2 astrocytoma and oli-
28(2):405–13. goastrocytoma in adults: long-term results and prognostic factors.
38. van Veelen ML, Avezaat CJ, Kros JM, van Putten W, Vecht Cancer. 2006;106(6):1372–81.
C. Supratentorial low grade astrocytoma: prognostic factors, dedif- 57. Bernstein M, Laperriere N, Glen J, Leung P, Thomason C, Landon
ferentiation, and the issue of early versus late surgery. J Neurol AE. Brachytherapy for recurrent malignant astrocytoma. Int J
Neurosurg Psychiatry. 1998;64(5):581–7. Radiat Oncol Biol Phys. 1994;30(5):1213–7 [Clinical Trial].
39. Soffietti R, Chio A, Giordana MT, Vasario E, Schiffer D. Prognostic 58. Laperriere NJ, Leung PM, McKenzie S, Milosevic M, Wong S,
factors in well-differentiated cerebral astrocytomas in the adult. Glen J, et al. Randomized study of brachytherapy in the initial man-
Neurosurgery. 1989;24(5):686–92. agement of patients with malignant astrocytoma. Int J Radiat Oncol
40. Pouratian N, Schiff D. Management of low-grade glioma. Curr Biol Phys. 1998;41(5):1005–11 [Clinical Trial Randomized
Neurol Neurosci Rep. 2010;10(3):224–31 [Review]. Controlled Trial Research Support, Non-U.S. Gov’t].
41. Wisoff JH, Sanford RA, Heier LA, Sposto R, Burger PC, Yates AJ, 59. Selker RG, Shapiro WR, Burger P, Blackwood MS, Arena VC,
et al. Primary neurosurgery for pediatric low-grade gliomas: a pro- Gilder JC, et al. The Brain Tumor Cooperative Group NIH Trial
spective multi-institutional study from the Children’s Oncology 87-01: a randomized comparison of surgery, external radiother-
Group. Neurosurgery. 2011;68(6):1548–54. [Clinical Trial Multicenter apy, and carmustine versus surgery, interstitial radiotherapy
Study Research Support, N.I.H., Extramural]; discussion 54-5. boost, external radiation therapy, and carmustine. Neurosurgery.
42. Smith JS, Chang EF, Lamborn KR, Chang SM, Prados MD, Cha S, 2002;51(2):343–55. [Clinical Trial Comparative Study
et al. Role of extent of resection in the long-term outcome of low- Randomized Controlled Trial Research Support, U.S. Gov't,
grade hemispheric gliomas. J Clin Oncol. 2008;26(8):1338–45. P.H.S.]; discussion 55–7.
60. Ostertag CB. Brachytherapy–interstitial implant radiosurgery. Acta in patients with gliomas involving pyramidal tracts. Neurosurgery.
Neurochir Suppl (Wien). 1993;58:79–84 [Research Support, 2007;61(5):935–48. [Randomized Controlled Trial Research
Non-U.S. Gov’t]. Support, Non-U.S. Gov't]; discussion 48-9.
61. De Benedictis A, Moritz-Gasser S, Duffau H. Awake mapping opti- 69. Schumacher M. Long-term results of brachytherapy with tempo-
mizes the extent of resection for low-grade gliomas in eloquent rary iodine-125 seeds in children with low-grade gliomas. Int J
areas. Neurosurgery. 2010;66(6):1074–84; discussion 84. Radiat Oncol Biol Phys. 2011;80(5):1604. [Comment Letter];
62. Suchorska B, Ruge M, Treuer H, Sturm V, Voges J. Stereotactic author reply.
brachytherapy of low-grade cerebral glioma after tumor resection. 70. Schnell O, Scholler K, Ruge M, Siefert A, Tonn JC, Kreth
Neuro Oncol. 2011;13(10):1133–42 [Clinical Trial]. FW. Surgical resection plus stereotactic 125I brachytherapy in adult
63. Schupak K, Malkin M, Anderson L, Arbit E, Lindsley K, Leibel patients with eloquently located supratentorial WHO grade II gli-
S. The relationship between the technical accuracy of stereotactic oma - feasibility and outcome of a combined local treatment con-
interstitial implantation for high grade gliomas and the pattern of cept. J Neurol. 2008;255(10):1495–502.
tumor recurrence. Int J Radiat Oncol Biol Phys. 1995;32(4): 71. Kreth FW, Faist M, Rossner R, Birg W, Volk B, Ostertag CB. The
1167–76. risk of interstitial radiotherapy of low-grade gliomas. Radiother
64. Gabayan AJ, Green SB, Sanan A, Jenrette J, Schultz C, Papagikos Oncol. 1997;43(3):253–60 [Clinical Trial Controlled Clinical Trial].
M, et al. GliaSite brachytherapy for treatment of recurrent malignant 72. Fabrini MG, Perrone F, De Franco L, Pasqualetti F, Grespi S,
gliomas: a retrospective multi-institutional analysis. Neurosurgery. Vannozzi R, et al. Perioperative high-dose-rate brachytherapy in the
2006;58(4):701–9. [Comparative Study]; discussion 9. treatment of recurrent malignant gliomas. Strahlenther Onkol.
65. Gobitti C, Borsatti E, Arcicasa M, Roncadin M, Franchin G, 2009;185(8):524–9.
Minatel E, et al. Treatment of recurrent high-grade gliomas with 73. Gutin PH, Prados MD, Phillips TL, Wara WM, Larson DA, Leibel
GliaSite brachytherapy: a prospective mono-institutional Italian SA, et al. External irradiation followed by an interstitial high activ-
experience. Tumori. 2011;97(5):614–9 [Clinical Trial]. ity iodine-125 implant “boost” in the initial treatment of malignant
66. Welsh J, Sanan A, Gabayan AJ, Green SB, Lustig R, Burri S, et al. gliomas: NCOG study 6G-82-2. Int J Radiat Oncol Biol Phys.
GliaSite brachytherapy boost as part of initial treatment of glioblas- 1991;21(3):601–6 [Clinical Trial Multicenter Study Research
toma multiforme: a retrospective multi-institutional pilot study. Int Support, U.S. Gov’t, P.H.S.].
J Radiat Oncol Biol Phys. 2007;68(1):159–65 [Evaluation Studies 74. Wen PY, Alexander 3rd E, Black PM, Fine HA, Riese N, Levin JM,
Multicenter Study]. et al. Long term results of stereotactic brachytherapy used in the
67. Chino K, Silvain D, Grace A, Stubbs J, Stea B. Feasibility and initial treatment of patients with glioblastomas. Cancer.
safety of outpatient brachytherapy in 37 patients with brain tumors 1994;73(12):3029–36 [Clinical Trial Research Support, U.S. Gov’t,
using the GliaSite Radiation Therapy System. Med Phys. 2008; P.H.S.].
35(7):3383–8. 75. Julow J, Kolumban Z, Viola A, Major T, Kolumban G. Prediction of
68. Wu JS, Zhou LF, Tang WJ, Mao Y, Hu J, Song YY, et al. Clinical volumetric change in the “triple ring” caused by glioma I-125
evaluation and follow-up outcome of diffusion tensor imaging- brachytherapy. Neuro Oncol. 2008;10(4):583–92 [Research
based functional neuronavigation: a prospective, controlled study Support, Non-U.S. Gov’t].
Malignant Glioma: Viewpoint—
Chemotherapy 19
Roger Stupp, Krisztian Homicsko,
and J. Gregory Cairncross
protective barrier. In most malignant gliomas, the blood–brain

Introduction barrier is at least partially disrupted, evidenced by contrast
enhancement on computed tomography (CT) and magnetic
Despite optimal surgery and adjuvant radiotherapy, malig- resonance imaging (MRI) studies. The blood–brain barrier
nant gliomas almost always recur [1]. One cause of failure is may not be an impediment to the identification of effective
explained by the diffusely infiltrating pattern of growth dis- drug therapies, because radiographic responses can be seen,
played by most gliomas, with malignant cells disseminated but it is a significant obstacle to achieve adequate tissue pen-
far beyond the initial bulky tumor. Although most relapses etration and to the development of curative drug therapies.
occur at or near the initial site of tumor, either along the Infiltrating microscopic glioma hidden behind intact portions
resection margin or at the edge of the radiation field, distant of the blood–brain barrier escape the exposure to drugs that
recurrences or seeding of the cerebrospinal fluid (CSF) do not cross the barrier and set the stage for tumor regrowth,
occurs in long-surviving patients. Malignant gliomas are not despite effective multimodality local therapies. Furthermore,
a localized disease but affect the whole brain and may dis- gliomas may induce aberrant capillary growth, which in turn
play wide dissemination even at early stages. Thus, thera- may limit the penetration of chemotherapeutics into the
peutic strategies aiming only at bulky local disease are tumor and adjacent brain. Other factors such as tumor size,
doomed to fail. The successful treatments of the future will cell density, and increased intratumoral and intracerebral
need to eradicate microscopic disease in many sites within pressure can inhibit blood flow and reduce drug delivery to
the brain and do so without neurotoxic effects. malignant gliomas.
Hence, a prerequisite for a highly efficacious antiglioma
drug, administered intravenously or by mouth, is the ability
Obstacles to Drug Therapy of Brain Tumors to cross the blood–brain barrier. Barrier permeability can be
inferred from the chemical structure of the drug; however,
The Blood–Brain Barrier measurements of drug concentrations in brain and glioma
tissue are difficult. It is neither easy, nor ethical to perform
The brain is considered a pharmacological sanctuary by vir- repeated brain biopsies to establish drug tissue levels, and
tue of the blood–brain barrier. Many chemotherapeutics, animal models of glioma are often too artificial to be helpful.
those that are large or hydrophilic, are unable to cross this Measurements of drug levels in the CSF may be a valuable
surrogate for brain and glioma exposure, but false readings
may occur due to the effects of tumor edema or compartmen-
talization of CSF spaces that result from pressure effects or
R. Stupp, M.D. (*) circulating tumor cells that disrupt the flow of CSF [2].
Department of Oncology and Cancer Center,
University Hospital Zurich, Zurich, Switzerland
e-mail: roger.stupp@usz.ch
The Complex Biology of Malignant Gliomas
K. Homicsko, M.D., Ph.D.
Department of Oncology, Centre Hospitalier Universitaire Vaudois,
Lausanne, Vaud, Switzerland The complexity of glioma biology is insufficiently reflected
by the conventional pathological classification as glioblas-
J.G. Cairncross, M.D.
Southern Alberta Cancer Research Institute, University of Calgary, toma (World Health Organization grade IV), anaplastic
Calgary, AB, Canada (grade III) astrocytoma, oligodendroglioma and mixed
L.S. Chin, W.F. Regine (eds.), Principles and Practice of Stereotactic Radiosurgery, 279
280 R. Stupp et al.
oligoastrocytoma. Mutational analyses, gene expression and Concomitant Medications

methylation (profiling allowed identification of glioblastoma
subclasses proneural, neural, classical, mesenchymal), or Many chemotherapy agents are metabolized by P450-
methylator phenotypes (CIMP) [3–5]. These molecular clas- dependent hepatic enzymes. Antiepileptics and corticoste-
sifications explain, in part the observed prognostic differ- roids, commonly prescribed for patients with gliomas,
ences and due to differences in the underlying genetic events induce these hepatic enzymes, leading to rapid degradation
tumors from specific subtypes are likely to respond differ- of certain classes of chemotherapeutics, undermining effec-
ently to biological therapies. Consequently development tive cancer treatment. The optimal concentrations of chemo-
of novel, targeted therapies needs to be tailored to specific therapeutics may be difficult to achieve when P450-inducing
subtype and associated aberrations. In addition to this inter- drugs are co-administered. The pharmacokinetic profiles of
patient heterogeneity, the importance of intratumoral hetero- new drug therapies for malignant glioma must be assessed in
geneity is increasingly recognized [6, 7]. Different regions of the presence of concomitant antiepileptic and steroid medi-
the same tumor may harbor different genetic variations due cations especially. Although the increased metabolism can
to clonal evolution within the tumor and hence while one be overcome by increasing the doses of the chemotherapy
region of the tumor is sensitive to the therapy, other regions agents by several-fold, this approach will make drug ther-
can be resistant and hence response is limited and recurrence apy prohibitively expensive and—more importantly—
is evident. An example of relevance in glioblastoma is the potentially dangerous. Should the antiepileptic treatment be
constitutively activated variant of the epidermal growth fac- discontinued inadvertently, a massive overdose of chemo-
tor receptor (EGFRviii) that is only expressed in a fraction of therapy could ensue. To avoid such pitfalls, the use of
tumor cells [8]. nonenzyme-inducing antiepileptics [e.g., valproic acid
The tumor stem cell hypothesis and hierarchical structure (Depakene, Abbott, Chicago, IL), lamotrigine (Lamictal,
of tumors reflects also intratumoral heterogeneity [9]. These GlaxoSmithKline, Research Triangle Park, NC), gabapentin
tumor-initiating cells are capable of self-renewal and pro- (Neurontin, Pfizer, New York, NY), levetiracetam, (Keppra,
moting tumor regrowth (stemness). These cells are also UCB, Brussels, Belgium), topiramate (Topamax, Ortho-
resistant to conventional therapy due to high expression of McNeil Neurologics, Titusville, NJ)] should be considered
multi-drug resistance proteins (MDRs, ABC transporters [14–16]. In clinical practice, the enzyme-inducing effect of
also known as P-glycoproteins), and they are protected from dexamethasone and the enzyme-inhibiting action of the
cytotoxic and ionizing radiation therapy in their frequent anticonvulsant valproic acid are of lesser concern, and for
dormant G0-cell cycle state [9]. In contrast to the initial valproic acid circumstantial evidence suggests a possible
belief, stem-like cancer cells are generated in a dynamic pro- synergistic antitumor effect [17, 18].
cess also from more differentiated non-stem-like cells [10].
Thus combination treatment is required to target both stem-
ness and more differentiated tumor cells by conventional How to Measure Antitumor Activity?
therapy [11].
The tumor microenvironment is another source of hetero- In oncology, new agents are initially tested for antitumor
geneity. Cancer associated fibroblasts (CAFs) and tumor activity in trials that assess whether the tumor type in ques-
endothelial cells are not only important in the development tion shrinks noticeably after drug exposure. Shrinkage is
of tumors but also in the maintenance and in response to usually determined radiographically using CT or MRI and
therapy [12]. While anti-angiogenic therapy is greatly lim- described as a rate of response. By convention, a 20 %
ited by both tumor endothelial and cancer cell adaptation, response rate means that 20 % of the tumors tested were
blockade of oncogenic signaling pathways in tumor cells by smaller after drug exposure. The World Health Organization
tyrosine kinase inhibitors (TKIs) can be attenuated by the (WHO) criteria specify measurement of perpendicular tumor
expression of paracrine signaling proteins (like hepatocyte diameters, generating a baseline maximum surface area; a
growth factor, HGF) from the surrounding CAFs, that in turn reduction of the surface area by ≥50 % is considered a par-
can activate pro-survival pathways (e.g., cMET) in tumor tial response; disappearance of all tumor is defined as com-
cells [13]. plete response, and an increase by 25 % is considered tumor
Taken together, the understanding of the complexity of progression. Regressions of <50 % or increases <25 % are
tumor biology in tumor development and also in response to called stable disease. For neurooncology, these stringent
therapy is crucial in our aim to overcome innate and adaptive imaging criteria have been supplemented by an assessment
resistance mechanisms and to achieve more pronounced and of the use of corticosteroids, which must be stable or decreasing
sustained responses [7]. in dose in order to qualify for a response or progression to be
19 Malignant Glioma: Viewpoint—Chemotherapy 281
declared (the so-called Macdonald criteria) [19]. In recent recommendations of the RANO consortium should be
years in general oncology, unidimensional tumor measure- considered.
ments have been proposed to simplify and speed up the pro- Because many gliomas are difficult to measure using per-
cess of assessing response, however in neurooncology these pendicular diameters, and to acknowledge the possibility
so-called RECIST criteria have not yet been adopted or vali- that prolonged disease stabilization may indeed be a genuine
dated [20]. With the increased use of anti-angiogenic agents indicator of anti-glioma activity, the percentage of patients
and their effect on normalizing the blood–brain barrier, who are progression-free at 6 months has been proposed as
increased attention is to be given on the tumor extension on an alternative measurement of drug efficacy in neurooncology
T2 and FLAIR MRI images rather than solely assessing the [24]. This endpoint, although never formally validated as a
decrease in contrast uptake. An international working party surrogate of overall survival, has been widely adopted in
is addressing methodological issues surrounding response recent years in chemotherapy trials for recurrent malignant
evaluation with modern treatments. A first set of RANO- glioma, and is often also used in the context of newly diag-
criteria (Response Assessment in Neuro-Oncology) has been nosed disease, as well. Nevertheless, overall survival remains
published [21]. the most important and reliable measurement of treatment
Because objective responses are uncommon in neuroon- outcome and randomized trials the best way to test a new
cology (i.e., most tumors are resistant to most drugs) and drug for glioma [25].
may be difficult to measure, even when they occur, research
reports on the medical treatment of brain tumor frequently
interpret stable disease or minor response (less than 50 %) as Active Agents
evidence of antitumor activity. This practice inflates response
rates and is of uncertain merit. With MRI having replaced Table 19.1 shows selected active cytotoxic agents.
CT, the matter of response assessment has been further com-
plicated. Enhanced CT is a relatively simple imaging tool.
MRI, on the other hand, is a better and more complicated Nitrosoureas
machine with a wide assortment of T1, T2, fluid-attenuated
inversion recovery (FLAIR), diffusion, and other sequences: Nitrosoureas were the principal drugs for malignant glioma for
Which ones should be used to assess the response of gliomas over 20 years. These lipophilic alkylating agents readily cross
to new drugs? For now, the WHO and Macdonald style the blood–brain barrier and have shown some activity in
response criteria, currently usually applied to T1 sequences patients with recurrent malignant glioma. Efficacy of lomustine
on gadolinium-enhanced MRI, continue to be useful [22, 23], [CCNU] has recently been confirmed in prospective random-
when novel vasculature-modifying agents are employed, the ized trials where lomustine served as the control arm [26, 27].
Mechanism of action Toxicity

Carmustine (BCNU) Chlorethylating agent Myelosuppression
Nausea/vomiting
Pulmonary fibrosis
Lomustine (CCNU) Chlorethylating agent Myelosuppression
Nausea/vomiting
(Pulmonary fibrosis)
Nimustine (ACNU) Chlorethylating agent Myelosuppression
Nausea/vomiting
Thrombocytopenia
Fotemustine Chlorethylating agent Nausea/vomiting
Temozolomide Methylating agent Thrombocytopenia
Nausea/vomiting
Irinotecan (investigational) Topoisomerase 1 inhibitor Diarrhea
Nausea/vomiting
Neutropenia
Bevacizumab Anti-VEGF monoclonal antibody Hypersensitivity reactions
Hypertension, Proteinuria
Table 19.1 Selected
active cytotoxic agents Hemorrhage
282 R. Stupp et al.
Carmustine 1 cycle of chemotherapy [36]. The superiority of PCV could

Carmustine (BCNU) has been the most widely used chemo- not be confirmed in a subsequent pooled analysis of 432
therapy for malignant glioma. In recent years, lomustine patients treated on Radiation Therapy Oncology Group
(CCNU) or nimustine (ACNU; 1-[(4-amino-2-methyl-5- (RTOG) protocols [37]. The Medical Research Council (UK)
pyrimidinyl)methyl]-1-(2-chloroethyl)-3-nitrosourea) have, conducted a randomized trial of standard focal radiotherapy
in part, replaced BCNU, to lessen pulmonary toxicity. (RT) vs. RT plus adjuvant PCV chemotherapy [38], finding
Fotemustine is a second-generation nitrosourea that has no difference in overall survival. Patients with grade 3 tumors
single-agent activity against recurrent glioma [28]. In Japan also did not benefit from PCV.
and Germany, nimustine (ACNU) is preferred, and fotemus-
tine is most popular in France and Italy [28–32]. Other tox-
icities include moderate and controllable nausea/vomiting, Temozolomide
and unpredictable and late myelosuppression (in particular in
patients who had previously received other chemotherapy). Over the past decade, temozolomide (TMZ) has emerged as
an active agent against malignant glioma [39]. TMZ is an
Lomustine oral alkylating agent, which is rapidly and completely
Lomustine (CCNU) is an orally available nitrosourea and is absorbed and spontaneously converts at physiological pH to
part of the PCV combination regimen. In recent years single the active metabolite, MTIC. TMZ has linear pharmacoki-
agent lomustine was used as the control arm in randomized netics with maximum plasma concentrations 30–90 min
trials investigating novel agents. While those trials failed to after oral intake. Although recommended to be taken in a
demonstrate superiority of the investigational drug, it con- fasting state (at least 1 h before and after intake), food will
firmed in contemporary trials the efficacy of lomustine in only lead to a 10 % reduced area under the concentration
recurrent glioblastoma after first line chemoradiotherapy curve (AUC) and a delayed peak concentration [40]. TMZ
[26, 27]. Response rates were low (4–9 %) with a PFS6 of has an excellent penetration into all body tissues.
19–25 %, comparable to what had previously been reported Pharmacokinetics of TMZ in the CSF have been reported.
for temozolomide in this setting. The AUC in the CSF as a surrogate for brain tissue penetra-
tion corresponded with approximately 20 % of the area under
the curve (AUC) in plasma [41]. Similar to nitrosoureas,
Procarbazine TMZ is a DNA alkylating agent. Methylation of the O-6
position of guanine by TMZ is an especially important bio-
Procarbazine is an alkylating agent that is given orally but logical action, although not the most frequent adduct. If left
requires hepatic activation. In a randomized trial comparing unrepaired, the guanine O-6 lesion triggers cytotoxic
procarbazine vs. BCNU vs. BCNU and methylprednisolone responses and apoptosis.
vs. methylprednisolone alone, the chemotherapy arms were Three pivotal phase II trials led to the approval of temo-
equivalent and slightly superior to methylprednisolone alone zolomide (Temodar; Temodal; Schering-Plough, Kenilworth,
[33]. Median survival was only 9–12 months from diagnosis. NJ, now MSD/Merck & Co, Whitehouse Station, NJ). The
In the setting of recurrent malignant glioma, response and U.S. FDA granted provisional (accelerated) FDA approval
tumor control rates of 30 % have been reported [34, 35]. in 1999 for the treatment of recurrent anaplastic astrocy-
toma. In Europe, it was also approved for the treatment of
recurrent glioblastoma [42–44]. Although the objective
The PCV Regimen response rate in the glioblastoma trials was low (5 % and
7 %, respectively) and was only 35 % in the anaplastic
In the 1970s, investigators from University of California San astrocytoma studies, each trial suggested an increase in the
Francisco (UCSF) developed the PCV regimen, a combination fraction of patients being progression-free at 6 months
of procarbazine, lomustine (CCNU), and vincristine. compared with a historical database [24]. This observation
In 1990, Levin et al. reported that adjuvant chemotherapy encouraged further evaluation of TMZ as a therapy for both
with procarbazine, CCNU, and vincristine (PCV) yielded recurrent and newly diagnosed malignant glioma. For recur-
better survival results than adjuvant BCNU for newly diag- rent glioma, a randomized phase III trial of TMZ vs. PCV in
nosed malignant glioma. This assertion was based on a previously irradiated, chemonaive patients has been con-
reanalysis of a randomized trial conducted between 1977 and ducted by the British National Cancer Research Institute.
1983 [36]. The survival difference favoring PCV was only In a second randomization this trial also compared two
statistically significant for the subgroup of patients (n = 73) different TMZ administration schedules (standard 5-day
with anaplastic astrocytomas (total patients randomized, regimen vs. dose-dense 3 weeks out of 4). In this trial TMZ
148) and good performance status, who had received at least efficacy was equivalent to PCV with less toxicity, however
dose intensification did not demonstrate an improved out- in steroid requirements and median survival of around 9
come, numerically the standard regimen was slightly supe- months [54, 55]. The European Medicines Agency rejected
rior [45]. Similarly, a German randomized NOA-04 trial the application in the absence of controlled data [56].
also compared the standard to a dose-intensified TMZ regi- Two independent phase III randomized trials have been
men and could not show a benefit for the alternative dosing initiated by the RTOG (trial#0825) and by Roche (AvaGlio)
schedule [46]. Intensification of the TMZ regimen failed to in newly diagnosed GBM patients. Bevacizumab is added to
demonstrate improved outcome in a large randomized trial standard chemoradiotherapy (TMZ/RT → TMZ, see below)
in newly diagnosed GBM patients (RTOG0525/Intergroup first line treatment. Both trials failed to demonstrate improve-
trial, discussed below) [47]. ment in overall survival, as expected radiological progression-
The definitive role of TMZ in the initial treatment of free survival was prolonged [57, 58]. Intuitively it appears
newly diagnosed glioblastoma was demonstrated in a ran- that prolongation in progression-free survival and a decrease
domized phase III trial by the European Organisation for in steroid requirements would translate into improved
Research and Treatment of Cancer (EORTC) and the NCI quality of life. However, the detailed analysis of net clinical
Canada Clinical Trials Group [48, 49]. This landmark trial benefits in the Radiation Therapy Oncology Group (RTOG)-
established the role of early use of chemotherapy in the man- sponsored study suggested the contrary, while quality of life
agement of glioma, and led to definitive regulatory approval assessments in the Roche-sponsored AvaGlio study stopped
of TMZ both in the United States, and worldwide. Further at the time of radiological tumor progression, and does not
details are discussed in the next main paragraph. allow for definitive conclusions. Furthermore, toxicity and
Temozolomide has also been extensively investigated in serious adverse events (hypertension, thromboembolic com-
recurrent anaplastic astrocytoma, mixed oligoastrocytoma plications, wound dehiscence and hemorrhage) were all
and oligodendroglioma. The response rate in these grade III more prevalent in the bevacizumab-treated patients.
tumors was in the range of 20–40 %, with 50–70 % of The integrin inhibitor cilengitide has also been investi-
patients likely deriving some benefit [42, 50–53] gated in a large prospective phase III trial in newly diagnosed
patients. Despite promising results in non-comparative phase
II trials, no benefit could be demonstrated in the definitive
Bevacizumab and Anti-angiogenic Therapy trial [59–61].
Unprecedented rapid disappearance of contrast enhance-

ment, reduction in peritumoral edema, and temporary clini- Miscellaneous Agents
cal improvement have been observed with the anti-VEGF
(vascular endothelial growth factor) monoclonal antibody, Various other agents have been investigated and are occa-
bevacizumab and the VEGF receptor tyrosine kinase inhibi- sionally used in the treatment of recurrent glioma.
tor cediranib [54, 55]. The neutralizing antibody bevaci- Randomized trials failed to demonstrate any relevant and
zumab will abrogate the vascular endothelial growth factor A reproducible antitumor activity of the tyrosine kinase inhibi-
(VEGF)-mediated inflammation and restore the blood brain tors imatinib and erlotinib when used as single agents [51,
barrier. This leads to less peritumoral edema and reduces 62]. Irinotecan is occasionally used as single agent [63, 64]
gadolinium-diffusion into the brain. Consequently, radio- or in combination with bevacizumab [54], some activity has
logical tumor extension appears to be lesser, and steroids been reported for cis- and carboplatin [65–67].
can be reduced or completely discontinued. This led to
enthusiastic large-scale use and investigation of anti-angio-
genic and vasculature-modifying agents in malignant glioma. Primary Treatment of Glioblastoma
For cediranib, a randomized phase III trial compared single
agent cediranib vs. the combination of cediranib and lomus- The value of radiotherapy in the treatment of malignant gli-
tine vs. lomustine alone. Although progression-free survival oma (both glioblastoma and anaplastic astrocytoma) was
favored the two experimental arms, no significant difference established in randomized trials almost 30 years ago [68–
in survival could by demonstrated with the addition of cedi- 71]. Compared with supportive care or nitrosourea-based
ranib [27]. Actually, patients receiving lomustine fared best chemotherapy alone, radiotherapy increased median survival
with a median overall survival of 9.8 months (compared to from 4 to 8 months. Subgroup analyses suggested that the
8.0 and 9.4 months in the cediranib alone or in combination addition of chemotherapy to RT might further improve sur-
with lomustine arms, respectively, p = NS). In the United vival for patients with grade 3 and 4 tumors, but none of the
States bevacizumab received regulatory approval for recur- randomized trials demonstrated an unequivocal beneficial
rent glioblastoma, based on two uncontrolled phase II trials role for nitrosourea-based chemotherapy. A meta-analysis
demonstrating a high radiological response rate, a decrease based on published studies, and in the year 2002 [72] and in
284 R. Stupp et al.
PcP prophylaxis
4 weeks
TMZ daily x 42d 5d 5d 5d x 6 cycles
4 weeks
1 6 10 14 18 22 week
RT; focal radiotherapy, 60 Gy over 6 weeks

RT 30 x 2 Gy to tumor volume + 2-3 cm margin
TMZ; temozolomdide (Temodal®, Temodar®)1

during RT: 75 mg/m2 daily (including weekend) for up to 49 days, on an empty stomach2,
PO 1-2 hours before RT, or in the a.m. on days without RT
adjuvant/maintenance: 200 mg/m2 (first cycle 150mg/m2) daily x 5 days, for 6 cycles
PcP; Pneumocystis jiroveci pneumonia prophylaxis

recommended during continuous TMZ chemoradiotherapy
pentamidine inhalations 1x/months 300 mg or trimethroprim/sulfametoxazole forte
3x/week
Alternative: monitoring absolute lymphocyte count, start prophylaxis only when lymphocytes
< 500/mm3, or CD4 < 200/mm3.
1 antiemeticprophylaxis: during chemoradiotherapy with low-dose TMZ usually simpe antiemetic sufficient, e.g.
metoclopramide or loperamide (5HT3 antagonist suggested for the first 2 days)
Adjuvant TMZ therapy: moderately emetogenic, primary prophylaxis with half-dose 5HT3 antagonist
recommended (e.g. ondansetron 4 mg, granisetron 1 mg)
2 empty stomach: fluids and light snack admissible
Fig. 19.1 Treatment scheme for newly diagnosed glioblastoma. From: Mino, Homicsko & Stupp: central nervous system In: M.A. Dicato (ed.),
Side effects of medical cancer therapy, Springer London 2013, doi:10.1007/978-0-85729-787-7_7; used with permission
the year 2002, an analysis of pooled individual data from RT and for 6 months afterward {TMZ/RT → TMZ})
more than 3,000 patients in 12 randomized trials [73] found improved overall survival in patients with glioblastoma [48].
a small but significant survival advantage, favoring chemo- In this phase III trial, 573 patients were randomized in less
therapy in addition to surgery and RT at initial diagnosis. For than 18 months. Patients who received initial treatment with
patients with both grade 3 and 4 tumors, a hazard ratio of RT alone (but may have received TMZ or other chemother-
0.85 for chemotherapy equated to a 5 % improvement in apy at progression) had a median survival of 12 months com-
overall survival from 15 to 20 % at 2 years [73]. pared with 15 months for patients treated with TMZ/RT at
In a German multicenter trial with patient accrual from diagnosis. More importantly, the chances of survival at 2
1994 to 1999 BCNU was replaced by nimustine (ACNU), years were only 11 % in the RT arm compared with 27 % for
another nitrosourea with less pulmonary toxicity [32]. those treated with TMZ/RT [49]. This trial established the
Three hundred seventy-five patients were randomized to current standard of care for patients with glioblastoma and
radiotherapy and either ACNU/cytarabine or ACNU/teni- serves as the benchmark to which newer strategies must be
poside; unfortunately there was no radiotherapy-alone compared (Fig. 19.1). These results compare favorably with
control arm. No difference in survival between the two reported results among patients with small tumors eligible
chemotherapy regimens could be shown; nevertheless the for stereotactic radiosurgery. The good outcome in the
survival outcomes were encouraging with a median sur- control arm also reflects the, albeit modest, activity of the
vival of 16 months and a 2-year survival rate of 27 % TMZ when administered as second-line therapy after tumor
among 302 patients with glioblastoma. More than 90 % of progression.
patients had undergone debulking surgery, and accrual
occurred over 6 years in 15 centers (average of three
patients per year per center). Thus, a strong selection bias Methylguanine Methyltransferase
cannot be excluded.
In 2005, the EORTC and NCIC (National Cancer Institute Methylguanine methyltransferase (MGMT) gene promoter
of Canada Clinical Trials Group) Intergroup trial demon- methylation has been established as the single most impor-
strated beyond doubt that chemotherapy (i.e., TMZ during tant predictive and prognostic marker in malignant glioma
Fig. 19.2 Survival according to MGMT status and treatment alloca- ylated according to randomized treatment allocation. Reprinted with
tion (combined modality vs. RT alone). Left panel: survival of patients permission from Stupp R, Mason WP, van den Bent MJ et al.
with a methylated MGMT promoter if treated with TMZ/RT → TMZ Radiotherapy plus concomitant and adjuvant temozolomide for glio-
(blue) or initial RT only (red). Right panel: survival if MGMT unmeth- blastoma. N Engl J Med 2005; 352: 987–996; used with permission
[74, 75]. Patients whose tumors have a methylated MGMT were randomized after determination of and stratification
gene promoter, resulting in gene silencing, decreased for MGMT status. The prognostic value of MGMT promoter
expression of MGMT, and less effective repair of DNA methylation status was confirmed with a median survival of
damage caused by TMZ are benefitting most from the 21.2 months for methylated tumors, compared to 14 months
addition of chemotherapy [76]. This predictive value has for the unmethylated group [47]. However, neither treat-
been repeatedly confirmed in glioblastoma patients, while ment intensification nor prolongation of the maintenance
in patients with anaplastic glioma, MGMT is a strong chemotherapy for up to 12 cycles appeared to improve
favorable prognostic factor independent of treatment, but outcome, overall median survival was 16 months from
loses in predictive strength [46]. In the pivotal EORTC/ registration (approx. 2–6 weeks after surgery). Ongoing
NCIC trial glioblastoma patients with tumors in which the investigations are either stratifying patients according to
MGMT gene promoter was unmethylated and thus presum- the MGMT status, or devising separate treatment strategies
ably fully expressed MGMT, no apparent benefit was for patients with and without MGMT promoter methylation
derived from the addition of TMZ during and after radio- (see Elderly below).
therapy [76]. Patients with an unmethylated tumor treated
with TMZ/RT had a median survival 13 months (2-year
survival rate of 14 %), compared to 12 months (2-year sur- Treatment of Anaplastic Glioma
vival rate 2 %) when treated initially with RT alone (at
recurrence the great majority of patients received chemo- Table 19.2 shows selected treatment options and potential
therapy) [76]. In patients with an MGMT methylated alternatives for anaplastic glioma.
tumor, median survival was 22 months (2-year survival
46 %) when treated with TMZ/RT → TMZ, compared to 15
months (2-year survival: 23 %) if received RT alone at ini- Anaplastic Astrocytoma
tial diagnosis (Fig. 19.2).
The RTOG0525/EORTC/NCCTG (North Central Cancer Unlike glioblastoma, there are no randomized data providing
Treatment Group) Intergroup trial aimed at overcoming the level I evidence on the role of chemotherapy in the initial
MGMT-mediated resistance by dose intensification and treatment of anaplastic astrocytoma. Some trials, however,
exhaustion of MGMT by continuous chemotherapy expo- have suggested a benefit from adjuvant nitrosourea-based
sure. Almost 1,200 patients were included and 833 patients therapy for patients with grade III tumors [70, 77].
286 R. Stupp et al.
Table 19.2 Selected treatment options and potential alternatives

Treatment Evidence level References Alternatives Evidence level References Remarks
Glioblastoma
TMZ/RT → TMZ × 6 for I [48] ACNU/RT → ACNU + VM26 III [21]
elderly patients
Methylated MGMT: TMZ II [90, 91] Hypofractionated RT II [88, 89] For elderly patients
Unmethylated MGMT: RT
Anaplastic astrocytoma RT I [68, 69, 71] BCNU/RT → BCNU II [33]
TMZ/RT → TMZ V [50] Analogy to GBM
Anaplastic grade III oligoastro/oligodendroglioma with LOH 1p/19a
RT → PCV x 6 II [83, 84] PCV × 4 before RT II [82, 85] Improved survival
RT → TMZ V Replace PCV by TMZ
or
TMZ → RT
TMZ/RT → TMZ V [50] Analogy to GBM
Recurrent glioma
TMZ II [42–45, 53] Procarbazine IV [46]
PCV IV [81]
Bevacizumab III [54–56] Irinotecan* IV [63, 64]
One of the few positive trials was reported by the Anaplastic Oligoastrocytoma
EORTC. Two hundred sixty-nine patients with malignant gli- and Oligodendroglioma
oma were randomized to focal RT vs. RT + dibromodulcitol
and BCNU [77]. The longer survival in the experimental arm Since the year 2000 WHO criteria and greater awareness of
was mainly due to the subgroup of 29 (!) patients with non- this subgroup, oligoastrocytoma and oligodendroglioma are
glioblastoma high-grade tumors. A confirmatory randomized increasingly identified as separate and distinct tumor entities
EORTC trial in grade III tumors failed to demonstrate a [80]. While mixed oligoastrocytomas behave largely like
statistically significant improvement in outcome [78]. A astrocytomas, pure oligodendrogliomas have a more favor-
randomized trial comparing adjuvant PCV with adjuvant PCV able natural history. Specific molecular changes, in particular
plus difluoromethyl ornithine (DFMO), an ornithine decar- the loss of genetic information on chromosomes 1p and 19q,
boxylase inhibitor, has been reported [79]. Patients were ran- have been associated with high sensitivity to chemotherapy
domized at the end of RT. Those who had already progressed and with particularly long survival [81]. Two randomized tri-
or had declining performance status were not included in the als investigated neoadjuvant (before RT) or adjuvant (after
study. Patients with grade 3 gliomas experienced longer statis- completion of RT) PCV chemotherapy in patients with newly
tically nonsignificant survival and time to progression in the diagnosed anaplastic oligoastrocytomas and oligodendroglio-
PCV + DFMO arm [79]. An identical trial in patients with gliomas [82, 83]. Both trials confirmed a more favorable progno-
blastoma was negative [79]. sis for patients with loss of heterozygosity (LOH) of 1p/19q.
In one phase II trial, in patients with recurrent anaplastic While at initial analysis after a median follow-up of more
astrocytoma, TMZ has shown a high objective response rate of than 6 years, no benefit in overall survival was detected [82,
35 %. A randomized trial (RTOG9813) comparing radiother- 83], a subsequent analysis, performed in 2012 after a median
apy plus adjuvant BCNU or CCNU vs. adjuvant TMZ in newly follow-up of 11 years, found a significant survival benefit for
diagnosed patients was closed early due to insufficient accrual. patients with oligodendroglioma harboring a co-deletion of
The German NOA-04 trial compared two treatment strat- chromosomes 1p and 19q [84, 85]. These findings are sup-
egies: patients with anaplastic glioma were randomized to ported by the earlier observation of prolonged progression-
either RT alone, and chemotherapy at recurrence vs. the free survival in this genetically defined subset, suggesting
opposite sequence with a primary endpoint of time to the that prolongation of progression-free survival may be a mean-
second failure. The data suggest that it does not matter which ingful surrogate endpoint. These trials also emphasize that the
strategy is employed first. MGMT promoter methylation was subgroup of patients with 1p/19q co-deletion represent a dis-
again a strong prognostic marker, however it was not predic- tinct pathologic entity and future revisions of the WHO clas-
tive for benefit from chemotherapy when analyzed by time to sification may include molecular characterization as a
first progression [46]. mandatory condition of pathological categorization.
radiotherapy (28 × 1.8 Gy, 50.4 Gy) with best supportive care
How to Select Patients for Chemotherapy? in 84 patients over the age of 70 years (median, 73 years,
range 70–85 years). Overall survival in patients receiving
When considering patients with malignant glioma for a spe- radiotherapy was 6.8 months compared with 4 months with
cific therapy, treatment-related, tumor-related, and patient- best supportive care only (p = 0.002) [88]. Roa et al. reported
related factors need to be considered. One of the most equivalent survival when treating glioblastoma patients over
important deterrents to chemotherapy is poor performance the age of 60 years (mean 72 years) with either standard frac-
status. Patients with poor neurologic function are unlikely to tionated radiotherapy (30 × 2 Gy, 60 Gy in 6 weeks) or with
improve substantially with chemotherapy. Similarly, patients hypofractionated radiotherapy (15 × 2.66 Gy, 40 Gy in 3
with poor general functioning are unlikely to tolerate chemo- weeks) [89]. Median survival durations were 5.1 and
therapy sufficiently well to achieve a benefit, especially when 5.6 months, respectively (p = 0.57). In the EORTC/NCIC
the response of malignant gliomas is often slow or protracted. TMZ/RT → TMZ trial, patients over the age of 50 years had
Tumor size and resectability are prognostic factors [86]. inferior overall survival, but the benefit afforded by TMZ/RT
For large, deep tumors and tumors crossing the midline, only was maintained across all age groups, including patients > 60
biopsy with stereotactic guidance is suggested. Tumors in or 65 years (patients >70 years were not eligible for trial par-
these locations are often accompanied by rapidly progressive ticipation) [49]. Therefore, combined treatment should not
neurologic symptoms, steroid dependence, and deteriorating be withheld based on age alone.
performance status. In all clinical trials, these types of Two randomized trials compared exclusive radiotherapy
patients have very poor outcomes irrespective of treatment with chemotherapy alone [90, 91]. In the 3-arm Nordic trial
with TMZ/RT or other treatment measures. In the EORTC/ 342 patients (median age of 70 years, range 60–88) were
NCIC trial, a trend for longer survival after TMZ/RT could randomly assigned to one of two RT (10 × 3.4 Gy or
also be demonstrated among patients with a tumor biopsy 30 × 2 Gy) schedules or standard dose TMZ monotherapy
only (median survival 9.4 vs. 7.4 months, p = 0.09) (see sup- [90]. The study showed that standard fractionated RT was
plementary appendix to [48] at http://content.nejm.org/cgi/ particularly poorly tolerated in patients older than 70 years,
content/full/352/10/987/DC1). and that single agent TMZ may be a valid alternative to
In our clinical practice, we aim at providing all care in the radiotherapy. Median overall survival was 6.0 months with
outpatient clinic both with chemotherapy and radiotherapy. RT60 (30 × 2 Gy), 7.5 months with RT34 (10 × 3.4 Gy), and
For patients with severely impaired performance status, we 8.3 months with TMZ (5/28 days) [90]. The German NOA-
frequently recommend single-modality therapy or support- 08 trial compared standard fractionated RT (30 × 2 Gy) with
ive care only. Nevertheless, French investigators reported on 1 week on, 1 week off, dose-dense TMZ in 373 elderly
a phase II trial including elderly patients (median age 77 patients (median age 72 years, range 66–84) [91]. Overall
years, range 70–87) with a Karnofsky performance status survival analysis favored RT (median 9.6 vs. 8.6 months,
below 70. Seventy patients were treated with single agent p = 0.033), however exclusive radiotherapy appears to be det-
TMZ (5/28 days). One third of the patients improved their rimental for patients with a methylated MGMT. Taken
performance status to ≥70 % and regained some indepen- together elderly patients with methylated MGMT tumors are
dence and quality of life [87]. The median survival was best treated with TMZ chemotherapy, while unmethylated
5.8 months, while subgroup analyses of patients with meth- fare better when treated with RT alone [91].
ylated or unmethylated MGMT showed median survivals of An ongoing trial coordinated by the NCIC in collabora-
7.2 and 4.4 months, respectively. tion with the EORTC is evaluating combined modality hypo-
fractionated TMZ/RT → TMZ vs. RT (15 × 3 Gy) alone in
patients over the age of 65 years but unfortunately, this trial
Treatment of the Elderly lacks a chemotherapy alone arm.
Although the median age of glioblastoma patients in clinical

trials ranges from 52 to 57 years, the median age in daily RPA Analysis
practice is 62–65 years. Glioblastoma is a disease of the
elderly. The age at diagnosis has been repeatedly shown to Grouping of the patients according to the RTOG recursive
be an important prognostic factor for survival. Based on poor partitioning analysis based on age, performance status, neu-
outcomes for patients over the age of 60 or 70 years, it had rologic function, treatment, and tumor grade allows for some
been suggested that these patients should only be managed comparison between trials and variable distribution of prog-
with supportive care. However, contemporary randomized nostic factors [92–94]. Table 19.3 shows recursive partition-
trials have revisited this attitude. A French trial compared ing analysis of the EORTC/NCIC trial.
288 R. Stupp et al.
Table 19.3 Recursive Standard arm: radiotherapy Experimental arm: combined TMZ/RT
partitioning analysis of the
RPA class Median survival Two-year survival Median survival Two-year survival
EORTC/NCIC trial
3 14.8 (10.9–17.0) 19.6 (6.9–32.3) 21.4 (15.3–N*) 43.4 (28.0–58.9)
4 13.3 (12.0–15.0) 10.9 (5.8–16.0) 16.3 (14.1–18.3) 27.9 (20.5–35.3)
5 9.1 (8.0–11.7) 6.2 (1.1–11.2) 10.3 (8.6–12.4) 16.5 (8.8–24.2)
RPA recursive partitioning analysis, TMZ/RT temozolomide and radiotherapy, N* upper confidence boundary
not reached
From: Mirimanoff RO, Gorlia T, Mason W, et al. Radiotherapy and temozolomide for newly diagnosed glio-
blastoma; recursive partitioning analysis of the EORTC 26981/22981-NCIC CE3 phase III randomized trial.
J Clin Oncol 2006; 24(16):2563–2569
Molecular Profiling Although temozolomide is considered a particularly

well-tolerated chemotherapy, severe myelosuppression and
Molecular markers play an increasing role in understanding, particularly thrombocytopenia, may occur in up to 5 % of
prognostication, prediction of response and treatment target patients. Continuous low-dose TMZ (with or without
identification of glioma [95]. The chromosomal co-deletion concomitant use of corticosteroids) will frequently induce
1p/19q is mediated by an unbalanced translocation, and is profound lymphocytopenia [107–109]. Counts of CD4 posi-
characteristic for oligodendroglial tumors and associated tive lymphocytes may fall well below 200/mm3, the thresh-
with benefit from chemotherapy [96]; EGFRviii mutation old recommended for Pneumocystis jiroveci (ex carinii)
and EGFR overexpression are associated with inferior prog- prophylaxis in HIV-positive patients [110]. By extrapolation,
nosis [8, 97–99], while IDH mutations, more commonly close monitoring of absolute lymphocyte and CD4 counts, or
observed in lower grade glioma, secondary glioblastoma and prophylaxis with pentamidine inhalations or trimethoprim-
younger patients, with a longer survival [100]. The role of sulfamethoxazole is advised for those on continuous
MGMT gene promoter methylation as a prognostic and pre- low-dose TMZ.
dictive marker has been discussed above. Gene expression Quality of life using the EORTC quality of life question-
signatures have allowed dividing tumors into distinct molec- naire QLQ-30 and a specific brain cancer module (BM-20)
ular subclasses: proneural, neural, classical, and mesenchy- has been evaluated in recent TMZ trials [111–113]. In a ran-
mal [4, 101]. Lately, analyses of the tumor genome domized phase II trial, TMZ administration after first line
methylation allow to further subdivide these subclasses [3]. radiotherapy (±chemotherapy) was compared with procarba-
The CpG island methylator phenotype (CIMP) correlates zine [43]. Quality of life was superior in several domains in
largely with a subgroup of tumors of the proneural subclass, patients treated with TMZ compared with procarbazine
enriched for IDH mutated tumors [3]. These few examples [111]. In the EORTC/NCIC trial, no detrimental effect on
illustrate the complexity of correlations and interactions of quality of life was observed in the longer surviving TMZ/RT
genes, mutations, epigenetic modifications, and tumor pro- patient group [113]. Except for a slight increase in fatigue
files. Individual molecular tumor characteristics allow the (6 % with RT vs. 9 % with TMZ/RT), there was no treatment
selection of an optimized treatment strategy in some dependent difference in non-hematologic toxicities [48].
instances. Personalized cancer therapy is on the horizon, also
for glioma [3, 76, 101–103].
Chemotherapy or Stereotactic Radiosurgery
Chemotherapy Toxicity and Quality of Life Stereotactic radiosurgery, despite promising reports on
selected patients in phase II trials and in retrospective series,
Systemic administration of chemotherapy is inherently asso- failed to demonstrate any prolongation of survival, qual-
ciated with the risk of nontarget organ toxicity. All cyto- ity of life, or local control and was the object in 2005 of a
toxic agents currently used for the treatment of malignant systematic review by the American Society for Therapeutic
glioma may induce moderate or severe myelosuppression. Radiology and Oncology (ASTRO) [114]. Apparent
Carmustine may also cause severe interstitial pneumonitis in improvements in survival in some non-controlled trials may
20–30 % of patients [104–106]. A major limitation of the simply reflect patient selection [25]. To date two randomized
PCV chemotherapy regimen is the cumulative hematotoxic- trials were started, but only one was completed. A stereotac-
ity, which frequently prohibits delivering the planned num- tic boost of 15–24 Gy administered to tumors of ≤4 cm in
ber of cycles and dose intensity. size before standard external beam radiotherapy showed no
Table 19.4 Advantages of chemotherapy and SRS

Chemotherapy SRS Remarks
Level of evidence I (glioblastoma) IV
Tumor size Any Tumors <3–4 cm Chemotherapy will treat microscopic
infiltrative disease, whereas SRS will only
eliminate visible tumor
Karnofsky performance status >60 % >80 % SRS requires anesthesia
Substitute for surgery No Yes If tissue previously available for histologic
diagnosis
Toxicity SRS risk of late radionecrosis
Systemic Manageable Absent
Local None Moderate
Drug interaction Possible None
Edema induction None Possible
Costs High Moderate
Treatment duration Long Short
Availability Large scale, anytime Limited, restrictive
Patient selection Minor, mainly Highly selected, applicable to only
performance status 10–25 % of high-grade glioma patients
improvement in any of the outcome parameters [115]. The diotherapy and adjuvant chemotherapy of newly diagnosed
other randomized trial evaluating fractionated stereotactic glioblastoma [48, 49]. In this trial with 85 centers, many
boost (5 × 4 Gy) administered after the end of external beam smaller hospitals participated, confirming the applicability
radiotherapy was prematurely discontinued due to insuffi- of this treatment to everyday practice. The subsequent large
cient patient accrual [116]. and pragmatic RTOG0525/EORTC/NCCTG Intergroup
Why increasing doses of radiation do not [117]? study with more than 250 participating RT-departments (in
Glioblastoma is an infiltrating tumor affecting the whole the majority North American hospitals) included 1,173
brain, and the contrast enhancement and tumor extension patients and confirmed the results of previous trials [47].
detected on MRI is “only the tip of the iceberg.” Stereotactic Stereotactic radiosurgery has the ability to deliver with
radiosurgery can only be applied to tumors with a maximum utmost precision high radiation doses, either as a single frac-
diameter of 3–4 cm, while malignant gliomas are large and tion or as fractionated radiotherapy. However, in the case of
diffuse. Tumor cells outside of the field of irradiation can ill-defined targets, high precision is of limited value. In these
serve as sanctuaries for tumor regrowth and maintenance, cases novel neuro-imaging techniques like tumor metabolic
this will inevitably lead to failure of all exclusively local treat- evaluation with positron-emission tomography (PET) using
ment approaches. Before the era of modern radiation therapy, amino-acid tracer, or improved anatomical visualization
surgery was the sole treatment modality and heroic lobec- using diffusion tensor tractography can be useful.
tomies and hemispherectomies were performed in vain. The different treatment approaches of systemic and local
Even after extensive resections tumor recurred at distant treatment are not necessarily mutually exclusive. The combi-
brain sites. The dogma of regional tumor recurrence popular- nation of stereotactic radiosurgery with TMZ in the frame of
ized by Hochberg and Pruitt [118] may need to be revisited. the standard chemoradiotherapy is likely feasible and should
As patients now live longer thanks to combined chemoradio- be properly studied in prospective randomized trials. The
therapy and other truly effective therapies, distant recurrences repeated reports on feasibility of this approach in uncon-
are observed at increasing frequency. Table 19.4 shows the trolled phase II studies, and frequently with inappropriate
advantages of chemotherapy and SRS. historical comparators are misleading and a waste of the lim-
Stereotactic radiosurgery not only requires equipment ited resources [120]. In vitro, temozolomide has been shown
and expertise, but also limited to a minority of patients with to inhibit tumor cell migration. Similarly, the combination of
malignant glioma (10–25 %) (reviewed in [119]). Only temozolomide and radiotherapy together with carmustine-
patients with adequate performance status and with tumors impregnated wafers (Gliadel; MCI Pharma, Baltimore, MD)
distant from vital structures are considered for radiosurgery. might improve local control while eradicating infiltrating
In contrast, TMZ chemotherapy can be initiated in 90 % of tumor cells [121].
patients with malignant glioma without substantial toxicity. In select patients with small and circumscribed recurrent
Furthermore, level I evidence is available for the chemora- high-grade glioma, re-irradiation is increasingly used.
290 R. Stupp et al.
Median survival rates of 8–11 months have been reported, 18. Weller M, Stupp R, Wick W. Epilepsy meets cancer: when, why,
but the absence of long-term survivors with grade IV tumors and what to do about it? Lancet Oncol. 2012;13:e375–82.
19. Macdonald DR, Cascino TL, Schold Jr SC, Cairncross
indicate that this is not a curative strategy [122, 123]. JG. Response criteria for phase II studies of supratentorial malig-
Ongoing trials investigate the combination of stereotactic re- nant glioma. J Clin Oncol. 1990;8:1277–80.
irradiation with bevacizumab. The marked anti-inflammatory 20. Therasse P, Arbuck SG, Eisenhauer EA, et al. New guidelines to
effect of bevacizumab may allow larger RT-fields, better tol- evaluate the response to treatment in solid tumors. European
Organization for Research and Treatment of Cancer, National
erance, and less toxicity [124]. Cancer Institute of the United States, National Cancer Institute of
Canada. J Natl Cancer Inst. 2000;92:205–16.
21. Wen PY, Macdonald DR, Reardon DA, et al. Updated response
References assessment criteria for high-grade gliomas: response assessment in
neuro-oncology working group. J Clin Oncol. 2010;28: 1963–72.
1. Stupp R, Hegi ME, Gilbert MR, Chakravarti A. Chemoradiotherapy 22. Leimgruber A, Ostermann S, Yeon E, et al. Perfusion and diffu-
in malignant glioma: standard of care and future directions. J Clin sion MRI of glioblastoma progression in a 4 years prospective
Oncol. 2007;25:4127–36. temozolomide clinical trial. Int J Radiat Oncol Biol Phys.
2. Balis FM, Blaney SM, McCully CL, et al. Methotrexate distribu- 2006;64:869–75.
tion within the subarachnoid space after intraventricular and intra- 23. Moffat BA, Chenevert TL, Lawrence TS, et al. Functional diffu-
venous administration. Cancer Chemother Pharmacol. 2000;45: sion map: a noninvasive MRI biomarker for early stratification of
259–64. clinical brain tumor response. Proc Natl Acad Sci U S A.
2005;102:5524–9.
3. Noushmehr H, Weisenberger DJ, Diefes K, et al. Identification of
24. Wong ET, Hess KR, Gleason MJ, et al. Outcomes and prognostic
a CpG island methylator phenotype that defines a distinct sub-
factors in recurrent glioma patients enrolled onto phase II clinical
group of glioma. Cancer Cell. 2010;17:510–22.
trials. J Clin Oncol. 1999;17:2572.
4. Verhaak RG, Hoadley KA, Purdom E, et al. Integrated genomic
25. Irish WD, Macdonald DR, Cairncross JG. Measuring bias in
analysis identifies clinically relevant subtypes of glioblastoma
uncontrolled brain tumor trials–to randomize or not to randomize?
characterized by abnormalities in PDGFRA, IDH1, EGFR, and
Can J Neurol Sci. 1997;24:307–12.
NF1. Cancer Cell. 2010;17:98–110.
26. Wick W, Puduvalli VK, Chamberlain MC, et al. Phase III study of
5. Sturm D, Witt H, Hovestadt V, et al. Hotspot mutations in H3F3A
enzastaurin compared with lomustine in the treatment of recurrent
and IDH1 define distinct epigenetic and biological subgroups of
intracranial glioblastoma. J Clin Oncol. 2010;28:1168–74.
glioblastoma. Cancer Cell. 2012;22:425–37.
27. Batchelor TT, Mulholland P, Neyns B, et al. Phase III randomized
6. Gerlinger M, Rowan AJ, Horswell S, et al. Intratumor heterogene- trial comparing the efficacy of cediranib as monotherapy, and in
ity and branched evolution revealed by multiregion sequencing. combination with lomustine, versus lomustine alone in patients
N Engl J Med. 2012;366:883–92. with recurrent glioblastoma. J Clin Oncol. 2013;31:3212–8.
7. Fisher R, Pusztai L, Swanton C. Cancer heterogeneity: implica- 28. Frenay M, Giroux B, Khoury S, et al. Phase II study of fotemus-
tions for targeted therapeutics. Br J Cancer. 2013;108(3):479–85. tine in recurrent supratentorial malignant gliomas. Eur J Cancer.
8. Sampson JH, Heimberger AB, Archer GE, et al. Immunologic 1991;27:852–6.
escape after prolonged progression-free survival with epidermal 29. Takakura K, Abe H, Tanaka R, et al. Effects of ACNU and radio-
growth factor receptor variant III peptide vaccination in patients therapy on malignant glioma. J Neurosurg. 1986;64:53–7.
with newly diagnosed glioblastoma. J Clin Oncol. 2010;28: 30. Yoshida J, Kajita Y, Wakabayashi T, Sugita K. Long-term follow-
4722–9. up results of 175 patients with malignant glioma: importance of
9. Huse JT, Holland EC. Targeting brain cancer: advances in the radical tumour resection and postoperative adjuvant therapy with
molecular pathology of malignant glioma and medulloblastoma. interferon, ACNU and radiation. Acta Neurochir. 1994;127:55–9.
Nat Rev Cancer. 2010;10:319–31. 31. Kochii M, Kitamura I, Goto T, et al. Randomized comparison of
10. Roesch A, Fukunaga-Kalabis M, Schmidt EC, et al. A temporarily intra-arterial versus intravenous infusion of ACNU for newly
distinct subpopulation of slow-cycling melanoma cells is required diagnosed patients with glioblastoma. J Neuro Oncol. 2000;49:
for continuous tumor growth. Cell. 2010;141:583–94. 63–70.
11. Gupta PB, Onder TT, Jiang G, et al. Identification of selective 32. Weller M, Müller B, Koch R, et al. Neuro-Oncology Working
inhibitors of cancer stem cells by high-throughput screening. Cell. Group 01 trial of nimustine plus teniposide versus nimustine plus
2009;138:645–59. cytarabine chemotherapy in addition to involved-field radiother-
12. Hanahan D, Coussens LM. Accessories to the crime: functions of apy in the first-line treatment of malignant glioma. J Clin Oncol.
cells recruited to the tumor microenvironment. Cancer Cell. 2003;21:3276–84.
2012;21:309–22. 33. Green SB, Byar DP, Walker MD, et al. Comparisons of carmus-
13. Straussman R, Morikawa T, Shee K, et al. Tumour micro- tine, procarbazine, and high-dose methylprednisolone as additions
environment elicits innate resistance to RAF inhibitors through to surgery and radiotherapy for the treatment of malignant glioma.
HGF secretion. Nature. 2012;487:500–4. Cancer Treat Rep. 1983;67:121–32.
14. Dichter MA, Brodie MJ. New antiepileptic drugs. N Engl J Med. 34. Rodriguez LA, Prados M, Silver P, Levin VA. Reevaluation of
1996;334:1583–90. procarbazine for the treatment of recurrent malignant central ner-
15. Kanner AM, Balabanov AJ. The use of monotherapy in patients vous system tumors. Cancer. 1989;64:2420–3.
with epilepsy: an appraisal of the new antiepileptic drugs. Curr 35. Newton HB, Junck L, Bromberg J, et al. Procarbazine chemother-
Neurol Neurosci Rep. 2005;5:322–8. apy in the treatment of recurrent malignant astrocytomas after
16. Vecht CJ, Wagner GL, Wilms EB. Interactions between antiepilep- radiation and nitrosourea failure. Neurology. 1990;40:1743–6.
tic and chemotherapeutic drugs. Lancet Neurol. 2003;2: 404–9. 36. Levin VA, Silver P, Hannigan J, et al. Superiority of post-
17. Weller M, Gorlia T, Cairncross JG, et al. Prolonged survival with radiotherapy adjuvant chemotherapy with CCNU, procarbazine,
valproic acid use in the EORTC/NCIC temozolomide trial for and vincristine (PCV) over BCNU for anaplastic gliomas: NCOG
glioblastoma. Neurology. 2011;77:1156–64. 6G61 final report. Int J Radiat Oncol Biol Phys. 1990;18:321–4.
37. Prados MD, Scott C, Curran Jr WJ, et al. Procarbazine, lomustine, 55. Kreisl TN, Kim L, Moore K, et al. Phase II trial of single-agent
and vincristine (PCV) chemotherapy for anaplastic astrocytoma: bevacizumab followed by bevacizumab plus irinotecan at tumor
A retrospective review of radiation therapy oncology group proto- progression in recurrent glioblastoma. J Clin Oncol. 2009;27:
cols comparing survival with carmustine or PCV adjuvant chemo- 740–5.
therapy. J Clin Oncol. 1999;17:3389–95. 56. Wick W, Weller M, van den Bent M, Stupp R. Bevacizumab and
38. Medical Research Council Brain Tumor Working Party. recurrent malignant gliomas: a European perspective. J Clin
Randomized trial of procarbazine, lomustine and vincristine in the Oncol. 2010; 28: e188–189; author reply e190–182
adjuvant treatment of high-grade astrocytoma: a Medical Research 57. Chinot OL, Wick W, Mason W, et al. Bevacizumab plus
Council trial. J Clin Oncol. 2001;19:509–18. radiotherapy-temozolomide for newly diagnosed glioblastoma.
39. Stupp R, Gander M, Leyvraz S, Newlands E. Current and future N Engl J Med. 2014;370:709–22.
developments in the use of temozolomide in the treatment of brain 58. Gilbert MR, Dignam JJ, Armstrong TS, et al. A randomized trial
tumors. Lancet Oncol. 2001;2:552–60. of bevacizumab for newly diagnosed glioblastoma. N Engl J Med.
40. Brada M, Judson I, Beale P, et al. Phase I dose-escalation and 2014;370:699–708.
pharmacokinetic study of temozolomide (SCH 52365) for refrac- 59. Reardon DA, Fink KL, Mikkelsen T, et al. Randomized phase II
tory or relapsing malignancies. Br J Cancer. 1999;81:1022–30. study of cilengitide, an integrin-targeting arginine-glycine-
41. Ostermann S, Csajka C, Buclin T, et al. Plasma and cerebrospinal aspartic acid peptide, in recurrent glioblastoma multiforme. J Clin
fluid population pharmacokinetics of temozolomide in malignant Oncol. 2008;26:5610–7.
glioma patients. Clin Cancer Res. 2004;10:3728–36. 60. Stupp R, Hegi ME, Neyns B, et al. Phase I/IIa study of cilengitide
42. Yung WK, Prados MD, Yaya-Tur R, et al. Multicenter phase II and temozolomide with concomitant radiotherapy followed by
trial of temozolomide in patients with anaplastic astrocytoma or cilengitide and temozolomide maintenance therapy in patients
anaplastic oligoastrocytoma at first relapse. Temodal Brain Tumor with newly diagnosed glioblastoma. J Clin Oncol. 2010;28:
Group. J Clin Oncol. 1999;17:2762–71. 2712–8.
43. Yung WK, Albright RE, Olson J, et al. A phase II study of temo- 61. Stupp R, Hegi ME, Gorlia T et al. Cilengitide combined with stan-
zolomide vs. procarbazine in patients with glioblastoma multi- dard treatment for patients with newly diagnosed glioblastoma
forme at first relapse. Br J Cancer. 2000;83:588–93. and methylated O6-methylguanine-DNA methyltransferase
44. Brada M, Hoang-Xuang K, Rampling R, et al. Multicenter phase (MGMT) gene promoter: Key results of the multicenter, random-
II trial of temozolomide in patients with glioblastoma multiforme ized, open-label, controlled, phase III CENTRIC study. Lancet
at first relapse. Ann Oncol. 2001;12:259–66. Oncol 2014, in press
45. Brada M, Stenning S, Gabe R, et al. Temozolomide versus procar- 62. Dresemann G, Weller M, Rosenthal MA, et al. Imatinib in combi-
bazine, lomustine, and vincristine in recurrent high-grade glioma. nation with hydroxyurea versus hydroxyurea alone as oral therapy
J Clin Oncol. 2010;28:4601–8. in patients with progressive pretreated glioblastoma resistant to
46. Wick W, Hartmann C, Engel C, et al. NOA-04 randomized phase standard dose temozolomide. J Neurooncol. 2010;96:393–402.
III trial of sequential radiochemotherapy of anaplastic glioma with 63. Friedman HS, Petros WP, Friedman AH, et al. Irinotecan therapy
procarbazine, lomustine, and vincristine or temozolomide. J Clin in adults with recurrent or progressive malignant glioma. J Clin
Oncol. 2009;27:5874–80. Oncol. 1999;17:1516–25.
47. Gilbert MR, Wang M, Aldape KD, et al. Dose-dense temozolo- 64. Raymond E, Fabbro M, Boige V, et al. Multicentre phase II study
mide for newly diagnosed glioblastoma: a randomized phase III and pharmacokinetic analysis of irinotecan in chemotherapy-
clinical trial. J Clin Oncol. 2013;10:4085–91. naive patients with glioblastoma. Ann Oncol. 2003;14:603–14.
48. Stupp R, Mason WP, van den Bent MJ, et al. Radiotherapy plus 65. Grossman SA, O’Neill A, Grunnet M, et al. Phase III study com-
concomitant and adjuvant temozolomide for glioblastoma. N Engl paring three cycles of infusional carmustine and cisplatin followed
J Med. 2005;352:987–96. by radiation therapy with radiation therapy and concurrent car-
49. Stupp R, Hegi ME, Mason WP, et al. Effects of radiotherapy with mustine in patients with newly diagnosed supratentorial glioblas-
concomitant and adjuvant temozolomide versus radiotherapy toma multiforme: Eastern Cooperative Oncology Group Trial
alone on survival in glioblastoma in a randomised phase III study: 2394. J Clin Oncol. 2003;21:1485–91.
5-year analysis of the EORTC-NCIC trial. Lancet Oncol. 2009;10: 66. Brandes AA, Basso U, Reni M, et al. First-line chemotherapy with
459–66. cisplatin plus fractionated temozolomide in recurrent glioblastoma
50. Stupp R, van den Bent MJ, Hegi ME. Optimal role of temozolo- multiforme: a phase II study of the Gruppo Italiano Cooperativo di
mide in the treatment of malignant gliomas. Curr Neurol Neurosci Neuro-Oncologia. J Clin Oncol. 2004;22:1598–604.
Rep. 2005;5:198–206. 67. Bafaloukos D, Tsoutsos D, Kalofonos H, et al. Temozolomide and
51. van den Bent M, Brandes A, Rampling R, et al. Randomized phase cisplatin versus temozolomide in patients with advanced mela-
II trial of erlotinib versus temozolomide or carmustine in recurrent noma: a randomized phase II study of the Hellenic Cooperative
glioblastoma. Report from EORTC Brain Tumor Group study Oncology Group. Ann Oncol. 2005;16:950–7.
26034. J Clin Oncol. 2009;27:1268–74. 68. Walker MD, Alexander Jr E, Hunt WE, et al. Evaluation of BCNU
52. van den Bent MJ, Chinot O, Boogerd W, et al. Second-line chemo- and/or radiotherapy in the treatment of anaplastic gliomas.
therapy with temozolomide in recurrent oligodendroglioma after A cooperative clinical trial. J Neurosurg. 1978;49:333–43.
PCV (procarbazine, lomustine and vincristine) chemotherapy: 69. Walker MD, Green SB, Byar DP, et al. Randomized comparisons
EORTC Brain Tumor Group phase II study 26972. Ann Oncol. of radiotherapy and nitrosoureas for the treatment of malignant
2003;14:599–602. glioma after surgery. N Engl J Med. 1980;303:1323–9.
53. van den Bent MJ, Taphoorn MJ, Brandes AA, et al. Phase II study 70. Kristiansen K, Hagen S, Kollevold T, et al. Combined modality
of first-line chemotherapy with temozolomide in recurrent oligo- therapy of operated astrocytomas grade III and IV. Confirmation of
dendroglial tumors: the European Organization for Research and the value of postoperative irradiation and lack of potentiation of
Treatment of Cancer Brain Tumor Group Study 26971. J Clin bleomycin on survival time: a prospective multicenter trial of the
Oncol. 2003;21:2525–8. Scandinavian Glioblastoma Study Group. Cancer. 1981;47:649–52.
54. Friedman HS, Prados MD, Wen PY, et al. Bevacizumab alone and 71. Walker MD, Strike TA, Sheline GE. An analysis of dose-effect
in combination with irinotecan in recurrent glioblastoma. J Clin relationship in the radiotherapy of malignant gliomas. Int J Radiat
Oncol. 2009;27:4733–40. Oncol Biol Phys. 1979;5:1725–31.
292 R. Stupp et al.
72. Fine HA, Dear KB, Loeffler JS, et al. Meta-analysis of radiation 90. Malmstrom A, Gronberg BH, Marosi C, et al. Temozolomide ver-
therapy with and without adjuvant chemotherapy for malignant sus standard 6-week radiotherapy versus hypofractionated radio-
gliomas in adults. Cancer. 1993;71:2585–97. therapy in patients older than 60 years with glioblastoma: the
73. Stewart LA. Chemotherapy in adult high-grade glioma: a system- Nordic randomised, phase 3 trial. Lancet Oncol. 2012;13: 916–26.
atic review and meta- analysis of individual patient data from 12 91. Wick W, Platten M, Meisner C, et al. Temozolomide chemother-
randomised trials. Lancet. 2002;359:1011–8. apy alone versus radiotherapy alone for malignant astrocytoma in
74. Weller M, Stupp R, Hegi ME, et al. Personalized care in neuro- the elderly: the NOA-08 randomised, phase 3 trial. Lancet Oncol.
oncology coming of age: why we need MGMT and 1p/19q testing 2012;13:707–15.
for malignant glioma patients in clinical practice. Neuro Oncol. 92. Curran Jr WJ, Scott CB, Horton J, et al. Recursive partitioning
2012;14 Suppl 4:iv100–8. analysis of prognostic factors in three Radiation Therapy Oncology
75. Weller M, Stupp R, Reifenberger G, et al. MGMT promoter meth- Group malignant glioma trials. J Natl Cancer Inst. 1993;85:
ylation in malignant gliomas: ready for personalized medicine? 704–10.
Nat Rev Neurol. 2010;6:39–51. 93. Scott CB, Scarantino C, Urtasun R, et al. Validation and predictive
76. Hegi ME, Diserens AC, Gorlia T, et al. MGMT gene silencing and power of Radiation Therapy Oncology Group (RTOG) recursive
benefit from temozolomide in glioblastoma. N Engl J Med. partitioning analysis classes for malignant glioma patients: a
2005;352:997–1003. report using RTOG 90-06. Int J Radiat Oncol Biol Phys. 1998;40:
77. Hildebrand J, Sahmoud T, Mignolet F, et al. Adjuvant therapy 51–5.
with dibromodulcitol and BCNU increases survival of adults with 94. Mirimanoff RO, Gorlia T, Mason W, et al. Radiotherapy and
malignant gliomas. EORTC Brain Tumor Group. Neurology. temozolomide for newly diagnosed glioblastoma: recursive parti-
1994;44:1479–83. tioning analysis of the EORTC 26981/22981-NCIC CE3 phase III
78. Hildebrand J, Gorlia T, Kros JM, et al. Adjuvant dibromodulcitol randomized trial. J Clin Oncol. 2006;24:2563–9.
and BCNU chemotherapy in anaplastic astrocytoma: results of a 95. Weller M, Pfister SM, Wick W, et al. Molecular neuro-oncology in
randomised European Organisation for Research and Treatment of clinical practice: a new horizon. Lancet Oncol. 2013;14:e370–9.
Cancer phase III study (EORTC study 26882). Eur J Cancer. 96. Jenkins RB, Blair H, Ballman KV, et al. A t(1;19)(q10;p10) medi-
2008;44:1210–6. ates the combined deletions of 1p and 19q and predicts a better
79. Levin VA, Hess KR, Choucair A, et al. Phase III Randomized prognosis of patients with oligodendroglioma. Cancer Res.
Study of Postradiotherapy Chemotherapy with Combination 2006;66:9852–61.
alpha-Difluoromethylornithine-PCV versus PCV for Anaplastic 97. Brandes AA, Franceschi E, Tosoni A, et al. Epidermal growth fac-
Gliomas. Clin Cancer Res. 2003;9:981–90. tor receptor inhibitors in neuro-oncology: hopes and disappoint-
80. Kleihues P, Cavenee WK. Pathology and Genetics of tumours of ments. Clin Cancer Res. 2008;14:957–60.
the nervous system. Lyon: IARC Press; 2000. 98. Hegi ME, Rajakannu P, Weller M. Epidermal growth factor recep-
81. Cairncross JG, Ueki K, Zlatescu MC, et al. Specific genetic tor: a re-emerging target in glioblastoma. Curr Opin Neurol.
predictors of chemotherapeutic response and survival in patients 2012;25(6):774–9.
with anaplastic oligodendrogliomas. J Natl Cancer Inst. 1998;90: 99. Murat A, Migliavacca E, Gorlia T, et al. Stem cell-related “self-
1473–9. renewal” signature and high epidermal growth factor receptor
82. Cairncross G, Berkey B, Shaw E, et al. Phase III Trial of expression associated with resistance to concomitant chemoradio-
Chemotherapy Plus Radiotherapy Compared With Radiotherapy therapy in glioblastoma. J Clin Oncol. 2008;26:3015–24.
Alone for Pure and Mixed Anaplastic Oligodendroglioma: 100. Yan H, Parsons DW, Jin G, et al. IDH1 and IDH2 mutations in
Intergroup Radiation Therapy Oncology Group Trial 9402. J Clin gliomas. N Engl J Med. 2009;360:765–73.
Oncol. 2006;24:2707–14. 101. Phillips HS, Kharbanda S, Chen R, et al. Molecular subclasses of
83. van den Bent MJ, Carpentier AF, Brandes AA, et al. Adjuvant high-grade glioma predict prognosis, delineate a pattern of disease
procarbazine, lomustine, and vincristine improves progression- progression, and resemble stages in neurogenesis. Cancer Cell.
free survival but Not overall survival in newly diagnosed anaplas- 2006;9:157–73.
tic oligodendrogliomas and oligoastrocytomas: a randomized 102. Hegi ME, Diserens AC, Godard S, et al. Clinical trial substantiates
European organisation for research and treatment of cancer phase the predictive value of O-6-methylguanine-DNA methyltransfer-
III trial. J Clin Oncol. 2006;24:2715–22. ase promoter methylation in glioblastoma patients treated with
84. van den Bent MJ, Brandes AA, Taphoorn MJ, et al. Adjuvant procar- temozolomide. Clin Cancer Res. 2004;10:1871–4.
bazine, lomustine, and vincristine chemotherapy in newly diagnosed 103. Godard S, Getz G, Delorenzi M, et al. Classification of human
anaplastic oligodendroglioma: long-term follow-Up of EORTC astrocytic gliomas on the basis of gene expression: a correlated
brain tumor group study 26951. J Clin Oncol. 2013;31: 344–50. group of genes with angiogenic activity emerges as a strong pre-
85. Cairncross G, Wang M, Shaw E, et al. Phase III trial of chemora- dictor of subtypes. Cancer Res. 2003;63:6613–25.
diotherapy for anaplastic oligodendroglioma: long-term results of 104. O’Driscoll BR, Hasleton PS, Taylor PM, et al. Active lung fibrosis
RTOG 9402. J Clin Oncol. 2013;31:337–43. up to 17 years after chemotherapy with carmustine (BCNU) in
86. Lacroix M, Abi-Said D, Fourney DR, et al. A multivariate analysis childhood. N Engl J Med. 1990;323:378–82.
of 416 patients with glioblastoma multiforme: prognosis, extent of 105. Kreisman H, Wolkove N. Pulmonary toxicity of antineoplastic
resection, and survival. J Neurosurg. 2001;95:190–8. therapy. Semin Oncol. 1992;19:508–20.
87. Gallego Perez-Larraya J, Ducray F, Chinot O, et al. Temozolomide 106. Schmitz N, Diehl V. Carmustine and the lungs. Lancet. 1997;349:
in elderly patients with newly diagnosed glioblastoma and poor 1712–3.
performance status: an ANOCEF phase II trial. J Clin Oncol. 107. Brock CS, Newlands ES, Wedge SR, et al. Phase I trial of temo-
2011;29:3050–5. zolomide using an extended continuous oral schedule. Cancer
88. Keime-Guibert F, Chinot O, Taillandier L, et al. Radiotherapy for Res. 1998;58:4363–7.
glioblastoma in the elderly. N Engl J Med. 2007;356:1527–35. 108. Su YB, Sohn S, Krown SE, et al. Selective CD4+ lymphopenia in
89. Roa W, Brasher PM, Bauman G, et al. Abbreviated course of melanoma patients treated with temozolomide: a toxicity with
radiation therapy in older patients with glioblastoma multiforme: therapeutic implications. J Clin Oncol. 2004;22:610–6.
a prospective randomized clinical trial. J Clin Oncol. 2004;22: 109. Stupp R, Dietrich P, Ostermann Kraljevic S, et al. Promising
1583–8. survival for patients with newly diagnosed glioblastoma multiforme
treated with concomitant radiation plus temozolomide followed by a randomized phase III study. A joint study of the EORTC
adjuvant temozolomide. J Clin Oncol. 2002;20:1375–82. (22972) and the MRC (BR10). Front Radiat Ther Oncol. 1999;33:
110. Kovacs JA, Masur H. Prophylaxis against opportunistic infections 241–3.
in patients with human immunodeficiency virus infection. N Engl 117. Stupp R, Weber DC. The role of radio- and chemotherapy in glio-
J Med. 2000;342:1416–29. blastoma. Onkologie. 2005;28:315–7.
111. Osoba D, Brada M, Yung W, Prados M. Health-related quality of 118. Hochberg FH, Pruitt A. Assumptions in the radiotherapy of glio-
life in patients treated with temozolomide versus procarbazine for blastoma. Neurology. 1980;30:907–11.
recurrent glioblastoma multiforme. J Clin Oncol. 2000;18: 119. Romanelli P, Conti A, Pontoriero A, et al. Role of stereotactic
1481–91. radiosurgery and fractionated stereotactic radiotherapy for the
112. Osoba D, Brada M, Yung WK, Prados MD. Health-related quality treatment of recurrent glioblastoma multiforme. Neurosurg Focus.
of life in patients with anaplastic astrocytoma during treatment 2009;27:E8.
with temozolomide. Eur J Cancer. 2000;36:1788–95. 120. Einstein DB, Wessels B, Bangert B, et al. Phase II trial of radio-
113. Taphoorn MJ, Stupp R, Coens C, et al. Health-related quality of surgery to magnetic resonance spectroscopy-defined high-risk
life in patients with glioblastoma: a randomised controlled trial. tumor volumes in patients with glioblastoma multiforme. Int J
Lancet Oncol. 2005;6:937–44. Radiat Oncol Biol Phys. 2012;84:668–74.
114. Tsao MN, Mehta MP, Whelan TJ, et al. The American Societey 121. Stupp R, Hegi ME. Treatment of brain tumors (author reply).
for Therapeutic Radiology and Oncology (ASTRO) evidence- N Engl J Med. 2005;352:2350–3.
based review of the role of radiosurgery for malignant glioma. Int 122. Fogh SE, Andrews DW, Glass J, et al. Hypofractionated stereotac-
J Radiat Oncol Biol Phys. 2005;63:47–55. tic radiation therapy: an effective therapy for recurrent high-grade
115. Souhami L, Seiferheld W, Brachman D, et al. Randomized com- gliomas. J Clin Oncol. 2010;28:3048–53.
parison of stereotactic radiosurgery followed by conventional 123. Koga T, Maruyama K, Tanaka M, et al. Extended field stereotactic
radiotherapy with carmustine to conventional radiotherapy with radiosurgery for recurrent glioblastoma. Cancer. 2012;118:
carmustine for patients with glioblastoma multiforme: report of 4193–200.
Radiation Therapy Oncology Group 93-05 protocol. Int J Radiat 124. Shapiro LQ, Beal K, Goenka A, et al. Patterns of failure after con-
Oncol Biol Phys. 2004;60:853–60. current bevacizumab and hypofractionated stereotactic radiation
116. Brada M, Baumert B. Focal fractionated conformal stereotactic therapy for recurrent high-grade glioma. Int J Radiat Oncol Biol
boost following conventional radiotherapy of high-grade gliomas: Phys. 2012;85(3):636–42.
Meningioma
20
Alessandra Gorgulho, Carlos A. Mattozo,
and Antonio A.F. De Salles
Meningiomas are considered indolent tumors, but the nat-

Introduction ural history of any particular case is unpredictable. Although
three well-defined groups of meningiomas are currently used
Meningiomas are generally benign lesions that account for (named grade I, benign; grade II, atypical; and grade III,
15–20 % of primary brain tumors, affect predominately malignant), there were a great number of classifications
middle-aged patients, and occur predominately in females adopted in the literature, and, as a consequence, detailed cor-
[1, 2]. The atypical and malignant meningiomas are charac- relative clinicopathologic studies on this subject are difficult
terized by successive recurrences and an aggressive behav- to perform, and it is not possible to compare data from differ-
ior. Among all meningiomas, their incidence varies in the ent centers. The two reviews of the World Health Organization
literature ranging from 4.7 to 7.1 % and 1.0 to 3.7 % for (WHO) 2007 classifications provides histologic criteria of
atypical and malignant, respectively [3–6]. There is absence meningiomas with reasonably good prognostic correlation
of female predominance in malignant meningiomas suggest- [8]. However there is a large body of the literature not based
ing that endocrinologic influences apparently correlated with on the WHO 2000 or 2006 WHO classifications. These clas-
the genesis of benign meningiomas are not active in malig- sifications include increased mitotic count (2000) as the main
nant ones [4]. differentiation between atypical and anaplastic lesions. and
In general, meningiomas are usually well circumscribed, brain invasion as a criteria for atypia (2007).
globular, and slow-growing tumors that are thought to arise Additionally, another characteristic of meningiomas is the
from mesodermal arachnoid cells. These arachnoid cap cells, histologic dedifferentiation, increasing the grade as the suc-
localized on the apex of arachnoid granulations, are exposed cessive recurrences occur. This aspect is not often discussed
to the venous blood of the sinus and have a low rate of cell in the literature, and the existing series report an incidence of
division. The induction of cell division is thought to be an 4.4–18 % of cases [3, 5, 9, 10]. Ikeda et al. [11] demonstrated
early important step in the neoplastic transformation of these that the major phenotypic changes in the progression of
cells. They are structurally similar to those of meningiomas, meningiomas from the classic to the anaplastic type are the
and both tend to form whorls. However, this presumed origin loss of the meningioma’s architecture, a decreased expression
does not explain some unusual sites such as the ventricles or of epithelial membrane antigen (EMA), an increased expres-
within the cerebral parenchyma, where they probably arise sion of vimentin, and an expression of abnormal intermediate
from perivascular arachnoidal cells [7]. filament proteins. Recently, al-Mefty et al. [12] published a
series of 11 (6 %) meningiomas with malignant progression
out of 175 recurrent lesions. Interestingly, they found a com-
A. Gorgulho, M.D., M.Sc. (*) • A.A.F. De Salles, M.D., Ph.D. plex karyotype present in the histologically lower-grade
Department of Neurosurgery, University of California Los Angeles,
Los Angeles, CA, USA
tumor, contradicting the stepwise clonal evolution model.
Department of Neurosurgery, Hospital de Coração-Associação
do Sanatório Sírio, Sao Paulo, Brazil
e-mail: a_gorgulho@yahoo.com; agorgulho@mednet.ucla.edu Surgical Considerations
C.A. Mattozo, M.D.
Department of Surgery, Division of Neurosurgery, David Geffen Microsurgery remains the best option for symptomatic intra-
School of Medicine at UCLA, Los Angeles, CA, USA cranial meningiomas if complete resection can be achieved
Assistant Professor of Neurosurgery, Department of Neurosurgery, with low morbidity. Simpson [13] has proposed a system that
Pontificia Universidade Catolica de Curitiba, Brazil grades the extent of tumor removal and correlates with tumor
296 A. Gorgulho et al.
relapse. Grade 1 resection required gross total removal of all [36] studied the role of XRT for those lesions and found that
visible tumor plus dural attachments and was associated with young age (<58 years), modern imaging and treatment
9 % local recurrence. Gross total removal with dural cautery planning techniques, and a postoperative radiation dose of at
constituted grade 2 resection and was associated with a 19 % least 50 Gy contribute to improved outcome in patients with
relapse. A partial resection was a grade 3 procedure with a atypical and malignant meningiomas. They also recom-
relapse rate of 40 %. Poorer control rates occurred after mended that all patients should receive radiation therapy
grade 4 (decompression) or 5 (biopsy) procedures. immediately after initial surgery. Dziuk et al. [37] stated that
Since this system was proposed, many other reports have adjuvant XRT, recurrence status, and extent of resection
confirmed that local relapse of meningiomas depends upon were independent predictors of recurrence for malignant
the extent of first resection [14–16]. After gross total removal, meningiomas. They found that the 5-year disease-free sur-
the local recurrence rate at 5 and 10 years approximate 10 % vival after an initial total resection increased from 15 to 80 %
and 20 %, respectively [16]. After less than gross complete with radiation therapy. Recently, Ware et al. [38] published
resection, relapse rates at 5, 10, and 15 years were recently their experience treating atypical and malignant meningio-
reported at 25 %, 70 %, and 90 %, respectively [16]. mas with resection followed by brachytherapy with perma-
Jääskeläinen et al. [3] reported a recurrence rate 5 years after nent low-dose 125I. The median freedom from progression of
complete resection of 3 % for benign, 38 % for atypical, and the combined group of patients was 10.4 months.
78 % for anaplastic meningiomas. Furthermore, this approach resulted in a high complication
Nevertheless, total excision including the dural site of ori- rate. Their radiation necrosis rate was 27 %, with 13 % of
gin is seldom possible, especially when important vascular patients requiring additional surgery (Case Study 20.1).
and/or neural structures are encased within the tumor or if The main criticism for XRT as a complement to surgical
the tumor infiltrates adjacent anatomic structures [17–19]. resection has been a high long-term complication rate, which
The success rates associated with microsurgical excision of has been reported to be from 17 to 38 % when patients were
benign skull base tumors has varied from 26 to 89 % [20– followed longer than 5 years [39]. Nutting et al. [35] in their
25]. Although the mortality rate has decreased in recent skull base benign meningioma series reported that complica-
reports, a significant number of patients continue to experi- tions such as retinopathy, cataract, hypopituitarism, and
ence adverse effects related to surgical procedures with rates short-term memory deficit were likely to have been related to
ranging from 15 to 26 % [24, 26–29]. irradiation, occurring in 13.4 % of the patients. The true inci-
dence of neuropsychologic side effects for patients who sur-
vive more than 6 months without tumor recurrence, as in the
Radiation Therapy case of meningiomas, has been significantly underestimated.
A study of 748 patients showed some degree of neuropsy-
The benefit of conventional radiation therapy (XRT) to com- chologic deficit in 29 %, and 12 % were classified as having
plement subtotal resection of meningiomas is well estab- post-radiation dementia [40].
lished [10, 20, 30]. External beam radiation therapy in doses
of 45–50 Gy reduces the progression rate after partial resec-
tion of meningiomas [30, 31]. A control rate of 72–98 % has Stereotactic Radiation
been reported, depending on the timing of XRT after surgery,
the extent of resection, the presence of cell atypia, the XRT Radiation therapy to large portions of the brain is necessary
dose, the tumor location, and the XRT technique used [30– only in selected meningiomas. Malignant meningiomas with
34]. Glaholm et al. [32] reported 5-, 10-, and 15-year partial resection, as well as atypical meningiomas with risk
relapse-free survival rates of 78 %, 67 %, and 56 %, respec- of dural spread, require more extensive radiation fields.
tively, after postoperative radiotherapy. Well-demarcated benign lesions must be managed with focal
Goldsmith et al. [33], in a retrospective review of 117 radiation, either stereotactic radiosurgery (SRS) or stereotac-
patients receiving conventional RT after subtotal resection, tic radiotherapy (SRT) depending on involvement of elo-
reported a significant impact on progression-free survival quent structures or volume of the lesion.
(PFS) after XRT related to total dose. The 10-year relapse- The application of SRS and SRT for treatment of menin-
free rate in those receiving ≤52 Gy was 65 % compared with giomas has been extensively studied at our institution. For
93 % for those receiving >52 Gy. For skull base meningio- the radiosurgery procedures at UCLA, a Clinac 18 linear
mas, a 5- and 10-year PFS rate of 92 % and 83 %, respec- accelerator (Varian, Inc., Palo Alto, CA) adapted to a Philips
tively, has been reported after XRT [35]. SRS 200 was used from 1990 to 1996. A dedicated Varian
Atypical and malignant meningiomas differ from their 600 SR and X-Knife software (Radionics, Inc., Burlington,
benign counterparts in that they frequently recur after MA) were introduced for the development of SRT, being
complete excision and are more often fatal. Milosevic et al. used from 1996 to 1998. The Novalis dedicated system with
20 Meningioma 297
Table 20.1 Tumor growth control after stereotactic radiation of benign Table 20.4 Tumor control rates for cavernous sinus meningiomas
meningiomas according to the UCLA grading system
Tumor size SRS (%) SRT (%) Total (%) Grade Meningioma radiologic aspect Tumor control (%)
Decreased 22 (34.9) 24 (33.3) 46 (34.1) I Confined to the cavernous sinus 90
No change 36 (57.1) 46 (63.8) 82 (60.7) II Involvement of the petroclival region 86
Increased 5 (7.9) 2 (2.7) 7 (5.2) without brain-stem compression
Total 63 (100) 72 (100) 135 (100) III Extension to and compression of the 86
Tumor control 58 (92) 70 (97.2) 128 (94.8) optic nerve, chiasm, or tract
IV Involvement of the petroclival region 42
with compression of the brain stem
Table 20.2 Clinical follow-up results V Extensive involvement of both 0
Neurologic findings SRS (%) SRT (%) Total (%) cavernous sinus
Improved 22 (35) 21 (32.3) 43 (33.5)
Unchanged 36 (57.1) 42 (64.6) 78 (60.9)
Worsened 5 (7.9) 2 (3.1) 7 (5.4) De Salles et al. [43] studied the role of SRS and SRT in
Number of cases 63 (100) 65 (100) 128 (100) the management of skull base meningiomas and proposed a
grade system (see the section “Patient Algorithm” later).
Table 20.3 Tumor growth control after stereotactic radiation of atypical They studied 40 patients (34 SRS; 6 SRT) treated during an
meningiomas 8-year period. Tumor control rates were 90 % for grade I,
Tumor size SRS (%) SRT (%) Total (%) 86 % for grade II, 86 % for grade III, 42 % for grade IV, and
Decreased 1 (6.2) 1 (20) 2 (9.5) no control for grade V (Table 20.4). The control rate corre-
No change 4 (25) 2 (40) 6 (28.5) lated well with the tumor grade (p < 0.036). Additionally, the
Increased 11 (68.7) 2 (40) 13 (62) complication rate was 7.5 % (three cases) occurring only in
Total 16 (100) 5 (100) 21 (100) SRS treatments. Since this report, grades III, IV, and V are
Tumor control 5 (31.2) 3 (60) 8 (38.1) treated with SRT at our institution.
The experience of treatment of cavernous sinus meningi-
omas with SRT was recently reviewed by Selch et al. [44].
miniature multileaf collimator capability has been used since They studied 45 patients treated with a median prescribed
1997 (BrainLAB, Munich, Germany). dose of 50.4 Gy delivered to a median tumor volume of
The whole UCLA meningioma series was published by 14.5 cm3. The 3-year progression-free survival rate was
Torres et al. [41] with a review of 161 patients who under- 97.4 % and the preexisting neurologic complaints improved
went SRS or SRT for treatment of 194 meningiomas in all in 20 % of patients. One patient had an ipsilateral cerebro-
locations. Clinical and radiologic follow-up were obtained in vascular accident 6 months after SRT, although this compli-
128 patients harboring 156 lesions. Stereotactic radiosurgery cation is not likely related to the procedure.
was used to treat 79 lesions, and SRT was used to treat 77
lesions. The overall mean follow-up period was 32.5 months.
The mean prescribed dose for the SRS group was 15.6 Gy, Stereotactic Radiosurgery
whereas for the SRT group, the dose was 48.4 Gy. For benign
tumors, the overall local control rate was 94.8 %. The control SRS has become a well-accepted modality for the treatment
rates for the SRS and SRT were 92 % and 97.2 %, respec- of a number of neurologic indications, ranging from primary
tively (Table 20.1). The overall rate of clinical improvement and metastatic malignancies to benign tumors, arteriovenous
was 33.5 %. Worsening of the clinical symptoms was malformations (AVMs), pain, and other neurologic condi-
observed in 7.9 % of patients who underwent SRS and in tions. Unlike conventional fractionated radiotherapy, SRS
3.1 % who underwent SRT (Table 20.2). The incidence of does not exploit differences in the repair capacities of normal
clinical complications in the SRS and SRT groups was 5 % and neoplastic cells after repeated exposure to sublethal
and 5.2 %, respectively. For the group of 21 atypical menin- doses of irradiation. SRS achieves a therapeutic ratio by vir-
giomas, the tumor growth control was 31.2 % and 60 % for tue of the steep dose gradient between the isocenter and sur-
the SRS and SRT groups, respectively (Table 20.3). rounding normal parenchyma, inevitably produced by the
The analysis of the long-term follow-up (mean, 72.5 convergence of multiple small radiotherapy beams. The bio-
months) group of 32 patients revealed tumor growth control logic effect is irreparable cellular damage and vascular
rate of 92.3 % and 100 % for the SRS- and SRT-treated occlusion within the high dose volume.
patients, respectively. Worsening of previous neurologic defi- Meningiomas have several characteristics that make them
cit was identified in two (7.9 %) patients treated with SRS. No excellent tumors for radiosurgery [45]. First, they usually
complications were found in the SRT patients [42]. grow slowly, which allows time for the effects of irradiation
Fig. 20.1 (a) MRI scan from Case 1 showing recurrence of an atypical 1.07 cm3. Note that there is a side inversion in the brainlab treatment
meningioma in left tentorium 45 months after initial resection. (b) plan software
Treatment plan with shaped beans, 14 Gy, 90 % isodose line, volume
to materialize. Second, they are easily imaged with both homogeneity has been sacrificed to obtain conformal tumor
computed tomography (CT) and magnetic resonance imag- planning, thereby avoiding nearby normal structures.
ing (MRI), thereby facilitating accurate stereotactic targeting However, those structures encased by the tumor, such as
and appropriate dose planning. Finally, a high dose with cranial nerves, the carotid artery, or even the optic apparatus,
rapid fall-off can be directed to the imaging-defined tumor may be jeopardized with inhomogeneous radiation treat-
margin with confidence because meningiomas are usually ment. Multiple isocenters restrict the prescription to low iso-
well demarcated and rarely invade surrounding brain tissues dose volumes with a very high gradient of dose between the
(Fig. 20.1). The application of stereotactic radiation using all periphery and interior of the tumor. According to Nedzi et al.
available techniques, including linear accelerator (linac) [41, [60], dose inhomogeneity ≥5 Gy was the most significant
43, 44, 46–48]. Gamma Knife [49–54], and proton beam factor associated with morbidity after SRS.
[55], has been reported in the literature. Within the past few years, however, the development of
The effectiveness of SRS and the risks of complication miniature multileaf collimators has altered the way in which
depend on the size of the tumor. Lesions larger than 3 cm in radiosurgery is performed with linac systems. This technique
diameter must be treated with significantly lower doses of allows the beam to conform to the target closely and exclude
radiation to avoid excessive risk of complications. Such small normal structures. In addition to high target conformity, the
radiation doses may make radiosurgery ineffective. Tumor “single isocenter shaped beam approach” results in a homo-
location also affects the treatment decision, depending on the geneous dose within the target volume, as well as a highly
proximity to eloquent structures (especially the optic appara- efficient planning and treatment process.
tus). Single-dose SRS is indicated for meningiomas smaller SRS has already proved to be effective in the treatment of
than 3 cm or 20 cm3 in volume and with a minimal distance meningiomas. Kondziolka et al. [52] evaluated the long-
from the optic apparatus of between 2 and 4 mm [54, 56, 57]. term outcome of 85 patients (90 % of them skull base
Furthermore, tumors causing significant brain-stem compres- located) who underwent Gamma Knife radiosurgery. Tumor
sion may not be suitable for single-dose radiosurgery because control was obtained in 93 % of patients. Tumor resection
a slight swelling during the course of tumor necrosis induced was necessary in five (5.8 %) patients because of delayed
by radiosurgery may not be tolerable. Many alternatives were tumor growth or persistent symptoms. New or worsened
attempted in order to overcome such limitations, usually neurologic deficits occurred in five (5.8 %) patients during
resulting in treatment failure. This included two-staged proce- the follow-up.
dures [49, 58] or a decrease in the ideal peripheral dose [59]. Thirty-four percent to 71 % of meningiomas treated by
A lack of homogeneity in the radiosurgery dose has been SRS are expected to decrease in size, and 37–57 % are
a controversial issue in SRS treatment of brain tumors. Dose expected to stabilize [41, 51, 61, 62]. DiBiase et al. [63]
20 Meningioma 299
Fig. 20.2 (a) MRI scan at 33-month follow-up showing a new lesion in the floor of the left medial cranial fossa. (b) Treatment plan with shaped
beans, 18 Gy, 90 % isodose line, volume 0.65 cm3
studied the factors predicting local tumor control after (p < 0.001). The complication rate was 10 % for SRS and
Gamma Knife radiosurgery in 162 patients harboring menin- 22 % for the surgical group (p = 0.06).
giomas. The median prescribed dose was 14 Gy to the 50 % Patients treated by radiosurgery for meningioma must be
isodose line and the median tumor volume was 4.5 cm3. The followed carefully. Flickinger et al. [65] studied a group of
5-year disease-free survival was 86.2 %. The factors corre- 219 meningiomas without pathologic verification. The actu-
lated with a worse prognosis in their study were male arial rate of identifying a diagnosis other than meningioma
patients, conformity index less than 1.4, and tumor volume was 2.3 % at 5 and 10 years. Nakatomi et al. [66] reported
greater than 10 cm3. However, patients who were treated one patient treated by SRS for a supposed cavernous sinus
with less conformal plans that covered the dural tail had bet- meningioma that required surgical resection 17 months after
ter outcomes (Fig. 20.2). because of recurrence. The postoperative diagnosis was pri-
Hakim et al. [46] published the results of 127 patients mary cavernous sinus malignant lymphoma.
with 155 meningiomas treated by SRS with linac, including Kim et al. [67] recently analyzed the role of Gamma
atypical and malignant meningiomas. The median marginal Knife radiosurgery in 23 patients with 26 superficially
dose was 15 Gy and the median volume was 4.1 cm3. located meningiomas. The median tumor volume was
Freedom from progression was observed in 84.3 % of 4.7 cm3, and the mean margin dose was 16 Gy at the 50 %
patients at a median time of 23 months. The actuarial tumor isodose line. The local control rate was 95 % with stable
control for the patients with benign meningiomas was 89.3 % MRI imaging in 65 % of the cases. Symptomatic and tran-
at 5 years. Permanent complications were observed in 4.7 % sient peritumoral high signal on T2-weighted magnetic reso-
of patients. nance (MR) images was observed in 43 % of patients. This
SRS for patients with small- to medium-sized meningio- complication was associated with a high integral dose and a
mas was recently reported to provide tumor control equiva- large tumor volume. Moreover, tumor shrinkage was more
lent to Simpson grade 1 resection. Pollock et al. [64] prominent in the patients with symptomatic high signal
compared two groups of patients with lesions smaller than (p = 0.03).
35 mm. The 62 patients who underwent SRS with a mean MRI showed peritumorous imaging changes in 24 % of
margin dose of 17.7 Gy had no statistically significant differ- patients treated by SRS in the series published by Chang
ence in the 3- and 7-year actuarial PFS when compared with et al. [61]. Their overall rate of symptomatic edema was
57 patients who underwent complete surgical resection, 9.3 %. Furthermore, these imaging findings were related to
including their dural attachments and bone abnormalities tumor dose in univariate analysis, and with tumor location in
(Simpson 1) with no radiation. Furthermore, SRS provided a the multivariate analysis. Radiation-induced imaging
higher PFS rate compared with that in patients in whom changes were more often seen in convexity, parasagittal, and
resections were Simpson grade 2 (p < 0.05) and grade 3–4 falx meningiomas than in those of the skull base.
Table 20.5 Doses utilized in radiosurgery for intracranial meningiomas

Skull Base Meningiomas–SRS Peripheral
Author, Year Dose Variation (Gy) Technique
Meningiomas in the skull base account for the largest group Hakim, 1998 [46] 15 9–20 LINAC
of incompletely resected lesions usually due to cavernous Liscak, 1999 [59] 12 10–14 GK
sinus involvement, brain-stem adhesion, and infiltration of Kondziolka, 1999 [51] 16 9–25 GK
adjacent structures. Complete excision in these situations is Shafron, 1999 [47] 12.7 10–20 LINAC
unlikely without causing morbidity [68–70]. Stafford, 2001 [54] 16 12–36 LINAC
Villavicencio, 2001 [114] 15 12–18.5 LINAC
Meningiomas are the most frequent benign tumors treated
Friedman, 2005 [115] 16 10–20 LINAC
by radiosurgery, and the majority of them are located on the
Hasegawa, 2007 [116] 13 7.5–17 GK
skull base (Fig. 20.3). Liscak et al. [71] studied 176 patients
Pollock, 2012 [117] 16 N/A GK
with meningiomas located on the skull base. They obtained
N/A Not Available, LINAC Linear Accelerator, GK Gamma Knife
local control in 98 % of patients and a persistent morbidity
rate of 4.5 %. The rate of radiation-induced edema was 11 %.
Significantly lower edema occurrence was observed when the sinus treated by SRS. New cranial neuropathies were found
dosage to the tumor margin was lower than or equal to 14 Gy. in 19 % of patients. Additionally, all patients who developed
Chuang et al. [72] studied the linac-based SRS for 43 optic system neuropathy (6.4 %) received radiation doses
unresectable or partially resected meningiomas in the base of higher than 8 Gy to optic apparatus.
the skull. The 7-year local control in the SRS-alone group Leber et al. [75] correlated the doses delivered in the
and surgical excision with SRS group was 100 % and 84.4 %, treatment of cavernous sinus lesions by SRS with incidence
respectively (p = 0.21). of optic neuropathy. The actuarial rate was zero for patients
Using a mean margin dose of 13 Gy with a Gamma Knife who received a dose of less than 10 Gy, 26.7 % for patients
unit for treatment of 32 petroclival meningiomas, Roche receiving 10 to less than 15 Gy, and 77.8 % for those who
et al. [73] achieved local control in all patients. Although no received doses of 15 Gy or more (p < 0.0001).
new cranial deficit was observed, two patients developed Stafford et al. [76] reported 218 parasellar SRS proce-
hemiparesis due to pontine infarction. dures with 73 % of patients receiving more than 8 Gy to a
Spiegelmann et al. [48] studied 42 patients with menin- short segment of the optic apparatus. The overall incidence
giomas involving the cavernous sinus treated with linac. of radiation neuropathy was 1.9 %; the risk of neuropathy
A mean dose of 14 Gy was delivered to the tumor margin and was 1.1 % for patients receiving 12 Gy or less.
the median tumor volume was 8.2 cm3. Tumor control was Morbidity after SRS is strongly associated with tumor
observed in 97.5 % of the patients. Cranial nerve complica- volume and location [15, 77]. For these reasons, application
tions included new trigeminal neuropathy in 4.7 % and a new of SRS to cavernous sinus meningiomas has generally been
visual defect in 2.8 %. restricted to tumors <3 cm in greatest dimension and located
Despite the dosimetric advantages of SRS, permanent several millimeters from the optic apparatus [56, 77–79].
cranial nerve deficit is more common in the treatment of cav- These are the grade I and grade II of the UCLA classification
ernous sinus meningiomas. Tishler et al. [74] studied 62 [43]. Tumors grade III, IV, and V are candidates for SRT.
patients (42 with meningiomas) inside or near the cavernous See Table 20.5 for doses used in stereotactic radiosurgery
for intracranial meningiomas.
Stereotactic Radiotherapy
SRT is a treatment option that may be undertaken when the

risk of SRS is unacceptably high; for instance, in cases in
which tumors involve the optic pathways, the brain stem, or
lesions with larger volume near eloquent structures.
This technique combines the physical dose localization
advantage of SRS with the radiobiologic benefits of dose
fractionation [80]. Unlike SRS, the treatment planning fidu-
cial system and patient immobilization device can be applied
noninvasively [81]. The accuracy of patient repositioning
Fig. 20.3 MRI scans 6 months after the second treatment by SRS during SRT has been documented [82–84]. A total fraction-
showing both lesions with decreased size ated dose of ionizing radiation, known from conventional
20 Meningioma 301
radiation therapy experience to be effective for meningioma Table 20.6 Doses utilized in stereotactic radiotherapy for intracranial
yet tolerated by the central nervous system, can be delivered meningiomas
by the SRT technique. Author, year Dose/variation (Gy) Fractions (Gy)
Lo et al. [85] studied a group of 18 patients treated by SRT Andrews, 2002 [93] 51 (50–54) 1.8
for meningiomas with a large volume or adjacent to eloquent Selch, 2004 [44] 50.4 (N/A) 1.8
structures. The median dose was 54 Gy and the median vol- Henzel, 2006 [118] 56 (50.4–60) 1.8–2.0
ume was 8.8 cm3. The 3-year local control rate was 92.3 %; Adler, 2006 [119] 20.3 (N/A) 2–5
one complication was reported (5.5 %). These findings were Litré 2009 [120] 45 (N/A) 1.8
Combs, 2012 [121] N/A (25–68) 1.6–5
not statistically significant, different when compared with
another group of 35 patients who underwent SRS with a Gy Gray, N/A Not Available
median dose of 14 Gy delivered to a median volume of 6.8 cm3.
The application of SRT was recently reviewed by Zabel early side effects, including one VII nerve palsy and one
et al. [86] in 317 patients with benign and atypical meningio- Addisonian state.
mas in all intracranial sites. The median total dose was Debus et al. [87] published a large series of 189 patients
57.6 Gy delivered to a median volume of 33.6 cm3. The over- treated by SRT for skull base meningiomas with a mean radia-
all local control rate was 93.1 %. As expected, the local tion dose of 56.8 Gy. The median target volume was 52.5 cm3.
tumor failure was significantly greater in patients with WHO The tumor control rates for WHO grade 1 and WHO grade 2
grade 2 meningiomas than in patients with WHO grade 1 or tumors were 98.3 % and 77.7 %, respectively. Preexisting cra-
unconfirmed histology (p < 0.002). Moreover, patients with a nial nerve symptoms resolved completely in 28 % of the
tumor volume larger than 60 cm3 had a higher recurrence rate patients. Only 7 % of patients showed an impairment of pre-
(15.5 %) when compared with patients with tumor volumes existing neurologic symptoms, and 1.6 % experienced new
smaller than 60 cm3 (p < 0.001). They observed worsening of clinically significant deficits. These excellent results corrobo-
preexisting symptoms and new clinical symptoms in 8.2 % rate the UCLA experience with skull base meningiomas
and 2.5 % of patients, respectively. treated with SRT [44]. See Table 20.6 for doses used in stereo-
tactic radiosurgery for intracranial meningiomas.
Skull Base Meningiomas–SRT

Intensity-Modulated Radiation Therapy
Skull base meningioma patients for whom microsurgery
and/or SRS are contraindicated seem to be the ideal applica- The intensity-modulated radiation therapy (IMRT) technique
tion of SRT (Fig. 20.4). Alheit et al. [82] studied a group of modulates the intensity of the photon beam across the treat-
24 patients (21 of them harboring skull base meningiomas) ment portal. It has shown the ability to conform the dose to
treated by SRT with doses ranging from 50 to 55 Gy deliv- complex-shaped target volumes. Baumert et al. [88] made
ered to a median volume of 21 cm3. The progression-free comparisons between the treatment plans of SRT and IMRT
survival rate was 100 %, and two (8 %) patients experienced for skull base lesions. The conformity index and the sparing of
the organs at risk were better for the IMRT plan. The largest
improvement was observed for multifocal and irregular
tumors.
Pirzkall et al. published the results of IMRT treatment on
20 patients with benign skull base meningiomas. They used
a median minimum dose of 40 Gy delivered to a median vol-
ume of 108 cm3. In a median follow-up of 36 months, the
local tumor control rate was 100 %. The preexisting neuro-
logic symptoms improved in 60 % and worsened in 5 % of
patients. They also observed no radiation-induced peritu-
moral edema or new onset of neurologic deficits.
Interestingly, the application of intensity-modulated
stereotactic radiosurgery (IMSRS) for treatment of small
skull base lesions was compared with dynamic conformal
Fig. 20.4 (a) MRI scan showing an incidental finding of a small falx arc technique (DCA). Superior homogeneity and cover-
meningioma in a 77-year-old woman with high operative risk. The
lesion had a slight increase in size during 1 year of follow-up and was age for DCA was observed. The IMSRS also increased the
treated with 14 Gy, 90 % isodose line. (b) Stable size of the tumor 40 time for planning, dose delivery, and integral dose to the
months after treatment brain [89].
The intensity-modulated stereotactic radiation therapies

(IMSRT or IMSRS) have limited application when treating
intracranial meningiomas; specific cases where the eyes or
the brain stem are surrounded by the lesion may benefit from
this technique (Case Study 20.2).
Special Applications of Stereotactic

Radiation
Optic Sheath Fig. 20.5 (a) MRI scan from a 74-year-old German woman with
an incidental found meningioma in the tip of the left temporal fossa.
Optic nerve sheath meningiomas are uncommon, slow- She was followed with serial MRIs and growth was documented.
The patient refused surgery and accepted radiosurgery. (b) Treatment
growing tumors that can be categorized as primary and sec-
plan showing inclusion of the dural tail within the prescription isodose
ondary according to their origin. Although conservative line (90 %, 14 Gy, shaped beam)
management can be considered, the most common treatment
is resection, which may lead to blindness due to damage to
the pial vascular supply of the optic nerve [90]. Radiation
therapy has been used in recent decades to treat these tumors application in atypical and malignant histologic features
[91, 92]. (Fig. 20.5). At our institution, SRS and SRT are used to sal-
SRT has been used as one alternative to avoid radiation- vage recurrences after surgery and regional radiation.
induced optic neuropathy. Andrews et al. [93] treated 30 Ojemann et al. [95], using a Gamma Knife unit, treated 22
patients with 33 optic nerve sheath meningiomas with patients with primary or recurrent malignant meningiomas.
SRT. Local tumor control was observed in all patients, and The overall 5-year survival and progression-free survival
they observed 92 % of preserved vision among the optic rates were 40 % and 26 %, respectively. Age (<50 years) and
nerves with vision before treatment. Improvement in visual tumor volume (<8 cm3) were significant predictors of better
acuity and/or visual field occurred in 42 % of nerves. Visual prognosis. Complications were observed in 5 (23 %) patients
loss was observed in 2 (6.6 %) patients and 1 (3.3 %) devel- who developed radiation necrosis.
oped optic neuritis responsive to steroids. In the meningioma series published by Stafford et al. [54],
Becker et al. studied the role of SRT in 39 patients with 22 patients harboring either atypical (n = 13) or malignant
either primary or secondary optic nerve sheath meningiomas (n = 9) meningiomas were treated by Gamma Knife
using a median dose of 54 Gy. The tumor local control was SRS. They obtained 5-year survival rates of 76 % and 0 %
100 % after a median follow-up of 35 months. The rates of for atypical and malignant meningiomas, respectively. The
visual field stabilization, improvement, and worsening were 5-year local control rates were 68 % and 0 % for atypical and
70 %, 25 %, and 4 %, respectively. Additionally, new endocri- malignant tumors, respectively. Univariate analyses of tumor
nologic deficits after treatment were found in 14 % of patients histologic features demonstrated no correlation between
with primary tumors and in 8 % of patients with secondary complications arising after radiosurgery.
tumors. The visual improvement observed in patients treated Harris et al. [96] studied the results of Gamma Knife
by SRT for optic sheath meningiomas was recently reported SRS for treatment of 18 atypical and 12 malignant menin-
to occur within 1–3 months after treatment in 81 % of patients giomas. The 5- and 10-year overall actuarial survival rates
who experienced visual improvement [94]. Homogeneity of for atypical meningiomas were 59 % (±13), and malignant
the dose distribution becomes important when planning SRT lesions had 5- and 10-year overall actuarial survival rates of
for these tumors, as frequently the optic nerve is surrounded 59 % (±16) and 0. These rates were not significantly differ-
by the tumor. This is possible with shaped beams. ent from one another. Atypical meningiomas had a 5-year
PFS of 83 % (±7), and malignant tumors had a PFS of 72 %
(±10) (p = 0.018). They also found a significantly better
Atypical and Malignant Meningiomas PFS for lesions treated without evidence of progression
before SRS.
The role of conventional radiation therapy and brachyther- The results published recently by Huffmann et al. [97]
apy in the management of atypical and malignant meningio- using Gamma Knife SRS for treatment of atypical menin-
mas is well established [33, 36–39]. Although many series giomas is the first report where the WHO 2000 classifica-
are focused on stereotactic radiation for treatment of menin- tion criteria was clearly applied in a radiosurgery study.
giomas, relatively few published studies have defined its They treated 15 patients with 21 lesions using a median
20 Meningioma 303
prescription dose of 16 Gy. The local tumor control and logic verification. The residual tumors underwent radiosur-
marginal/distant tumor control were 95 % and 60 %, respec- gery or fractionated radiotherapy depending on whether they
tively. The only patient with local relapse was treated with were larger or smaller than 3 cm. Another group of 17
a prescription dose of 15 Gy. patients underwent SRS-only, given the presence of tumors
Zabel et al. [86] published a long-term follow-up to smaller than 3 cm that were distant from optic apparatus. The
meningiomas treated by SRT. A group of 26 patients harbor- 5-year actuarial local control rate was 94 %.
ing atypical lesions had 3-, 5-, and 10-year recurrence-free Recently, cytology of tumors located in the cavernous
survival rates of 96 %, 89 %, and 67 %, respectively. Atypical sinus became possible [100]. Frameless stereotaxis allows
histologic grade and tumor volume larger than 60 cm3 were for needle biopsy of these tumors through the foramen ovale.
correlated with a significantly worse local control in com- Biopsy of these tumors became part of the management of
parison with benign meningiomas. tumors in the cavernous sinus requiring a differential diagnosis
(Case Study 20.3).
Combination of Surgery and Radiosurgery

Recommendations
The concern of some neurosurgeons about better functional
outcomes has led some groups to adopt more conservative Patient Algorithm
approaches for skull base meningiomas. Radiosurgery has
become more accepted and recently is being combined with Treatment decisions for meningiomas are best made by a
surgery. Aziz et al. [98] reviewed 38 patients who underwent team of neurosurgeons and radiation oncologists representing
surgical resection for large sphenoid wing meningiomas that expertise in microsurgery, radiosurgery, and radiation ther-
invaded the cavernous sinus. They proposed that tumor por- apy. A combined approach to a tumor with microsurgery and
tions involving the medial compartment of cavernous sinus, radiosurgery is promising. Decompression of critical
as well as portions that infiltrate the superior orbital fissure, structures by microsurgery facilitates and increases the effec-
should be left for observation or stereotactic radiation. tiveness of radiosurgery. The literature suggests that conser-
Maruyama et al. [99] proposed a treatment strategy com- vative resection followed by radiation therapy leads to better
bining nonradical surgery and radiosurgery for the manage- preservation of neural function [24, 26, 101, 102].
ment of cavernous sinus meningiomas. Twenty-three patients When choosing a course of treatment, the patient’s age
had partial or subtotal surgical resection with the main aim to and clinical status must also be considered. Radiosurgery
remove the tumor components extending outside the cavern- should be considered the primary option in elderly patients
ous sinus. This was performed when the tumor was attached in which the risks of surgery are too great or when the poten-
to or displacing the optic apparatus and the brainstem, in tial morbidity of surgery may be unacceptable (Fig. 20.6).
patients with tumors larger than 3 cm in mean diameter, or in Because radiosurgery only requires local anesthesia, several
tumors suspected of malignancy with the aim to obtain histo- patients in their eighties and one in her nineties have
Fig. 20.6 (a) MRI scan showing a cavernous sinus and petro-clinoid isodose line. (c) MRI scan showing stable image after 9 years of fol-
ligament meningioma in a 71-year-old woman with pain in V2 and V3 low-up. The patient is requiring only sporadic medication for facial
region. (b) Treatment plan with multiple isocenters, 16 Gy, 50 % pain
Fig. 20.7 (a) MRI scans of a 52-year-old woman experiencing double treated by SRT. (b) Treatment plan using shaped beams with 50.4 Gy in
vision and lack of balance showing a large meningioma involving the 28 fractions, 90 % isodose line, volume 39.22 cm3. (c) MRI scans
left cavernous sinus, left petroclival region extending into the prepon- showing decrease in tumor size 43 months after treatment
tine and premedullary cisterns. The patient refused surgery and was
successfully undergone radiosurgery for their meningiomas Such small radiation doses may make radiosurgery ineffec-
at UCLA. tive; therefore surgical resection remains the best treatment
The clinical status of a patient should always be thoroughly for large tumors. Patients who are not surgical candidates
assessed prior to treatment. When choosing the best treatment, who have tumors with large size and/or close to optic appa-
it is important to note the onset, severity, and rate of progres- ratus or brainstem are given stereotactic radiotherapy instead
sion of any symptoms. A patient who is asymptomatic may of SRS. Both modalities are used in different clinical sce-
choose to postpone any treatment until symptoms develop or narios in our department and should be available to comple-
until imaging provides evidence of growth. A patient’s clinical ment each other, providing safer treatment options for
status may weigh heavily on the choice between radiosurgery patients (Fig. 20.7).
and microsurgery. Additionally, for cavernous sinus meningiomas, the deci-
Patients with parasellar lesions with acute neurologic sion between SRS or SRT is made with the help of our grad-
deterioration and previous history of malignancy, immuno- ing system [43]. Tumors confined to the cavernous sinus
deficiency, or inflammatory disease could benefit from histo- (grade I) or involving the petroclival region without brain-
logic diagnosis before treatment. Since 2003, we have used stem compression (grade II) are suitable for SRS. Lesions
an image-guided frameless biopsy procedure via the fora- with extension to the optic apparatus (grade III), involvement
men ovale, avoiding a major microsurgical procedure. This of petroclival region with compression of the brain stem
procedure is well tolerated by the patient, with a short hospi- (grade IV), or involvement of both cavernous sinus (grade V)
tal stay [100]. are recommended to receive SRT (Table 20.4). Surgical
The effectiveness of radiosurgery and the risks of compli- efforts should be concentrated in turning a high-grade cav-
cation depend upon the size of the tumor. Lesions larger than ernous sinus meningioma into a lower-grade tumor that may
3 cm in diameter must be treated with significantly lower be controlled by stereotactic radiation. Optic sheath menin-
doses of radiation to avoid excessive risk of complications. giomas are also indicated for SRT unless the lesion has small
20 Meningioma 305
size in a patient without useful vision, when SRS or micro- Complication Avoidance
surgery may be used.
The dose is chosen based on tumor volume using an isoeffect
line generated from an integrated logistic formula. The for-
Procedure and Treatment Planning mula relates dose and volume to the incidence of complica-
tions after SRS for AVMs [103]. In general, the authors
At UCLA, the SRS procedure begins early in the morning attempted to remain below the 3 % isoeffect line for
with the fixation of a stereotactic frame to the patient’s prevention of permanent symptomatic brain injury. On a log-
skull. Routinely, a BRW (Radionics) or a BrainLAB stereo- log plot of dose versus volume, this line is straight with a
tactic frame is applied after administration of a local anes- negative slope and has proved to be a reliable predictor of
thetic. A stereotactic contrast-enhanced CT scan is obtained SRS complications, as previously described by Kjelberg
and fused with a previously acquired 3-mm slice MR et al. [104] Dose was also tailored to the surrounding anat-
image. Image fusion is performed using a mutual informa- omy and typically reduced for lesions near the brain stem or
tion algorithm, which allows geometric alignment between cranial nerves; the optic chiasm was restricted to 8 Gy.
the two image sets. Thus, the CT scan is used for the preci- Using the beam’s-eye-view technique, the radiation arcs
sion of their anatomic localization, whereas high-resolution or beams avoid the eyes and eloquent structures whenever
MR images provide the anatomic details necessary for possible. Relative doses and sizes are designed to prevent
treatment planning. maximum dose to critical structures within or in close rela-
The same protocol is used for the SRT-treated patients; tionship to the tumor.
however, the patients are immobilized with a custom-fitted
thermoplastic face mask (U-PLAST; BrainLAB) instead of a
regular frame [82]. Before each daily treatment, reproduc- Future Directions
ibility of patient positioning is evaluated using the depth hel-
met method of Kooy et al. [83]. For both SRS and SRT Decisions regarding conservative management versus
techniques, the three-dimensional treatment plans were aggressive surgery and adjuvant therapy should rely on pre-
generated using a BrainLAB planning system (versions cise parameters. The use of dynamic contrast-enhanced per-
3.5–4.03; BrainLAB). fusion MRI with T2-weighted sequences measuring the
The treatment plan is safely performed using the endothelial permeability is useful in distinguishing atypical
“beam’s-eye-view” conformal technique (Fig. 20.8). meningiomas from typical meningiomas [105].
According to the tumor size and location, the radiation is Although the new WHO classification recommends more
delivered using dynamic arcs or static shaped beams using a objective criteria [8], there is still lack of correlation between
dedicated 6-MV linear accelerator (Novalis; BrainLAB). grade and behavior in some cases. Recently, a new marker to
This equipment utilizes a micro-multileaf collimator, predict meningioma behavior has been used. The human
obviating the need for multiple isocenters for dose telomerase catalytic subunit (hTERT) expression has shown
conformation. significant correlation with recurrent meningiomas [106,
107]. Another promising tool is the use of DNA microarray
assay technology, which detects numerous genetic abnor-
malities, to differentiate meningioma subtypes [108].
Likewise, the surgery field has been taking advantage of
new technology in recent years. The advances in neuronavi-
gational computer technologies have allowed surgeons to
display three-dimensional reconstructions of the lesions in
relation to the vessels, nerves, and eloquent tracts [109]. The
diffusion-tensor imaging tractography has also been used in
treatment planning for radiosurgery [110]. There have been
significant advances in minimally invasive skull base sur-
gery. Standard approaches to difficult regions have been
developed to allow safer and more complete resection [111–
113]. As we better understand meningioma biology and
genetics, and master the surgical techniques, we can improve
Fig. 20.8 MRI scans 1 year after treatment showing the lesion stable in
patient outcomes and quality of life.
size
Case Study 20.1 1.07 cm3 (Fig. 20.9b). The follow-up scan 33 months after
treatment showed a stable tentorial lesion (not shown) but a
A 47-year-old woman had a left frontotemporal atypical new lesion (arrow) in the floor of the left medial cranial fossa
meningioma that was resected on January 1998. On October (Fig. 20.10a). The patient was retreated with 18 Gy delivered
2001, she presented with headaches, and the MRI scan to 90 % isodose line, volume of 0.65 cm3 (Fig. 20.10b). The
revealed a small recurrence (Fig. 20.9a). She was treated by last follow-up MRI scan in January 2005 showed both
SRS, 14 Gy delivered to 90 % isodose line, volume of lesions decreasing in size (Fig. 20.11).
Fig. 20.9 Treatment plan with intensity-modulated stereotactic radiotherapy showing the sparing of eyes. The total dose of 59.4 Gy was
given with 33 fractions of 1.8 Gy delivered to 90 % isodose line for a volume of 24.83 cm3
Fig. 20.10 Coronal and sagittal MRI scans from Case 2 showing residual meningioma in frontal fossa and olfactory groove
20 Meningioma 307
Fig. 20.11 (a) MRI scan showing a left frontoparietal dural based increasing in size after 20 months of follow-up. The pathology
recurrent atypical meningioma. (b) Treatment plan with shaped report revealed malignant progression to WHO grade III
beans (with a left to right inversion), 47.6 Gy, 85 % isodose line in meningioma
28 fractions, volume 20.68 cm3. (c) MRI scan showing the tumor
Case Study 20.2 report revealed meningioma with atypical features, and
the MRI scan showed residual lesion in frontal fossa
A 51-year-old woman had surgical removal of a frontal (Fig. 20.12). Given the aggressive behavior of this resid-
fossa meningioma in January 2001 at an outside facility. ual tumor and the vision preservation, the treatment
The patient was stable until early 2003 when she started option was intensity-modulated stereotactic radiotherapy
experiencing fatigue and headaches. The MRI scan (IMSRT). The patient was treated with 33 fractions of
showed a residual tumor arising from the olfactory groove 1.8 Gy with a total radiation dose of 59.4 Gy delivered to
and extending into the right cribriform plate and right eth- 90 % isodose line, volume of 24.83 cm3 (Fig. 20.13). The
moid. She underwent a craniofacial approach with incom- MRI scans 1 year after the treatment showed the lesion
plete tumor resection in August 2003. The pathology stable in size (Fig. 20.14).
Fig. 20.12 MRI scans from case 3 showing a cavernous sinus region tumor extending into the left cerebellopontine angle cistern and the
prepontine cistern. The images were suggestive of a meningioma
Fig. 20.13 (a) Screen shot of the navigation system showing the radiotherapy with a total dose of 45 Gy, 85 % isodose line. The volume
needle trajectory for the frameless foramen ovale biopsy. (b) Treatment treated was 17.0 cm3
plan with dynamic arcs. The tumor was treated with stereotactic
Fig. 20.14 MRI scans 6 months after treatment showing a remarkable tumor response to radiation. This behavior is compatible with the
diagnosis of a cavernous sinus lymphoma
20 Meningioma 309
Case Study 20.3 men ovale (Fig. 20.16). The pathology report revealed
A 48-year-old man with diagnosis of AIDS had a 3-month that the tumor was likely of lymphocytic and histiocytic
history of double vision and pain in the left periorbital origin. It was believed that the mass represented a lym-
area as well as some decreased sensation in the left face. phoma. The patient underwent stereotactic radiotherapy
The MRI scan showed an extraaxial mass in the cavern- with a total dose of 45 Gy in 25 fractions of 1.8 Gy deliv-
ous sinus region extending into the left cerebellopontine ered to the 85 % isodose line, volume of 17.0 cm3
angle cistern and the prepontine cistern. The radiology (Fig. 20.16). The MRI scans 6 months after treatment
report was suggestive of a meningioma (Fig. 20.15). showed a remarkable decrease in size of the lesion. This
Given the patient’s history of HIV, the option was to per- rapid response confirms the diagnosis of a cavernous
form a frameless cavernous sinus biopsy through the fora- sinus lymphoma (Fig. 20.17).
Fig. 20.15 (a) MRI scan from a 71-year-old man with diagnosis of sylvian fissure lesion after 30 months. Only this lesion was treated
Parkinson disease showing multiples meningiomas. (b) The patient with 12 Gy, 85 % isodose line. (c) MRI scan showing the tumor
was followed and the MRI scan showed increase in size of the right slightly smaller in size 32 months after treatment
Intracranial
Meningioma
Symptomatic
Mass Effect Incidental Finding
Observation
No High Operative
Microsurgery Progression
Risk
Yes
Complete No
Stereotactic
Resection Radiation
Yes
• Not Eloquent • Eloquent

• Grades I - II • Grades III - V
• < 30 mm • > 30 mm
Observation SRS SRT
Fig. 20.16 Flowchart depicting selection criteria for stereotactic radiation of intracranial meningiomas used in our clinic
Fig. 20.17 The beam’s-eye-view image (above) as seen in the whenever possible. Relative doses and sizes are designed to prevent
treatment plan of a skull base meningioma (below) using the maximum dose to critical structures within or in close relationship
shaped beam technique with micro-multileaf collimator structures to the tumor
6. Perry A, Scheithauer BW, Stafford SL, et al. “Malignancy” in

References meningiomas: a clinicopathologic study of 116 patients, with
grading implications. Cancer. 1999;85:2046–56.
1. Bondy M, Ligon BL. Epidemiology and etiology of intracranial 7. Drummond KJ, Zhu JJ, Black PM. Meningiomas: updating basic
meningiomas: a review. J Neurooncol. 1996;29:197–205. science, management, and outcome. Neurologist. 2004;10:113–30.
2. Rohringer M, Sutherland GR, Louw DF, et al. Incidence and clini- 8. Kleihues P, Louis DN, Scheithauer BW, et al. The WHO classifi-
copathological features of meningioma. J Neurosurg. 1989;71: cation of tumors of the nervous system. J Neuropathol Exp Neurol.
665–72. 2002;61:215–25. discussion 226–219.
3. Jaaskelainen J, Haltia M, Servo A. Atypical and anaplastic menin- 9. Hug EB, Devries A, Thornton AF, et al. Management of atypical
giomas: radiology, surgery, radiotherapy, and outcome. Surg and malignant meningiomas: role of high-dose, 3D-conformal
Neurol. 1986;25:233–42. radiation therapy. J Neurooncol. 2000;48:151–60.
4. Mahmood A, Caccamo DV, Tomecek FJ, et al. Atypical and malig- 10. Pourel N, Auque J, Bracard S, et al. Efficacy of external fraction-
nant meningiomas: a clinicopathological review. Neurosurgery. ated radiation therapyin the treatment of meningiomas: a 20-year
1993;33:955–63. experience. Radiother Oncol. 2001;61:65–70.
5. Palma L, Celli P, Franco C, et al. Long-term prognosis for atypical 11. Ikeda H, Yoshimoto T. Immunohistochemical study of anaplastic
and malignant meningiomas: a study of 71 surgical cases. meningioma with special reference to the phenotypic change of
J Neurosurg. 1997;86:793–800. intermediate filament protein. Ann Diagn Pathol. 2003;7:214–22.
20 Meningioma 311
12. al-Mefty O, Kadri PA, Pravdenkova S, et al. Malignant progres- 33. Goldsmith BJ, Wara WM, Wilson CB, et al. Postoperative irradia-
sion in meningioma: documentation of a series and analysis of tion for subtotally resected meningiomas. A retrospective analysis
cytogenetic findings. J Neurosurg. 2004;101:210–8. of 140 patients treated from 1967 to 1990. J Neurosurg.
13. Simpson D. The recurrence of intracranial meningiomas after sur- 1994;80:195–201.
gical treatment. J Neurochem. 1957;20:22–39. 34. Wilson CB. Cavernous sinus meningiomas. Surg Neurol.
14. Adegbite AB, Khan MI, Paine KW, et al. The recurrence of intra- 1996;46:191–2.
cranial meningiomas after surgical treatment. J Neurosurg. 35. Nutting C, Brada M, Brazil L, et al. Radiotherapy in the treatment
1983;58:51–6. of benign meningioma of the skull base. J Neurosurg.
15. Marks LB, Spencer DP. The influence of volume on the tolerance 1999;90:823–7.
of the brain to radiosurgery. J Neurosurg. 1991;75:177–80. 36. Milosevic MF, Frost PJ, Laperriere NJ, et al. Radiotherapy for
16. Mirimanoff RO, Dosoretz DE, Linggood RM, et al. Meningioma: atypical or malignant intracranial meningioma. Int J Radiat Oncol
analysis of recurrence and progression following neurosurgical Biol Phys. 1996;34:817–22.
resection. J Neurosurg. 1985;62:18–24. 37. Dziuk TW, Woo S, Butler EB, et al. Malignant meningioma: an
17. Sekhar LN, Patel S, Cusimano M, et al. Surgical treatment of indication for initial aggressive surgery and adjuvant radiotherapy.
meningiomas involving the cavernous sinus: evolving ideas based J Neurooncol. 1998;37:177–88.
on a ten year experience. Acta Neurochir Suppl. 1996;65:58–62. 38. Ware ML, Larson DA, Sneed PK, et al. Surgical resection and
18. Sen C, Hague K. Meningiomas involving the cavernous sinus: his- permanent brachytherapy for recurrent atypical and malignant
tological factors affecting the degree of resection. J Neurosurg. meningioma. Neurosurgery. 2004;54:55–63. discussion 63–54.
1997;87:535–43. 39. al-Mefty O, Kersh JE, Routh A, et al. The long-term side effects
19. Sindou MP, Alaywan M. Most intracranial meningiomas are not of radiation therapy for benign brain tumors in adults. J Neurosurg.
cleavable tumors: anatomic-surgical evidence and angiographic 1990;73:502–12.
predictibility. Neurosurgery. 1998;42:476–80. 40. Crossen JR, Garwood D, Glatstein E, et al. Neurobehavioral
20. al-Mefty O, Ayoubi S, Smith RR. The petrosal approach: indica- sequelae of cranial irradiation in adults: a review of radiation-
tions, technique, and results. Acta Neurochir Suppl (Wien). induced encephalopathy. J Clin Oncol. 1994;12:627–42.
1991;53:166–70. 41. Torres RC, Frighetto L, De Salles AA, et al. Radiosurgery and
21. Bricolo AP, Turazzi S, Talacchi A, et al. Microsurgical removal of stereotactic radiotherapy for intracranial meningiomas. Neurosurg
petroclival meningiomas: a report of 33 patients. Neurosurgery. Focus. 2003;14:e5.
1992;31:813–28. discussion 828. 42. Torres RC, De Salles AA, Frighetto L, et al. Long-term follow-up
22. Couldwell WT, Fukushima T, Giannotta SL, et al. Petroclival using linac radiosurgery and stereotactic radiotherapy as a mini-
meningiomas: surgical experience in 109 cases. J Neurosurg. mally invasive treatment for intracranial meningiomas.
1996;84:20–8. Radiosurgery. 2004;5:115–23.
23. DeMonte F, Smith HK, al-Mefty O. Outcome of aggressive 43. De Salles AA, Frighetto L, Grande CV, et al. Radiosurgery and ste-
removal of cavernous sinus meningiomas. J Neurosurg. reotactic radiation therapy of skull base meningiomas: proposal of a
1994;81:245–51. grading system. Stereotact Funct Neurosurg. 2001;76:218–29.
24. O’Sullivan MG, van Loveren HR, Tew Jr JM. The surgical resect- 44. Selch MT, Ahn E, Laskari A, et al. Stereotactic radiotherapy for
ability of meningiomas of the cavernous sinus. Neurosurgery. treatment of cavernous sinus meningiomas. Int J Radiat Oncol
1997;40:238–44. discussion 245–237. Biol Phys. 2004;59:101–11.
25. Sekhar LN, Swamy NK, Jaiswal V, et al. Surgical excision of 45. Kondziolka D, Lunsford LD. Radiosurgery of meningiomas.
meningiomas involving the clivus: preoperative and intraoperative Neurosurg Clin N Am. 1992;3:219–30.
features as predictors of postoperative functional deterioration. 46. Hakim R, Alexander 3rd E, Loeffler JS, et al. Results of linear
J Neurosurg. 1994;81:860–8. accelerator-based radiosurgery for intracranial meningiomas.
26. De Jesus O, Sekhar LN, Parikh HK, et al. Long-term follow-up of Neurosurgery. 1998;42:446–53. discussion 453–444.
patients with meningiomas involving the cavernous sinus: recur- 47. Shafron DH, Friedman WA, Buatti JM, et al. Linac radiosurgery for
rence, progression, and quality of life. Neurosurgery. 1996;39: benign meningiomas. Int J Radiat Oncol Biol Phys. 1999;43:321–7.
915–9. discussion 919–920. 48. Spiegelmann R, Nissim O, Menhel J, et al. Linear accelerator
27. Little KM, Friedman AH, Sampson JH, et al. Surgical manage- radiosurgery for meningiomas in and around the cavernous sinus.
ment of petroclival meningiomas: defining resection goals based Neurosurgery. 2002;51:1373–9. discussion 1379–1380.
on risk of neurological morbidity and tumor recurrence rates 49. Iwai Y, Yamanaka K, Ishiguro T. Gamma knife radiosurgery for
in 137 patients. Neurosurgery. 2005;56:546–59. discussion the treatment of cavernous sinus meningiomas. Neurosurgery.
546–559. 2003;2:517–24. discussion 523–514.
28. Nanda A, Jawahar A, Sathyanarayana S. Microsurgery for poten- 50. Kobayashi T, Kida Y, Mori Y. Long-term results of stereotactic
tial radiosurgical skull base lesions: a retrospective analysis and gamma radiosurgery of meningiomas. Surg Neurol.
comparison of results. Skull Base. 2003;13:131–8. 2001;55:325–31.
29. Soyuer S, Chang EL, Selek U, et al. Radiotherapy after surgery for 51. Kondziolka D, Levy EI, Niranjan A, et al. Long-term outcomes
benign cerebral meningioma. Radiother Oncol. 2004;71:85–90. after meningioma radiosurgery: physician and patient perspec-
30. Taylor Jr BW, Marcus Jr RB, Friedman WA, et al. The meningi- tives. J Neurosurg. 1999;91:44–50.
oma controversy: postoperative radiation therapy. Int J Radiat 52. Kondziolka D, Nathoo N, Flickinger JC, et al. Long-term results
Oncol Biol Phys. 1988;15:299–304. after radiosurgery for benign intracranial tumors. Neurosurgery.
31. Barbaro NM, Gutin PH, Wilson CB, et al. Radiation therapy in the 2003;53:815–21. discussion 821–812.
treatment of partially resected meningiomas. Neurosurgery. 53. Leksell L. Stereotactic radiosurgery. J Neurol Neurosurg
1987;20:525–8. Psychiatry. 1983;46:797–803.
32. Glaholm J, Bloom HJ, Crow JH. The role of radiotherapy in the 54. Stafford SL, Pollock BE, Foote RL, et al. Meningioma radiosur-
management of intracranial meningiomas: the Royal Marsden gery: tumor control, outcomes, and complications among 190
Hospital experience with 186 patients. Int J Radiat Oncol Biol consecutive patients. Neurosurgery. 2001;49:1029–37. discussion
Phys. 1990;18:755–61. 1037–1028.
55. Vernimmen FJ, Harris JK, Wilson JA, et al. Stereotactic proton 76. Stafford SL, Pollock BE, Leavitt JA, et al. A study on the radiation
beam therapy of skull base meningiomas. Int J Radiat Oncol Biol tolerance of the optic nerves and chiasm after stereotactic radio-
Phys. 2001;49:99–105. surgery. Int J Radiat Oncol Biol Phys. 2003;55:1177–81.
56. Nicolato A, Foroni R, Alessandrini F, et al. The role of Gamma 77. Chen JC, Giannotta SL, Yu C, et al. Radiosurgical management of
Knife radiosurgery in the management of cavernous sinus menin- benign cavernous sinus tumors: dose profiles and acute complica-
giomas. Int J Radiat Oncol Biol Phys. 2002;53:992–1000. tions. Neurosurgery. 2001;48:1022–30. discussion 1030–1022.
57. Subach BR, Lunsford LD, Kondziolka D, et al. Management of 78. De Salles AA, Bajada CL, Goetsch S, et al. Radiosurgery of cav-
petroclival meningiomas by stereotactic radiosurgery. ernous sinus tumors. Acta Neurochir Suppl (Wien). 1993;58:
Neurosurgery. 1998;42:437–43. discussion 443–435. 101–3.
58. Pendl G, Schrottner O, Eustacchio S, et al. Stereotactic radiosur- 79. Lee JY, Niranjan A, McInerney J, et al. Stereotactic radiosurgery
gery of skull base meningiomas. Minim Invasive Neurosurg. providing long-term tumor control of cavernous sinus meningio-
1997;40:87–90. mas. J Neurosurg. 2002;97:65–72.
59. Liscak R, Simonova G, Vymazal J, et al. Gamma knife radiosur- 80. Tome WA, Mehta MP, Meeks SL, et al. Fractionated stereotactic
gery of meningiomas in the cavernous sinus region. Acta radiotherapy: a short review. Technol Cancer Res Treat.
Neurochir (Wien). 1999;141:473–80. 2002;1:153–72.
60. Nedzi LA, Kooy H, Alexander 3rd E, et al. Variables associated 81. Gill SS, Thomas DG, Warrington AP, et al. Relocatable frame for
with the development of complications from radiosurgery of intra- stereotactic external beam radiotherapy. Int J Radiat Oncol Biol
cranial tumors. Int J Radiat Oncol Biol Phys. 1991;21:591–9. Phys. 1991;20:599–603.
61. Chang JH, Chang JW, Choi JY, et al. Complications after gamma 82. Alheit H, Dornfeld S, Dawel M, et al. Patient position reproduc-
knife radiosurgery for benign meningiomas. J Neurol Neurosurg ibility in fractionated stereotactically guided conformal radiother-
Psychiatry. 2003;74:226–30. apy using the BrainLab mask system. Strahlenther Onkol.
62. Valentino V, Schinaia G, Raimondi AJ. The results of radiosurgi- 2001;177:264–8.
cal management of 72 middle fossa meningiomas. Acta Neurochir 83. Kooy HM, Dunbar SF, Tarbell NJ, et al. Adaptation and verifica-
(Wien). 1993;122:60–70. tion of the relocatable Gill-Thomas-Cosman frame in stereotactic
63. DiBiase SJ, Kwok Y, Yovino S, et al. Factors predicting local radiotherapy. Int J Radiat Oncol Biol Phys. 1994;30:685–91.
tumor control after gamma knife stereotactic radiosurgery for 84. Solberg T, Ford J, Medin P, et al. Reproducibility of frame posi-
benign intracranial meningiomas. Int J Radiat Oncol Biol Phys. tioning for fractionated stereotactic radiosurgery. J Radiosurg.
2004;60:1515–9. 1999;2:57–64.
64. Pollock BE, Stafford SL, Utter A, et al. Stereotactic radiosurgery 85. Lo SS, Cho KH, Hall WA, et al. Single dose versus fractionated
provides equivalent tumor control to Simpson grade 1 resection stereotactic radiotherapy for meningiomas. Can J Neurol Sci.
for patients with small- to medium-size meningiomas. Int J Radiat 2002;29:240–8.
Oncol Biol Phys. 2003;55:1000–5. 86. Milker-Zabel S, Zabel A, Schulz-Ertner D, et al. Fractionated ste-
65. Flickinger JC, Kondziolka D, Maitz AH, et al. Gamma knife reotactic radiotherapy in patients with benign or atypical intracra-
radiosurgery of imaging-diagnosed intracranial meningioma. Int J nial meningioma: long-term experience and prognostic factors. Int
Radiat Oncol Biol Phys. 2003;56:801–6. J Radiat Oncol Biol Phys. 2005;61:809–16.
66. Nakatomi H, Sasaki T, Kawamoto S, et al. Primary cavernous 87. Debus J, Wuendrich M, Pirzkall A, et al. High efficacy of fraction-
sinus malignant lymphoma treated by gamma knife radiosurgery: ated stereotactic radiotherapy of large base-of-skull meningiomas:
case report and review of the literature. Surg Neurol. 1996;46:272– long-term results. J Clin Oncol. 2001;19:3547–53.
8. discussion 278–279. 88. Baumert BG, Norton IA, Davis JB. Intensity-modulated stereotac-
67. Kim DG, Kim Ch H, Chung HT, et al. Gamma Knife surgery of tic radiotherapy vs. stereotactic conformal radiotherapy for the
superficially located meningioma. J Neurosurg. 2005;102(Suppl): treatment of meningioma located predominantly in the skull base.
255–8. Int J Radiat Oncol Biol Phys. 2003;57:580–92.
68. Kotapka MJ, Kalia KK, Martinez AJ, et al. Infiltration of the 89. Ernst-Stecken A, Lambrecht U, Ganslandt O, et al. Radiosurgery
carotid artery by cavernous sinus meningioma. J Neurosurg. of small skull-base lesions no advantage for intensity-modulated
1994;81:252–5. stereotactic radiosurgery versus conformal arc technique.
69. Larson JJ, van Loveren HR, Balko MG, et al. Evidence of menin- Strahlenther Onkol. 2005;181:336–44.
gioma infiltration into cranial nerves: clinical implications for cav- 90. Gabibov GA, Blinkov SM, Tcherekayev VA. The management of
ernous sinus meningiomas. J Neurosurg. 1995;83:596–9. optic nerve meningiomas and gliomas. J Neurosurg. 1988;68:
70. Zentner J, Meyer B, Vieweg U, et al. Petroclival meningiomas: is 889–93.
radical resection always the best option? J Neurol Neurosurg 91. Kupersmith MJ, Warren FA, Newall J, et al. Irradiation of menin-
Psychiatry. 1997;62:341–5. giomas of the intracranial anterior visual pathway. Ann Neurol.
71. Liscak R, Kollova A, Vladyka V, et al. Gamma Knife radiosurgery 1987;21:131–7.
of skull base meningiomas. Acta Neurochir Suppl. 2004;91: 92. Wara WM, Sheline GE, Newman H, et al. Radiation therapy of
65–74. meningiomas. Am J Roentgenol Radium Ther Nucl Med.
72. Chuang CC, Chang CN, Tsang NM, et al. Linear accelerator- 1975;123:453–8.
based radiosurgery in the management of skull base meningiomas. 93. Andrews DW, Faroozan R, Yang BP, et al. Fractionated stereotac-
J Neurooncol. 2004;66:241–9. tic radiotherapy for the treatment of optic nerve sheath meningio-
73. Roche PH, Pellet W, Fuentes S, et al. Gamma Knife radiosurgical mas: preliminary observations of 33 optic nerves in 30 patients
management of petroclival meningiomas results and indications. with historical comparison to observation with or without prior
Acta Neurochir (Wien). 2003;145:883–8. discussion 888. surgery. Neurosurgery. 2002;51:890–902. discussion 903–894.
74. Tishler RB, Loeffler JS, Lunsford LD, et al. Tolerance of cranial 94. Baumert BG, Villa S, Studer G, et al. Early improvements in vision
nerves of the cavernous sinus to radiosurgery. Int J Radiat Oncol after fractionated stereotactic radiotherapy for primary optic nerve
Biol Phys. 1993;27:215–21. sheath meningioma. Radiother Oncol. 2004;72:169–74.
75. Leber KA, Bergloff J, Pendl G. Dose-response tolerance of the 95. Ojemann SG, Sneed PK, Larson DA, et al. Radiosurgery for
visual pathways and cranial nerves of the cavernous sinus to ste- malignant meningioma: results in 22 patients. J Neurosurg.
reotactic radiosurgery. J Neurosurg. 1998;88:43–50. 2000;93 Suppl 3:62–7.
20 Meningioma 313
96. Harris AE, Lee JY, Omalu B, et al. The effect of radiosurgery dur- 110. Kamada K, Todo T, Masutani Y, et al. Combined use of
ing management of aggressive meningiomas. Surg Neurol. tractography-integrated functional neuronavigation and direct
2003;60:298–305. discussion 305. fiber stimulation. J Neurosurg. 2005;102:664–72.
97. Huffmann BC, Reinacher PC, Gilsbach JM. Gamma Knife sur- 111. Cook SW, Smith Z, Kelly DF. Endonasal transsphenoidal removal
gery for atypical meningiomas. J Neurosurg. 2005;102(Suppl): of tuberculum sellae meningiomas: technical note. Neurosurgery.
283–6. 2004;55:239–44. discussion 244–236.
98. Abdel-Aziz KM, Froelich SC, Dagnew E, et al. Large sphenoid 112. Couldwell WT, Weiss MH, Rabb C, et al. Variations on the stan-
wing meningiomas involving the cavernous sinus: conservative dard transsphenoidal approach to the sellar region, with emphasis
surgical strategies for better functional outcomes. Neurosurgery. on the extended approaches and parasellar approaches: surgical
2004;54:1375–83. discussion 1383–1374. experience in 105 cases. Neurosurgery. 2004;55:539–47. discus-
99. Maruyama K, Shin M, Kurita H, et al. Proposed treatment strategy sion 547–550.
for cavernous sinus meningiomas: a prospective study. 113. Jho HD, Alfieri A. Endoscopic glabellar approach to the anterior
Neurosurgery. 2004;55:1068–75. skull base: a technical note. Minim Invasive Neurosurg.
100. Frighetto L, De Salles AA, Behnke E, et al. Image-guided frame- 2002;45:185–8.
less stereotactic biopsy sampling of parasellar lesions. Technical 114. Villavicencio AT, Black PM, Shrieve DC, Fallon MP, Alexander
note. J Neurosurg. 2003;98:920–5. E, Loeffler JS. Linac radiosurgery for skull base meningiomas.
101. Duma CM, Lunsford LD, Kondziolka D, et al. Stereotactic radio- Acta Neurochir (Wien). 2001;143(11):1141–52.
surgery of cavernous sinus meningiomas as an addition or alterna- 115. Friedman WA, Murad GJ, Bradshaw P, Amdur RJ, Mendenhall
tive to microsurgery. Neurosurgery. 1993;32:699–704. discussion WM, Foote KD, Bova FJ. Linear accelerator surgery for menin-
704–695. giomas. J Neurosurg. 2005;103(2):206–9.
102. Suzuki M, Mizoi K, Yoshimoto T. Should meningiomas involving 116. Hasegawa T, Kida Y, Yoshimoto M, Koike J, Iizuka H, Ishii
the cavernous sinus be totally resected? Surg Neurol. 1995;44:3– D. Long-term outcomes of Gamma Knife surgery for cavernous
10. discussion 10–13. sinus meningioma. J Neurosurg. 2007;107(4):745–51.
103. Flickinger JC. An integrated logistic formula for prediction of 117. Pollock BE, Stafford SL, Link MJ, Brown PD, Garces YI, Foote
complications from radiosurgery. Int J Radiat Oncol Biol Phys. RL. Single-fraction radiosurgery of benign intracranial meningio-
1989;17:879–85. mas. Neurosurgery. 2012;71(3):604–12. doi:10.1227/NEU.
104. Kjellberg RN, Davis KR, Lyons S, et al. Bragg peak proton beam 0b013e31825ea557. discussion 613.
therapy for arteriovenous malformation of the brain. Clin 118. Henzel M, Gross MW, Hamm K, Surber G, Kleinert G, Failing T,
Neurosurg. 1983;31:248–90. Strassmann G, Engenhart-Cabillic R. Significant tumor volume
105. Yang S, Law M, Zagzag D, et al. Dynamic contrast-enhanced per- reduction of meningiomas after stereotactic radiotherapy: results
fusion MR imaging measurements of endothelial permeability: of a prospective multicenter study. Neurosurgery. 2006;59(6):
differentiation between atypical and typical meningiomas. AJNR 1188–94. discussion 1194.
Am J Neuroradiol. 2003;24:1554–9. 119. Adler Jr JR, Gibbs IC, Puataweepong P, Chang SD. Visual field
106. Kalala JP, Maes L, Vandenbroecke C, et al. The hTERT protein as preservation after multisession cyberknife radiosurgery for peri-
a marker for malignancy in meningiomas. Oncol Rep. 2005; optic lesions. Neurosurgery. 2006;59(2):244–54. discussion
13:273–7. 244–54.
107. Maes L, Lippens E, Kalala JP, et al. The hTERT-protein and Ki-67 120. Litré CF, Colin P, Noudel R, Peruzzi P, Bazin A, Sherpereel B,
labelling index in recurrent and non-recurrent meningiomas. Cell Bernard MH, Rousseaux P. Fractionated stereotactic radiotherapy
Prolif. 2005;38:3–12. treatment of cavernous sinus meningiomas: a study of 100 cases.
108. Wrobel G, Roerig P, Kokocinski F, et al. Microarray-based gene Int J Radiat Oncol Biol Phys. 2009;74(4):1012–7.
expression profiling of benign, atypical and anaplastic meningio- 121. Combs SE, Adeberg S, Dittmar JO, Welzel T, Rieken S, Habermehl
mas identifies novel genes associated with meningioma progres- D, Huber PE, Debus J. Skull base meningiomas: Long-term
sion. Int J Cancer. 2005;114:249–56. results and patient self-reported outcome in 507 patients treated
109. Rohde V, Spangenberg P, Mayfrank L, et al. Advanced neuro- with fractionated stereotactic radiotherapy (FSRT) or intensity
navigation in skull base tumors and vascular lesions. Minim modulated radiotherapy (IMRT). Radiother Oncol. 2013;106(2):
Invasive Neurosurg. 2005;48:13–8. 186–91.
Meningioma: Viewpoint—Surgery
21
Lawrence S. Chin, David J. Padalino, Pulak Ray,
and John M. Caridi
into three grades based on microscopic appearance [5]. WHO

Introduction grade I includes several variants, the most common being
the transitional, fibroblastic, and meningiothelial. Grade I
The first successful documented resection of this benign, slow tumors generally follow a benign course and make up nearly
growing tumor was performed by Zanobi Pecchiolo (1801– 90 % of all meningiomas. Grade II tumors, which account
1866) at Siena University. In a vast surgical series published in for 5–7 % of meningiomas, are associated with greater cel-
1847, 1,524 cases were described, one of which described a lular atypia and include the clear cell, chordoid, and atypical
large meningioma removed from the right sinciput through a variants (see Table 21.1). These tumors have a recurrence
triangular flap. Then on December 15, 1887, William W. Keen rate 1.5–5 times greater than that of grade I varieties. Grade
(1837–1932) was the first neurosurgeon to successfully resect III meningiomas (1–3 % of total meningiomas) include the
a meningioma in the United States. The patient proceeded to anaplastic, papillary, and rhabdoid variants. These tumors
survive for 30 years after without any clinical signs of recur- have a high rate of invasion into surrounding tissues (including
rence [1]. Due to much debate over the etiology and pathogen- brain), which leads to a higher incidence of recurrence and
esis of this tumor, it was not until 1922 in his text entitled may metastasize.
Meningiomas, Their Classification, Regional Behaviour, Life Comprising over 30 % of primary intracranial tumors, the
History and Surgical End Results that Harvey Cushing charac- incidence of meningiomas in the population is 4.5 per
terized the tumor as a meningioma, thus leaving room for the 100,000 person-years with females outnumbering males by
histologic type to be clarified [2, 3]. It is now known that 2:1 according to some series, but up to even 6:1 in others [6].
meningiomas arise from transformed arachnoid cap cells The incidence has also been shown to increase with age.
that form slow growing tumors that are attached to the overly- Autopsy studies show that approximately 2.5 % of people
ing dura [4]. While most are well circumscribed, approxi- have an incidental meningioma [7]. Anecdotal evidence has
mately 10 % spread diffusely following and sometimes suggested several correlations with meningiomas. Cushing
invading the contours of the underlying bone. These are believed that head trauma was affiliated with tumorogenesis,
termed meningioma en plaque and are much more difficult to but no definitive study has confirmed this despite multiple
resect. Histologically, the WHO has separated meningiomas attempts to find a correlation [3, 8, 9]. Although several stud-
ies have attempted to link viral infection to the development
of meningiomas, none have proved the causative effect [10].
L.S. Chin, M.D., F.A.C.S., F.A.A.N.S. (*) • D.J. Padalino, M.D.
There is also a link between breast cancer and meningioma
Department of Neurosurgery, SUNY Upstate Medical University, that has been observed in women, but not in men. A study by
750 East Adams Street, Syracuse, NY 13210, USA Rao et al. showed that of 6,527 meningiomas and 153,599
e-mail: chinL@upstate.edu; padalind@upstate.edu breast cancers diagnosed in women and 2,327 meningiomas
P. Ray, M.D. and 1,668 breast cancers diagnosed in men found in five can-
Department of Neurosurgery, Delaware Neurosurgical Group, cer registries between 1995 and 2003, the rate of breast can-
779 Christiana Road, Suite 202, Newark, DE 19713, USA
e-mail: pulak12@gmail.com
cer in women already diagnosed with meningioma was 80
times expected, and the rate of meningioma in women
J.M. Caridi, M.D.
Department of Neurosurgery, Ichan School of Medicine at Mount
already diagnosed with breast cancer was 58 times the
Sinai, One Gustav Levy Place, New York, NY 10029, USA expected rate. No such association could be found in the
e-mail: John.caridi@mountsinai.org male patients [11].
316 L.S. Chin et al.
Table 21.1 Meningioma classification by WHO grade increases [15]. These tumors are nearly always extra-axial,
% of total showing displacement of the cortex. Reactive sclerosis
WHO grade Histological variants meningiomas (hyperostosis) of the overlying bone can also be seen in
I—Benign Transitional, fibroblastic, >90 % some cases. Conventional catheter angiography, although
meningothelial, increasingly supplanted by MR angiography and CT angi-
psammomatous, angiomatous,
microcystic, secretory, ography, is often critical in the surgical planning because it
lymphoplasmacyte-rich, can determine the blood supply to the tumor, degree of vas-
metaplastic cularity, patency of dural venous sinuses, and suitability for
II—Atypical Clear cell, chordoid, atypical 5–7 % endovascular embolization. Meningiomas are primarily fed
III—Malignant Anaplastic, papillary, rhabdoid 1–3 % by branches from the external carotid artery, commonly the
middle meningeal or occipital artery, although large tumors
usually parasitize a significant blood supply from underly-
ing small pial arteries.
Apart from female gender, there are two other major
factors that do seem to lead to meningioma formation: radiation
and chromosome 22 mutation. The incidence of meningioma Surgical Decision-Making and Techniques
has been found to be four times higher in patients who had
tinea capitis treated with local radiation compared to control Numerous factors must be addressed when evaluating a
patients [12]. Another definitive correlation with meningio- patient for surgery: age, location, size, symptoms, neurologi-
mas is deletions in the long arm of chromosome 22. cal deficit, tumor growth rate, proximity to critical structures,
Numerous patients with meningiomas have been found to blood supply, venous drainage and sinus patency, previous
have monosomy 22 and the gene thought responsible for treatments, and most importantly, patient expectations. A
meningioma formation has been localized to the same region multidisciplinary approach with input from other disciplines
as that of type 2 neurofibromatosis [13]. This may explain such as ENT, radiation oncology, and radiology, when appro-
the high correlation between NF2 and the development of priate, is important.
meningiomas. The distinguishing advantage of surgery over radiosur-
The signs and symptoms associated with a meningioma gery is the ability to achieve a gross total resection. The
depend primarily on its site of origin. The most common pre- incidence of tumor shrinkage following radiosurgery varies
sentation is a new onset seizure, occurring approximately between 20 and 40 %, and the actual amount of decrease is
30 % of the time [14]. These seizures are most often general- fairly minor. Thus, overriding any other consideration is the
ized, but can be partial or complex partial. Headache, mental fact that some patients will insist on having even a small
status changes, hemiplegia, isolated cranial nerve deficits, benign tumor surgically removed. These patients should be
and signs of increased intracranial pressure may also be seen offered an opportunity for surgical resection provided that
in meningioma patients. the risks are acceptable. An optimal resection requires com-
The radiographic diagnosis of meningioma is usually plete removal of the meningioma with all of its dural and
made with CT or MRI. Unenhanced CT shows meningio- bony attachments. In a classic paper by Simpson from 1957,
mas to be slightly hyperdense compared to normal brain, he describes the recurrence rates for meningiomas follow-
often with surrounding edema and variable mass effect; ing surgical resection [16]. A five tier grading system was
adding intravenous contrast results in homogeneous devised to describe the extent of tumor removal, with grade
enhancement of the mass. MRI is the preferred method of I being the most complete and grade V being a simple
diagnosis because of its superior soft tissue resolution. The decompression. Simpson noted that the recurrence rate fol-
tumor appears isointense to brain on an unenhanced lowing grade I resection was 9 %, grade II was 19 %, and
T1-weighted image (T1WI), but hyperintense with sur- grade III was 29 %. Later, a grade 0 was added to the
rounding white matter edema on a T2-weighted image Simpson scale [17, 18]. This added a 2 cm margin of normal
(T2WI). The addition of gadolinium contrast reveals a uni- appearing dura to a Simpson I resection. The recurrence rate
formly enhancing mass with a broad-based connection to of a patient undergoing a grade 0 resection of a benign
the surrounding dura. A dural tail is variably seen and rep- meningioma is essentially zero. Clearly, the goal of surgery
resents growth of meningioma along the dura away from is to perform a grade 0 or a grade 1 resection; unfortunately,
the main tumor. The significance of this finding is that if this may not be possible despite surgical expertise because
left untreated by either radiosurgery or surgery, since both the most important variable that influences extent of resec-
are focal treatment modalities, the chance for recurrence tion is tumor location.
21 Meningioma: Viewpoint—Surgery 317
the blood supply can influence the decision to embolize a tumor.

Convexity Region Typically, there is no need for preoperative embolization of
smaller size meningiomas at the convexity with dural feed-
Convexity meningiomas are found on the dura over the ing arteries since there is no significant benefit with regard
hemispheres without extension into the dural sinuses or inva- to making the surgery safer for the patient or making the
sion of the skull base. They represent approximately 15 % of resection technically easier. However, large skull base
meningiomas and are the least likely to recur because they meningiomas with deep feeding arteries that are large
are nearly always amenable to a grade 0 or 1 resection. enough to cannulate in order to inject embolic material such
Intraoperative stereotactic guidance is helpful in localizing as microspheres or liquid embolic agent may benefit from
the projection of the tumor on the scalp. A generous scalp embolization by reducing blood loss and causing tumor
and craniotomy flap can then be tailored to allow for a 2 cm necrosis, which will soften the tissue and lead to a techni-
margin around the tumor. Preoperative embolization is less cally easier removal of the devascularized tumor [19, 20].
helpful than in other areas as the blood supply from dural Care must be taken to closely watch the patient after embo-
feeders is easily interrupted by incising the dura circumfer- lization because post-embolization swelling with significant
entially as the first step in tumor removal. Pial feeders are too mass effect on surrounding structures can occur. Some
small to be embolized and are divided during the course of authors have suggested performing the preoperative emboli-
tumor dissection off the brain. Large tumors may be adherent zation within 24 h prior to the operation to allow for some
to the pia causing unavoidable cortical injury in the process early necrosis and softening of the tumor tissue while mini-
of a complete removal. This superficial invasion does not mizing the duration of time that the patient could potentially
constitute a sign of malignancy, but neurological deficit is develop dangerous post-embolization edema. It is known
possible and patients should be counseled appropriately. that tumor vessels can recanalize if left for greater than
Surgery should be considered anytime a patient wants a cure 24–48 h after embolization, but this is more likely with
of their convexity meningioma regardless of size and also some of the less durable liquid embolic agents that were
recommended strongly in patients with a tumor larger than used in the past. Current technique with microspheres and
3 cm. Radiosurgery is reserved for asymptomatic, small liquid embolic agents provide a much more durable occlu-
meningiomas, and the rare cases of recurrence following sion and recanalization is not as much of a concern. Some
resection (see Case 21.1). have suggested waiting as long as a week in order to maximize
the devascularizing necrosis effects after the embolization
procedure [19, 20]. Embolization is usually not recommended
as a sole treatment.
Parasagittal Region
Parasagittal meningiomas account for 20–30 % of meningio-

mas and describe a location without intervening brain tissue Sphenoid Wing and Anterior Skull Base
or dura between the tumor and the superior sagittal sinus (see
Case 21.2). Conventional angiography or (CT or) MR venog- Sphenoid wing meningiomas are divided into lateral and
raphy is helpful in determining if the sinus is patent. When medial depending on the site of origin. Lateral sphenoid
there is invasion of a venous sinus, a complete resection wing meningiomas may invade the sphenoid bone, but com-
requires removal of the infiltrated sinus as well. This is ordi- plete resection is still possible provided the tumor is not big
narily safe only with the anterior one-third segment of the and the dura of the temporal or frontal floor not extensively
sagittal sinus. The scalp and bone flap will need to cross the invaded (see Case 21.3). Complete removal of a meningioma
midline if the sagittal sinus is to be resected. Patients with originating from the medial sphenoid wing is a difficult task
middle and posterior third sagittal sinus involvement and a for the surgeon because the tumor is likely involving the dura
large tumor can benefit from a combined approach where the around the optic nerve, superior orbital fissure, and cavern-
bulk of the tumor is removed, which relieves mass effect on ous sinus (see Case 21.4).
the brain while leaving the residual tumor in the sinus to be Meningiomas from the anterior cranial floor (olfactory
treated by radiosurgery. Preoperative embolization may be groove, planum sphenoidale, tuberculum sella, and dia-
useful in reducing blood loss during the debulking of large phragma sella) offer similar challenges because of the nearby
tumors. optic nerves, chiasm, and pituitary stalk (see Case 21.5). In
Each case should be reviewed individually, and not all these tumors, a grade 2 resection (coagulation of the dural
meningiomas need preoperative endovascular embolization. attachment) is usually the best possible outcome. An overly
Factors such as the size and location of the tumor as well as aggressive resection may injure cranial nerves II, III, IV, V,
and VI and/or the carotid artery and pituitary stalk. Identifying

and maintaining the arachnoid planes around the meningi- Petroclival Region
oma and observing the lateral border of the cavernous sinus
can help protect neurovascular structures. Early interruption This broad category includes meningiomas found in the
of feeding arteries (e.g., anterior ethmoidal branches) at the posterior cavernous sinus, tentorium, cerebellopontine angle
base of the meningioma diminishes tumor bleeding during (CPA), and foramen magnum. Nearly all of these tumors,
the debulking phase of the operation. Preoperative emboliza- particularly when large, will require a skull base approach.
tion of these tumors can significantly reduce blood supply to A combination of transpetrosal, combined subtemporal and
the tumor, but may have a relatively high risk of blindness transmastoid, retromastoid, and extreme lateral transcondy-
from embolization of collateral blood supply to the optic lar approaches may be needed and should be part of the
nerve from external carotid branches. In general, the tech- skull base neurosurgeon’s arsenal. In most situations, a skull
nique for meningioma removal involves dissection around base specialist from ENT will be needed for their expertise
the periphery of the tumor with coagulation of feeding arter- in the bony anatomy of the temporal bone. Preoperative
ies while avoiding significant brain retraction. Once this can embolization is frequently helpful for larger tumors in this
no longer be accomplished safely because of tumor bulk, the region. Although radiographic gross total resections are
meningioma is incised and internally debulked using coagu- possible, it is rare to accomplish better than a grade II or III
lation and ultrasonic aspirators. The capsule of the tumor can resection (see Case 21.6). Thus, frequent MRI follow-up
then be infolded and dissected off the remaining brain and (yearly intervals) even after a gross total resection is essen-
neurovascular structures. The attachment of the tumor is tial to catch an early recurrence. When residual tumor is
coagulated and, if possible, cut off the skull base. seen on postoperative imaging, it should be treated with
The top priority when removing a meningioma from the radiosurgery. When residual tumor is known to be left
sellar/parasellar region is to decompress the optic nerves behind but cannot be seen radiographically or if the histol-
and chiasm. Not only does this address the most serious ogy is grade II or III, fractionated radiation therapy should
neurological deficit caused by the tumor, but any residual be considered.
tumor can be treated by radiosurgery. A separation of at
least 2–3 mm is needed to permit an adequate dose of radia-
tion to the tumor edge (12 Gy or higher) while keeping the Conclusion
optic nerve dose to safe levels (<8 Gy). Since the cranial
nerves in the cavernous sinus are relatively insensitive to The ideal approach to a patient with a meningioma is to con-
radiation effects, there is little need to aggressively resect sider both surgical and radiosurgical options. A multidisci-
tumor in this location. Visible residual tumor should be plinary team often accomplishes this goal best although good
treated aggressively with early radiosurgery because having balanced opinions can be given in any clinical setting.
a maximum amount of separation between the tumor edge Surgery and radiosurgery should be thought of as comple-
and optic nerve or brainstem is critical in determining the mentary procedures rather than competing treatment modali-
dose that can be safely delivered. A mistake is often made in ties. When relief of mass effect is needed because of tumor
waiting to see if the residual tumor will start to grow. size or patient’s symptoms, then craniotomy is the only
Invariably it does and when finally recognized, the tumor option. Radiosurgery can be used in smaller tumors and can
may be too close to the optic nerve to be treated by radiosur- treat residual tumor, thus freeing the surgeon from having to
gery. At this point, the only options are fractionated radia- resect tumors from areas of the brain and skull that may
tion therapy or repeat surgery. cause significant morbidity.
Case Reports
Case 21..1
underwent gamma knife radiosurgery to a dose of 16 Gy @
A 40-year-old man presented with seizures and mild left-sided 50 % (Fig. 21.1d). In the first 6 months after radiosurgery he
weakness. His MRI scan showed a right convexity meningi- developed transient radiation-induced swelling associated
oma (Fig. 21.1a, b). He underwent a gross total resection with focal seizures and left arm weakness (Fig. 21.1e, f).
(Fig. 21.1c), but pathology indicated an atypical meningioma Approximately 1 year after treatment, the tumor has regressed
(grade 2). He developed a recurrence 9 months later and and the patient is neurologically normal (Fig. 21.1g).
(continued)
Case 21.1 (continued)
Fig. 21.1 MRI scans of a convexity atypical meningioma. (a, b) recurrence of convexity tumor. (e, f) Three months after treatment
Preoperative axial and coronal scans showing a convexity menin- the tumor is unchanged in size, but there is an increase in surround-
gioma with mass effect. (c) Postoperative axial scan showing coming edema as seen on flair images. (g) One year later the tumor has
plete resection of tumor. (d) Gamma knife treatment MRI shows regressed
Case 21.2 Case 21.3

This 37-year-old woman presented with seizures and altered A 63-year-old woman was admitted with severe
mental status. A parasagittal meningioma occluding the sag- headaches and confusion. MRI showed a large right
ittal sinus was seen in the anterior 1/3 sinus (Fig. 21.2a, b). A sphenoid wing meningioma (Fig. 21.3a). After preop-
bifrontal craniotomy was performed and the tumor was erative embolization, a right pterional craniotomy was
totally resected along with the sagittal sinus (Fig. 21.2c, d). performed with total excision of the tumor (Fig. 21.3b).
One year later, she has no recurrence of tumor. One-year follow-up shows no recurrence of tumor.
Fig. 21.3 A right lateral sphenoid wing meningioma. (a)

Preoperative axial MRI shows a large tumor compressing the right
frontal lobe. (b) Postoperative MRI shows a gross total resection
Fig. 21.2 An anterior parasagittal meningioma involving the

sagittal sinus. (a, b) Coronal and sagittal MRI show a meningi-
oma arising from the sagittal sinus and anterior falx. (c, d)
Postoperative MRIs indicate complete tumor resection
Case 21.4 was removed except for the portion in the cavernous
This 47-year-old woman presented with headaches and sinus. This ensured adequate separation between the
was found to have a right medial sphenoid wing menin- tumor and the optic nerves, thus allowing the residual
gioma (Fig. 21.4a). A right pterional craniotomy was per- tumor to be treated by gamma knife radiosurgery
formed to attempt a gross total resection. The entire tumor (Fig. 21.4b). Follow-up 5 years later shows no change in
the small residual cavernous sinus mass (Fig. 21.4c).
Fig. 21.4 A large right medial sphenoid wing meningioma. (a) amount of tumor is left in the cavernous sinus that is treated by
Preoperative axial MRI shows a large tumor arising from the gamma knife radiosurgery. (c) Five years later, there is no change
medial sphenoid wing and cavernous sinus. (b) Only a small in the residual tumor
Case 21.5
A 28-year-old woman presented with a bitemporal

hemianopsia and headaches. Preoperative MRI showed
a diaphragma sella meningioma (Fig. 21.5a). A right
pterional craniotomy with gross total resection of
tumor was performed (Fig. 21.5b). The optic nerves
were well decompressed and her vision returned to
normal. Three years later, she has no recurrence of
tumor and her vision remains normal.
Fig. 21.5 (a) Tuberculum sella meningioma causing a bitemporal
visual field cut. (a) sagittal MRI shows a meningioma arising from
the diaphragma sella with elevation of the optic chiasm. (b)
Postoperative MRI shows that the entire tumor has been removed
21 Meningioma: Viewpoint—Surgery 321
Case 21.6 subtemporal/transmastoid approach (Fig. 21.6d, e). The

This 39-year-old woman presented with severe headaches bulk of the tumor was removed en bloc along with a por-
and a large tumor based on the left petrous portion of the tion of the tentorium. The pathology was atypical menin-
temporal bone and tentorium (Fig. 21.6a–c). The tumor gioma (grade 2). Although no obvious residual tumor was
appeared to involve the left mastoid aircells and had a seen, fractionated radiation therapy was administered
dural tail along the tentorium. After preoperative emboli- because likely some tumor was left and the atypical pathol-
zation, a radical excision was performed using a combined ogy greatly increased the risk of recurrence.
Fig. 21.6 A very large petrous/tentorial meningioma. (a, b) Axial carotid angiogram shows feeders from the middle meningeal artery
and coronal MRI shows a meningioma infiltrating the mastoid and and occipital artery. (d, e) Postoperative MRIs show a gross total
compressing the left cerebellum and temporal lobe. (c) External resection
nervous system. J Neuropathol Exp Neurol. 2002;61:215–25; dis-

References cussion 226–219.
6. CBTRUS: 2006 Annual Report Central Brain Tumor Registry of
1. Keen W. Three successful cases of cerebral surgery. J Med Sci. the United States; 2006.
1888;96:329–57, 452–65. 7. Nakasu S, Hirano A, Shimura T, Llena JF. Incidental meningiomas
2. Al-Mefty O. Meningiomas. New York: Raven; 1991. in autopsy study. Surg Neurol. 1987;27:319–22.
3. Cushing H, 1869–1939.; Eisenhardt, Louise, 1891-: Meningiomas, 8. Phillips LE, Koepsell TD, van Belle G, Kukull WA, Gehrels JA,
their classification, regional behaviour, life history, and surgical end Longstreth Jr WT. History of head trauma and risk of intracranial
results, by Harvey Cushing with the collaboration of Louise meningioma: population-based case-control study. Neurology.
Eisenhardt., 1938 2002;58:1849–52.
4. Kallio M, Sankila R, Hakulinen T, Jaaskelainen J. Factors affecting 9. Preston-Martin S, Pogoda JM, Schlehofer B, Blettner M, Howe
operative and excess long-term mortality in 935 patients with intra- GR, Ryan P, Menegoz F, Giles GG, Rodvall Y, Choi NW, Little J,
cranial meningioma. Neurosurgery. 1992;31:2–12. Arslan A. An international case-control study of adult glioma and
5. Kleihues P, Louis DN, Scheithauer BW, Rorke LB, Reifenberger G, meningioma: the role of head trauma. Int J Epidemiol. 1998;27:
Burger PC, Cavenee WK. The WHO classification of tumors of the 579–86.
10. Krieg P, Scherer G. Cloning of SV40 genomes from human brain benign intracranial meningiomas. Int J Radiat Oncol Biol Phys.
tumors. Virology. 1984;138:336–40. 2004;60:1515–9.
11. Rao G, Giordano SH, Liu J, McCutcheon IE. The association 16. Simpson D. The recurrence of intracranial meningiomas after
of breast cancer in men and women. Neurosurgery. 2009;65: surgical treatment. J Neurol Neurosurg Psychiatry. 1957;20:22–39.
483–9. 17. Borovich B, Doron Y. Recurrence of intracranial meningiomas: the
12. Modan B, Baidatz D, Mart H, Steinitz R, Levin SG. Radiation- role played by regional multicentricity. J Neurosurg. 1986;64:
induced head and neck tumours. Lancet. 1974;1:277–9. 58–63.
13. Joachim T, Ram Z, Rappaport ZH, Simon M, Schramm J, Wiestler 18. Borovich B, Doron Y, Braun J, Guilburd JN, Zaaroor M, Goldsher D,
OD, von Deimling A. Comparative analysis of the NF2, TP53, Lemberger A, Gruszkiewicz J, Feinsod M. Recurrence of intracranial
PTEN, KRAS, NRAS and HRAS genes in sporadic and radiation- meningiomas: the role played by regional multicentricity. Part 2:
induced human meningiomas. Int J Cancer. 2001;94:218–21. Clinical and radiological aspects. J Neurosurg. 1986;65:168–71.
14. Lieu AS, Howng SL. Intracranial meningiomas and epilepsy: inci- 19. Dowd CF, Halbach VV, Higashida RT. Meningiomas: the role of
dence, prognosis and influencing factors. Epilepsy Res. 2000;38: preoperative angiography and embolization. Neurosurg Focus.
45–52. 2003;15(1):Article 10.
15. DiBiase SJ, Kwok Y, Yovino S, Arena C, Naqvi S, Temple R, 20. Probst EN, Grzyska U, Westphal M, Zeumer H. Preoperative embo-
Regine WF, Amin P, Guo C, Chin LS. Factors predicting local lization of intracranial meningiomas with a fibrin glue preparation.
tumor control after gamma knife stereotactic radiosurgery for AJNR Am J Neuroradiol. 1999;20:1695–702.
Meningioma—Viewpoint:
Fractionated Radiotherapy 22
Igor J. Barani, Arie Perry, and C. Leland Rogers
in all age brackets from the second to ninth decades [12, 13].
Introduction Age may have prognostic relevance. More aggressive
clinical and histologic features have been noted in children
Meningiomas are the most common intracranial neoplasm in and in young adults [14]. Jung et al. found that the growth
adults [1, 2]. The majority are benign, and collectively, they rate after subtotal resection (STR) was greater in patients
form the most common group of intracranial neoplasms in younger than 50 years of age (7.51 cm3 per year) than in
adults, comprising over 30 % of primary central nervous sys- older patients (0.83 cm3 per year) [15].
tem (CNS) tumors [2–5]. The reported incidence varies from Overall, meningiomas are more commonly diagnosed in
<[less than] 1 to>[greater than] 6 per 100,000 depending upon females, approximately 2:1 (female:male) [2, 16]. Yet, atypi-
the population and method of identification [3, 4, 6]. An cal and anaplastic meningiomas may occur more frequently
aggregate incidence of 2.6 per 100,000 has been calculated in men [17]. The reason for this gender distribution is unclear.
from surgical series [7]. Extrapolating this to the population Progesterone and estrogen receptors have been identified in
of the United States, it can be expected that approximately greater than 70 and 30 % of meningiomas, respectively [17].
8,000 meningiomas will be diagnosed per year [7]. Many are Increased estrogenicity is believed to be associated with an
identified solely on the basis of imaging findings; in recent increase in tumor-related symptoms during pregnancy or the
studies, 35–62 % were diagnosed on the basis of imaging luteal phase of the menstrual cycle [18]. A heightened risk
alone [8, 9]. Using the University of Pittsburgh data, image- has been described with obesity, and an increased relative
diagnosed meningiomas may account for an additional 2,800 risk of breast cancer has been described in meningioma
patients, bringing the total estimate to nearly 11,000 annually. patients, as has the converse [18]. These endocrinologic
Even this sum is likely an underestimate. Meningiomas have associations are intriguing, but to date, no hormone-based
been identified in as many as 2.3 % of autopsies, many of manipulations have been of therapeutic benefit [19, 20].
which are subclinical [10]. Meningiomas may present with focal neurologic symp-
Except in the setting of neurofibromatosis type 2 (NF2), toms depending on their intracranial location. Seizure is one
meningiomas are rare in the pediatric population [11]. The of the more common presenting signs, occurring in 26 % of
peak occurrence is during the sixth and seventh decades of meningioma patients [21]. Many meningiomas of the skull
life, but the age range is broad, with a 5 % greater incidence base present with lateralized cranial nerve deficits [22].
Raised intracranial pressure is uncommon at presentation,
but can occur secondary to obstructive hydrocephalus.
I.J. Barani, M.D.
Therapy is therefore aimed at relieving neurologic symptoms
Department of Radiation Oncology, University of California, and providing long-term tumor control.
San Francisco, 505 Parnassus Avenue, Room L08, San Francisco, Meningiomas are neoplasms of the meningothelial cells of
CA, USA the arachnoid layer and occur most frequently where these
A. Perry, M.D. cells are most numerous, namely within the arachnoid villi or
Department of Pathology and Neurological Surgery, granulations that lie along the dural venous sinuses [23].
University of California, San Francisco, 505 Parnassus Ave.,
M551, San Francisco, CA, USA
Furthermore, arachnoid cap cell clusters become more prom-
inent with age, corresponding to the age-related increase in
C.L. Rogers, M.D. (*)
Department of Radiation Oncology, Virginia Commonwealth
incidence of meningioma [25]. The most common locations
University, 401 College St., Richmond, VA, USA include the convexity (parasagittal dura, or attached to the
e-mail: Leland@GammaWest.com sagittal sinus), sphenoid ridge, and planum sphenoidale. Other
324 I.J. Barani et al.
locations include the sylvian fissure, parasellar region, and Classification

olfactory grooves. Meningiomas can also arise intraventricu-
larly from pia-arachnoid rests, and about 10 % occur infraten- Although the majority of meningiomas are considered
torially along the free edge of the tentorium, clivus, foramen benign (~80 %), it is well recognized that the biological
magnum, pretroclival ligament, and petrous ridge [24]. Less spectrum is wide and, for a substantial minority, the clinical
frequently, optic nerve sheaths and spinal canal may be course difficult to predict. Even histopathologically benign
involved. meningiomas can recur despite seemingly complete resec-
tion. It has been postulated that microscopic foci of tumor in
dural strips away from the main tumor mass may be respon-
Etiology sible [35]. This finds support in the fact that patients with
multiple meningiomas often display the identical NF2 muta-
The etiology of meningiomas remains largely unknown. tion in all tumors [36]. These studies suggest that even
Cushing and others invoked trauma as potentially caus- benign meningiomas may spread peripherally through the
ative, but this has not been substantiated in large epide- dura, possibly in a discontinuous fashion in some cases. If
miologic studies [26, 27]. Meningiomas occur with greater microscopic nests are left behind, then recurrence is possible
frequency in patients with certain rare genetic conditions, despite a relatively indolent biology.
such as NF2, but this explains only a small minority of High-grade meningiomas (atypical and anaplastic) com-
cases [11, 14]. Ironically, despite its role as an effective prise roughly 25 % of cases and are considerably more
therapy for meningiomas, prior exposure to ionizing radia- aggressive, being associated with increased morbidity and
tion is also a risk factor for developing meningioma [7, mortality. Thus, in addition to the prognostic significance of
28–33]. Similarly, radiation-induced meningioma accounts extent of resection, one of the most important factors to pre-
for only a very small fraction of cases. The vast majority dict tumor behavior is histopathologic grading [37, 38].
of meningiomas arise in the absence of any known predis- Atypical meningioma (WHO grade II) has a 7–8-fold
posing factors and are often referred to as “sporadic.” increased risk of recurrence over benign tumors (WHO grade I)
and a slightly, but statistically significant increased risk of
mortality compared with age and sex matched controls.
Radiation-Induced Meningioma Anaplastic meningiomas (WHO grade III) are fortunately
rare (1–3 %), but are associated with considerable risk of
The data establishing radiation as the etiologic agent for death from disease, with the mean survival being less than 2
meningioma formation derives from experience with atomic years in one study [39]. Therefore, following the diagnosis of
bomb survivors, and with cranial irradiation for tumors or an atypical meningioma, more aggressive treatment may be
scalp irradiation for maladies such as tinea capitis [7, 28–33]. indicated [40]. This would similarly apply to anaplastic
Based on an experience of tinea capitis in children who meningiomas, although nearly all will eventually recur, and
immigrated to Israel in the aftermath of the World War II, a often exhaust both surgical and radiation therapy options.
sevenfold increase in neoplasms of the nervous system, diag- In the 2000 WHO classification scheme [41], the histo-
nosed, on average, 18 years after treatment, with 19 menin- logical grading of meningiomas underwent substantial revi-
giomas and 7 gliomas in nearly 11,000 subjects [29]. In a sion and standardization of criteria based on the
separate study, Strojan and colleagues projected the risk of clinicopathological series from Mayo Clinic [37, 38]. This
developing meningioma after radiation therapy to be 0.53 % has resulted in some upgrading of previously diagnosed
at 5 years and 8.18 % at 25 years [32]. Indeed, radiation- benign meningiomas to the atypical category [42]. To a lesser
induced meningiomas remain the most common form of extent, there has been a modest decrease in the number of
radiation-induced neoplasm in literature [33]. In contrast, a anaplastic or malignant meningiomas due to stricter diagnos-
large retrospective cohort study comparing radiosurgery out- tic requirements for these subtypes. In comparison, the most
comes with national cancer registries (approximately 5,000 recent WHO classification revision published in 2007 [43]
patients with 30,000 patient-years of follow-up) did not made minimal additional changes to meningioma grading,
detect increased risk of malignancy [34]. This study supports the most significant being that brain-invasive meningiomas
the safety of radiosurgery, and suggests a dose-volume effect are regarded as WHO grade II based on similar recurrence
for radiation-induced meningioma. It may thus be expected and mortality to atypical meningiomas in general. However,
that modern irradiation techniques, aimed at high degrees of WHO chose not to make brain invasion itself another crite-
conformality and at reducing high-dose as well as low-dose rion for the diagnosis of atypical meningioma, since benign-
volumes, would be beneficial. appearing examples often do not share the same genetic
alterations seen in other high-grade meningiomas.
22 Meningioma—Viewpoint: Fractionated Radiotherapy 325
Histopathology and morbidity if left untreated. Therefore, therapeutic inter-

vention is often needed for most diagnosed tumors. In clini-
The current WHO classification lists 16 variants or subtypes, cal practice, surgery has been the treatment of choice for
falling into three grade designations (Table 22.1) [25, 44]. most meningiomas. Resection enables histologic diagnosis,
Four subtypes are higher grade by definition and include relieves mass effect and associated symptoms, alleviates
clear cell (WHO grade II), chordoid (WHO grade II), tumor-induced seizures, and achieves high local control
rhabdoid (WHO grade III), and papillary (WHO grade III) rates. However, not all tumors can be completely resected
meningiomas. However, the required fraction of such variant without considerable morbidity. The likelihood of safe and
histology in the tumor specimen to warrant a higher grade complete resection depends, in part, upon site and surgical
designation has not been or standardized. Such determination accessibility, and is frequently precluded by involvement of
is at the discretion of the neuro-pathologist and often neces- the cranial nerves, arteries, or by venous sinus invasion.
sitates secondary review in a specialty center. Gross total resection (GTR) is often not attempted in the cav-
With appropriate treatment, approximately 80 % of ernous sinus, clivus, cerebello-pontine angle, sellar region,
benign (WHO grade I) or image-diagnosed meningiomas or optic nerve sheath. This is significant since in his seminal
remain progression-free at 10 years (Fig. 22.1). In contrast, publication in 1957, Simpson correlated the extent of tumor
only about 35 % of atypical (WHO grade II) patients remain resection, associated dural attachments, and any hyperostotic
disease-free at 10 years even with definitive therapy [37, 39, 43]. bone to local recurrence risk, and defined 5 grades of resec-
Anaplastic (WHO grade III) meningiomas have a mean over- tion, so-called Simpson grades (Table 22.2) [45]. The surgi-
all survival of less than 2–3 years and often progress quickly cal resection extent has been consistently identified as an
through serial treatments [39]. important prognostic feature [7, 12, 46, 47], and majority of
centers continue to use Simpson’s criteria.
More recently Sughrue et al. challenged the relevance of
External Beam Radiation Therapy: Simpson’s classification in the modern era. With 373 WHO
Benign Meningiomas grade I meningioma patients followed for a median of
3.7 years, they found no significant difference in 5-year
Benign Meningiomas progression-free survival (PFS) following Simpson grade I–
IV resections, with respective 5-year PFS rates of 95, 85, 88,
Although the majority of meningiomas are benign and 81 % (p = ns) [48]. Similar findings were previously
(WHO grade I) and slow-growing, they may become large reported by Condra et al. [46] and more recently by Oya
enough over time to cause devastating neurologic deficits et al. [49]. Even though these studies did not identify a dif-
ference in local control after Simpson grade I–III resections,
they did generally reveal shorter PFS following Simpson
Table 22.1 Grading criteria for meningiomas [43] grade IV surgery [46, 49]. To directly address the predictive
Benign meningioma (WHO grade I)
value of resection extent, Hasseleid et al. [50] reviewed 391
Any predominant histology other than clear cell, chordoid, papillary, patients with convexity meningiomas and were able to iden-
or rhabdoid tify significant outcome differences between Simpson grade
Lacks criteria of atypical, brain invasive, and anaplastic meningioma I, grades II + III, and grade IV + V resections, continuing to
Atypical meningioma (WHO grade II) (any of two major criteria) support the relevance and applicability of Simpson’s criteria.
Mitotic index ≥4 mitoses/10 high-powered fields (HPF) Therefore, gross tumor resection (GTR, Simpson I–III)
At least three of the following five parameters:
• Sheeting architecture (two dimensional sheets with loss of whorls
remains the objective of meningioma surgery, and is achiev-
and fascicles) able in approximately one-half to two-thirds of patients [47,
• Small cell formation (clusters of lymphocyte-like cells with high 51], depending on the tumor location [52]. For these situa-
N/C ratio) tions, where complete and safe tumor removal is not possi-
• Hypercellularity
• Macronucleoli
ble, an effective strategy is STR followed by postoperative
• Spontaneous necrosis (e.g., not induced by embolization or radiation therapy or radiation therapy alone.
radiation) For benign meningioma, GTR is considered as definitive
Brain-invasive meningioma (WHO grade II) treatment, but with extended follow-up, recurrences develop
Tongue-like protrustions of meningioma into adjacent brain in approximately 5–30 % of patients (Table 22.3 and
parenchyma Fig. 22.1). For example, among 145 patients in whom GTR
Meningioma containing islands of entrapped GFAP-positive
parenchyma was achieved, Mirimanoff et al. [51] reported 5-, 10-, and
Anaplastic (malignant) meningioma (WHO grade III) (either of two 15-year recurrence rates of 7 %, 20 %, and 32 % and second
criteria) operation rates of 6 %, 15 %, and 20 %, respectively. These
Mitotic index ≥20 mitoses/10 HPF rates were confirmed by a large series from the Mayo Clinic
Frank anaplasia (sarcoma-, carcinoma-, or melanoma-like histology) [12]. Condra et al. of the Unversity of florida, [46] reported
Fig. 22.1 Progression-free survival reported in 35 selected studies that and more consistent outcomes in both cohorts of patients treated with
included fractionated external beam radiation therapy by year of publi- combined therapy (STR + RT) and RT alone. Advancements in surgical
cation. Please note that technological advances in imaging, treatment techniques over the same time period did not yield similar improve-
planning and delivery primarily after the year 2000 resulted in superior ments in outcomes
Table 22.2 Simpson grades of resection based on 265 patients Recurrence rates following STR alone are substantially
Resection grade Definition Recurrence (%) higher (Table 22.3 and Fig. 22.1). For example, among 116
1 GTR of tumor, dural attachments 9 patients with STR, Stafford et al. [12] found recurrences in
and abnormal bone 39 % at 5 years and 61 % at 10 years. Condra et al. [46]
2 GTR of tumor, coagulation 19 documented local recurrences in 47 %, 60 %, and 70 % of
of dural attachments patients with STR at 5, 10, and 15 years, respectively.
3 GTR of tumor without resection or 29
Overall, it is estimated that 40–50 % of patients with STR
coagulation of dural attachments or
extradural extensions (e.g., invaded develop local progression within 5 years, 60 % within 10
or hyperostotic bone) years, and at least 70 % within 15 years [46, 47]. In a recent
4 Partial resection of tumor 44 study, Jensen et al. [54] reported that STR was associated
5 Simple decompression (biopsy) – with both poorer PFS and overall survival (OS). Despite
GTR gross total resection these data, patients are frequently observed following
From: Simpson D. The Recurrence of Intracranial Meningiomas After STR. For example, in a Mayo Clinic series of 581 patients,
Surgical Treatment. J Neurol Neurosurg Psychiatry. 1957;20:22–39;
116 (20 %) of whom has STR, only 10 (9 %) received adju-
used with permission
vant radiation therapy [12].
For cases where GTR is not possible, radiation therapy
the importance of complete excision but found no associa- (RT), whether fractionated external beam radiation therapy
tion between Simpson grade and local control or cause-spe- (EBRT) or stereotactic radiosurgery (SRS), is currently the
cific survival, as long as GTR (Simpson grades 1, 2, and 3) only accepted nonsurgical treatment option. Historically,
was achieved. For patients treated with surgery alone, GTR meningiomas were considered to be resistant to RT, and
resulted in 5-, 10-, and 15-year actuarial recurrence rates of authors of many older retrospective studies reported infre-
7 %, 20 %, and 24 %, respectively [46]. In a more recent quent radiographic regression [55]. Additionally, concerns
report by Soyuer et al., the respective 5-, 10-, and 15-year about malignant degeneration were cited as a reason to
rates were 23 % 39 %, and 60 %, the higher recurrence rates refrain from either definitive or postoperative RT [29, 56].
(Table 22.3) likely reflecting the more routine use and Consequently, many patients with inoperable or subtotally
improved quality of neuroimaging [53]. resected meningiomas were observed. At present, malignant
Table 22.3 Selected studies of fractionated external beam radiation therapy for patients largely with WHO grade I or image-diagnosed (presumed
WHO grade I) meningiomas
Progression-free survival Clinical Tumor Late Time
GTR STR STR + RT improvement shrinkage toxicity point
Author Year N Technique (%) (%) RT (%) (%) (%) (%) (%) (years)
Adegbite et al. 1983 114 EBRT 74 34 82 10
Mirimanoff et al. 1985 225 EBRT 80 45 10
Barbaro et al. 1987 135 EBRT 96 40 68 0 Crude
Taylor et al. 1988 132 EBRT 77 18 82 10
Glaholm et al. 1990 117 EBRT 96 43 77 46 38 10
Miralbell et al. 1992 115 EBRT 48 88 16 8
Mahmood et al. 1994 254 EBRT 98 62 5
Goldsmith et al. 1994 117 EBRT 77 3.6 10
98a
Peele et al. 1996 86 EBRT 52 100 5 Crude
Condra et al. 1997 246 EBRT 80 40 87 24 10
Stafford et al. 1998 581 EBRT 75 39 10
Nutting et al. 1999 82 EBRT 83 14 10
Vendrely et al. 1999 156 EBRT 79 59 29 8
Pourel et al. 2001 26 EBRT 76 2.2 8
Dufour et al. 2001 31 EBRT 93 71 29 3.2 10
Jalali et al. 2002 41 FSRT 100 26.8 22 12.1 3
Uy et al. 2002 40 IMRT 93 23 5 5
Pirzkall et al. 2003 20 IMRT 100 60 25 0 3
Soyuer et al. 2004 92 EBRT 77 38 91 2.5 10
Selch et al. 2004 45 FSRT 97 20 18 0 3
Milker-Zabel et al. 2005 317 IMRT 89 42.9 23 0 10
Henzel et al. 2006 224 FSRT 97 43.4 46 0 3
Milker-Zabel et al. 2007 94 IMRT 94 39.8 20 4 4.4
Hamm et al. 2008 183 FSRT 93 23.2 2.7 3
Litre et al. 2009 100 FSRT 94 94 50–81 9 0 5
Korah et al. 2010 41 FSRT 94 3 5
Metellus et al. 2010 53 FSRT 94 94 58.5 30 1.9 10
Bria 2011 60 FSRT 95 60 1
96 3
Minniti 2011 52 FSRT 93 20 23 5.5 5
Mahadevan 2011 16 FSRT 100 19 4 2
Morimoto 2011 31 FSRT 87 1 5
Onodera 2011 27 FSRT 96.2 5.3
Ohba 2011 281 FSRT/SRS 88.3 63.7 92.3 5
96 99 5
Tanzler 2011 146 EBRT/ 93 99 6.8 10
FSRT/SRS
Paulsen 2012 109 FSRT 98 21 5 5
a
Goldsmith et al., 10-year PFS 98 % with treatment after 1980 when CT and MRI began to be used for treatment planning, versus 77 % before 1980
GTR gross tumor resection, STR subtotal tumor resection, RT radiotherapy, EBRT external beam radiation therapy, IMRT intensity-modulated
radiation therapy, FSRT fractionated stereotactic radiation therapy, SRS stereotactic radiosurgery (1–5 fractions)
degeneration has not been explicitly linked to RT [32, 51, and frequently in tumor-related neurologic symptoms, was
57]; rather, it is believed to be related to the natural history of observed despite the fact that only about one-fourth to one-
a subset of meningiomas [58, 59]. third of treated meningiomas exhibit tumor shrinkage
Most retrospective studies since the 1980s have demon- (Table 22.3).
strated that, for benign meningioma, postoperative RT after Radiotherapy can also be used as a primary treatment for
STR improves local control (Table 22.3 and Fig. 22.1), and meningiomas diagnosed radiographically or by biopsy.
perhaps even survival [46, 60]. Improvement in local control, In one of the earliest reports, Glaholm et al. [29, 56] reported
the outcome of 32 patients treated with primary RT without noted complications in 3.6 % of 140 patients treated with RT
resection at the Royal Marsden Hospital; 47 % disease-free after STR from 1967 to 1990. The observed RT-related tox-
survival (DFS) was observed at 15 years. This was only mod- icity in this study included retinopathy, optic neuropathy, and
estly lower than a corresponding 56 % DFS in patients cerebral necrosis. Further analysis of the data by Goldsmith
treated with STR + RT. More recent series demonstrate market al. [67] permitted construction of a model to predict that
edly improved local control rates and PFSs, >93 % after optic nerve tolerance is 890 optic ret, equivalent to 54 Gy/30
5–10 years of follow-up (Table 22.3) [61–64]. These out- fractions. This threshold dose is supported by only rare
comes compare favorably to those observed after GTR for observations of optic nerve injury with doses lower than
benign meningiomas (Fig. 22.1) and, to a certain degree, 54 Gy, particularly with fractional doses ≤2.0 Gy [56, 68,
challenge the paradigm that surgery is the mainstay of 69]. Additionally, Uy et al. [70] noted no anterior optic path-
meningioma management. way injury with a median dose of 50.4 Gy and fractions of
The discrepant outcomes of older studies (before 2000) 1.7–2.0 Gy. Shrieve et al. [71] confirmed these findings as
versus more recent RT studies, whether in the postoperative well as the benefit of fractional doses at or lower than
or primary setting, may be attributed to the technical advance- approximately 2.1 Gy.
ments in RT treatment planning and delivery techniques. Other cranial neuropathies (non-optic) are uncommon
This trend is clearly illustrated in Fig. 22.1 for PFS, and in with modern treatment techniques but have been reported in
Table 22.3 for RT-related late neurologic toxicity as well as older series. For example, Selch et al. [72] found no
neurologic improvement. These observations are more RT-related neuropathy in 45 patients treated for cavernous
directly supported by the analysis of Goldsmith et al. [65] of sinus meningiomas to median dose of 50.4 Gy (1.8 Gy/frac-
140 patients treated at the University of California, San tion). Similar findings were noted earlier by Urie et al. [73].
Francisco to a median dose of 54 Gy in the postoperative set- Brain or brainstem injury, including necrosis, is quite uncom-
ting where a PFS rate of 98 % was noted in those treated after mon in the modern era with conventional doses and tech-
1980 (when computed tomography [CT] and/or magnetic niques. Similarly, edema is generally not observed following
resonance imaging [MRI] was used for planning therapy) fractionated RT, unlike single-session SRS treatment. For
compared to a rate of 77 % for patients treated before 1980 this reason, fractionated RT may be preferred over SRS for
when advanced planning techniques were not available. tumors in locations where edema is likely to develop and to
Newer treatment series also report the use of intensity- significant morbidity (e.g., parasagittal, cerebello-pontine
modulated RT (IMRT) as well as fractionated stereotactic RT angle, and/or large meningiomas).
(FSRT) (Table 22.3). IMRT techniques allow delivery of
dose via multiple convergent beams (Fig. 22.2), each of
which can be divided into zones of different intensities, so Optic Nerve Sheath Meningiomas
that the integral dose can tightly conform to the shape of the
target. This can potentially minimize unnecessary irradiation The management of optic nerve sheath meningiomas
of the surrounding normal tissues, including the critical ante- (ONSMs) is almost exclusively nonsurgical. These rare
rior optic structures. FSRT is characterized by highly accu- tumors comprise <2 % of all meningiomas and arise from the
rate patient immobilization setup, use of a stereotactic meningeal lining of the optic nerves [74, 75]. They generally
coordinate system, and sub-millimeter mechanical and beam grow very slowly within the subarachnoid space and the con-
tolerances. All modern RT techniques rely on the use of fines of the optic canal, eventually compressing the optic
sophisticated imaging for treatment planning and verifica- nerve itself and/or its vasculature. Patients usually present
tion. It is believed, but not yet definitively proven, that these with progressive vision loss, although some exhibit rapid loss
newer treatment methods will continue to improve patient of vision. In a minority of patients, ONSMs are found inci-
outcomes and reduce RT-related toxicity. dentally on imaging. Less common complaints include visual-
It is worth noting that serious side effects of modern field defects, loss/alterations of color vision, and orbital pain.
methods of treatment planning and delivery are already quite Resection of ONSMs while keeping optic nerve in place
uncommon. For example, Debus et al. [66] reported on 189 has been tried [76–80], but carries an unacceptably high
patients treated with FSRT for large skull base meningiomas, risk of visual complications and local recurrence. Resection
using median daily fractions of 1.8 Gy to a mean dose of commonly impairs the blood supply and results in blindness,
56.8 Gy. With a nearly 3-year median follow-up, the rate of even though the nerve is left grossly intact. For these rea-
clinically significant neuro-toxicity (grade 3) was 2.2 % sons, fractionated RT has become the centerpiece of ONSM
(four patients), and 1.7 % (three patients) in the absence of a management. In one of the largest series published, Turbin
preexisting neurologic deficit. The neurologic deficits et al. [80] reported on 64 patients with ONSMs and found
reported in the study included reduced vision, a new visual- that fractionated RT alone provided more favorable out-
field loss, and trigeminal neuropathy. Goldsmith et al. [65] comes than observation, surgery alone, or a combination of
Fig. 22.2 A representative intensity-modulated radiation therapy GTV is shaded red; a dose of 54 Gy/30 fractions was prescribed to the
(IMRT) plan for a right cavernous sinus meningioma (WHO grade I) in 95 % isodose line to minimize the risk of injury to the anterior optic
a 52 years old patient with slight right temporal field deficit thought to be structures. Note: Maximal point dose for the plan (56.38 Gy) is located
related to disease encroachment on the right portion of the optic chiasm. within the tumor volume, and the maximum dose to the chiasm is <54 Gy
surgery + RT. Patients in the RT only group were the only ones from Japan, Narayan et al. [81] evaluated 14 patients with
that did not demonstrate a statistically significant decline in ONSMs who were treated with conformal, fractionated RT
visual acuity after a mean follow-up of 11.5 years (range, to 50.4–56 Gy. After a median follow-up of 4.1 years, 86 %
4.75–23 years); all other groups demonstrated a significant of the treated patients had either improved or stable visual
decrement in visual acuity, including complete visual loss. acuity. These studies and six additional studies with simi-
Irradiated patients received 40–55 Gy of fractionated RT lar results have established highly conformal, fractionated
(1.8–2.0 Gy/fraction) over approximately 5 weeks. In a study RT as the first-choice treatment modality for patients with
ONSMs. In aggregate, the local control is excellent (95 %); ations with RFS and OS were highly significant, with grade II
early visual improvement (<3 months from the completion meningiomas having a nearly eightfold increased risk of sub-
of RT) is attained in about half of the patients (54.7 %); and sequent recurrence after GTR and a slightly increased mor-
complications of therapy are relatively uncommon [74, 82]. tality rate.
It is our policy to treat symptomatic ONSM patients with Our ability to define impact of treatment on outcomes of
highly conformal, fractionated RT to doses of 50.4–54 Gy atypical meningioma patients is problematic since much of
using daily fractions of 1.8 Gy; however, there is no uni- the published literature consists of retrospective reports and
form agreement as to whether patients with stable ONSMs predates the new WHO classification criteria, which results
and stable vision should be approached with observation or in relative under-reporting of atypical meningioma inci-
early RT. Patients with bilateral disease are generally treated dence. There is general agreement, but not consensus, that
whereas those with unilateral tumors are observed provided STR is insufficient treatment for WHO grade II meningio-
that they can be expected to comply with serial visual field mas. For example, surveys of neurosurgeons in Germany
testing, and regular clinical and imaging follow-up. [85] and United Kingdom [86] indicated that 26 and 41 %,
respectively, do not recommend adjuvant radiotherapy after
STR of an atypical meningioma. These reports are corrobo-
Recurrent Meningiomas rated by publications such that by Goyal et al. [87] who
reported 17 % 10-year local control rate for atypical menin-
Recurrent meningiomas after surgery and/or RT exhibit giomas at a single institution, but despite this, was unable to
markedly increased progression rates over newly diagnosed document a significant benefit associated with the use of
tumors [46, 51, 68, 83], with mean interval to clinical recur- postoperative radiation therapy. Neurosurgeons often employ
rence of 4 years [51, 55]. Miralbell et al. [68] reported a a strategy of serial re-resection to manage tumor recurrence.
78 % PFS at 8 years in patients treated with surgery + RT for There is even less agreement regarding adjuvant treat-
recurrent tumors, compared to only 11 % with surgery alone. ment following GTR. In Germany [85], 84 % of centers
Similarly, Taylor et al. [83] reported on 30 patients who (47 of 56) recommended surgery alone for initially diag-
developed recurrent disease after their initial treatment; 15 nosed, gross-totally resected WHO grade II meningioma,
underwent surgery alone, 10 surgery + RT, and 5 either and in United Kingdom [86], 80 % of centers espoused such
refused further treatment or were deemed inoperable. Local a practice. Several retrospective reports [84, 87–90] have
control was 30 % at 10 years in patients who had surgery supported this approach by suggesting that GTR alone is
alone at first recurrence, versus 89 % for patients who sufficient and that adjuvant RT adds no utility. Jaaskelainen
received salvage radiation, with or without resection. The et al. [58] reported a 38 % 5-year local recurrence after
Taylor et al. [83] study provides compelling support for the GTR, and did not find a benefit to adjuvant RT. However, no
use of salvage RT to doses of 54–59.4 Gy (1.8 Gy/fraction, randomized trials have been completed to define an appro-
or similar regimens at 2 Gy/fraction) not only because of priate care pattern in this patient population, and many of
superior local control, but also improved survival, at 10 years the cited studies are small, retrospectively analyzed, and
89 % for surgery + RT compared to 43 % with surgery alone. predate current WHO grading criteria. Additionally, many
These data corroborate aggressive treatment for recurrent of the RT doses used in these studies were too low to be
meningiomas, even at first recurrence. effective.
A modern study by Aghi et al. [40] in 108 patients with
atypical meningioma (per 2000/2007 WHO diagnostic crite-
EBRT: Atypical (WHO Grade II) Meningiomas ria) noted a 5-year local recurrence rate of 50 % following
Simpson grade I surgical resection alone, whereas there were
Atypical, WHO grade II, meninigiomas account for approxi- no recurrences in the small cohort (n = 8) receiving postop-
mately 20 % to 25 % of meningiomas [17]. In one recent erative RT. Another recent report by Komotar et al. [91]
analysis, the rate of atypical meningioma rose to 35 % after reviewed outcomes among 45 patients with a gross-totally
adoption of the modern WHO grading standards [84]. resected (Simpson grade I and II) atypical meningioma, con-
However, in the literature atypical meningiomas are a hetero- firmed by postoperative imaging, and found a 65 % actuarial
geneous group, owing largely to the imprecise pathologic local recurrence rate at 6 years without, versus 20 % with
classification criteria and varying classification schemes over adjuvant RT. For patients with recurrent WHO grade II
time. Given the recent changes in diagnostic criteria [43, 44], meningioma, Mair et al. found that neither salvage surgical
the natural history of atypical meningiomas is now better resection extent nor salvage RT was predictive of outcome.
characterized. Systematic application of these criteria across This is consistent with results from Aghi and colleagues [40]
multiple centers will be critical toward achieving concordance where 10-year disease-specific survival after first recurrence
and defining subsequent management. The statistical associ- of 69 % even with salvage therapies. Furthermore, Kotomar
et al. [91] concluded that recurrences resulted in shortened In contrast, other studies have reached different conclu-
overall survival, and significant additional treatment burden. sions regarding the role of adjuvant RT. Mair et al. [90] sug-
Many investigators have recommended irradiation, irre- gested that EBRT was not appropriate following GTR, and
spective of resection extent in this patient population [31, advised SRS rather than EBRT following STR. Despite these
46], but others, like Goyal et al. [86], argued that fractionated differing conclusions, their report confirmed improved PFS
RT has no significant impact on either local control and over- in patients who received adjuvant EBRT (STR + EBRT) com-
all survival. Their retrospective analysis was based on a pared to STR alone cohort. Four-year PFS rates were 72 %
cohort of 22 patients, 15 of whom underwent GTR (Simpson versus 13 % (p = 0.043), respectively. Stratification of resec-
grades 1–3), 4 STR, and 3 resection of unknown extent. tion extent was not done in this study, and the EBRT dose
Eight patients received RT (2 after initial resection and 6 at was comparatively low (51.8 Gy in 28 fractions). Hardesty
the time of recurrence). The median radiation dose was et al. [93] reached a similar conclusion, reporting that adju-
54 Gy (range 35–59.4 Gy). After a median follow-up of vant RT (EBRT or SRS) following “aggressive microsurgical
5.5 years, local control was 87 % at 5 and 10 years following resection” of an atypical meningioma did not significantly
GTR, and RT had no significant impact on local control or impact recurrence risk. However, it is worth noting that no
overall survival. This study highlights some of the difficul- patient with GTR + postoperative RT experienced recurrence,
ties relevant to atypical tumors, including inconsistent use, with a median 52 month’s follow-up. The median EBRT dose
dosing, and timing of RT as well as wide variations in the in this study was 54 Gy in 27–30 fractions.
resection extent [31]. Stessin et al. [94] performed an analysis of SEER data of
In the previously cited study by Aghi et al. [40] of 108 657 patients treated for “non-benign” meningioma between
patients with atypical meningioma and Simpson grade 1 1988 and 2007 (predating the updated WHO classification
GTR, actuarial tumor recurrence rates were 7 % at 1 year, system). Of these patients, 37 % (244 patients) received
41 % at 5 years, and 48 % at 10 years. Most recurrences adjuvant EBRT. When controlling for WHO grade (II versus
occurred within 5 years of surgical resection. Of the 108 III), tumor size, resection extent, and diagnosis date, EBRT
patients, 8 received immediate, adjuvant RT to 59.4–61.2 Gy did not appear to impart a statistically significant overall sur-
(1.5–1.8 Gy/fraction) to the resection bed with a 1 cm margin, vival or DFS benefit. Paradoxically, overall survival was
and experienced no recurrences during the follow-up period. lower in patients who received adjuvant EBRT compared to
This difference did not reach statistical significance (p = 0.1), those who were treated with surgery alone, perhaps reflect-
likely due to the small number of patients (n = 8) in the RT ing selection bias. This study did not consider details of RT,
cohort. The study also noted that 30 patients with recurrence including dose, target volume definition, and other treatment
ultimately received EBRT or SRS, and 73 % underwent repeat parameters.
surgery, with a mean number of craniotomies of 2.7. Only one These apparent contradictions concerning adjuvant RT
meningioma transformed to WHO grade III tumor, but at 7 following either GTR or STR of an atypical (WHO grade II)
years, 33 % of patients died due to recurrent disease. meningioma suggest that RT technique and therapeutic
Similarly, Komotar’s study [91] of 45 patients with atypi- details may be important, including target volume definition,
cal meningioma (following Simpson grade I/II resection) dose/fractionation, and possibly, even mode of delivery. For
suggested a benefit to immediate adjuvant RT at a dose of example, Park et al. [92] found that PFS was improved with
59.4 Gy to the tumor cavity plus 0.5–1.0 cm margin. In this median RT dose of 61.2 Gy. Aghi et al. [40] noted no local
study, 32 patients had GTR alone and an additional 13 under- recurrences with doses of 59.4–61.2 Gy, and Komotar et al.
went immediate adjuvant external beam RT after GTR [91] had noticeably better outcomes with median EBRT dose
(GTR + EBRT). With GTR alone, 41 % (13 patients) recurred of 59.4 Gy. Hug et al. [31] found that local control for an
at a median of 19 months whereas only 8 % (one patient) atypical meningioma was significantly enhanced by cumula-
recurred in the GTR + EBRT group at 52.5 months. This tive doses of at least 60 Cobalt-Gray Equivalents (CGE),
translated into 6-year actuarial recurrence risks of 65 % ver- using protons. Hoffmann et al. [95] identified 10 WHO grade
sus 20 % (p = 0.085) with GTR and GTR + EBRT, respec- II patients and noted a postoperative recurrence rate of 50 %
tively. Other recent analyses similarly supported the use of and suggested a benefit to postoperative EBRT, especially
immediate EBRT in this setting [92]. when radical surgery cannot be achieved, and recommended
These studies provide evidence for upfront postoperative a total dose of 60 Gy.
fractionated RT following GTR of an atypical meningioma. The recently completed RTOG 0539 study used a dose
However, the possibility remains that close clinical and of 54 Gy in 30 fractions for newly diagnosed atypical
radiographic follow-up with delayed salvage RT as needed meningioma following GTR; 60 Gy/30 fractions following
may also be feasible. The key question in this patient subset STR and/or for recurrent WHO grade II tumor of any extent.
is whether early adjuvant or late salvage RT is more appro- Also recently completed was the EORTC 22042–26042 trial,
priate. Existing publications do not answer this question. which employed 60 Gy following GTR, and added a 10 Gy
boost after STR. The results of these trials are expected to multi-modality treatment (surgery + RT). These studies also
generate high-quality evidence to help guide therapy in this demonstrated that immediate, adjuvant EBRT offers a greater
patient population. benefit than EBRT at progression. Milosevic et al. [103]
Optimal dosing regimens, target volume definitions, and found that patients treated to doses <50 Gy experienced infe-
choice among many different radiation modalities are in rior cause-specific survival, as did those treated prior to 1975
need of further study. Data suggest that dose escalation may (before routine use of CT-based planning). Dziuk et al. [97]
be an important factor in non-benign meningiomas but such determined that EBRT improved 5-year PFS from 50 % to
escalation needs to be approached with caution since higher 80 %, and recommended a final dose of 60 Gy in the immedi-
rates of complication are possible [96]. ate post-op setting.
The issue of target volumes in the postoperative setting
for grade III meningiomas remains unresolved, particularly
EBRT: Anaplastic (WHO Grade III) when comparing SRS and EBRT treatments. With SRS,
Meningiomas Pollock et al. [99] reported tumor progression “away from
the original irradiated tumor” in 30 % of patients with atypi-
These aggressive tumors are often diagnosed after surgical cal and anaplastic meningioma, occurring at a median of 15
resection. As with other meningiomas, surgical resection months after SRS. Most reported recurrences were described
extent correlates with recurrence rates [39, 87, 89, 97]. as “marginal,” that is, occurring “adjacent to the irradiated
Outcomes following surgery alone are quite poor; 78 % tumor.” Mattozo et al. [104] evaluated recurrences more sys-
5-year recurrence rate has been reported following GTR for tematically after treatment with SRS and FSRT and found
patients with anaplastic meningioma [58]. In a study from that 77 % of recurrences were within the original surgical
1998 by Dzuik et al. [97], 38 patients with “malignant resection cavity, and recommended that the entire surgical
meningioma” were reported. Eleven of the patients had cavity be targeted, with an SRS boost to the recurrent nodule
hemangiopericytoma. With that caveat, adjuvant irradiation if desired, to reduce the risk of local relapse.
following initial resection improved 5-year DFS to 80 %, As mentioned above, the timing of RT in these patients
versus 15 % with surgery alone (p = 0.002). EBRT also appears to be important. Most studies suggest limited benefit
improved 2-year DFS for recurrent meningioma, but had from salvage RT (compared to immediate post-op RT).
no impact upon 5-year DFS for this subgroup. For this rea- Dziuk et al. [97] reported improved local control with the
son, most practitioners have recommended adjuvant RT addition of post-op RT after initial resection and found
[48, 98, 99]. 2-year PFS improvement with salvage RT. However, there
Sughrue et al. [100] recently evaluated the benefit of was no benefit at 5 years in the salvage setting. Several inves-
resection extent for WHO grade III meningioma; all patients tigators have demonstrated improvement in outcomes with
were referred for postoperative RT. They found that aggres- RT, but the magnitude of improvement has been highly vari-
sive attempts to achieve GTR did not improve survival and able [31, 97–100, 102–107].
often resulted in compromised neurologic outcome. Instead, As with atypical meningiomas, higher RT doses appear to
they found that near-total resection (NTR), defined as >90 % improve local control for patients with anaplastic (WHO
tumor debulking, resulted in improved survival when fol- grade III) histology. On review of WHO grade II/III patients,
lowed by adjuvant RT. In this report, salvage surgery also Milosevic et al. [103] noted improved cause-specific survival
appeared beneficial; a median survival of 53 months with of 42 % with ≥50 Gy versus 0 % with <50 Gy. Similarly,
salvage surgery was noted, as opposed to 25 months without Goldsmith et al. [65] reported 5-year PFS of 63 % using
(p = 0.02). Salvage surgery was often followed by RT or even >53 Gy versus 17 % with ≤53, and Dziuk et al. [97] recom-
I-125 brain brachytherapy [101]. This is in contrast to other mended a dose of 60 Gy, even after GTR. Using either pro-
studies [102] suggesting that mode of salvage therapy for tons or combined protons and photons, DeVries et al. [108]
WHO grade III meningiomas does not favorably impact sub- and Hug et al. [31] demonstrated increased local control and
sequent progression. survival with total doses exceeding 60 Gy. In a mixed group
With only retrospective studies to guide treatment, the of WHO grade II/III tumors, Hug et al. reported a 100 %
true benefits of multi-modality therapy are not known. Many 5-year local control for patients receiving ≥60 CGE versus
published studies use varying definitions of anaplasia, surgi- 0 % for lower doses (p = 0.0006). This corresponded to
cal extent, RT parameters, and outcomes. In aggregate, most 8-year local control of 33 % and 0 %, respectively. For the
studies [46, 53, 97, 102, 103] appear to report measurable WHO grade III subgroup, this translated into an overall sur-
benefit to adjuvant EBRT ± SRS following surgery in ana- vival difference at 5-years of 87 % for ≥60 CGE versus 15 %
plastic meningiomas. for <60 CGE [31]. As mentioned in the atypical meningioma
Milosevic et al. [103] and Dziuk et al. [97] published section, dose escalation should be approached with caution
some of the earliest studies demonstrating benefit from since significant toxicity can occur [96].
device-specific, usually at least 3 mm. Gross-totally resected

EBRT: Technical Factors WHO grade II meningiomas (Simpson grades I–III) may be
treated similar to recurrent grade I meningiomas, although a
Technical improvements have favorably impacted the out- modestly higher dose of approximately 60 (1.8–2.0 Gy) has
come and side effect profile of patients treated for meningi- been proposed in some recent studies [31, 40, 91, 108].
oma with fractionated EBRT. Image-based techniques allow In contrast, patients at high risk for local recurrence
treatments to be delivered with greater precision and confor- require aggressive therapy. These patients include WHO
mality [72, 109]. Local control improvements were substanti- grade III meningioma of any resection extent, a recurrent
ated by Goldsmith et al. [65] and Milosevic et al. [103] during WHO grade II meningioma of any resection extent, or subto-
analysis of their series that spanned the transition from CT- to tally resected, newly diagnosed WHO grade II meningioma
MRI-based planning. Goldsmith et al. [65, 110] found that (Simpson grade IV–V). In these patients, GTV is defined as
CT- and/or MRI-based target definition, along with appropri- the tumor bed on postoperative, contrast-enhanced MR
ate immobilization, resulted in 21 % improvement in PFS imaging and includes any residual nodular enhancement.
(p = 0.002). This can be seen more broadly in published stud- Again, neither cerebral edema nor the dural tail should be
ies of meningioma treatments, as shown in Fig. 22.1, with specifically included in the GTV. Two target volumes are
better clustering of reported treatment outcomes in the mod- subsequently defined: CTV54Gy = GTV + 2 cm, and
ern era (post 2000) in which MRI-based targeting and image- CTV60Gy = GTV + 1 cm. Appropriate PTV expansions must
guided delivery have been commonplace. be added to CTV as above. Using IMRT, a concomitant boost
technique is possible, for instance—similar to the approach
used in the recent RTOG trial (RTOG-0539)—a PTV54Gy
EBRT: Target Volumes and Dose may be treated to 54 Gy, while a larger PTV60Gy receives
60 Gy, both in 30 daily fractions. Some studies have reported
Target volumes are usually based on postoperative, contrast- improved outcomes for atypical and malignant meningioma,
enhanced MRI. Preoperative imaging is sometimes helpful particularly following STR, using higher doses [108]. The
in identifying disease extent when postoperative changes recently completed EORTC trial (22042–26042) employed
complicate interpretation. 60 Gy, but added a 10 Gy boost for a final dose of 70 Gy in
Image-diagnosed meningiomas can be treated with defin- 35 fractions, following STR of either a WHO grade II or III
itive intent using fractionated EBRT to a dose of 54 Gy in meningioma. As in any RT planning, care must be exercised
27–30 daily fractions. In these cases, the gross tumor volume regarding location of doses beyond 60 Gy as this may cause
(GTV) is defined as the enhancing tumor mass on a significant neurologic toxicity [96].
T1-weighted contrast-enhanced MR imaging. Clinical target
volume (CTV) is generally the same as GTV. Treatment
device-appropriate planning target volume (PTV) expansion EBRT: Toxicity
should be added to GTV to ensure precise treatment (e.g.,
GTV + 0.3 cm = PTV). EBRT side effects experienced by patients treated for menin-
For gross-totally resected WHO grade I meningioma, adju- giomas are few with approximately 5 % of patients experi-
vant RT is generally not indicated (pending results of the encing more than transient neurologic morbidity. Early
RTOG 0539). However, the benefits of RT following initial publications identified no deleterious treatment-related side
STR demand further clarification. At present either observa- effects [83, 113], but with longer follow-up injury to anterior
tion or early adjuvant RT can find support. Recurrent WHO visual structures and select cranial nerves can develop. This
grade I meningioma tends to exhibit a more aggressive clinical was apparent in a report by Debus et al. [66] of 189 patients
course and, irrespective of resection extent, should likely be treated with FSRT using median daily fractions of 1.8 Gy to
treated with EBRT to a dose of at least 54 Gy in 27–30 daily a mean cumulative dose of 56.8 Gy. With a median follow up
fractions, or with SRS using a single dose or single-fraction of nearly 3 years, they identified clinically significant (grade
equivalent of approximately 14 Gy. In the recurrent setting, 3) toxicity in 2.2 % [4] of patients, lower at 1.7 % [3] in the
the GTV is defined by the tumor bed and any gross residual/ absence of preexisting neurologic deficit. Reduced vision, a
persistent nodular enhancement. Neither cerebral edema nor visual-field cut, or trigeminal neuropathy was most common.
the dural tail [111, 112] need be specifically included in the This represents substantial improvement over 38 % rate
GTV. CTV represents a 1 cm expansion (anatomically con- reported by Al-Mefty et al. [114] using older RT techniques.
strained) beyond the GTV, and may be reduced around natural Similarly, Goldsmith et al. reported a 3.6 % (5 of 140) rate
barriers to growth such as uninvolved, non-hyperostotic skull. of toxicity attributed to treatment in patients treated with
The PTV consists of a geometric expansion beyond the CTV EBRT for subtotally resected meningioma. They noted reti-
to account for setup and treatment delivery uncertainties and is nopathy in two, optic neuropathy in one, and cerebral necro-
sis in two patients [65]. In a separate publication, Goldsmith and up to 20 years after primary treatment. Most patients
et al. [67] constructed a model to predict optic nerve toler- were treated using a three-field technique (one AP and two
ance and recommended a maximum dose of 54 Gy in 30 lateral beams). Memory impairment was found in 80 % of
fractions. the patients, and one-third manifested difficulty with visual-
Non-visual pathway deficits can occur [66] but are motor speed, frontal lobe executive functions, and fine motor
uncommon. Selch et al. found no treatment-related cranial coordination. Cognitive outcomes appeared to correlate with
neuropathies of any kind in 45 cavernous sinus meningioma the total dose of radiation delivered but not with the volume
patients treated with fraction sizes of 1.7 or 1.8 Gy per day of the brain irradiated. The pattern of test findings was most
to a total dose of 50.4 Gy [72]. Urie et al. [73] concurred consistent with radiation injury to subcortical white matter.
that treatment-related neurologic deficits in this dose range These two apparently conflicting studies suggest that pre-
are rare. treatment cognitive deficits, the timing of evaluations, the
Cerebral necrosis is also uncommon but has been length of follow-up, and the choice of assessments are
observed [68, 70, 114]. Many of the findings in the Al-Mefty important. To date, there are no robust prospective studies of
et al. [114] study in a series of 58 patients with a variety of cognitive function in patients treated with EBRT for skull
skull base, parasellar, or pineal region tumors who devel- base tumors. Given the relatively long survival of these
oped late brain parenchymal changes (29 %) ranging from 4 patients, it is expected that long-term sequelae of therapy
months to 23 years after treatment are likely related to anti- will manifest and will guide therapeutic choices in the future.
quated treatment techniques. Modern series do not report
significant toxicity at doses <60 Gy delivered over 30 treat-
ment days provided that “hotspots” in excess of the prescrip- Conclusion
tion dose are placed within tumor or other safe locations
(e.g., surgical cavity). In general, integral doses to the tem- EBRT is highly effective in the management of meningiomas
poral lobes and to large volumes of untargeted brain are and long-term data clearly indicate excellent tumor control
much lower with current conformal and image-guided tech- rates, with an acceptable incidence of complications. FSRT
niques. In particular, image-guided therapy allows radiation techniques are capable of delivering highly localized irradia-
oncologists to reduce treatment (PTV) margins by ensuring tion compared with conventional radiotherapy and have the
precise daily setup. potential to reduce long-term radiation morbidity. EBRT and
Additionally, pituitary dysfunction [68, 69, 114, 115], SRS result in very similar local control rates, and either can
cerebrovascular events [56, 68, 72], second malignancy, be recommended for many patients. EBRT is suitable for a
orbital fibrosis [116], edema [72], and other sporadic toxici- broader range of patients, whereas SRS achieves excellent
ties have been reported. Toxicity rates (at various time points) local control outcomes in a well-defined cohorts. SRS is usu-
are summarized in Table 22.3 and range from 0 % to 8 % in ally limited by tumor diameter and volume restrictions to
the modern era (after year 2000). It is worth noting that the ensure safe treatments. Common relative constraints (vary
retrospective nature of most series often precludes reliable by institution) on diameter have been 3.0–3.5 cm, with total
accounting of treatment-related toxicity since reliable grad- volumes of 7.5–15 cm3 [106, 119]. Pollock et al. [119] found
ing of these events is often not possible in retrospect. that SRS provides equivalent tumor control to a Simpson
grade I resection with small- to medium-sized meningiomas
of <3.5 cm in average diameter and <15 cm3. Kondziolka
EBRT: Cognitive Morbidity et al. [106] reported excellent outcomes for tumors of
<3.0 cm in diameter or 7.5 cm3 volume. The primary reasons
Some retrospective studies have identified personality for volume restriction, as suggested by Ramsey et al. [120],
changes [70] and memory loss [46, 68, 115, 116] as compli- is to reduce edema risk that is often associated with SRS
cations of EBRT. Steinvorth et al. [117] prospectively evalu- treatment of convexity or large meningioma, or those that
ated cognitive outcomes in patients with skull base exhibit significant pretreatment edema. Edema risks are
meningiomas after FSRT. They used a comprehensive bat- markedly lower with EBRT, even for large tumors.
tery of neurocognitive tests before, after first fraction, at An additional constraint on SRS is the anterior visual
completion, and 12 months after FSRT. They observed, after pathway which is exquisitely radiation sensitive. At many
the first fraction, a transient decline in memory and a relative radiosurgery centers, point doses to the optic nerves and chi-
increase in attention. There was no cognitive deterioration asm are limited to 8 or 9 Gy, although some suggest that
with additional follow-up. In another study of cognitive out- small volume of the optic nerve can tolerate up to 12 Gy
comes in skull base tumors, Meyers et al. [118] evaluated 19 [121]. Even at a higher tolerance limit, single-fraction SRS
patients who received treatment to the paranasal sinuses to a would be excluded from treatment of ONSMs, as well as
median radiation dose of 60 Gy in 30 daily fractions with a some parasellar and orbital tumors. In aggregate, SRS can be
comprehensive neurocognitive test battery at least 20 months safely and effectively applied to smaller meningiomas at
adequate distance from the anterior visual pathways with 14. Perry A, Dehner LP. Meningeal tumors of childhood and infancy.
limited risk and side effects profiles. An update and literature review. Brain Pathol. 2003;13:386–408.
15. Jung HW, Yoo H, Paek SH, Choi KS. Long-term outcome
EBRT does not suffer from these limitations when com- and growth rate of subtotally resected petroclival meningiomas:
monly accepted fractionation regimens are used. EBRT car- experience with 38 cases. Neurosurgery. 2000;46:567–74.
ries only a small risk of edema or optic neuropathy. discussion 574.
Furthermore, FSRT allows highly conformal and precise 16. Bondy M, Ligon BL. Epidemiology and etiology of intracranial
meningiomas: a review. J Neurooncol. 1996;29:197–205.
delivery, with excellent local control and expected improve- 17. Louis DN, Ohgaki H, Wiestler OD, et al. The 2007 WHO classifi-
ments in observed neurologic and cognitive toxicities. cation of tumours of the central nervous system. Acta Neuropathol.
Despite proven effectiveness of EBRT, there is a relatively 2007;114:97–109.
paucity of prospective, high-quality, randomized data neces- 18. Wahab M, Al-Azzawi F. Meningioma and hormonal influences.
Climacteric. 2003;6:285–92.
sary to define optimal care pattern for patients with menin- 19. Perry A. Unmasking the secrets of meningioma: a slow but
giomas. RTOG 0539 and EORTC 22042–26042 have rewarding journey. Surg Neurol. 2004;61:171.
recently closed to accrual, and are a step in the right direction 20. Ragel B, Jensen RL. New approaches for the treatment of refrac-
as we expect these studies to help define standard-of-care for tory meningiomas. Cancer Control. 2003;10:148–58.
21. Vernooij MW, Ikram MA, Tanghe HL, et al. Incidental findings on
these patients. There still remain many questions as well as brain Mri in the general population. N Engl J Med. 2007;357: 1821–8.
questions relating to optimal management, quality-of-life 22. Heth JA, Al-Mefty O. Cavernous sinus meningiomas. Neurosurg
and cognitive function. As these and other questions are Focus. 2003;14:e3.
addressed, the management of patients with meningioma and 23. Commins DL, Atkinson RD, Burnett ME. Review of meningioma
histopathology. Neurosurg Focus. 2007;23:E3.
the role of EBRT to be more clearly defined. 24. Buetow MP, Buetow PC, Smirniotopoulos JG. Typical, atypical,
and misleading features in meningioma. Radiographics. 1991;11:
1087–106.
References 25. Perry A. Meningiomas. In: McLendon RE, Rosenblum MK,
Bigner DD, editors. Russell and Rubinstein’s pathology of tumors
of the nervous system. London: Hodder Arnold; 2006. p. 427–74.
1. Hoffman S, Propp JM, McCarthy BJ. Temporal trends in inci- 26. Annegers JF, Laws Jr ER, Kurland LT, Grabow JD. Head trauma
dence of primary brain tumors in the United States, 1985–1999. and subsequent brain tumors. Neurosurgery. 1979;4:203–6.
Neuro Oncol. 2006;8:27–37. 27. Inskip PD, Mellemkjaer L, Gridley G, Olsen JH. Incidence of
2. Claus EB, Bondy ML, Schildkraut JM, Wiemels JL, Wrensch M, intracranial tumors following hospitalization for head injuries
Black PM. Epidemiology of intracranial meningioma. (Denmark). Cancer Causes Control. 1998;9:109–16.
Neurosurgery. 2005;57:1088–95. discussion 1088. 28. Ron E, Modan B, Boice Jr JD. Mortality after radiotherapy for
3. Kuratsu J, Kochi M, Ushio Y. Incidence and clinical features of ringworm of the scalp. Am J Epidemiol. 1988;127(4):713–25.
asymptomatic meningiomas. J Neurosurg. 2000;92:766–70. 29. Ron E, Modan B, Boice Jr JD, et al. Tumors of the brain and ner-
4. Kuratsu J, Ushio Y. Epidemiological study of primary intracranial vous system after radiotherapy in childhood. N Engl J Med.
tumors: a regional survey in Kumamoto Prefecture in the southern 1988;319(16):1033–9.
part of Japan. J Neurosurg. 1996;84:946–50. 30. Ghim TT, Seo J, O’Brien M, Meacham L, Crocker I, Krawiecki
5. Wrensch M, Minn Y, Chew T, Bondy M, Berger MS. Epidemiology N. Childhood intracranial meningiomas after high‐dose irradia-
of primary brain tumors: current concepts and review of the litera- tion. Cancer. 1993;71:4091–5.
ture. Neuro Oncol. 2002;4:278–99. 31. Hug EB, Devries A, Thornton AF, et al. Management of atypical
6. CBTRUS. Statistical report: Primary brain tumors in the United and malignant meningiomas: role of high-dose, 3D-conformal
States, 1997–2001. Hinsdale: Central Brain Tumor Registry of the radiation therapy. J Neurooncol. 2000;48:151–60.
United States. 2004 32. Strojan P, Popovic M, Jereb B. Secondary intracranial
7. DeMonte F, Marmor E, Al-Mefty O. Meningiomas. Brain tumors: meningiomas after high-dose cranial irradiation: report of five
an encyclopedic approach. New York: Churchill Livingstone; 2001. cases and review of the literature. Int J Radiat Oncol Biol Phys.
8. DiBiase SJ, Kwok Y, Yovino S, et al. Factors predicting local 2000;48:65–73.
tumor control after gamma knife stereotactic radiosurgery for 33. Al-Mefty O, Topsakal C, Pravdenkova S, Sawyer JR, Harrison
benign intracranial meningiomas. Int J Radiat Oncol Biol Phys. MJ. Radiation-induced meningiomas: clinical, pathological, cyto-
2004;60:1515–9. kinetic, and cytogenetic characteristics. J Neurosurg. 2004;100:
9. Flickinger JC, Kondziolka D, Maitz AH, Lunsford LD. Gamma 1002–13.
knife radiosurgery of imaging-diagnosed intracranial meningi- 34. Rowe J, Grainger A, Walton L, Silcocks P, Radatz M, Kemeny A.
oma. Int J Radiat Oncol Biol Phys. 2003;56:801–6. Risk of malignancy after gamma knife stereotactic radiosurgery.
10. Nakasu S, Hirano A, Shimura T, Llena JF. Incidental meningio- Neurosurgery. 2007;60:60–5. discussion 65.
mas in autopsy study. Surg Neurol. 1987;27:319–22. 35. Borovich B, Doron Y. Recurrence of intracranial meningiomas:
11. Perry A, Giannini C, Raghavan R, et al. Aggressive phenotypic the role played by regional multicentricity. J Neurosurg.
and genotypic features in pediatric and Nf2-associated meningio- 1986;64:58–63.
mas: a clinicopathologic study of 53 cases. J Neuropathol Exp 36. Stangl AP, Wellenreuther R, Lenartz D, et al. Clonality of multiple
Neurol. 2001;60:994–1003. meningiomas. J Neurosurg. 1997;86:853–8.
12. Stafford SL, Perry A, Suman VJ, et al. Primarily resected menin- 37. Perry A, Stafford SL, Scheithauer BW, Suman VJ, Lohse
giomas: outcome and prognostic factors in 581 mayo clinic CM. Meningioma grading: an analysis of Histologic parameters.
patients, 1978 through 1988. Mayo Clin Proc. 1998;73:936–42. Am J Surg Pathol. 1997;21:1455–65.
13. Adegbite AB, Khan MI, Paine KW, Tan LK. The recurrence of 38. Durand A, Labrousse F, Jouvet A, et al. WHO grade II and III
intracranial meningiomas after surgical treatment. J Neurosurg. meningiomas: a study of prognostic factors. J Neurooncol.
1983;58:51–6. 2009;95:367–75.
39. Perry A, Scheithauer BW, Stafford SL, Lohse CM, Wollan PC. 59. Niranjan A, Kondziolka D, Lunsford LD. Neoplastic transforma-
“Malignancy” in meningiomas: a clinicopathologic study of 116 tion after radiosurgery or radiotherapy: risk and realities.
patients, with grading implications. Cancer. 1999;85:2046–56. Otolaryngol Clin North Am. 2009;42:717.
40. Aghi MK, Carter BS, Cosgrove GR, et al. Long-term recurrence 60. McCarthy BJ, Davis FG, Freels S, et al. Factors associated with
rates of atypical meningiomas after gross total resection with or survival in patients with meningioma. J Neurosurg. 1998;88:
without postoperative adjuvant radiation. Neurosurgery. 831–9.
2009;64:56–60. discussion 60. 61. Hamm K, Henzel M, Gross MW, Surber G, Kleinert G, Engenhart-
41. Kleihues P, Louis DN, Scheithauer BW, et al. The WHO classifi- Cabillic R. Radiosurgery/stereotactic radiotherapy in the thera-
cation of tumors of the nervous system. J Neuropathol Exp Neurol. peutical concept for skull base meningiomas. Zentralbl Neurochir.
2002;61:215–25. 2008;69:14–21.
42. Smith SJ, Boddu S, Macarthur DC. Atypical meningiomas: WHO 62. Korah MP, Nowlan AW, Johnstone PA, Crocker IR. Radiation
moved the goalposts? Br J Neurosurg. 2007;21:588–92. therapy alone for imaging-defined meningiomas. Int J Radiat
43. Perry A, Louis DN, Scheithauer BW, Budka H, Von Deimling Oncol Biol Phys. 2010;76:181–6.
A. Meningiomas. In: Louis DN, Ohgaki H, Wiestler OD, Cavenee 63. Litre CF, Colin P, Noudel R, et al. Fractionated stereotactic radio-
WK, editors. WHO classification of tumors of the central nervous therapy treatment of cavernous sinus meningiomas: a study of 100
system. Lyon: IARCC; 2007. p. 164–72. cases. Int J Radiat Oncol Biol Phys. 2009;74:1012–7.
44. Fuller GN, Scheithauer BW. The 2007 revised World Health 64. Metellus P, Batra S, Karkar S, et al. Fractionated conformal radio-
Organization (WHO) classification of tumours of the central ner- therapy in the management of cavernous sinus meningiomas:
vous system: newly codified entities. Brain Pathol. long-term functional outcome and tumor control at a single insti-
2007;17:304–7. tution. Int J Radiat Oncol Biol Phys. 2010;78(3):836–43.
45. Simpson D. The recurrence of intracranial meningiomas after sur- 65. Goldsmith BJ, Wara WM, Wilson CB, Larson DA. Postoperative
gical treatment. J Neurol Neurosurg Psychiatry. 1957;20:22–39. irradiation for subtotally resected meningiomas. A retrospective
46. Condra KS, Buatti JM, Mendenhall WM, Friedman WA, Marcus analysis of 140 patients treated from 1967 to 1990. J Neurosurg.
RBJ, Rhoton AL. Benign meningiomas: primary treatment selec- 1994;80:195–201.
tion affects survival. Int J Radiat Oncol Biol Phys. 66. Debus J, Wuendrich M, Pirzkall A, et al. High efficacy of fraction-
1997;39:427–36. ated stereotactic radiotherapy of large base-of-skull meningiomas:
47. Pollock BE, Stafford SL, Link MJ. Gamma knife radiosurgery for long-term results. J Clin Oncol. 2001;19:3547–53.
skull base meningiomas. Neurosurg Clin N Am. 67. Goldsmith BJ, Rosenthal SA, Wara WM, Larson DA. Optic neuropa-
2000;11:659–66. thy after irradiation of meningioma. Radiology. 1992;185:71–6.
48. Sughrue ME, Kane AJ, Shangari G, et al. The relevance of 68. Miralbell R, Linggood RM, de la Monte S, Convery K,
Simpson Grade I and II resection in modern neurosurgical treat- Munzenrider JE, Mirimanoff RO. The role of radiotherapy in
ment of World Health Organization Grade I meningiomas: clinical the treatment of subtotally resected benign meningiomas.
article. J Neurosurg. 2010;113:1029–35. J Neurooncol. 1992;13:157–64.
49. Oya S, Kawai K, Nakatomi H, Saito N. Significance of Simpson 69. Pirzkall A, Debus J, Haering P, et al. Intensity modulated radio-
grading system in modern meningioma surgery: integration of the therapy (IMRT) for recurrent, residual, or untreated skull-base
grade with MIB-1 labeling index as a key to predict the recurrence meningiomas: preliminary clinical experience. Int J Radiat Oncol
of WHO grade I meningiomas: clinical article. J Neurosurg. Biol Phys. 2003;55:362–72.
2012;117:121–8. 70. Uy NW, Woo SY, Teh BS, et al. Intensity-modulated radiation
50. Hasseleid BF, Meling TR, Rønning P, Scheie D, Helseth therapy (IMRT) for meningioma. Int J Radiat Oncol Biol Phys.
E. Surgery for convexity meningiomas: Simpson Grade I resec- 2002;53:1265–70.
tion should still be the goal: clinical article. J Neurosurg. 71. Shrieve DC, Hazard L, Boucher K, Jensen RL. Dose fractionation
2012;117(6):999–1006. in stereotactic radiotherapy for parasellar meningiomas: radiobio-
51. Mirimanoff RO, Dosoretz DE, Linggood RM, Ojemann RG, logical considerations of efficacy and optic nerve tolerance.
Martuza RL. Meningioma: analysis of recurrence and progres- J Neurosurg. 2004;101 Suppl 3:390–5.
sion following neurosurgical resection. J Neurosurg. 1985; 72. Selch MT, Ahn E, Laskari A, et al. Stereotactic radiotherapy for
62:18–24. treatment of cavernous sinus meningiomas. Int J Radiat Oncol
52. Morokoff AP, Zauberman J, Black PM. Surgery for convexity Biol Phys. 2004;59:101–11.
meningiomas. Neurosurgery. 2008;63:427–34. 73. Urie MM, Fullerton B, Tatsuzaki H, et al. A dose response analy-
53. Soyuer S, Chang EL, Selek U, Shi W, Maor MH, DeMonte sis of injury to cranial nerves and/or nuclei following proton beam
F. Radiotherapy after surgery for benign cerebral meningioma. radiation therapy. Int J Radiat Oncol Biol Phys. 1992;23:27–39.
Radiother Oncol. 2004;71:85–90. 74. Berman D, Miller NR. New concepts in the management of optic
54. Jensen R, Lee J. Predicting outcomes of patients with intracranial nerve sheath meningiomas. Ann Acad Med Singapore. 2006;35:
meningiomas using molecular markers of hypoxia, vascularity, 168–74.
and proliferation. Neurosurgery. 2012;71:146–56. 75. Liu JK, Forman S, Moorthy CR, Benzil DL. Update on treatment
55. King DL, Chang CH, Pool JL. Radiotherapy in the management modalities for optic nerve sheath meningiomas. Neurosurg Focus.
of meningiomas. Acta Radiol Ther Phys Biol. 1966;3:26–33. 2003;14:e7.
56. Glaholm J, Bloom HJ, Crow JH. The role of radiotherapy in the 76. Andrews BT, Wilson CB. Suprasellar meningiomas: the effect of
management of intracranial meningiomas: the Royal Marsden tumor location on postoperative visual outcome. J Neurosurg.
Hospital experience with 186 patients. Int J Radiat Oncol Biol 1988;69:523–8.
Phys. 1990;18:755–61. 77. Cristante L. Surgical treatment of meningiomas of the orbit and
57. King DL, Chang CH, Pool JL. Radiotherapy in the management optic canal: a retrospective study with particular attention to the
of meningiomas. Acta Radiol. 1966;5:26–33. visual outcome. Acta Neurochir (Wien). 1994;126:27–32.
58. Jaaskelainen J, Haltia M, Servo A. Atypical and anaplastic menin- 78. Ito M, Ishizawa A, Miyaoka M, Sato K, Ishii S. Intraorbital
giomas: radiology, surgery, radiotherapy, and outcome. Surg meningiomas. Surgical management and role of radiation therapy.
Neurol. 1986;25:233–42. Surg Neurol. 1988;29:448–53.
79. Kennerdell JS, Maroon JC, Malton M, Warren FA. The manage- 98. El-Khatib M, El Majdoub F, Hoevels M, et al. Stereotactic LINAC
ment of optic nerve sheath meningiomas. Am J Ophthalmol. radiosurgery for incompletely resected or recurrent atypical and
1988;106:450–7. anaplastic meningiomas. Acta Neurochir (Wien). 2011;153:
80. Turbin RE, Thompson CR, Kennerdell JS, Cockerham KP, 1761–7.
Kupersmith MJ. A long-term visual outcome comparison in 99. Pollock BE, Stafford SL, Link MJ, Garces YI, Foote
patients with optic nerve sheath meningioma managed with obser- RL. Stereotactic radiosurgery of World Health Organization grade
vation, surgery, radiotherapy, or surgery and radiotherapy. II and III intracranial meningiomas: treatment results on the basis
Ophthalmology. 2002;109:890–9. discussion 899. of a 22-year experience. Cancer. 2012;118:1048–54.
81. Narayan S, Cornblath WT, Sandler HM, Elner V, Hayman 100. Sughrue ME, Sanai N, Shangari G, Parsa AT, Berger MS,
JA. Preliminary visual outcomes after three-dimensional confor- McDermott MW. Outcome and survival following primary and
mal radiation therapy for optic nerve sheath meningioma. Int J repeat surgery for World Health Organization Grade III meningio-
Radiat Oncol Biol Phys. 2003;56:537–43. mas. J Neurosurg. 2010;113:202–9.
82. Bloch O, Sun M, Kaur G, Barani IJ, Parsa AT. Fractionated radio- 101. Ware ML, Larson DA, Sneed PK, Wara WW, McDermott
therapy for optic nerve sheath meningiomas. J Clin Neurosci. MW. Surgical resection and permanent brachytherapy for recur-
2012;19:1210–5. rent atypical and malignant meningioma. Neurosurgery.
83. Taylor BWJ, Marcus RBJ, Friedman WA, Ballinger WEJ, Million 2004;54:55–63. discussion 63.
RR. The meningioma controversy: postoperative radiation ther- 102. Rosenberg LA, Prayson RA, Lee J, et al. Long-term experience with
apy. Int J Radiat Oncol Biol Phys. 1988;15:299–304. World Health Organization grade III (malignant) meningiomas at a
84. Pearson BE, Markert JM, Fisher WS, et al. Hitting a moving target: single institution. Int J Radiat Oncol Biol Phys. 2009;74:427–32.
evolution of a treatment paradigm for atypical meningiomas amid 103. Milosevic MF, Frost PJ, Laperriere NJ, Wong CS, Simpson
changing diagnostic criteria. Neurosurg Focus. 2008;24(5):E3. WJ. Radiotherapy for atypical or malignant intracranial meningi-
85. Simon M, Boström J, Koch P, Schramm J. Interinstitutional vari- oma. Int J Radiat Oncol Biol Phys. 1996;34:817–22.
ance of postoperative radiotherapy and follow up for meningio- 104. Mattozo CA, De Salles AAF, Klement IA, et al. Stereotactic radia-
mas in Germany: impact of changes of the WHO classification. tion treatment for recurrent nonbenign meningiomas. J Neurosurg.
J Neurol Neurosurg Psychiatry. 2006;77:767–73. 2007;106:846–54.
86. Marcus HJ, Price SJ, Wilby M, Santarius T, Kirollos 105. Harris AE, Lee JY, Omalu B, Flickinger JC, Kondziolka D,
RW. Radiotherapy as an adjuvant in the management of intracra- Lunsford LD. The effect of radiosurgery during management of
nial meningiomas: are we practising evidence-based medicine? Br aggressive meningiomas. Surg Neurol. 2003;60:298–305. discus-
J Neurosurg. 2008;22:520–8. sion 305.
87. Goyal LK, Suh JH, Mohan DS, Prayson RA, Lee J, Barnett 106. Kondziolka D, Flickinger JC, Perez B. Judicious resection and/or
GH. Local control and overall survival in atypical meningioma: a radiosurgery for parasagittal meningiomas: outcomes from a mul-
retrospective study. Int J Radiat Oncol Biol Phys. 2000;46: ticenter review. Gamma Knife Meningioma Study Group.
57–61. Neurosurgery. 1998;43:405–13. discussion 413.
88. Mahmood A, Caccamo DV, Tomecek FJ, Malik GM. Atypical and 107. Boskos C, Feuvret L, Noel G, et al. Combined proton and photon
malignant meningiomas: a clinicopathological review. conformal radiotherapy for intracranial atypical and malignant
Neurosurgery. 1993;33:955–63. meningioma. Int J Radiat Oncol Biol Phys. 2009;75:399–406.
89. Palma L, Celli P, Franco C, Cervoni L, Cantore G. Long-term 108. DeVries A, Munzenrider JE, Hedley-Whyte T, Hug EB. The role
prognosis for atypical and malignant meningiomas: a study of 71 of radiotherapy in the treatment of malignant meningiomas.
surgical cases. J Neurosurg. 1997;86:793–800. Strahlenther Onkol. 1999;175:62–7.
90. Mair R, Morris K, Scott I, Carroll TA. Radiotherapy for atypical 109. DC S, JS L. Radiotherapy in the treatment of meningiomas. In:
meningiomas: clinical article. J Neurosurg. 2011;115:811–9. Black PM, Loeffler JS, editors. Cancer of the nervous system.
91. Komotar RJ, Iorgulescu JB, Raper DMS, et al. The role of radio- Philadelphia: Lippincott Williams & Wilkins; 2005. p. 357–66.
therapy following gross-total resection of atypical meningiomas: 110. Goldsmith BJ, Larson DA. Conventional radiation therapy for
clinical article. J Neurosurg. 2012;117:679–86. skull base meningiomas. Neurosurg Clin N Am. 2000;11:605–15.
92. Park H, Kim I, Jung H. Atypical meningioma: outcomes and prog- 111. Rogers L, Jensen R, Perry A. Chasing your dural tail: Factors pre-
nostic factors. Int J Radiat Oncol Biol Phys. 2009;75:S238. dicting local tumor control after gamma knife stereotactic
93. Hardesty DA, Wolf AB, Brachman DG, et al. The impact of adju- radiosurgery for benign intracranial meningiomas: In regard to
vant stereotactic radiosurgery on atypical meningioma recurrence DiBiase et al. (Int J Radiat Oncol Biol Phys 2004;60:1515–1519).
following aggressive microsurgical resection: clinical article. Int J Radiat Oncol Biol Phys. 2005;62:616–8. author reply 618.
J Neurosurg. 2013;119(2):475–81. 112. Qi ST, Liu Y, Pan J, Chotai S, Fang LX. A radiopathological clas-
94. Stessin AM, Schwartz A, Judanin G, et al. Does adjuvant external- sification of dural tail sign of meningiomas. J Neurosurg.
beam radiotherapy improve outcomes for nonbenign meningio- 2012;117:645–53.
mas? A Surveillance, Epidemiology, and End Results (SEER)–based 113. Barbaro NM, Gutin PH, Wilson CB, Sheline GE, Boldrey EB,
analysis: clinical article. J Neurosurg. 2012;117:669–75. Wara WM. Radiation therapy in the treatment of partially resected
95. Hoffmann W, Mühleisen H, Hess CF, et al. Atypical and anaplas- meningiomas. Neurosurgery. 1987;20:525–8.
tic meningiomas—does the new WHO-classification of brain 114. Al-Mefty O, Kersh JE, Routh A, Smith RR. The long-term side
tumours affect the indication for postoperative irradiation? Acta effects of radiation therapy for benign brain tumors in adults.
Neurochir. 1995;135:171–8. J Neurosurg. 1990;73:502–12.
96. Katz TS, Amdur RJ, Yachnis AT, Mendenhall WM, Morris 115. Nutting C, Brada M, Brazil L, et al. Radiotherapy in the treatment
CG. Pushing the limits of radiotherapy for atypical and malignant of benign meningioma of the skull base. J Neurosurg.
meningioma. Am J Clin Oncol. 2005;28:70–4. 1999;90:823–7.
97. Dziuk TW, Woo S, Butler EB, et al. Malignant meningioma: an 116. Maguire PD, Clough R, Friedman AH, Halperin EC. Fractionated
indication for initial aggressive surgery and adjuvant radiotherapy. external-beam radiation therapy for meningiomas of the cavern-
J Neurooncol. 1998;37:177–88. ous sinus. Int J Radiat Oncol Biol Phys. 1999;44:75–9.
117. Steinvorth S, Welzel G, Fuss M, et al. Neuropsychological out- Simpson grade 1 resection for patients with small- to medium-size
come after fractionated stereotactic radiotherapy (FSRT) for base meningiomas. Int J Radiat Oncol Biol Phys. 2003;55:1000–5.
of skull meningiomas: a prospective 1-year follow-up. Radiother 120. Ramsey AF, Blurton M, Ekstrand K, et al. Edema following
Oncol. 2003;69:177–82. Gamma-Knife® radiosurgery for intracranial meningiomas. Int J
118. Meyers CA, Geara F, Wong PF, Morrison WH. Neurocognitive Radiat Oncol Biol Phys. 2002;54:146–47.
effects of therapeutic irradiation for base of skull tumors. Int J 121. Stafford SL, Pollock BE, Foote RL, Link MJ, Schomberg PJ, Pollock
Radiat Oncol Biol Phys. 2000;46:51–5. BE. Stereotactic radiosurgery for meningioma. Contemporary ste-
119. Pollock BE, Stafford SL, Utter A, Giannini C, Schreiner reotactic radiosurgery. Armonk: Futura; 2002. p. 157–71.
SA. Stereotactic radiosurgery provides equivalent tumor control to
Radiosurgery of Acoustic
Schwannomas 23
John C. Flickinger, Hideyuki Kano, and L. Dade Lunsford
neoplasia syndrome resulting from mutations in the NF2

Introduction tumor suppressor gene on chromosome 22q. NF2 has a fre-
quency of one in 25,000 live births and nearly 100 % pene-
Acoustic schwannomas, also referred to as acoustic neuro- trance by age 60. Half of patients inherit a germline mutation
mas or vestibular schwannomas, are benign tumors arising from an affected parent and the remainder acquire a de novo
from Schwann cells that ensheath the eighth cranial nerve. mutation for neurofibromatosis type 2 [2].
Presenting symptoms can include hearing loss, tinnitis, pres- While bilateral vestibular schwannoma development is
sure sensation in the ear, imbalance, spells of dizziness (more the hallmark of type-2 neurofibromatosis (NF2), these
common than true vertigo), facial weakness or twitching, patients can sometimes present with only a single acoustic
and facial pain or paresthesias. They are usually relatively schwannoma at presentation, although usually at a younger
homogeneously enhancing tumors, although they sometimes age than spontaneous unilateral acoustic schwannoma
can form cysts. They tend to arise within the internal audi- patients. Usually NF2 patients will have one or more other
tory canal and while they can slowly enlarge the canal, they associated tumors. NF2 patients have a propensity to develop
do not invade bone as meningiomas often do. Once they other central nervous system tumors including non-acoustic
extend outside the internal auditory canal, acoustic tumors schwannomas, meningiomas, ependymomas, astrocytomas,
can expand more easily in the cerebellopontine angle where and neurofibromas. They may also develop peripheral neu-
they are not constrained by bone. Acoustic schwannomas ropathy, ophthalmological lesions (cataracts, epiretinal
can be differentiated in CT and MR images from acoustic membranes, and retinal hamartomas), and cutaneous lesions
region meningiomas by their lack of dural attachment and (intracutaneous plaque-like lesions or more deep-seated sub-
invasion along the petrous bone at the porus acusticus [1]. cutaneous nodular tumors).
Differentiation of acoustic schwannomas from meningio- The cellular transformation of Schwann cells into schwan-
mas is not particularly critical for management of these nomas is caused from a lack of the functional product of the
tumors by radiosurgery or radiotherapy since similar radia- NF2 gene, which is the protein, schwannomin/merlin (S/M).
tion techniques are used for either tumor and tumor control Lack of normal S/M protein in the schwannoma cell occurs
rates are high for either tumor. from a detectable direct genetic mutation in 50 % of sporadic
Acoustic schwannomas can occur in a unilateral sponta- vestibular schwannomas, while in the other cases, epigenetic
neous form as part of Neurofibromatosis type 2 (NF-2), factors or activation of protease cascade results in ineffective
where there is a high propensity to develop bilateral vestibu- S/M. The exact interactions of S/M with extracellular matrix,
lar schwannomas. NF2 is an autosomal-dominant multiple membranous glycoprotein, and cytoskeleton appear to acti-
vate several pathways that could regulate the cell-cycle pro-
cess, apoptosis, and intercellular interaction [3].
J.C. Flickinger, M.D. (*)

Department of Radiation Oncology, University of Pittsburgh, Patient Assessment
Pittsburgh, PA USA
e-mail: flickingerjc@upmc.edu
Vestibular schwannomas most commonly arise in the ves-
H. Kano • L.D. Lunsford
Department of Neurological Surgery,
tibular branch of that nerve. In spite of the site of their origin,
University of Pittsburgh School of Medicine, vertigo and/or balance problems are less likely to be the pre-
200 Lothrop Street, Pittsburgh, PA 15213, USA senting symptom than hearing loss or tinnitis because there
L.S. Chin, W.F. Regine (eds.), Principles and Practice of Stereotactic Radiosurgery, 339
340 J.C. Flickinger et al.
Table 23.1 Gardner-Robertson hearing (G-R) classification based on Table 23.2 House-Brackmann facial nerve grades [5]
the poorer of either pure tone average (PTA) or speech discrimination
Grade 1 Normal facial function in all regions
score (SDS) class divisions [4]
Grade 2 Mild dysfunction: light weakness not obvious at rest
G-R hearing class Criteria Grade 3 Moderate weakness that is obvious at rest but not
Class I (good) PTA 0–30dB and/or SDS 70–100 % disfiguring and with complete eyelid closure
Class II (serviceable/useful) PTA 31–50dB and/or SDS 50–69 % Grade 4 Moderate to severe weakness: obvious disfiguring
Class III (non-serviceable) PTA 51–90dB and/or SDS 5–49 % weakness with incomplete eyelid closure
Class IV (poor) PTA 91dB-max and/or SDS 1–4 % Grade 5 Severe weakness: barely perceptible muscle movement
Class V (none/deaf) Untestable for PTA and/or SDS with incomplete eyelid closure
Grade 6 Complete paralysis, no movement, loss of tone
are compensating overlapping control mechanisms that help

with balance. Acoustic schwannoma patients who deny any Table 23.3 Koos acoustic schwannoma tumor size grading criteria
balance problems can often be found to have vestibular Koos Grade I Purely intracanalicular tumor limited to the internal
abnormalities with more extensive vestibular testing and auditory canal only
even simple tests like tandem gait testing in darkened rooms. Koos Grade II < 2 cm extracanalicular/CPA extension without
Hearing is usually graded with the Gardner-Robertson brainstem compression
Classification system [4] as shown in Table 23.1. Pure tone Koos Grade III Extracanalicular/CPA extension >2 cm, with no
brainstem compression
average (PTA) is usually defined as the average of the dB lev-
Koos Grade IV Extracanalicular/CPA extension with any degree
els at which subjects identify pure tones at 500, 1,000, and of brainstem compression
2,000 Hz, although some audiologists and otologists prefer to
include a fourth level of 3,000 Hz with the average to include
more of the high frequencies that help in speech discrimina- acoustic schwannomas) managed by radiosurgery. Four of
tion. Speech recognition threshold (SRT) is defined as the 6 obtained complete relief and one patient partial relief
volume at which 50 % of the speech is correctly identified. of hemifacial spasm with accompanying tumor shrinkage.
Speech discrimination score is the percentage of words recog- The sixth patient with no relief had no tumor shrinkage.
nized at a particular volume. Speech discrimination testing is New onset transient facial twitching has been observed in
done at an amplification above the PTA and SRT but can vary rare cases following acoustic schwannoma radiosurgery but
slightly with the testing volume. Serviceable hearing is a term seems to be a temporary response without any facial weak-
used to classify Gardner-Robertson Class 1–2 hearing that is ness developing.
useful in understanding speech. “Serviceable” is a better term As acoustic schwannomas continue to grow, they may
that “Useful” hearing since any residual hearing (even poor), compress the trigeminal nerve causing facial numbness,
may often be useful to a patient. Extensive high frequency pain, and/or paresthesias. Tumor growth may also lead to
hearing loss can leave a patient with 0 % speech discrimina- brain stem compression and possibly obstructive hydroceph-
tion but a testable pure tone average. Such hearing may still alus. The Koos classification (Table 23.3) is used to catego-
be helpful in identifying and locating lower frequency (non- rize acoustic schwannoma tumors according to size.
speech) noises and thus have some use and can be incredibly
useful in patients with no hearing in the other ear.
Facial weakness occurs less commonly from tumor growth Management Options
or from iatrogenic (surgical or radiation) injury than hearing
loss in acoustic schwannoma patients even though the nerves Management options for acoustic schwannoma include
run side by side through the internal auditory canal. The observation, complete or partial resection, single fraction or
House-Brackmann classification for facial weakness [5] multiple fraction radiosurgery (with 2–5 fractions) and frac-
(Table 23.2) was devised to grade postoperative facial weak- tionated radiotherapy (>5–30 fractions) with photons (X-rays
ness following acoustic schwannoma surgery. Fortunately the or Co-60 gamma rays) or with protons. Surgical intervention
15–20 % rates of developing any facial neuropathy (tempo- is usually the preferred option for large acoustic tumors (par-
rary plus permanent) reported in early radiosurgery series ticularly >4 cm) with no residual hearing in that ear and
with higher doses of 16–18 Gy and less conformal treatment problems such as obstructive hydrocephalus.
plans have dropped to significantly less than 1 % in modern Kondziolka’s literature review of acoustic schwannoma
series with lower doses of 12–13 Gy and better treatment observation reported growth rates of 0–4 mm per year in
planning. Transient facial spasms or twitching without per- different series with average volume doubling times of
sistent weakness may sometimes occur in acoustic schwan- 1.65–4.4 years [7].
noma patients from facial nerve irritation. Chang et al. [6] When observation is chosen as the management strategy,
reported a series of six patients with hemifacial spasm from patients must stick to regular follow up audiometry and CT
cerebellopontine angle tumors (four meningiomas and two or MR scans at the same imaging center with comparisons of
23 Radiosurgery of Acoustic Schwannomas 341
Fig. 23.1 Subtraction co-registration (fusion) T1-contrast enhanced MR images of a patient with bilateral acoustic tumors from NF-2 with 2009
images subtracted from 2011 images to highlight 2-year growth within the right sided tumor
the newest to the oldest scans. As with any slow-growing deterioration is documented exposes patients to the risks of
tumor, slow but steady growth may sometimes be reported as hearing loss during the follow-up intervals as well as poten-
no interval growth for several successive scans when only the tially higher hearing loss risk from irradiation of a larger
two latest scans are compared by the radiologists, when com- tumor while initial radiation treatment has only the risk of
parison of the first to last scan identifies unmistakable growth. hearing loss from treating the tumor at the time it is diag-
High resolution imaging with fine cuts (1–2 mm cuts) and nosed. Kondziolka’s review of the medical literature on
software that allows volume calculation can be extremely observation versus radiation treatment provides a detailed
useful for early identification of limited tumor growth. One argument for early radiosurgical intervention [7].
technique for identifying tumor growth that can be useful at
the time of radiosurgery in NF2 patients with multiple tumors
(usually meningiomas and acoustic schwannomas) is image Radiobiology of Tumor Control
co-registration (fusion) of old and new scans with subtrac- and Complications
tion of the old images from the new images to highlight
changes as shown in Fig. 23.1. Prior to the advent of radiosurgery, vestibular schwannomas
Higher growth rates have been reported for younger were managed almost exclusively by observation or surgical
patients and those with extracanalicular and cystic tumors. resection because of a belief that they were radio-resistant,
Observation of small acoustic schwannomas in NF2 patients with the first article indicating effectiveness of radiotherapy
with good hearing that appears to be stable may allow patients in controlling acoustic schwannomas (following incomplete
to enjoy more time with good hearing and give patients with resection) published in 1987, following early reports of
tumors in their only hearing ear time to learn sign language successful control with gamma knife radiosurgery [8].
before their hearing is lost. Initial observation may also be the Since then, multiple radiosurgery and radiotherapy series
best option for elderly patients with significant comorbidities with long-term follow up demonstrated that vestibular
and small tumors. Initial observation also is a good strategy schwannomas have control rates equal or better than other
for managing patients who want to delay treatment because benign CNS tumors like WHO Grade-1 meningiomas and
of anxiety about any potential side effects from radiosurgery. nonfunctional pituitary adenomas using similar or lower
A strategy of choosing to delay radiation treatment of an doses (in the range of 45–54 Gy at 1.8 Gy per fraction, 18–21
acoustic schwannoma until significant growth and/or hearing Gy in three fractions or 10–13 Gy in a single fraction).
Vernimmen and Slabbert calculated alpha-beta ratios of within the brainstem could not be ruled out to development
2.4 and 1.8 Gy for control of acoustic schwannomas from of the 3.5 % and 9.9 % rates of facial and trigeminal neu-
published literature using the Fraction Equivalent plot (FE) ropathies reported, respectively.
and the Tucker methods, respectively [9]. Since the accepted Adams et al. reported brainstem and cranial nerve injury
alpha-beta values for late reacting CNS injury (cranial developing after LINAC radiosurgery to a dose of 12 Gy pre-
nerves, parenchymal brain injury or spinal cord) fall in the scribed to an 80 % isodose for treatment of an acoustic
range of 1–2.5 Gy, these findings would support limited, if schwannoma measuring 27 × 19 × 18 mm [13]. Four months
any, therapeutic gain from fractionation compared to single later she developed subjective right facial weakness and
fraction treatment. In contrast to this radiobiological analy- reduced taste on the right side of the tongue (seventh cranial
sis, some centers report of better hearing preservation with nerve—CNVII) and by 8 months post-SRS, new right sided
fractionated radiotherapy as opposed to single fraction radio- facial numbness and tingling (fifth cranial nerve—CNV). An
surgery but all of these comparisons have difficulties. All MRI at 8 months showed reduction in tumor size to
CNS tissue does not seem to be equally likely to manifest 20 × 16 × 14 mm with an adjacent area of enhancement (con-
clinically symptomatic radiation injury. Changes in function sistent with parenchymal radiation injury) of the right lateral
of somatic sensory nerves (optic and auditory) are detected a pons/cerebellar peduncle which subsequently resolved. They
lower doses than somatic sensory nerves (trigeminal), while performed a γ-H2AX focus formation assay which revealed
motor nerves (Cranial nerve 7, as well as 3, 4, 6, etc.) appear impaired DNA damage kinetics in ex vivo-irradiated primary
less likely to show radiation injury. Following radiation tissues which suggest that the radiosurgery patient had an
treatment of acoustic schwannomas (by radiotherapy or inherent deficiency in double-stranded DNA breakage repair
radiosurgery), hearing loss is the most frequent sequelae mechanisms.
(30–60 % in most series) explained by the anatomy of the
tumor and/or the sensitivity of the special sensory nerve,
while the rates of trigeminal neuropathy (usually 1–4 %) Hearing Preservation and Cochlear Tolerance
exceed those of facial neuropathy (usually <1 %). The greater
risks of trigeminal versus facial nerve injuries are observed Hearing preservation rates after acoustic schwannoma radio-
even though a shorter segment of the trigeminal nerve is irra- surgery vary with prescription dose, length of follow up,
diated with generally lower doses than the facial nerve. tumor size, pretreatment hearing, and cochlear dose. Kano
Analysis of the initial experience of acoustic schwannoma analyzed factors affecting hearing preservation in 77 acous-
radiosurgery at the University of Pittsburgh (including doses tic schwannoma patients with Gardner-Robertson (GR)
of 12–20 Gy) correlated rates of trigeminal and facial injury Grade 1–2 hearing who underwent gamma knife radiosur-
with dose and length of nerve irradiated [10]. gery at the University of Pittsburgh with doses of 12–13 Gy
between 2004 and 2007 [14]. He correlated hearing loss with
Grade 1 hearing (versus Grade 2), age <60 years, and mean
Brainstem Tolerance cochlear dose <4.2 Gy. 12/12 patients <60 years old with
cochlear doses <4.2 Gy retained serviceable (Grade 1–2)
The parenchymal tissue in the brainstem tissue does not hearing for 2 years or more. Although a single fraction
seem to be inherently more sensitive to radiation injury than cochlear dose of 4.2 Gy seems rather low to expect a correla-
other locations when assessed by the development of post- tion with hearing loss. Zuur et al. were able to identify mod-
radiosurgery T2 signal changes, yet those changes are much est decreases in hearing within 101 head and neck cancer
more likely to result in neurological dysfunction than in patients undergoing 6–7 week courses of IMRT with median
other sites [11]. cochlear doses of only 11.4 Gy [15]. Yomo correlated hear-
Nakaya et al. reported the University of Pittsburgh radio- ing loss with a cochlear dose of 4 Gy or more in their series
surgery experience with 202 vestibular schwannomas and 44 of 154 acoustic schwannoma patients undergoing gamma
meningiomas with brainstem compression after a median knife radiosurgery at Marseille [16]. Hasegawa et al. also
marginal dose of 13 Gy [12]. Median tumor volumes were correlated hearing loss with cochlear dose [17]. Although
3.9 cm3 (range, 0.8–39.0 mL) and 6.6 mL (range, 1.6– Kim et al., were unable to find a significant correlation
25.1 mL) for acoustic schwannomas and meningiomas, between hearing loss and cochlear dose in their series of 66
respectively. No clear brainstem injury (with brainstem unilateral acoustic schwannoma patients undergoing radio-
edema and/or enhancement) was identified with median surgery to a mean dose of 12 Gy, but did correlate hearing
follow-ups 65 and 60 months for the vestibular schwanno- loss with transient post-radiosurgery volume expansion of
mas and meningiomas, respectively. Although 13 Gy to the ≥20 % from baseline [18].
edge of the brainstem seems reasonably safe from this report, While most cases of radiation injury to cranial nerves 5
some contribution by radiation exposure to the nerve roots and 7 usually show up within 2 or 3 years after radiosurgery,
hearing decline following radiosurgery requires longer

follow up. Hydrocephalus After Radiosurgery
Yomo’s analysis of the Marseille SRS experience found
rates of hearing decline of 5.86 dB/year <2 years after SRS Hydrocephalus after radiosurgery reported 0–8 % of patients
to a mean dose of 12.1 Gy dropped to a mean of 1.86 dB/year [24]. Han et al. [25] reported the rate of symptomatic com-
>2 years after SRS (p < 0.001) [16]. The 10-year Kaplan– municating hydrocephalus after radiosurgery was 5.6 %.
Meier rates for preserving the same Gardner-Robertson hear- Factors associated with higher rate of hydrocephalus
ing levels, serviceable hearing, and any testable hearing were included large tumor volume, female gender, and large vol-
44.0 % ±11.7 %, 44.5 % ±10.5 %, and 85.3 % ±6.2 %, ume of the lateral ventricle at the time of radiosurgery. Lee
respectively. Carlson reported the long-term hearing out- et al. [26] reported that surgical intervention for hydrocepha-
come of acoustic schwannoma SRS with 12–13 Gy marginal lus after radiosurgery required 5 % of patients at a median of
doses at Mayo Clinic for 44 subjects with a median audio- 15.5 months after radiosurgery.
metric follow-up of 9.3 years [19]. The Kaplan–Meier rates
of serviceable hearing preservation at 1, 3, 5, 7, and 10 years
following SRS were 80 %, 55 %, 48 %, 38 %, and 23 %, Tumor Control
respectively. Multivariate analysis correlated pretreatment
ipsilateral pure tone average (p < 0.001) and tumor size Long-term tumor control rates of acoustic schwannomas
(p = 0.009) with hearing loss. Roos reported the Royal with radiosurgery vary from 90 % to 98 % in different series.
Adelaide Hospital low dose (12–14 Gy) linac SRS experi- Chopra reported long-term results with acoustic schwan-
ence in 84 evaluable acoustic schwannoma patients. Their noma radiosurgery to 12–13 Gy for 216 patients at Pittsburgh
Kaplan–Meier hearing preservation rate of 50 % at 5 years with a Kaplan–Meier 98 % tumor control (freedom from
declined to 23 % by 10 years. The estimated risk of hearing resection) at 10 years. Hasegawa reported on a series of 440
loss after SRS for patients with initial PTA ≥ 20 dB was 5.0 vestibular schwannoma (median tumor volume 2.8 cu cm)
(95 % CI 2.2–11.2) times than with PTA < 20 dB [20] Kano. patients, (including 93 or 21 % who had undergone prior
resection and 13 or 3 % NF-2 patients) with SRS to a median
marginal dose of 12.8 Gy with a median follow-up of
Trigeminal Neuropathy After Radiosurgery 12.5 years. The 5- and ≥10-year Kaplan–Meier progression-
free survivals (PFS) were 93 % and 92 %. Their multivariate
Trigeminal nerve function can be preserved in the majority of analysis, correlated poorer progression-free survival with
patients. Lunsford et al. [21] reported that patients with ves- brainstem compression with deviation of the fourth ventricle
tibular schwannoma who received 18–20 Gy to the tumor (p < 0.0001), marginal dose ≤13 Gy (p = 0.01), prior treat-
margin had a 73 % of preservation rate of trigeminal nerve ment (p = 0.02), and female sex (p = 0.02). The Kaplan–Meier
function. However, patients with vestibular schwannoma who 10-year tumor control rates were 94 % for tumors <10 cm3,
received 12–13 Gy to the tumor margin had a 3.1 % risk of versus 77 % for tumors ≥10 cm3 (p < 0.0001). Among 363
trigeminal neuropathy. Hasegawa et al. [22] reported that three patients with tumors <10 cm3, the actuarial 10-year Kaplan–
(1 %) of 287 vestibular schwannoma patients with a median Meier tumor control was 96 % with >13 Gy (n = 123) versus
margin dose of ≤ 13 Gy developed facial numbness, including 94 % with 12–13 Gy (n = 240), (p = 0.62).
two with transient and one with persistent facial numbness.
Radiosurgery of Larger Acoustic Tumors

Facial Nerve Preservation
In addition to the 77 % tumor control reported by Hasegawa,
Facial nerve preservation rate varied between 91 and 100 %. for SRS of acoustic tumors >10 cm3, other centers have
Lunsford et al. [21] reported that normal facial function was reported separate results for radiosurgery of larger vestibular
preserved in 79 % of patients after 5 years in patients received schwannomas. Milligan reported a 5-year Kaplan–Meier
18–20 Gy to the tumor margin. But patients who received tumor control rate of 82 % for 22 acoustic tumors >2.5 cm in
12–13 Gy to the tumor margin had less than 1 % of facial the CPA with SRS to 12–14 Gy at the Mayo Clinic. They also
neuropathy. Hasegawa et al. [22] reported that three (1 %) of found 5-year Kaplan–Meier rates of 15 % for facial weak-
287 vestibular schwannoma patients with a median margin ness and 23 % for hearing preservation [27] Kano reported
dose of ≤ 13 Gy developed facial palsy, including two with the Pittsburgh experience with 65 patients with acoustic
transient and one with persistent palsy after a repeat schwannomas 3–4 cm in diameter managed by radiosurgery.
SRS. Regis reported transient facial palsy lower than 1 % in Two patients (3 %) underwent resection within 6 months due
1,000 patients [23]. to progressive symptoms. Two years later, with 63 tumors
overall after the two post-SRS resections, 16 tumors (25 %) (non-NF2) who underwent planned combined subtotal micro-
had a volume reduction of more than 50 %, 22 (35 %) tumors surgical removal followed by SRS [28]. Surgery was per-
had a volume reduction of 10–50 %, 18 (29 %) were stable in formed through translabyrinthine (n = 25) or retrosigmoid
volume (volume change < 10 %), and seven (11 %) had larger (n = 25) approaches. The median follow-up was 33.8 months.
volumes (five of the 7 patients underwent resection and one They found clinical tumor control in 92 % of the cases and
of the 7 underwent repeat SRS). Eighteen (82 %) of 22 radiological control in 90 %. One year post-radiosurgery,
patients with serviceable hearing before SRS still had ser- facial nerve function was good (House-Brackmann Grade I or
viceable hearing after SRS more than 2 years later. Three II) in 94 % of the patients. One of the TWO patients who
patients (5 %) developed symptomatic hydrocephalus and underwent surgery to preserve hearing maintained serviceable
underwent placement of a ventriculoperitoneal shunt. In four hearing after resection followed by GKS.
patients (6 %) trigeminal sensory dysfunction developed,
and in one patient (2 %) mild facial weakness (House-
Brackmann Grade II) developed after SRS. In univariate Radiation Treatment Planning Issues
analysis, patients who had a previous resection (p = 0.010),
those with a tumor volume exceeding 10 mL (p = 0.05), and Besides the usual concerns with producing highly conformal
those with Koos Grade 4 tumors (p = 0.02) had less likeli- radiation treatment plans for acoustic schwannomas several
hood of tumor control after SRS. The 5–10 year Kaplan– specific treatment planning issues arise. Based on the analy-
Meier tumor control rates were approximately 91 % and ses for Pittsburgh and Marseille, the mean cochlear dose
74 % for tumors <10 cm3 and >10 cm3, respectively. should be kept below 4.2 Gy for single fraction radiosurgery.
Another approach used with larger acoustic schwannomas Often the facial nerve is difficult to see on a treatment plan-
is to start with initial subtotal resection followed by radiosur- ning CT or MR scan particularly with larger acoustic
gery to residual tumor. Van de Langenberg reported on 50 con- schwannomas. As shown in Fig. 23.2, where it can be seen
secutive patients for large spontaneous acoustic schwannoma with a small intracanalicular acoustic schwannoma, the
Fig. 23.2 Radiosurgery plan for an intracanalicular acoustic schwannoma. The cochlea is labeled as “c,” the facial nerve as “7”
Fig. 23.3 Treatment plan for an acoustic schwannoma using selective cochlear blocking
facial nerve (labeled as 7) runs superiorly and anteriorly in Yomo reported the Marseille experience with repeat
the internal auditory canal. Care should be taken to avoid radiosurgery (median initial and retreatment doses of 12 Gy)
high dose regions and generous treatment margins in this in eight acoustic schwannomas that progressed after initial
region of the internal auditory canal. Since the auditory nerve radiosurgery [29]. The minimum and median follow up after
tends to run inferiorly and posteriorly in the internal auditory repeat radiosurgery was 24 and 64 months. The median
canal (labeled as 8 in Fig. 23.2), it would be prudent to limit interval between radiosurgery procedures was 46 months.
radiation dose in that location to maximize chances of hear- The median tumor volumes were 0.51 and 1.28 mL at the
ing preservation. Figure 23.3 shows a treatment plan for a initial and second SRS treatments, respectively. No tumor
larger tumor that uses blocking of beam channels that would progressed and 6/8 showed a significant decrease in tumor
have contributed significantly to cochlear dose. volume. Serviceable hearing was preserved in only one of
the 3 patients with Gardner-Robinson I-II hearing at the time
of the second SRS. No other new neurological deficits were
Salvage of Radiosurgery Failures observed after repeat SRS.
Hasegawa’s report included the management of 36 acous-
Tumors that show sustained growth after radiosurgery should tic schwannoma patients who underwent further treatment
be managed with either complete surgical resection or radio- after initial radiosurgery [17]. Seven patients who underwent
surgery. Tumor control rates for repeat radiosurgery of resection did not have clear evidence of sustained tumor pro-
benign tumors seem relatively high considering that they gression. Six underwent repeat radiosurgery but three of
have failed prior radiation treatment and usually are retreated them subsequently underwent tumor resection. Twenty five
to a lower dose. Kano reported the Pittsburgh experience patients (with mean tumor volumes of 10.7 cu cm) needed
with repeat radiosurgery to a median dose of 11 Gy in six surgery for ventricular peritoneal shunts to manage post-SRS
patients (at a median of 63 months after a median initial dose hydrocephalus. After their second intervention with surgery
of 13 Gy) [28]. After a median follow up of 29 months (range or radiosurgery the 5- and 10-year Kaplan–Meier rates for
13–71) all tumors were controlled and no new neurological freedom from further intervention were 92 % and 91 %,
sequelae were observed. respectively.
16. Yomo S, Carron R, Thomassin JM, Roche PH, Regis J. Longitudinal

References analysis of hearing before and after radiosurgery for vestibular
schwannoma. J Neurosurg. 2012;117:877–85.
17. Hasegawa T, Kida Y, Kato T, Iizuka H, Yamamoto T. Factors asso-
1. Tator CH, Duncan EG, Charles D. Comparisons of the clinical and ciated with hearing preservation after Gamma Knife surgery for
radiological features and surgical management of posterior fossa vestibular schwannomas in patients who retain serviceable hearing.
meningiomas and acoustic neuromas. Can J Neurol Sci. J Neurosurg. 2011;115:1078–86.
1990;17:170–6. 18. Kim YH, Kim DG, Han JH, Chung HT, Kim IK, Song SW, Park JH,
2. Asthagiri AR, Parry DM, Butman JA, Kim HJ, Tsilou ET, Zhuang Z, Kim JW, Park CK, Kim CY, Paek SH, Jung HW. Hearing outcomes
Lonser RR. Neurofibromatosis type 2. Lancet. 2009;373: after stereotactic radiosurgery for unilateral intracanalicular ves-
1974–86. tibular schwannomas: implication of transient volume expansion.
3. Roche PH, Bouvier C, Chinot O, Figarella-Branger D. Genesis and Int J Radiat Oncol Biol Phys. 2013;85:61–7.
biology of vestibular schwannomas. Prog Neurol Surg. 2008;21:24– 19. Carlson ML, Jacob JT, Pollock BE, Neff BA, Tombers NM, Driscoll
31. doi:10.1159/000156556. CL, Link MJ. Long-term hearing outcomes following stereotactic
4. Gardner G, Robertson JH. Hearing preservation in unilateral acous- radiosurgery for vestibular schwannoma: patterns of hearing loss
tic neuroma surgery. Ann Otol Rhinol Laryngol. 1988;97:55–66. and variables influencing audiometric decline. J Neurosurg.
5. House JW, Brackmann DE. Facial nerve grading system. 2013;118(3):579–87.
Otolaryngol Head Neck Surg. 1985;93:146–7. 20. Roos DE, Potter AE, Zacest AC. Hearing preservation after low
6. Chang CS, Chuang CC, Wu MF, Liu WS, Tu HT, Huang CF. Gamma dose linac radiosurgery for acoustic neuroma depends on initial
Knife surgery for hemifacial spasm related to cerebellopontine hearing and time. Radiother Oncol. 2011;101(3):420–4.
angle tumors. J Neurosurg. 2012;117(Suppl):170–4. 21. Lunsford LD, Niranjan A, Flickinger JC, Maitz A, Kondziolka D.
7. Kondziolka D, Mousavi SH, Kano H, Flickinger JC, Lunsford Radiosurgery of vestibular schwannomas: summary of experience
LD. The newly diagnosed vestibular schwannoma: radiosurgery, in 829 cases. J Neurosurg. 2005;102(Suppl):195–9.
resection, or observation? Neurosurg Focus. 2012;33:E8. 22. Hasegawa T, Kida Y, Kato T, Iizuka H, Kuramitsu S, Yamamoto
doi:10.3171/2012.6.FOCUS12192. T. Long-term safety and efficacy of stereotactic radiosurgery for
8. Wallner KE, Sheline GE, Pitts LH, Wara WM, Davis RL, Boldrey vestibular schwannomas: evaluation of 440 patients more than 10
EB. Efficacy of irradiation for incompletely excised acoustic years after treatment with Gamma Knife surgery. J Neurosurg.
neurilemomas. J Neurosurg. 1987;67:858–63. doi:10.3171/ 2013;118:557–65.
jns.1987.67.6.0858. 23. Regis J, Roche PH, Delsanti C, Thomassin JM, Ouaknine M,
9. Vernimmen FJ, Slabbert JP. Assessment of the alpha/beta ratios for Gabert K, Pellet W. Modern management of vestibular schwanno-
arteriovenous malformations, meningiomas, acoustic neuromas, mas. Prog Neurol Surg. 2007;20:129–41.
and the optic chiasma. Int J Radiat Biol. 2010;86:486–98. 24. Hayhurst C, Monsalves E, Bernstein M, Gentili F, Heydarian M,
doi:10.3109/09553001003667982. Tsao M, Schwartz M, van Prooijen M, Millar BA, Menard C,
10. Flickinger JC, Kondziolka D, Lunsford LD. Dose and diameter Kulkarni AV, Laperriere N, Zadeh G. Predicting nonauditory
relationships for facial trigeminal and acoustic neuropathies fol- adverse radiation effects following radiosurgery for vestibular
lowing acoustic neuroma radiosurgery. Radiother Oncol. schwannoma: a volume and dosimetric analysis. Int J Radiat Oncol
1996;41(3):215–9. S0167814096018312 [pii]. Biol Phys. 2012;82:2041–6. doi:10.1016/j.ijrobp.2011.02.017.
11. Flickinger JC, Kondziolka D, Pollock BE, Maitz AH, Lunsford 25. Han JH, Kim DG, Chung HT, Paek SH, Park CK, Kim CY, Hwang
LD. Complications from arteriovenous malformation radiosurgery: SS, Park JH, Kim YH, Kim JW, Song SW, Kim IK, Jung HW. The
multivariate analysis and risk modeling. Int J Radiat Oncol Biol risk factors of symptomatic communicating hydrocephalus after
Phys. 1997;38:485–90. S0360301697894813 [pii]. stereotactic radiosurgery for unilateral vestibular schwannoma: the
12. Nakaya K, Niranjan A, Kondziolka D, Kano H, Khan AA, Nettel B, implication of brain atrophy. Int J Radiat Oncol Biol Phys.
Koebbe C, Pirris S, Flickinger JC, Lunsford LD. Gamma knife 2012;84:937–42. doi:10.1016/j.ijrobp.2012.01.048.
radiosurgery for benign tumors with symptoms from brainstem 26. Lee SH, Seol HJ, Kong DS, Nam DH, Park K, Kim JH, Lee JI. Risk
compression. Int J Radiat Oncol Biol Phys. 2010;77:988–95. factors and tumor response associated with hydrocephalus after
13. Adams G, Martin OA, Roos DE, Lobachevsky PN, Potter AE, gamma knife radiosurgery for vestibular schwannoma. Acta Neurochir
Zacest AC, Bezak E, Bonner WM, Martin RF, Leong T. Enhanced (Wien). 2012;154:1679–84. doi:10.1007/s00701-012-1350-0.
intrinsic radiosensitivity after treatment with stereotactic radiosur- 27. Milligan BD, Pollock BE, Foote RL, Link MJ. Long-term tumor con-
gery for an acoustic neuroma. Radiother Oncol. 2012;103:410–4. trol and cranial nerve outcomes following γ knife surgery for larger-
14. Kano H, Kondziolka D, Khan A, Flickinger JC, Lunsford volume vestibular schwannomas. J Neurosurg. 2012;116(3):598–604.
LD. Predictors of hearing preservation after stereotactic radiosur- 28. Kano H, Kondziolka D, Niranjan A, Flannery TJ, Flickinger JC,
gery for acoustic neuroma. J Neurosurg. 2009;111:863–73. Lunsford LD. Repeat stereotactic radiosurgery for acoustic neuro-
15. Zuur CL, Simis YJ, Lamers EA, Hart AA, Dreschler WA, Balm AJ, mas. Int J Radiat Oncol Biol Phys. 2010;76(2):520–7.
Rasch CR. Risk factors for hearing loss in patients treated with 29. Yomo S, Arkha Y, Delsanti C, Roche PH, Thomassin JM, Régis J.
intensity-modulated radiotherapy for head-and-neck tumors. Int J Repeat gamma knife surgery for regrowth of vestibular schwanno-
Radiat Oncol Biol Phys. 2009;74:490–6. mas. Neurosurgery. 2009;64(1):48–54. discussion 54–5.
Acoustic Neuroma: Viewpoint—Surgery
24
Gabriel Zada and Steven L. Giannotta
including patient preference, surgeon bias, cost, patient age,

Introduction and lifestyle issues. In our practice, the roles of providing
recommendations for neurosurgical treatment and/or stereo-
Acoustic schwannomas are the most common neoplasms tactic radiosurgical treatment of acoustic neuromas depend
arising in the cerebello-pontine angle. Treatment paradigms primarily on four major factors: (1) patient age (2) tumor
for patients with acoustic schwannomas may be inherently size (3) hearing levels, and (4) recurrence.
complex in nature, because of the multimodality treatment
options available and numerous clinical considerations,
including overall safety, hearing preservation, and facial Conservative Management
nerve function, among others. Over the past several decades,
the decision-making process for treating patients with acous- As the availability of neuro-imaging studies increases, many
tic neuromas has taken on even more complexity, given patients will have new tumors diagnosed in an incidental
trends toward improved radiosurgical treatment modalities fashion, without any subjective clinical symptoms. If tumors
and earlier detection with advanced neuro-imaging. In expe- are small, a logical strategy may be to simply follow with
rienced hands, surgical results provide rapid and reasonable surveillance imaging.
success rates to those seeking an excisional cure. On the There are several studies, both short and long term, under-
other hand, as longer-term follow-up has become available, scoring the fact that some tumors will change very little over
radiosurgical tumor control data, both fractionated and time [1–5]. Depending on the study, up to 85 % of follow-up
unfractionated, have shown more impressive safety and cases may show little or no growth of acoustic neuromas. In
tumor control rates. In selected cases, simple observation addition, several authors have demonstrated decreased tumor
may also be a preferred strategy. Clearly, there is no best way volume over time. Annual diameter growth rates have ranged
to treat all patients with acoustic schwannomas, and various between 0.15 and 4 mm [6–9]. Some practitioners, however,
treatment strategies should be optimally used in a compli- have witnessed alarming growth in rare cases, sometimes result-
mentary fashion to achieve the best overall outcomes. ing in hydrocephalus and/or brainstem compression, highlight-
In the ideal situation, a simple comparison between effi- ing the unpredictability of a strategy that relies simply on
cacy and safety data from several different therapeutic observation [5]. Cystic schwannomas, in particular, have been
options should be sufficient to make an informed decision. reported to have faster growth rates and present with more con-
However, aside from the various treatment options that exist, cerning signs related to brainstem compression [10].
additional factors beyond effectiveness often come into play, Conservative management for young patients may therefore
be fraught with hazard. More often than not, their tumors will
grow, thus exposing them to greater risk associated with treat-
G. Zada, M.D. (*)
Department of Neurosurgery, Keck School of Medicine, ment, or even eliminating the potential for treatment alterna-
University of Southern California, 1520 San Pablo Street, tives. We have a large number of patients over the age of 70 who
Suite 3800, Los Angeles, CA 90033, USA are simply examined on a yearly basis and undergo imaging.
e-mail: gzada@usc.edu
These patients infrequently require intervention, and radiosur-
S.L. Giannotta, M.D. gery is often employed if they do [11, 12]. Tumors that are too
Department of Neurological Surgery, Keck School of Medicine
large to be treated with radiosurgery in patients within this age
of USC, 1200 North State Street, Suite 3300,
Los Angeles, CA 90033, USA group are almost always offered operative therapy, unless there
e-mail: giannott@usc.edu are substantial comorbidities that preclude general anesthesia.
348 G. Zada and S.L. Giannotta
low, occurring 0–2 % of cases [9, 13–15, 18–20]. Mortality

Surgical Management from surgery is rare, occurring in 0.4–2 % of patients. The
two most important factors relating to safety and efficacy
The gold standard for treatment of any benign tumor is cura- associated with acoustic neuroma surgery are tumor size and
tive total removal. The concept of complete removal without the surgeon’s experience.
the need for lifelong surveillance is appealing to many
patients, making patient preference a major factor in the
selection of primary therapy. Thus, when deciding between
radiosurgical versus surgical alternatives, the realistic expec- Facial Nerve Preservation Outcomes
tation of a safe and total surgical obliteration is a powerful
incentive. Long-term follow-up data are available for selected Of all potential negative factors, a majority of patients are
surgical series documenting the potential for long-term free- concerned about facial nerve function following surgery. In
dom from recurrence in experienced surgeons’ hands [9, fact, this is clearly how the most accomplished practitioners
13–15]. In 379 surgical cases performed by the senior author, of the surgical art keep score. It is important for the surgeon
no recurrences resulted when the nerve of origin of the tumor advising a patient to be familiar with statistical outcomes
was identified and sacrificed. Surgical treatment of acoustic from recent surgical series and deem whether he or she can
schwannomas is often the preferred intervention in patients offer similar results. Most reports define an “excellent” or
with larger tumors (>3 cm diameter), younger age, prior “good” outcome as a House-Brackmann facial nerve score of
radiosurgical treatment, those with a preference towards sur- 1 or 2 [21].
gical excision. A review of the senior author’s results in 379 acoustic
Prior to the advent of radiosurgery, it was the authors’ tumors of all sizes revealed a 79 % H-B 1 or two facial nerve
philosophy to offer microscopic total removal to all patients function rate. Depending on the size of the tumor, excellent
with acoustic tumors. This strategy mandated, in many cases, facial nerve functional results have been reported in up to
protracted dissection of a thinned out facial nerve at the 90 % of cases [9, 13–15, 22–30].
porus acousticus. Review of the senior author’s facial nerve The Acoustic Neuroma Registry, started in the late 1980s,
results shows 19 % of cases with long-term House Brackmann catalogues the national results of various surgeons through
scores of 3 or worse. Given the high rate of radiosurgical questionnaires and surveys. In a summary of 1,579 cases
control of small lesions, subtotal removal in certain narrow recorded in this registry between 1989 and 1994, Wiegand
circumstances may be a preferable alternative. et al. found that in 92 % of cases surgeons felt the removal to
Over the past two decades, however, a new strategy has be total with 94 % preservation of the facial nerve continuity
been adopted in a majority of patients, in which the shift of [20]. Of these, 69 % had excellent facial nerve function after
the operative strategy focuses on aggressive tumor debulking follow-up at 1 year.
with a major emphasis on maintaining facial nerve function. In order to attempt a more direct comparison between sur-
Depending on the surgical approach utilized, it is not uncom- gery and radiosurgery, we sifted through the various publica-
mon to deliberately leave a residual amount of microscopic tions to identify those who stratified facial nerve results
tumor along selected functional anatomical regions, such as based on tumor size. Facial nerve outcomes after surgery for
the facial nerve, brainstem, or within the internal auditory tumors less than 3 cm were tabulated as single fraction radio-
canal (during retrosigmoid approaches). This strategy is surgery only addresses this subgroup of patients. As
most often employed in very large or vascular tumors. expected, these facial nerve outcomes are laudable, with
Frequently, no evidence of residual tumor is identifiable on favorable facial nerve outcomes in the 80–99 % range [13–
postoperative MR imaging. If a visualized region of residual 15, 25, 30, 31].
tumor is identifiable on postoperative MRI, this can be sub- Another way to capture surgical results that can be com-
sequently treated with radiosurgery, especially in cases pared with radiosurgical data is to look at cases where the
where documented tumor growth of the residual component surgeon attempted to preserve hearing. The majority of these
is observed [16, 17]. lesions would be in the size range for single fraction radio-
Long-term tumor control or cure is desirable to many surgery. Shelton et al. reported that 89 % of their patients
patients, but at what price? Among larger series, surgical experienced good facial function in a series of 106 cases,
complication rates have become acceptably low. The most performed through the middle fossa approach [32]. Other
common complication is CSF leaks, occurring in 2–25 % of groups have elevated their surgical techniques to where
cases [9, 13–15, 18–20]. Major morbidity, including lower facial nerve preservation and cochlear nerve preservation are
cranial nerve palsy, ataxia, or long-tract dysfunction, is also an expected outcome [33–38].
24 Acoustic Neuroma: Viewpoint—Surgery 349
Table 24.1 Approach selection.

Hearing Preservation Retrosigmoid Middle fossa Translabyrinthine
Size
The data on hearing is a bit difficult to interpret as patient <1 cm
outcomes are sometimes reported as simply at or near pre-op Lateral impact + +++ ++
levels. However, per the Gardner–Robertson scale a class I or Medial +++ +++ ++
<2.5 cm +++ 0 +++
II is considered as serviceable or better hearing [39]. In this
>2.5 cm ++ 0 +
scale pure-tone average of 0–30 db and speech discrimina-
Only-hearing ear +++ 0 +
tion of 70–100 % corresponds to class 1 hearing. Class 2 Hearing
hearing assumes 31–50 dB pure-tone average and 50–69 % >50/50 +++ +++ +
speech discrimination. The AAO-HNS Guidelines on the <50/50 + + +++
evaluation of hearing preservation also developed a system Recurrent + 0 +++
of classification with class A or B generally indicating useful
hearing [40]. Class A hearing includes <30 dB pure-tone
thresholds and >70 % speech discrimination. Class B
includes 30–50 dB pure-tone thresholds and >50 % speech
discrimination. Most reports use one of these two methods
for recording hearing outcomes.
The data with regard to hearing preservation among the
large series is somewhat variable. Good hearing can range
from 24 to 60 % depending on the series [9, 13–15, 18, 25,
26, 28, 30, 31, 36]. It is noteworthy that the Acoustic
Neuroma Registry data described by Wiegand et al. reports a
Fig. 24.1 (a) A 5 cm acoustic neuroma prior to resection. The patient
22 % hearing preservation rate [20]. presented with tinnitus, hearing loss, and signs of brainstem compres-
A cursory look at data from groups who report large series sion. (b) Postoperative image after a retrosigmoid craniotomy. All pre-
of middle fossa approaches would suggest that this strategy operative symptoms resolved
might be more effective in terms of hearing preservation.
Most accomplished groups report good hearing rates on the best approach for a given situation. Hearing preservation,
range of 50 % [32–35, 38, 41–43]. However, these series are facial nerve function, and incidence of total removal seem to
all biased towards the smallest of tumors. be practitioner or team related as opposed to approach
related. This suggests that technical expertise and strategic
operative decision-making can negate most disadvantages
Surgical Approaches related to the various approach strategies.
A retrospective review of our institution’s last 50 cases
Each of the three traditional approaches to the CP angle has of retrosigmoid resections was compared to the last 50
their own putative strategic advantages (Table 24.1). The translabyrinthine resections. Tumors were included that
translabyrinthine approach minimizes retraction, maximizes were 3 cm or less and results were categorized in terms of
exposure of the internal auditory canal, and facilitates defin- facial nerve outcome, incidence of total resection, and
ing the facial nerve within the canal. The retrosigmoid major complications. At discharge, 37 of the retrosigmoid
approach has the advantage of being stereotypical for most patients had an H-B score of 1 as compared to 28 of the
neurosurgeons while enabling the preservation of the hearing translabyrinthine patients. Combining H-B 1 and 2 scores
apparatus (Fig. 24.1). The middle fossa approach, while showed 82 % with the retrosigmoid cases and 70 % of the
admittedly for smaller tumors, also facilitates exposure of translabyrinthine. This did not reach statistical significance.
the lateral most extent of the internal auditory canal. There was one major neurological complication in the ret-
Proponents of each approach are able to post impressive rosigmoid group and none in the other group. Three cases
numbers in terms of safety and cure rates. Many studies have of subtotal removal resided in each group. Long-term fol-
looked at outcomes based on approach and no discernible low-up of these patients showed even less of a difference in
advantage has been documented for one approach over facial nerve function. This suggests that surgical technique
another [15, 33]. However, for purely intracanalicular tumors may be more important than approach strategies in terms of
several reports have shown that the middle fossa approach is outcome for acoustic tumors.
safer for hearing preservation than the retrosigmoid approach Ideally, each practitioner should have a working knowl-
[25, 35, 43]. A cursory review of the literature would suggest edge of all approaches. This can instill the wary patient with
no way to resolve controversies that may emerge as to the some confidence that decision-making is based on assess-
ment of the patient’s best interests. For those who are experi-
encing unacceptable results, altering approach strategy may
show modest benefits, but doing such will not make up for
deficiencies in surgical technique as it relates to the brain-
stem vascularity and cranial nerves.
Choice of Treatment Method (Radiosurgery

Versus Excision)
At our institution, the three traditional approaches for surgi-

cal excision of acoustic tumors are practiced. Further, addi-
tional technologies for radiosurgery (such as Gamma Knife,
Linac, Cyberknife, etc.) are offered. The surgical decision- Fig. 24.2 (a) A 2.5 cm acoustic neuroma in a patient who presented
with slight hearing loss. The fourth ventricle is distorted. This patient
making is less predicated on available technologies or reli- elected to have radiosurgery. (b) Post gamma knife radiosurgical treat-
ance on a practiced single surgical approach, than it is on ment. Central necrosis is present as well as decreased pressure on pos-
patient age, tumor size, and need for hearing preservation terior fossa structures
(Table 24.2).
that surgical excision is required following radiosurgery, it

Age and Tumor Size has been reported that tumor excision is much more chal-
lenging, and the resiliency of the facial nerve is compro-
For younger patients, emphasis is placed on surgical removal. mised, compared to nonirradiated patients [48].
A surveillance strategy in this group is likely to be futile, For large lesions greater than 3 cm, single fraction
since inevitably the lesion will cause further symptoms and radiosurgery has no role [49]. For these lesions, gross sur-
require treatment. Radiosurgery has a long follow-up period gical removal or in certain situations subtotal removal,
and the window of vulnerability for recurrence is potentially with follow-up radiosurgery or observation, is advisable.
wide. Reliable data on lifelong tumor control for patients in Subtotal removal for a patient with a single tumor and good
their 30s or 40s is lacking. For those in their 50s and 60s, hearing in the other ear should be an unusual event. A small
single fraction radiosurgery is an attractive alternative. residual may be left behind in an effort to avoid a major
Efficacy and safety statistics are available for this group and complication such as facial nerve sacrifice or brainstem
are highly acceptable (Fig. 24.2). Patients in older age groups injury. For younger patients with large tumors, surgical
rarely need any therapy unless their tumor is large enough to removal is the preferred strategy. The length of vulnerabil-
be threatening [11, 12]. Furthermore, radiosurgery is not ity for recurrence is too great for younger patients to rely
without its complications and failures, which develop in on subtotal removal. Some further therapeutic endeavor
approximately 3–34 % of cases in even the most experienced will ultimately be necessary, multiplying the potential for
centers [44, 45]. Serviceable hearing is lost in 30–43 % of complications.
patients following radiosurgery [46, 47]. Finally, in the event Large lesions in older patients can present some strategic
problems. This would seem like an ideal situation for hypo-
fractionated radiosurgery. However, with lesions greater than
Table 24-2. Treatment modality.
4 cm or somewhat smaller lesions with associated arachnoid
Radiosurgery Surgery Observe cysts usurping much of the available reserve in the posterior
Size fossa, radiosurgery with its attendant edema formation may
>2.5 cm + +++ 0
produce unacceptable risks 6–12 months posttreatment. Data
1.5 to 2.5 cm ++ ++ +
is sorely lacking for this modality in larger tumors. Until
<1.5 cm +++ ++ ++
Age
better long-term studies are available, older patients in good
<40 + +++ 0 health with large lesions should be offered the option of sur-
40 to 60 ++ ++ + gical removal. The decision for total versus subtotal removal
>60 +++ + +++ is made at the time of surgery and predicated on the likeli-
Hearing hood of complications. Radiosurgery as an adjunct can be
>50/50 +++ ++ ++ offered for any threatening residual. For large tumors in
<50/50 ++ ++ + older patients who are poor surgical risks, hypofractionated
Recurrent +++ + + radiosurgery is a logical option.
Hearing Preservation
Hearing preservation in the context of surgical removal can

be expected in experienced hands to be successful 50 % of
the time with intracanalicular tumors and 30 % with larger
lesions that are generally under 3 cm. This presupposes good
functional hearing to begin with. Attempts to save hearing in
a marginal or poorly hearing ear will be unrewarding. That
ear will be a constant source of distraction to the patient as it
picks up unstructured background noise and reduces the
Fig. 24.3 (a) Intraoperative view of acoustic neuroma during a left ret-
overall functionality of the hearing. Thus, compromises in rosigmoid craniotomy. (b) After tumor removal, the nerves of the porus
surgical strategy to preserve the function of a poorly hearing acousticus are seen. Note to transected superior vestibular nerve from
ear should be vigorously resisted. where the tumor originated
Single fraction radiosurgery is fast growing in popularity and
may become the treatment of choice in the absence of an expe- may complicate decision making in patients who cannot
rienced surgeon with a proven track record for the safe and understand a complex set of options. Patient and family
effective removal. For those lesions less than 3 cm, one can biases for or against surgery or radiation may direct the
expect at least 50 % hearing preservation or better assuming patient’s thinking contrary to the physician’s best judgment.
accepted proven radiosurgical techniques are utilized. The Access to the internet, influence from patients who have had
major drawback in prescribing it for all small acoustic tumors is one form of therapy or another, and loyalty to a particular
the lack of long-term efficacy data. For many patients, the need institution may be relevant factors in decision-making.
for continued surveillance and the thought of the continued One can guide the decision making by trying to simplify
presence of the lesion are negative satisfiers. principles. Explaining away misconceptions is a place to
If an acoustic neuroma recurs and is deemed in need of start. Identifying patient and family biases and dealing
treatment, the first option should be radiosurgery. This with them in a forthright way will also help. If it is per-
assumes the tumor regrowth has been detected before it has ceived that this discussion is simply a device to steer the
grown too large for radiosurgery. In these cases a translaby- decision-making toward the surgeon or radiosurgeon’s
rinthine surgical approach will offer the largest corridor bias, confusion and mistrust can develop. If a patient har-
while minimize the need for retraction or dissection of previ- bors a tumor that may be amenable to either surgical
ously scarred brain. A repeat retrosigmoid craniotomy may removal or radiosurgery, a simple construct can be pre-
prove difficult. sented to the family and patient to facilitate their decision-
making. Does the patient insist that the tumor be gone?
Benefits include diagnostic certainty and the lack of need
Choice of Surgical Approach for long-term surveillance.
Ultimately patients will decide on what treatment option
Our preference is to use the translabyrinthine approach for they prefer with some guidance as to the risk/benefit ratio
all tumors where hearing preservation is unlikely (Table 24.1). from the physician. The surgical removal of acoustic neu-
Thus, it is used in all large tumors and those with poor hear- roma has been refined over time to achieve impressive
ing. Certainly those greater than 3 cm would be treated this results. However, obvious risk of major morbidity remains.
way and most with speech discrimination scores of less than Radiosurgical methods are progressively less invasive and
50 %. The only time we would favor a retro sigmoid approach less likely to cause major morbidity. However, large lesions
in a large tumor would be the case of a large lesion in an only are clearly not amenable to this therapy and lifelong observa-
hearing ear where a subtotal removal is contemplated. tion is a requirement, even for small lesions, and especially
In tumors that protrude from the porus acousticus, the in younger patients. No perfect algorithm exists and each
retrosigmoid approach is preferred for hearing preserva- patient scenario has unique challenges. The first question the
tion (Fig. 24.3). In our most recent 80 cases using this physician and patient must answer is how important is the
approach, functional hearing resulted in 30 %. We limit the removal of the tumor to one’s overall comfort. If knowing
use of the middle fossa strategy for those small intracana- the tumor is still present makes one unable to participate in
licular lesions that are impacted in the lateral end of the life activities then surgery seems logical. Otherwise the
internal auditory canal. radiosurgical alternatives must be explored. With sound
Other factors come into play as patients try to make advice and research, the patient will ultimately decide his
informed decisions. Socioeconomic and educational status own treatment algorithm.
22. Arriaga MA, Gorum M, Kennedy A. Clinical pathways in acoustic

References tumor management. Laryngoscope. 1997;107(5):602–6.
23. Harner SG, Beatty CW, Ebersold MJ. Retrosigmoid removal of
acoustic neuroma: experience 1978–1988. Otolaryngol Head Neck
1. Charabi S, Thomsen J, Mantoni M, Charabi B, Jorgensen B, Surg. 1990;103(1):40–5.
Borgesen SE, et al. Acoustic neuroma (vestibular schwannoma): 24. Ojemann RG. Management of acoustic neuromas (vestibular
growth and surgical and nonsurgical consequences of the wait-and- schwannomas) (honored guest presentation). Clin Neurosurg.
see policy. Otolaryngol Head Neck Surg. 1995;113(1):5–14. 1993;40:498–535.
2. Charabi S, Tos M, Thomsen J, Charabi B, Mantoni M. Vestibular 25. Sterkers JM, Morrison GA, Sterkers O, El-Dine MM. Preservation
schwannoma growth—long-term results. Acta Otolaryngol Suppl. of facial, cochlear, and other nerve functions in acoustic neuroma
2000;543:7–10. treatment. Otolaryngol Head Neck Surg. 1994;110(2):146–55.
3. Rosenberg SI. Natural history of acoustic neuromas. Laryngoscope. 26. Sugita K, Kobayashi S. Technical and instrumental improvements
2000;110(4):497–508. in the surgical treatment of acoustic neurinomas. J Neurosurg.
4. Tschudi DC, Linder TE, Fisch U. Conservative management of uni- 1982;57(6):747–52.
lateral acoustic neuromas. Am J Otol. 2000;21(5):722–8. 27. Symon L, Bordi LT, Compton JS, Sabin IH, Sayin E. Acoustic neu-
5. Nikolopoulos TP, Fortnum H, O’Donoghue G, Baguley D. Acoustic roma: a review of 392 cases. Br J Neurosurg. 1989;3(3):343–7.
neuroma growth: a systematic review of the evidence. Otol 28. Tonn JC, Schlake HP, Goldbrunner R, Milewski C, Helms J, Roosen
Neurotol. 2010;31(3):478–85. K. Acoustic neuroma surgery as an interdisciplinary approach: a
6. Deen HG, Ebersold MJ, Harner SG, Beatty CW, Marion MS, Wharen neurosurgical series of 508 patients. J Neurol Neurosurg Psychiatry.
RE, et al. Conservative management of acoustic neuroma: an outcome 2000;69(2):161–6.
study. Neurosurgery. 1996;39(2):260–4; discussion 4–6. 29. Tos M, Thomsen J, Harmsen A. Results of translabyrinthine
7. Hoistad DL, Melnik G, Mamikoglu B, Battista R, O’Connor CA, removal of 300 acoustic neuromas related to tumour size. Acta
Wiet RJ. Update on conservative management of acoustic neuroma. Otolaryngol Suppl. 1988;452:38–51.
Otol Neurotol. 2001;22(5):682–5. 30. Wiet RJ, Mamikoglu B, Odom L, Hoistad DL. Long-term results of
8. Levo H, Pyykko I, Blomstedt G. Non-surgical treatment of vestibu- the first 500 cases of acoustic neuroma surgery. Otolaryngol Head
lar schwannoma patients. Acta Otolaryngol Suppl. 1997;529:56–8. Neck Surg. 2001;124(6):645–51.
9. Samii M, Matthies C. Management of 1000 vestibular schwannomas 31. Gormley WB, Sekhar LN, Wright DC, Kamerer D, Schessel
(acoustic neuromas): the facial nerve–preservation and restitution of D. Acoustic neuromas: results of current surgical management.
function. Neurosurgery. 1997;40(4):684–94; discussion 94–5. Neurosurgery. 1997;41(1):50–8; discussion 8–60.
10. Yashar P, Zada G, Harris B, Giannotta SL. Extent of resection and 32. Shelton C, Brackmann DE, House WF, Hitselberger WE. Middle
early postoperative outcomes following removal of cystic vestibular fossa acoustic tumor surgery: results in 106 cases. Laryngoscope.
schwannomas: surgical experience over a decade and review of the 1989;99(4):405–8.
literature. Neurosurg Focus. 2012;33(3):E13. 33. Arriaga MA, Chen DA, Fukushima T. Individualizing hearing pres-
11. Glasscock 3rd ME, Pappas Jr DG, Manolidis S, Von Doersten PG, ervation in acoustic neuroma surgery. Laryngoscope. 1997;107(8):
Jackson CG, Storper IS. Management of acoustic neuroma in the 1043–7.
elderly population. Am J Otol. 1997;18(2):236–41; discussion 41–2. 34. Haines SJ, Levine SC. Intracanalicular acoustic neuroma: early sur-
12. Silverstein H, McDaniel A, Norrell H, Wazen J. Conservative man- gery for preservation of hearing. J Neurosurg. 1993;79(4):515–20.
agement of acoustic neuroma in the elderly patient. Laryngoscope. 35. Irving RM, Jackler RK, Pitts LH. Hearing preservation in patients
1985;95(7 Pt 1):766–70. undergoing vestibular schwannoma surgery: comparison of middle
13. Ebersold MJ, Harner SG, Beatty CW, Harper Jr CM, Quast fossa and retrosigmoid approaches. J Neurosurg. 1998;88(5):
LM. Current results of the retrosigmoid approach to acoustic neuri- 840–5.
noma. J Neurosurg. 1992;76(6):901–9. 36. Nadol Jr JB, Chiong CM, Ojemann RG, McKenna MJ, Martuza
14. Koos WT, Matula C, Levy D, Kitz K. Microsurgery versus radio- RL, Montgomery WW, et al. Preservation of hearing and facial
surgery in the treatment of small acoustic neurinomas. Acta nerve function in resection of acoustic neuroma. Laryngoscope.
Neurochir Suppl. 1995;63:73–80. 1992;102(10):1153–8.
15. Sampath P, Holliday MJ, Brem H, Niparko JK, Long DM. Facial 37. Post KD, Eisenberg MB, Catalano PJ. Hearing preservation in ves-
nerve injury in acoustic neuroma (vestibular schwannoma) surgery: tibular schwannoma surgery: what factors influence outcome? J
etiology and prevention. J Neurosurg. 1997;87(1):60–6. Neurosurg. 1995;83(2):191–6.
16. Modugno GC, Pirodda A, Ferri GG, Fioravanti A, Calbucci F, Pezzi 38. Slattery 3rd WH, Brackmann DE, Hitselberger W. Middle fossa
A, et al. Small acoustic neuromas: monitoring the growth rate by approach for hearing preservation with acoustic neuromas. Am J
MRI. Acta Neurochir (Wien). 1999;141(10):1063–7. Otol. 1997;18(5):596–601.
17. Silverstein H, Rosenberg SI, Flanzer JM, Wanamaker HH, Seidman 39. Gardner G, Robertson JH. Hearing preservation in unilateral acous-
MD. An algorithm for the management of acoustic neuromas tic neuroma surgery. Ann Otol Rhinol Laryngol. 1988;97(1):55–66.
regarding age, hearing, tumor size, and symptoms. Otolaryngol 40. Committee on Hearing and Equilibrium guidelines for the evalua-
Head Neck Surg. 1993;108(1):1–10. tion of hearing preservation in acoustic neuroma (vestibular
18. Gjuric M, Wigand ME, Wolf SR. Enlarged middle fossa vestibular schwannoma). American Academy of Otolaryngology-Head and
schwannoma surgery: experience with 735 cases. Otol Neurotol. Neck Surgery Foundation, INC. Otolaryngol Head Neck Surg.
2001;22(2):223–30; discussion 30–1. 1995;113(3):179–80.
19. Van Leeuwen JP, Cremers CW, Theunissen EJ, Marres EH, Meyer 41. Brackmann DE, Owens RM, Friedman RA, Hitselberger WE, De la
E. Translabyrinthine and transotic surgery for acoustic neuroma. Cruz A, House JW, et al. Prognostic factors for hearing preservation
Clin Otolaryngol Allied Sci. 1994;19(6):491–5. in vestibular schwannoma surgery. Am J Otol. 2000;21(3):417–24.
20. Wiegand DA, Ojemann RG, Fickel V. Surgical treatment of acous- 42. Glasscock 3rd ME, Hays JW, Minor LB, Haynes DS, Carrasco
tic neuroma (vestibular schwannoma) in the United States: report VN. Preservation of hearing in surgery for acoustic neuromas. J
from the Acoustic Neuroma Registry. Laryngoscope. 1996; Neurosurg. 1993;78(6):864–70.
106(1 Pt 1):58–66. 43. Rowed DW, Nedzelski JM. Hearing preservation in the removal of
21. House JW, Brackmann DE. Facial nerve grading system. intracanalicular acoustic neuromas via the retrosigmoid approach. J
Otolaryngol Head Neck Surg. 1985;93(2):146–7. Neurosurg. 1997;86(3):456–61.
44. Roche PH, Noudel R, Regis J. Management of radiation/radiosurgi- 47. Yang I, Aranda D, Han SJ, Chennupati S, Sughrue ME, Cheung
cal complications and failures. Otolaryngol Clin North Am. SW, et al. Hearing preservation after stereotactic radiosurgery for
2012;45(2):367–74, ix. vestibular schwannoma: a systematic review. J Clin Neurosci.
45. Hayhurst C, Monsalves E, Bernstein M, Gentili F, Heydarian M, 2009;16(6):742–7.
Tsao M, et al. Predicting nonauditory adverse radiation effects fol- 48. Slattery 3rd WH. Microsurgery after radiosurgery or radiotherapy
lowing radiosurgery for vestibular schwannoma: a volume and for vestibular schwannomas. Otolaryngol Clin North Am.
dosimetric analysis. Int J Radiat Oncol Biol Phys. 2012; 2009;42(4):707–15.
82(5):2041–6. 49. Myrseth E, Moller P, Pedersen PH, Lund-Johansen M. Vestibular
46. Kano H, Kondziolka D, Khan A, Flickinger JC, Lunsford schwannoma: surgery or gamma knife radiosurgery? A prospec-
LD. Predictors of hearing preservation after stereotactic radiosur- tive, nonrandomized study. Neurosurgery. 2009;64(4):654–61;
gery for acoustic neuroma. J Neurosurg. 2009;111(4):863–73. discussion 61–3.
Acoustic Tumors:
Viewpoint—Stereotactic Radiotherapy 25
Christopher J. Farrell and David W. Andrews
small- and medium-sized acoustic neuromas, Sughrue et al.

Introduction found that tumors exhibiting a growth rate of greater than
2.5 mm/year were significantly (75 % vs. 32 %) less likely to
With the widespread availability of MRI, patients with retain serviceable hearing compared to those with slower
acoustic neuromas are being increasingly diagnosed earlier growing tumors [9]. In a large series of patients from
in their clinical course, and the need for high-quality data to Denmark managed conservatively, Stangerup et al. identified
guide both patients and clinicians through the treatment deci- initial speech discrimination score as predictive of long-term
sion-making process has become paramount [1, 2]. Complicating hearing retention. Among patients with a speech discrimina-
this decision-making process is the unpredictable natural tion score of 100 % at the time of tumor diagnosis, 69 %
history of acoustic neuromas with wide variability in terms maintained good hearing after more than 10 years of obser-
of growth rate and symptomatic progression. Several meta- vation compared to only 38 % of patients with a slight
analyses have been performed demonstrating that only discrimination loss at the time of initial diagnosis [10]. The
43–54 % of tumors will show radiographic enlargement over results of any radiation or surgical intervention aimed at
an observational period of approximately 3 years and that a hearing preservation must be carefully balanced against the
small percentage (4–6 %) of tumors will regress in size with- anticipated rate of hearing maintenance during a period of
out intervention [3, 4]. Many studies have attempted to more observation based on the individual clinical, radiographic,
accurately predict the natural history of untreated acoustic and audiometric data.
neuromas; however, no resolute clinical or radiographic indi-
cators of progression have been identified. Hajoff et al. dem-
onstrated that extracanalicular tumors may be more likely to Stereotactic Radiosurgery
fail conservative management compared to intracanalicular
tumors, while other studies have suggested that younger age Since Lars Leksell treated the first acoustic neuroma patient
and growth within the first year of radiographic surveillance with stereotactic radiosurgery (SRS) at the Karolinska
may be predictive of subsequent progression [5–7]. Unfortu- Institute in 1969, a steady increase in the utilization of radia-
nately, these studies and others have consistently shown that tion therapy as either primary therapy or for residual disease
hearing may irrevocably deteriorate even in the absence of following microsurgical resection has been observed with
radiographic tumor growth, contributing to the fear that an the results of modern SRS series consistently demonstrating
observation-type approach may lead to missed opportunities its efficacy and safety [11–18]. Stereotactic techniques cur-
for hearing preservation [7, 8]. In a comprehensive review rently employed include SRS and fractionated stereotactic
of published studies containing hearing outcome data for radiotherapy (FSR) with SRS most commonly delivered
patients who underwent conservative management for using either a Gamma Knife (GK) or linear accelerator
(LINAC). Early experience with SRS delivered prescription
doses ranging from 16 to 20 Gy and achieved excellent rates
This is an adaptation of a chapter that was originally published in of local tumor control (89–100 %). However, the observed
Controversies in Stereotactic Radiosurgery edited by Jason P. Sheehan rates of facial and trigeminal neuropathy were 29–67 % and
and Peter C. Gerszten published by Thieme Medical Publishers, 2013.
33–59 %, respectively, with multivariate analysis revealing
C.J. Farrell, M.D. (*) • D.W. Andrews, M.D. that tumor size and prescription dose correlated with the
Department of Neurological Surgery, Thomas Jefferson University,
901 Walnut Street, 3rd floor, Philadelphia, PA 19107, USA development of cranial neuropathy [19–21]. Following intro-
e-mail: Christopher.farrell@jefferson.edu duction of MRI-based treatment planning and reduction of
356 C.J. Farrell and D.W. Andrews
the prescription dose, facial and trigeminal nerve morbidity level while still prescribing an effective dose necessary to
was significantly decreased without change in tumor control achieve tumor control [32–36]. Despite differences in dose
[22]. Multiple large SRS series with at least 5-year median conformality and homogeneity, the results of LINAC-based
follow-up periods have yielded local tumor control rates of SRS are highly comparable to those of GK regarding both
89–100 % using prescription doses of 12–14 Gy; however, tumor control and hearing preservation [16, 28, 37–40]. The
attempts at further SRS dose reduction have met with unac- long-term decline in hearing experienced by patients
ceptable levels of tumor recurrence [23]. following SRS has led our group and others to investigate the
While the efficacy and safety of SRS for small- and role of FSR to increase the therapeutic threshold and better
medium-sized acoustic neuromas has been reliably estab- preserve function in patients with acoustic neuromas and
lished, less clear has been the long-term results of SRS treat- serviceable hearing.
ment with regard to functional hearing preservation defined
as Gardner-Robertson (GR) classification I or II or American
Academy of Otolaryngology-Head and Neck Surgery (AAO- Fractionated Stereotactic Radiotherapy
HNS) class A or B hearing [24]. In a prospective study
assessing quality of life (QOL) measures in patients under- As summarized above, SRS represents the pioneering inter-
going observational, surgical, and radiation therapy for vention that broadened the treatment alternatives for patients
acoustic neuromas, nearly one-third of patients cited hearing with acoustic neuromas. Originally practiced on the Gamma
as the single most important factor affecting their QOL [25]. Knife, radiosurgery techniques were soon adapted to retrofit-
Similar to the natural history of hearing loss with these ted LINACs [40] and subsequently to linear accelerators
tumors, hearing loss associated with SRS treatment has also designed for and dedicated to radiosurgery [41]. It was this
shown great variability with the overall results remaining latter development that enabled the design of fractionated
inconclusive in terms of their superiority relative to conser- stereotactic techniques with noninvasive relocatable frames
vative management. In a systematic analysis of the literature, and marked the beginning of modern stereotactic techniques
Yang et al. reported a hearing preservation rate following described below. The sections below discuss the radiobio-
SRS of 57 % with an average follow-up period of 41 months logical rationale and advantages of FSR as well as published
[26]. The only significant variable found to positively corre- outcomes as they relate to acoustic tumors.
late with hearing preservation was marginal treatment dose
≤12.5 Gy. Hearing loss may occur as early as 3 months fol-
lowing SRS with the potential to decline many years after Radiobiologic Principles of Fractionation
treatment [27, 28]. Chopra et al. reported that after 3 years of for Selected Late-Responding Tissues
follow-up, the hearing preservation rate was 74 % but had
declined to 44 % at 10 years [11, 21]. Similarly, Roos We previously reviewed the advantages of FSR and derived
reported a 5- and 10-year hearing preservation rate of 57 % a hearing RET formula as a formula to guide dose-
and 24 %, respectively [28]. fractionation strategies [42] discussed below. SRS has made
Radiation therapy is believed to induce sensorineural the treatment of skull base tumors far more precise with
hearing loss (SNHL) via injury to the cochlea, cochlear modern imaging and versatile three-dimensional treatment
nerve, or brainstem auditory nuclei, and many recent studies planning software that, through dose-volume histograms,
have focused on identifying patient-related and dosimetry maximizes dose to target with high conformality and mini-
variables that may allow for improved preservation of func- mizes dose to contiguous normal structures. Two exceptions
tional hearing. Several studies have demonstrated that are optic nerve sheath meningiomas and acoustic neuromas
patients with GR class I were more likely to retain functional where these cranial nerves are intrinsic to the target volume.
hearing compared to GR class 2 patients suggesting that As special sensory cranial nerves, injury to sensory function
interventions aimed at hearing preservation should be con- occurs much more frequently than either sensory or motor
sidered as early as possible, particularly for younger patients function in mixed cranial nerves, reflecting a lower thresh-
[29–31]. Additionally, retrospective studies have also sug- old for injury. To make matters even more complicated
gested that radiation dose to the cochlea may be a critical when considering acoustic neuromas, recent literature has
determinant of preserved hearing function. Kano et al. focused on the cochlea itself which may have a different
reported the results of 77 patients treated with a marginal radiobiological tolerance [31]. We will advance arguments
dose of 12.5 Gy and found that after controlling for GR class, based on published data that tumors involving or near spe-
patients who received a central cochlear dose <4.2 Gy were cial sensory cranial nerves, when treatment is indicated,
more likely to retain serviceable hearing [31]. Unfortunately, should be treated with FSR utilizing daily low conventional
it is not always possible to limit the cochlear dose to a safe fraction sizes.
25 Acoustic Tumors: Viewpoint—Stereotactic Radiotherapy 357
de-escalation for FSR, we have found the lower dose cohort

The Example of Optic Nerve Sheath of 46.8 Gy achieves an excellent tumor control rate while
Meningiomas improving hearing preservation [30]. It remains unclear,
however, if further dose reductions improve hearing deficits
Single doses of 15 or 54 Gy in 30 daily fractions both result even more while still maintaining high rates of tumor
in excellent control of meningiomas [43]. The SRS dose of control.
15 Gy, however, exceeds the generally accepted single-dose To guide and optimize future treatments, we need to
tolerance of the optic nerves (8–10 Gy) and may be associ- understand the radiobiological rationale for fractionation for
ated with a risk of optic neuropathy approaching 78 % [44]. acoustic neuromas. We must first acknowledge that the target
For fractionated radiotherapy, decades of experience have is a late-responding tissue radiographically inseparable from
led to guidelines for normal tissue tolerance. For optic nerve a special sensory cranial nerve that could be compromised
and chiasm, guidelines have evolved that have proven to be by radiation treatment. Fractionation allows time between
extremely safe and are generally based on the optic RET fractions for normal tissues (i.e., cochlear nerve) to repair
dose of 8.9 Gy recommended by Goldsmith et al. [45], later sublethal damage, while tumor tissue is less efficient at repair
corroborated by a large retrospective analysis by Parsons due to faulty repair mechanisms. This leads to the separation
et al. [46]. We adopted the observations set forth by Parsons of tumor and normal tissue cell survival curves, thereby min-
and dropped the daily dose to the optic nerve to 1.8 Gy and imizing normal tissue toxicity while maximizing tumor cell
the total dose to 54 Gy—both below the threshold for injury kill. This is depicted in Fig. 25.1 below.
established by the optic RET formula and corroborated by The radiobiology of high dose single fraction involves
Parsons. Our results yielded excellent tumor control and a more than sublethal damage, and the major effects have been
compelling preservation or improvement in vision in these documented to involve both lethal DNA damage and endo-
patients [47]. These were the first data supporting FSR as a thelial damage. While favorable for tumor cell kill, these
means of controlling a benign tumor while simultaneously
preserving or restoring the function of a special sensory cra-
nial nerve.
The Biologically Effective Dose

for Acoustic Neuromas
We previously noted that reports of hearing preservation are

not uniform, ranging from subjective hearing outcomes to
standard audiometric outcomes scales. The most widely
cited scale is the Gardner-Robertson scale [24], but despite
this scale there still exist disparities in reported outcomes
due to differences, for example, in the derivation of pure tone
average. We reviewed published literature and selected only
those studies that featured audiometric outcomes utilizing
the Gardner-Robertson scale confirmed by audiometric data
[42]. Other scales or outcomes based on subjective hearing
results were excluded. The most informative reports included
the actual audiometric data, but most reports did not. As we
previously noted, radiosurgery doses that reflected excellent
tumor control rates also reflected Gamma Knife dose de-
escalations over time to improve rates of cranial neuropathy.
Other than our recent publication [30], literature does not
Fig. 25.1 Graph illustrating the effect of fractionation on the cell sur-
reflect dose iterations for FSR. From past literature, high vival of tumor cells (A) as well as normal cells (B). Since normal cells
rates of tumor control have been achieved for both a single have intact repair mechanisms, whereas tumor cells do not, fraction-
dose of 12–13 and 45–50 Gy in 1.8 to 2 Gy daily fractions, ation effectively separates the cell survival curves for tumor cells and
suggesting a dose equivalence for treatment of acoustic neu- normal cells, thereby sparing normal cells while killing tumor cells.
Published with permission from Balagamwala EH, Chao ST, Suh
romas. Unlike the optic apparatus, however, acceptably JH. Principles of radiobiology of stereotactic radiosurgery and clinical
low rates of hearing loss have not yet been generally applications in the central nervous system. Technol Cancer Res Treat.
agreed upon. In our single-institution experience of dose 2012;11(1):3–13. Epub 2011/12/21. http://www.tcrt.org
dual effects may have adverse effects on the normal neuropil steeper and hence shorter route. Carrying the analogy
and vasculature of the cochlear nerve. forward, the steeper route represents a shorter distance, or in
To better understand an optimal dose regimen for the biological terms, the use of larger fractions results in a
treatment of acoustic neuromas, it is important to review the shorter time course with a smaller total dose.
concept of a biologically effective dose, or BED, which is Currently, disparate focused radiation dose/fractionation
useful for quantifying treatment expectations [49] and is an schedules are practiced worldwide for acoustic tumors,
inherent part of the linear quadratic formula for radiation including single fraction SRS, hyperfractionated FRS, and
effect. For a review of the derivation and significance of α/β hypofractionated FSR. Doses of 12 Gy in a single fraction,
[alpha/beta] ratios and the linear quadratic equation, the 46.8 Gy in 1.8 Gy daily fractions, or 18 Gy in three fractions
reader is referred to an excellent recent review by all achieve excellent tumor control rates with serviceable
Balagamwala et al. [48]. The radiation cell kill (or effect E) hearing preservation rates that vary roughly from 60 to 80 %.
can be expressed as: One might argue that these dose/fractionation schedules are
not biologically equivalent because of variable outcomes in
(
Cell kill = E = n a [alpha ] d + b [ beta ] d 2 ) (25.1) hearing preservation, but for purposes of this model, we will
assume they are. We can correspondingly assign values of
where α [alpha] and β [beta] are the radiosensitivity coeffi- BED1, BED2, and BED3, respectively, to these dose/fraction
cients, n the number of fractions, and d the dose per fraction schemes. Assuming BED1 = BED2 for tissue of an unknown
such that the total dose (D) = nd. α/β, or in terms of total dose D and fraction dose d:
Historically, radiosurgical dose reductions for acoustic
tumors have been performed empirically to improve hearing D1[1 + d1/(α/β[alpha/beta])] = D2[1 + d2/(α/β[alpha/
outcomes. Consistent with this exercise, consider a progres- beta])] or α/β[alpha/beta] = (D1 × d1 − D2 × d2)/D2 − D1
sive reduction in d approaching but not reaching 0. Although
For acoustic neuromas using the assumption of equivalent
the number of fractions n will then need to be increased to
BED for 12 Gy in a single fraction (d1 = 12 Gy) and 46.8 Gy
maintain the same effect, β[beta]d2 will be very small in
in 26 fractions (d2 = 1.8 Gy):
comparison to α[alpha]d, and when d is very small, the above
equation can be simplified to: a / b [alpha / beta ] = éë(12 ´ 12 ) - ( 46.8 ´ 1.8 ) ùû ( 46.8 - 12 )
E = n a [alpha ] d = a [alpha ] D (25.2) = 1.72
and the corresponding BED based on this α/β[alpha/beta]
This demonstrates that the total dose D of radiotherapy ratio is calculated as:
given at a very low dose per fraction represents the highest
total dose required to obtain a specific effect, in this case BED = 12 (1 + 12 / 1.72 ) = 46.8 (1 + 1.8 / 1.72 ) = 95.7 Gy1.72
simultaneously high rates of tumor control and hearing pres-
ervation. This total dose represents the biologically effective Now having derived an isoeffective BED for acoustic
dose or BED in situations where neoplastic cellular repopu- tumors, we can test an alternate dose-fractionation schedule
lation can be ignored, which is the case for acoustic neuro- by substituting actual values from Eq. (25.3) above for BED3,
mas. Equation (25.2) above can be redefined as: an actual dose-fractionation schedule recently reported [50]
to see if our derived α/β [alpha/beta] ratio actually yields the
BED = D = E / a [alpha ] same dose per fraction solving for d as the unknown:
é d ù
With this assumption, BED for any practical radiotherapy 3d ê1 + ú = 95.7 Gy1.72
application can be derived by first dividing both sides of Eq. ë 1.72 û
(25.1) by α [alpha] such that: which is rearranged to:
3d + 1.74 d 2 - 95.7 = 0
BED = E / a =
(
n a d + b d2 ) = nd é1 + d ù
(25.3)
a ê a / b úû For a general quadratic equation such as ad2 + bd + c = 0,
ë
the positive solution for d is given by:
It is perhaps difficult to understand that one BED can be
obtained by different dose/fractionation schedules. Drawing -b + (b 2
- 4 ac )
from an analogy by Jones and colleagues [49], the height of 2a
a mountain could be viewed as a BED. BED represents the
absolute height using the longest route (i.e., the lowest dose) where in this case a = 1.74, b = 3, and c = −95.7. The solution
to the summit, but it is possible to climb to the summit by a for d above is 6.6 Gy which is within a 10 % error of 6 Gy.
As a working model, therefore, an α/β[alpha/beta] ratio of where N is the number of fractions. This can be folded into:
1.72 yielding an isoeffective BED of 95.7 Gy is a reasonable D = 1,120 × N0.43 or 1,120 hearing RET = D × N − 0.43 for
derivation from current literature regarding focused radiation dose/fraction schemes with high hearing preservation rates.
for acoustic neuromas. For one fraction: D = 11.2 Gy. This implies that the radio-
While an α/β[alpha/beta] ratio of 1.72 is low and con- surgical dose should not exceed a threshold around 11.2 Gy
sistent with late-responding normal CNS tissues [51], this to obtain high rates of hearing preservation. Based on this
model has potential pitfalls including an oversimplification hearing RET formula, single fraction doses above 11.2 Gy or
of single Gamma Knife treatments which necessarily involve a cumulative FSR dose above 45 Gy in a 25 fraction regimen
multiple isocenter hot spots, and a two-target model (DNA could result in higher rates of hearing loss. For SRS, any
damage and endothelial cell apoptosis) may be necessary, dose below 13 Gy results in loss of tumor control [23], so
especially at high doses per fraction. In the former case, a even with an approximate 60 % hearing preservation rate,
single one-point calculation will not be representative of the 13 Gy may represent the most optimal dose achievable with
biological effect throughout a larger acoustic tumor with SRS. Based on our previous literature review, a median value
multiple hot spots, and multiple BED evaluations would be of 13 Gy has, in fact, resulted in reported serviceable hearing
necessary [49]. In the latter case, a modification of the linear losses ranging from 17 to 67 % [42]. A median FSR dose of
quadratic equation such as a generalized LQ (gLQ) [52] or a 47.5 Gy, while maintaining excellent tumor control, has
universal survival curve (USC) [53] might be more appropri- resulted in reported serviceable hearing losses ranging from
ate. As a starting point, however, we felt this model would be 29 to 43 % [42]. As noted in Fig. 25.2b below, the dose pre-
useful for dose-fractionation schedules designed for treat- scription schemes with the highest yield of hearing preserva-
ment of acoustic tumors. Its utility as a predictive model of tion fall within the box that includes the 10 % variance of the
serviceable hearing loss, however, remains unclear due to the α/β [alpha/beta] ratio we derived. The single point represents
significant variations in hearing outcomes mentioned above. a recently published paper which describes a high probabil-
ity of hearing loss with a dose-fractionation scheme for
acoustic neuromas described as a total dose of 54 Gy in 27
Derivation of a Hearing Ret Formula fractions [55]. Applying an α/β [alpha/beta] ratio of 1.63 for
for Acoustic Neuromas tumor control and the hearing RET formula as a guide
for hearing preservation, the total dose should not exceed
We felt a more systematic approach to serviceable hearing 46.2 Gy for 27 fractions to maximize the chance of hearing
preservation would be the derivation and application of a preservation while maintaining excellent tumor control.
hearing RET formula [42]. A prospective trial by Pan et al.
[54] documented that for almost all cases in which significant
hearing loss occurred in the affected ear receiving radiation, Further Radiobiological Advantages:
the dose was ≥45 Gy. Among published data available, we The Example of Large Tumors
were able to plot the log of dose versus the log of the number
of fractions drawing from Goldsmith’s model [3] for reports With the progressive shift in treatment patterns toward use of
with hearing preservation rates of at least 70 % based on radiation therapies for acoustic neuromas, some authors have
Gardner-Robertson criteria. The linear regression gave us a suggested that relative indications for microsurgical resec-
formula for dose/fraction size regimens with a high correlation include the presence of medically refractory trigeminal
tion coefficient (R2 = 0.9705). Drawing from Goldsmith’s neuralgia, hydrocephalus, symptomatic mass effect includ-
optic RET derivation, we postulated that dose/fraction ing disabling ataxia or hemiparesis, intractable headache,
schemes with very high hearing preservation rates could be and large tumor size. Although the definition of large tumor
represented by a parallel line on the same plot, intersecting a size remains somewhat arbitrary, maximal tumor diameter
point corresponding with the dosage/fraction scheme docu- >3 cm has conventionally been used as a cutoff for SRS
mented as safe. Data from Pan et al. [54] suggest that 45 Gy treatment based on studies demonstrating rates of effective
in 25 × 1.8 Gy fractions (or less) is such a point. By substitu- tumor control as low as 57 % and increased risk of complica-
tion to the linear regression equation: tion compared to SRS treatment of small- and medium-sized
tumors [56]. Milligan et al. described their experience using
Intercept = log ( 4500 ) - 0.4308 ´ log ( 25 ) SRS for tumors with volumes more than one standard devia-
with an intercept at 3.051. tion above the mean during a 9-year period and demonstrated
a 5-year actuarial tumor control rate of 83 % [57]. Impairment
So the line equation is: of facial nerve function was observed in 14 % of patients and
15 % required prolonged corticosteroid administration for sym-
log D = 0.431 ´ log N + 3.051 ptomatic radiation-induced brainstem or cerebellum edema.
Fig. 25.2 (a) Logarithmic plot of total dose versus number of fractions or below the hearing RET curve depending on which α/β [alpha/beta]
for cases reporting preservation of serviceable hearing; (b) plots of bio- ratio is utilized. The black arrow represents the point beyond which dose-
logically equivalent doses utilizing the linear quadratic formula with α/β fractionation schemes predict a higher likelihood of hearing preservation
[alpha/beta] ratios as +10 % of the calculated α/β [alpha/beta] value, in for a lower α/β [alpha/beta] ratio, the gray arrow for a higher α/β [alpha/
comparison with the hearing RET plot. The dose prescription schemes beta] ratio. The black point represents a recent study which documented
predicting the highest yield of hearing preservation fall within the box at a high rate of hearing loss [56] and falls above the hearing RET range
Table 25.1 Results of the ANA surveys over three decades

Treatment reported by respondents 2007–2008 (%) 1998 (%) 1983 (%)
Translabyrinthine approach 33 51 72
RETrosigmoid approach 17 28 11
Middle fossa approach 10 6 3
Don’t know approach 0 0 14
Total, microsurgical resection 60 85 100
Stereotactic radiosurgery (SRS) 12 NA NA
Fractionated stereotactic radiotherapy (FSR) 8 NA NA
Total SRS/FSR 20 5 0
Watch and wait 20 4 0
Don’t know what type of treatment NA 6 0
Total 100 100 100
Based on their results, the authors recommended consideration 1 year posttreatment. Interestingly, of the 12 patients included
of surgical resection for tumors larger than 2.7 cm in poste- in the study who reported pretreatment symptoms of trigem-
rior fossa greatest dimension. A group from the Netherlands inal hypoesthesia, 11 (92 %) experienced resolution of their
recently reported primary SRS treatment of 33 patients with condition following SRS likely related to the 67 % inci-
large tumors defined as tumor volume >6 cm3 with indenta- dence of tumor volume shrinkage observed on radiographic
tion of brainstem (Koos Grade III & IV). Patients with NF2, follow-up (Table 25.1).
symptomatic mass effect, or maximum extracanalicular The use of FSR has been recommended as an alternative
dimension ≥4 cm were excluded from primary SRS treat- radiation modality for the treatment of larger tumors adja-
ment. A radiographic tumor control rate of 88 % was cent to the brainstem although minimal data exists to suggest
reported over a median follow-up time of 30 months, improved outcome relative to SRS for this population of
although 7 of 33 patients required further microsurgical tumors. Mandl et al. reported a cohort of patients with tumors
or SRS treatment for development of either progressive >3 cm in diameter treated primarily with multisession SRS
neurologic symptoms or radiographic enlargement. using a maximum dose of 25 Gy delivered in 5 Gy fractions
Transient facial and trigeminal neuropathy occurred in 9 % [58]. The actuarial 5-year tumor control and facial nerve
and 14 % of patients, respectively, with good facial nerve function rates were 82 % and 80 %, respectively. We have
function (HB Grades I and II) preserved in all patients at updated our review of FSR treatment for acoustic neuromas,
Fig. 25.3 Serial magnetic resonance imaging scans (upper) with asso- of 100 %, representing a decrease to Gardner-Robertson level 2 service-
ciated audiograms obtained at same time (lower) in patient treated with able hearing. (c) Posttreatment T1-weighted gadolinium-enhanced
FSR to 46.8 Gy. (a) Pretreatment T1-weighted gadolinium-enhanced axial image at 42 months featuring marked diminution in tumor size.
axial image featuring right acoustic neuroma. Audiogram reflects pure Audiogram at 42 months after treatment reflects stable audiogram with
tone average of 22 and speech discrimination score of 95 %, represen- pure tone average of 30 and speech discrimination score of 92 %, back
ting Gardner-Robertson level 1 pretreatment serviceable hearing. to lowest range of Gardner-Robertson level 1 serviceable hearing. From
(b) Posttreatment T1-weighted gadolinium-enhanced axial image at 6 Andrews DW, Werner-Wasik M, Den RB, Paek SH, Downes-Phillips B,
months featuring some enlargement and extensive central necrosis of Willcox TO, et al. Toward dose optimization for FSR for acoustic neu-
tumor. Audiogram at 6 months after treatment reflects some audiomet- romas: comparison of two dose cohorts. Int J Radiat Oncol Biol Phys.
ric decay with pure tone average of 35 and speech discrimination score 2009;74(2):419–26
and of 154 tumors treated in the 46.8 Gy cohort, 15 were We will first address dose prescription conventions (see
>6 cc in size. In this subgroup, we have documented no treat- Fig. 25.4). While a Gamma Knife isosurface prescription is
ment failures, and of 6 patients with serviceable hearing, classically at the 50 % isodose line, LINAC-based treatments
3 have maintained hearing in a serviceable range at a median typically fall between an 85 and 95 % isodose line. Assuming
follow-up of 35 months. At conventional fraction sizes, we in both cases circular collimation is utilized, the 50 % isosur-
have also noted no other cranial neuropathies in this sub- face yields a steeper dose gradient in the vicinity of the
group of patients. See Fig. 25.3 as an example. cochlear nerve when compared to higher LINAC-based iso-
The theoretical advantages of LINAC-based fractionation surface prescriptions and thus results in a higher dose to the
for large tumors are depicted in Figs. 25.4 and 25.5 below. cochlear nerve (see Fig. 25.4). Our analysis is based on an
For larger treatment volumes, the penumbra affecting con- actual comparison of a multi-shot/high conformality Gamma
tiguous normal tissues is often judged to be prohibitive for Knife SRS treatment to a single isocenter 5 dynamic arc FSR
single fraction SRS that is substantiated in a broad literature. treatment on a Novalis Classic. The dose distribution penum-
In a review of 149 patients with posterior fossa tumors bra is a depiction of a circular collimator instead of a dynamic
treated with SRS, for example, factors that predicted cranial arc penumbra to allow a uniform comparison of dose pre-
nerve palsy included Dmax of >17.5 Gy, a prescribed isosur- scription conventions with both techniques.
face dose of >12.5 Gy, length of cranial nerve of >16 mm, We will next address dose delivery conventions (see
and tumor volume of 1.7 cc [59]. Particularly for larger Fig. 25.5). The prior section reviewed the advantages of frac-
acoustic tumors, both the dose prescription convention and tionation and staying within the hearing RET. Beyond con-
the design of multiple isocenters to achieve conformality cerns of total dose and fraction size, the manner in which the
increase the likelihood of overdosing the cochlear nerve. dose is delivered, particularly for larger tumors, will affect
Fig. 25.4 Comparison of dose distribution to the cochlear nerve with acusticus (tumor is shaded; outer line is 90 % isodose 10 prescription
Gamma Knife and FSR treatments based on isosurface prescription con- line; inner line is 95 % isodose line). Assuming cochlear nerve is around
ventions. (a) Axial T1 gadolinium-enhanced MRI scan of right acoustic the 7 o’clock position, the nerve is within a 5 % dose gradient above
neuroma. (b) Artist’s rendering of translucent acoustic tumor with cra- isodose prescription. (g) Artist’s caricature of (d) (magnified sagittal
nial nerves VII and VIII adherent to the anterior and caudal surface of the cross section of intracanalicular portion of tumor treated with Gamma
tumor, coursing to internal auditory canal. (c) Eight-shot Gamma Knife Knife). The VIII nerve at the 7 0’clock position is within the prescribed
radiosurgery treatment plan with a 12 Gy prescription to the 50 % iso- isodose line and exposed to higher dose gradients. (h) Profile of a focused
dose line (yellow) for right acoustic neuroma. (d) Magnified sagittal view radiation beam with typical isodose prescriptions at 50 % (Gamma
of actual treatment plan in the distal porus acusticus [outer line (yellow) Knife) and 90 % (FSR). Vertical columns in either scenario could repre-
is 50 % isodose prescription line; middle line (magenta) is tumor surface; sent narrow location range of cochlear nerve relative to tumor (e.g.,
inner line (green) is 60 % isodose line]. Assuming cochlear nerve is within 1 mm of tumor surface). Broad inferior horizontal bar represents
around the 7 o’clock position, the nerve is within a 10 % dose gradient potential actual dose range delivered to cochlear nerve with 50 % isodose
above isodose prescription. (e) Single-shot Novalis FSR treatment plan prescription; narrow superior horizontal bar represents potential actual
with a 1.8 Gy prescription to the 90 % isodose line for a right acoustic dose range delivered to cochlear nerve with 90 % isodose prescription.
neuroma (tumor is shaded; outer line is 90 % isodose prescription line). The potential dose gradient at a 50 % isodose prescription is more than
(f) Magnified sagittal view of actual treatment plan in the distal porus three times greater than 90 % prescription at the same distance
the function of the organs at risk (OARs), in this case the penumbra resulting from multiple shots yielding a much
cochlea and the cochlear nerve. The latest innovation in more homogeneous dose of radiation to the cochlear nerve.
LINAC-based radiosurgery platforms is the use of mini-leaf Dose homogeneity promises to minimize injury to the
collimation. Rather than sphere packing with variably sized cochlear nerve (see Fig. 25.5) and high conformality prom-
shots from circular collimators, mini-leaf collimation auto- ises to minimize injury to the cochlea, a structure too small
mates target isosurface conformality with automated and for reliable dose-volume histograms.
paired tungsten leaves which move dynamically through an The other important structure affecting hearing outcomes
arc of radiation to match target geometry. This technique is the cochlea itself and both SRS and FSRT techniques
eliminates the dose inhomogeneity within the cochlear nerve achieve high conformality limiting dose to the cochlea.
Fig. 25.5 Comparison of dose distribution to the cochlear nerve with shots. 4.2 Gy is threshold beyond which injury to the cochlea results in
Gamma Knife and FSR treatments based on dose delivery conven- hearing loss [31]. (c) After a single isocenter dynamic arc treatment
tions. (a) Axial T1 gadolinium-enhanced right acoustic neuroma, with mini-multileaf collimation treatment, the dose penumbra includ-
highlighted in red. (b) Cartoon of the same tumor now featuring the ing the cochlear nerve is now much more homogeneous and close to
cochlear nerve and cochlea (OARs are organs at risk). After a multi- the isosurface prescription dose. A total dose of 45 Gy with conven-
ple-shot, high conformality treatment, the dose penumbra including tional fraction sizes is a threshold beyond which injury to the cochlea
the cochlear nerve necessarily includes hot spots from overlapping results in hearing loss [55]
Recent data suggests that a dose <4.2 Gy for an SRS At the present time, a multitude of studies reporting
treatment [31] or 45 Gy for conventional radiation doses [54] outcomes using FSR have consistently demonstrated excel-
is favorable for cochlear function. We have since learned in lent tumor control rates comparable to those achieved with
our experience that the dose fall off outside the PTV isosur- SRS and low risk of cranial nerve morbidity. The optimal
face prescription is important when considering cochlear dose and fractionation scheme for FSR remains uncertain
dose, and an 80–85 % isosurface prescription for LINAC can with the widespread variation in treatment practices at differ-
achieve the best dose distribution for both the cochlea and ent institutions, precluding definitive comparisons of FSR to
cochlear nerve. SRS. Several recent studies have described outcomes using
FSR with maximum doses ranging from 46.8 to 57.6 Gy,
typically administered in conventional 1.8–2 Gy fractions.
History and Progress in Clinical Outcomes Overall, functional hearing preservation was observed in
Utilizing FSR 61–94 % of patients over an average follow-up time of
approximately 50 months [60–67]. In 2001, our group at
In 1987, Wallner et al. described the use of fractionated con- Thomas Jefferson University performed a nonrandomized
ventional radiation therapy (RT) for the treatment of acoustic prospective cohort analysis comparing outcomes after SRS
neuromas following subtotal resection. They demonstrated and FSR using a maximum dose of 50.4 Gy for 122 patients.
that when doses ≥45 Gy were administered, the tumor recur- Tumor control, facial nerve preservation, and trigeminal
rence rate could be significantly decreased. Several years neuropathy rates were equivalent among the two cohorts;
later, a group from France described their experience using however, hearing preservation rates were 2.5-fold higher in
RT as primary therapy for acoustic neuromas, including a the FSR group with a crude hearing preservation rate of
small subset of patients with neurofibromatosis type 2 (NF2) 81 % compared to 33 % in the SRS cohort [60]. Limitations
and preexisting unilateral deafness secondary to previous of this prospective study included short-term length of audi-
surgical resection of a contralateral tumor. At an average ometry data and an SRS hearing preservation rate toward the
follow-up of 70 months, hearing was preserved in 3 of the 5 low range of reported outcomes. Combs et al. subsequently
patients with no evidence of tumor progression. Based on reported their single-institution prospective experience com-
these encouraging early results, a number of groups posited paring SRS to FSR using marginal dose of 57.6 Gy which
that combining the radiobiologic advantages of dose frac- revealed no overall difference in terms of hearing preserva-
tionation with the precision of stereotactic techniques may tion between the two modalities when the SRS prescription
enable effective treatment of acoustic neuromas with reduced dose was ≤13 Gy [68], but the FSR group represented only
cranial neuropathy and improved hearing preservation. 10 patients.
Table 25.2 Summary of treatment outcomes for SRS and FSR 2004–2009
Treatment N Tumor volume (cc) Dose (Gy) Follow-up (months) Tumor control rate (%) Hearing preservation rate (%)
SRS 1,850 2.3 12.6 71 89 60
FSR 404 4.1 52.4 53 95.8 79
From: Arthurs BJ, Fairbanks RK, Demakas JJ, Lamoreaux WT, Giddings NA, Mackay AR, et al. A review of treatment modalities for vestibular
schwannoma. Neurosurg Rev. 34(3):265–77; discussion 77–9
Table 25.3 Summary of serviceable hearing outcomes from the 2007–2008 ANA survey
Treatment Hearing status (N) before treatment after treatment Rate of serviceable hearing preservation (%)
SRS 108 22 20
FSR 84 41 49
Similar to the early stages of SRS, the optimal FSR volume) ratios as measures, respectively, of dose homogeneity
prescription dose necessary to achieve long-term tumor and dose conformality. For circular collimation, a uniform
control while minimizing damage to normal structures such reporting mechanism should include mean number of isocen-
as the cochlea and cochlear nerve remains to be established. ters stratified by tumor sizes, and these data might build a rela-
We previously demonstrated that FSR dose reduction from tionship between such variables as tumor size, serviceable
50.4 to 46.8 Gy enabled improved hearing outcomes without hearing loss, and mean number of isodose centers as a mea-
sacrifice to tumor control. These results are consistent with sure of conformality. Newer techniques such as mini-multileaf
data using RT for nasopharyngeal cancer revealing that hear- collimation with dynamic arc or intensity modulation (Novalis/
ing loss is significantly increased when the mean cochlear TrueBeam) or non-isocentric smearing with pencil beams
dose is >48 Gy [30, 69]. Conversely, in the same patient (CyberKnife) [70] both promise to maintain conformality and
population, hearing is preserved for a mean cochlear dose increase dose homogeneity, and should also include MDPD
<45 Gy [54]. Importantly, hearing impairment after fraction- and PTV values.
ated RT appears to have a delayed time to onset relative to Despite a broad variety of dose-fractionation schedules
SRS and longer-term data will be necessary to fully evaluate without a clear rationale, recent systematic literature review
this treatment modality [27]. by Arthurs and colleagues, however, favors FSR when
Several groups have reported their experience using a assessing hearing preservation [66] (see Table 25.2). These
multisession SRS (≤5 treatment fractions) scheme in an data are remarkably similar to data we previously compiled
effort to balance the advantages of fractionation with the con- from published series [42].
venience of SRS. Hansasuta et al. described the extensive Drawing from the patients’ perspective, we once again
experience of the Stanford group delivering 18 Gy over three refer to the ANA 2007–2008 patient survey, recognizing that
treatment sessions using the CyberKnife [50]. During a all data collected were self-reported by patients and reported
median follow-up period of 3 years, they found a 98 % tumor as is. Table 25.3 indicates the self-reported Gardner-
control rate and accompanying 76 % hearing preservation Robertson Class 1 and 2 results of respondents who under-
rate. Meijer et al., however, reported the results of a pros- went either SRS or FSR.
pective cohort study comparing single versus multisession If this survey is any indication, FSR provides a more
LINAC-based SRS and observed no differences in either favorable outcome when assessing serviceable hearing
tumor control or hearing preservation rates between the two preservation.
cohorts which were 75 % and 61 %, respectively, for service-
able hearing [67]. A slight but statistically significant reduc-
tion in trigeminal neuropathy incidence was noted with Future Directions
multisession SRS.
As mentioned above, we have modeled the radiobiology of Future treatment possibilities for acoustic neuromas remain
fractionation for acoustic neuromas, including a hearing RET both challenging and hopeful. The challenge will remain the
formula as a guideline for treatment. Summarizing the above design of appropriate prospective trials that guide practitio-
literature, there remains no systematic application of radiobio- ners with useful level I data. While radiosurgery and FSR
logic principles to the treatment of acoustic neuromas. In fact, have become common practice, physicians have become
none of the FSR or SRS papers we previously analyzed pro- entrenched in the commercial platform they use which
vided uniform treatment planning data such as MDPD (maximum makes prospective randomized comparisons of different treat-
dose to prescribed dose) ratios or PTV (prescription to tumor ment techniques on different platforms extremely difficult.
Also, further refinements in our understanding of the 3. Smouha EE, Yoo M, Mohr K, Davis RP. Conservative management
radiobiological tolerances of the cochlea and the cochlear of acoustic neuroma: a meta-analysis and proposed treatment algo-
rithm. Laryngoscope. 2005;115(3):450–4.
nerve, including the hearing RET formula, will be important 4. Selesnick SH, Johnson G. Radiologic surveillance of acoustic neu-
new information critical to the design of new radiation treat- romas. Am J Otol. 1998;19(6):846–9.
ment strategies. It has been well established that atrophy of 5. Hajioff D, Raut VV, Walsh RM, Bath AP, Bance ML, Guha A, et al.
the cochlear nerve (loss of cochlear neurons) disproportion- Conservative management of vestibular schwannomas: third review
of a 10-year prospective study. Clin Otolaryngol. 2008;33(3):
ately affects speech discrimination ability rather than pure 255–9.
tone hearing thresholds [71–73]. Associating pure tone aver- 6. Matthies C, Samii M. Management of 1000 vestibular schwannomas
ages with cochlear doses and speech discrimination scores (acoustic neuromas): clinical presentation. Neurosurgery. 1997;
with cochlear nerve doses might provide further insight into 40(1):1–9; discussion 10.
7. Flint D, Fagan P, Panarese A. Conservative management of spo-
the radiobiology of these structures and guide future treat- radic unilateral acoustic neuromas. J Laryngol Otol. 2005;119(6):
ment plans. Uniform analyses of reliable audiometric data 424–8.
immediately before and at routine intervals after treatment 8. Bakkouri WE, Kania RE, Guichard JP, Lot G, Herman P, Huy
will yield better outcome data. In one recent retrospective PT. Conservative management of 386 cases of unilateral vestibular
schwannoma: tumor growth and consequences for treatment.
analysis of temporal bone histopathology in 32 patients with J Neurosurg. 2009;110(4):662–9.
acoustic neuromas, the tumors caused significantly more 9. Sughrue ME, Yang I, Aranda D, Lobo K, Pitts LH, Cheung SW,
inner and outer hair cell loss, cochlear neuronal loss, precipi- et al. The natural history of untreated sporadic vestibular schwan-
nomas: a comprehensive review of hearing outcomes. J Neurosurg.
tate in endolymph and perilymph, and decreased pure tone
2010;112(1):163–7.
average, when compared with the opposite ear. Tumor size, 10. Stangerup SE, Thomsen J, Tos M, Caye-Thomasen P. Long-term
distance from the cochlea, and nerve of origin did not corre- hearing preservation in vestibular schwannoma. Otol Neurotol.
late with structural changes in the cochlea or the hearing 2010;31(2):271–5.
11. Chopra R, Kondziolka D, Niranjan A, Lunsford LD, Flickinger
threshold [74]. With these provocative observations, cochlear
JC. Long-term follow-up of acoustic schwannoma radiosurgery
pathology rather than retrocochlear pathology may delineate with marginal tumor doses of 12 to 13 Gy. Int J Radiat Oncol Biol
important prognostic groups prior to treatment. Phys. 2007;68(3):845–51.
The optimal treatment outcome will include the excellent 12. Leksell L. A note on the treatment of acoustic tumours. Acta Chir
Scand. 1971;137(8):763–5.
tumor control rates already achieved with better hearing pres-
13. Chen DA. Acoustic neuroma in a private neurotology practice:
ervation and possibly recovery of hearing. The latter achieve- trends in demographics and practice patterns. Laryngoscope. 2007;
ment may include a combination of targeted chemotherapy and 117(11):2003–12.
radiation. In one recent application of chemotherapy, VEGF 14. Flickinger JC, Kondziolka D, Niranjan A, Maitz A, Voynov G,
Lunsford LD. Acoustic neuroma radiosurgery with marginal tumor
blockade with bevacizumab improved hearing in some, but
doses of 12 to 13 Gy. Int J Radiat Oncol Biol Phys. 2004;
not all, patients with neurofibromatosis type 2 and was associ- 60(1):225–30.
ated with a reduction in the volume of most growing acoustic 15. Kalogeridi MA, Georgolopoulou P, Kouloulias V, Kouvaris J,
neuromas [75]. The impact bevacizumab may have for spo- Pissakas G. Long-term results of LINAC-based stereotactic radio-
surgery for acoustic neuroma: the Greek experience. J Cancer Res
radic tumors remains unknown. The molecular mechanisms at
Ther. 2009;5(1):8–13.
play in sporadic tumors recently reviewed by Neff et. al., nota- 16. Hasegawa T, Fujitani S, Katsumata S, Kida Y, Yoshimoto M, Koike
bly Merlin’s interactions with other proteins, and regulation of J. Stereotactic radiosurgery for vestibular schwannomas: analysis
the NF2 gene [76] could spawn new targeted molecular thera- of 317 patients followed more than 5 years. Neurosurgery. 2005;
57(2):257–65; discussion 265.
pies which may complement existing focused radiation strate-
17. Murphy ES, Barnett GH, Vogelbaum MA, Neyman G, Stevens GH,
gies or supersede them altogether. Cohen BH, et al. Long-term outcomes of Gamma Knife radiosur-
gery in patients with vestibular schwannomas. J Neurosurg. 2011;
Acknowledgment We thank our colleagues from the Department of 114(2):432–40.
Radiation Oncology, Drs. Wenyin Shi and Haisong Liu, for their help- 18. Pollock BE, Driscoll CL, Foote RL, Link MJ, Gorman DA, Bauch
ful comments during the preparation of this manuscript. CD, et al. Patient outcomes after vestibular schwannoma manage-
ment: a prospective comparison of microsurgical resection and ste-
reotactic radiosurgery. Neurosurgery. 2006;59(1):77–85; discussion
77–85.
19. Flickinger JC, Lunsford LD, Linskey ME, Duma CM, Kondziolka
References D. Gamma knife radiosurgery for acoustic tumors: multivariate
analysis of four year results. Radiother Oncol. 1993;27(2):91–8.
1. Stangerup SE, Tos M, Caye-Thomasen P, Tos T, Klokker M, 20. Foote RL, Coffey RJ, Swanson JW, Harner SG, Beatty CW, Kline
Thomsen J. Increasing annual incidence of vestibular schwannoma RW, et al. Stereotactic radiosurgery using the gamma knife for
and age at diagnosis. J Laryngol Otol. 2004;118(8):622–7. acoustic neuromas. Int J Radiat Oncol Biol Phys. 1995;32(4):
2. Propp JM, McCarthy BJ, Davis FG, Preston-Martin S. Descriptive 1153–60.
epidemiology of vestibular schwannomas. Neuro Oncol. 2006;8(1): 21. Murphy ES, Suh JH. Radiotherapy for vestibular schwannomas: a
1–11. critical review. Int J Radiat Oncol Biol Phys. 2011;79(4):985–97.
22. Flickinger JC, Kondziolka D, Lunsford LD. Dose and diameter 40. Winston KR, Lutz W. Linear accelerator as a neurosurgical tool for
relationships for facial, trigeminal, and acoustic neuropathies stereotactic radiosurgery. Neurosurgery. 1988;22(3):454–64. Epub
following acoustic neuroma radiosurgery. Radiother Oncol. 1988/03/01.
1996;41(3):215–9. 41. Das IJ, Downes MB, Corn BW, Curran WJ, Werner-Wasik M,
23. Pollock BE, Link MJ, Foote RL. Failure rate of contemporary low- Andrews DW. Characteristics of a dedicated linear accelerator-
dose radiosurgical technique for vestibular schwannoma. J Neurosurg. based stereotactic radiosurgery-radiotherapy unit. Radiother Oncol.
2009;111(4):840–4. 1996;38(1):61–8. Epub 1996/01/01.
24. Gardner G, Robertson JH. Hearing preservation in unilateral acous- 42. Andrews DW, Bednarz G, Werner-Wasik M, Downes-Phillips
tic neuroma surgery. Ann Otol Rhinol Laryngol. 1988;97(1):55–66. B. Fractionated stereotactic radiotherapy: rationale, indications,
Epub 1988/01/01. and treatment technique. In: Chin L, Regine W, editors. Principles
25. Di Maio S, Akagami R. Prospective comparison of quality of life and practice of stereotactic radiosurgery. New York: Springer;
before and after observation, radiation, or surgery for vestibular 2008. p. 289–98.
schwannomas. J Neurosurg. 2009;111(4):855–62. 43. Shrieve DC, Hazard L, Boucher K, Jensen RL. Dose fractionation
26. Yang I, Aranda D, Han SJ, Chennupati S, Sughrue ME, Kane AJ, in stereotactic radiotherapy for parasellar meningiomas: radiobio-
et al. Hearing preservation after stereotactic radiosurgery for ves- logical considerations of efficacy and optic nerve tolerance.
tibular schwannoma: a systematic review. J Clin Neurosci. 2009; J Neurosurg. 2004;101 Suppl 3:390–5. Epub 2004/11/13.
16(6):742–4. 44. Leber KA, Bergloff J, Pendl G. Dose–response tolerance of the visual
27. Bhandare N, Jackson A, Eisbruch A, Pan CC, Flickinger JC, pathways and cranial nerves of the cavernous sinus to stereotactic
Antonelli P, et al. Radiation therapy and hearing loss. Int J Radiat radiosurgery. J Neurosurg. 1998;88(1):43–50. Epub 1998/01/07.
Oncol Biol Phys. 2010;76(3 Suppl):S50–7. 45. Goldsmith BJ, Rosenthal SA, Wara WM, Larson DA. Optic neuropa-
28. Roos DE, Potter AE, Brophy BP. Stereotactic radiosurgery for thy after irradiation of meningioma. Radiology. 1992; 185(1):71–6.
acoustic neuromas: what happens long term? Int J Radiat Oncol 46. Parsons JT, Bova FJ, Fitzgerald CR, Mendenhall WM, Million
Biol Phys. 2012;82(4):1352–5. RR. Radiation optic neuropathy after megavoltage external-beam
29. Hasegawa T, Kida Y, Kato T, Iizuka H, Yamamoto T. Factors asso- irradiation: analysis of time-dose factors. Int J Radiat Oncol Biol
ciated with hearing preservation after Gamma Knife surgery for Phys. 1994;30(4):755–63.
vestibular schwannomas in patients who retain serviceable hearing. 47. Andrews DW, Faroozan R, Yang BP, Hudes RS, Werner-Wasik M,
J Neurosurg. 2011;115(6):1078–86. Kim SM, et al. Fractionated stereotactic radiotherapy for the treat-
30. Andrews DW, Werner-Wasik M, Den RB, Paek SH, Downes- ment of optic nerve sheath meningiomas: preliminary observations
Phillips B, Willcox TO, et al. Toward dose optimization for frac- of 33 optic nerves in 30 patients with historical comparison to
tionated stereotactic radiotherapy for acoustic neuromas: observation with or without prior surgery. Neurosurgery.
comparison of two dose cohorts. Int J Radiat Oncol Biol Phys. 2002;51(4):890–902; discussion 903–4. Epub 2002/09/18.
2009;74(2):419–26. 48. Balagamwala EH, Chao ST, Suh JH. Principles of radiobiology
31. Kano H, Kondziolka D, Khan A, Flickinger JC, Lunsford of stereotactic radiosurgery and clinical applications in the central
LD. Predictors of hearing preservation after stereotactic radiosur- nervous system. Technol Cancer Res Treat. 2012;11(1):3–13. Epub
gery for acoustic neuroma. J Neurosurg. 2009;111(4):863–73. 2011/12/21.
32. Massager N, Nissim O, Delbrouck C, Delpierre I, Devriendt D, 49. Jones B, Dale RG, Deehan C, Hopkins KI, Morgan DA. The role of
Desmedt F, et al. Irradiation of cochlear structures during vestibular biologically effective dose (BED) in clinical oncology. Clin Oncol
schwannoma radiosurgery and associated hearing outcome. (R Coll Radiol). 2001;13(2):71–81. Epub 2001/05/26.
J Neurosurg. 2007;107(4):733–9. 50. Hansasuta A, Choi CY, Gibbs IC, Soltys SG, Tse VC, Lieberson
33. Paek SH, Chung HT, Jeong SS, Park CK, Kim CY, Kim JE, et al. RE, et al. Multisession stereotactic radiosurgery for vestibular
Hearing preservation after gamma knife stereotactic radiosurgery schwannomas: single-institution experience with 383 cases. Neuro-
of vestibular schwannoma. Cancer. 2005;104(3):580–90. surgery. 2011;69(6):1200–9. Epub 2011/05/12.
34. Brown M, Ruckenstein M, Bigelow D, Judy K, Wilson V, Alonso- 51. Jones B, Dale RG. Mathematical models of tumour and normal tis-
Basanta M, et al. Predictors of hearing loss after gamma knife sue response. Acta Oncol. 1999;38(7):883–93. Epub 1999/12/22.
radiosurgery for vestibular schwannomas: age, cochlear dose, and 52. Wang JZ, Huang Z, Lo SS, Yuh WT, Mayr NA. A generalized
tumor coverage. Neurosurgery. 2011;69(3):605–13; discussion linear-quadratic model for radiosurgery, stereotactic body radiation
613–4. therapy, and high-dose rate brachytherapy. Sci Transl Med. 2010;
35. Linskey ME, Johnstone PA, O’Leary M, Goetsch S. Radiation expo- 2(39):39ra48. Epub 2010/07/09.
sure of normal temporal bone structures during stereotactically guided 53. Park C, Papiez L, Zhang S, Story M, Timmerman RD. Universal
gamma knife surgery for vestibular schwannomas. J Neurosurg. survival curve and single fraction equivalent dose: useful tools in
2003;98(4):800–6. understanding potency of ablative radiotherapy. Int J Radiat Oncol
36. Tamura M, Carron R, Yomo S, Arkha Y, Muraciolle X, Porcheron Biol Phys. 2008;70(3):847–52. Epub 2008/02/12.
D, et al. Hearing preservation after gamma knife radiosurgery for 54. Pan CC, Eisbruch A, Lee JS, Snorrason RM, Ten Haken RK, Kileny
vestibular schwannomas presenting with high-level hearing. PR. Prospective study of inner ear radiation dose and hearing loss
Neurosurgery. 2009;64(2):289–96; discussion 296. in head-and-neck cancer patients. Int J Radiat Oncol Biol Phys.
37. Andrews DW, Bednarz G, Evans JJ, Downes B. A review of 3 cur- 2005;61(5):1393–402. Epub 2005/04/09.
rent radiosurgery systems. Surg Neurol. 2006;66(6):559–64. 55. Rasmussen R, Claesson M, Stangerup SE, Roed H, Christensen IJ,
38. Likhterov I, Allbright RM, Selesnick SH. LINAC radiosurgery and Caye-Thomasen P, et al. Fractionated stereotactic radiotherapy of
radiotherapy treatment of acoustic neuromas. Otolaryngol Clin vestibular schwannomas accelerates hearing loss. Int J Radiat
North Am. 2007;40(3):541–70, ix. Oncol Biol Phys. 2012;83(5):e607–11. Epub 2012/05/09.
39. Hsu PW, Chang CN, Lee ST, Huang YC, Chen HC, Wang CC, et al. 56. van de Langenberg R, Hanssens PE, Verheul JB, van Overbeeke JJ,
Outcomes of 75 patients over 12 years treated for acoustic neuro- Nelemans PJ, Dohmen AJ, et al. Management of large vestibular
mas with linear accelerator-based radiosurgery. J Clin Neurosci. schwannoma. Part II Primary Gamma Knife surgery: radiological and
2010;17(5):556–60. clinical aspects. J Neurosurg. 2011;115(5):885–93. Epub 2011/08/16.
57. Milligan BD, Pollock BE, Foote RL, Link MJ. Long-term tumor 67. Meijer OW, Vandertop WP, Baayen JC, Slotman BJ. Single-fraction
control and cranial nerve outcomes following gamma knife surgery vs fractionated linac-based stereotactic radiosurgery for vestibular
for larger-volume vestibular schwannomas. J Neurosurg. 2012; schwannoma: a single-institution study. Int J Radiat Oncol Biol
116(3):598–604. Epub 2011/12/20. Phys. 2003;56(5):1390–6.
58. Mandl ES, Meijer OW, Slotman BJ, Vandertop WP, Peerdeman 68. Combs SE, Burkholder I, Edler L, Rieken S, Habermehl D, Jakel O,
SM. Stereotactic radiation therapy for large vestibular schwanno- et al. Randomised phase I/II study to evaluate carbon ion radio-
mas. Radiother Oncol. 2010;95(1):94–8. Epub 2010/02/09. therapy versus fractionated stereotactic radiotherapy in patients
59. Foote KD, Friedman WA, Buatti JM, Meeks SL, Bova FJ, Kubilis with recurrent or progressive gliomas: the CINDERELLA trial.
PS. Analysis of risk factors associated with radiosurgery for BMC Cancer. 2010;10:533.
vestibular schwannoma. J Neurosurg. 2001;95(3):440–9. Epub 69. Chen WC, Jackson A, Budnick AS, Pfister DG, Kraus DH, Hunt
2001/09/22. MA, et al. Sensorineural hearing loss in combined modality treat-
60. Andrews DW, Suarez O, Goldman HW, Downes MB, Bednarz G, ment of nasopharyngeal carcinoma. Cancer. 2006;106(4):820–9.
Corn BW, et al. Stereotactic radiosurgery and fractionated stereo- 70. Gevaert T, Levivier M, Lacornerie T, Verellen D, Engels B,
tactic radiotherapy for the treatment of acoustic schwannomas: Reynaert N, et al. Dosimetric comparison of different treatment
comparative observations of 125 patients treated at one institution. modalities for stereotactic radiosurgery of arteriovenous malforma-
Int J Radiat Oncol Biol Phys. 2001;50(5):1265–78. tions and acoustic neuromas. Radiother Oncol. 2013;106:192–7.
61. Chan AW, Black P, Ojemann RG, Barker FG, II, Kooy HM, Lopes 71. Otte J, Schunknecht HF, Kerr AG. Ganglion cell populations in
VV, et al. Stereotactic radiotherapy for vestibular schwannomas: normal and pathological human cochleae. Implications for cochlear
favorable outcome with minimal toxicity. Neurosurgery. 2005; implantation. Laryngoscope. 1978;88(8 Pt 1):1231–46. Epub
57(1):60–70; discussion 60–70. 1978/08/01.
62. Combs SE, Volk S, Schulz-Ertner D, Huber PE, Thilmann C, Debus 72. Pauler M, Schuknecht HF, Thornton AR. Correlative studies of
J. Management of acoustic neuromas with fractionated stereotactic cochlear neuronal loss with speech discrimination and pure-
radiotherapy (FSRT): long-term results in 106 patients treated in a tone thresholds. Arch Otorhinolaryngol. 1986;243(3):200–6. Epub
single institution. Int J Radiat Oncol Biol Phys. 2005;63(1):75–81. 1986/01/01.
63. Koh ES, Millar BA, Menard C, Michaels H, Heydarian M, Ladak S, 73. Schuknecht HF. Auditory and cytocochlear correlates of inner ear
et al. Fractionated stereotactic radiotherapy for acoustic neuroma: disorders. Otolaryngol Head Neck Surg. 1994;110(6):530–8. Epub
single-institution experience at The Princess Margaret Hospital. 1994/06/01.
Cancer. 2007;109(6):1203–10. 74. Roosli C, Linthicum Jr FH, Cureoglu S, Merchant SN. Dysfunction
64. Maire JP, Huchet A, Milbeo Y, Darrouzet V, Causse N, Celerier D, of the cochlea contributing to hearing loss in acoustic neuromas: an
et al. Twenty years’ experience in the treatment of acoustic neuro- underappreciated entity. Otol Neurotol. 2012;33(3):473–80. Epub
mas with fractionated radiotherapy: a review of 45 cases. Int J Radiat 2012/03/02.
Oncol Biol Phys. 2006;66(1):170–8. 75. Plotkin SR, Stemmer-Rachamimov AO, Barker II FG, Halpin C,
65. Thomas C, Di Maio S, Ma R, Vollans E, Chu C, Clark B, et al. Padera TP, Tyrrell A, et al. Hearing improvement after bevacizumab
Hearing preservation following fractionated stereotactic radiother- in patients with neurofibromatosis type 2. N Engl J Med. 2009;
apy for vestibular schwannomas: prognostic implications of 361(4):358–67. Epub 2009/07/10.
cochlear dose. J Neurosurg. 2007;107(5):917–26. 76. Neff BA, Welling DB, Akhmametyeva E, Chang LS. The molecular
66. Arthurs BJ, Fairbanks RK, Demakas JJ, Lamoreaux WT, Giddings biology of vestibular schwannomas: dissecting the pathogenic pro-
NA, Mackay AR, et al. A review of treatment modalities for vestibular cess at the molecular level. Otol Neurotol. 2006;27(2):197–208.
schwannoma. Neurosurg Rev. 2011;34(3):265–77; discussion 277–9. Epub 2006/01/27.
for Pituitary Adenomas 26
Jason P. Sheehan and Brian Williams
and directed toward neighborhood symptoms alone, and in

Introduction the case of an advancing struma are in no sense curative, it
is possible that radiotherapy should often be employed as a
Pituitary adenomas are common tumors found in approxi- supplementary measure” [3]. Later, Cushing and his col-
mately 10–27 % of the general population [1, 2]. leagues used a device called a radium bomb to deliver
Microadenomas which are less than 1 cm in maximum radiation therapy to intracranial tumors [4, 5]. Since that
dimension are typically diagnosed as a result of hormone time, neurological surgeons and radiation oncologists, in
overproduction. Macroadenomas which are pituitary adeno- conjunction with medical physicists, have utilized ionizing
mas greater than 1 cm in dimension may be discovered as a radiation to treat patients with recurrent or residual pitu-
result of direct mass effect leading to hypopituitarism, eleva- itary adenomas.
tion in prolactin output, or a focal neurological deficit (e.g., In the modern era, most centers have turned to stereotac-
cranial nerve dysfunction). Increasingly with the availability tic radiosurgery (SRS) for treating patients with a recurrent
of CT and MRI, pituitary adenomas are also being detected or residual pituitary adenoma. Tremendous effort in the field
incidentally. Microadenomas are divided fairly evenly of SRS has been placed on the preservation of surrounding
between functioning and nonfunctioning lesions, while non- neuronal, vascular, and hormonal structures in an effort to
functioning lesions comprise approximately 80 % of all improve the therapeutic ratio. Technical refinements for
macroadenomas [1]. Pituitary adenoma patients typically treating pituitary adenoma patients have been achieved
exhibit symptoms of headache (40–60 %), visual distur- through advances in radiobiology, neuroimaging, medical
bance, hypopituitarism, or rarely apoplexy [1, 2]. physics, and biomedical engineering. We review the contem-
Pituitary adenomas remain difficult to cure with micro- porary role of SRS for pituitary adenomas.
surgical techniques alone, and they often require multimo-
dality treatment, which includes surgery, radiation therapy,
radiosurgery, and medical management. In his seminal Stereotactic Radiosurgical Concept
work The Pituitary Body and Its Disorders, Harvey and Techniques
Cushing recognized the limits of microsurgery and the
utility of ionizing radiation for the treatment of patients In 1951, SRS was described by Lars Leksell [6] as the
with pituitary adenomas. He wrote “in view of the fact that “closed skull destruction of an intracranial target using ion-
the surgical measures at our disposal are merely palliative izing radiation.” In 1968, Leksell treated the first pituitary
adenoma patient with the Gamma Knife. Subsequently, SRS
has been utilized to treat thousands of patients with pituitary
J.P. Sheehan, M.D., Ph.D. (*) adenomas.
Department of Neurological Surgery, Health Sciences Center, SRS delivers a highly focused, large dose of radiation
University of Virginia, Box 800212, Charlottesville, to the target, while sparing surrounding structures from an
VA 22908, USA appreciable dose of radiation. Stereotactic image guid-
e-mail: jps2f@virginia.edu; jps2f@hscmail.mcc.virginia.edu
ance is utilized to achieve this. Radiosurgery is tradition-
B. Williams, M.D. ally delivered in a single session but may be delivered in
Department of Neurosurgery, Health Sciences Center, University
of Virginia, Box 800212, Charlottesville, VA 22908, USA up to five sessions (i.e., fractions) [7]. It is characterized
e-mail: BJW6M@hscmail.mcc.virginia.edu by a steep dose falloff to the surrounding normal tissues.
370 J.P. Sheehan and B. Williams
For cobalt-based SRS devices, the steepest gradient index

(i.e., falloff) is achieved around a 50 % isodose line, SRS for Pituitary Adenomas
whereas for linear accelerator (LINAC)-based systems, it
is usually at the 80–90 % isodose line. Patients are immo- SRS and Nonfunctioning Pituitary Adenomas
bilized using rigid frames fixed to the skull or other immo-
bilization devices (e.g., aquaplast masks or bite blocks); After a microsurgical resection most commonly through a
each immobilization device has its own stereotactic coor- transsphenoidal approach, control of a pituitary tumor will
dinate system. Radiosurgery reliably achieves submilli- be achieved in 50–80 % of adenomas [10]. SRS is an excel-
meter accuracy in intracranial space. Onboard imaging lent treatment approach for patients who have progression or
systems (OBI) may be used to further track and compen- recurrence. It may also be utilized in cases of residual tumors
sate for sources of error (e.g., setup error, patient move- that are known to be more aggressive (e.g., silent adrenocor-
ment, etc.). ticotropic hormone (ACTH) or TSH adenomas). Based upon
There are several types of radiosurgical delivery devices, the published literature, radiosurgery affords the vast major-
including the Gamma Knife® (Elekta AB), modified LINAC’s ity of pituitary adenoma patients with effective, long-term
such as the CyberKnife (Accuray), Novalis (Brainlab), tumor control as well as a low rate of complications [10].
TrueBeam STX (Varian), and Axesse (Elekta), or proton In Table 26.1, we list the major radiosurgical series since
beam units (e.g., IBA or Mevion). Single session radiosurgi- 2002 that detail outcomes in nonfunctioning adenoma
cal margin doses for nonfunctioning adenomas range from patients [10–36]. Single session radiosurgery margin doses
12 to 18 and 15 to 30 Gy for functioning adenomas. For of 12–18 Gy are generally used for nonfunctioning adeno-
multi-session radiosurgery, these doses may be divided over mas. Tumor control rates with radiosurgery vary from 83 to
two to five fractions. 100 % (average 95.5 %) (Table 26.1). Neurological deficits
Gamma Knife radiosurgery involves the use of multiple following radiosurgery are uncommon (average 2.7 %, range
isocenters to achieve a highly conformal dose plan. The 0–7.1 %). The most common complication is hypopituita-
number of isocenters or “shots” varies based upon the size, rism (average 13 %, range 0–39 %) which is a manageable
shape, and location of the pituitary adenoma. With the cur- problem provided that it is detected and appropriate hor-
rent version of the Gamma Knife Perfexion™, each isocen- monal replacement therapy is initiated. At the University of
ter is comprised of eight independent sectors of 24 beams, Virginia, we reported 90 % tumor control in a series of 140
and each isocenter contains up to 192 simultaneous beams. patients with nonfunctioning pituitary adenomas [36]. Tumor
Beam diameters for the current Gamma Knife Perfexion unit control was statistically less probable when the tumor vol-
range from 0 (i.e., blocked) to 16 mm. Other cobalt-based ume treated was greater than 5 cm3 [36]. In a separate study,
radiosurgical devices are also in use and include the Infini™ we examined the outcomes of a smaller group of patients
system (MASEP). with a minimum follow-up of 4 years after radiosurgery. The
LINAC-based radiosurgery (e.g., CyberKnife®, overall progression-free survival was 83 %, and three-
TrueBeam™ STx, Trilogy, Tomotherapy™, and Axesse™) quarters of the patients demonstrated a marked decrease in
uses multiple radiation arcs to cross-fire photon beams at a the adenoma volume by at least 20 % or more [37]. This
target volume [8]. Most systems use dynamic techniques in same study demonstrated that tumor control was related to
which the arc is rotated around its radius to deliver radiation treatment volume and that those patients with a treatment
that enters many different vantage points. Technical volume of less than 5 cm3 were far more likely to have tumor
refinements with LINAC-based radiosurgery include beam control after radiosurgery. In a recent analysis by the North
shaping, intensity modulation, multileaf collimation, and American Gamma Knife consortium, younger patient age,
onboard CT or fluoroscopic imaging. smaller tumor volume, and no prior radiation therapy were
Proton therapy has been adapted as a radiosurgical tool significantly related to the favorable outcomes of tumor con-
for intracranial pathology. It takes advantage of the inher- trol and neurological stability after radiosurgery for a non-
ently superior dose distribution of protons compared with functioning pituitary adenoma. These findings underscore
that of photons because of the Bragg peak phenomenon [9]. the value of a maximum safe resection prior to SRS. Please
Currently, there are a few centers using proton beam technol- see Table 26.1 for a detailed summary.
ogy to perform radiosurgery (one session treatment); many
proton centers perform fractionated stereotactic radiotherapy
(FSRT). The number of proton beam centers has increased SRS and Cushing’s Disease
significantly in the past 10 years as the technology becomes
more cost-effective particularly with the advent of compact In the vast majority of cases, endogenous Cushing’s syn-
proton beam units such as the one by Mevion Medical drome results from oversecretion of ACTH from a pituitary
Systems. adenoma [38]. Although microsurgical resection is the mainstay
26 Stereotactic Radiosurgery for Pituitary Adenomas 371
Table 26.1 Summary of the literature review for the radiosurgical management of nonfunctioning pituitary adenomas
Mean/median Mean/median Radiologic control Neurological Delayed
Year Authors follow-up (months) margin dose (Gy) of tumor (%) deficit (%) hypopituitarism (%)
2002 Feigl et al. 55.2 15 94 NR 40
2002 Sheehan et al. 31.2 16 97.6 4.8 0
2002 Wowra and Stummer 57.7 16 93.3 0 10
2003 Petrovich et al. 34 15 100 3 NR
2004 Losa et al. 41.1 16.6 96.3 0 9.3
2004 Muacevic et al. 21.7 16.5 95 0 3.9
2005 Kajiwara et al. 32.1 12.6 92.9 7.1 7.1
2005 Picozzi et al. 40.6 16.5 96.1 NR NR
2005 Iwai et al. 36.4 12.3 93 0 7
2006 Mingione et al. 46.4 18.5 92.2 0 25
2006 Voges et al. 56.6 13.4 100 4.2 12.3
2007 Liscak et al. 60 20 100 0 1.4
2008 Pollock et al. 64 16 96.8 1.6 27
2009 Hoybye and Rahn 78 20 100 4.3 0
2009 Kobayashi 50.2 NR 96.7 2.8 8.2
2010 Castro et al. 42 12.5 100 0 0
2010 Hayashi et al. 36 18.2 100 0 0
2011 Gopalan et al. 95 18.4 83 6.3 39
2011 Iwata et al. 33 21 Gy/3 Fr 98 1 2
(CyberKnife) 25 Gy/5 Fr
2011 Park et al. 62 13 90 2.4 24
2012 Starke et al. 50.4 18 90 13.7 30.3
Average 48.7 16 95.5 2.7 13
NR not reported, FR fraction
of treatment for Cushing’s disease, many adenomas show may be due in part to the higher margin doses delivered to
invasion of the surrounding dura and/or cavernous sinus Cushing’s patients and the resulting higher doses to collat-
or are difficult to delineate on MRI, thereby making a eral structures, such as the normal pituitary gland and stalk.
surgical cure unlikely. Radiosurgery proves a valuable Long-term radiologic and endocrine follow-up is crucial
treatment option for patients with persistent Cushing’s because late recurrences have been seen in several Cushing’s
disease after a resection. disease series [46, 58]. Please see Table 26.2 for a detailed
In Table 26.2, the major radiosurgical series for Cushing’s summary.
patients since 2000 are listed (Table 26.2) [11–15, 18, 21, 23,
24, 27, 31, 33, 39–62]. In these studies, most authors utilized
the 24 h urinary free cortisol (UFC) or serum cortisol to SRS and Acromegaly
define an endocrine remission. In addition, radiosurgical
margin doses of 18 Gy to as high as 30 Gy were delivered to Acromegaly has an approximate prevalence of 60 per million
the adenomas of Cushing’s disease patients. people [63]. Poorly controlled acromegaly is associated with
The remission rates vary tremendously from 16.7 to significant morbidity (e.g., hypertension, diabetes, cardio-
87 %, but the majority of reports note a remission rate of myopathy, and sleep apnea) and mortality [63]. Given the
greater than 50 %. In particular, the endocrine remission potential for cure, surgical resection remains the initial treat-
rate after radiosurgery does not match the excellent rate of ment for most acromegalic patients. Nevertheless, many
radiologic control of the adenoma. Based on our institu- patients will not achieve a normalization of IGF-1 and
tional experience, endocrine remission of Cushing’s disease growth hormone after resection. Some of these patients
is typically achieved 12 months after SRS [46]. In most exhibit adenoma invasion of surrounding structures (e.g., the
series, after radiosurgery the rates of newly developed or dura or the cavernous sinus). Also, many patients with acro-
worsened cranial neuropathies, including visual deteriora- megaly have macroadenomas; complete resection is not
tion, are low (average 3.4 %; Table 26.2). The risk of delayed always feasible with the large, infiltrative lesions.
hypopituitarism after radiosurgery for Cushing’s disease Table 26.3 details the major radiosurgical series for acrome-
appears to be slightly higher (average 24.9 %; Table 26.2) galic patients published since 2000 (Table 26.3) [11–15, 18, 21,
compared to that of nonfunctioning adenoma series; this 24, 27, 31, 33, 40, 41, 44, 47, 48, 51, 52, 54, 58–61, 64–79].
Table 26.2 Summary of the literature review for the radiosurgical management of Cushing’s syndrome
Mean/median Mean/median Biochemical Neurological
Year Author margin dose (Gy) follow-up (months) remission (%) deficit (%) Hypopituitarism (%)
2000 Izawa et al. 23.8 26.4 16.7 0 0
2000 Sheehan et al. 20 39.1 63 2.3 16
2000 Shin et al. 32.3 88.2 50 6.3 16.7
2001 Hoybye et al. NR 16.8 44 0 68.8
2002 Feigl et al. 15 55.2 60 NR 40
2002 Kobayashi et al. 28.7 64 23.3 NR NR
2002 Laws et al. 20 NR 74 2.5 24
2002 Pollock et al. 20 42.4 78 22.2 16
2003 Choi et al. 28.5 42.5 55.6 0 0
2003 Petrovich et al. 15 34 NR 3 NR
2003 Wong et al. NR 38 100 0 20
2003 Witt 24 24 0 0 NR
2005 Devin et al. 14.7 42 49 0 40
2006 Voges et al. 16.4 58.7 52.9 4.2 12.3
2007 Castinetti et al. 29.5 54.7 42.5 5 15
2007 Jagannathan et al. 23 45 54 5 22
2007 Kajiwara et al. 26 38.5 50 0 50
2007 Petit et al. 20 62 52 0 52
2008 Pollock et al. 20 73 87 0 36
2008 Tinnel et al. 25 37 50 0 50
2009 Castinetti et al. 28 94 50 5.3 21
2009 Wan et al. 23 67.3 27.9 2.9 1.7
2009 Kobayashi 28.7 64.1 35 NR NR
2010 Hayashi 25.2 36 38 15.4 0
2011 Sheehan et al. 24 31 54 NR 22
Average 23 48.9 49.9 3.4 23.6
NR not reported
Similar margin doses of 18–30 Gy are routinely delivered to SRS and Prolactinomas
the adenoma as part of single session radiosurgery. Endocrine
remission after radiosurgery is achieved in an average of Prolactinomas represent one of the more common types of
60.7 % of acromegalic patients (range 0–82 %). Neurological secretory pituitary adenomas. However, unlike patients with
deficits and hypopituitarism occur after radiosurgery in acromegaly or Cushing’s disease, prolactinomas are usually
1.8 % (range 0–11 %) and 14.7 % (range 0–40 %) of patients, managed medically. For those patients with a medically
respectively. Temporary cessation of pituitary suppressive refractory prolactinoma or those suffering from side effects
medications near the time of radiosurgery appears to increase of the medications, radiosurgery may be an option. Since
the chance of endocrine remission in acromegalic patients most patients are successfully managed with medical ther-
[80]. Also, patients who had a lower adenoma volume at the apy, patients that undergo radiosurgery likely represent a
time of radiosurgery were more likely to achieve endocrine more aggressive tumor biology.
remission after radiosurgery [80]. Thus, for those acrome- The tumor control rate after radiosurgery for a prolacti-
galic patients with a macroadenoma, maximal safe resection noma is typically is over 90 %. However, the endocrine remis-
should be attempted prior to radiosurgery. sion rates vary after radiosurgery [12, 15, 27, 40, 47, 54, 55,
At our center, the mean time to endocrine remission 58–60, 74, 81–90]. Biochemical remission of antisecretory
after radiosurgery for acromegalic patients was 24 months medications after radiosurgery ranges from 0 to 83 %. The
which is twice as long as the same milestone for Cushing’s chance of biochemical remission of a prolactinoma after
disease patients. This and other findings suggest a dif- radiosurgery tends to be lower than for those with Cushing’s
ferential response of various subtypes of functioning disease and acromegaly even when taking into account
adenomas to SRS [54]. Please see Table 26.3 for a detailed pre-radiosurgical attributes such as tumor volume [54]. Some
summary. of this variation in endocrine remission rates emanate from
Table 26.3 Summary of the literature review for the radiosurgical management of acromegaly
Mean/median Mean/median margin Biochemical Neurological
Year Authors follow-up (months) dose (Gy) remission (%) deficit (%) Hypopituitarism (%)
2000 Izawa et al. 26.4 23.8 41.4 0 0
2000 Shin et al. 42.7 34.4 66.7 6.3 0
2000 Zhang et al. 34 31.3 36.8 2.9 0
2001 Fukuoka et al. 42 20 50 0 0
2001 Ikeda et al. 55.8 25 82 0 0
2002 Feigl et al. 55.2 15 60 NR 40
2002 Pollock et al. 42.4 20 42 0 16
2003 Attanasio et al. 46 20 23 0 6.6
2003 Choi et al. 42.5 28.5 50 0 0
2003 Muramatsu et al. 30 27.5 50 0 0
2003 Petrovich et al. 34 15 NR 3 NR
2003 Witt 24 24 25 0 NR
2005 Castinetti et al. 49.5 25 17 0 17.1
2005 Gutt et al. 22.8 18 47.7 0 NR
2005 Kajiwara et al. 53.5 13.5 0 0 0
2005 Koybayashi et al. 63.3 18.9 4.8 11.1 14.6
2006 Jezkova et al. 53.7 35 50 0 26
2006 Voges et al. 54.3 16.5 37.5 4.2 12.3
2007 Pollock et al. 63 20 50 2.2 33
2007 Roberts et al. 25.4 21 44.4 0 33.3
2007 Vik-Mo et al. 66 26.5 17 3.3 13.1
2008 Jagannathan et al. 57 22 53 4.2 34
2008 Losa et al. 69 21.5 60.2 0 8.5
2008 Pollock et al. 46.9 20 67 0 36
2008 Tinnel et al. 35 25 44.4 11 22
2009 Castinetti et al. 102 24 42 5.3 21
2009 Wan et al. (MASEP 67.3 21.4 36.9 1 1.7
GKS)
2009 Ronchi et al. 120 20 46.0 0 40
2010 Iwai et al. 84 20 38.0 0 8
2010 Hayashi et al. 36 25.2 40.0 0 0
2010 Poon et al. 73.8 20–35 75 0 11.4 % (GKS1);
27.3 % (repeat GKS)
2011 Sheehan et al. 31 24 53 NR 34
2012 Franzin et al. 71 22.5 60.7 0 7.8
Average 52.1 22.6 44.1 1.8 14.7
NR not reported, GKS Gamma Knife surgery, GKS1 first gamma knife procedure
selection biases at radiosurgical centers. As with acromegalic patients are typically bridged with suppressive medications.
patients, the chance of endocrine remission appears to be They are periodically taken off of the antisecretory medica-
enhanced when radiosurgery is delivered during a period of tion, and endocrine function is checked. Antisecretory medi-
temporary cessation of antisecretory medications [81, 84]. cations can be halted at the time when a post-radiosurgery
Please see Table 26.4 for a detailed summary. endocrine normalization has been attained. The time interval
in which remission may occur varies from 3 months to 8
years [31, 57, 75]. The majority of radiosurgical reports indi-
SRS-Induced Endocrine Remission and Late cated endocrine remission in Cushing’s disease and acrome-
Recurrence galic patients within 1–3 years.
Several investigators have identified factors that increase the
SRS yields hormone normalization within a period of time probability of endocrine remission. Castinetti and colleagues
that is longer than that achieved after surgical extirpation but [65] contended that preoperative GH and IGF-1 levels are
shorter than with radiation therapy [91]. After radiosurgery, significantly associated with the rate of post-SRS remission.
Table 26.4 Summary of the literature for the radiosurgical management of prolactinomas
Mean/median Margin Biochemical Neurological
Year Authors follow-up (months) dose (Gy) remission (%) deficit (%) Hypopituitarism (%)
2000 Izawa et al. 28 22 20 0 NR
2000 Landolt et al. 29 25 25 0 NR
2000 Pan et al. 33 32 41 0 NR
2002 Feigl et al. 55 15 NR NR NR
2002 Pollock et al. 42 20 29 14 16
2003 Choi et al. 42.5 28.5 24 0 0
2003 Muramatsu et al. 30 15 0 7.7 0
2003 Petrovich et al. 41 15 83 0 NR
2005 Kajiwara et al. 35.3 17.5 33 4.7 9.5
2006 Pouratian et al. 55 18.6 26 7 28
2006 Voges et al. 56 20 15.4 4.2 18.3
2007 Ma et al. 37 26.1 40 NR 17.6
2008 Pollock et al. 48 30 18 2 45
2008 Tinnel et al. 19.5 30 50 11 22
2009 Castinetti et al. 85.5 30 46.6 5.3 21
2009 Jezkova et al. 75.5 49 37.1 0 14.3
2009 Wan et al. 67.3 35 23.3 1.7 1.7
2009 Kobayashi et al. 37.4 18.4 43.5 in 23 pts 0 0
2010 Tanaka et al. 60 25 17 8 42
2011 Sun et al. 48 23 0 0 0
2012 Liu et al. 36 15 27.3 0 4.5
Average 45.8 24.3 29.2 3.5 15
NR not reported, pts patients
Pollock and colleagues [75] evaluated 46 acromegalic The effect of pituitary suppressive medications on
patients and identified two significant associations. A pre- radiosurgical outcome remains controversial. Reports have
radiosurgical IGF-1 level greater than 2.25 times the upper not been entirely consistent regarding the importance of a
limit of normal range was significantly associated with a lower temporary cessation of suppressive medications at the time
rate of endocrine normalization (HR 2.9, 95 % CI 1.2–6.9). of radiosurgery [55, 64, 65, 75, 76, 81, 92]. Two groups ana-
For patients who had IGF-1 levels less than 2.25 times the lyzed remission rates after radiosurgery among patients on
upper limit of normal and were not taking somatostatin ago- somatostatin agonists, and they failed to identify an associa-
nists at radiosurgery, the rates of biochemical remission tion between the use of somatostatin agonists and endocrine
exceeded 80 %. A similar finding was noted by Landolt and remission [64, 65]. However, the definition of endocrine
colleagues [92], who demonstrated that the post-SRS remis- remission has not been well defined across series [52, 66].
sion rate was lowered from 60 to 11 % for patients taking Second, the length of follow-up in retrospective series varies
octreotide during the radiosurgical period. Somatostatin ago- widely. Pollock et al. [75] demonstrated that remission con-
nists may lower the radiosensitivity of adenoma cells because tinued to occur for up to 5 years following radiosurgery,
of decreased cell division. Also, somatostatin agonists, such whereas in the report by Castinetti and colleagues [65], 44 %
as octreotide, can act as free radical scavengers and reduce of their patient population received their final endocrine
the DNA damage following ionizing radiation. evaluation less than 3 years after radiosurgery. At the
This counterproductive effect of antisecretory medica- University of Virginia, we advocate the cessation of suppres-
tions on radiosurgical outcomes is not exclusive to acromeg- sive medication for 6–8 weeks around the radiosurgery.
aly. Landolt and colleagues [92] found a trend toward worse There does appear to be a differential radiosensitivity
outcomes in prolactinoma patients on dopamine agonist. between specific types of functioning adenomas [54, 59, 86].
Pouratian et al. [84] analyzed 23 patients with refractory pro- Cushing’s disease demonstrates the highest rates of bio-
lactinomas and noted a significant increase in the rates of chemical remission, followed by acromegaly, prolactinomas,
remission in patients who were not taking dopamine agonists and Nelson’s syndrome. This difference in response after
at the time of radiosurgery. We have also observed a similar radiosurgery has been attributed to patient selection, tumor
improvement in endocrine remission for acromegalic patients volume, radiation dose, use of suppressive medications,
who were taken off of pituitary suppressive medications at and duration of follow-up [54, 86]. Pollock et al. [54]
the time of their radiosurgery [68]. reviewed a retrospective series of 46 pituitary adenoma patients.
This case-controlled study demonstrated variations in a hypersecretory state is a more significant threat to most
endocrine remission after SRS for various types of secretory patients than delayed hypopituitarism. Hypopituitarism can
adenomas. The study and others suggest a differential radio- usually be managed with hormonal replacement.
sensitivity for adenoma types, but the underlying cause for The second most common toxicity is a cranial neuropa-
this variable response remains unclear. thy. Cranial nerves II, III, IV, V, and VI are located in the
A few cases of recurrence following radiosurgical- parasellar or suprasellar regions and are at risk for injury
induced remission have been reported [46, 76]. However, after radiosurgery. In the absence of tumor growth, such neu-
recurrence rates of up to 20 % have been reported. This ropathies after radiosurgery occur in very few patients.
underscores the importance of long-term radiographic and Improved conformality, steeper dose gradients, and adequate
endocrine follow-up for patients with secretary pituitary shielding help to minimize this risk [91]. Other rare toxici-
adenomas. ties include radiation necrosis of the adjacent parenchyma
[54, 59, 60, 76], internal carotid artery narrowing [76, 94],
and radiation-induced secondary malignancy [95]. No cases
Adverse Events from Radiosurgery of radiation-induced secondary malignancies have been
reported to date after SRS for pituitary tumors. Based upon
Adverse events following radiosurgery for a pituitary ade- the available literature, the risk of serious and permanent
noma occur in only a minority of patients. Hypopituitarism complications after radiosurgery for a pituitary adenoma is
is the most frequently occurring side effect from radiosur- quite low.
gery. Most radiosurgical series demonstrate that 30 % of
patients will eventually develop some form of pituitary defi-
ciency after radiosurgery. Post-radiosurgical dysfunction if it Conclusions
occurs generally involves the anterior pituitary axis but can
involve the posterior pituitary axis or very rarely lead to pan- SRS plays a significant role in the contemporary manage-
hypopituitarism. Hypopituitarism has been correlated to the ment of pituitary adenoma patients. SRS is typically per-
radiosurgical treatment volume, with those patients having a formed in patients with substantial residual tumor or
tumor volume ≤ 4.0 cm3 exhibiting an 18 % 5-year risk of recurrence after resection. It is also used for patients with
hypopituitarism versus 58 % for those with larger lesions functioning adenomas that fail to achieve endocrine remis-
[54]. The risk of hypopituitarism after radiosurgery is likely sion after a prior resection. Even when the adenoma resides
to be related to the pre-radiosurgical status of the normal in the cavernous sinus, neurological function after SRS is
pituitary gland, the type and timing of prior treatments, the usually preserved. Delayed hypopituitarism is the most com-
radiosurgical dose delivered to the normal gland, the dose mon complication but still occurs in a minority of patients.
delivered to the pituitary stalk, and the rigorousness and Permanent and uncorrectable complications after radiosur-
length of the endocrine follow-up assessment period. An gery are rare. Long-term follow-up for pituitary adenoma
absolutely safe radiosurgical dose or dose per volume so as patients undergoing radiosurgery is recommended.
to avoid hypopituitarism probably does not exist. Some have Case 26.1: A patient had a nonsecretory adenoma involv-
advocated surgical placement of a spacer between the resid- ing the right side of the sella and cavernous sinus (Fig. 26.1a).
ual adenoma and pituitary gland if postoperative radiosur- The patient underwent Gamma Knife radiosurgery in 2005.
gery is contemplated [93]. Delivery of an optimal dose to the The adenoma was treated with 16 Gy to the periphery
adenoma should not be compromised with the hope of avoid- (Fig. 26.1b). The patient had marked regression of his tumor
ing hypopituitarism. Adenoma progression or persistence of as shown on this MRI from 2013.
Fig. 26.1 A 57-year-old male presented with a recurrent, nonfunction- size and demonstrates diminished contrast enhancement. (c) Six and
ing pituitary adenoma. He had two prior transsphenoidal resections. (a) one-half years after radiosurgery, the patient’s adenoma continues to
The patient underwent SRS in which 15 Gy was delivered to the tumor regress. All images are post-contrast, coronal MRIs
margin. Four years later (b), the pituitary adenoma had decreased in
23. Mitsumori M, Shrieve DC, Alexander 3rd E, Kaiser UB, Richardson

References GE, Black PM, et al. Initial clinical results of LINAC-based stereo-
tactic radiosurgery and stereotactic radiotherapy for pituitary ade-
nomas. Int J Radiat Oncol Biol Phys. 1998;42(3):573–80.
1. Dekkers OM, Pereira AM, Romijn JA. Treatment and follow-up of 24. Mokry M, Ramschak-Schwarzer S, Simbrunner J, Ganz JC, Pendl
clinically nonfunctioning pituitary macroadenomas. J Clin G. A six year experience with the postoperative radiosurgical man-
Endocrinol Metab. 2008;93(10):3717–26. agement of pituitary adenomas. Stereotact Funct Neurosurg.
2. Vance ML. Treatment of patients with a pituitary adenoma: one 1999;72 Suppl 1:88–100.
clinician’s experience. Neurosurg Focus. 2004;16(4):1–6. 25. Muacevic A, Uhl E, Wowra B. Gamma knife radiosurgery for non-
3. Cushing H. The pituitary body and its disorders. Philadelphia: functioning pituitary adenomas. Acta Neurochir. 2004;91:51–4.
Lippincott; 1912. 26. Muramatsu J, Yoshida M, Shioura H, Kawamura Y, Ito H, Takeuchi
4. Seymour ZA, Cohen-Gadol AA. Cushing’s lost cases of “radium H, et al. [Clinical results of LINAC-based stereotactic radiosurgery
bomb” brachytherapy for gliomas. J Neurosurg. 2010;113(1):141–8. for pituitary adenoma]. Nippon Igaku Hoshasen Gakkai Zasshi.
5. Schulder M, Loeffler JS, Howes AE, Alexander 3rd E, Black 2003;63(5):225–30.
PM. Historical vignette: the radium bomb: Harvey Cushing and the 27. Petrovich Z, Jozsef G, Yu C, Apuzzo ML. Radiotherapy and stereo-
interstitial irradiation of gliomas. J Neurosurg. 1996;84(3):530–2. tactic radiosurgery for pituitary tumors. Neurosurg Clin N Am.
6. Leksell L. The stereotaxic method and radiosurgery of the brain. 2003;14(1):147–66.
Acta Chir Scand. 1951;102(4):316–9. 28. Picozzi P, Losa M, Mortini P, Valle MA, Franzin A, Attuati L, et al.
7. Barnett GH, Linskey ME, Adler JR, Cozzens JW, Friedman WA, Radiosurgery and the prevention of regrowth of incompletely
Heilbrun MP, et al. Stereotactic radiosurgery—an organized removed nonfunctioning pituitary adenomas. J Neurosurg. 2005;
neurosurgery-sanctioned definition. J Neurosurg. 2007;106(1):1–5. 102(Suppl):71–4.
8. Friedman WA, Foote KD. Linear accelerator radiosurgery in the 29. Pollock BE, Cochran J, Natt N, Brown PD, Erickson D, Link MJ,
management of brain tumours. Ann Med. 2000;32(1):64–80. et al. Gamma knife radiosurgery for patients with nonfunctioning
9. Chen CC, Chapman P, Petit J, Loeffler J. Proton radiosurgery in pituitary adenomas: results from a 15-year experience. Int J Radiat
neurosurgery. Neurosurg Focus. 2007;23(6):E5. Oncol Biol Phys. 2008;70(5):1325–9.
10. Höybye C, Rähn T. Adjuvant gamma knife radiosurgery in non- 30. Sheehan JP, Kondziolka D, Flickinger J, Lunsford LD. Radiosurgery
functioning pituitary adenomas; low risk of long-term complica- for residual or recurrent nonfunctioning pituitary adenoma. J
tions in selected patients. Pituitary. 2009;12(3):211–6. Neurosurg. 2002;97(5 Suppl):408–14.
11. Pan L, Zhang N, Wang E, Wang B, Xu W. Pituitary adenomas: the 31. Shin M, Kurita H, Sasaki T, Tago M, Morita A, Ueki K, et al.
effect of gamma knife radiosurgery on tumor growth and endocri- Stereotactic radiosurgery for pituitary adenoma invading the cav-
nopathies. Stereotact Funct Neurosurg. 1998;70 Suppl 1:119–26. ernous sinus. J Neurosurg. 2000;93 Suppl 3:2–5.
12. Feigl GC, Bonelli CM, Berghold A, Mokry M. Effects of gamma 32. Voges J, Kocher M, Runge M, Poggenborg J, Lehrke R, Lenartz D,
knife radiosurgery of pituitary adenomas on pituitary function. J et al. Linear accelerator radiosurgery for pituitary macroadenomas:
Neurosurg. 2002;97(5 Suppl):415–21. a 7-year follow-up study. Cancer. 2006;107(6):1355–64.
13. Hayashi M, Izawa M, Hiyama H, Nakamura S, Atsuchi S, Sato H, 33. Witt TC, Kondziolka D, Flickinger J, Lunsford LD. Gamma knife
et al. Gamma knife radiosurgery for pituitary adenomas. Stereotact radiosurgery for pituitary tumors. In: Lunsford LD, Kondziolka D,
Funct Neurosurg. 1999;72 Suppl 1:111–8. Flickinger J, editors. Gamma knife brain surgery. Basel: Karger;
14. Inoue HK, Kohga H, Hirato M, Sasaki T, Ishihara J, Shibazaki T, 1998. p. 114–27.
et al. Pituitary adenomas treated by microsurgery with or without 34. Wowra B, Stummer W. Efficacy of gamma knife radiosurgery for
gamma knife surgery: experience in 122 cases. Stereotact Funct nonfunctioning pituitary adenomas: a quantitative follow up with
Neurosurg. 1999;72 Suppl 1:125–31. magnetic resonance imaging-based volumetric analysis. J Neurosurg.
15. Izawa M, Hayashi M, Nakaya K, Satoh H, Ochiai T, Hori T, et al. 2002;97(5 Suppl):429–32.
Gamma knife radiosurgery for pituitary adenomas. J Neurosurg. 35. Yoon SC, Suh TS, Jang HS, Chung SM, Kim YS, Ryu MR, et al.
2000;93 Suppl 3:19–22. Clinical results of 24 pituitary macroadenomas with linac-based
16. Kajiwara K, Saito K, Yoshikawa K, Kato S, Akimura T, Nomura S, stereotactic radiosurgery. Int J Radiat Oncol Biol Phys. 1998;41(4):
et al. Image-guided stereotactic radiosurgery with the CyberKnife for 849–53.
pituitary adenomas. Minim Invasive Neurosurg. 2005;48(2):91–6. 36. Starke RM, Williams BJ, Jane Jr JA, Sheehan JP. Gamma knife
17. Kobayashi T. Long-term results of stereotactic gamma knife radiosurgery for patients with nonfunctioning pituitary macroadenomas:
surgery for pituitary adenomas. Specific strategies for different predictors of tumor control, neurological deficits, and hypopituita-
types of adenoma. Prog Neurol Surg. 2009;22:77–95. rism. J Neurosurg. 2012;117(1):129–35.
18. Lim YL, Leem W, Kim TS, Rhee BA, Kim GK. Four years’ 37. Gopalan R, Schlesinger D, Vance ML, Laws E, Sheehan J. Long-term
experiences in the treatment of pituitary adenomas with gamma outcomes after gamma knife radiosurgery for patients with a nonfunc-
knife radiosurgery. Stereotact Funct Neurosurg. 1998;70 Suppl tioning pituitary adenoma. Neurosurgery. 2011;69(2):284–93.
1:95–109. 38. Biller BMK, Grossman AB, Stewart PM, Melmed S, Bertagna X,
19. Liscak R, Vladyka V, Marek J, Simonova G, Vymazal J. Gamma Bertherat J, et al. Treatment of adrenocorticotropin-dependent
knife radiosurgery for endocrine-inactive pituitary adenomas. Acta Cushing’s syndrome: a consensus statement. J Clin Endocrinol
Neurochir (Wien). 2007;149(10):999–1006; discussion. Metab. 2008;93(7):2454–62.
20. Losa M, Valle M, Mortini P, Franzin A, da Passano CF, Cenzato M, 39. Castinetti F, Nagai M, Dufour H, Kuhn JM, Morange I, Jaquet P,
et al. Gamma knife surgery for treatment of residual nonfunction- et al. Gamma knife radiosurgery is a successful adjunctive treat-
ing pituitary adenomas after surgical debulking. J Neurosurg. ment in Cushing’s disease. Eur J Endocrinol. 2007;156(1):91–8.
2004;100(3):438–44. 40. Castinetti F, Nagai M, Morange I, Dufour H, Caron P, Chanson P,
21. Martinez R, Bravo G, Burzaco J, Rey G. Pituitary tumors and et al. Long-term results of stereotactic radiosurgery in secretory
gamma knife surgery. Clinical experience with more than two years pituitary adenomas. J Clin Endocrinol Metab. 2009;94(9):3400–7.
of follow-up. Stereotact Funct Neurosurg. 1998;70 Suppl 1:110–8. 41. Choi JY, Chang JH, Chang JW, Ha Y, Park YG, Chung
22. Mingione V, Yen CP, Vance ML, Steiner M, Sheehan J, Laws ER, SS. Radiological and hormonal responses of functioning pituitary
et al. Gamma surgery in the treatment of nonsecretory pituitary adenomas after gamma knife radiosurgery. Yonsei Med J. 2003;
macroadenoma. J Neurosurg. 2006;104(6):876–83. 44(4):602–7.
42. Degerblad M, Rahn T, Bergstrand G, Thoren M. Long-term results 62. Wong GK, Leung CH, Chiu KW, Ma R, Cockram CS, Lam MJ,
of stereotactic radiosurgery to the pituitary gland in Cushing’s dis- et al. LINAC radiosurgery in recurrent Cushing’s disease after
ease. Acta Endocrinol. 1986;112(3):310–4. transsphenoidal surgery: a series of 5 cases. Minim Invasive
43. Devin JK, Allen GS, Cmelak AJ, Duggan DM, Blevins LS. The Neurosurg. 2003;46(6):327–30.
efficacy of linear accelerator radiosurgery in the management of 63. Melmed S. Acromegaly. N Engl J Med. 2006;355(24):2558–73.
patients with Cushing’s disease. Stereotact Funct Neurosurg. 64. Attanasio R, Epaminonda P, Motti E, Giugni E, Ventrella L, Cozzi
2004;82(5–6):254–62. R, et al. Gamma-knife radiosurgery in acromegaly: a 4-year follow-
44. Ganz JC, Backlund EO, Thorsen FA. The effects of gamma knife up study. J Clin Endocrinol Metab. 2003;88(7):3105–12.
surgery of pituitary adenomas on tumor growth and endocrinopa- 65. Castinetti F, Taieb D, Kuhn J-M, Chanson P, Tamura M, Jaquet P,
thies. Stereotact Funct Neurosurg. 1993;61 Suppl 1:30–7. et al. Outcome of gamma knife radiosurgery in 82 patients with
45. Hoybye C, Grenback E, Rahn T, Degerblad M, Thoren M, Hulting acromegaly: correlation with initial hypersecretion. J Clin
AL. Adrenocorticotropic hormone-producing pituitary tumors: 12- Endocrinol Metab. 2005;90(8):4483–8.
to 22-year follow-up after treatment with stereotactic radiosurgery. 66. Fukuoka S, Ito T, Takanashi M, Hojo A, Nakamura H. Gamma
Neurosurgery. 2001;49(2):284–91; discussion 91–2. knife radiosurgery for growth hormone-secreting pituitary adeno-
46. Jagannathan J, Sheehan JP, Pouratian N, Laws ER, Steiner L, Vance mas invading the cavernous sinus. Stereotact Funct Neurosurg.
ML. Gamma knife surgery for Cushing’s disease. J Neurosurg. 2001;76(3–4):213–7.
2007;106(6):980–7. 67. Ikeda H, Jokura H, Yoshimoto T. Transsphenoidal surgery and
47. Kajiwara K, Saito K, Yoshikawa K, Kato S, Akimura T, Nomura S, adjuvant gamma knife treatment for growth hormone-secreting
et al. Image-guided stereotactic radiosurgery with the CyberKnife for pituitary adenoma. J Neurosurg. 2001;95(2):285–91.
pituitary adenomas. Minim Invasive Neurosurg. 2005;48(2):91–6. 68. Jagannathan J, Sheehan JP, Pouratian N, Laws Jr ER, Steiner L,
48. Kim SH, Huh R, Chang JW, Park YG, Chung SS. Gamma knife Vance ML. Gamma knife radiosurgery for acromegaly: outcomes
radiosurgery for functioning pituitary adenomas. Stereotact Funct after failed transsphenoidal surgery. Neurosurgery. 2008;62(6):
Neurosurg. 1999;72 Suppl 1:101–10. 1262–9; discussion 9–70.
49. Kobayashi T, Kida Y, Mori Y. Gamma knife radiosurgery in the 69. Jezkova J, Marek J, Hana V, Krsek M, Weiss V, Vladyka V, et al.
treatment of Cushing disease: long-term results. J Neurosurg. Gamma knife radiosurgery for acromegaly—long-term experience.
2002;97(5 Suppl):422–8. Clin Endocrinol (Oxf). 2006;64(5):588–95.
50. Laws ER, Reitmeyer M, Thapar K, Vance ML. Cushing’s disease 70. Kobayashi T, Mori Y, Uchiyama Y, Kida Y, Fujitani S. Long-term
resulting from pituitary corticotrophic microadenoma. Treatment results of gamma knife surgery for growth hormone-producing
results from transsphenoidal microsurgery and gamma knife radio- pituitary adenoma: is the disease difficult to cure? J Neurosurg.
surgery. Neurochirurgie. 2002;48(2–3 Pt 2):294–9. 2005;102(Suppl):119–23.
51. Levy RP, Fabrikant JI, Frankel KA, Phillips MH, Lyman JT, 71. Landolt AM, Haller D, Lomax N, Scheib S, Schubiger O, Siegfried J,
Lawrence JH, et al. Heavy-charged-particle radiosurgery of the et al. Stereotactic radiosurgery for recurrent surgically treated acro-
pituitary gland: clinical results of 840 patients. Stereotact Funct megaly: comparison with fractionated radiotherapy. J Neurosurg.
Neurosurg. 1991;57(1–2):22–35. 1998;88(6):1002–8.
52. Morange-Ramos I, Regis J, Dufour H, Andrieu JM, Grisoli F, 72. Laws Jr ER, Vance ML. Radiosurgery for pituitary tumors and cra-
Jaquet P, et al. Gamma-knife surgery for secreting pituitary adeno- niopharyngiomas. Neurosurg Clin N Am. 1999;10(2):327–36.
mas. Acta Neurochir. 1998;140(5):437–43. 73. Losa M, Gioia L, Picozzi P, Franzin A, Valle M, Giovanelli M, et al.
53. Petit JH, Biller BMK, Coen JJ, Swearingen B, Ancukiewicz M, The role of stereotactic radiotherapy in patients with growth
Bussiere M, et al. Proton stereotactic radiosurgery in management hormone-secreting pituitary adenoma. J Clin Endocrinol Metab.
of persistent acromegaly. Endocr Pract. 2007;13(7):726–34. 2008;93(7):2546–52 [see comment].
54. Pollock BE, Brown PD, Nippoldt TB, Young Jr WF. Pituitary tumor 74. Muramatsu J, Yoshida M, Shioura H, Kawamura Y, Ito H, Takeuchi
type affects the chance of biochemical remission after radiosurgery H, et al. [Clinical results of LINAC-based stereotactic radiosurgery
of hormone-secreting pituitary adenomas. Neurosurgery. 2008;62(6): for pituitary adenoma]. Nihon Igaku Hoshasen Gakkai Zasshi.
1271–6; discussion 6–8. 2003;63(5):225–30.
55. Pollock BE, Nippoldt TB, Stafford SL, Foote RL, Abboud 75. Pollock BE, Jacob JT, Brown PD, Nippoldt TB. Radiosurgery of
CF. Results of stereotactic radiosurgery in patients with hormone- growth hormone-producing pituitary adenomas: factors associated
producing pituitary adenomas: factors associated with endocrine with biochemical remission. J Neurosurg. 2007;106(5):833–8.
normalization. J Neurosurg. 2002;97(3):525–30. 76. Pollock BE, Nippoldt TB, Stafford SL, Foote RL, Abboud
56. Seo Y, Fukuoka S, Takanashi M, Sasaki T, Suematsu K, Nakamura CF. Results of stereotactic radiosurgery in patients with hormone-
J. Gamma knife surgery for Cushing’s disease. Surg Neurol. producing pituitary adenomas: factors associated with endocrine
1995;43(2):170–6. normalization. J Neurosurg. 2002;97(3):525–30.
57. Sheehan JM, Vance ML, Sheehan JP, Ellegala DB, Laws Jr 77. Thoren M, Rahn T, Guo WY, Werner S. Stereotactic radiosurgery
ER. Radiosurgery for Cushing’s disease after failed transsphenoidal with the cobalt-60 gamma unit in the treatment of growth
surgery. J Neurosurg. 2000;93(5):738–42. hormone-producing pituitary tumors. Neurosurgery. 1991;29(5):
58. Tinnel BA, Henderson MA, Witt TC, Fakiris AJ, Worth RM, Des 663–8.
Rosiers PM, et al. Endocrine response after gamma knife-based ste- 78. Vik-Mo EO, Oksnes M, Pedersen P-H, Wentzel-Larsen T, Rodahl
reotactic radiosurgery for secretory pituitary adenoma. Stereotact E, Thorsen F, et al. Gamma knife stereotactic radiosurgery for acro-
Funct Neurosurg. 2008;86(5):292–6. megaly. Eur J Endocrinol. 2007;157(3):255–63.
59. Voges J, Kocher M, Runge M, Poggenborg J, Lehrke R, Lenartz D, 79. Franzin A, Spatola G, Losa M, Picozzi P, Mortini P. Results of
et al. Linear accelerator radiosurgery for pituitary macroadenomas: gamma knife radiosurgery in acromegaly. Int J Endocrinol.
a 7-year follow-up study. Cancer. 2006;107(6):1355–64. 2012;2012:342034.
60. Wan H, Chihiro O, Yuan S. MASEP gamma knife radiosurgery for 80. Sheehan JP, Pouratian N, Steiner L, Laws ER, Vance ML. Gamma
secretory pituitary adenomas: experience in 347 consecutive cases. knife surgery for pituitary adenomas: factors related to radiological
J Exp Clin Cancer Res. 2009;28:36. and endocrine outcomes. J Neurosurg. 2011;114(2):303–9.
61. Witt TC. Stereotactic radiosurgery for pituitary tumors. Neurosurg 81. Landolt AM, Lomax N. Gamma knife radiosurgery for prolactino-
Focus. 2003;14(5):e10. mas. J Neurosurg. 2000;93 Suppl 3:14–8.
82. Pan L, Zhang N, Wang EM, Wang BJ, Dai JZ, Cai PW. Gamma 89. Ma ZM, Qiu B, Hou YH, Liu YS. [Gamma knife treatment for
knife radiosurgery as a primary treatment for prolactinomas. J pituitary prolactinomas]. Zhong Nan Da Xue Xue Bao Yi Xue Ban.
Neurosurg. 2000;93 Suppl 3:10–3. 2006;31(5):714–6.
83. Choi JY, Chang JH, Chang JW, Ha Y, Park YG, Chung SS. 90. Liu X, Kano H, Kondziolka D, Park KJ, Iyer A, Shin S, et al.
Radiological and hormonal responses of functioning pituitary ade- Gamma knife stereotactic radiosurgery for drug resistant or intoler-
nomas after gamma knife radiosurgery. Yonsei Med J. 2003;44(4): ant invasive prolactinomas. Pituitary. 2013;16(1):68–75.
602–7. 91. Jagannathan J, Yen C-P, Pouratian N, Laws ER, Sheehan JP.
84. Pouratian N, Sheehan J, Jagannathan J, Laws Jr ER, Steiner L, Stereotactic radiosurgery for pituitary adenomas: a comprehensive
Vance ML. Gamma knife radiosurgery for medically and surgically review of indications, techniques and long-term results using the
refractory prolactinomas. Neurosurgery. 2006;59(2):255–66; dis- gamma knife. J Neurooncol. 2009;92(3):345–56.
cussion −66. 92. Landolt AM, Haller D, Lomax N, Scheib S, Schubiger O, Siegfried
85. Jezkova J, Hana V, Krsek M, Weiss V, Vladyka V, Liscak R, et al. J, et al. Octreotide may act as a radioprotective agent in acromegaly.
Use of the Leksell gamma knife in the treatment of prolactinoma J Clin Endocrinol Metab. 2000;85(3):1287–9.
patients. Clin Endocrinol (Oxf). 2009;70(5):732–41. 93. Taussky P, Kalra R, Coppens J, Mohebali J, Jensen R, Couldwell WT.
86. Kobayashi T. Long-term results of stereotactic gamma knife radio- Endocrinological outcome after pituitary transposition (hypophyso-
surgery for pituitary adenomas. Specific strategies for different pexy) and adjuvant radiotherapy for tumors involving the cavernous
types of adenoma. Prog Neurol Surg. 2009;22:77–95. sinus. J Neurosurg. 2011;115(1):55–62.
87. Tanaka S, Link MJ, Brown PD, Stafford SL, Young Jr WF, Pollock 94. Lim YJ, Leem W, Park JT, Kim TS, Rhee BA, Kim GK. Cerebral
BE. Gamma knife radiosurgery for patients with prolactin-secreting infarction with ICA occlusion after gamma knife radiosurgery for
pituitary adenomas. World Neurosurg. 2010;74(1):147–52. pituitary adenoma: a case report. Stereotact Funct Neurosurg.
88. Sun DQ, Cheng JJ, Frazier JL, Batra S, Wand G, Kleinberg LR, 1999;72 Suppl 1:132–9.
et al. Treatment of pituitary adenomas using radiosurgery and 95. Loeffler JS, Niemierko A, Chapman PH. Second tumors after
radiotherapy: a single center experience and review of literature. radiosurgery: tip of the iceberg or a bump in the road? Neurosurgery.
Neurosurg Rev. 2011;34(2):181–9. 2003;52(6):1436–40; discussion 40–2.
Pituitary Tumors: Viewpoint—Surgery
27
Khaled M. Krisht, William T. Couldwell,
and Martin H. Weiss
concerned with treating pituitary adenomas, with special

Introduction emphasis on surgical approaches and technique and the role
of stereotactic radiosurgery (SRS).
Since the first recorded attempt at resection of a pituitary
tumor in an acromegalic patient via a staged lateral subtem-
poral approach by Caton and Paul [1], surgical approaches Indications for Surgery
to the sellar, suprasellar, and parasellar regions have under-
gone a profound evolutionary change, with refinements in The gold standard for treating both symptomatic extra-
both microsurgical technique and instrumentation. In addi- glandular nonsecreting and glandular secreting pituitary ade-
tion, the advent of both microscopic and endoscopic means nomas, with the exception of prolactin-secreting tumors, is
for improved visualization and understanding of the surgical resection. A trial of medical therapy with octreotide or
nuances of sellar and suprasellar pathology has created an cabergoline for growth hormone (GH)-secreting tumors may
algorithmic and structured approach to treating sellar be warranted (especially in GH–prolactin (PRL)-cosecreting
pathology, especially in dealing with pituitary adenomas. tumors). Ketoconazole may be used for adrenocorticotropic
The first successful removal of a pituitary tumor via a supe- hormone (ACTH)-secreting tumors as an initial mode of ther-
rior nasal transsphenoidal approach was reported in 1907 apy in patients with multiple comorbidities or in those patients
by Schloffer (Fig. 27.1) [2–4]. Owing to the disfiguring who are reluctant to undergo surgery [8–12]. In patients who
effects of this approach, Cushing used the sublabial trans- fail to respond to medical therapy or those who experience
septal technique, which he later abandoned in favor of the intolerable side effects, surgical resection becomes necessary.
transcortical route in part because of the large number of In certain instances of progressive visual loss caused by
recurrent tumor cases he observed in his practice [5, 6]. It increasing mass effect or in cases of sudden onset of visual
was not until 50 years later that Guiot and Hardy further loss or ophthalmoplegia associated with severe headache and
refined and popularized the transsphenoidal technique, acute adrenal insufficiency as with pituitary apoplexy, emer-
reintroducing it back to the United States [7]. This chapter gent surgical decompression of the optic apparatus along with
will review the management strategies and decision making steroid replacement is warranted [13–16]. The general goals
of surgery are removal of the tumor mass with decompression
K.M. Krisht, M.D. • W.T. Couldwell, M.D., Ph.D. (*) of the optic apparatus, normalization of hormonal hypersecre-
Department of Neurosurgery, Clinical Neurosciences Center, tion, and preservation of normal pituitary function (Fig. 27.2).
University of Utah, 175 North Medical Drive East, Radiation therapy, either fractionated radiosurgery or gamma
Salt Lake City, UT 84132, USA knife surgery, may be used as part of a planned combination
e-mail: neuropub@hsc.utah.edu
therapy for large tumors extending into the cavernous sinus
M.H. Weiss, M.D. with internal carotid artery encasement, as a salvage treatment
Department of Neurological Surgery, Keck School of Medicine,
University of Southern California, 1200 North State Street, Suite for recurrent tumor, or as the primary mode of treatment for
3300, Los Angeles, CA 90089, USA patients who are poor surgical candidates [17–20].
380 K.M. Krisht et al.
with the help of an endocrinologist. Hyperthyroidism should

be treated with thyroxine, with the aim of reducing cardiac and
sympathetic overactivity. Elevated blood sugar levels caused
by GH- and ACTH-secreting tumors should be managed.
Conversely, patients with hypoactive pituitary function should
receive appropriate replacement therapy to reestablish homeo-
stasis before undergoing surgery [12, 21, 22]. All patients
should have baseline fasting pituitary laboratory measurements
including prolactin, thyroid-stimulating hormone (TSH), free
T4, ACTH, and fasting AM cortisol; follicle-stimulating hor-
mone, luteinizing hormone, GH, and insulin-like growth factor
(IGF-1); and testosterone (in males) and estradiol (in females)
levels [11, 23, 24]. Patients with visual disturbance should
undergo a formal baseline visual field and acuity testing by a
neuro-ophthalmologist [25]. Special attention should be paid to
thyroid and steroid replacement; low levels of these hormones
may result in increased surgical and perioperative risk, and they
should be replaced accordingly. Hypocortisolemia may be
replaced acutely with intravenous corticosteroids, but thyroid
hormone deficit may take several days to replace.
In planning for surgery, the anesthesiologist should be
cognizant of the medical and physical stigmata of the pitu-
itary tumor–related endocrinopathies because they have
important implications with regard to preoperative care and
readiness. Patients on bromocriptine therapy may suffer from
gastroparesis and have occasional vomiting [16, 23, 26].
Such patients should be appropriately identified and suitable
measures taken to control their nausea and prevent aspira-
tion. Acromegalic patients should be evaluated for sleep
apnea, snoring, and airway changes prior to intubation
[27–29]. In addition, electrocardiography and a transthoracic
echocardiogram to evaluate for baseline left ventricular
Fig. 27.1 Schloffer’s transnasal transsphenoidal operation. (a) Incision function are warranted since patients with acromegaly are at
made along the left nasolabial furrow around the left ala nasi and con- risk for hypertrophic cardiomyopathy [30, 31]. If an arterial
tinued up to the glabella. (b) The incision cuts through the skin, nasal line is required throughout the case, especially in patients
bone, philtrum, and the anterior part of the septum. The whole external with known hypertension, some anesthesiologists will advo-
nose is reflected to the right, exposing the remainder of the septum. (c)
The rest of the nasal septum has been removed, exposing the rostrum of cate placing the line in the dorsalis pedis artery since acro-
the sphenoid sinus. (d) The anterior wall of the sphenoid sinus is megalic patients may have an impaired ulnar artery
opened, the mucosa lining of the sinus is removed with a sharp spoon, circulation. In the event that a radial arterial line is contem-
and the floor of the sella is removed with a small chisel or punch for- plated, an Allen test should be performed to document good
ceps. (From: Cope VZ: The pituitary fossa, and the methods of surgical
approach thereto. Br J Surg 4:107–144, 1916) collateral flow [32]. Cushingoid patients with hypercorti-
solemia are easily bruised, and their peripheral veins can be
difficult to cannulate [16]. Special care in both positioning
Endocrinologic, Anesthetic, and venous access should be practiced in such patients.
and Radiographic Considerations Intravenous hydrocortisone (100 mg) may be given at the
and Planning start of the procedure to prevent diabetes insipidus resulting
from antidiuretic hormone deficiency [23, 24, 33]. The anes-
The radiographic features of the pituitary tumor and its associ- thesiologist also needs to be mindful of the very short clo-
ated endocrinopathies have important medical, anesthetic, and sure time involved at the conclusion of the resection to
surgical implications. Before considering surgical intervention, adjust the anesthetic dose accordingly in preparation for
control and medical management of any preexisting medical awakening. In the rare instance that air is to be injected
conditions and endocrinologic abnormalities should be insti- through a lumbar drain to elevate the cerebrospinal fluid
tuted. This is best accomplished in an interdisciplinary manner pressure and aid in the delivery and resection of a suprasellar
27 Pituitary Tumors: Viewpoint—Surgery 381
Fig. 27.2 (a) MR images of the brain with gadolinium enhancement of the left cavernous sinus in a 38-year-old Cushingoid woman who
demonstrating a large sellar and suprasellar GH-secreting pituitary presented with an elevated fasting serum cortisol level of 99 μg/dL and
macroadenoma in a 36-year-old acromegalic woman who presented significantly elevated 24-h urinary cortisol. The patient underwent a
with a 4-year history of amenorrhea and a GH level of 60 ng/mL. Gross microscopic transsphenoidal resection of her tumor with endoscopic
total resection was achieved via a transnasal transsphenoidal resection. assistance. A cure was achieved with a drop in her postoperative
(b) MR images of the brain with gadolinium enhancement showing a cortisol level to 1.7 μg/dL. She was subsequently placed on hormone
sellar ACTH-secreting pituitary adenoma with extension into the region replacement therapy
component of the tumor (a practice that is not employed at amenable to transsphenoidal resection with wider removal of
our facility), the anesthesiologist should eliminate nitrous the skull base for exposure. Some lesions are fibrous in consis-
oxide from the anesthetic gas mixture to prevent tension tency, making them difficult to resect through a transsphenoidal
pneumocephalus [34]. route, especially in cases where the tumor may be adherent to
The surgical route of choice is best determined by careful important nearby neurovascular structures, or masking impor-
inspection and evaluation of the morphology of the tumor of tant arterial feeders such as the superior hypophyseal artery sup-
interest and the area it occupies on thin-sliced coronal and sag- plying the optic chiasm [35, 37, 39–41]. Another important
ittal magnetic resonance images through the sellar region consideration is the proximity of the tumor to the intracavern-
while considering ease of access and patient safety. The vast ous carotid arteries. This is particularly important for identify-
majority of pituitary tumors are removed via the transsphenoi- ing deviations of the eccentric carotid arteries, which may
dal approach. Lesions confined to the sella without any supra- deviate into a midline trajectory and preclude a transsphenoidal
sellar extension can be easily resected via a transsphenoidal approach [37]. Fine-cut computed tomography (CT) of the head
route. Despite the suprasellar extension of some sellar lesions, with three-dimensional reconstruction helps identify important
however, transsphenoidal surgery may still be used as long as bony landmarks, demonstrate the extent of sellar expansion,
the tumor appears to be centrally located and symmetric. The and elucidate any evidence of tumor erosion through the sellar
“hour-glass” or “dumb-bell” appearance of a tumor may wall. This becomes increasingly important in repeat transsphe-
necessitate a combined approach for maximal safe resec- noidal surgeries in which the normal anatomy is perturbed and
tion, if the tumor does not descend from below [35–38]. may therefore be misleading [37, 42]. In such a scenario, a
Asymmetric extension of the tumor either anteriorly or stereotactic CT scan is useful for stereotactic neuronavigation,
posteriorly may pose a challenge for the surgeon attempt- which helps to confirm midline and avoid inadvertent injury to
ing to tackle it via a transsphenoidal approach but may be the carotid or violation of the anterior fossa space.
and places the sphenoid ostia in the direct line of sight of

Surgical Approaches to the Pituitary the surgeon. The head may be fixed in a three-point Mayfield
frame or on a horseshoe head holder and secured with tape.
The Direct Microscopic Endonasal At least 10 min prior to incision, the nostrils are packed with
Transsphenoidal Approach oxymetazoline-soaked cottonoids to allow shrinkage of the
turbinates and to optimize hemostasis during initial expo-
Hirsch was the first to adopt the endonasal route to the sure. We no longer routinely prep the nose. The ipsilateral
sphenoid sinus using multiple-staged operations including thigh or abdomen is prepped and draped for harvest of fat
resection of the middle turbinate and ethmoidectomy [43]. and fascia. Some surgeons use intraoperative C-arm fluo-
Despite his success, however, he changed his technique to a roscopy for guidance; however, we prefer to use such assis-
transseptal approach. Almost eight decades later, in 1987, tance only in selected cases. The rostrum of the sphenoid is
Griffith and Veerapen revived the endonasal approach to the identified at the level of the superior aspect of the distal
sphenoid sinus [44]. middle turbinate with direct vision using a handheld specu-
The patient is placed in the supine “reflex” position with lum. The sphenoid ostium is found as the superior landmark
the knee and body slightly bent at the mid-position to pre- of the opening of the rostrum just at the level of the inferior
vent entrapment neuropathy, venous pooling, and patient border of the superior turbinate. The perpendicular plate of
slippage (Fig. 27.3). The head is slightly tilted to the contra- the ethmoid is fractured at the junction of the sphenoid
lateral side, the neck is flexed, and the head is extended at rostrum. A self-retaining nasal retractor is placed with the
the occipitocervical joint and slightly turned toward the remnant of the perpendicular plate of the ethmoid in the
surgeon. This ensures that the face is parallel to the floor midline. The mucosa are elevated away from each side of
the sphenoid rostrum. The sphenoid ostia should become
visible, and the sphenoid rostrum is removed with the aid of
micro rongeurs or a high-speed drill. The sella is identified,
and the bone over the sella is removed with micro rongeurs.
A #11 blade is used to create a durotomy in an X-fashion,
and then small, angled microscissors are used to further
open the dura to the extremes of exposure desired (Fig. 27.4).
With large tumors, the intracranial pressure helps deliver the
tumor when the dura is opened. First, the floor of the tumor
is removed with curettes. Then, the lateral aspects of the
tumor are removed to enable descent of the large center part
of the tumor. Four-quadrant curettage with various Hardy
ring curettes allows further tumor removal from lateral,
superior, and inferior recesses. Closure begins with packing
the resection site with adipose tissue. The septum is reduced
to the midline, and the authors elevate the middle turbinates
on both sides to ensure unimpeded drainage of the maxillary
sinuses. The authors employ well-lubricated, soft, curved
Rhino Rocket tubes covered with a piece of latex as airway
tubes, which are carefully placed posteriorly into the naso-
pharynx [45, 46].
Advancements in Endoscopic Surgery
The successful reintroduction of the pure endonasal route

paved the way for the use of endoscopic surgery in address-
Fig. 27.3 Patient and operating room positioning for the transsphenoi- ing sellar and suprasellar lesions [47, 48]. There are many
dal approach. The head is elevated 15° above to reduce venous bleeding. advantages to the employment of endoscopy as a sole or
The head is tilted toward the left shoulder to facilitate endonasal expo-
supplemental tool for the resection of pituitary lesions. The
sure. (From: Couldwell WT, Weiss MH. The transnasal transsphenoidal
approach. In: Apuzzo MLJ, ed. Surgery of the third ventricle. 2nd ed. endoscope provides a panoramic view of the sphenoid sinus
Baltimore: Williams & Wilkins; 1998. p. 553–74, with permission) and its important bony impressions. It also affords the
Diaphragm Anterior Optic

Subfrontal arachnoid of sella cerebral a. chiasm Infundibulum
spaces
Hypothalamus
Tumor
Basilar a.
Dura
A
Sphenoid Arachoid
sinus
Tumor
Window in
sella turcica
B
Dura
C
Arachnoid
Diaphragm
E
F
Fig. 27.4 Tumor removal. Tumor is readily removed with various fluid fistula. (From: Couldwell WT, Weiss MH. The transnasal transsphe-
curettes and scoops. It is imperative that the integrity of the arachnoid is noidal approach. In: Apuzzo MLJ, ed. Surgery of the third ventricle. 2nd
preserved to minimize the potential for a postoperative cerebrospinal ed. Baltimore: Williams & Wilkins; 1998. p. 553–74, with permission)
surgeon with the ability to zoom in on areas of interest such reflected on a screen as a two-dimensional image, which
as the tumor–gland interface and obtain a sharpened close-up results in loss of the sharp and focused stereoscopic view and
view. Lenses of various angles on the endoscope allow the depth perception provided by the microscope. In addition,
operator to look into anatomic corners of the suprasellar there is a steep learning curve for the neurosurgeon inexperi-
region that otherwise cannot be seen with the conventional enced in endoscopy [53–55]. In our hybrid practice, we have
microscope. Additionally, there is no displacement of the been increasingly employing the endoscope as a supplemen-
midline septum, and nasal packing and retraction are not tal tool for inspecting areas that are hidden from the field of
needed [49–52]. Notwithstanding the improved and versatile view of the microscope in all of skull base surgery. In lateral
visibility afforded by endoscopic transsphenoidal surgery, and extreme anterior and posterior tumor rests, and those that
the technique is not without pitfalls. In general, more of the are buried behind redundant folds of suprasellar arachnoid as
nasal septum is resected to allow adequate visual access. well as in cases of tumor extending into the cavernous sinus,
With endoscopic surgery, the surgeon’s field of view is the endoscope is a valuable tool.
Modifications of the Transsphenoidal Optic chiasm Dura of

Tumor tuberculum
Approach
Arachnoid Recess of sellae Circular
infundibulum sinus
The standard transsphenoidal approach is used in most cases A
of pituitary tumors with sellar and suprasellar extension;
however, regions of the skull base that were once thought
only accessible “from above” are now being approached
transfacially. With better knowledge of microsurgical anat-
omy and modern microinstrumentation, neurosurgeons have
Dura over
modified the transsphenoidal approach to gain better access
B pituitary
to regions such as the cavernous sinus and the suprasellar
cisterns.
Fraioli et al. [56] treated 11 patients with sellar tumors
invading the medial wall of the cavernous sinus with a trans-
maxillosphenoidal approach, which involves a unilateral, bilat-
eral, or Le Fort maxillary osteotomy and removal of the
medial wall of the maxillary sinus in addition to the standard Dural opening
transsphenoidal exposure. It allows direct visualization of the
C
intracavernous carotid artery during tumor resection but is
limited by extension of tumor lateral to the carotid artery. Sabit
et al. [57] described a safe, minimally invasive combined trans-
maxillary transsphenoidal approach to the cavernous sinus
that is both extradural and extranasal. This approach provides
adequate lateral-to-medial reach in the parasellar and infrasel- Tumor
Tumor seen
lar regions with visualization of the entire ipsilateral cavern- through arachnoid
ous sinus and the medial aspect of the contralateral cavernous Pituitary
sinus. The extended transsphenoidal approach to tumors of
Circular sinus
the tuberculum sellae and suprasellar cistern was first described
by Weiss [58]. Suprasellar tumors without sellar enlargement D Dural flap
have been successfully resected using a modified transsphe-
Fig. 27.5 (a–d) Modification of the transsphenoidal approach: expo-
noidal approach that involves a wide bone exposure of the
sure of the anterior skull base (extended transsphenoidal). A pure supra-
anterior surface of the sella and removal of the posterior por- sellar tumor may be approached by extending the bony resection
tion of the planum sphenoidale [59, 60]. This novel approach anteriorly over the tuberculum sellae, thus exposing the dura mater
has allowed resection of craniopharyngiomas, central nervous lying anterior to the circular sinus. An incision is made in the dura ante-
riorly and inferiorly to the circular sinus. The sinus is then coagulated
system hemangioblastomas, and ectopic ACTH-producing
and transected to gain a direct view of the suprasellar cistern without
adenomas arising from the pituitary stalk [59, 60]. disturbing the pituitary gland. (From: Couldwell WT, Weiss MH. The
Additional exposure of the skull base for lesions of the transnasal transsphenoidal approach. In: Apuzzo MLJ, ed. Surgery of
parasellar and clival region can be achieved by extended the third ventricle. 2nd ed. Baltimore: Williams & Wilkins; 1998.
p. 553–74, with permission)
transsphenoidal approaches [36]. Various portions of the
skull base may be exposed, and bone resection can be
extended by repositioning the patient’s head and the self- Jho and Carrau [52] reported encouraging results in a
retaining speculum. The standard transsphenoidal approach series of 50 patients who underwent endoscopic endonasal
can be extended anteriorly to resect suprasellar lesions, infe- transsphenoidal surgery. One of the main advantages of
riorly to expose clival lesions, and inferolaterally to access this approach is excellent panoramic visualization of the
cavernous sinus lesions (Figs. 27.5, 27.6, and 27.7) [36]. sellar and suprasellar anatomy with increased illumination
These variations on the transsphenoidal approach provide a and magnification [48]. Anatomic studies have demon-
minimally invasive technique that avoids prolonged surgery strated that the endoscope provides a volume of exposure
and brain retraction. superior to that of the operating microscope [54], but dis-
Technological advances in the areas of endoscope- advantages include the lack of stereoscopic vision, the
assisted microneurosurgery [47, 49] have been applied to lack of adequate instrumentation, the limited space of
the classical transsphenoidal operation in an attempt to working through one nostril [49], and a learning curve for
further decrease morbidity and mortality risks [47, 52, 55]. using this technique.
Sphenoid
sinus Rostrum Chordoma
Clivus
Soft
palate
Pharynx
Fig. 27.6 Modifications of the transsphenoidal approach: inferior exposure additional exposure to the mid and lower clivus requires more inferior
of the clivus. Exposure of the clivus is facilitated by slight flexion of the exposure. (From: Couldwell WT, Weiss MH. The transnasal transsphenoidal
patient’s head and repositioning of the nasal self-retaining retractor to point approach. In: Apuzzo MLJ, ed. Surgery of the third ventricle. 2nd ed.
inferiorly. The upper clivus lies directly posterior to the sphenoid sinus, but Baltimore: Williams & Wilkins; 1998. p. 553–74, with permission)
A B
Floor of Dura of sella
sella Floor of sella Dura over pituitary gland,
cavernous sinus, and ICA
Area of bone
removal
Sphenoid
sinus
Wall of sphenoid
Cavernous sinus
Pituitary
gland sinus
ICA
NII
Dural
NIII
incision
NIV
enlarged
NV
Sphenoid
sinus
Tumor
exposed
Area of
bone removal
Fig. 27.7 Modification of the transsphenoidal approach: inferolateral incised with a size 11 blade and opened with curved alligator microscis-
exposure of the cavernous sinus. (a) After exposure of the dura overly- sors. Removal of the intracavernous portion of the tumor is carried out
ing the sella, the bone overlying the cavernous sinus, including that with a microcurette. (From: Couldwell WT, Weiss MH. The transnasal
overlying the carotid grooves, is carefully removed. This removal transsphenoidal approach. In: Apuzzo MLJ, ed. Surgery of the third
defines the lateral extent of the exposure limited by the cavernous cra- ventricle. 2nd ed. Baltimore: Williams & Wilkins; 1998. p. 553–74,
nial nerves. (b) The dura medial to the internal carotid artery is first with permission)
Fig. 27.8 A giant hemorrhagic pituitary macroadenoma in a 44-year- on the hypothalamus, right medial temporal lobe, and right basal
old man who presented with progressive left-sided weakness and ganglia with encasement of bilateral cavernous carotid arteries. This
visual disturbance. There are areas of scant enhancement with exten- lesion was resected via a transcranial route using a standard pterional
sion into the suprasellar and prepontine cisterns exerting mass effect craniotomy
Transcranial Surgery for Pituitary Tumors

Radiation Therapy and Radiosurgery
With the increasing popularity of the transsphenoidal micro- of Pituitary Tumors
scopic or endoscopic approach, the indications for transcra-
nial approaches have become few. There are instances where Improvements in radiation techniques with the development
the transsphenoidal route cannot be employed despite agree- of medical linear accelerator systems and the enhancement in
able tumor morphology and location. Such is the case in stereotactic brain imaging have led to more effective and
patients with severe sphenoid sinusitis or midline “kissing” safer radiation treatments. Today, both radiation therapy
internal carotid arteries [40, 41]. In other instances, a large (XRT) and SRS play an important role in treating recurrent
eccentric suprasellar component of the tumor may be difficult pituitary tumors in patients with hypopituitarism who may
to access from a transsphenoidal approach without putting not tolerate surgical re-resection and residual tumors invad-
the patient at risk (especially extensions lateral to the supra- ing the cavernous sinus [62–67]. In the latter case, fraction-
sellar carotid artery) (Fig. 27.8). For these scenarios, a tran- ated radiotherapy can provide excellent tumor control without
scranial approach would be a safer and more suitable option. risking injury to the nearby cranial nerves. XRT has the
The two most widely used transcranial approaches are the potential of delivering high dosages of radiation (up to 45 Gy)
pterional (frontotemporal) approach, which was first directly to the tumor site while reducing radiation to the sur-
described in detail and popularized by Yasargil [61], and the rounding structures. With fractionated radiotherapy, the
anterior subfrontal approach. The pterional approach presents 45-Gy dose is given over 5 weeks in fractionated doses of
the shortest transcranial trajectory to the suprasellar cistern 1.8 Gy/day, 5 days a week [68, 69]. SRS uses stereotactically
and should be the method of choice in a patient with a pre- focused high-energy proton particles generated by a cyclo-
fixed chiasm since the tumor can be resected inferior, supe- tron to irradiate the tumor. The highest dose possible is deliv-
rior, and posterior to the chiasm. In the case of a postfixed ered in a focused manner while keeping the dose to the optic
chiasm where the tumor lies between the two optic nerves, apparatus less than or equal to 8 Gy [69–71]. The preliminary
the subfrontal approach gives the surgeon a straight frontal data for tumor control and normalization of hypersecretory
trajectory with direct visualization of the tumor [35, 39–41]. states after radiosurgery appears promising for GH- and
In performing the latter approach, the surgeon has to be mind- ACTH-secreting pituitary adenomas [69–73]. XRT and SRS,
ful of the location of the frontal sinus to avoid a breach and to however, rarely normalize prolactin levels in patients with
make every effort to preserve the olfactory nerves. On rare prolactinomas. In fact, radiotherapy may have the opposite
occasions when a large proportion of the tumor lies within the effect of elevating the prolactin level secondary to disruption
third ventricle, an interhemispheric transcallosal approach of the tuberohypophysial pathway [66, 70, 72]. Despite the
may be warranted either as a stand-alone approach or in com- positive and encouraging results experienced with XRT, sur-
bination with a pterional craniotomy [36, 40, 41]. gical resection remains the initial mainstay treatment for both
Fig. 27.9 Pituitary transposition (hypophysopexy) with planned ered to surrounding tissues on coronal (b) and axial images (c). Note
radiosurgical treatment of residual cavernous sinus tumor. Acromegalic lateral displacement of the gland by the interposed fat graft on coronal
young female with residual tumor within the cavernous sinus follow- image. Diagram representing dosimetry target in relationship to brain-
ing hypophysopexy. Coronal (a) T1-weighted gadolinium-enhanced stem and optic apparatus with transposed pituitary in green (d). (From
MR images demonstrating the relationship of the cavernous sinus Couldwell et al., Hypophysopexy technique for radiosurgical treat-
tumor (T), fat graft (F), optic chiasm (OC), and transposed pituitary ment of cavernous sinus pituitary adenoma. Pituitary 5:169–173, 2002;
gland (P). Isodose curves on coronal image demonstrating doses deliv- used with permission)
ACTH- and GH-secreting tumors [62–65, 67, 74, 75]. XRT tumors as a stand-alone treatment or in combination with a
and SRS do not lead to any substantial visual improvement in planned transsphenoidal re-resection as well as in cases of
cases where the tumor is compressing the optic apparatus recurrent tumors with hypopituitarism and difficult to access
[18, 71, 72, 76]. In such instances, immediate surgical decom- locations as discussed above [17–20]. The proximity of the
pression is required. The reported improvements in hormonal pituitary gland to the region of interest increases the risk for
levels and anatomic features after radiotherapy of hyperse- developing hypopituitarism even with SRS. To curtail the
cretory tumors take up to 5 years to occur; however, since effect of radiation on the pituitary gland in a planned adjuvant
metabolic effects of the glandular tumors may be taxing on radiosurgery case after transsphenoidal resection, the authors
many organ systems, especially the cardiovascular system, have developed a technique (hypophysopexy) to transpose the
prompt and precipitous reduction in the hormonal level pituitary gland away from the planned radiation site (residual
through surgical resection remains an important first-line tumor within the cavernous sinus) [77–79]. This technique
treatment measure [70, 71]. Radiosurgery is, however, an involves interposing a fat graft between the pituitary gland
attractive treatment option for recurrent ACTH-secreting and the residual tumor in the cavernous sinus (Fig. 27.9).
Treatment of the residual tumor with SRS is facilitated by the 23. Messick Jr JM, Laws Jr ER, Abboud CF. Anesthesia for
greater distance between the pituitary gland and the tumor, transsphenoidal surgery of the hypophyseal region. Anesth Analg.
1978;57(2):206–15.
therefore reducing the radiation exposure to the gland and the 24. Roizen M. Diseases of the endocrine system. In: Katz B, Beumof J,
likelihood for developing subsequent hypopituitarism. Kadis L, editors. Anesthesia and uncommon diseases. Philadelphia:
WB Saunders; 1990. p. 265.
25. Rush SC, Kupersmith MJ, Lerch I, Cooper P, Ransohoff J, Newall
J. Neuro-ophthalmological assessment of vision before and after
References radiation therapy alone for pituitary macroadenomas. J Neurosurg.
1990;72(4):594–9.
1. Caton R. Notes of a case of acromegaly treated by operation. Br 26. Black PM, Zervas NT, Candia GL. Incidence and management of
Med J. 1893;2(1722):1421–3. complications of transsphenoidal operation for pituitary adenomas.
2. Schloffer H. Zur frage der operationen an der hypophyse. Beitr Klin Neurosurgery. 1987;20(6):920–4.
Chir. 1906;50:767–817. 27. Ovassapian A, Doka JC, Romsa DE. Acromegaly—use of fiberop-
3. Schloffer H. Erfolgreiche operation eines hypophysentumor auf tic laryngoscopy to avoid tracheostomy. Anesthesiology. 1981;
nasalem wege. Wien Med Wochenschr. 1907;20:621–4. 54(5):429–30.
4. Schloffer H. Weiterer bericht uber den fall von operiertem 28. Perks WH, Horrocks PM, Cooper RA, Bradbury S, Allen A,
hypophysen-tumor. Wien Med Wochenschr. 1907;20:1075–8. Baldock N, et al. Sleep apnoea in acromegaly. Br Med J. 1980;
5. Cope V. The pituitary fossa, and the methods of surgical approach 280(6218):894–7.
thereto. Br J Surg. 1916;4:107–44. 29. Siegler J. Acromegaly associated with laryngeal obstruction.
6. Cushing H. Surgical experiences with pituitary disorders. JAMA. J Laryngol Otol. 1952;66(12):620–1.
1914;63:1515. 30. McGuffin Jr WL, Sherman BM, Roth F, Gorden P, Kahn CR,
7. Guiot G, Rougerie J, Fourestler A, Fournier A, Comoy C, Vulmiere Roberts WC, et al. Acromegaly and cardiovascular disorders. A
J, et al. Une nouvelle technique endoscopique: exploration prospective study. Ann Intern Med. 1974;81(1):11–8.
endoscopiques intracraniennes. Presse Med. 1963;71:1225–8. 31. Sugihara N, Shimizu M, Kita Y, Shimizu K, Ino H, Miyamori I,
8. Couldwell WT, Weiss M. Pituitary macroadenomas. In: Apuzzo M, et al. Cardiac characteristics and postoperative courses in Cushing’s
editor. Brain surgery: complication avoidance and management. syndrome. Am J Cardiol. 1992;69(17):1475–80.
New York: Churchill Livingstone; 1993. p. 295–312. 32. Allen E. Thromboangitis obliterans: methods of diagnosis of
9. Davis DH, Laws Jr ER, Ilstrup DM, Speed JK, Caruso M, Shaw chronic occlusive arterial lesions to wrist with illustrative cases. Am
EG, et al. Results of surgical treatment for growth hormone- J Med Sci. 1929;178:237–44.
secreting pituitary adenomas. J Neurosurg. 1993;79(1):70–5. 33. Winstone NE, Brooke BN. Effects of steroid treatment on patients
10. Freda PU, Wardlaw SL, Post KD. Long-term endocrinological fol- undergoing operation. Lancet. 1961;1(7184):973–5.
low-up evaluation in 115 patients who underwent transsphenoidal 34. Spaziante R, de Divitiis E. Forced subarachnoid air in transsphenoi-
surgery for acromegaly. J Neurosurg. 1998;89(3):353–8. dal excision of pituitary tumors (pumping technique). J Neurosurg.
11. Laws Jr ER. Pituitary surgery. Endocrinol Metab Clin North Am. 1989;71(6):864–7.
1987;16(3):647–65. 35. Barrow D, Tindall G, Tindall S. Combined simultaneous transsphe-
12. Laws Jr ER, Piepgras DG, Randall RV, Abboud CF. Neurosurgical noidal transcranial operative approach to selected sellar tumors.
management of acromegaly. Results in 82 patients treated between Perspect Neurol Surg. 1992;3:49–57.
1972 and 1977. J Neurosurg. 1979;50(4):454–61. 36. Couldwell WT, Weiss M. The transnasal transsphenoidal approach.
13. Altay T, Krisht KM, Couldwell WT. Sellar and parasellar meta- In: Apuzzo M, editor. Surgery of the third ventricle. 2nd ed.
static tumors. Int J Surg Oncol. 2012;2012:647256. Philadelphia: Williams & Wilkins; 1998. p. 553–74.
14. Bills DC, Meyer FB, Laws Jr ER, Davis DH, Ebersold MJ, 37. Couldwell WT. Transsphenoidal and transcranial surgery for
Scheithauer BW, et al. A retrospective analysis of pituitary apo- pituitary adenomas. J Neurooncol. 2004;69(1–3):237–56.
plexy. Neurosurgery. 1993;33(4):602–8; discussion 608–9. 38. Tindall G, Barrow D. Tumors of the sellar and parasellar area in
15. Ebersold MJ, Laws Jr ER, Scheithauer BW, Randall RV. Pituitary adults. In: Youmans J, editor. Neurological surgery. Philadelphia:
apoplexy treated by transsphenoidal surgery. A clinicopathologi- WB Saunders; 1990. p. 3447.
cal and immunocytochemical study. J Neurosurg. 1983;58(3): 39. Kobayashi T, Nakane T, Kageyama N. Combined trans-sphenoidal
315–20. and intracranial surgery for craniopharyngioma. Prog Exp Tumor
16. Matjasko J. Perioperative management of patients with pituitary Res. 1987;30:341–9.
tumors. Semin Anesth. 1984;3:155–67. 40. Patterson RH. The role of transcranial surgery in the management
17. Ludecke DK, Lutz BS, Niedworok G. The choice of treatment after of pituitary adenoma. Acta Neurochir. 1996;65:16–7.
incomplete adenomectomy in acromegaly: proton–versus high 41. Wilson C. Neurosurgical management of large invasive pituitary
voltage radiation. Acta Neurochir (Wien). 1989;96(1–2):32–8. tumors. In: Tindall G, Collins W, editors. Clinical management of
18. Rush SC, Newall J. Pituitary adenoma: the efficacy of radiotherapy pituitary disorders. New York: Raven; 1979. p. 335–42.
as the sole treatment. Int J Radiat Oncol Biol Phys. 1989;17(1): 42. Laws Jr ER, Fode NC, Redmond MJ. Transsphenoidal surgery fol-
165–9. lowing unsuccessful prior therapy. An assessment of benefits and
19. Sheehan JP, Kondziolka D, Flickinger J, Lunsford LD. Radiosurgery risks in 158 patients. J Neurosurg. 1985;63(6):823–9.
for residual or recurrent nonfunctioning pituitary adenoma. 43. Hirsch O. Eine neue methode der endonasalen operation von
J Neurosurg. 2002;97(5 Suppl):408–14. hypophysentumoren. Wien Med Wochenschr. 1909;59:636.
20. Shin M, Kurita H, Sasaki T, Tago M, Morita A, Ueki K, et al. 44. Griffith HB, Veerapen R. A direct transnasal approach to the sphe-
Stereotactic radiosurgery for pituitary adenoma invading the cav- noid sinus. Technical note. J Neurosurg. 1987;66(1):140–2.
ernous sinus. J Neurosurg. 2000;93 (Suppl 3):2–5. 45. Laws Jr ER, Trautmann JC, Hollenhorst Jr RW. Transsphenoidal
21. Hirsch IB, McGill JB, Cryer PE, White PF. Perioperative manage- decompression of the optic nerve and chiasm. Visual results in 62
ment of surgical patients with diabetes mellitus. Anesthesiology. patients. J Neurosurg. 1977;46(6):717–22.
1991;74(2):346–59. 46. Laws Jr ER. Transsphenoidal approach to pituitary tumors. In:
22. Shucart WA, Jackson I. Management of diabetes insipidus in neu- Schmidek H, Sweet W, editors. Operative neurosurgical techniques.
rosurgical patients. J Neurosurg. 1976;44(1):65–71. Philadelphia: WB Saunders; 1996. p. 283
47. Cappabianca P, Alfieri A, Thermes S, Buonamassa S, de Divitiis 64. Kim SH, Huh R, Chang JW, Park YG, Chung SS. Gamma knife
E. Instruments for endoscopic endonasal transsphenoidal surgery. radiosurgery for functioning pituitary adenomas. Stereotact Funct
Neurosurgery. 1999;45(2):392–5; discussion 395–6. Neurosurg. 1999;72 Suppl 1:101–10.
48. Jho HD, Carrau RL, Ko Y, Daly MA. Endoscopic pituitary surgery: 65. Kobayashi T, Kida Y, Mori Y. Gamma knife radiosurgery in the
an early experience. Surg Neurol. 1997;47(3):213–22; discussion treatment of Cushing disease: long-term results. J Neurosurg. 2002;
222–3. 97(5 Suppl):422–8.
49. Alfieri A. Endoscopic endonasal transsphenoidal approach to the 66. Kondziolka D, Flickinger JC, Lunsford LD. Radiation therapy and
sellar region: technical evolution of the methodology and refine- radiosurgery of pituitary tumors. In: Krisht A, Tindall G, editors.
ment of a dedicated instrumentation. J Neurosurg Sci. 1999;43(2): Pituitary disorders: comprehensive management. Philadelphia:
85–92. Williams & Wilkins; 1999. p. 407–15.
50. Carrau RL, Jho HD, Ko Y. Transnasal-transsphenoidal endoscopic 67. Landolt AM, Lomax N. Gamma knife radiosurgery for prolactino-
surgery of the pituitary gland. Laryngoscope. 1996;106(7): mas. J Neurosurg. 2000;93 Suppl 3:14–8.
914–8. 68. Flickinger JC, Lunsford LD, Kondziolka D. Dose prescription and
51. Jankowski R, Auque J, Simon C, Marchal JC, Hepner H, Wayoff dose-volume effects in radiosurgery. Neurosurg Clin North Am.
M. Endoscopic pituitary tumor surgery. Laryngoscope. 1992;102(2): 1992;3(1):51–9.
198–202. 69. Littley MD, Shalet SM, Beardwell CG, Robinson EL, Sutton
52. Jho HD, Carrau RL. Endoscopic endonasal transsphenoidal sur- ML. Radiation-induced hypopituitarism is dose-dependent. Clin
gery: experience with 50 patients. J Neurosurg. 1997;87(1):44–51. Endocrinol (Oxf). 1989;31(3):363–73.
53. Sethi DS, Pillay PK. Endoscopic management of lesions of the sella 70. Brada M, Rajan B, Traish D, Ashley S, Holmes-Sellors PJ, Nussey
turcica. J Laryngol Otol. 1995;109(10):956–62. S, et al. The long-term efficacy of conservative surgery and radio-
54. Spencer WR, Das K, Nwagu C, Wenk E, Schaefer SD, Moscatello therapy in the control of pituitary adenomas. Clin Endocrinol (Oxf).
A, et al. Approaches to the sellar and parasellar region: anatomic 1993;38(6):571–8.
comparison of the microscope versus endoscope. Laryngoscope. 71. Morange-Ramos I, Regis J, Dufour H, Andrieu JM, Grisoli F,
1999;109(5):791–4. Jaquet P, et al. Short-term endocrinological results after gamma
55. Yaniv E, Rappaport ZH. Endoscopic transseptal transsphenoidal knife surgery of pituitary adenomas. Stereotact Funct Neurosurg.
surgery for pituitary tumors. Neurosurgery. 1997;40(5):944–6. 1998;70 Suppl 1:127–38.
56. Fraioli B, Esposito V, Santoro A, Iannetti G, Giuffre R, Cantore 72. Landolt AM, Haller D, Lomax N, Scheib S, Schubiger O, Siegfried
G. Transmaxillosphenoidal approach to tumors invading the medial J, et al. Stereotactic radiosurgery for recurrent surgically treated
compartment of the cavernous sinus. J Neurosurg. 1995;82(1):63–9. acromegaly: comparison with fractionated radiotherapy. J Neurosurg.
57. Sabit I, Schaefer SD, Couldwell WT. Extradural extranasal com- 1998;88(6):1002–8.
bined transmaxillary transsphenoidal approach to the cavernous 73. Yoon SC, Suh TS, Jang HS, Chung SM, Kim YS, Ryu MR, et al.
sinus: a minimally invasive microsurgical model. Laryngoscope. Clinical results of 24 pituitary macroadenomas with linac-based stereo-
2000;110(2 Pt 1):286–91. tactic radiosurgery. Int J Radiat Oncol Biol Phys. 1998;41(4):849–53.
58. Weiss M. Transnasal transsphenoidal approach. In: Apuzzo M, edi- 74. Pan L, Zhang N, Wang EM, Wang BJ, Dai JZ, Cai PW. Gamma
tor. Surgery of the third ventricle. Philadelphia: Williams & knife radiosurgery as a primary treatment for prolactinomas.
Wilkins; 1987. p. 476–94. J Neurosurg. 2000;93 Suppl 3:10–3.
59. Kouri JG, Chen MY, Watson JC, Oldfield EH. Resection of supra- 75. Wowra B, Stummer W. Efficacy of gamma knife radiosurgery for
sellar tumors by using a modified transsphenoidal approach. Report nonfunctioning pituitary adenomas: a quantitative follow up
of four cases. J Neurosurg. 2000;92(6):1028–35. with magnetic resonance imaging-based volumetric analysis.
60. Mason RB, Nieman LK, Doppman JL, Oldfield EH. Selective exci- J Neurosurg. 2002;97(5 Suppl):429–32.
sion of adenomas originating in or extending into the pituitary stalk 76. Boelaert K, Gittoes NJ. Radiotherapy for non-functioning pituitary
with preservation of pituitary function. J Neurosurg. 1997;87(3): adenomas. Eur J Endocrinol. 2001;144(6):569–75.
343–51. 77. Couldwell WT, Rosenow JM, Rovit RL, Benzil DL. Hypophysopexy
61. Yasargil M. Microneurosurgery. In: Microsurgical anatomy of the technique for radiosurgical treatment of cavernous sinus pituitary
basal cisterns and vessels of the brain, diagnostic studies, general adenoma. Pituitary. 2002;5(3):169–73.
operative techniques, and pathological considerations of the intra- 78. Taussky P, Kalra R, Coppens J, Mohebali J, Jensen R, Couldwell
cranial aneurysms, vol I. Stuttgart: Georg Thieme Verlag; 1984. WT. Endocrinological outcome after pituitary transposition
62. Izawa M, Hayashi M, Nakaya K, Satoh H, Ochiai T, Hori T, et al. (hypophysopexy) and adjuvant radiotherapy for tumors involving
Gamma knife radiosurgery for pituitary adenomas. J Neurosurg. the cavernous sinus. J Neurosurg. 2011;115(1):55–62.
2000;93 Suppl 3:19–22. 79. Liu JK, Schmidt MH, MacDonald JD, Jensen RL, Couldwell
63. Kim MS, Lee SI, Sim JH. Gamma knife radiosurgery for function- WT. Hypophysial transposition (hypophysopexy) for radiosurgical
ing pituitary microadenoma. Stereotact Funct Neurosurg. 1999;72 treatment of pituitary tumors involving the cavernous sinus.
Suppl 1:119–24. Technical note. Neurosurg Focus. 2003;14(5):e11.
Pituitary and Pituitary Region Tumors:
Viewpoint—Fractionated Radiation 28
Therapy
Jonathan P.S. Knisely and Paul W. Sperduto
A newer goal may be the avoidance of delivery of radiation

Introduction to the hippocampal structures, to avoid impairing their func-
tions in learning and memory. Radiation may be delivered in
Fractionated radiotherapy for pituitary tumors and tumors a single radiosurgical dose or a variable number of smaller
arising in the parasellar region has been used for over a cen- doses. Radiation is frequently employed for incompletely
tury. The multidisciplinary nature of optimal management of resected tumors, for recurrent tumors, and as a primary treat-
tumors of the anterior skull base has been clearly established; ment for medically inoperable patients and for patients who
the expertise of endocrinologists, neurosurgeons, radiation refuse a surgical intervention.
oncologists, neuroradiologists, neuro-ophthalmologists, and This chapter will primarily review the use of convention-
laboratory medicine specialists is commonly required to ally fractionated photon radiation therapy (25–30 weekday
optimally manage tumors that arise in the pituitary or pitu- treatments, given over a 5- to 6-week time period, in doses of
itary region. The skull base structures surrounding the pitu- approximately 1.8 Gy/day) for pituitary adenomas, meningi-
itary gland are anatomically complex and include a number omas, and craniopharyngiomas, all benign anterior skull base
of important vascular structures and cranial nerves [1], which tumors arising in the sella and parasellar tissues. Many of the
can present significant challenges to the surgical manage- management principles and concerns are largely the same for
ment of pituitary region tumors. Complete resections with other benign or for malignant tumors arising in this area.
acceptable operative morbidity can be difficult to achieve.
Iatrogenic damage may occur during the treatment of an
anterior skull base tumor that may result in endocrinologic, Epidemiology and Histology
visual, or other neurologic deficits. Radiation can induce
second tumors, accelerate atherogenesis, and cause endo- Pituitary Adenomas
crine insufficiencies. It is easy to appreciate the potential of
iatrogenic damage to decrease the quality of life for patients Pituitary adenomas represent approximately 10–12 % of all
that may live for decades after an intervention. adult primary intracranial neoplasms; about half of these
Treatment options include the use of microsurgery, medi- pituitary adenomas are hormonally active [2]. Several
cal therapy, and irradiation. Therapeutic goals include the autopsy series have detected pituitary adenomas in as many
destruction of a tumor or control of its growth, controlling as 22–25 % of cases [3, 4]. Equal numbers of men and
hormonal hypersecretion (if present), and the restoration of women develop macroadenomas, and the incidence peaks in
vision or lost function without injuring surrounding normal the third and fourth decades of life. Only 10 % of cases arise
neurovascular tissues or inducing pituitary hypofunction. in children or adolescents. Individuals with multiple endo-
crine neoplasia type I have a hereditarily increased risk of
J.P.S. Knisely, M.D. developing a pituitary adenoma, and kindreds with an
Department of Radiation Medicine, North Shore LIJ Health increased risk of pituitary adenomas without other endocrine
System, Hofstra North Shore LIJ School of Medicine, organ neoplasia have also been described.
300 Community Drive, Manhasset, NY 11030, USA
Pituitary adenomas are classified as either functioning
P.W. Sperduto, M.D., M.P.P., F.A.S.T.R.O. (*) (endocrine active) or nonfunctioning (endocrine inactive),
Minneapolis Radiation Oncology & Gamma Knife Center,
with functioning tumors subclassified by the endocrine axis or
University of Minnesota, 560 South Maple Street, Suite 10,
Waconia, MN 55391, USA axes upon which they act; nonfunctioning tumors are usually
e-mail: psperduto@mropa.com classified as either oncocytic or non-oncocytic (Table 28.1).
392 J.P.S. Knisely and P.W. Sperduto
Table 28.1 Pituitary tumor classification recurrence rate that is very low, even if bone adjacent to the
Pituitary adenoma tumor is invaded.
cell type Hormone product Clinical syndrome
Corticotroph ACTH Cushing’s disease
Lactotroph Prolactin Galactorrhea/ Craniopharyngiomas
amenorrhea/impotence
Somatotroph Growth hormone Acromegaly, pituitary
Craniopharyngiomas are rare, representing only a few percent
gigantism
Somatotroph/ Growth hormone Acromegaly and
of all intracranial primary tumors [10]. Craniopharyngiomas
lactotroph and prolactin galactorrhea/amenorrhea arise from either embryonic remnants of an incompletely
Thyrotroph Thyrotropin Hyperthyroidism involuted hypophyseal-pharyngeal duct or from metaplastic
Gonadotroph FSH, LH Hypopituitarism ectodermal cells in the anterior hypophysis [11]. They account
Mixed cell None Hypopituitarism for a significant portion of the intracranial tumors of child-
(non-oncocytic) hood but are as common in adults as in children and adoles-
Oncocytic None Hypopituitarism cents [12]. The age distribution shows a bimodal pattern, with
FSH follicle-stimulating hormone, LH luteinizing hormone a first larger peak between the ages of 5 and 10 years and a
second less prominent peak in the fifth decade, and there is a
The detection of these hormones may be possible by testing very slight male preponderance reported in some large series.
peripheral blood or by doing immunohistochemical staining No recognized hereditary syndromes are associated with
of the tumor. craniopharyngiomas.
Two major histopathologic categories of craniopharyngi-
oma are recognized. There is evidence that they may have
Meningiomas different etiologies [10]. The adamantinomatous variant is
more commonly seen in pediatric patients, and papillary
Meningiomas represent approximately 18–20 % of all pri- squamous tumors are more commonly seen in adults and are
mary brain tumors. Twenty-five percent to 30 % of menin- apparently more likely to be durably controlled by surgery
giomas arise along the base of the anterior and middle alone. There is apparently some occasional intergrading of
cranial fossae in locations such as the cavernous sinus, craniopharyngioma tumor types [12].
tuberculum sellae, Meckel’s cave, or sphenoid wing [5].
The incidence of intracranial meningioma increases with
increasing age, and women are more commonly affected Clinical Presentation
than men by a ratio of approximately 2:1. Meningiomas
are more common in individuals with neurofibromatosis Pituitary Adenomas
type 2, in individuals who have a remote history of cranial
irradiation, and in patients with pulmonary lymphangi- Endocrine abnormalities are the most common presentation
oleiomyomatosis [6]. of pituitary adenomas. Functioning adenomas will manifest
Histopathologically, resected meningiomas from the skull themselves by causing one of the classic endocrine syn-
base should be graded as per the 2007 World Health dromes (Table 28.1). Hormonal insufficiency may arise from
Organization (WHO) guidelines [7]. If there are ≥4 mitoses compression of the normal pituitary gland or of the pituitary
per ten high-powered microscopy fields (HPF) or at least stalk. Insufficiencies may be more subtle than hypersecretory
three features believed to correspond with aggressiveness are states, but either can develop insidiously and only be discov-
seen—small cells, macronucleoli, hypercellularity, necrosis, ered after a protracted period of symptoms. Amenorrhea in
and sheeting architecture—a meningioma is deemed atypical young women and a decreased libido in men may be the signs
(grade II). Brain invasion is an indicator of grade II rather and signals that trigger an evaluation of the hypothalamic-
than grade III behavior in the absence of any histologic ana- pituitary axis.
plasia. A grade III meningioma must have frank anaplasia or Endocrine-active adenomas are more commonly diag-
a mitotic rate of >20/10 HPF. Four variants are considered nosed as microadenomas (<1.0 cm in diameter) than as mac-
innately more aggressive by definition: clear cell (grade II), roadenomas (≥1.0 cm in diameter), but somatotroph or
chordoid (grade II), papillary (grade III), and rhabdoid (grade lactotroph macroadenomas are common. Lateral extension
III). Grade II meningiomas recur at a rate of ~40% at 5 years of pituitary tumors into the cavernous sinus can cause diplo-
time, and grade III meningiomas have a median overall pia or facial sensory symptoms. Superior extension will
survival of <2 years [8, 9]. Grade I tumors have a short-term cause pressure on the optic nerves and chiasm, with a resul-
28 Pituitary and Pituitary Region Tumors: Viewpoint—Fractionated Radiation Therapy 393
tant bitemporal superior quadrantanopsia or bitemporal Table 28.2 Diagnostic evaluation of patient with a sellar or parasellar
hemianopsia. Frontal headaches may be caused by irritation tumor
of meningeal sensory fibers as the adenoma enlarges. History and physical examination with particular attention to signs,
symptoms, and stigmata from hormonal hypersecretion or
hyposecretion and a detailed cranial nerve examination
Laboratory endocrine evaluation (for appropriate patients with
Meningiomas apparent endocrinologic problems, based on clinical assessment, or if
imaging reveals a tumor arising in or extending to the pituitary sella
The presenting symptoms of meningiomas may be diverse in or suprasellar-hypothalamic region)
these locations but are most commonly from mass effect Provocative testing of apparently affected axes, with assessment of:
upon adjacent normal tissues. The exact symptoms will • Serum prolactin level
depend upon what structures are compressed. The clinical • Fasting growth hormone level
• IGF-1 level and growth hormone dynamic testing with insulin
symptomatology may not be as distinctive as for a functional
tolerance, glucose suppression testing, and thyrotropin-
pituitary adenoma, and high-quality imaging is frequently releasing hormone testing
required to determine whether a patient may have a pituitary • Serum ACTH, 24-h urinary free cortisol and
adenoma, a meningioma, a schwannoma, or other regional 17-hydroxycorticosteroid levels, and response
pathology such as an aneurysm, sarcoid, or another condi- to dexamethasone suppression testing
tion. Involvement of the cavernous sinus will commonly • Gonadal function testing with serum LH, FSH levels,
and plasma estradiol and testosterone levels
cause diplopia or facial sensory symptoms; a meningioma
• Thyrotropin (TSH), thyroxine, and triiodothyronine levels
arising from the optic nerve sheath will cause unusual visual
• Complete blood count (CBC) and coagulation studies
symptoms. Radiological evaluation including an MRI scan with coronal thin cuts
through the sphenoid bone with and without gadolinium contrast
and possibly also including CT imaging to evaluate bony changes.
Craniopharyngiomas Specialized neurointerventional imaging and petrosal venous sinus
sampling may be required in cases of Cushing disease with a
radiographically occult microadenoma
These tumors most commonly are diagnosed as a result of Neuro-ophthalmologic examination for patients with tumors that
their mass effect upon the optic nerves or from mass effect impinge upon the optic apparatus or cause cranial nerve abnormalities
upon normal hypothalamic or pituitary function. As these are affecting vision
generally slow-growing tumors, an insidious loss of function
may occur, so that the degree of compensation for loss of
vision or other deficits may be remarkable at the time of For tumors that present with mass effect, surgical debulk-
diagnosis. Increased intracranial pressure may lead to head- ing of the tumor may be needed promptly. A neurosurgeon
aches and projectile vomiting. The mass effect in adults usu- skilled in skull base and pituitary surgery and knowledgeable
ally results in amenorrhea for women and in loss of libido for about the use of fractionated irradiation and stereotactic
men, and if the tumor extends far enough into the cranial radiosurgery should be consulted. This surgeon should be
vault, hypothalamic, frontal, or temporal lobe symptoms aware of the need to establish a distance between the rostral
may develop. Diabetes insipidus, obesity, and neurocogni- surface of the tumor and the optic nerves.
tive or developmental abnormalities may result from a High-dose glucocorticoids may provide temporary
craniopharyngioma. improvement of mass effect while additional evaluations are
performed. An endocrinologic evaluation can determine
whether there are hormonal abnormalities that need to be
Diagnostic Workup and Staging addressed before surgery and may also reveal that medical
management is possible as an initial therapy.
All patients with tumors arising in this region should be Discussion of each case with a neuroradiologist when
managed with a multidisciplinary approach. Not every spe- imaging studies are being ordered will facilitate obtaining
cialist needs to see every case, but with a team of specialists appropriately detailed studies. A dedicated pituitary mag-
who are comfortable with multidisciplinary care and who are netic resonance imaging (MRI) scan with and without
aware of available therapeutic options, there will be less of a gadolinium contrast administration will show details of
chance of overlooking tests or therapies that could help the relation of the tumor to the optic nerves, chiasm, and
develop individualized safe and effective treatment plans tracts and other critical structures that would affect radia-
(Table 28.2). tion management options being considered. Pulse
sequences that suppress the bright signal from fat can
clarify whether or not enhancing tumor extends into the irradiation, stereotactic radiosurgery, and instillation of
orbit. Computed tomography (CT) imaging may show radioisotope colloids into cystic tumors.
bone remodeling.
A neuro-ophthalmologic consultation is often warranted
to evaluate the patient for subtle funduscopic abnormalities Usual Therapeutic Approaches
or tumor-related vision changes. Serial perimetry is able to
follow visual fields over time in a way that clinical exami- Radiation therapy is usually deferred until after any oper-
nation cannot. Clinical evidence of injury to the cranial ative neurosurgical management has been completed.
nerves controlling extraocular motion should be docu- This is because surgery is able to address both mass effect
mented by a careful observer; improvement is often seen (if present) as well as hormonal hypersecretion (if the
after irradiation. tumor in question is a hormonally active pituitary ade-
When a patient appears to have long-standing neurocog- noma). For some pituitary adenomas (prolactinomas,
nitive issues, it is worth the effort to have the patient evalu- growth hormone-secreting adenomas, and thyroid hor-
ated with neurocognitive testing appropriate to the clinical mone-secreting adenomas), it is possible to use pharma-
situation. This can help identify learning issues that may be cologic management with specific action at the
addressable with interventions or therapy. hypothalamic-pituitary axis and avoid or defer surgical
A radiation oncologist should also be asked to evaluate therapy or radiation therapy [19–21]. Neither meningio-
the patient so that the strengths and shortcomings of different mas nor craniopharyngiomas have systemic management
radiation management strategies may be presented to arrive options similar to those for functioning pituitary adeno-
at a satisfactory management approach. mas. Meningiomas may be detected incidentally and
Neurosurgical staging systems for anterior skull base found to be without clinical signs or symptoms. As these
tumors have little applicability to radiation management tumors are generally slow growing, documenting the rate
decisions [13–15]. The critical issues for radiation manage- of growth prior to pursuing definitive interventions may
ment decisions for pituitary adenomas and all tumors in the be useful, as no intervention is without risk. Some menin-
sellar and parasellar area are the proximity of the tumor to giomas and most craniopharyngiomas are detected
the optic apparatus and the size of the tumor. Pituitary adeno- because of clinical signs or symptoms from the tumor’s
mas and other tumors that are large or that very closely presence at the skull base. In some of these cases, a crani-
approach the optic nerves or chiasm are not appropriate for otomy may be needed to try to alleviate mass effect and
stereotactic radiosurgery. Fractionated irradiation can be bring about a recovery of function, but it may not be
used for all tumors in this location. Not surprisingly, tumor appropriate for all cases, when radiation management
volume has been shown to be inversely correlated with long- options exist that are likely to be highly effective.
term control in patients with pituitary adenomas that are After an initial therapeutic intervention (usually surgery),
treated with fractionated irradiation [16]. it is common to repeat many elements of the battery of tests
For meningiomas, the most common staging system that were performed at the time of diagnosis to document any
was developed over 50 years ago to provide some guid- changes in neurologic, ophthalmologic, or endocrinologic
ance as to the risk of recurrence after surgical manage- function (Table 28.1). Documentation of improvements and
ment [15]. Molecular markers may be assessed to help deficits will allow them to be accurately attributed to the
predict the risk for recurrence after surgical resection [17, effects of the tumor, to the initial intervention, or to subse-
18]. Radiotherapy’s value in controlling meningioma quent therapy.
growth is broadly recognized. If a meningioma is less If there is consideration being given to the use of stereo-
than 5 mm from any part of the optic nerves or chiasm, tactic radiosurgery for a tumor that preoperatively was docu-
safe management with radiosurgery is extremely chal- mented to be very close to the optic nerves, volumetric
lenging and in the eyes of many experts carries an unac- contrast-enhanced MRI must be repeated to ensure that the
ceptable risk of radiation optic neuritis, although tumor has been adequately debulked away from the optic
fractionated radiation can be used. apparatus. This geometric separation of several millimeters
There is no accepted staging system for craniopharyn- between the tumor and the optic apparatus is not as critical if
giomas, which are frequently densely adherent to adjacent fractionated irradiation will be used as a postsurgical ther-
normal tissues. Aggressive surgical resections may be apy. The optic nerves will tolerate fractionated doses of radi-
complicated by intraoperative damage to normal tissues ation to ~50–55 Gy used to durably control most tumors in
involved with the tumor. Irradiation may also cause mor- this area, while a radiation dose >8 to 10 Gy given to the
bidity but generally is not as likely to cause an acute loss optic nerves in a single fraction will cause vision loss in an
of function. Radiation options include fractionated unacceptably high number of patients [22–25].
Simulation is the process of determining the geometric

Radiation Therapy factors relating to fractionated irradiation delivery, and treat-
ment planning uses information about the radiation beams
The role of radiotherapy in the management of anterior skull and these geometric parameters to generate data on radiation
base tumors has waxed and waned over the past century. doses to the tumor and normal tissues. The first step in this
Radiotherapy treatment for pituitary adenomas and parasel- process is to fabricate a patient-immobilizing device that will
lar tumors is currently delivered with a very high degree of be used for all subsequent treatments. A thermoplastic mask
precision, is extremely well tolerated, and provides excellent is commonly used for this purpose, together with a baseplate
outcomes by delivering a cytostatic or cytocidal dose of radi- that holds the patient’s head in a position that tucks the chin
ation to the tumor while not exceeding the radiation toler- to the chest. Many centers use baseplates that have the ability
ance of adjacent normal tissues. to be set at a predetermined angle for reproducibly achieving
Current radiation therapy techniques are greatly improved a high degree of chin tuck. A well-constructed mask and
relative to those of earlier eras. Superior resolution on imaging baseplate pair should provide setup accuracies of ~2 mm.
of normal anatomy and pathology is a major component of More elaborate immobilization techniques that incorporate
this, as tumors may be diagnosed earlier in their natural his- custom-molded bite blocks and custom-molded posterior
tory. Modern treatment techniques reproducibly immobilize head supports can provide immobilization accuracy that
patients to achieve setup accuracies of approximately 2 mm, approaches a millimeter or less [28]. Rigid and reproducible
and radiation fields may be shaped to match tumor geometries immobilization techniques are mandatory for the most con-
and include less normal tissues. Less normal tissue irradiation formal radiation therapy delivery techniques. The position of
will contribute to lower rates of late endocrine hypofunction the patient’s head at the time of the fabrication of this device
and other normal tissue damage. Temporal lobe damage has will determine whether simple vertical beams or arcs can be
been described as a complication associated with historical used with the couch in a neutral position. Inattention to this
radiation therapy techniques [26]. The use of higher-energy point may reduce the ease of daily treatment or prevent some
photons and multiple treatment portals to treat centrally well-characterized and simple treatment techniques from
located tumors should decrease the radiation dose delivered to being easily adopted.
the temporal lobes. Using more radiation portals may attenu- For pituitary radiation therapy, a traditional simulation
ate late neurocognitive sequelae, but the possibility exists that method uses a fluoroscopic apparatus with the same geome-
late radiation-induced tumorigenesis may be increased with try as a linear accelerator to set up an anterior and two lateral
treatment techniques that increase the volume of normal tis- radiation portals of approximately 4 × 4 cm, centered just
sues exposed to low doses of irradiation; useful data regarding above the pituitary fossa. The immobilizing mask can be
incidence rates and dose-volume histograms that might pro- marked with the entry points, central axes, and corners for
vide clinical guidance regarding this particular complication these fields. A contour of the skull is obtained along the line
are essentially nonexistent [27]. Despite these advances, if a that joins the central points of these fields. A dosimetrist can
tumor is not clearly separated by several mm from any part of use this contour information and X-ray beam data to calcu-
the optic nerves or chiasm, it is generally safer to deliver frac- late radiation parameters with computerized treatment plan-
tionated radiation therapy than radiosurgery. ning software. Obviously, the location of the radiation portals
and their size depends upon the size and location of the
tumor; the size of smaller portals is frequently limited by a
Simulation and Radiotherapy Planning physicist’s ability to accurately measure the beam output
through a small and irregular aperture.
Before radiation treatment can commence, preparatory steps CT simulation simplifies some aspects of this process.
must occur, and the exact nature of these steps will vary The patient needs to be immobilized in the treatment posi-
slightly with the equipment being used to treat the patient. tion, and setup points need to be marked on the mask for later
Radiotherapy treatment techniques have evolved from a dis- reference. A noncontrast CT scan with fine cuts (1.25–
tant era where the distance between the patient’s right and 2.5 mm thick, extending from the vertex of the head to the
left temple along the central axis of a pair of opposing X-ray foramen magnum) is obtained. Contouring (identification of
beams aimed at the presumed location of the tumor was the structures in three-dimensional space in the imaging study)
critical factor involved in setting up, planning, and deliver- needs to be done for all the important tissues, particularly if
ing radiation therapy. Current sophisticated radiotherapy intensity-modulated radiation therapy (IMRT) techniques
techniques use meticulous care in a series of steps before will be used to treat the patient. High-resolution treatment
any treatment is given to ensure that treatment will be deliv- planning CT studies generally show the pituitary fossa, optic
ered to the tumor being targeted only as prescribed by the nerves, and lenses well enough that these structures can be
radiation oncologist. clearly defined. The radiation therapy portals are set up on
digitally reconstructed radiographs after the contouring pro- incomplete coverage of intracranial target tissue (“cold
cess has been completed, and from this information, a spots”) that can lead to tumor recurrences [29, 30].
dosimetrist can directly use this digital data to calculate the Radiotherapy technology advances that permit more than a
beam weights and dosimetric parameters. few coplanar fields to be used to treat a tumor have been used
The use of image fusion software to register a high-quality by researchers to determine an optimal number of radiation
MRI study to the treatment planning CT scan may help in beams for treatment of skull base tumors and to conformally
delineation of the tumor and adjacent normal tissues. Image avoid unnecessary irradiation of normal tissues such as the
fusion software rigidly registers the two imaging studies hippocampi [31–33].
through the use of normalized mutual information, chamfer Treatments that are three-dimensionally conformal to a radi-
matching, or other algorithms and then reformats and redis- ation target can be routinely delivered with a high degree of
plays the voxels from the MRI in the CT scan’s geometric precision. This has been termed 3DCRT (three-dimensional
space. The MRI scan does not need to be done with the conformal radiation therapy). The use of rigid, reproducible
patient in the immobilization mask. Structures better seen on immobilization and a large number of radiation beams or of arc-
an MRI, like the hippocampi, can be contoured on the MRI ing beams has led some to term this approach fractionated ste-
and can be transferred to the CT for calculation purposes. reotactic radiation therapy (FSRT), although this terminology is
Clinicians have realized from more detailed dosimetric inaccurate in that true stereotaxy (in a neurosurgical sense) is
analyses that there can be areas of unacceptably high dose not achieved [34]. An example of an FSRT dosimetric plan for
(“hot spots”) that may injure normal tissues or areas of a patient with a pituitary adenoma is shown in Fig. 28.1a–c.
Fig. 28.1 (a) Axial fractionated stereotactic radiation therapy (FSRT) develop a more conformal FSRT plan than an IMRT plan. Tissues outside
dose distribution. (b) Coronal FSRT dose distribution. (c) Sagittal FSRT the target volume are treated to higher doses in the IMRT plan than in the
dose distribution. (d) Axial intensity-modulated radiation therapy (IMRT) FSRT plan. Both plans would be acceptable for fractionated irradiation
dose distribution. (e) Coronal IMRT dose distribution. (f) Sagittal IMRT using conventional fractionation, but neither could be accepted for radio-
dose distribution. It is clear from these dose distributions (shown on MR surgical treatment, because the tolerance of the anterior visual pathways
images) that for this patient with a pituitary adenoma, it was possible to would be exceeded by doses that would control the adenoma
Software and hardware technology to dynamically con- Table 28.3 Results of radiation therapy for nonfunctional adenomas
trol the point-to-point intensity of incident radiotherapy No. of Results
beams has been developed and is known as IMRT. Radiation Reference patients Treatment End point Duration (%)
therapy plans generated and delivered using IMRT tech- Breen et al. [47] 120 S + RT, RT LC 10 y, 88;
niques are significantly different than plans developed in the 20 y, 78;
past. For an IMRT plan to be generated, a computer must be 30 y, 65
Tsang et al. [16] 160 S + RT LC 10 y, 87
told what the overall goals of the plan should be. These
Grigsby et al. [52] 111 S + RT, RT DFS 13 y, 84
include limitations of doses to certain structures, acceptable
Brada et al. [37] 252 S + RT, RT PFS 10 y, 97;
ranges of doses to other structures, and the basic parameters 20 y, 92
of beam energy and beam entry points. The computer soft- Woollons et al. [43] 50 S + RT PFS 5 y, 72
ware then selects radiation beam intensities that can achieve Park et al. [49] 44 S + RT PFS 5 y, 97.7;
the desired therapeutic goals. 10 y, 97.7
Better integration of these software and hardware advances Gittoes et al. [53] 66 S + RT PFS 15 y, 93
and faster computer microprocessors has allowed the use of Sasaki et al. [54] 65 S + RT LC 10 y, 98
changes in radiation beam intensity in multiple portals to Grabenbauer 50 S + RT LC 54 mo, 94 %
et al. [42]
achieve better radiation dose distributions, with fewer hot or
S surgery, RT radiation therapy, DFS disease-free survival, PFS pro-
cold spots inside the target volume and no hot spots outside the
gression-free survival, LC local control
target volume. These technological advances may decrease
treatment-related morbidity and improve treatment results
overall, but this has not yet been demonstrated to be the case Patients must understand the need of long-term close surveil-
for most intracranial tumors, including pituitary adenomas. An lance if radiation therapy is not employed as a routine
example of an IMRT dosimetric plan for a patient with a pitu- adjunctive therapy [48, 49].
itary adenoma is shown in Fig. 28.1d–f. Radiotherapy alone achieves control rates of approxi-
Proton beam therapy has been gradually increasing in mately 80–95 % at 10-year follow-up for recurrent nonfunc-
availability in some markets and there are now beginning to tional adenomas or for primary treatment of nonfunctional
be small series reported in the medical literature of patients adenomas [16, 37, 47]. Longer-term follow-up has docu-
treated with protons for pituitary adenomas, craniopharyn- mented continued progression up to >20 years from therapy,
giomas, and meningiomas. Modeling studies indicate that documenting the need for lifelong surveillance of patients
there may be significant benefit to the use of protons, but this with pituitary adenomas [50, 51].
has not yet been demonstrated in clinical series [35, 36]. A critical factor in the long-term local control after post-
operative irradiation of pituitary adenomas may well be the
postoperative volume of tumor. This may conceivably range
Radiation Therapy Results from a small amount of tumor remaining in a cavernous
sinus to a macroscopic suprasellar extension into the mid-
Pituitary Macroadenomas brain. Local control can now be more accurately determined
than in the past because of improvements in imaging tech-
Fractionated doses of approximately 45–50 Gy in daily nology and availability. Side-by-side comparisons can be
doses of 1.8 Gy are currently used for treating pituitary ade- made between studies obtained a decade or more apart, and
nomas at most centers because of the knowledge that has subtle changes in size can be detected. In addition to the pre-
been gained over the past decades about the low risk of caus- vention of adenoma growth, many radiotherapy series have
ing visual complications when these guidelines are followed, reported an improvement in mass effect and mass effect-
the dose response seen for doses lower than this range, and related symptoms such as vision, but neurosurgical manage-
the apparent absence of improved results and increasing rate ment is much more effective in promptly relieving mass
of visual and other complications when higher total doses effect. Several reports on fractionated irradiation for non-
and plans with hot spots are used [16, 37–42]. functioning pituitary adenomas are collected in Table 28.3
For patients who have undergone a surgical resection of a [16, 37, 42, 43, 47, 49, 52–54].
pituitary adenoma, long-term local control is improved when
radiation treatment is administered before a macroscopic
recurrence of tumor develops, though a patient with a com- Growth Hormone-Secreting Adenomas
plete resection of tumor may not require postoperative irra- (Acromegaly)
diation [16, 43–45]. The rates of recurrence are not at all
trivial in many series, and there is a very significant potential Acromegaly is still a common indication for fractionated
for progression at 5–10 years after surgical resection [46, 47]. irradiation. These tumors are frequently large, and as a result,
surgical debulking of suprasellar extension may be inade- management possibly used to help decrease tumor bulk
quate to permit consideration of stereotactic radiosurgery before commencing irradiation and also to help suppress
because the tumor may still be in close juxtaposition to the hormonal excess after irradiation while waiting for the even-
optic apparatus. There are some who believe acromegaly is tual declines that are expected. As there is some data that
associated with a higher risk of radiation optic neuropathy hormonal suppression prior to radiosurgery may depress the
than for nonfunctional adenomas, with a cited risk of as high ultimate control rates for some functional pituitary adeno-
as 1.4 % with conventional treatment techniques [25, 50]. A mas, these agents may need to be evaluated for any protec-
large tumor mass and high pretreatment growth hormone tive effect in the setting of conventionally fractionated and
levels may predict a worse prognosis for patients with acro- delivered radiotherapy [60]. There is little doubt that the use
megaly [45, 55]. of medical therapies can help patients while awaiting the
Older literature on the effectiveness of radiation therapy effects of radiotherapy [61].
for acromegaly cannot provide all the data that is believed Adequate follow-up has been obtained in several series
necessary to evaluate the effectiveness of the therapy. The of patients with growth hormone-secreting pituitary ade-
definitions of normal have been lowered with more precise nomas that document a progressive response over time
laboratory testing over the past several decades. Current [38, 62–65]. The completeness of surgical resection
guidelines on the definition of cure include a random growth appears to contribute to the long-term results achievable
hormone <1 μ[mu]g/L, or <0.4 μ[mu]g/L after an with radiation therapy [63, 65]. The results of FSRT are
oral glucose-tolerance test, and insulin-like growth factor 1 only now becoming available; further follow-up will doc-
(IGF-1) levels that are within age-adjusted normative ranges ument whether or not superior results are possible with
[56]. There is data that the IGF-1 levels after therapy may the technological advances that have been used to treat
correlate better with cure than GH values [57]. these patients [66]. The results of several series of acro-
Modern multidisciplinary management for many patients megalic patients treated with fractionated irradiation are
will include medical management with somatostatin receptor listed in Table 28.4 [38, 55, 61–67].
ligands such as octreotide or lanreotide. Growth hormone
receptor antagonists such as pegvisomant are generally
reserved for patients who do not achieve biochemical control Prolactin-Secreting Adenomas
with somatostatin receptor ligands [58]. The use of such
medical management strategies may alter the rate at which Prolactinomas are less commonly treated with irradiation
conventionally delivered radiation can achieve durable con- at present, and the role of surgical intervention in this dis-
trol of hormonal excess and may provide additional opportu- order may also be declining with the availability of rea-
nities for improving the therapeutic ratio by effecting sonably well-tolerated medications that can be taken for
regressions in tumor bulk prior to commencing definitive protracted periods of time. Historical series show that for
irradiation [59]. patients treated with irradiation, there is a decreased rate
Radiotherapy should be considered for all patients with a of durable control without the addition of medical therapy
postoperative basal GH level of >5 μ[mu]g/L, with medical but that over time, the proportion of patients with hor-
Table 28.4 Results of radiation therapy for growth hormone-secreting adenomas

Results
Reference No. of patients Treatment End point Duration (%)
Eastman et al. [62] 87 S + RT, RT GH <5 μ[mu]g/L 5 y, 30; 10 y, 53; 15 y, 77; 20 y, 89
Tsang et al. [38] 52 S + RT, RT GH <10 μ[mu]g/L 10 y, 46
Milker-Zabel et al. [66] 20 S + RT, RT “Normalized GH level”; 26 mo, 80; 25 mo, 45
“normalized IGF-1 level”
Biermasz et al. [64] 40 S + RT GH <5 μ[mu]g/L; “normalized 5 y, 75; 10 y, 76; 15 y, 87.
IGF-1 level” Last follow-up, 73
Thalassinos et al. [65] 46 RT GH <5 μ[mu]g/L 5 y, 30; 10 y, 32; >10 y, 55
Dowsett et al. [55] 15 S + RT GH <5 μ[mu]g/L 3–5 y, 25; 5–10 y, 29; >10 y, 100 %
Minniti et al. [63]
Roug et al. [61] 34 S + RT + M GH <1 μ[mu]g/L; IGF-1 < +2 SD 1 y, 24; 3 y, 38; 5 y, 64
González et al. [67] 40 S + RT + M GH <1 μ[mu]g/L; “normalized 5 y, 46; 10 y, 57; 5 y, 36; 10 y, 43
IGF-1 level”
S surgery, RT radiation therapy, GH growth hormone, IGF insulin-like growth factor 1, M medical therapy, SD standard deviation
Table 28.5 Results of radiation therapy for prolactinomas

Results
Ozgen et al. [69] 106 S + RT PRL <30 μ[mu]g/L Not stated, 58
Tsang et al. [38] 64 S + RT PRL <20 μ[mu]g/L 10 y, 25
Wallace and Holdaway [71] 25 S + RT Normal PRL 4 y, 33
Tsagarakis et al. [72] 36 RT PRL <360 mU/L 8 y, 50
Williams et al. [68] 28 RT Normal PRL Not stated, 29
Littley et al. [70] 58 S + RT, RT PRL <500 mU/L 10 y, 50 % (predicted)
Sun et al. [73] 7 S + RT PRL <20 ng/mL 21 mo, 100
PRL prolactin, S surgery, RT radiation therapy
Table 28.6 Results of radiation therapy for Cushing’s disease

Results
Howlett et al. [80] 21 RT Normal mean serum cortisol 9.5 y, 57
Tsang et al. [38] 29 S + RT, RT Urinary free cortisol (UFC) <220 nmol/24 h 10 y, 53
Vicente et al. [74] 14 S + RT UFC <303.5 nmol/24 h 2 y, 70
Nagesser et al. [81] 86 RT and unilateral Low-dose dexamethasone suppression test 21 y, 64
adrenalectomy (serum cortisol <80 nmol/L)
Estrada et al. [83] 30 S + RT UFC <303.5 nmol/24 h; low-dose 42 mo, 83 %
dexamethasone suppression test normal
S surgery, RT radiation therapy, UFC urinary free cortisol
monal control without medical management does increase, Meningiomas

with perhaps 50 % of patients achieving a serum level
<500 mU/L at 10 years after irradiation [38, 68–70]. Radiotherapy has been shown to be highly effective at con-
Hormonal normalization may only be achieved in many trolling meningiomas in numerous series [30, 84–88]. Large,
patients with continuing medical management after sur- recurrent, and histologically aggressive tumors are more
gery and irradiation. Radiotherapy does reliably achieve likely to progress after treatment [30, 86]. The morbidity
growth control, where this is an important goal (to avoid associated with fractionated irradiation is acceptably low,
visual or other complications). The results of several particularly when compared with surgical series, but can cer-
series of prolactinomas treated with fractionated irradia- tainly be minimized for individual patients by not using
tion are summarized in Table 28.5 [38, 68–73]. Recently radiotherapy plans that have dosimetric hot spots or cold
reported series are very small, in part because of the suc- spots that may lead to radiation injury to normal tissues or
cess of medical management and in part because radiosur- early recurrences [86–89].
gical treatment is increasingly commonly used for patients
who need irradiation.
Craniopharyngiomas
Adrenocorticotropic Hormone-Secreting Radiotherapy has been used in the treatment of craniopha-

Adenomas (Cushing’s Disease) ryngiomas for many decades. Detailed evaluations of surgi-
cal outcomes and of radiation therapy outcomes have been
Adrenocorticotropic hormone (ACTH)-hypersecreting pituitary compiled [90–92]. Aggressive surgery has been associated
adenomas should be irradiated if neurosurgical treatment is with higher rates of visual loss and diabetes insipidus, and
unsuccessful [74–78]. Fractionated irradiation has historically larger tumors are recognized to be more likely to recur [90].
been used as primary therapy for Cushing’s disease as well, but The use of radiotherapy has been shown in retrospective
in current practice, only a rare case will be referred for irradia- series to decrease the probability of recurrence requiring
tion without a pituitary operation [78, 79]. In patients medically additional surgery after a subtotal resection and has led to the
unable to undergo microsurgery, medical therapy can suppress acceptance of the role of radiation therapy. Novel radiation
cortisol excess until the effects of irradiation on hypercorti- delivery technologies have been used over the years to treat
solemia can be achieved, commonly seen in a year or 2 [80–82]. craniopharyngiomas [93–96] because of concern about
The results of several radiotherapy series for Cushing’s disease delayed radiation-related morbidity being added to some-
are detailed in Table 28.6 [38, 74, 80, 81, 83]. times considerable morbidity from the tumor and surgical
Table 28.7 Results of radiation therapy for craniopharyngioma

Results
Combs et al. [97] 40 FSRT Tumor growth control 10 y, 100
Minniti et al. [98] 39 FSRT Tumor growth control 5 y, 92
Smee et al. [99] 41a RT, FSRT Tumor growth control 23 y, 95.5
Masson-Cote et al. [100] 53 RT Tumor growth control 5 y, 85; 10 y, 69
S surgery, RT radiation therapy, FSRT fractionated stereotactic radiation therapy
a
Five of 41 patients received stereotactic radiosurgery or intracystic radiotherapy—results by treatment approach were not presented separately
management. More recently, stereotactic radiotherapy and 7. Perry A, Louis DN, Scheithauer BW, Budka H, von Deimling
A. Meningiomas. In: Louis DN, Ohgaki H, Wiestler OD, Cavenee
proton beam radiotherapy have been used increasingly fre-
WK editors. WHO classification of tumours of the central nervous
quently. There is insufficient data to assess long-term results system. 4th ed. Lyon: The International Agency for Research on
of proton beam therapy at present. These treatment tech- Cancer; 2007. p. 164–72.
niques may limit the dose of irradiation given to contiguous 8. Perry A, Stafford SL, Scheithauer BW, et al. Meningioma grad-
ing: an analysis of histologic parameters. Am J Surg Pathol. 1997;
normal brain and pituitary, but as always, care must be taken
21:1455–65.
to avoid radiation injury to the optic apparatus [93] or not 9. Perry A, Scheithauer BW, Stafford SL, et al. “Malignancy” in
irradiating the entire tumor (cysts may increase in size dur- meningiomas: a clinicopathologic study of 116 patients. Cancer.
ing radiotherapy) [94, 95]. The results of several series of 1999;85:2046–56.
10. Kahn EA, Gosch HH, Seeger JF, et al. Forty-five years experience
patients irradiated for craniopharyngiomas are presented in
with the craniopharyngiomas. Surg Neurol. 1973;1:5–12.
Table 28.7 [97–100]. 11. Yasargil MG, Curcic M, Kis M, et al. Total removal of craniopha-
ryngiomas. Approaches and long-term results in 144 patients.
J Neurosurg. 1990;73:3–11.
12. Adamson TE, Wiestler OD, Kleihues P, et al. Correlation of clini-
Conclusion cal and pathological features in surgically treated craniopharyn-
giomas. J Neurosurg. 1990;73:12–7.
Numerous advances in imaging and treatment delivery have 13. Hardy J. Transphenoidal microsurgery of the normal and patho-
made fractionated irradiation for pituitary adenomas and logical pituitary. Clin Neurosurg. 1969;16:185–217.
14. Leavens ME, McCutcheon IF, Samaan NA. Management of pitu-
other anterior skull base tumors an extremely good therapeu-
itary adenomas. Oncology (Williston Park). 1992;6(6):69–79; dis-
tic option. Evaluation of radiation management approaches cussion 79–80.
for individual patients with tumors in the pituitary and para- 15. Simpson D. The recurrence of intracranial meningiomas after sur-
sellar region should bear in mind the potential for significant gical treatment. J Neurol Neurosurg Psychiatry. 1957;20:22–39.
16. Tsang RW, Brierley JD, Panzarella T, et al. Radiation therapy for
complications from the attempt to use the most advanced
pituitary adenoma: treatment outcome and prognostic factors. Int
technology available. Not every patient should be treated with J Radiat Oncol Biol Phys. 1994;30:557–65.
radiosurgery or IMRT; extremely good and predictable results 17. Perry A, Gutmann DH, Reifenberger G. Molecular pathogenesis
are possible with highly conformal fractionated irradiation, of meningiomas. J Neurooncol. 2004;70:183–202.
18. Perry A. Unmasking the secrets of meningioma: a slow but
an option that should be nearly universally available.
rewarding journey. Surg Neurol. 2004;61:171–3.
19. Molitch ME. Medical management of prolactin-secreting pituitary
adenomas. Pituitary. 2002;5:55–65.
References 20. Racine MS, Barkan AL. Medical management of growth hormone-
secreting pituitary adenomas. Pituitary. 2002;5:67–76.
1. Rhoton Jr AL. The anterior and middle cranial base. Neurosurgery. 21. Beck-Peccoz P, Persani L. Medical management of thyrotropin-
2002;51:S273–302. secreting pituitary adenomas. Pituitary. 2002;5:83–8.
2. Laws Jr ER, Thapar K. Brain tumors. CA Cancer J Clin. 22. Tishler RB, Loeffler JS, Lunsford LD, et al. Tolerance of cranial
1993;43:263–71. nerves of the cavernous sinus to radiosurgery. Int J Radiat Oncol
3. Costello RT. Subclinical adenoma of the pituitary gland. Am J Biol Phys. 1993;27:215–21.
Pathol. 1936;12:205–16. 23. Leber KA, Bergloff J, Pendl G. Dose–response tolerance of the
4. Kernohan JW, Sayre GP. Tumors of the pituitary gland and infun- visual pathways and cranial nerves of the cavernous sinus to ste-
dibulum. In: Atlas of tumor pathology. Washington: Armed Forces reotactic radiosurgery. J Neurosurg. 1998;88:43–50.
Institute of Pathology, Section X, Fascicle 36; 1956. 24. Stafford SL, Pollock BE, Leavitt JA, et al. A study on the radiation
5. DeMonte F, Al-Mefty O. Meningiomas. In: Kaye AH, Laws ER, tolerance of the optic nerves and chiasm after stereotactic radio-
editors. Brain tumors. New York: Churchill Livingstone; 1995. surgery. Int J Radiat Oncol Biol Phys. 2003;55:1177–81.
p. 675–704. 25. van den Bergh AC, Schoorl MA, Dullaart RP, et al. Lack of radia-
6. Moss J, DeCastro R, Patronas NJ, et al. Meningiomas in lymphan- tion optic neuropathy in 72 patients treated for pituitary adenoma.
gioleiomyomatosis. JAMA. 2001;286:1879–81. J Neuroophthalmol. 2004;24:200–5.
26. al-Mefty O, Kersh JE, Routh A, et al. The long-term side effects markers of tumour quiescence and regrowth. Clin Endocrinol
of radiation therapy for benign brain tumors in adults. J Neurosurg. (Oxf). 2003;58:763–9.
1990;73:502–12. 46. Grigsby PW, Simpson JR, Fineberg B. Late regrowth of pituitary
27. Hall EJ, Wuu CS. Radiation-induced second cancers: the impact adenomas after irradiation and/or surgery. Hazard function analy-
of 3D-CRT and IMRT. Int J Radiat Oncol Biol Phys. 2003; sis. Cancer. 1989;63:1308–12.
56:83–8. 47. Breen P, Flickinger JC, Kondziolka D, et al. Radiotherapy for non-
28. Ohtakara K, Hayashi S, Tanaka H, et al. Clinical comparison of functional pituitary adenoma: analysis of long-term tumor control.
positional accuracy and stability between dedicated versus conven- J Neurosurg. 1998;89:933–8.
tional masks for immobilization in cranial stereotactic radiotherapy 48. Chen Y, Wang CD, Su ZP, et al. Natural history of postoperative
using 6-degree-of-freedom image guidance system-integrated plat- nonfunctioning pituitary adenomas: a systematic review and
form. Radiother Oncol. 2012;102:198–205. meta-analysis. Neuroendocrinology. 2012;96(4):333–42.
29. Marks JE, Baglan RJ, Prassad SC, et al. Cerebral radionecrosis: 49. Park P, Chandler WF, Barkan AL, et al. The role of radiation ther-
incidence and risk in relation to dose, time, fractionation and vol- apy after surgical resection of nonfunctional pituitary macroade-
ume. Int J Radiat Oncol Biol Phys. 1981;7:243–52. nomas. Neurosurgery. 2004;55:100–6.
30. Goldsmith BJ, Larson DA. Conventional radiation therapy for 50. van den Bergh AC, Hoving MA, Links TP, et al. Radiation optic
skull base meningiomas. Neurosurg Clin N Am. 2000;11:605–15. neuropathy after external beam radiation therapy for acromegaly:
31. Graham JD, Nahum AE, Brada M. A comparison of techniques for report of two cases. Radiother Oncol. 2003;68:101–3.
stereotactic radiotherapy by linear accelerator based on 3-dimen- 51. Goffman TE, Dewan R, Arakaki R, et al. Persistent or recurrent
sional dose distributions. Radiother Oncol. 1991;22:29–35. acromegaly. Long-term endocrinologic efficacy and neurologic
32. Perks JR, Jalali R, Cosgrove VP, et al. Optimization of safety of postsurgical radiation therapy. Cancer. 1992;69:271–5.
stereotactically-guided conformal treatment planning of sellar and 52. Grigsby PW, Stokes S, Marks JE, et al. Prognostic factors and
parasellar tumors, based on normal brain dose volume histograms. results of radiotherapy alone in the management of pituitary ade-
Int J Radiat Oncol Biol Phys. 1999;45:507–13. nomas. Int J Radiat Oncol Biol Phys. 1988;15:1103–10.
33. Taylor BF, Picone J, Kimmett J, et al. Improved hippocampal- 53. Gittoes NJ, Bates AS, Tse W, et al. Radiotherapy for non-function
sparing using an asymmetric conformal arc technique for treat- pituitary tumours. Clin Endocrinol (Oxf). 1998;48:331–7.
ment of pituitary lesions. Int J Radiat Oncol Biol Phys. 2011; 54. Sasaki R, Murakami M, Okamoto Y, et al. The efficacy of conven-
81(Suppl):S291–2. tional radiation therapy in the management of pituitary adenoma.
34. Kondziolka D, Lunsford LD, Loeffler JS, et al. Radiosurgery and Int J Radiat Oncol Biol Phys. 2000;47:1337–45.
radiotherapy: observations and clarifications. J Neurosurg. 2004; 55. Dowsett RJ, Fowble B, Sergott RC, et al. Results of radiotherapy
101:585–9. in the treatment of acromegaly: lack of ophthalmologic complica-
35. Boehling NS, Grosshans DR, Bluett JB, et al. Dosimetric com- tions. Int J Radiat Oncol Biol Phys. 1990;19:453–9.
parison of three-dimensional conformal proton radiotherapy, 56. Giustina A, Chanson P, Bronstein MD, et al. A consensus on criteria
intensity-modulated proton therapy, and intensity-modulated for cure of acromegaly. J Clin Endocrinol Metab. 2010;95:3141–8.
radiotherapy for treatment of pediatric craniopharyngiomas. Int J 57. Gullu S, Keles H, Delibasi T, et al. Remission criteria for the fol-
Radiat Oncol Biol Phys. 2012;82:643–52. low-up of patients with acromegaly. Eur J Endocrinol. 2004;150:
36. Arvold ND, Niemierko A, Broussard GP, et al. Projected second 465–71.
tumor risk and dose to neurocognitive structures after proton ver- 58. Giustina A, Bronstein MD, Casanueva FF. Current management
sus photon radiotherapy for benign meningioma. Int J Radiat practices for acromegaly: an international survey. Pituitary. 2011;
Oncol Biol Phys. 2012;83(4):e495–500. 14:125–33.
37. Brada M, Rajan B, Traish D, et al. The long-term efficacy of con- 59. Melmed S, Sternberg R, Cook D, et al. A critical analysis of pitu-
servative surgery and radiotherapy in the control of pituitary ade- itary tumor shrinkage during primary medical therapy in acromeg-
nomas. Clin Endocrinol (Oxf). 1993;38:571–8. aly. J Clin Endocrinol Metab. 2005;90:4405–10.
38. Tsang RW, Brierley JD, Panzarella T, et al. Role of radiation ther- 60. Landolt AM, Haller D, Lomax N, et al. Octreotide may act as a
apy in clinical hormonally-active pituitary adenomas. Radiother radioprotective agent in acromegaly. J Clin Endocrinol Metab.
Oncol. 1996;41:45–53. 2000;85:1287–9.
39. McCord MW, Buatti JM, Fennell EM, et al. Radiotherapy for pitu- 61. Roug S, Rasmussen AK, Juhler M, et al. Fractionated stereotactic
itary adenoma: long-term outcome and sequelae. Int J Radiat radiotherapy in patients with acromegaly: an interim single-centre
Oncol Biol Phys. 1997;39:437–44. audit. Eur J Endocrinol. 2010;162:685–94.
40. Zierhut D, Flentje M, Adolph J, et al. External radiotherapy of pitu- 62. Eastman RC, Gorden P, Glatstein E, et al. Radiation therapy of
itary adenomas. Int J Radiat Oncol Biol Phys. 1995;33:307–14. acromegaly. Endocrinol Metab Clin North Am. 1992;21:693–712.
41. Milker-Zabel S, Debus J, Thilmann C, et al. Fractionated stereo- 63. Minniti G, Jaffrain-Rea ML, Osti M, et al. The long-term efficacy
tactically guided radiotherapy and radiosurgery in the treatment of of conventional radiotherapy in patients with GH-secreting pitu-
functional and nonfunctional adenomas of the pituitary gland. Int itary adenomas. Clin Endocrinol (Oxf). 2005;62:210–6.
J Radiat Oncol Biol Phys. 2001;50:1279–86. 64. Biermasz NR, van Dulken H, Roelfsema F. Long-term follow-up
42. Grabenbauer GG, Fietkau R, Buchfelder M, et al. Hormonally results of postoperative radiotherapy in 36 patients with acromeg-
inactive hypophyseal adenomas: the results and late sequelae after aly. J Clin Endocrinol Metab. 2000;85:2476–82.
surgery and radiotherapy. Strahlenther Onkol. 1996;172:193–7. 65. Thalassinos NC, Tsagarakis S, Ioannides G, et al. Megavoltage
43. Woollons AC, Hunn MK, Rajapakse YR, et al. Non-functioning pituitary irradiation lowers but seldom leads to safe GH levels in
pituitary adenomas: indications for postoperative radiotherapy. acromegaly; a long-term follow-up study. Eur J Endocrinol. 1998;
Clin Endocrinol (Oxf). 2000;53:713–7. 138:160–3.
44. Lillehei KO, Kirschman DL, Kleinschmidt-DeMasters BK, et al. 66. Milker-Zabel S, Zabel A, Huber P, et al. Stereotactic conformal
Reassessment of the role of radiation therapy in the treatment of radiotherapy in patients with growth hormone-secreting pituitary
endocrine-inactive pituitary macroadenomas. Neurosurgery. adenoma. Int J Radiat Oncol Biol Phys. 2004;59:1088–96.
1998;43:432–8. 67. González B, Vargas G, Espinosa-de-los-Monteros AL, et al.
45. Greenman Y, Ouaknine G, Veshchev I, et al. Postoperative surveil- Efficacy and safety of radiotherapy in acromegaly. Arch Med Res.
lance of clinically nonfunctioning pituitary macroadenomas: 2011;42:48–52.
68. Williams M, van Seters AP, Hermans J, et al. Evaluation of the 85. Dufour H, Muracciole X, Metellus P, et al. Long-term tumor
effects of radiotherapy on macroprolactinomas using the decline control and functional outcome in patients with cavernous sinus
rate of serum prolactin levels as a dynamic parameter. Clin Oncol meningiomas treated by radiotherapy with or without previous
(R Coll Radiol). 1994;6:102–9. surgery: is there an alternative to aggressive tumor removal?
69. Ozgen T, Oruckaptan HH, Ozcan OE, et al. Prolactin secreting Neurosurgery. 2001;48:285–94.
pituitary adenomas: analysis of 429 surgically treated patients, 86. Milker-Zabel S, Zabel A, Schulz-Ertner D, et al. Fractionated
effect of adjuvant treatment modalities and review of the litera- stereotactic radiotherapy in patients with benign or atypical intra-
ture. Acta Neurochir (Wien). 1999;141:1287–94. cranial meningioma: long-term experience and prognostic factors.
70. Littley MD, Shalet SM, Reid H, et al. The effect of external pitu- Int J Radiat Oncol Biol Phys. 2005;61:809–16.
itary irradiation on elevated serum prolactin levels in patients with 87. Jalali R, Loughrey C, Baumert B, et al. High precision focused
pituitary macroadenomas. Q J Med. 1991;81:985–98. irradiation in the form of fractionated stereotactic conformal radio-
71. Wallace EA, Holdaway IM. Treatment of macroprolactinomas at therapy (SCRT) for benign meningiomas predominantly in the
Auckland Hospital 1975–91. N Z Med J. 1995;108:50–2. skull base location. Clin Oncol (R Coll Radiol). 2002;14:103–9.
72. Tsagarakis S, Grossman A, Plowman PN, et al. Megavoltage pitu- 88. Debus J, Wuendrich M, Pirzkall A, et al. High efficacy of fraction-
itary irradiation in the management of prolactinomas: long-term ated stereotactic radiotherapy of large base-of-skull meningiomas:
follow-up. Clin Endocrinol (Oxf). 1991;34:399–406. long-term results. J Clin Oncol. 2001;19:3547–53.
73. Sun DQ, Cheng JJ, Frazier JL, Batra S, Wand G, Kleinberg LR, 89. Uy NW, Woo SY, Teh BS, et al. Intensity-modulated radiation
et al. Treatment of pituitary adenomas using radiosurgery and therapy (IMRT) for meningioma. Int J Radiat Oncol Biol Phys.
radiotherapy: a single center experience and review of literature. 2002;53:1265–70.
Neurosurg Rev. 2011;34:181–9. 90. Hetelekidis S, Barnes PD, Tao ML, et al. 20-year experience in
74. Vicente A, Estrada J, de la Cuerda C, et al. Results of external childhood craniopharyngioma. Int J Radiat Oncol Biol Phys.
pituitary irradiation after unsuccessful transsphenoidal surgery in 1993;27(2):189–95.
Cushing’s disease. Acta Endocrinol (Copenh). 1991;125:470–4. 91. Kalapurakal JA. Radiation therapy in the management of pediatric
75. Plowman PN. Pituitary adenoma radiotherapy—when, who and craniopharyngiomas—a review. Childs Nerv Syst. 2005;21(8–9):
how? Clin Endocrinol (Oxf). 1999;51:265–71. 808–16.
76. Becker G, Kocher M, Kortmann RD, et al. Radiation therapy in 92. Schoenfeld A, Pekmezci M, Barnes MJ, et al. The superiority of
the multimodal treatment approach of pituitary adenoma. conservative resection and adjuvant radiation for craniopharyngi-
Strahlenther Onkol. 2002;178:173–86. omas. J Neurooncol. 2012;108:133–9.
77. Mahmoud-Ahmed AS, Suh JH. Radiation therapy for Cushing’s 93. Hasegawa T, Kondziolka D, Hadjipanayis CG, et al. Management
disease: a review. Pituitary. 2002;5:175–80. of cystic craniopharyngiomas with phosphorus-32 intracavitary
78. Melby JC. Therapy of Cushing [sic] disease: a consensus for pitu- irradiation. Neurosurgery. 2004;54:813–20.
itary microsurgery. Ann Intern Med. 1988;109:445–6. 94. Glod J, Koch B, Myseros J, Breneman J, et al. Issues concerning
79. Mampalam TJ, Tyrrell JB, Wilson CB. Transsphenoidal microsur- the treatment of a child with a craniopharyngioma. Med Pediatr
gery for Cushing [sic)] disease. A report of 216 cases. Ann Intern Oncol. 2002;38:360–7.
Med. 1988;109:487–93. 95. Freeman CR, Patrocinio H, Farmer JP. Highly conformal radio-
80. Howlett TA, Plowman PN, Wass JA, et al. Megavoltage pituitary therapy for craniopharyngioma: (potentially) throwing the baby
irradiation in the management of Cushing’s disease and Nelson’s out with the bathwater. Med Pediatr Oncol. 2003;40:340–1.
syndrome: long-term follow-up. Clin Endocrinol (Oxf). 1989;31: 96. Selch MT, DeSalles AA, Wade M, et al. Initial clinical results of
309–23. stereotactic radiotherapy for the treatment of craniopharyngiomas.
81. Nagesser SK, van Seters AP, Kievit J, et al. Treatment of pituitary- Technol Cancer Res Treat. 2002;1:51–9.
dependent Cushing’s syndrome: long-term results of unilateral 97. Combs SE, Thilmann C, Huber PE, et al. Achievement of long-
adrenalectomy followed by external pituitary irradiation com- term local control in patients with craniopharyngiomas using high
pared to transsphenoidal pituitary surgery. Clin Endocrinol (Oxf). precision stereotactic radiotherapy. Cancer. 2007;109:2308–14.
2000;52:427–35. 98. Minniti G, Saran F, Traish D, et al. Fractionated stereotactic con-
82. Miller JW, Crapo L. The medical treatment of Cushing’s syn- formal radiotherapy following conservative surgery in the control
drome. Endocr Rev. 1993;14:443–58. of craniopharyngiomas. Radiother Oncol. 2007;82:90–5.
83. Estrada J, Boronat M, Mielgo M, et al. The long-term outcome of 99. Smee RI, Williams JR, Kwok B, et al. Modern radiotherapy
pituitary irradiation after unsuccessful transsphenoidal surgery in approaches in the management of craniopharyngiomas. J Clin
Cushing’s disease. N Engl J Med. 1997;336:172–7. Neurosci. 2011;18:613–7.
84. Selch MT, Ahn E, Laskari A, Lee SP, et al. Stereotactic radio- 100. Masson-Cote L, Masucci GL, Atenafu EG, et al. Long-term out-
therapy for treatment of cavernous sinus meningiomas. Int J comes for adult craniopharyngioma following radiation therapy.
Radiat Oncol Biol Phys. 2004;59:101–11. Acta Oncol. 2013;52(1):153–8.
Pituitary Tumors: Viewpoint— Medical
Therapy 29
Rachel L. Hopkins
found in these cases upon detailed questioning of the patient,

Introduction even when they are not the presenting complaints.
It is important to distinguish tumors of non-pituitary ori-
The management of tumors originating from the pituitary gin that can occur in the sellar region from those that arise
requires consideration of both hormonal functionality and mass from the pituitary. Most lesions of non-pituitary origin have
effect of the tumor. Tumors can arise out of any of the pituitary classic radiologic appearance, including meningiomas, ham-
cell types and each can produce a clinical syndrome with a typi- artomas, and craniopharyngiomas. Infiltrative processes
cal corresponding response to therapeutic efforts. It is impor- such as lymphocytic hypophysitis and sarcoidosis can usu-
tant that endocrine function be evaluated prior to any treatment ally be easily distinguished as well, but can look like tumors.
because some pituitary tumors, such as prolactinomas, will The relative role of surgical and medical therapy and
respond very well to medical treatment, obviating the need for radiotherapy in pituitary adenomas depends on the specific
surgery. Even when surgery is the primary treatment for a pitu- subtype of tumor. Many pituitary tumors require more than
itary tumor, the nature of the tumor can determine what pre- one treatment modality, and optimal therapy involves
and postsurgical therapy will be indicated. In addition, large collaboration between the surgeon, endocrinologist, radia-
tumors can cause hypopituitarism that needs to be addressed tion oncologist, and neuroradiologist.
before surgical intervention is attempted. Surgical and radiation
therapy can result in pituitary dysfunction that requires medical
management and replacement hormone therapy. Pituitary Tumor Subtypes
Prolactinomas
Diagnosis and Classification
Diagnosis
By convention, pituitary tumors are usually classified into Prolactinomas, or prolactin-secreting pituitary adenomas,
microadenomas, those less than 1 cm in diameter, and mac- are the most common type of functioning pituitary lesion and
roadenomas which are greater than or equal to 1 cm. account for about 40 % of all pituitary tumors. In most cases,
Microadenomas that are not hypersecreting hormones are patients present with typical signs and symptoms. In women,
often discovered incidentally on imaging. Small or large these include amenorrhea and galactorrhea. Men most
hypersecreting tumors usually come to attention because of often present with symptoms of hypogonadism including
symptoms of a hormone excess syndrome such as acromeg- decreased energy and sexual dysfunction. The prevalence of
aly or Cushing’s syndrome. In addition, macroadenomas prolactinomas is somewhat higher in women than in men [1].
may present with symptoms such as visual changes, head- Bone loss may occur as a result of suppression of sex steroid
aches, or cranial nerve dysfunction as a result of mass effect production [2].
of the tumor. Symptoms of pituitary hypofunction are often Diagnosis is usually easily established by the presence of
elevated serum prolactin levels. Levels greater than 500 μg/L
are essentially diagnostic of prolactinoma; those greater than
R.L. Hopkins, M.D. (*) 250 μg/L are usually prolactinomas (although some medica-
Division of Endocrinology & Metabolism,
State University of New York Upstate Medical University,
tions can cause elevations this high). Prolactin levels less
3229 East Genesee Street, Syracuse, NY 13214, USA than 250 μg/L can result from stalk compression or other
e-mail: hopkinra@upstate.edu causes, but can represent a small prolactinoma as well.
404 R.L. Hopkins
It is important to note that in the presence of a large pitu- treatment, this is a disease with high morbidity and mortality
itary tumor, mildly elevated prolactin levels (<100–150 μg/L) largely due to cardiovascular effects and abnormal glucose
may be the result of stalk compression by the tumor. Prolactin metabolism. Patients with cortisol excess, or Cushing’s syn-
secretion is tonically suppressed by dopamine delivered to the drome, usually come to medical attention because of typical
pituitary from the hypothalamus. Disruption of this pathway clinical features. The most specific of these features are easy
by compression of the stalk results in elevated serum prolac- bruising, facial plethora, proximal myopathy, and wide vio-
tin levels. Treatment with medication in this situation will laceous striae on the abdomen or upper thighs. In addition,
decrease prolactin levels, but will not control tumor growth. there are a host of less discriminating features present in the
general population that, in the right clinical setting, might
Management prompt investigation for Cushing’s syndrome such as central
Of all pituitary tumors, prolactinomas are the most respon- obesity and new-onset diabetes mellitus.
sive to medical therapy. The goal of therapy is to normalize The diagnosis of Cushing’s disease involves first confirm-
prolactin levels and reduce tumor size, and this goal can usu- ing the presence of hypercortisolism. Because none of our
ally be attained quickly with one of two dopamine agonists current tests for hypercortisolism has 100 % specificity or
available in the United States. Dopamine agonist therapy is sensitivity, this is a process that requires multiple confirma-
efficacious in treating even large tumors that are compress- tory tests, especially in more subtle cases [4]. Once estab-
ing the optic chiasm and causing visual changes. This suc- lished, it must be determined whether the hypercortisolism is
cess in treatment with medication has made surgery for ACTH dependent. This can usually be achieved with mea-
prolactinomas generally unnecessary, a development that has surement of a serum ACTH level.
occurred over the past several decades. The majority of ACTH-dependent hypercortisolism is
The two dopamine agonists available in the United States caused by corticotroph adenomas. However, evaluation of
are bromocriptine and cabergoline. Cabergoline is generally possible Cushing’s disease is complicated by the high inci-
preferred because of evidence of higher efficacy and a lower dence (about 10 %) of nonfunctioning pituitary incidentalo-
side effect profile. Bromocriptine can be used in patients mas. It is possible to misattribute ACTH-dependent
who do not tolerate cabergoline or if cost is a significant con- Cushing’s syndrome to a nonfunctioning pituitary adenoma.
sideration, as cabergoline is a more expensive medication. Care must be taken to avoid missing a diagnosis of ectopic
Over the past several years, there has been concern about ACTH production (for instance, from a bronchial carcinoid),
evidence of cardiac valve regurgitation in patients taking high which might result in unnecessary pituitary surgery and fail-
doses of dopamine agonists for treatment of Parkinson’s dis- ure to cure the underlying process.
ease. The data are somewhat equivocal, but so far there is no
strong evidence that such damage occurs in patients taking Management
the significantly lower doses used to treat prolactinomas [3]. The goal of therapy is to rapidly normalize serum cortisol
Surgical intervention is generally reserved for patients levels. Surgery is the primary treatment of corticotroph ade-
with tumors resistant to, or for patients who are intolerant of, nomas. Reported rates of remission resulting from transsphe-
dopamine agonists, patients considering pregnancy during noidal surgery range from 69 to 98 % [5]. The highest
which tumors might increase in size, or those who suffer remission rates are in cases of noninvasive microadenomas,
pituitary apoplexy. Approximately 10 % of patients are resis- while lower rates of success are seen in macroadenomas with
tant to cabergoline therapy even after dose optimization. invasion into the cavernous sinus. Conventional radiotherapy
Radiotherapy is recommended for patients who are intol- and stereotactic radiotherapy are used as adjuvant therapy,
erant of dopamine agonist therapy and in whom surgery has but their effects are slow and do not achieve prompt control
failed to control tumor growth. There are rare cases of very of hypercortisolism.
aggressive or malignant prolactinomas. These usually require There are a number of medical therapies that decrease
surgical and/or radiation therapy. Temozolomide therapy has either ACTH or cortisol secretion. Most have significant side
been used as chemotherapeutic treatment of malignant pro- effect profiles that can limit their use, but they may prove
lactinomas as well. beneficial in certain situations. In particular, medical therapy
is often employed in patients who have received radiation
therapy and are awaiting the often delayed results of that,
Corticotroph Adenomas those who are not candidates for surgery or in whom the
source of ACTH is uncertain, in very ill patients to ameliorate
Diagnosis the effects of hypercortisolism before surgery, or in patients
Cushing’s disease is defined as the condition of excess corti- with recurrent or persistent hypercortisolism after surgery.
sol production caused by an adrenocorticotropic hormone Inhibitors of steroidogenesis include ketoconazole,
(ACTH)-secreting pituitary adenoma. In the absence of metyrapone, mitotane, and etomidate. Ketoconazole has
29 Pituitary Tumors: Viewpoint—Medical Therapy 405
been used extensively in control of cortisol secretion. It is to confirm the diagnosis. Inadequate suppression of GH on
often well tolerated, but use can be limited by derangements this test indicates the presence of a somatotroph adenoma.
of liver enzymes. Metyrapone inhibits 11-beta-hydroxylase
activity. Its use is limited by androgenic effects in women Management
and mineralocorticoid-related adverse effects. It is also not The goal of treatment is both tumor shrinkage and biochemi-
available commercially and the producing company must be cal control of IGF-1 levels. Because of the insidious clinical
contacted directly to obtain it. Mitotane has significant gas- onset, most patients present with macroadenomas at the time
trointestinal and neurologic adverse effects. Also, once of diagnosis. As a result, the surgical cure rate is relatively
taken, it is stored in adipose tissue for about 2 years. low, around 50 %, in cases of macroadenomas greater than
Etomidate is used often in induction of anesthesia and is gen- 2 cm in size. In microadenomas showing no invasion into the
erally too sedating for practical long-term use. cavernous sinus, the surgical cure rate is as high as 80 % [9].
Centrally acting agents include cabergoline and pasireo- When surgery is not curative, radiotherapy, medical ther-
tide. While cabergoline, a dopamine agonist discussed previ- apy, and quite often both are used as adjunct treatment. First-
ously, does have a relatively low side effect profile, it is line medical therapy usually involves somatostatin analogs
generally not very efficacious in controlling serum cortisol (SSAs), either octreotide or lanreotide. Some tumors resis-
levels and any effect seen is not always sustained [6]. tant to usual doses of SSAs respond well to dose optimiza-
Pasireotide is a somatostatin receptor ligand that has shown tion using higher-than-usual doses [10]. The dopamine
some potential for use in Cushing’s syndrome, but results of agonists cabergoline and bromocriptine have proven useful
ongoing research are pending [7]. There is some evidence in some cases of acromegaly, even those without cosecretion
that combined therapy with cabergoline, pasireotide, and of prolactin. Cabergoline is preferred because there is more
ketoconazole may be useful. The glucocorticoid receptor evidence of efficacy [9, 11].
antagonist mifepristone (RU486) has proven useful in some Pegvisomant is a human growth hormone analog that
cases of hypercortisolism [8]. blocks the action of GH at the site of its cognate receptor and
Bilateral adrenalectomy is necessary in some patients with is quite successful at decreasing IGF-1 levels. Because of its
refractory hypercortisolism or who have been intolerant to mechanism of action, treatment with pegvisomant actually
several other interventions, or in whom the source of ACTH causes an increase in GH levels. There is concern about
overproduction cannot be identified. This option is compli- reflex increase in tumor size resulting from the lack of nega-
cated by the risk of Nelson’s syndrome (enlargement of the tive feedback from IGF-1. This has proven to be an uncom-
pituitary adenoma secondary to lack of suppressive feedback mon event that has not been clearly attributable to
from circulating cortisol). If bilateral adrenalectomy is pegvisomant itself [12]. Nonetheless, patients being treated
employed, lifelong steroid replacement will be required. with pegvisomant should undergo regular surveillance with
MRI during initial treatment. It is also prudent to avoid its
use in the setting of large tumors near the optic chiasm.
Somatotroph Adenomas Primary medical, as opposed to surgical, therapy may be
appropriate in patients with macroadenomas showing no evi-
Diagnosis dence of mass effect in which cure is very unlikely, those
Growth hormone-secreting, or somatotroph, adenomas cause who carry a high surgical risk and those who for their own
gigantism in children and acromegaly in adults. Acromegaly reasons prefer medication over surgery.
is a rare disease associated with increased morbidity and early The question has been raised as to whether use of SSAs
mortality. Clinical onset can be insidious and the disease can to debulk large tumors prior to surgeries will improve out-
take years to be recognized. Suspicion for the condition comes. At this point the data are mixed and it is not clear
should be raised in patients with two or more of the following whether there is any advantage to this approach. Further
comorbidities: new-onset diabetes, diffuse arthralgias, new- research is needed to explore this issue [9, 13].
onset or difficult-to-control hypertension, cardiac disease
including biventricular hypertrophy and diastolic or systolic
dysfunction, fatigue, headaches, carpal tunnel syndrome, Thyrotropin Adenomas
sleep apnea syndrome, diaphoresis, loss of vision, colon pol-
yps, and progressive jaw malocclusion [9]. The finding of an Diagnosis
elevated insulin-like growth factor-1 (IGF-1) level, combined Thyrotropin, or TSH-producing, adenomas are quite rare rep-
with suspicious clinical features and a pituitary lesion on resenting less than 2 % of pituitary adenomas. Classically, pre-
MRI, is sometimes adequate to make the diagnosis of acro- senting symptoms are those of hyperthyroidism and patients
megaly. However, in equivocal cases an oral glucose tolerance almost always have a goiter. Patients may also present with
test with serial growth hormone (GH) measurements is used symptoms of mass effect of the lesion or hypopituitarism.
406 R.L. Hopkins
The diagnosis is suggested by an elevated TSH level in Gonadotroph and Nonfunctioning Adenoma
the setting of elevated thyroxin or free thyroxin as opposed
to the suppressed TSH that would be found in primary hyper- Diagnosis
thyroidism. This lab pattern is also seen in the syndrome of Gonadotroph adenomas are the most common pituitary mac-
resistance to thyroid hormone (RTH). In this condition, fam- roadenomas and represent the majority of nonfunctioning
ily history will reveal other family members with thyroid adenomas. Many of these benign tumors are discovered inci-
problems. Studies that can help differentiate between these dentally on imaging obtained for other reasons, and others
two conditions include the TRH stimulation test and the T3 come to medical attention when they grow large enough to
suppression test in which lack of response would indicate a cause symptoms of mass effect. Usually circulating levels of
thyrotropin adenoma. In addition, a high alpha-subunit to follicle-stimulating hormone (FSH) and luteinizing hormone
TSH molar ratio is seen with these lesions. Genetic testing (LH) are normal because most gonadotroph adenomas
can evaluate for the presence of RTH. It is important to rule secrete inefficiently or variably. However, circulating levels
out RTH even if a pituitary lesion is seen on MRI because of of the free alpha-subunit component of these hormones are
the previously mentioned 10 % incidence of nonfunctioning often elevated.
pituitary adenomas in the general population. Patients being evaluated for incidentally discovered pitu-
The diagnosis of thyrotropin adenomas can be delayed by itary adenomas, or those who present with symptoms of
the coexistence of autoimmune hypothyroidism [14] or mass effect, should undergo biochemical evaluation for pos-
hyperthyroidism, both of which are far more common. sible hormone hypersecretion or hyposecretion. In patients
Classically, by the time of diagnosis, most patients with TSH- with large adenomas, symptoms of mass effect are often
producing adenomas have had a long history of thyroid dys- more obvious, but careful clinical evaluation and testing may
function and usually have quite large tumors often with reveal hypopituitarism in many. Replacement of pituitary
suprasellar extension or sphenoidal sinus invasion. This may hormones is an important aspect of the medical management
be changing with the advent of highly sensitive TSH assays. of pituitary adenomas.
There is some evidence that more thyrotropin adenomas have Hypersecreting gonadotroph adenomas are rare but do
been diagnosed as microadenomas in the past decade or so, occur. There are several case reports of ovarian hyperstimu-
although not all centers have found this to be the case [15, 16]. lation syndrome (OHSS) in patients shown to have FSH-
hypersecreting pituitary adenomas [19]. Diagnosis of
Management hyperfunctioning gonadotroph adenomas is based on mea-
Surgery is the primary treatment for thyrotropin adenomas. surement of elevated FSH and/or LH in the setting of ele-
When tumors are large at diagnosis, complete surgical resec- vated estradiol or testosterone levels.
tion of these tumors is often quite difficult. Adjunct therapy
with radiation is considered when there is significant resid- Management
ual tissue or persistent hyperthyroidism after surgery. Clinically nonfunctioning pituitary microadenomas do not
Medical therapy with the long-acting SSAs octreotide generally require surgery as long as they are not causing
LAR and lanreotide has been reported with fairly high suc- mass effect or endocrine dysfunction. Most nonfunctioning
cess. The combination of surgery and radiation therapy microadenomas will not grow significantly, but some will.
results in complete normalization of thyroid hormone levels Therefore, patients should be followed clinically and with
about 1/3 of the time, while 1/3 of cases achieve a level of repeat imaging, usually at yearly intervals. Macroadenomas
control considered adequate [17]. Treatment with SSAs is are more likely to grow, and for those that don’t initially
successful in about 95 % of patients. Dopamine agonists meet criteria for surgery, follow-up imaging is recommended
cabergoline and bromocriptine have been used to treat TSH- 6 months after initial presentation. Evaluation for pituitary
secreting adenomas with far less consistent results. hypofunction should also be repeated periodically in
Treatment with either thyroidectomy or thyroid ablation macroadenomas.
with radioactive iodine is generally not advised because of For nonfunctioning tumors that are causing compression
concern that absence of thyroid hormone feedback suppres- of the optic chiasm or other mass effect, surgery is the first
sion will result in tumor expansion (much as in Nelson’s syn- line of treatment. Endocrine function needs to be carefully
drome). One case report following two patients over 8–12 monitored after surgery to look for both resolution of preop-
years saw no such response [18]. Nonetheless, this is still not erative pituitary deficits and the development of new hypopi-
considered to be a standard approach to treatment. tuitarism as a result of surgery. In addition, long-term
Successful control of hyperthyroidism may result in per- surveillance is necessary because nonfunctioning pituitary
manent or transient central hypothyroidism, and levothyrox- adenomas can recur after transsphenoidal surgery [20, 21].
ine replacement therapy may be necessary after surgery and/ The risk of recurrence is thought to be highest in cases with
or radiotherapy. incomplete resection leaving residual tumor. In these cases,
29 Pituitary Tumors: Viewpoint—Medical Therapy 407
adjuvant radiation therapy might lower the recurrence rate,

although some argue for observation without radiotherapy in References
these patients [21]. In patients who undergo prophylactic
radiation therapy, there is some risk of developing hypopitu- 1. Kars M, Souberein PC, Herings RM, Romijn JA, Vandenbroucke
JP, DeBoer A, Dekkers OM. Estimated age- and sex-specific inci-
itarism so these patients need to undergo long-term monitor-
dence and prevalence of dopamine agonist-treated hyperprolac-
ing of endocrine function. Decline in endocrine function tinemia. J Clin Endocrinol Metab. 2009;94(8):2729–34.
may occur months or even years after radiation treatment. 2. Schlechte J, Walkner L, Kathol M. A longitudinal analysis of pre-
No drug therapy has been shown to consistently control menopausal bone loss in healthy women and women with hyperp-
rolactinemia. J Clin Endocrinol Metab. 1992;75(3):698–703.
tumor growth in nonfunctioning adenomas [22, 23]. New
3. Melmed S, Casanueva FF, Hoffman AR, Kleinberg DL, Montori
medical treatments are being explored, but have not been VM, Schlechter JA, Wass JA. Diagnosis and treatment of hyperpro-
through adequate clinical trials to be included in routine lactinemia: an Endocrine Society clinical practice guideline. J Clin
practice [24]. Therefore, the main role of medical therapy for Endocrinol Metab. 2011;96(2):273–88.
4. Nieman L, Biller BMK, Findling JW, Newell-Price J, Savage MO,
nonfunctioning pituitary adenomas is to address pituitary
Stewart PM, Montori VM. The diagnosis of Cushing’s syndrome:
hypofunction caused by either the tumor itself or by surgical an Endocrine Society clinical practice guideline. J Clin Endocrinol
or radiation therapy. Metab. 2008;93(5):1526–40.
Tumors causing gonadotropin hyperfunction are treated 5. Kelly DF. Transsphenoidal surgery for Cushing’s disease: a review
of success rates, remission predictors, management of failed
primarily with surgery and radiotherapy as above. There
surgery, and Nelson’s syndrome. J Neurosurg Focus. 2007;23(3):
have been reports of hormonal control with dopamine ago- 1–6.
nists, but again, only rare tumor shrinkage [22]. 6. Petrossians P, Thonnard A, Beckers A. Medical treatment in
Although the majority (about 80 %) of nonfunctioning Cushing’s syndrome: dopamine agonists and cabergoline.
Neuroendocrinology. 2010;92 Suppl 1:116–9.
pituitary adenomas are of gonadotroph origin, all of the other
7. Pedroncelli A. Medical treatment of Cushing’s disease: somatosta-
pituitary tumor subtypes can present as nonfunctioning ade- tin analogues and pasireotide. Neuroendocrinology. 2010;92 Suppl
nomas as well. There is emerging evidence that these differ- 1:120–4.
ent subtypes may follow quite different clinical courses [25]. 8. Castinetti F, Brue T, Conte-Devolx B. The use of glucocorticoid
receptor antagonist mifepristone in Cushing’s syndrome. Curr Opin
A subtype referred to as silent corticogonadotroph adenomas
Endocrinol Diabetes Obes. 2012;19(4):295–9.
in particular seems to have a more aggressive clinical course 9. Katznelson L, Atkinson JLD, Cook DM, Ezzat SZ, Hamrahian AH,
with a higher rate of both recurrence and development of Miller KK. American Association of Clinical Endocrinologists
postsurgical hypopituitarism [26]. Immunostaining of medical guidelines for clinical practice for the diagnosis and treat-
ment of acromegaly—2011 update. Endocr Pract. 2011;17 Suppl
resected nonfunctioning pituitary tumors may aid in guiding
4:1–44.
postsurgical management of patients with nonfunctioning 10. Ludlam WH, Anthony L. Safety review: dose optimization of
pituitary adenomas. somatostatin analogs in patients with acromegaly and neuroendo-
crine tumors. Adv Ther. 2011;28(10):825–41.
11. Sandret L, Maison P, Chanson P. Place of cabergoline in acromeg-
aly: a meta-analysis. J Clin Endocrinol Metab. 2011;96(5):
Conclusion 1327–35.
12. Buhk JH, Jung S, Psychogios MN, Goricke S, Hartz S, Schulz-
Heise S, Klingebiel R, Forsting M, Bruckmann H, Dorfler A,
The role and success of medical therapy for pituitary tumors Jordan M, Buchfelder M, Knauth M. Tumor volume of growth
varies tremendously depending on the tumor cell type. hormone-secreting pituitary adenomas during treatment with
Medical therapy is most successful for prolactinomas where pegvisomant: a prospective multicenter study. J Clin Endocrinol
Metab. 2010;95(2):552–8.
available drugs are both tolerable and highly effective. For
13. Sherlock M, Woods C, Sheppard MC. Medical therapy in acromeg-
all other tumors, neurosurgery is the primary modality of aly. Nat Rev Endocrinol. 2011;7:291–300.
treatment. In these cases radiation and drug therapy play 14. Meyers A, Hatanpaa KJ, Madden C, Lingvay I. Thyrotropin-
adjuvant roles. Monitoring for hypopituitarism is crucial secreting adenoma in a patient with primary hypothyroidism.
Endocr Pract. 2011;17(6):135–9.
both pre- and postoperatively, especially in macroadenomas,
15. Socin HV, Chanson P, Delemer B, Tabarin A, Rohmer V, Mockel J,
and after radiation treatment to the pituitary. Medical man- Stevenaert A, Beckers A. The changing spectrum of TSH-secreting
agement is crucial in achieving appropriate pituitary hor- pituitary adenomas: diagnosis and management in 43 patients. Eur
mone replacement therapy which is both life sustaining J Endocrinol. 2003;148(4):433–42.
16. Macchia E, Gasperi M, Lombardi M, Morselli L, Pinchera A,
and leads to improved quality of life in patients with
Acerbi G, Rossi G. Clinical aspects and therapeutic outcome in
hypopituitarism. thyrotropin-secreting pituitary adenomas: a single center experi-
Finally, advances in drug therapy targeted to more spe- ence. J Endocrinol Invest. 2009;32:773–9.
cific receptor subtypes in pituitary tumors may lead to even 17. Beck-Peccoz P, Persani L, Mannavola D, Campi I. TSH-secreting
adenomas. Best Pract Res Clin Endocrinol Metab. 2009;23:597–606.
more successful medical therapy of the pituitary tumor sub-
18. Daousi C, Foy PM, MacFarlane IA. Ablative thyroid treatment for
types. Currently, a multidisciplinary approach affords the thyrotoxicosis due to thyrotropin-producing pituitary tumors.
best advantage to the patient. J Neurol Neurosurg Psychiatry. 2007;78(1):93–5.
408 R.L. Hopkins
19. Knoepfelmacher M, Danilovic DLS, Nasser RHRR, Mendonca 23. Dekkers OM, Pereira AM, Romijn JA. Treatment and follow-up of
BB. Effectiveness of treating ovarian hyperstimulation syndrome clinically nonfunctioning pituitary macroadenomas. J Clin
with cabergoline in two patients with gonadotropin-producing pitu- Endocrinol Metab. 2008;93(10):3717–26.
itary adenomas. Fertil Steril. 2006;86(3):15–8. 24. Pereirra AM, Biermasz NR. Treatment of nonfunctioning pituitary
20. Chen Y, Wang CD, Su ZP, Chen YX, Cai L, Zhuge QC, Wu adenomas: what were the contributions of the last 10 years? Ann
ZB. Natural history of postoperative nonfunctioning pituitary ade- Endocrinol (Paris). 2012;73(2):111–6.
nomas: a systematic review and meta-analysis. Neuroendocrinology. 25. Cooper O, Melmed S. Subclinical hyperfunctioning pituitary ade-
2012;96(4):333–42. nomas: the silent tumors. Best Pract Res Clin Endocrinol Metab.
21. Chang EF, Zada G, Kim S, Lamborn KR, Quinones-Hinojosa A, 2012;26:447–60.
Tyrrell JB, Wilson CB, Kunwar S. Long-term recurrence and mor- 26. Cooper O, Ben-Shlomo A, Bonert V, Bannykh S, Mirocha J,
tality after surgery and adjuvant radiotherapy for nonfunctional Melmed S. Silent corticogonadotroph adenomas: clinical and cel-
pituitary adenomas. J Neurosurg. 2008;108:736–45. lular characteristics and long-term outcomes. Horm Cancer. 2010;
22. Shomali ME, Katznelson L. Medical therapy of gonadotropin- 1(2):80–92.
producing and nonfunctioning pituitary adenomas. Pituitary. 2002;
5(2):89–98.
Pediatric Radiosurgery
30
Arthur K. Liu
Linac-based radiosurgery uses a single X-ray source

Introduction which is moved around the patient. Collimation can be
accomplished using cones or multi-leaf collimators.
In the 1950s, Drs. Leksell and Larsson first described the use Treatment can be delivered using static fields or moving arcs.
of stereotactic radiosurgery (SRS), a technique to deliver an Patient head positioning uses either an invasive frame similar
ablative dose of radiation to an intracranial target [1, 2]. Since to the frame used for Gamma Knife radiosurgery, or non-
then, extensive experience has been obtained with SRS in invasive frames. Additional positioning accuracy is obtained
adults, demonstrating safety and efficacy in benign and with robotic couches that allow for correction of both trans-
malignant diseases [3–5]. Over the past three decades, there lational and rotational errors.
has been increasing interest in the use of SRS in pediatric
patients. In this chapter, we discuss the treatment of children
with SRS and review existing pediatric SRS literature. Unique Aspects of Radiosurgery in Children
There are some unique practical aspects of delivering radio-

Radiosurgery surgery in children. In children less than 24 months old, the
lack of skull thickness complicates the fixation of a stereo-
The goal of radiosurgery is to deliver high doses of radiation tactic localization frame to the skull. Skull thickness is not an
to a discrete target while minimizing radiation to surrounding issue when using a non-invasive frame. An inappropriately
tissue. This requires accurate patient and tumor localization, applied head frame pin may penetrate the skull of a young
and the ability to have steep dose falloff from the target. child. To avoid this complication, some institutions utilize a
Radiosurgery can be delivered with a Gamma Knife system special torque wrench that allows accurate measurement of
and with linear accelerator (linac)-based systems. torque force applied to a child’s skull during head frame
The Gamma Knife system accomplishes this by directing placement, which should be no more than 2–4 inch-pounds
multiple (approximately 200 cobalt-60 sources, with the spe- of torque for younger children and slightly more for a child
cific number of sources depending on the type of system) over 3. Another potential complication of a child’s thin skull
beams of gamma rays to a single target. The gamma rays are is that it is deformable. This may result in a loss of position-
directed to the targets by collimators ranging in size from 4 ing accuracy. In younger children (2–10 years of age), preci-
to 18 mm that focus the beams of radiation from the source sion is maintained by using special posts of the head frame
to the target with an accuracy of 0.3 mm at the isocenter. that are tailored to the curve of a small child’s head. Often,
Using a stereotactic head frame, the target is placed at the longer pins are used because of their small head circumfer-
isocenter. Depending on the complexity of the target shape, ence, making placement of the frame more critical than in
there may be multiple isocenters that are treated to ade- adolescents and adults. Because of the need to use long pins,
quately cover the target. special care must be used when moving the child during the
procedure.
Most adult patients receive Gamma Knife radiosurgery
A.K. Liu, M.D., Ph.D. (*) under local anesthesia with light sedation. However, many chil-
Department of Radiation Oncology, University of Colorado
dren younger than 12 years old cannot tolerate the procedure
Cancer Center, 1665 Aurora Court, Suite 1032 MS F-706,
Aurora, CO 80045, USA awake and therefore undergo radiosurgery using general anes-
e-mail: arthur.liu@ucdenver.edu thesia. This requires special expertise with anesthesiologists
410 A.K. Liu
who are comfortable with imaging and treating patients under and the mean obliteration time was 79.6 months. Radiation
general anesthesia [6]. For Gamma Knife SRS, the anesthe- necrosis was seen in 1.1 % of patients. Another large series
tized child is moved multiple times during planning and treat- from the University of Virginia reported on 186 children
ment with the head frame in place. [13]. Similar to the UK experience, hemorrhage was the
The anesthetic concerns are decreased for linac-based most common presenting sign, seen in 71.5 % of children.
radiosurgery. Treatment planning scans are typically per- Mean initial SRS dose was 21.9 Gy. In 41 children who
formed on a different day, eliminating the need to move chil- underwent repeat SRS for incomplete obliteration, the mean
dren under general anesthesia. However, it is still critical to SRS dose at retreatment was 20.7 Gy. With a mean follow-up
closely coordinate with the anesthesia team to maintain a of 98 months, complete obliteration was seen in 49.5 % after
safe environment for treatment of the child. initial SRS and 58.6 % after repeat SRS. Small nidus volume
Finally, there is a significant difference in normal tissue and higher SRS dose were associated with higher rates of
tolerance between children and adults. There is mixed data obliteration. MRI changes were seen in 37.8 %, but were
showing cognitive deficits in adults who receive radiation to accompanied by symptoms in only 7.2 %. Another series
the brain [7]. However, there are multiple studies showing from the University of Pittsburgh reports on 135 children.
significant neurocognitive deficits in children, with increasing The median SRS dose was 20 Gy. With a median follow-up
risk and severity for younger children and high doses [7, 8]. of 71.3 months, the obliteration rate at 5 years was 67 %. On
multivariate analysis, only higher SRS dose was associated
with improved obliteration rates. Adverse radiation effects
Published Experiences of Pediatric SRS were seen in 5.9 %. There are two other series with 100 or
more children, one from Taiwan [14] and the other from
SRS has been used in children with arteriovenous malforma- France [15], that report similar results.
tions, malignant brain tumors, and benign brain tumors. The extensive use of SRS in children with AVMs has
Some reports are specific to pediatric patients, but many also allowed some comparisons between pediatric and adult
include adult patients. Whenever possible, the pediatric spe- patients. Hemorrhage as a presenting symptoms is more
cific data is presented. Current literature is reviewed here and common in children [16]. While the overall obliteration rate
grouped by diagnosis. is comparable between children and adults, it appears that
children respond earlier to SRS [16, 17]. There may also be
a differential response based on the size of the AVM. The
Arteriovenous Malformations group from Taiwan found that AVMs smaller than 3 cm and
larger than 20 cm have similar obliteration rates between
Arteriovenous Malformations (AVM) are congenital vascu- children and adults, with much better response in smaller
lar abnormalities that are at risk of hemorrhage. Primary AVMs [14]. However, children with AVMs of intermediate
management is typically surgery, with embolization as an size between 3 and 10 cm had lower obliteration rates than
option. SRS is a well-accepted treatment option when surgi- adults with similarly sized AVMs.
cal risk is unacceptable. The earliest descriptions of SRS for
children with AVMs come from Italy [9], the University of
Pittsburgh [10], and Children’s Hospital, Boston and Ependymoma
Brigham and Women’s Hospital [11]. Since that time, there
have been numerous publications reporting the experience of While ependymomas make up a relatively small proportion
using SRS for children with AVMs, demonstrating both effi- of pediatric brain tumors, they may be well suited to treat-
cacy and safety. The largest series are reviewed here. ment with SRS. The standard management of these tumors is
The largest pediatric SRS experience for AVMs comes maximally safe resection followed by postoperative fraction-
from the UK [12]. Three hundred and sixty three children ated radiation therapy. Chemotherapy may be of limited
underwent 410 treatments, including 4 with planned two- effectiveness and is currently being studied in a randomized
stage treatments and 43 who were retreated. The most com- trial by the Children’s Oncology Group. Local failures occur
mon presentation was hemorrhage seen in 80 % of the in approximately 2/3 of recurrences and typically occur
children. Mean SRS dose was 22.7 Gy (range 15–25 Gy). within the treatment field [18]. SRS has been used in both
Follow-up duration for the group was not specified. residual tumor and recurrent disease.
Obliteration rates were calculated from 220 children with In the upfront setting, at St. Jude Children’s Research
sufficient follow-up, while toxicity was reported for the Hospital five children received an SRS boost (median dose
entire cohort. For children treated upfront, the obliteration rate 10 Gy) after conventionally fractionated radiation (50.4–
was 71 % and the mean obliteration time was 32.4 months. 52.2 Gy) for residual disease [19]. At a median follow-up
For children retreated, the obliteration rate was 62.5 % of 24 months, one child had a local recurrence and died
30 Pediatric Radiosurgery 411
of progressive disease. No significant toxicities were seen.

In another series from Boston, three children with ependy- Embryonal Tumors
momas received SRS (dose 10–15 Gy) as part of initial
management after conventionally fractionated radiation [20]. Embryonal tumors include medulloblastoma, primitive neuro-
Two of those children are alive with no evidence of disease ectodermal tumors (PNET), medulloepithelioma, pineoblastoma,
30 and 62 months from SRS. ependymoblastoma, and atypical teratoid/rhabdoid tumor
There are larger series with ependymoma treated at (ATRT). Standard treatment for these embryonal tumors
recurrence. Twenty six patients (12 were under 18 years at includes surgical resection, chemotherapy, and radiation ther-
diagnosis) with 49 lesions were treated at the Mayo Clinic apy. Radiation treatment for these tumors typically includes
[21]. The median SRS dose was 18 Gy. With a median fol- craniospinal irradiation due to a risk of dissemination through-
low-up of 3.1 years, local control at 3 years was 72 % and out the CNS, along with a boost to initial sites of disease. SRS
median overall survival was 5.5 years. Two patients devel- has also been used in both initial therapy and recurrent disease.
oped radiation necrosis. At the University of Pittsburgh, The Boston SRS series includes 21 children with medullo-
39 patients (median age 22.8 years, ranging from 2.9 to blastoma and PNET, with six children being treated upfront
71.1 years) received SRS to 56 lesions [22]. Median SRS with SRS (median dose 12.5 Gy) [20]. Median PFS was 11
dose was 15 Gy. With a median follow-up of 23.5 months, months and 3 year local control was 57 %. A report from the
PFS at 3 years was 45.8 % and OS at 3 years was 36.1 %. University of Heidelberg on 20 patients with 29 lesions
Three patients developed radiation necrosis. included 6 patients under 14 years [26]. All patients had recur-
While these two series show good local control with min- rent disease after receiving prior chemotherapy and radiation
imal toxicity, there are other reports with less promising therapy. Eight lesions were treated with SRS with a median
results. The Boston series reports on 90 children treated dose of 15 Gy. There were no local failures or symptomatic
with SRS (median dose 12.5 Gy) included 25 children with radiation necrosis. However, the follow-up for the SRS lesions
ependymoma treated at recurrence and the three children was not specified. A series from Austria included 12 children
treated upfront mentioned earlier [20]. For children with with embryonal tumors [27]. Mean dose was 18.8 Gy. With a
ependymomas treated with SRS, median PFS was mean follow-up of 32 months, local control was 50 %. At the
8.5 months and local control at 3 years was 29 %. Radiation University of Pittsburgh, seven children with recurrent disease
necrosis and neurologic symptoms requiring surgery received SRS (median dose was 14.5 Gy) to 15 lesions [28].
occurred in 19 of the 90 patients. A smaller series of six With a median follow-up of 25.8 months, local failure was
children with recurrent ependymoma receiving SRS (median seen in 5 children. No radiation necrosis was seen. All children
dose 18 Gy) at St. Jude Children’s Research Hospital had a died from progressive disease, with a median survival of
median PFS of 15 months and median OS of 65 months. 15 months from SRS. A series from Texas included seven chil-
Four patients died of disease. One patient was alive 10 years dren with embryonal tumors (medulloblastoma, PNET, and
after SRS, but with symptomatic radiation necrosis requir- ATRT) who received SRS (median dose of 17 Gy) to 15
ing surgery and hyperbaric oxygen. The last patient died lesions [24]. With a median follow-up of 11.5 months, there
from radiation necrosis. was one in field failure and two cases of radiation necrosis.
There are other series that include small numbers of pedi- There are other small series of SRS in pediatric embryo-
atric patients. The Indiana University experience reported on nal tumors. A report from Denmark included three children
eight patients (five were children) with 13 tumors, treated with embryonal tumors who received SRS (15–18 Gy) [29].
upfront and at recurrence. Median SRS dose was 14 Gy and Two children are alive with distant disease at 76 and 70
median follow-up was 30 months. Three year in field control months from SRS and one died from metastatic disease at 13
was 61 % and OS was 75 %. Two patients, both of whom months from SRS. At the University of Arizona, two chil-
received prior radiation therapy, developed symptomatic radi- dren with medulloblastoma received SRS (4.5 and 6 Gy) for
ation necrosis. The Washington University experience of SRS a boost after fractionated radiation therapy. At 51 months
(dose ranged from 14 to 20 Gy) for nine patients with epen- and 16 months of follow-up, both children have no evidence
dymoma included an 11 and an 18 years old [23]. The older of disease and no radiation necrosis was seen.
child treated definitively is alive with no evidence of disease
56 months after SRS. The younger child had a local failure
and died 27 months after SRS. A series from Texas included High-Grade Gliomas
three children with ependymoma [24]. All three recurred
within 16 months, and two died from progressive disease. High-grade gliomas are WHO grade III and IV tumors.
Two children were treated in Japan to 16 lesions with local These are typically very aggressive with poor clinical out-
control for 21 months and no reported toxicity [25]. comes. Standard therapy includes resection and adjuvant
412 A.K. Liu
radiation and chemotherapy. In these children, SRS has been Another series from Texas included three children with
used as initial therapy and at recurrence. pilocytic astrocytomas who received SRS (dose of 10–25 Gy)
The pediatric SRS experience from Boston reports 18 [24]. No local failures were seen with a follow-up ranging
children with high-grade gliomas (median SRS dose 12.5 Gy from 1 to 27 months.
for the entire series) [20]. With a median follow-up of 24 There are a number of reported series of SRS for patients
months, PFS at 3 years was 12 months and local control at with low-grade astrocytomas that include children. However,
3 years was 50 %. An early series from the University of these do not report specific results for the children. A series
Pittsburgh of 25 children who received SRS included five from Japan of 12 patients with WHO grade I and 39 WHO
children with high grade gliomas [30]. Mean SRS dose for grade II patients included some children (mean age of 9.8
the entire group was 15.2 Gy. Two have no evidence of dis- and 30.9 years for patients with WHO grade I and WHO
ease, one had local progression only, one had disseminated grade II tumors, respectively) [33]. Mean SRS dose was 12.5
disease, and one died of disease. and 15.7 Gy for patients with WHO grade I and WHO grade
Two children with anaplastic astrocytomas treated to II tumors, respectively. For patients with WHO grade I
three lesions were included in the series from Texas [24]. tumors with a mean follow-up of 24.8 months, local control
SRS dose ranged from 17 to 19 Gy. With a follow-up of was 91.7 %. For patients with WHO grade II tumors with a
16–18 months, there was no progression and one child had mean follow-up of 28.5 months, local control was 87.2 %.
asymptomatic radiation necrosis. Post-SRS edema was seen in 16.7 % of WHO grade I patients
and 41 % of WHO grade II patients. Another series from
Taiwan of 21 patients with low- grade astrocytomas (includ-
Low-Grade Gliomas ing both WHO grade I and II tumors) includes nine children
[34]. For the entire cohort, the median SRS dose was 14.5 Gy.
Low-grade astrocytomas are WHO grade I or II and include With a median follow-up of 67 months, PFS at 10 years was
pilocytic astrocytomas and fibrillary astrocytomas. These 65 %. Eight patients (40 %) developed mild to moderate
tumors are typically slow growing. With a complete resec- adverse radiation effects. The University of Pittsburgh sepa-
tion, adjuvant therapy is usually not required. For incom- rately reported their SRS experience for 25 patients (median
pletely resected tumors, adjuvant chemotherapy and radiation age of 30 years, range 8–68 years) WHO grade II astrocyto-
therapy are treatment options. SRS has been used as initial mas [35]. Sixteen patients were treated at initial presentation
therapy and at recurrence. and nine were treated at progression. Median SRS dose
The largest series of children with low-grade gliomas was 14 Gy. With a median follow-up of 65 months, PFS at
comes from the University of Pittsburgh [31]. Fifty children 10 years was 37 %. No permanent treatment-related morbid-
with juvenile pilocytic astrocytomas received SRS (median ity was seen.
dose 14.5 Gy) at initial diagnosis for residual tumor (n = 16) or
recurrence (n = 34). With a median follow-up of 55.5 months,
PFS at 3 years was 82.8 %. Longer PFS was seen with solid Craniopharyngioma
tumors, smaller target volumes (<8 cc), treatment upfront
(versus at recurrence), and no brainstem involvement. Two Craniopharyngiomas are benign tumors that arise in Rathke’s
children developed asymptomatic radiation necrosis and three pouch. Typical management involves surgical resection with
had symptomatic radiation necrosis. At the Karolinska adjuvant fractionated radiation therapy used for residual dis-
Hospital in Sweden, their series included 16 children (of 19 ease. There are reports of SRS for either residual or recurrent
patients total in the report) with pilocytic astrocytomas who disease.
received SRS (median dose 10 Gy). With a median follow-up The largest report of SRS in pediatric patients with cra-
of the children of 5 years, there were no local failures. Three niopharyngioma comes from Japan [36]. Their report of 107
children had asymptomatic radiation necrosis and two had patients includes 38 children <15 years at diagnosis. For the
symptomatic radiation necrosis, with one requiring surgery. entire cohort, mean SRS dose was 11.5 Gy and mean follow-
A series reporting SRS in children from Austria included 12 up was 65.5 months. Progression-free survival and overall
children with low-grade gliomas [27]. Mean SRS dose was survival at 10 years were 53.8 % and 91 %, respectively.
15.2 Gy. With a mean follow-up of 38 months, tumor progres- Response was reported separately for children and adults,
sion was seen in two children and one child developed edema with 32 % of children progressing compared to 13 % of
requiring steroids. adults. A 37 patient series from the University of Virginia
A small series from Washington University includes included 17 children <18 years at diagnosis [37]. Median
six children with pilocytic astrocytomas that received SRS SRS dose was 14.5 Gy and median follow-up was 50 months.
(median dose 15.5 Gy) [32]. Median follow-up for the Progression-free survival and overall survival at 5 years were
SRS patients was not specified. PFS at 5 years was 80 %. 67 % and 75.6 %, respectively, with only one death attributed
30 Pediatric Radiosurgery 413
to tumor progression. In this group, local progression was 5. Monaco EA, Grandhi R, Niranjan A, Lunsford LD. The past, present
not associated with age. Another study from Sweden included and future of Gamma Knife radiosurgery for brain tumors: the
Pittsburgh experience. Expert Rev Neurother. 2012;12(4):437–45.
both children (11 under 16 years) and adults [10, 38]. Median Epub 2012/03/28.
follow-up was 3.5 years. In children, the median SRS dose 6. Bauman GS, Brett CM, Ciricillo SF, Larson DA, Sneed P, Staplers
was 5 Gy (range 2.5–15 Gy). Local progression was seen LJ, et al. Anesthesia for pediatric stereotactic radiosurgery.
in 82 % of the children, compared to 50 % of adults. This high Anesthesiology. 1998;89(1):255–7. Epub 1998/07/17.
7. Lawrence YR, Li XA, el Naqa I, Hahn CA, Marks LB, Merchant
rate of progression was attributed to the low SRS dose. A report TE, et al. Radiation dose-volume effects in the brain. Int J Radiat
on 12 patients with recurrent craniopharyngioma from the Oncol Biol Phys. 2010;76(3 Suppl):S20–7. Epub 2010/03/05.
University of Miami included 10 children <18 years [39]. 8. Mulhern RK, Merchant TE, Gajjar A, Reddick WE, Kun LE. Late
Mean SRS dose was 14 Gy. With a mean follow-up of 39 neurocognitive sequelae in survivors of brain tumours in childhood.
Lancet Oncol. 2004;5(7):399–408.
months, two local failures (one in a child and one in an adult) 9. Colombo F, Benedetti A, Casentini L, Pozza F, Avanzo R, Marchetti
were seen. Other series that included children report similar C, et al. Stereotactic radiosurgery of intracranial tumors in child-
response, with tumor control rates of 68–87 % [27, 40, 41]. hood. J Neurosurg Sci. 1985;29(3):233–7. Epub 1985/07/01.
10. Kondziolka D, Lunsford LD, Flickinger JC. Stereotactic radio-
surgery in children and adolescents. Pediatr Neurosurg. 1990;16(4–5):
219–21. Epub 1990/01/01.
Other Tumors 11. Loeffler JS, Rossitch Jr E, Siddon R, Moore MR, Rockoff MA,
Alexander 3rd E. Role of stereotactic radiosurgery with a linear
There are numerous reports of SRS for other diagnosis in accelerator in treatment of intracranial arteriovenous malforma-
tions and tumors in children. Pediatrics. 1990;85(5):774–82. Epub
children. In general, these are either very small or do not 1990/05/01.
specify results for the pediatric patients. These include 12. Dinca EB, de Lacy P, Yianni J, Rowe J, Radatz MW, Preotiuc-
meningiomas [24, 27, 42, 43], pituitary adenomas [44–50], Pietro D, et al. Gamma Knife surgery for pediatric arteriovenous
and vestibular schwannomas [51–57]. In general, these stud- malformations: a 25-year retrospective study. J Neurosurg Pediatr.
2012;10(5):445–50.
ies show minimal toxicity with good tumor control. 13. Yen CP, Monteith SJ, Nguyen JH, Rainey J, Schlesinger DJ,
Sheehan JP. Gamma Knife surgery for arteriovenous malforma-
tions in children. J Neurosurg Pediatr. 2010;6(5):426–34. Epub
Summary 2010/11/03.
14. Pan DH, Kuo YH, Guo WY, Chung WY, Wu HM, Liu KD, et al.
Gamma Knife surgery for cerebral arteriovenous malformations in
SRS in children has been reported for many diagnoses. The children: a 13-year experience. J Neurosurg Pediatr. 2008;1(4):
most robust data supporting SRS as a treatment option in 296–304. Epub 2008/04/02.
children exists for AVMs. Overall obliteration rates appear 15. Reyns N, Blond S, Gauvrit JY, Touzet G, Coche B, Pruvo JP, et al.
Role of radiosurgery in the management of cerebral arteriovenous mal-
similar between adults and children, although children may formations in the pediatric age group: data from a 100-patient series.
respond earlier. Neurosurgery. 2007;60(2):268–76; discussion 76; Epub 2007/02/10.
For malignant or benign brain tumors, the data is more 16. Tanaka T, Kobayashi T, Kida Y, Oyama H, Niwa M. Comparison
limited. The majority of reports are small with short follow- between adult and pediatric arteriovenous malformations treated by
Gamma Knife radiosurgery. Stereotact Funct Neurosurg. 1996;66
up. The largest series exist for craniopharyngioma, low- Suppl 1:288–95. Epub 1996/01/01.
grade glioma, and recurrent ependymoma. These reports 17. Nicolato A, Lupidi F, Sandri MF, Foroni R, Zampieri P, Mazza C,
taken together suggest low toxicity but efficacy is difficult et al. Gamma Knife radiosurgery for cerebral arteriovenous malfor-
to interpret given the small numbers and limited follow-up. mations in children/adolescents and adults. Part II: differences in
obliteration rates, treatment-obliteration intervals, and prognostic
At this time, SRS in children with brain tumors can be con- factors. Int J Radiat Oncol Biol Phys. 2006;64(3):914–21.
sidered in patients with recurrent or residual disease, but 18. Merchant TE, Li C, Xiong X, Kun LE, Boop FA, Sanford
ideally requires multi-disciplinary evaluation by neurosur- RA. Conformal radiotherapy after surgery for paediatric ependy-
gery, radiation oncology, and medical oncology. moma: a prospective study. Lancet Oncol. 2009;10(3):258–66.
19. Aggarwal R, Yeung D, Kumar P, Muhlbauer M, Kun LE. Efficacy
and feasibility of stereotactic radiosurgery in the primary manage-
ment of unfavorable pediatric ependymoma. Radiother Oncol.
References 1997;43(3):269–73. Epub 1997/06/01.
20. Hodgson DC, Goumnerova LC, Loeffler JS, Dutton S, Black PM,
1. Leksell L. The stereotaxic method and radiosurgery of the brain. Alexander 3rd E, et al. Radiosurgery in the management of pediat-
Acta Chir Scand. 1951;102(4):316–9. Epub 1951/12/13. ric brain tumors. Int J Radiat Oncol Biol Phys. 2001;50(4):929–35.
2. Larsson B, Leksell L, Rexed B, Sourander P, Mair W, Andersson Epub 2001/06/29.
B. The high-energy proton beam as a neurosurgical tool. Nature. 21. Stauder MC, Ni Laack N, Ahmed KA, Link MJ, Schomberg PJ,
1958;182(4644):1222–3. Epub 1958/11/01. Pollock BE. Stereotactic radiosurgery for patients with recurrent
3. Friedman WA, Bova FJ. Radiosurgery for arteriovenous malforma- intracranial ependymomas. J Neurooncol. 2012;108(3):507–12.
tions. Neurol Res. 2011;33(8):803–19. Epub 2011/10/19. 22. Kano H, Yang HC, Kondziolka D, Niranjan A, Arai Y, Flickinger JC,
4. Pollock BE. Stereotactic radiosurgery of benign intracranial et al. Stereotactic radiosurgery for pediatric recurrent intracranial epen-
tumors. J Neurooncol. 2009;92(3):337–43. Epub 2009/04/10. dymomas. J Neurosurg Pediatr. 2010;6(5):417–23. Epub 2010/11/03.
414 A.K. Liu
23. Mansur DB, Drzymala RE, Rich KM, Klein EE, Simpson JR. The 42. Im SH, Wang KC, Kim SK, Oh CW, Kim DG, Hong SK, et al.
efficacy of stereotactic radiosurgery in the management of intracra- Childhood meningioma: unusual location, atypical radiological
nial ependymoma. J Neurooncol. 2004;66(1–2):187–90. findings, and favorable treatment outcome. Childs Nerv Syst. 2001;
24. Giller CA, Berger BD, Pistenmaa DA, Sklar F, Weprin B, Shapiro 17(11):656–62.
K, et al. Robotically guided radiosurgery for children. Pediatr 43. Pollock BE, Stafford SL, Link MJ, Brown PD, Garces YI, Foote
Blood Cancer. 2005;45(3):304–10. Epub 2004/11/24. RL. Single-fraction radiosurgery of benign intracranial meningio-
25. Endo H, Kumabe T, Jokura H, Shirane R, Tominaga T. Stereotactic mas. Neurosurgery. 2012;71(3):604–13.
radiosurgery for nodular dissemination of anaplastic ependymoma. 44. Thoren M, Rahn T, Hallengren B, Kaad PH, Nilsson KO, Ravn H,
Acta Neurochir (Wien). 2004;146(3):291–8; discussion 8. et al. Treatment of Cushing’s disease in childhood and adolescence
26. Milker-Zabel S, Zabel A, Thilmann C, Zuna I, Hoess A, Wannenmacher by stereotactic pituitary irradiation. Acta Paediatr Scand. 1986;
M, et al. Results of three-dimensional stereotactically-guided radio- 75(3):388–95.
therapy in recurrent medulloblastoma. J Neurooncol. 2002;60(3): 45. Kobayashi T, Kida Y, Mori Y. Gamma knife radiosurgery in the
227–33. treatment of Cushing disease: long-term results. J Neurosurg. 2002;
27. Eder HG, Leber KA, Eustacchio S, Pendl G. The role of gamma 97(5 Suppl):422–8.
knife radiosurgery in children. Childs Nerv Syst. 2001;17(6): 46. Park KJ, Kano H, Parry PV, Niranjan A, Flickinger JC, Lunsford
341–6; discussion 7; Epub 2001/06/22. LD, et al. Long-term outcomes after gamma knife stereotactic
28. Flannery T, Kano H, Martin JJ, Niranjan A, Flickinger JC, Lunsford radiosurgery for nonfunctional pituitary adenomas. Neurosurgery.
LD, et al. Boost radiosurgery as a strategy after failure of initial man- 2011;69(6):1188–99.
agement of pediatric primitive neuroectodermal tumors. J Neurosurg 47. Castro DG, Cecilio SA, Canteras MM. Radiosurgery for pituitary
Pediatr. 2009;3(3):205–10. Epub 2009/04/03. adenomas: evaluation of its efficacy and safety. Radiat Oncol. 2010;
29. Mirza B, Monsted A, Harding J, Ohlhues L, Roed H, Juhler M. 5:109.
Stereotactic radiotherapy and radiosurgery in pediatric patients: 48. Iwai Y, Yamanaka K, Yoshimura M, Kawasaki I, Yamagami K,
analysis of indications and outcome. Childs Nerv Syst. 2010; Yoshioka K. Gamma knife radiosurgery for growth hormone-
26(12):1785–93. Epub 2010/07/30. producing adenomas. J Clin Neurosci. 2010;17(3):299–304.
30. Grabb PA, Lunsford LD, Albright AL, Kondziolka D, Flickinger 49. Castinetti F, Nagai M, Morange I, Dufour H, Caron P, Chanson P,
JC. Stereotactic radiosurgery for glial neoplasms of childhood. et al. Long-term results of stereotactic radiosurgery in secretory
Neurosurgery. 1996;38(4):696–701; discussion 701–2. pituitary adenomas. J Clin Endocrinol Metab. 2009;94(9):3400–7.
31. Kano H, Niranjan A, Kondziolka D, Flickinger JC, Pollack IF, Epub 2009/06/11.
Jakacki RI, et al. Stereotactic radiosurgery for pilocytic astrocyto- 50. Wan H, Chihiro O, Yuan S. MASEP gamma knife radiosurgery for
mas part 2: outcomes in pediatric patients. J Neurooncol. 2009; secretory pituitary adenomas: experience in 347 consecutive cases.
95(2):219–29. Epub 2009/05/27. J Exp Clin Cancer Res. 2009;28:36. Epub 2009/03/17.
32. Mansur DB, Rubin JB, Kidd EA, King AA, Hollander AS, Smyth 51. Sharma MS, Singh R, Kale SS, Agrawal D, Sharma BS, Mahapatra
MD, et al. Radiation therapy for pilocytic astrocytomas of childhood. AK. Tumor control and hearing preservation after Gamma Knife
Int J Radiat Oncol Biol Phys. 2011;79(3):829–34. Epub 2010/04/28. radiosurgery for vestibular schwannomas in neurofibromatosis type
33. Kida Y, Kobayashi T, Mori Y. Gamma knife radiosurgery for low- 2. J Neurooncol. 2010;98(2):265–70. Epub 2010/04/21.
grade astrocytomas: results of long-term follow up. J Neurosurg. 52. Phi JH, Kim DG, Chung HT, Lee J, Paek SH, Jung HW. Radiosurgical
2000;93 Suppl 3:42–6. treatment of vestibular schwannomas in patients with neurofibro-
34. Wang LW, Shiau CY, Chung WY, Wu HM, Guo WY, Liu KD, et al. matosis type 2: tumor control and hearing preservation. Cancer.
Gamma Knife surgery for low-grade astrocytomas: evaluation of 2009;115(2):390–8. Epub 2008/12/26.
long-term outcome based on a 10-year experience. J Neurosurg. 53. Wentworth S, Pinn M, Bourland JD, Deguzman AF, Ekstrand K,
2006;105(Suppl):127–32. Ellis TL, et al. Clinical experience with radiation therapy in the
35. Park KJ, Kano H, Kondziolka D, Niranjan A, Flickinger JC, management of neurofibromatosis-associated central nervous sys-
Lunsford LD. Early or delayed radiosurgery for WHO grade II tem tumors. Int J Radiat Oncol Biol Phys. 2009;73(1):208–13.
astrocytomas. J Neurooncol. 2011;103(3):523–32. Epub 2008/08/09.
36. Kobayashi T, Kida Y, Mori Y, Hasegawa T. Long-term results of 54. Meijer OW, Vandertop WP, Lagerwaard FJ, Slotman BJ. Linear
gamma knife surgery for the treatment of craniopharyngioma in 98 accelerator-based stereotactic radiosurgery for bilateral vestibular
consecutive cases. J Neurosurg. 2005;103(6 Suppl):482–8. schwannomas in patients with neurofibromatosis type 2.
37. Xu Z, Yen CP, Schlesinger D, Sheehan J. Outcomes of Gamma Knife Neurosurgery. 2008;62(5 Suppl):A37–42; discussion A-3. Epub
surgery for craniopharyngiomas. J Neurooncol. 2011;104(1):305–13. 2008/07/17.
38. Ulfarsson E, Lindquist C, Roberts M, Rahn T, Lindquist M, Thoren 55. Mathieu D, Kondziolka D, Flickinger JC, Niranjan A, Williamson R,
M, et al. Gamma knife radiosurgery for craniopharyngiomas: long- Martin JJ, et al. Stereotactic radiosurgery for vestibular schwanno-
term results in the first Swedish patients. J Neurosurg. 2002; mas in patients with neurofibromatosis type 2: an analysis of tumor
97(5 Suppl):613–22. control, complications, and hearing preservation rates. Neurosurgery.
39. Amendola BE, Wolf A, Coy SR, Amendola MA. Role of radiosur- 2007;60(3):460–8; discussion 8–70; Epub 2007/03/01.
gery in craniopharyngiomas: a preliminary report. Med Pediatr 56. Rowe JG, Radatz MW, Walton L, Soanes T, Rodgers J, Kemeny
Oncol. 2003;41(2):123–7. AA. Clinical experience with gamma knife stereotactic radiosurgery
40. Chung WY, Pan DH, Shiau CY, Guo WY, Wang LW. Gamma knife in the management of vestibular schwannomas secondary to
radiosurgery for craniopharyngiomas. J Neurosurg. 2000;93 Suppl type 2 neurofibromatosis. J Neurol Neurosurg Psychiatry. 2003;
3:47–56. 74(9):1288–93. Epub 2003/08/23.
41. Niranjan A, Kano H, Mathieu D, Kondziolka D, Flickinger JC, 57. Kida Y, Kobayashi T, Tanaka T, Mori Y. Radiosurgery for bilateral
Lunsford LD. Radiosurgery for craniopharyngioma. Int J Radiat neurinomas associated with neurofibromatosis type 2. Surg Neurol.
Oncol Biol Phys. 2009;78(1):64–71. 2000;53(4):383–9; discussion 9–90; Epub 2000/05/29.
Pediatric Disorders:
Stephanie L. Da Silva and Mark D. Krieger
Introduction Surgical Management Overview: Pediatric

Brain Tumors
The pediatric brain is different from the adult brain in many
ways. Plasticity, myelination, and tolerance to radiation dos- Pediatric brain tumors are extremely heterogeneous. A team
ages vary greatly with age. Furthermore, pediatric disorders approach is usually taken and includes the pediatric neuro-
of the brain differ significantly from those in adults. These surgeon, neuro-oncologist, radiation oncologist, psycholo-
considerations are important in the discussion of the use of gist, and radiologist among others. In cases of low grade
stereotactic radiosurgery in children. This chapter will focus tumors, surgical intervention is recommended to obtain a tis-
on the treatment of tumors and vascular malformations of the sue sample for diagnosis and to maximally resect the tumor
pediatric brain from a surgical perspective. while limiting morbidity. However, surgical management
In contrast to brain tumors representing only 1–2 % of may not be the best method of treatment for highly invasive,
all new cancers in adults [1], pediatric tumors of the central malignant tumors or benign tumors that may infiltrate the
nervous system represent 20 % of all childhood cancers visual pathways of a child. Factors such as age, histology,
with an incidence of 2.5 per 100,000 children under the age and tumor location are generally considered in determining
of 15 [2]. Central nervous system (CNS) tumors are the the appropriate treatment, though in certain incidences these
most common solid tumors in children and second most fre- factors may be weak predictors of outcome.
quent of childhood malignancies after leukemia. Differences
are also apparent between pediatric and adult patients in the
tumors’ sites of origins, histological features, and clinical Low-Grade Tumors
presentation. A majority of childhood brain tumors origi-
nate infratentorially in the cerebellum, brainstem, or fourth Low-Grade Gliomas
ventricular region, while adult tumors have a proclivity to
arise in the cerebral cortex. Pathology of pediatric brain The most common primary central nervous system malig-
tumors varies by age, with the heterogeneity of these tumors nancies affecting the pediatric population are low-grade
attributed to their unique natural history, treatment options, gliomas, with an annual incidence of 0.6 cases per 100,000
and prognosis. Surgery and extent of resection plays a large children under 18 years of age [3]. Fortunately, the prognosis
role in the management of these tumors. Radiation therapy for children with these tumors is excellent; long-term
is used more cautiously given concerns about its effect progression-free survival (PFS) is reported as high as 80–90 %
on the developing brain and on patients with long life at 10 years [3]. Juvenile pilocytic astrocytomas (JPAs), the
expectancies. most common of these low-grade gliomas in children, are
characterized by early age of onset, lack of invasiveness, and
favorable prognosis in patients [4]. Cerebral hemispheric
and cerebellar JPAs are optimally treated by surgical resec-
tion as these tumors tend to be well-circumscribed with rea-
sonably well-defined borders [5]. As these tumors are usually
S.L. Da Silva • M.D. Krieger, M.D. (*) curable upon complete removal, surgical resection is not
Division of Neurosurgery, Children’s Hospital Los Angeles,
1300 N. Vermont Ave., Los Angeles, CA 90027, USA generally followed by adjuvant therapy. Even with incom-
e-mail: mkrieger@chla.usc.edu plete tumor resection, pilocytic astrocytomas are associated
416 S.L. Da Silva and M.D. Krieger
with prolonged survival. In a series of 361 pediatric patients of brain tumors in children [14, 15], are epithelial tumors
with low-grade gliomas (63 % pilocytic astrocytomas), varying in size that tend to arise in the sellar or suprasellar
Armstrong et al. [6] determined the 10-and 20-year overall location. Surgery is indicated in mostly all cases, with the
survival rates to be 87 % and 82 %, respectively. However, goal being to establish a tissue diagnosis and to remove as
prolonged survival in this series did not come without a cost, much tumor as is safely possible by limiting morbidity. With
especially when these tumors are closely associated with the technological advancements in surgical tools including
critical brain structures. A substantial number of these long- ultrasonic aspiration, image-guiding techniques, and neuro-
term survivors experienced poor outcome in other areas, such endoscopy, surgical success in resecting craniopharyngio-
as cognitive deficit, hearing loss, blindness, and endocrine mas generated a wave of enthusiasm through publications by
abnormalities. In patients with JPAs that cannot be totally Hoffman et al. [16], Choux and Lena [17], Pierre-Kahn et al.
resected or exhibit tumor regrowth, adjuvant therapy is needed. [18], Caldarelli et al. [19], and Zuccaro [20]. This enthusi-
Though radiation therapy may be used for older children, it is asm, however, was short-lived, as associated mortality was
undesirable in young patients due to conventional side effects reported to occur in up to 50 % at 10 years and was followed
of radiation, and in these patients chemotherapy is utilized by a high rate of recurrence (50 % in some cases) [21].
more frequently. For patients with newly diagnosed unresect- Following aggressive resection, complications ensued as
able, progressive, or recurrent JPAs, stereotactic radiosurgery patients suffered peri-operative injury to the hypothalamus
(SRS) is now considered a valuable option to control tumor leading to memory, behavioral, and endocrine disorders.
progression or long-term remission. As reported by Kano et al. As injury to the hypothalamus was found to be the main fac-
[5] in a series of 50 pediatric patients, newly diagnosed patients tor associated with morbidity, surgical treatment modalities
receiving SRS who had either biopsy or subtotal resection had were developed to avoid hypothalamic dysfunction by
a significantly better progression free survival at 10 years when approaching the tumor via the trans-nasal route [21]. A pre-
compared to patients who underwent delayed SRS for recur- operative grading system was devised by Puget et al. [22] to
rent tumors after initial total or subtotal resection (STR) balance the benefits of aggressive surgical resection with the
(88.9 % and 44.5 %, respectively, p = 0.049). risk of significant morbidity in children, dividing tumors into
While the mainstay of treatment for low-grade gliomas is three types based on extent of invasion of the hypothalamus.
maximal surgical resection, tumors that arise within critical Gross total resection (GTR) was proposed to be attempted in
brain regions such as the brain stem or basal ganglia are gener- type 0 tumors (representing no hypothalamic involvement)
ally not resectable without the risk of significant morbidity. and type 1 tumors (distorting or elevating the hypothalamus).
In these gliomas, minimally invasive neurosurgical techniques In type 2 tumors, in which the hypothalamus is no longer
including stereotactic biopsy for diagnosis and endoscopic visible, STR is recommended, leaving only the hypothalamic
third ventriculostomy for cerebrospinal fluid (CSF) diversion component. Radiation therapy would be the preferred
have been beneficial for surgical management [7–9]. Appro- therapy. In a series by Jane et al. [23] in 2010, among 22
priate adjuvant therapy is given once a diagnosis has been pediatric patients with some sellar component to their cra-
made. For tumors considered to be accessible, however, surgery niopharyngioma, transsphenoidal resection resulted in a high
should be pursued. In St. Jude’s experience of 52 patients with rate of GTR with a low risk of recurrence (2 of 15 patients
low-grade focal brainstem gliomas, 48 % of patients had surgi- after GTR). It was noted, however, that a majority of cases
cal resection as the primary treatment with a 5- and 10-year approached via this route were infra diaphragmatic in
PFS and overall survival of 59 %/98 % and 52 %/90 %, respec- location [23].
tively [10]. These results are comparable to other studies who In cases where gross total resection is not safely possible,
found 5-year survival rates between 87 and 100 % [11, 12], subtotal resection may be followed by radiation therapy.
though disease progression is not uncommon in these tumors. Patients in either case are followed by serial MRIs and clini-
Yen et al. [13] reported their use of Gamma knife surgery as an cal monitoring of pituitary and visual functioning. In patients
alternative to surgical resection for 20 patients with focal brain- with residual disease who had previously undergone subtotal
stem gliomas, finding that the tumors disappeared in 4 patients resection or in those with recurrence following an initially
and shrank in 12. Thus, they concluded that gamma knife sur- thought GTR, stereotactic radiotherapy, both in the form
gery may be effective as a primary treatment or adjunct to open of radiosurgery or fractionated stereotactic radiotherapy,
surgery for focal brainstem gliomas [13]. is increasingly showing encouraging results. Kiehna and
Merchant [24] recently published a review of the literature
for studies including radiation therapy for pediatric cranio-
Craniopharyngioma pharyngioma, reporting modern studies with 90 % disease
control in a 5-year follow-up and more than 85 % favorable
In pediatric neurosurgery, the surgical management of cra- functional outcome [22, 24–27]. However, long-term follow-
niopharyngiomas is one of the most highly controversial top- up beyond 5–10 years is necessary to assess tumor control
ics. Craniopharyngiomas, comprising approximately 5–10 % relative to quality of life and functional outcome as these
31 Pediatric Disorders: Viewpoint—Surgery 417
Fig. 31.1 Sagittal contrast-enhanced T1-weighted MRI scan showing Fig. 31.2 On last follow-up 1 year after Gamma knife radiosurgery for
a large cystic craniopharyngioma residual tumor, the patient remains stable and the tumor controlled
children remain vulnerable to late treatment failures and non-NF-1 children, suggesting a more favorable prognosis in
side effects from radiation therapy [24] (Case Study 31.1, the context of NF-1 [33]. Radiation therapy in patients with
Figs. 31.1 and 31.2). NF-1 is limited due to the increased risk of second malignan-
cies [34] and radiation-related cerebral vasculopathies [35].
Sharif et al. [34] reported a high incidence of second nervous
Neurofibromatosis-1 (NF-1) system tumors in NF1 OPG patients, with an increased risk
in irradiated patients (4 malignant peripheral-nerve sheath
Neurofibromatosis-1, or von Recklinghausen NF, is one of tumors in 12 irradiated NF-1 OPG patients vs. 3/48 NF-1
the most common autosomal-dominant CNS disorders with controls).
an incidence of about 1 per 2,500–3,000 live births [28, 29].
About half of all cases are inherited; the other half arise de
novo. The most common genetic abnormality associated Neurofibromatosis-2 (NF-2)
with NF results in the loss of a tumor suppressor gene. NF-1
patients are prone to the development of pilocytic astrocyto- Neurofibromatosis-2, or bilateral acoustic NF, is more rare
mas and gliomas involving the optic pathway in children than NF-1 with an incidence of 1 in 25,000 live births [36–38].
[30]. An estimated 15–20 % of children with NF-1 are Though NF-2 is often associated with adult onset, Evans
afflicted with an optic pathway glioma (OPG), with a major- et al. [39] showed that at least 18 % of NF-2 patients present
ity of symptomatic children diagnosed before the age of 6 [3, in childhood and are likely to have a more severely debilitat-
31, 32]. Rodriguez et al. [30] reported that in NF-1 patients ing disease course. Furthermore, 10–18 % of children pre-
with pilocytic astrocytomas and indeterminate subtype low- senting with a meningioma or schwannoma are likely to have
grade astrocytoma, extent of resection was not significantly NF-2 with prolonged follow-up [39]. Eighth nerve dysfunc-
associated with recurrence, radiographic progression, or tion is a much more common symptom in children than in
tumor location. However, patients who undergo GTR have a adults. Nunes and MacCollin [40] reported that hearing loss
better overall survival (OS) rate than those who undergo STR owing to vestibular tumors affected 9/12 children and 11/24
or biopsy (100 % vs. 52 % and 65 % at 10 years, respec- ears by age 18 years. Treatment for NF-2 is primarily surgi-
tively) [30]. Chemotherapy is considered a first-line therapy cal; optimal management includes presymptomatic resection
in children with NF-1 and afflicted with low-grade gliomas. for vestibular schwannomas [40]. Vestibular schwannomas
A Children’s Oncology Group phase III trial of 401 children in NF-2 patients are more difficult to treat than those with
treated with different chemotherapy regimens revealed sporadic unilateral VS given the association with the eighth
5 year OS rates of 98 % for patients with NF-1 vs. 86 % for cranial nerve (vestibulocochlear) and as such surgery almost
always leads to loss of the cochlear nerve and total deafness primitive neuroectodermal tumors (PNETs). GTR is achieved
[41]. Depending on the location of the tumor, surgical resec- less for eloquent brain areas such as the thalamus and the
tion via a middle fossa or retrosigmoid approach with partial basal ganglia than noneloquent areas such as the superficial
or complete tumor removal may preserve hearing [42]. SRS cortical locations not associated with essential functions
is a controversial treatment of vestibular schwannomas in (63 % vs. 91 %, respectively) [54].
NF-2 and is currently being extensively debated [43]. While
radiosurgery may preserve hearing function and arrest tumor
growth, the presence of acoustic neuroma is still of threat to Glioblastoma
the patient [42].
Glioblastomas are the most malignant astrocytic neoplasm
and characteristically have zones of coagulation necrosis and
Meningioma areas of vascular hyperplasia [55]. These tumors represent
3 % of CNS tumors in children with a 5-year survival of
Meningiomas are rare in children, accounting for 0.4–4.6 % about 20 % [56]. Due to the aggressive nature of glioblasto-
of all CNS tumors in the pediatric population [44, 45]. mas, prognosis is poor; the median survival of these patients
As previously discussed, meningiomas within the pediatric is only 12 months [55]. Glioblastomas are more commonly
age group are often the first sign of neurofibromatosis 2 [46]. found in the superficial cortex than in deep structures [57],
Pediatric meningiomas are most commonly intracranial and often as a consequence of intratumoral hemorrhage,
(90 %), followed by intraspinal (6 %), intraorbital (2 %), and approximately 5–10 % of these children present as the result
elsewhere (2 %). These tumors are frequently located in the of sudden deterioration in their neurological status [58]. In
cerebral convexity, followed by intraventricular, falcine, order to significantly increase overall survival of the patient,
parasagittal, anterior, and middle cranial fossa [47]. Perry gross total resection should be the goal. However, this may
et al. [48] noted that meningiomas in the pediatric population be more feasible in children than adults. In a study of 37
have a high frequency of brain invasion and are more pheno- pediatric patients with glioblastoma not of the brainstem,
typically and genotypically aggressive when compared Yang et al. [57] reported a median OS of 45.1 months in
to adults. Surgical management is difficult in children due to patients with GTR vs. <1 year in patients who had STR or
location of tumors, larger size at presentation, relatively less biopsy only. Furthermore, at 5 years, 42 % of patients who
blood volume in children, and risks inherent in long opera- had GTR were still alive, whereas none of the patients with-
tions such as massive blood transfusions and hypothermia out GTR survived [57]. Extent of resection but not location
[49]. The goal of treatment is considered gross total resec- of the lesion was reported as an independent predictive factor
tion with resection of the dural origin or attachment recom- of OS in a multivariate analysis performed by Song et al.
mended to lower the risk of recurrence [45, 50]. In the largest [59]. Though radiation and chemotherapy exist as current
series of 87 pediatric patients by Rushing et al. [50], mortal- treatment options, glioblastomas have an inherent resistance
ity rate and recurrence rate were reported as 10.6 % and of conventional therapy. Therefore, SRS is thought to be
19.4 %, respectively. There is limited data on the use of SRS more advantageous due to its improved set-up accuracy in
in children, reported only in case reports. In one such report, treating recurrent glioblastomas [60]. However, to date there
SRS was used to successfully treat a child with recurrent is no prospective study reviewing the efficacy of SRS in
meningioma in association with meningioangiomatosis [51]. pediatric patients with glioblastoma.
Recurrent meningioma has also been successfully treated in
three children using Gamma knife radiosurgery [52, 53].
Ependymoma
High-Grade Tumors Ependymomas are the third most common intracranial tumors
in children, representing up to 10 % of all CNS tumors in the
Treatment for high-grade tumors without brain stem invol- pediatric age range with a peak incidence between birth and
vement in children is generally approached by surgical 4 years [61]. The greatest determinant of survival in children
resection followed by adjuvant therapy. The main goals of with ependymoma is extent of resection. Supratentorial epen-
surgery for these malignant tumors include obtaining a his- dymomas, constituting about 30 % of ependymomas in chil-
tological diagnosis and reducing the burden of the tumor by dren [62], have a better prognosis than infrantentorial
decreasing mass effect. High-grade tumors that have the ependymomas [63–65]. GTR of supratentorial ependymomas
treatment option of stereotactic radiosurgery following surgi- is more likely to be achieved when compared to those in the
cal resection include glioblastomas, ependymomas, medul- infratentorial location [66, 67], especially those which do not
loblastomas, ependymoblastomas, pineoblastomas, and CNS invade the ventricular system [68, 69]. In these patients with
100% Following maximal safe resection, postoperative MRI should

90% be performed to radiologically confirm the extent of resec-
80% tion. If there is residual tumor found with no dissemination,
70% second-look surgery is strongly advised. Even with maximal
resection, however, recurrence rates may be as high as 50 %.
% survival
60%
50%
Adjuvant radiation therapy of 54–55.8 Gy to the tumor
bed is a commonly accepted postsurgical treatment for chil-
40%
dren over the age of 3 years old. One series from UCSF
30%
GTR showed that radiation therapy has been found to improve
20%
STR PFS even after adjusting for extent of resection and also
10% interestingly in those who received GTR when compared to
0% those without radiation therapy [70]. The adverse effects of
0 6 12 18 24 30 36 42 48
radiation on children under 3 years of age have traditionally
Months after surgery
limited this type of treatment, though a series by Koshy et al.
Fig. 31.3 Percent progression-free survival vs. months after surgery [71] used data from Surveillance Epidemiology and End
for ependymoma patients following GTR and STR Results of 184 children younger than 3 years to show
that postoperative radiation therapy significantly improved
100% 3-year overall survival when compared with those who did
90%
not receive RT (81 % and 58 %, respectively, p = 0.005).
Grundy et al. [72] reported in his series of 89 children
80%
younger than 3 years that primary postoperative chemother-
70%
apy led to a 5-year OS of 63.4 %, without radiation therapy
% survival
60%
in 42 % of those treated for nonmetastatic disease. Currently
50% under clinical evaluation is a randomized phase III trial
40% studying how well maintenance chemotherapy works in
30% comparison with observation following induction chemo-
20% GTR therapy and radiation therapy in treating newly diagnosed
10% STR pediatric ependymoma.
0% Patients with recurrent ependymomas may also want to
0 6 12 18 24 30 36 42 48 consider stereotactic radiosurgery. In evaluating the utility
Months after surgery of SRS in treating childhood ependymomas, two factors
must be considered: efficacy in disease control and toxicity
Fig. 31.4 Percent overall survival vs. months after surgery for ependy-
moma patients following GTR and STR
[73]. In 2008, Merchant et al. [74] described their expe-
rience with reirradiation for childhood ependymomas at
St. Jude’s Hospital. Six of the 38 patients in their series were
supratentorial ependymomas who undergo complete resection treated with stereotactic radiosurgery for recurrence, with
and exhibit no evidence of metastasis, surgery alone may age at time of initial presentation ranging between 1 and 6
be acceptable treatment option and radiation therapy can be years. At the time of initial disease presentation, all patients
avoided without affected OS, as evidenced by Venkatramani received radiation therapy while three received chemother-
et al. [68]. apy. Notably, only two of the patients treated with SRS had
Gross total resection has been found to dramatically prior gross total resection. Using a median dose of 18 Gy,
increase OS and PFS. In one study at Children’s Hospital Los ranging between 15 and 20 Gy, only one of the six patients
Angeles, 28 children with a mean age of 43 months (range survived. Four patients had local progression, though two of
5–51 months) were surgically treated for ependymomas in a these patients had a longer PFS after SRS than after initial
10-year period [61]. Twenty-one (75 %) patients underwent adjuvant therapy. This finding may suggest that SRS is bet-
GTR, whereas 7 (25 %) underwent STR or biopsy. In a ter for temporizing local disease progression than standard
median follow-up period of 30 months (range 1–95 months), radiotherapy when repeat surgery is not feasible. An even
the 28 patients had 3- and 5-year PFS rates of 21.4 % and more encouraging result came from Liu et al. [75], who
10.7 %, respectively, and 3- and 5-OS rates of 39.3 % used hypofractionated re-irradiation, “staged radiosurgery,”
and 17.9 %, respectively. Patients receiving radiographically to treat locally recurrent ependymomas in six children.
confirmed GTR had a much higher PFS and OS (33 and 44 All patients survived with no evidence of disease at median
months, respectively), than those receiving STR (6.5 and follow-up of 28 months from the time of re-irradiation
15 months, respectively), as shown in Figs. 31.3 and 31.4. (range 6–56 months), though it is important to note that five
Fig. 31.5 Axial contrast-enhanced T1-weighted MRI scan showing a Fig. 31.6 The patient remains stable with no evidence of disease 5 years
large right calcified and hemorrhagic paraventricular mass, diagnosed after surgical resection followed by intensity-modulated radiation therapy
after biopsy as a supratentorial ependymoma and concomitant chemotherapy
of the six patients had GTR upon disease recurrence prior 55–76 %, respectively [84]. Gamma knife surgery (GKS) has
to re-irradiation. Following re-irradiation, three of the six been found in anecdotal case series to have potential benefits
patients were reported to have a longer second disease-free especially for young children unsuitable for initial or addi-
interval than the time to initial progression. In both the tional fractionated radiation therapy, which carries the risk
Merchant and Liu series, however, evidence of radiation of neurotoxicity and poor neurocognitive outcome [85, 86].
necrosis was described in a few patients and thus remains a Flannery et al. [87] showed that patients who had a smaller
risk in radiosurgical treatment (Case Study 31.2, Figs. 31.5 tumor volume were more likely to respond to GKS treat-
and 31.6). ment. The longest survival, 126 and 67 months, occurred in
two patients who had repeat maximal resection followed by
early GKS.
PNET (Medulloblastoma, Ependymoblastoma,
Pineoblastoma)
Medulloblastoma
Although there is some controversy surrounding the histo-
pathological classification of childhood CNS embryonal Medulloblastomas account for 20 % of pediatric CNS tumors
tumors such as CNS primitive neuroectodermal tumors and 40 % of posterior fossa tumors in children, with an inci-
(PNETs), medulloblastoma, ependymoblastoma, and pineo- dence of 0.6 per 100,000 cases [88]. These tumors are most
blastoma (usually grouped with tumors of the pineal region), commonly found in fourth ventricle and tend to invade nor-
determining the appropriate surgical management plan and mal cerebellar tissue, with 15 % of cases infiltrating the brain
treatment is similar. Dissemination at the time of diagnosis stem [89]. Medulloblastomas are often seen as a midline
for medulloblastomas and pineoblastomas to other intracra- hyperdense cerebellar vermian mass upon review on a CT
nial sites and spinal cord is reported to occur in 10–30 % scan [90], and ventriculomegaly is noted in approximately
of patients and results in a worse prognosis [76–81]. CNS 85–90 % of all cases [91, 92]. Though the peak incidence of
PNETs account for 4.8 % of childhood brain tumors [82]. children diagnosed with medulloblastomas is between 5 and
Even with multimodal treatment, including maximal safe 7 years, 10 % of cases are diagnosed in the first year of life
resection, fractionated neuraxis radiation therapy, and sys- [93]. Patients with these tumors can be divided into standard
temic chemotherapy, local and metastatic recurrences occur risk and high risk. Standard risk patients are those over the
in 30–40 % of pediatric patients [83]. Five-year survival age of 3, have no evidence of neuraxis dissemination on the
for standard-risk and high-risk patients are 70–80 % and brain and spine MRI and lumbar CSF, and have less than
1.5 cm2 of residual tumor on post-MRI scans. In contrast,

high-risk patients are those with disseminated disease, ana-
plastic histology, postoperative CSF cytology, and a post-op
MRI showing greater than 1.5 cm2 residual tumor [93, 94].
For children with medulloblastoma, surgery is usually the
primary indication for treatment to remove as much of the
tumor as safely possible while also confirming histological
diagnosis. Maximal surgical resection has been associated
with improved rate of survival, chiefly in those with no
evidence of dissemination at the time of diagnosis [95, 96].
In infants, treatment is particularly challenging due to the
vascular nature of tumor, high incidence of leptomeningeal
spread, and lower chance of achieving a gross total resection
[93]. Patrice et al. [97] was the first to propose the SRS tech-
nique as an effective and safe treatment in the management
of medulloblastoma, showing that SRS was capable of con-
trolling small, locally recurrent disease in 5 out of 11 patients
with a median survival from the time of SRS of 10 (range 5
to 59+) months. The six remaining patients died of progres-
Fig. 31.7 A sagittal contrast-enhanced T1-weighted image reveals a
sive disease. Furthermore, this series proposed that SRS was an large preoperative pineoblastoma in the pineal region
effective addition to craniospinal irradiation for patients with
newly diagnosed medulloblastoma, as all three patients were
alive at last follow-up without evidence of disease [97]. worse prognosis for children aged ≤5 years compared with
older children; these patients had a 5-year survival rate of
15 % and 57 %, respectively (p < 0.00001). Among all
Ependymoblastoma patients, 5-year survival rate increased with degrees of resec-
tion: 29 % for debulking, 53 % for STR, and 84 % for GTR
Ependymoblastomas are considered a subgroup of CNS [100]. For children under the age of 3 years, the prognosis is
PNETs, most frequently located in the supratentorial region, even worse, with a survival rate at 1 year of 0 % [102].
but may also arise in the infratentorial or spinal sites [98]. Radiotherapy may have a significant impact on survival.
These tumors are classified as WHO grade IV due to their A series by Hinkes et al. [102] indicated that six older children
aggressive behavior with rapid growth and proclivity to dis- (median 6.75 y, range 3.6–16.9 y) who underwent radiother-
seminate into the craniospinal region [99]. Gerber et al. [99] apy in a maintenance chemotherapy regimen brought all
reported in their series of 10 patients a median OS survival of those who had not been in complete remission into remission
12 months, with longer survival of those with radiotherapy and kept all others in remission with no evidence of remain-
or chemotherapy. If intraventricular chemotherapy was part ing tumor (Case Study 31.3, Figs. 31.7 and 31.8).
of treatment, a subcutaneous reservoir with an intraventricu-
lar catheter was implanted in the anterior horn of lateral ven-
tricle. In this series, 5 year PFS and OS were 36.4 ± 14.5 % Vascular Malformations: Arteriovenous
and 30.3 ± 15.9 %, respectively, though all patients eventu- Malformation
ally died of tumor progression [99].
The most dangerous congenital vascular malformations are
arteriovenous malformations (AVMs). Prevalence in adults
Pineoblastoma is about 18 per 100,000 and AVMs account for between 1
and 2 % of all strokes, 3 % of strokes in young adults, and
Pineoblastomas are pineal region tumors that can metasta- 9 % of subarachnoid hemorrhages [103]. AVMs usually
size along the neuroaxis and may be locally invasive, spread- present between ages 10 and 40 with the most common
ing outside the pineal region through the subarachnoid space symptoms being headache, seizures, focal neurologic defi-
[100]. These tumors are relatively rare, accounting for 0.6 % cits, and intracranial hemorrhage; however, at least 15 % of
of all childhood brain tumors [82]. The standard treatment people affected by AVMs are asymptomatic [103]. The long-
includes maximal surgical resection followed by adjuvant term crude annual case fatality rates are between 1 and 1.5 %
cranial-spinal irradiation and systemic chemotherapy [101]. [103]. After a hemorrhage, the annual mortality can be as
A literature review of patients with pineoblastoma revealed a high as 18 % [104]. Ninety percent of AVMs in the brain are
Fig. 31.8 The patient remains without residual tumor following surgi-
cal resection and chemotherapy at last follow-up
Fig. 31.9 An axial contrast-enhanced T2-weighted MRI shows the
appearance of a right temporal lobe hematoma secondary to underlying
located supratentorially, while the rest are located in the AVM
posterior fossa. The Spetzler-Martin Intracranial AVM grading
scale [105] was devised based upon size, location, and deep
venous drainage. The lesion grade (I–V) is derived by sum-
ming assigned points in each category. Complete surgical
excision of a grade I lesion (small <3 cm, located in a non-
eloquent region such as the anterior frontal lobe, and with
solely superficial drainage) would have very little risk of any
resulting morbidity or mortality. On the other hand, a grade
V lesion (larger than 6 cm, located within or adjacent to elo-
quent brain areas and with some drainage into the deep
venous system) would have significant morbidity and mor-
tality [105]. Surgical resection is the treatment of choice for
grade I–III AVMs, due to its high cure rate (reported from 89
to 98.4 %) and low morbidity and mortality [106, 107].
In children, AVMs are most frequently located in the basal
ganglia, thalamus, corpus callosum, brainstem, or within the
motor, speech, or visual cortex [108, 109] and are the most
frequent cause of intracranial hemorrhage in children [110–
112]. Mortality from AVM hemorrhage in the pediatric pop-
ulation is much greater for cerebellar AVM than for AVM
in the cerebral hemisphere (57 % to 4.5 %, respectively,
p < 0.0001) [113]. While surgery is the mainstay of treatment Fig. 31.10 A year after recurrence and subsequent Cyber knife radio-
for AVMs, endovascular embolization has become a useful therapy, an axial contrast-enhanced T2-weighted MRI scan shows no
evidence of residual AVM
additional technique along with radiosurgery as an alterna-
tive for high-risk lesions. After primary stereotactic radiosur-
gery, Pan et al. [114] reported a pediatric AVM obliteration Gamma knife radiosurgery, as Dinca et al. [115] reported
rate of 65 % at 48 months following the procedure. When 71.3 % for a one time treatment and 82.7 % overall. Thus,
needed, additional SRS led to a total obliteration rate of GKS is considered to be a valid active management option
81 % [114]. Kano et al. [110] showed an even higher total for pediatric AVM, though limitations include delayed effect
obliteration with SRS of 70 % with at least 4 years of with persistence of bleeding risk for 2–4 years posttreatment
follow-up. Similar obliteration rates were obtained with [115] (Case Study 31.4, Figs. 31.9 and 31.10).
Case Studies Case 31.3
A 2-year-old boy presented with signs and symptoms of

elevated intracranial pressure. Imaging studies revealed
Case 31.1 a large pineal region tumor (Fig. 31.7). It was felt not
consistent with a germinoma, and given the large size of
The patient is a 15-year-old young man, who presented the tumor and associated hydrocephalus, operative
with an indolent history of progressive visual loss and resection was indicated. Pathology came back consis-
headaches. Imaging studies revealed a large cystic cra- tent with pineoblastoma. The tumor was removed in its
niopharyngioma (Fig. 31.1). A pterional craniotomy entirety using microsurgical dissection techniques and
was then completed. Multiple cystic areas were resec- the Cavitron ultrasonic aspirator. He received postop-
ted and aspirated. This was a very prolonged and diffi- erative chemotherapy but has not needed radiation ther-
cult dissection given the tumor’s investment in the apy. The postoperative MRI of the brain does not show
surrounding optic chiasm, contralateral optic nerve, any evidence of residual tumor (Fig. 31.8).
and posterior fossa. Two years later, the patient had evi-
dence of tumor recurrence in the temporal lobe. Using
microsurgical dissection techniques, the multiple cysts
of the tumor were removed in this location. Two and a
half years after the second resection, there was again Case 31.4
evidence of recurrence, which required tumor removal Patient is a 7-year-old young man who was found to
in the medial portion of the sylvian fissure. Five months have a hemorrhage in his right temporal area. Imaging
later, he underwent Gamma knife radiosurgery. The studies were consistent with an AVM (Fig. 31.9) and
tumor remains controlled 1 year after radiosurgery at angiography confirmed the diagnosis. The lesion was
last follow-up (Fig. 31.2). not amenable to endovascular obliteration and resec-
tion was indicated. The frameless stereotactic system
was used to localize the AVM and microsurgical dis-
section was performed. An intraoperative angiogram
Case 31.2 showed no evidence of residual AVM. Four years later,
A 6-year-old young man presented with left-sided the AVM recurred and dissection was carried down
hemiparesis, headache, nausea and vomiting, and in a subtemporal fashion to coagulate and divide the
lethargy. Imaging studies revealed a right parietal abnormal arteries, then removed in their entirety in the
tumor that was later diagnosed as a supratentorial temporal lobe. Though the intraoperative angiogram
ependymoma (Fig. 31.5). The large hemorrhagic and was once again negative, there was recurrence a year
calcified tumor was removed in its entirety using later and he subsequently underwent Cyber knife
microsurgical dissection techniques and the Stealth radiotherapy. MRI scans did not show any residual
frameless stereotactic system to confirm complete lesion at last-follow-up a year later (Fig. 31.10).
resection. Postoperatively, the patient became pro-
gressively more lethargic and was found to need a
ventriculoperitoneal shunt to manage new onset
References
hydrocephalus. One year and eight months later, he
had a recurrent ventricular lesion that was resected 1. Surawicz TS, McCarthy BJ, Kupelian V, Jukich PJ, Bruner JM,
without evidence of residual tumor. The patient under- Davis FG, et al. Descriptive epidemiology of primary brain and
went intensity-modulated radiation therapy along CNS tumors: results from the Central Brain Tumor Registry of the
United States, 1990–1994. Neuro Oncol. 1999;1(1):14–25.
with concomitant chemotherapy. Two years later, he
2. Reisner A, Szerlip NJ, Chin LS. Pediatric radiosurgery. In: Chin
had a T10 spinal cord lesion, a pilomatrixoma of the LS, Regine W, editors. Principles and practice of stereotactic
thoracic spine and a L1–L2 spinal cord lesion within a radiosurgery. New York: Springer; 2008.
span of 9 months that were each resected. At last fol- 3. Oh KS, Hung J, Robertson PL, Garton HJ, Muraszko KM, Sandler
HM, et al. Outcomes of multidisciplinary management in pediat-
low-up 3 years later, the lesion has been stable with no
ric low-grade gliomas. Int J Radiat Oncol Biol Phys. 2011; 81(4):
changes in the fourth ventricular lesion and lumbar e481–8.
lesions on MRI of the brain (Fig. 31.6) and spine. 4. Forsyth PA, Shaw EG, Scheithauer BW, O’Fallon JR, Layton
Jr DD, Katzmann JA. Supratentorial pilocytic astrocytomas.
A clinicopathologic, prognostic, and flow cytometric study of 51 Research Hospital experience 1984–2001. Int J Radiat Oncol Biol
patients. Cancer. 1993;72(4):1335–42. Phys. 2002;53(3):533–42.
5. Kano H, Niranjan A, Kondziolka D, Flickinger JC, Pollack IF, 27. Lin LL, El Naqa I, Leonard JR, Park TS, Hollander AS, Michalski
Jakacki RI, et al. Stereotactic radiosurgery for pilocytic astrocyto- JM, et al. Long-term outcome in children treated for craniopha-
mas. Part 2: Outcomes in pediatric patients. J Neurooncol. 2009; ryngioma with and without radiotherapy. J Neurosurg Pediatr.
95(2):219–29. 2008;1(2):126–30.
6. Armstrong GT, Conklin HM, Huang S, Srivastava D, Sanford R, 28. Williams VC, Lucas J, Babcock MA, Gutmaan DH, Korf B, Maria
Elison DW, et al. Survival and long-term health and cognitive out- BL. Neurofibromatosis type 1 revisited. Pediatrics. 2009;123(1):
comes after low-grade glioma. Neuro Oncol. 2011;13(2):223–34. 124–33.
7. Minturn JE, Fisher MJ. Gliomas in children. Curr Treat Options 29. Huson SM, Compston DA, Clark P, Harper PS. A genetic study of
Neurol. 2013;15(3):316–27. von Recklinghausen neurofibromatosis in south east Wales.
8. Selvapandian S. Endoscopic management of thalamic gliomas. I. Prevalence, fitness, mutation rate and effect of parental trans-
Minim Invasive Neurosurg. 2006;49(4):194–6. mission on severity. J Med Genet. 1989;26(11):704–11.
9. Li KW, Roonprapunt C, Lawson HC, et al. Endoscopic third ven- 30. Rodriguez FJ, Perry A, Gutmann DH, O’Neill BP, Leonard J,
triculostomy for hydrocephalus associated with tectal gliomas. Bryant S, et al. Gliomas in neurofibromatosis type 1: a clinico-
Neurosurg Focus. 2005;18(6A):E2. pathologic study of 100 patients. J Neuropathol Exp Neurol.
10. Kilmo P, Pai Panadiker AS, Thompson CJ, Boop FA, Qaddoumi I, 2008;67(3):240–9.
Gajjar A, et al. Management and outcome of focal low-grade 31. Listernick R, Charrow J, Greenwald M, et al. Natural history of
brainstem tumors in pediatric patients: the St. Jude experience. optic pathway tumors in children with neurofibromatosis type 1:
J Neurosurg Pediatr. 2013;11(3):274–81. a longitudinal study. J Pediatr. 1994;125(1):63–6.
11. Sandri A, Sardi N, Genitori L, Giordano F, Peretta P, Basso ME, 32. Listernick R, Ferner RE, Liu GT, et al. Optic pathway gliomas
et al. Diffuse and focal brain stem tumors in childhood: prognostic in neurofibromatosis-1: controversies and recommendations. Ann
factors and surgical outcome. Experience in a single institution. Neurol. 2007;61(3):189–98.
Childs Nerv Syst. 2006;22(9):1127–35. 33. Ater J, Holmes E, Zhou T, Mazewski C, Roberts W, Vezina G,
12. Teo C, Siu TL. Radical resection of focal brainstem gliomas: is it et al. Results of COG protocol A9952: a randomized phase 3 study
worth doing? Childs Nerv Syst. 2008;24(11):1307–14. of two chemotherapy regimens for incompletely resected low-
13. Yen CP, Sheenah J, Steiner M, Patterson G, Steiner L. Gamma grade glioma in young children (LGG 18). Neuro Oncol. 2008;
knife surgery for focal brainstem gliomas. J Neurosurg. 2007; 10(3):451.
106(1):8–17. 34. Sharif S, Ferner R, Birch JM, Gillespie JE, Gattamaneni HR,
14. Jane Jr JA, Laws ER. Craniopharyngioma. Pituitary. 2006;9(4):323. Baser ME, et al. Second primary tumors in neurofibromatosis 1
15. Garre ML, Cama A. Craniopharyngioma: modern concepts in patients treated for optic glioma: substantial risks after radiother-
pathogenesis and treatment. Curr Opin Pediatr. 2007;19(4):471–9. apy. J Clin Oncol. 2006;24(16):2570–5.
16. Hoffman HJ, De Silva M, Humphreys RP, Drake JM, Smith ML, 35. Grill J, Couanet D, Cappelli C, Habrand JL, Rodriguez D, Sainte-
Blaser SI. Aggressive surgical management of craniopharyngio- Rose C, et al. Radiation-induced cerebral vasculopathy in children
mas in children. J Neurosurg. 1992;76(1):47–52. with neurofibromatosis and optic pathway glioma. Ann Neurol.
17. Choux M, Lena G. Bases of surgical management of craniopha- 1999;45(3):393–6.
ryngioma in children [proceedings]. Acta Neurochir Suppl (Wien). 36. Evans DG, Moran A, King A, Saeed S, Gurusinghe N, Ramsden
1979;28(2):348. R. Incidence of vestibular schwannoma and neurofibromatosis
18. Pierre-Kahn A, Brauner R, Renier D, Sainte-Rose C, Gangemi 2 in the North West of England over a 10-year period: higher inci-
MA, Rappaport R, Hirsch JF. Treatment of craniopharyngiomas in dence than previously thought. Otol Neurotol. 2005;26(1):93–7.
children. Retrospective analysis of 50 cases. Arch Fr Pediatr. 37. Evans DG, Howard E, Giblin C, Clancy T, Spencer H, Huson SM,
1988;45(3):163–7. Lalloo F. Birth incidence and prevalence of tumor-prone syn-
19. Caldarelli M, di Rocco C, Papacci F, Colosimo Jr C. Management dromes: estimates from a UK family genetic register service. Am
of recurrent craniopharyngioma. Acta Neurochir (Wien). 1998; J Med Genet A. 2010;152A(2):327–32.
140(5):447–54. 38. Antinheimo J, Sankila R, Carpén O, Pukkala E, Sainio M,
20. Zuccaro G. Radical resection of craniopharyngioma. Childs Nerv Jääskeläinen J. Population-based analysis of sporadic and type 2
Syst. 2005;21(8–9):679–90. neurofibromatosis-associated meningiomas and schwannomas.
21. Puget S. Treatment strategies in childhood craniopharyngioma. Neurology. 2000;54(1):71–6.
Front Endrocrinol (Lausanne). 2012;3:64. 39. Evans DGR, Birch JM, Ramsden RT. Paediatric presentation of
22. Puget S, Garnett M, Wray A, Grill J, Habrand JL, Bodaert N, et al. type 2 neurofibromatosis. Arch Dis Child. 1999;81(6):496–9.
Pediatric craniopharyngiomas: classification and treatment accord- 40. Nunes F, MacCollin M. Neurofibromatosis 2 in the pediatric pop-
ing to the degree of hypothalamic involvement. J Neurosurg. 2007; ulation. J Child Neurol. 2003;18(10):718–24.
106(1 Suppl):3–12. 41. Evans DG. Neurofibromatosis type 2 (NF2): a clinical and molec-
23. Jane Jr JA, Prevedello DM, Alden TD, Laws Jr ER. The transsphe- ular review. Orphanet J Rare Dis. 2009;4:16.
noidal resection of pediatric craniopharyngiomas: a case series. 42. Subach BR, Kondziolka D, Lunsford LD, Bissonette DJ,
J Neurosurg Pediatr. 2010;5(1):49–60. Flickinger JC, Maitz AH. Stereotactic radiosurgery in the man-
24. Kiehna EN, Merchant TE. Review radiation therapy for pediatric agement of acoustic neuromas associated with neurofibromatosis
craniopharyngioma. Neurosurg Focus. 2010;28(4):E10. Type 2. J Neurosurg. 1999;90(5):815–22.
25. Merchant TE, Kiehna EN, Kun LE, Mulhern RK, Li CH, Xiong X, 43. Mathieu D, Kondziolka D, Flickinger JC, Niranjan A, Williamson
et al. Phase II trial of conformal radiation therapy for pediatric R, Martin JJ, et al. Stereotactic radiosurgery for vestibular schwan-
patients with craniopharyngioma and correlation of surgical nomas in patients with neurofibromatosis type 2: an analysis of
factors and radiation dosimetry with change in cognitive function. tumor control complications and hearing preservation rates.
J Neurosurg. 2006;104(2 Suppl):94–102. Neurosurgery. 2007;60(3):460–8; discussion 468–70.
26. Merchant TE, Kiehna EN, Sanford RA, Mulhern RK, Thompson 44. Di Rocco C, Di Rienzo A. Meningiomas in childhood. Crit Rev
SJ, Wilson MW, et al. Craniopharyngioma: the St. Jude Children’s Neurosurg. 1999;9(3):180–8.
45. Kotecha RS, Pascoe EM, Rushing EJ, Rorke-Adams LB, 63. Perilongo G, Massimino M, Sotti G, Belfontali T, Masiero L,
Zwerdling T, Gao X, et al. Meningiomas in children and adoles- Rigobello L, et al. Analyses of prognostic factors in a retrospec-
cents: a meta-analysis of individual patient data. Lancet Oncol. tive review of 92 children with ependymoma: Italian Pediatric
2011;12(13):1229–39. Neuro-oncology Group. Med Pediatr Oncol. 1997;29(2):79–85.
46. Ruggieri M, Iannetti P, Polizzi A, Mantia IL, Spalice A, Giliberto 64. Pollack IF, Gerszten PC, Martinez AJ, Lo KH, Shultz B, Albright
O, et al. Earliest clinical manifestations and natural history of neu- AL, et al. Intracranial ependymomas of childhood: long-term out-
rofibromatosis type 2 (NF2) in childhood: a study of 24 patients. come and prognostic factors. Neurosurgery. 1995;37(4):655–66;
Neuropediatrics. 2005;36(1):21–34. discussion 666–7.
47. Santos MV, Furlanetti L, Valera ET, Brassesco MS, Tone LG, de 65. Grill J, Le Deley M, Gambarelli D, Raguin MA, Couanet D,
Oliveira RS. Pediatric meningiomas: a single-center experience Pierre-Kahn A, et al. Postoperative chemotherapy without irradia-
with 15 consecutive cases and review of the literature. Childs Nerv tion for ependymoma in children under 5 years of age: a multi-
Syst. 2012;28(11):1887–96. center trial of the French Society of Pediatric Oncology. J Clin
48. Perry A, Giannini C, Raghavan R, Scheithauer BW, Banerjee R, Oncol. 2001;19(5):1288–96.
Margraf L, et al. Aggressive phenotypic and genotypic features in 66. Robertson PL, Zeltzer PM, Boyett JM, Rorke LB, Allen JC, Geyer
paediatric and NF2-associated meningiomas: a clinic-pathologic JR, et al. Survival and prognostic factors following radiation ther-
study of 53 cases. J Neuropathol Exp Neurol. 2001;60(10): apy and chemotherapy for ependymomas in children: a report of
994–1003. the Children’s Cancer Group. J Neurosurg. 1998;88(4):695–703.
49. Menon G, Nair S, Sudhir J, Rao BR, Mathew A, Bahuleyan 67. Bouffet E, Perilongo G, Canete A, Massimino M. Intracranial
B. Childhood and adolescent meningiomas: a report of 38 cases ependymomas in children: a critical review of prognostic factors
and review of literature. Acta Neurochir. 2009;151(3):239–44; and a plea for cooperation. Med Pediatr Oncol. 1998;30(6):319–29;
discussion 244. discussion 329–31.
50. Rushing EJ, Olsen C, Mena H, Rueda ME, Lee YS, Keating RF, 68. Venkatramani R, Dhall G, Patel M, Grimm J, Hawkins C,
et al. Central nervous system meningiomas in the first two decades McComb G, Krieger M, et al. Supratentorial ependymoma in chil-
of life: a clinicopathological analysis of 87 patients. J Neurosurg. dren: to observe or to treat following gross total resection? Pediatr
2005;103(6 Suppl):489–95. Blood Cancer. 2012;58(3):380–3.
51. Kim NR, Choe G, Shin SH, Wang KC, Cho BK, Choi KS, et al. 69. Ernestus RI, Wilcke O, Schroder R. Supratentorial ependymomas
Childhood meningiomas associated with meningioangiomatosis: in childhood: clincopathological findings and prognosis. Acta
report of five cases and literature review. Neuropathol Appl Neurochir. 1991;111(3–4):96–102.
Neurobiol. 2002;28(1):48–56. 70. Pejavar S, Polley MY, Rosenberg-Wohl S, Chennupati S, Prados
52. Im SH, Wang KC, Kim SK, Oh CW, Kim DG, Hong SK, et al. MD, Berger MS, et al. Pediatric intracranial ependymoma: the
Childhood meningioma: unusual location, atypical radiological roles of surgery, radiation and chemotherapy. J Neurooncol. 2012;
findings and favorable treatment outcome. Childs Nerv Syst. 106(2):367–75.
2001;17(11):656–62. 71. Koshy M, Rich S, Merchant TE, Mahmood U, Regine WF, Kwok
53. Li X, Zhao J. Intracranial meningiomas of childhood and adoles- Y. Post-operative radiation improves survival in children younger
cence: report of 34 cases with follow-up. Childs Nerv Syst. 2009; than 3 years with intracranial ependymoma. J Neurooncol.
25(11):1411–7. 2011;105(3):583–90.
54. Sawaya R, Hammoud M, Schoppa D, Hess KR, Wu SZ, Shi W-M, 72. Grundy RG, Wilne SA, Weston CL, Robinson K, Lashford LS,
et al. Neurosurgical outcomes in a modern series of 400 cranioto- Ironside J, et al. Primary postoperative chemotherapy without radio-
mies for treatment of parenchymal tumors. Neurosurgery. therapy for intracranial ependymoma in children: the UKCCSG/
1998;42(5):1044–55; discussion 1055–6. SIOP prospective study. Lancet Oncol. 2007;8(8):696–705.
55. Khalatbari MR, Hamidi M, Moharamzad Y. Glioblastoma multi- 73. Krieger MD, McComb JG. The role of stereotactic radiotherapy in
forme with very rapid growth and long-term survival in children: the management of ependymomas. Childs Nerv Syst. 2009;25(10):
report of two cases and review of the literature. Childs Nerv Syst. 1269–73.
2011;27(8):1347–52. 74. Merchant TE, Boop FA, Kun LE, Sanford RA. A retrospective
56. CBTRUS 2011 statistical report. Primary brain and central ner- study of surgery and reirradiation for recurrent ependymoma. Int J
vous system tumors diagnosed in the United States in 2004–2007. Radiat Oncol Biol Phys. 2008;71(1):87–97.
Available from: www.cbtrus.org/reports and tables. 75. Liu AK, Foreman NK, Gaspar LE, Trinidad E, Handler
57. Yang T, Temkin N, Barber J, Geyer JR, Leary S, Browd S, MH. Maximally safe resection followed by hypofractionated re-
Ojemann JG, et al. Gross total resection correlates with long-term irradiation for locally recurrent ependymoma in children. Pediatr
survival in pediatric patients with glioblastoma. World Neurosurg. Blood Cancer. 2009;52(7):804–7.
2013;79(3–4):537–44. 76. Fouladi M, Gajjar A, Boyett JM, Walter AW, Thompson SJ,
58. Tamber MS, Rutka JT. Pediatric supratentorial high-grade gliomas. Merchant TE, et al. Comparison of CSF cytology and spinal mag-
Neurosurg Focus. 2003;14(2):e1. netic resonance imaging in the detection of leptomeningeal disease
59. Song KS, Phi JH, Cho BK, Wang KC, Lee JY, Kim DG, et al. in pediatric medulloblastoma or primitive neuroectodermal tumor.
Long-term outcomes in children with glioblastoma. J Neurosurg J Clin Oncol. 1999;17(10):3234–7.
Pediatr. 2010;6(2):145–9. 77. Zeltzer PM, Boyett JM, Finlay JL, Albright AL, Rorke LB,
60. Biswas T, Okunieff P, Schell MC, Smudzin T, Pilcher WH, Bakos Milstein JM, et al. Metastasis stage, adjuvant treatment, and resid-
RS, et al. Stereotactic radiosurgery for glioblasotma: a retrospec- ual tumor are prognostic factors for medulloblastoma in children:
tive analysis. Radiat Oncol. 2009;4:11. conclusions from the Children’s Cancer Group 921 randomized
61. Krieger MD, Bowen IE. Effects of surgical resection and adjuvant phase III study. J Clin Oncol. 1999;17(3):832–45.
therapy on pediatric intracranial ependymomas. Expert Rev 78. Yao MS, Mehta MP, Boyett JM, Li H, Donahue B, Rorke LB,
Neurother. 2005;5(4):465–71. et al. The effect of M-stage on patterns of failure in posterior fossa
62. McGuire CS, Sainani KL, Fisher PG. Incidence patterns for epen- primitive neuroectodermal tumors treated on CCG-921: a phase
dymoma: a surveillance, epidemiology, and end results study. III study in a high-risk patient population. Int J Radiat Oncol Biol
J Neurosurg. 2009;110(4):725–9. Phys. 1997;38(3):469–76.
79. Jakacki RI, Zeltzer PM, Boyett JM, Albright AL, Allen JC, Geyer and residual medulloblastoma. Pediatr Neurosurg. 1995;22(4):
JR, et al. Survival and prognostic factors following radiation and/ 197–203.
or chemotherapy for primitive neuroectodermal tumors of the 98. Louis DN, Ohgaki H, Wiestler O, Cavenee WKE. WHO classifi-
pineal region in infants and children: a report of the Childrens cation of tumors of the central nervous system. Lyon: IARC Press;
Cancer Group. J Clin Oncol. 1995;13(6):1377–83. 2007.
80. Cohen BH, Zeltzer PM, Boyett JM, Geyer JR, Allen JC, Finlay JL, 99. Gerber NU, von Hoff K, von Bueren AO, Treulieb W, Warmuth-
et al. Prognostic factors and treatment results for supratentorial Metz M, Pietsch T, et al. Outcome of 11 children with ependymo-
primitive neuroectodermal tumors in children using radiation and blastoma treated within the prospective HIT-trials between 1991
chemotherapy: a Childrens Cancer Group randomized trial. J Clin and 2006. J Neurooncol. 2011;102(3):459–69.
Oncol. 1995;13(7):1687–96. 100. Tate M, Sughrue ME, Rutkowski MJ, Kane AJ, Aranda D,
81. Reddy AT, Janss AJ, Phillips PC, Weiss HL, Packer RJ. Outcome McClinton L, et al. The long-term postsurgical prognosis of
for children with supratentorial primitive neuroectodermal tumors patients with pineoblastoma. Cancer. 2012;118(1):173–9.
treated with surgery, radiation, and chemotherapy. Cancer. 2000; 101. Gilheeney SW, Saad A, Chi S, Turner C, Ullrich NJ, Goumnerova
88(9):2189–93. L, et al. Outcome of pediatric pineoblastoma after surgery, radia-
82. Kaatsch P, Rickert CH, Kuhl J, Schuz J, Michaelis J. Population- tion and chemotherapy. J Neurooncol. 2008;89(1):89–95.
based epidemiologic data on brain tumors in German children. 102. Hinkes BG, von Hoff K, Deinlein F, Warmurth-Metz M, Soerensen
Cancer. 2001;92(12):3155–64. N, Timmermann B, et al. Childhood pineoblastoma: experiences
83. Gold DR, Packer RJ, Cohen BH. Treatment strategies for medul- from the prospective multicenter trials HIT-SKK87, HIT-SKK92
loblastoma and primitive neuroectodermal tumors. Neurosurg and HIT 91. J Neurooncol. 2007;81(2):217–23.
Focus. 1999;7(2):E1. 103. AL-Shahi R, Warlow C. A systematic review of the frequency and
84. Rossi A, Caracciolo V, Russo G, Reiss K, Giordano A. Medullo- prognosis of arteriovenous malformations of the brain in adults.
blastoma: from molecular pathology to therapy. Clin Cancer Res. Brain. 2001;124(Pt 10):1900–26.
2008;14(4):971–6. 104. Brown Jr RD, Wiebers DO, Torner JC, O’Fallow WM. Frequency
85. Fouladi M, Gilger E, Kocak M, Wallace D, Buchana G, Reeves C, of intracranial hemorrhage as a presenting symptom and subtype
et al. Intellectual and functional outcome of children 43 years old analysis: a population-based study of intracranial vascular malfor-
or younger who have CNS malignancies. J Clin Oncol. 2005; mations in Olmstead Country, Minnesota. J Neurosurg. 1996;
23(28):7152–60. 85(1):29–32.
86. Kiltie AE, Lashford LS, Gattamaneni HR. Survival and late effects 105. Spetzler RF, Martin NA. A proposed grading system for arteriove-
in medulloblastoma patients treated with craniospinal irradiation nous malformations. J Neurosurg. 2008;108(1):186–93.
under three years old. Med Pediatr Oncol. 1997;28(5):348–54. 106. Kiris T, Sencer A, Sahinbas M, Sencer S, Imer M, Izgi N. Surgical
87. Flannery T, Kano H, Martin JJ, Niranjan A, Flickinger JC, results in pediatric Spetzler-Martin grades I–III intracranial arte-
Lunsford LD, et al. Boost radiosurgery as a strategy after failure of riovenous malformations. Childs Nerv Syst. 2005;21(1):69–74.
initial management of pediatric primitive neuroectodermal 107. Schaller C, Schramm J. Microsurgical results for small arteriove-
tumors. J Neurosurg Pediatr. 2009;3(3):205–10. nous malformations accessible for radiosurgical or embolization
88. CBTRUS, Central Brain Tumor Registry of the United States. treatment. Neurosurgery. 1997;40(4):664–72; discussion 672–4.
Primary brain tumors in the United States, statistical report (2000– 108. Di Rocco C, Tamburrini G, Rollo M. Cerebral arteriovenous
2004), years of data collected. Chicago: Author; 2007–2008. malformations in children. Acta Neurochir (Wien). 2000;142(2):
89. Jenkin O. The radiation treatment of medulloblastoma. J Neurooncol. 145–56; discussion 156–8.
1996;29(1):45–54. 109. Perini S, Zampieri P, Rosta L, Piovan E, Micheli ED, Nicolato A,
90. Nejat F, El Khashab M, Rutka JT. Initial management of child- et al. Endovascular treatment of pial AVMS: technical operations,
hood brain tumors: neurosurgical considerations. J Child Neurol. indications and limits in pediatric age patients. J Neurosurg Sci.
2008;23(10):1136–48. 1997;41(4):325–30.
91. Crawford JR, MacDonald TJ, Packer RJ. Medulloblastoma in 110. Kano H, Kondziolka D, Flickinger JC, Yang HC, Flannery TJ,
childhood: new biological advances. Lancet Neurol. 2007;6(12): Awan NR, et al. Stereotactic radiosurgery for arteriovenous mal-
1073–85. formations, part 2: management of pediatric patients. J Neurosurg
92. Packer RJ, Cogen P, Vezina G, Rorke LB. Medulloblastoma: clinical Pediatr. 2012;9(1):1–10.
and biologic aspects. Neuro Oncol. 1999;1(3):232–50. 111. Hladky JP, Lejeune JP, Blond S, Pruvo JP, Dhellemmes P. Cerebral
93. Dhall G. Medulloblastoma. J Child Neurol. 2009;24(11): arteriovenous malformations in children: report on 62 cases. Childs
1418–30. Nerv Syst. 1994;10(5):328–33.
94. Packer RJ, Rood BR, MacDonald TJ. Medulloblastoma: present 112. Nicolato A, Foroni R, Seghedoni A, Martines V, Lupidi F,
concepts of stratification into risk groups. Pediatr Neurosurg. Zampieri P, et al. Leksell gamma knife radiosurgery for cerebral
2003;39(2):60–7. arteriovenous malformations in pediatric patients. Childs Nerv
95. Albright AL, Wisoff JH, Zeltzer PM, Boyett JM, Rorke LB, Syst. 2005;21(4):301–7; discussion 308.
Stanley P. Effects of medulloblastoma resections on outcome in 113. Kondziolka D, Humphreys RP, Hoffman HJ, Hendrick EB, Drake
children: a report from the Children’s Cancer Group. Neurosurgery. JM. Arteriovenous malformations of the brain in children: a forty
1996;38(2):265–71. year experience. Can J Neurol Sci. 1992;19(1):40–5.
96. Taylor RE, Bailey CC, Robinson K, Weston CL, Ellison D, 114. Pan DH, Kuo YH, Guo WY, Chung WY, Wu HM, Liu KD, et al.
Ironside J, et al. Results of a randomized study of preradiation Gamma knife surgery for cerebral arteriovenous malformations in
chemotherapy versus radiotherapy alone for nonmetastatic medul- children: a 13-year experience. J Neurosurg Pediatr. 2008;1(4):
loblastoma: The International Society of Paediatric Oncology/ 296–304.
United Kingdom Children’s Cancer Study Group PNET-3 Study. 115. Dinca EB, de Lacy P, Yianni J, Rowe J, Radatz MW, Preotiuc-
J Clin Oncol. 2003;21(8):1581–91. Pietro D, Kemeny AA. Gamma knife surgery for pediatric arterio-
97. Patrice SJ, Tarbell NJ, Goumnerova LC, Shrieve DC, Black PM, venous malformations: a 25-year retrospective study. J Neurosurg
Loeffler JS. Results of radiosurgery in the management of recurrent Pediatr. 2012;10(5):445–50.
Pediatric Disorders: Viewpoint—
Fractionated Radiotherapy 32
Thomas E. Merchant and Erin S. Murphy
require further surgery, radiation therapy or both. Based on

Introduction published reports demonstrating major reductions in acute
side effects associated with radiation therapy for adult can-
Childhood cancer is the most common cause of disease cers, 3-dimensional (3D) radiation was introduced incre-
related death among children ages 0–19 years. It has not mentally to the pediatric population. For pediatric patients
been defined as a public health problem because fewer than with CNS tumors, age has been the major driving force
12,500 cases are diagnosed each year. CNS tumors represent behind the inclusion or exclusion of radiation therapy and
21 % of childhood cancers or about 2,700 cases per year [1]. the sequencing of treatment. The age minimum of 3 years
This rare group of tumors is comprised of astrocytoma 56 %, has often been used to determine whether a child should
embryonal tumors including medulloblastoma, supratento- receive radiation therapy as part of their frontline treatment.
rial primitive neuroectodermal tumor, and pineoblastoma The embryonal tumors and ependymoma serve as pertinent
20 %, ependymoma 7 %, craniopharyngioma 3 %, germ cell examples. For other tumors, such as low-grade astrocytoma,
tumors 3 %, and tumors involving the spinal cord, pituitary, the age at which radiation therapy has been considered
and choroid plexus 11 %. acceptable has varied from 3 years to 10 years of age. For
As recent as 15 years ago, radiotherapy avoidance was CNS germ cell tumors, craniopharyngioma and high-grade
the primary objective of clinical trials developed for CNS glioma, age has been considered important factor, although
tumors in children. Most treatment morbidity was attributed most patients with these diagnoses tend to present at an
to radiation therapy and there was a major focus on cogni- older age.
tive effects and second malignancies. These concerns were With the objective of reintroducing radiation therapy for
justified based on the results of the time and the available the treatment of young children, defined by the National
treatment. After decades of pursuing radiotherapy avoid- Cancer Institute as children under the age of 3 years, and
ance, investigators found that side effects remained despite improving outcomes for all patients, investigators have
the omission of radiation therapy and follow-up reports looked at ways in which the total prescribed dose might be
demonstrated that disease control was inferior without radi- reduced, the prescribed volume of radiation might be lim-
ation therapy as a component of frontline management. ited, or how newer methods of focal radiation delivery
Additionally, functional outcomes were relatively unchanged might be implemented to increase conformity of the high-
among long-term survivors: patients would pursue radio- est isodose volumes. To achieve the aforementioned goals,
therapy avoidance, experience side effects associated with clinical trials have been designed to test the safety of lower
those treatments, suffer progression of disease, and eventually doses and reduced target volumes for a variety of tumors
and to study radiation-related treatment effects focusing on
objective measures including neurologic, endocrine, and
T.E. Merchant, D.O., Ph.D. (*) cognitive function. Because newer methods of radiation
Department of Radiological Sciences, St. Jude Children’s Research therapy planning yield information about the 3D distribu-
Hospital, 262 Danny Thomas Place, Memphis, TN 38105, USA
e-mail: thomas.merchant@stjude.org tion of dose in normal tissues, the long-term goal of many
investigations has been to correlate treatment dosimetry
E.S. Murphy, M.D.
Department of Radiation Oncology, Cleveland Clinic, with functional outcomes and to model the effects of radiation
Cleveland, OH 44106, USA dose on normal tissues.
428 T.E. Merchant and E.S. Murphy
Contrasting Fractionated Irradiation Medulloblastoma

and Radiosurgery
Historically, the treatment for medulloblastoma has consisted
Although this chapter is meant to focus on the applications of surgery, postoperative radiation therapy, and during the
of fractionated irradiation, the use of radiosurgery in pediat- past 15 years, systematic post-irradiation chemotherapy for
ric CNS tumors requires comment. Radiosurgery is often those older than the age of 3 years at the time of diagnosis.
proposed as an alternative to fractionated irradiation for pri- Surgery followed by chemotherapy, in an effort to delay or
mary pediatric CNS tumors based on the assumption that it avoid radiation therapy, has been the path followed by most
would be less toxic. Most investigators would agree that children under the age of 3 years.
indications exist for the use of radiosurgery in the frontline For decades, radiation therapy for older children con-
adjuvant treatment of craniopharyngioma and certain low- sisted of craniospinal irradiation to 36 Gy followed by
grade astrocytomas, supplemental boost treatment in con- “boost” treatment of the anatomic posterior fossa to 54 Gy.
junction with fractionated irradiation of pineal region These guidelines were developed in an era where adjuvant
tumors, and recurrent or metastatic disease. The usual cave- therapy consisted of radiation therapy alone and they
ats regarding tolerance of critical normal tissue structures remained the standard until it was eventually demonstrated
apply, namely, tolerance of optic chiasm, nerves, brain stem, that the craniospinal dose could be reduced for low-risk
and diencephalic structures to high-dose single-fraction patients (i.e., no evidence of neuraxis dissemination and
irradiation. residual tumor measuring <1.5 cm2) with the addition of
The true value of radiosurgery in the treatment of pediatric multiagent chemotherapy [6]. Patients with neuraxis dissem-
patients is unknown because of a lack of prospective experi- ination or substantial residual tumor continue to receive
ence and long-term outcomes. For example, the published craniospinal irradiation to dose levels ≥36 Gy despite the
series that include pediatric patients with pilocytic astrocyto- addition of multiagent chemotherapy.
mas are all retrospective with median follow-up ranging from
2 to 8 years, patient numbers ranging from 12 to 50, a variable
number of patients who had undergone prior radiotherapy, Evolution of Treatment and Impact
marginal prescription dose ranging from 10 to 15 Gy, and on Cognitive Effects
no consensus of appropriate target volume [2–5]. The same
groups reported rates of tumor cyst expansion ranging from Fear of craniospinal irradiation and its effects on cognitive
11 to 25 %. Due to the lack of common inclusion criteria, function have always been and remain justifiable. The
target volume, and prescription dose, the applicability of this expectation for a child who receives craniospinal irradiation
data to patient treatment strategies is difficult. at an early age is a loss of IQ at a rate of approximately 3.9
The potential benefits of radiosurgery are often overshad- points per year without plateau over the time period of the
owed by poor patient selection or when long-term therapeu- reported data, about 6 years [7]. These data were derived
tic goals or the natural history of the tumor is not considered. from the St. Jude experience that included children initially
For the treatment of pediatric CNS tumors, radiosurgery has treated postoperatively with chemotherapy that eventually
not been proved to be equivalent to surgery or fractionated required conventional dose craniospinal irradiation.
irradiation in terms of outcomes and side effects. Questionable Craniospinal doses of 35.2 Gy and posterior fossa doses of
applications of radiosurgery include the following: low-dose 54–55.8 Gy were prescribed for these patients with a median
(<12 Gy) single-fraction irradiation in lieu of a full-dose age of approximately 3.2 years. Until this article by Walter
fractionated treatment course regardless of tumor size; irra- et al. [7], the data in the available literature was not mod-
diation of a tumor bed without an identifiable target; target- eled, included few subjects, did not include acceptable fol-
ing macroscopic residual tumor within a tumor bed known to low-up, and suffered from significant attrition. Working on
contain microscopic disease with no plan for fractionated the assumption that the craniospinal dose and treatment vol-
irradiation of the tumor bed; primary radiosurgery in lieu of ume were largely responsible for the effects observed in
craniospinal irradiation for seeding tumors; radiosurgery in these children, investigators sought ways to reduce the dose
lieu of relatively simple surgery when feasible; radiosurgery of craniospinal irradiation from 36 to 23.4 Gy for selected
to the solid component of a cystic/solid tumor complex; average-risk patients using adjuvant chemotherapy. In the
boost treatment with radiosurgery with subtherapeutic dose largest study reported for that time, and using longitudinal
attenuation of the fractionated component. Finally, the pedi- psychology data from the CCG-9892 trial, Ris et al. [8]
atric caregiver requires advance notice that imaging changes showed a decline in IQ of 4.3 points per year among patients
are prone to occur after radiosurgery that should not be inter- between the ages of 3 and 7 years treated with craniospinal
preted as progressive disease. irradiation to 23.4 Gy, posterior fossa irradiation to 55.8 Gy,
32 Pediatric Disorders: Viewpoint—Fractionated Radiotherapy 429
Fig. 32.1 (a) Classic lateral portal film outlining the posterior fossa boost for a patient with medulloblastoma (lower): composite craniospinal,
posterior fossa, and primary site dose distribution for a standard-risk patient with medulloblastoma treated on the St. Jude SJMB96 protocol
and adjuvant chemotherapy. In other words, reducing the using a similarly defined clinical and planning target volume
craniospinal dose to 23.4 Gy did not improve cognitive out- margins. This small and step-wise reduction in primary site
comes in young patients with average-risk medulloblas- targeting did not affect the rate or patterns of failure for
toma. This lack of improvement should have come as no patients with average-risk medulloblastoma. The 3-year
surprise given the volume of normal brain subtended by the event-free survival was reported at 78 ± 6 %. Although the
treatment of the entire posterior fossa. It has been estimated, gross tumor volume was significantly smaller than the ana-
depending on the individual patient, that nearly 40 % of the tomic posterior fossa in these patients, with the additional
entire brain receives the prescription dose of 54–55.8 Gy 2 cm clinical target volume margin and 0.3–0.5 cm planning
when the posterior fossa is the target and conventional radi- target volume margin, the final target approached, and in
ation therapy planning is used. some cases exceeded, the anatomic posterior fossa for these
The more recent focus for investigators has been the tar- same patients. The volumetric details from this study pro-
get volume for the “boost” component of treatment after cra- vided some insight into the effect of the actual postoperative
niospinal irradiation in an attempt to alter treatment from the primary site volume, location, and targeting guidelines on
entire posterior fossa to the primary site with a limited mar- our ability to further reduce the volume of irradiation after
gin. Although there are a number of single institution reports the craniospinal component. Despite the small reduction
that appear to have accomplished this goal, the largest study in the volume of irradiation, the decline in IQ for patients
of its kind was the SJMB96 trial conducted at St. Jude with average-risk medulloblastoma, ages 3–7 years, was
Children’s Research Hospital and collaborating centers at significantly improved to approximately –2.4 points per year.
Texas Children’s Hospital, the Royal Children’s Hospital in Indeed, the psychology data from this study forms the largest
Melbourne, and the New Children’s Hospital in Sydney from data set for patients treated with craniospinal irradiation for
1996 to 2003 [9] (Fig. 32.1). medulloblastoma and provides data from which parents and
This study included patients with a variety of embryonal caregivers may estimate the effects of craniospinal irradia-
tumors and presentations including average and high-risk tion. In summary, eliminating posterior fossa irradiation and
medulloblastoma, supratentorial PNET, and atypical teratoid treating the primary site with a limited margin after cranio-
rhabdoid tumor. All patients received similar post-irradiation spinal irradiation does improve cognitive outcomes for
high-dose chemotherapy with peripheral stem cell support patients with medulloblastoma.
and all patients received postoperative craniospinal irradia- The next step taken by the St. Jude consortium and sepa-
tion that was uniquely risk adapted. For example, patients rately by the Children’s Oncology Group was further dose
classified as average-risk (no evidence of metastases, resid- and volume reduction. The recently completed St. Jude trial
ual tumor <1.5 cm2), regardless of embryonal tumor type or known as SJMB03 (clinicaltrials.gov, NCT00085202)
location, received 23.4 Gy craniospinal irradiation, 36 Gy includes identical risk-based craniospinal dosing and
posterior fossa irradiation (infratentorial tumors only), and post-irradiation chemotherapy as the SJMB96 trial with
primary site irradiation using an anatomically confined 2 cm newer targeting guidelines for the primary site boost. The
clinical target volume margin surrounding the postopera- clinical target volume margin is an anatomically confined
tive tumor bed and residual tumor and 0.3–0.5 cm planning margin of 1 cm surrounding the postoperative tumor bed and
target volume margin. Patients with high-risk tumors received residual tumor. The planning target volume margins remain
36–39.6 Gy craniospinal irradiation and primary site irradiation 0.3–0.5 cm. The Children’s Oncology Group is conducting
Fig. 32.2 Schema from

the COG standard risk
medulloblastoma trial
ACNS0331 (clinicaltrials.
gov, NCT00085735)
a medulloblastoma trial (ACNS 0331, clinicaltrials.gov, • Will 18 Gy craniospinal irradiation prove to be as

NCT00085735) for patients 3–8 years of age and 8–21 years effective (similar disease control) as 23.4 Gy craniospinal
of age (Fig. 32.2). The younger children are enrolled on an irradiation?
ambitious four-arm trial with two steps of randomization. • Are the volumetric differences between the different
Postoperatively patients are randomized to 23.4 Gy cranio- treatment arms sufficiently large to detect a difference in
spinal irradiation with concurrent vincristine versus 18 Gy cognitive or audiometric outcomes?
craniospinal irradiation and boost treatment of the anatomic Figure 32.3 shows a graphical representation of the dose
posterior fossa to 5.4 Gy (total 23.4 Gy posterior fossa) and distribution for the total brain volume that would be expected
concurrent vincristine. Patients are further randomized to for a typical patient planned and treated using conformal
primary site treatment that may include 55.8 Gy to the con- radiation therapy on the Children’s Oncology Group trial.
ventional posterior fossa volume or 55.8 Gy conformal pri- There are several regions of these curves that are notable
mary site irradiation using a 1.5 cm clinical target volume moving from lowest to highest doses. The total brain is
margin. All patients receive post-irradiation chemotherapy expected to receive the entire craniospinal dose as prescribed,
consisting of cisplatin, CCNU, cyclophosphamide, and vin- 18 or 23.4 Gy. The dose to the total brain falls steeply and the
cristine every six weeks for a total of nine cycles. The che- curves separate according to those who receive conventional
motherapy regimen was designed based on the equivocal posterior fossa irradiation (parallel opposed portals in the
results from the A9961 intergroup study that compared plati- example) or those who receive conformal treatment of the
num, CCNU, and vincristine to cyclophosphamide, cisplatin, primary site—who tend to have an increase in the volume
and vincristine. The older patients all receive 23.4 Gy cranio- that receives the intermediate doses. The curves come
spinal irradiation and concurrent vincristine followed by a together at the point where 30 % of the total brain receives
randomization to 55.8 Gy to the conventional posterior fossa approximately 40 Gy, and for the remainder of the high dose
volume or 55.8 Gy conformal treatment of the primary site region, the curves diverge based on conventional targeting
using a 1.5 cm clinical target volume margin. They will and treatment of the posterior fossa versus conformal treat-
then receive post-irradiation chemotherapy identical to the ment of the primary site. These curves demonstrate the volu-
younger patients. metric effects of craniospinal dose reduction, the volumetric
A number of important questions should be answered effects of conformal irradiation, and finally volumetric
after completion of the Children’s Oncology Group study in effects of the reduction in the targeted volume from the ana-
late 2013: tomic posterior fossa to the primary site with the prescribed
• In the context of conventional chemotherapy is it possible clinical target volume margin of 1.5 cm. Based on the
to eliminate irradiation of the entire anatomic posterior improvement shown for the SJMB96 trial over conventional
fossa? radiation therapy, one would anticipate an improved outcome
Fig. 32.3 Graphical

representation of the
whole-brain dose
distribution for
representative cases
planned for treatment on
each of the four arms of
the COG standard-risk
medulloblastoma trial
ACNS0331 (clinicaltrials.
gov, NCT00085735)
for those treated with 18 Gy craniospinal irradiation and of CNS necrosis at 5 years was 3.7 % (SD, 1.3 %) [13].
perhaps those treated on the other arms. Those patients for whom CNS necrosis was observed had the
Emerging data strengthen the association between the dis- largest infratentorial volumes receiving dose in excess of 50,
tribution of dose to regional volumes of normal brain and 52, and 54 Gy, providing new limits of tolerance for children
decline in cognitive function in several domains. The negative with medulloblastoma and other CNS embryonal tumors
impact on IQ and academic test scores associated with increas- treated with aggressive surgery, craniospinal irradiation, and
ing mean dose was shown for the entire brain volume, supra- intensive chemotherapy.
tentorial brain volume, left and right temporal lobes, and left
hippocampus (IQ, math, and reading scores) [10]. Figure 32.4
shows example IQ scores modeled based on SJMB96 data. Endocrine Effects
Proton therapy is an excellent measure to reduce extra-CNS
and CNS dose in these patients: Photon dose volume data Baseline and prospective endocrine evaluations were per-
associated with a decline in IQ are graphically inferior to formed for children on the risk-adapted SJMB96 trial
proton dose volume data for similar patients [11]. described above. The 4-year cumulative incidence of growth
Early data from a prospective proton therapy trial for hormone deficiency, thyroid-stimulating hormone (TSH)
patients with medulloblastoma with a median follow-up of deficiency, adrenocorticotropic hormone (ACTH) defi-
3.2 years demonstrates a 3-year progression-free survival of ciency, and primary hypothyroidism was 93 ± 4 %, 23 ± 8 %,
83 % and 76 %, respectively, for standard- and high-risk 38 ± 6 %, and 65 ± 7 %, respectively [14]. Radiation dosime-
children [12]. Fifty-one children underwent neurocognitive try to the hypothalamic-pituitary axis was associated only
evaluations at baseline and 31 received follow-up testing. with the development of TSH deficiency; the 4-year cumula-
Full scale, verbal, and performance IQ evaluations were not tive incidence was 44 ± 19 % and 11 ± 8 % for those receiving
significantly different, but there was a decline in processing a median dose to the hypothalamus of <42 Gy compared
speed. Long-term neurocognitive outcome data is awaited to to ≥42 Gy, respectively.
be able to compare proton data with that of modern photon
data for which we have longer follow-up.
Treatment for Young Children
Neurovascular Effects Reintroducing radiation therapy to very young children

under the age of 3 years with embryonal tumors has meant
An important and devastating side effect of radiation therapy omitting craniospinal irradiation from their treatment
is CNS necrosis. Among 148 patients who received 23.4 Gy regimen. The obvious advantages of CSI elimination, com-
craniospinal irradiation and 88 patients who received more bined with conformal radiation therapy (CRT) of the poste-
than 36 Gy craniospinal irradiation, the cumulative incidence rior fossa and/or primary site, are apparent. The Children’s
Fig. 32.4 Modeled IQ scores in children with medulloblastoma treated with craniospinal irradiation on the SJMB96 trial (clinicaltrials.gov,
NCT00003211)
Oncology Group Study P9934 demonstrated the advantage methotrexate to an intensive induction chemotherapy regi-
of the addition of CRT to postoperative chemotherapy when men of vincristine, etoposide, cyclophosphamide, and cispl-
compared postoperative chemotherapy alone for children atin. The induction regimen is followed by second surgery if
with nonmetastatic medulloblastoma between the ages of 8 less than a complete response or consolidation chemotherapy
and 36 months [15]. Seventy-four children on this trial with stem cell rescue. Radiotherapy is not a part of this clini-
underwent maximal surgical resection followed by four cal trial but age, risk, and response-adapted guidelines are
cycles of induction chemotherapy (cyclophosphamide, vin- suggested after the completion of consolidation. CSI is
cristine, cisplatin, and etoposide) followed by age and avoided for patients without metastatic disease and the target
response-adjusted CRT to the posterior fossa (18 or 23.4 Gy) volume includes the tumor bed with a 1 cm clinical target
and tumor bed (cumulative 50.4 or 54 Gy) and maintenance volume margin with a total dose of 50 Gy (45 Gy < 24 month)
chemotherapy (four alternating cycles of cyclophosphamide for complete response or 54 Gy (50 Gy < 24 month) for par-
and vincristine followed by oral etoposide). The 4-year EFS tial response or stable disease. Patients presenting with meta-
and OS was 50 ± 6 % and 69 ± 5.5 %, respectively, which static disease are recommended to receive reduced dose CSI
compared favorably to results from Pediatric Oncology of 18 Gy followed by the same tumor bed boost as above.
Group (POG) 9233 trial which included postoperative mul-
tiagent chemotherapy alone and resulted in EFS and OS of
25 ± 5.5 % and 46 ± 6 %, respectively. It is important to note Primary Brain Tumors That Require Focal
that the P9934 study included more patients with a gross Irradiation
total resection and desmoplastic or nodular histology, which
are known to be good prognostic factors compared to POG The ICRU-50 Report Guidelines published in 1993 [16] and
9233. Of note there was no significant or consistent decline more recently the ICRU-62 [17] Report Guidelines pub-
in either cognitive or motor function that could be attributed lished in 1999 have been taken to heart by those involved in
to the chemotherapy or CRT. The majority of children did the design and conduct of brain tumor trials for children. The
exhibit a significant delay in development at baseline, most definitions of the gross tumor volume (GTV), clinical target
likely due to the tumor itself. volume (CTV), and planning target volume (PTV) follow
The current trial sponsored by the Children’s Oncology closely those of the ICRU with one notable exception. The
Group (ACNS0334, clinicaltrials.gov, NCT00336024) for GTV includes the tumor bed in addition to the residual
children with high-risk primitive neuroectodermal tumors tumor. In the ICRU-50 report, the tumor bed is not meant as
randomizes children under the age of 3 years who have part of the GTV and the CTV margin is the margin surround-
undergone maximal surgical resection to the addition of ing the GTV in consideration of microscopic extension.
Because margins are not assessed for brain tumors, all mar- tissue volume in question and when using advanced
gins of the tumor bed are considered to be at risk and likely modeling procedures, one may integrate dose over speci-
to harbor microscopic disease with extension into surround- fied intervals and treat these data as clinical variables in a
ing normal tissues. Until the day when neurosurgeons biopsy multivariate analysis. Most modeling procedures have used
the operative cavity and the pathologists determine margins specific doses chosen along the dose volume curves and
or the neuroradiologists have the means to identify subclini- entered these values into one of the many models meant to
cal microscopic disease, the margin of the tumor bed will be determine the probability of a specific effect.
considered to be the basic volume at risk and defined as In consideration of radiation-related CNS effects, four
the GTV. important points should be understood: (1) pre-irradiation
The governing principals of target volume definitions and morbidity has not been systematically studied; (2) most
beam’s eye view treatment planning, whether forward or reports describing radiation-related side effects are outdated;
inverse, are similar in children and adults. After the target (3) most studies include small numbers of subjects and have
volume definition, the next step in the planning process is the large rates of attrition; and (4) radiation has been treated as a
definition of normal tissue volumes or so-called critical categorical variable in most analyses. The governing hypoth-
structures. These volumes take on additional meaning in the esis for current clinical trials involving children with CNS
treatment of younger patients where there is an incentive to tumors that require only focal irradiation is that irradiation of
minimize dose to normal tissues and spare these patients smaller volumes will reduce radiation-related side effects
from the late effects of radiation on the central nervous without affecting tumor control. In support of this hypothesis,
system. For children destined to receive central nervous current studies include statistical monitoring of rate and
system irradiation, planners should have at minimum, dose patterns of failure and comparison of disease control rates to
volume data for the entire brain, supratentorial brain or those achieved for patients treated using historic guidelines.
temporal lobes, hypothalamic-pituitary unit, cochleae, optic In support of this hypothesis, most studies include longitu-
chiasm, cervical spinal cord, and brainstem. These patients dinal assessment to evaluate patients for CNS effects and
are at risk for radiation-related CNS effects including abnor- include baseline assessments and some prospect of compar-
malities in neurovascular function affecting hearing, vision, ing results with historical controls, although, as mentioned
and gross neurologic function, cognitive effects involving earlier, there are limited data for such a comparison. In a land-
learning, memory, attention, behavior, academic achieve- mark study conducted at St. Jude Children’s Research
ment, and global IQ, endocrine function affecting growth Hospital from 1997 to 2003, patients with localized primary
hormone secretion, thyroid hormone secretion, ACTH secre- brain tumors were treated on a Phase II study to estimate the
tion, and the gonadotropins. Patients who receive hypotha- local control and patterns of failure for pediatric patients with
lamic irradiation are at risk for hypothalamic obesity. Patients localized primary CNS tumors treated with 3D conformal and
treated with large volume cranial, superficial cranial, or cra- intensity-modulated radiation therapy. The study also sought
niospinal irradiation are at risk for abnormalities in growth to quantify radiation-related CNS effects looking at pre-
and development including deformity of the calvarium, irradiation morbidity and estimating the time to onset of
facial bones, and spine [18]. CNS effects, specifically evaluating the effects of radiation
The true value of past, present, and future 3D treatment dose and volume on longitudinal functional outcomes. This
trials will be the correlation of 3D radiation dosimetry with study, known as RT-1 (clinicaltrials.gov, NCT00187226),
functional outcomes. Three-dimension radiation planning included patients as young as 12 months of age and up to age
has greatly improved our understanding of radiation dose 25 years, tumors that required only focal irradiation, biopsy
distributions in tumor and normal tissues. The most practical proven tumors except for optic pathway glioma, no evidence
way for those involved in treatment planning to view dose of dissemination (patients requiring craniospinal radiation
volume data is with the 2D dose volume histogram (DVH), were not eligible), no prior fractionated external beam irradi-
which is a reduction of the 3D dose volume information ation history, and prior chemotherapy was not exclusionary.
based on the assumption that all elements of the brain vol- Patients were required to have a minimum performance status
ume in question are functionally similar. In other words, the and parents were required to sign a protocol consent indicat-
dose-volume histogram lacks spatial information. The cumu- ing that they understood the risks of target volume reduction
lative DVH graphs the percent volume of a normal tissue including an increase in the rate of failure. Notably, patients
structure receiving a specified dose in a cumulative manner. with brainstem glioma because of their poor prognosis and
While this is helpful to the planner, the most important graph patients treated with craniospinal irradiation were excluded
is the differential DVH, which represents the percent volume because the whole-brain component of this treatment pre-
of the normal tissue structure (y-axis) receiving the given cludes a high degree of normal tissue sparing (Fig. 32.5).
dose on the x-axis. Integrating these data one can determine The RT-1 protocol had two strata. The first included a
the distribution or burden of radiation therapy to the normal 1 cm CTV margin for ependymoma, WHO Grade I and
Fig. 32.5 (a) Axial T2-weighted MR image of cerebellar pilocytic astrocytoma. (b) Axial T2-weighted MR image of optic pathway glioma.
(c) Axial T2-weighted MR image of diffuse pontine glioma. (d) Axial T2-weighted MR image of gliomatosis cerebri
Grade II astrocytoma, and craniopharyngioma. The second for the spinal cord and optic chiasm. At our own center, we
included a 2 cm CTV margin for Grade III and IV astrocy- routinely prescribe 54 Gy to the upper cervical spinal cord
toma. Planning target volume margins for all tumors were which is meant to include the upper 10 cm3’s of spinal cord
0.3–0.5 cm based on the methods of immobilization of the defined as 30 2 mL CT slices multiplied by the area of the
time that included vacuum-bag and thermoplastic technol- cross-section of the spinal cord ranging from 3–5 cm2. No
ogy, relocatable stereotactic head frames, and frameless neurologic deficits ascribed to the spinal cord have been
stereotaxy using radiocamera systems. During the time identified in patients treated in this manner. Similarly, dose
period of the study, additional developments included levels of 54–55.8 Gy to the optic chiasm have also been pre-
improved IV-administered general anesthesia, routine use of scribed in patients who have tumors adjacent to or involving
prone position general anesthesia, spiral CT, non-coplanar this critical normal tissue structure. Again, no delayed vision
multi-field conformal irradiation, routine MR registration, loss has been observed in this patient population. It is impor-
linear accelerator treatment automation with independent tant to note that while we limit doses to the tolerance levels
gantry and couch movement, micro-leaf collimation as described, regardless of the method of volume reduction
(<1 cm), widely available inverse planning, and intensity- or “cone down,” these structures will continue to receive
modulated radiation therapy. One of the key success factors additional dose. In our own practice, we have limited the
in radiation planning has been the introduction of MR regis- cone down dose objective to be less than 70 % of the daily
tration into the planning process: treatment standards should prescription dose. Although a point for more lengthy discus-
include multi-sequence MR registration. MR-only planning sion, the variances in PTV or normal tissue (unspecified
has been proposed for patients with CNS tumors. tissue) inhomogeneity, while often included in cooperative
Compared to prior treatment eras, dose prescriptions for group or institutional guidelines, are unsupported by data.
patients with pediatric CNS tumors are now similar or identi- There are a number of critical aspects to planning (pre-
cal to those used for adults. One example is ependymoma: treatment), treatment (on treatment), and post-treatment
postoperatively, 59.4 Gy is prescribed with dose reductions imaging. At the present time, thin slice CT (<2 mm) is the
to 54 Gy only for children under the age of 18 months who fundamental data set. CT is used for radiation dose calcula-
have undergone gross total resection. The dose reduction for tion as well as to define cochlea and spinal cord. As poste-
those less than 18 months of age is not based on experi- rior fossa tumors are the most common site for CNS tumors
mental data, rather concerns about late effects in very young in children, the cervical spinal cord is often subtended by
children who have not previously been irradiated as a part of the irradiated volume and needs to be precisely defined.
their frontline management. Grade I and II astrocytoma and The cochlea, an adjacent and important structure, is easily
craniopharyngioma are prescribed 54 Gy and Grade III and defined within the temporal bone and needs to be precisely
IV astrocytoma 59.4 Gy. The data showing no difference in defined. Because these structures are small and prone to the
dose for adults with low-grade glioma [19] does not apply to errors of registration (e.g., flexion or extension errors for
children for pediatric patients are more likely to have juve- the spinal cord when registering diagnostic MR and treatment
nile pilocytic astrocytoma (WHO Grade I) compared to planning CT), it is our suggestion, while CT remains the
adults who are more likely to have WHO Grade II astrocy- fundamental data set, to define the cochleae within the tem-
toma. One important aspect of dose prescription for local- poral bone and the cervical spinal cord beginning below the
ized brain tumor protocols in children is the tolerance doses foramen magnum on CT. MR imaging is, of course, used to
define target volumes and other normal tissue structures.

Gadolinium-enhanced T-1-weighted MR imaging works Ependymoma
well for most tumor systems to identify post-operative tumor
bed, enhancing residual tumor and normal tissue structures. Focusing on children with ependymoma treated on the RT-1
The post-contrast T-1-weighted imaging sequence should be protocol, and as previously reported, there were 88 children
acquired three dimensionally to improve spatial resolution treated from July 1997 to January 2003. The age at the time
and to minimize errors in registration. Certain low-grade of irradiation was 2.85 years (median) with a standard devia-
astrocytomas, most high-grade astrocytomas, and nearly all tion of 4.47 years. Seventy-four patients underwent gross
brainstem gliomas are best seen on T-2-weighted and fluid- total resection, six underwent near total resection, and eight
attenuated inversion recovery (FLAIR) imaging. Fortunately subtotal resection. Thirty-two patients underwent second
T2-weighted and FLAIR sequences may now be required surgery prior to the initiation of radiation therapy. Sixteen
3-dimensionally with high resolution at 1.5 T. In developing patients had pre-irradiation chemotherapy and 20 patients
targeting guidelines, the imaging must have resolution lower had supratentorial tumor location. As reported in the Journal
than the prescribed margins. of Clinical Oncology the 3-year event-free survival was
Imaging during treatment is a concept that requires care- 75 ± 6 % with a cumulative incidence of local failure of
ful study and discussion. Imaging during treatment may be 14 ± 4 % [20]. The median time to failure was 16 months
critical for patients with craniopharyngioma and cystic low- ranging from 6 to 26 months. These data compared favorably
grade astrocytoma. These patients are prone to cyst expan- to the CCG9942 study, which reported a 5-year event-free
sion during radiation therapy and in the setting of highly survival of 57 ± 6 % for children 3–21 years of age with intra-
focused small margin irradiation have the potential to expand cranial ependymoma [21]. Through the RT-1 trial the volume
outside of the prescribed isodose volume. On the St. Jude of irradiation for ependymoma was shown to be reducible to
RT-1 protocol, imaging was performed at weeks 3 and 5 of a 1 cm clinical target volume margin without effecting tumor
irradiation therapy in order to acquire experimental data for control. The improved tumor control rate was attributed to
quantitative measure of the MR T1 value in normal tissues. the high rate of gross total resection, improved targeting
This maneuver proved to be critical for patients with cranio- through MR and CT registration, the relatively high total
pharyngioma, the majority of which required cyst aspiration dose 59.4 Gy, and the relatively high tolerance dose for the
during radiation therapy either stereotactically or through cervical spinal cord (54 Gy). Based on the RT-1 study, the
imbedded catheter and reservoir systems. Despite aggressive Children’s Oncology Group initiated a study (ACNS0121,
aspiration, 15 % of patients required replanning. In addition, clinicaltrials.gov, NCT00027846) in August 2003 and it
several cases of expansive low-grade astrocytoma were noted completed patient enrollment in January 2010, the results are
during treatment. Given the standard to perform MR imaging not yet reported.
at some point in time prior to the initiation of radiation ther-
apy, often 2–3 weeks, followed by 6 weeks of radiation
therapy and the usual 6-week follow-up after radiation, Cognitive Effects
nearly 4 months may elapse between preplanning MRI and
posttreatment MRI. Posttreatment imaging on the RT-1 pro- Patients on the RT-1 protocol submitted to a battery of psy-
tocol includes scans every 3 months for 2 years, every 6 chology testing including measures of intelligent, academic
months for an additional 3 years followed by yearly imaging. achievement, attention, impulsivity and reaction time, mem-
Patients with potentially seeding tumors such as ependy- ory, auditory learning, problem, and adaptive behavior.
moma underwent MR imaging of the spine every 12 months. Testing was performed at baseline and serially at 6, 12, 24,
Over a 5.5 year period of time, from July 1997 to January 36, 48, and 60 months. As previously reported, children
2003, 202 patients with localized CNS tumors were under the age of 3 years had a statistically lower baseline
enrolled on the RT-1 protocol including 88 children with IQ. Both younger and older patients showed no change in IQ
ependymoma, 49 children with low-grade astrocytoma, 35 with time after irradiation. These data are an improvement
patients with high-grade astrocytoma, and 28 patients with compared to historic St Jude data where much older children
craniopharyngioma. There were two additional patients had a decline in IQ of approximately 2.4 points per year.
eligible for study; one with a CNS germ cell tumor, the Academic achievement, memory, verbal learning, and adap-
other with a choroid plexus tumor. The rates and patterns of tive behavior have been stable in these same patients [22]
failure for all these patient groups have been previously (Fig. 32.6).
reported or presented and fell within the expected range of The DVH data for the total brain, supratentorial brain,
failure noted for patients treated in the conventional radia- left and right temporal lobes were acquired for 86 of the 88
tion therapy era. ependymoma patients included on the RT-1 protocol.
post-radiotherapy fell within population norms. There was a

correlation between mean infratentorial dose and IQ, math,
reading, and spelling scores. When evaluating specific
sub-volumes of the cerebellum, there was a correlation
between mean anterior cerebellum dose and reading scores
and mean posterior cerebellum dose and math, spelling, and
visual-auditory learning scores.
Endocrine Effects
Patients on the RT-1 study also underwent provocative endo-

crine testing to measure growth hormone secretion, thyroid
hormone secretion, ACTH reserve, and gonadotropin secre-
Fig. 32.6 Prospective evaluation of children receiving focal irradiation
demonstrates essentially stable IQ and adaptive behaviors, but a signifi-
tion in appropriate patients. These tests were performed at
cant decline in communication scores baseline and 6, 12, 36, and 60 months after the initiation of
irradiation. In the first report of its kind, 46 of 68 initial
patients were found to have preexisting hormone deficien-
These data were partitioned into three intervals representing cies including 16 of 32 patients with posterior fossa tumors
the percent volume receiving less than 20 Gy, the percent [25, 26]. Among these patients specific deficiencies included
volume receiving 20–40 Gy, and the percent volume growth hormone secretion abnormality in 38 %, thyroid hor-
receiving 40–60 Gy. The DVH data were combined with mone secretion abnormality in 43 %, ACTH secretion
clinical data to form a linear aggression model estimating abnormality in 22 %, and GNRH abnormality in 18 %. There
IQ as a function of time [23]. was one patient with hyperprolactinemia. Among children
Initially the correlation involved dosimetry and the IQ data with ependymoma, the baseline rate of growth hormone defi-
for all patients. A total brain dose volume estimating equation ciency using the ATT/L-dopa test was 29 %, thyroid hor-
was produced that included in the intercept age and in the mone abnormality using the TSH surge 27 % or TRH
slope time and dose volume. No other clinical variables were stimulation test 10 %, ACTH secretion abnormality using the
included in the equation. A similar equation was developed 1 μg ACTH test 12 %, and metyrapone test 12 %.
for the supratentorial brain; however, no correlation could be Change in peak growth hormone levels following ATT/L-
developed for the temporal lobes or with any of the clinical dopa have been studied in children from baseline through 12
co-variants. It is important to note in the total brain volume months after irradiation. The falloff in growth hormone for
dosimetry estimating equation that the low dose coefficient (v 25 “normal” patients was shown to be exponential and a
0–25 Gy) was negative and small, the high dose coefficient (v population-based estimating equation was developed for
45–65 Gy) was negative and small, and the intermediate dose which all dose terms (low, intermediate, and high) were sta-
volume term v 20–40 or v –25 to 45 was similar in weight to tistically significant and negative. The effect of low dose was
the high dose coefficient but with a positive term. To increase smallest and the effects of intermediate and high doses were
the specificity of our model we performed a correlation only similar. Similar changes in peak growth hormone after ATT/
with infratentorial patients and developed three equations; L-dopa were developed specifically for ependymoma.
one for the total brain, another for the supratentorial brain,
and one for the left temporal lobe. The dose volume equations
have similar coefficients, negative low dose, positive interme- Other Long-Term Side Effects
diate dose, and negative high dose terms. The low dose terms
were not statistically significant for the total brain or left tem- The data gathered from the St. Jude prospective trial of
poral lobe equations. Age continued to affect the intercept conformal radiotherapy for children with ependymoma
and time affected for the slope. has allowed us to better counsel our patients and their fami-
A recent study evaluated the impact of cerebellar radia- lies on the long-term side effects of modern radiotherapy.
tion dosimetry on cognitive and academic abilities for chil- The 7-year cumulative incidence of brainstem radionecrosis
dren with infratentorial ependymoma receiving adjuvant was found to be 1.6 % [27]. The 7-year cumulative incidence
conformal radiotherapy (54–59.4 Gy) [24]. Seventy-six chil- of any secondary malignancy and second malignant brain
dren underwent cognitive and academic testing at baseline, tumor was 4.1 % and 2.3 %, respectively. The secondary
6 months post-conformal radiotherapy, and annually for malignancies included two glioblastoma and one papillary
5 years. Overall for this population, the IQ scores at 5 years thyroid cancer. Only 1 of 153 patients required a revascular-
ization surgery and this patient’s high dose target volume medulloblastoma: clinical outcomes including hearing and neuro-
encompassed the Circle of Willis. cognitive. Int J Radiat Oncol Biol Phys. 2011;81(2):S113.
13. Murphy E, Merchant T, Wu S, Xiong X, Gajjar A. Necrosis after
In summary, the results of recent trials for localized brain craniospinal irradiation: results from a prospective series of chil-
tumors have made radiation therapy more acceptable for dren with CNS embryonal tumors. Int J Radiat Oncol Biol Phys.
young children with brain tumors. Improved targeting may 2012;83(5):e655–60.
increase tumor control rates for some tumors, which will in 14. Laughton SJ, Merchant TE, Sklar CA, Kun LE, Fouladi M,
Broniscer A, et al. Endocrine outcomes for children with embryo-
turn increase the validity of numerous prognostic factors. nal brain tumors after risk-adapted craniospinal and conformal
The next generation of volume reduction trials is currently primary-site irradiation and high-dose chemotherapy with stem-cell
underway for patients with medulloblastoma and localized rescue on the SJMB-96 trial. J Clin Oncol. 2008;26(7):1112–8.
brain tumors. We also anticipate outcomes from children 15. Ashley DM, Merchant TE, Strother D, Zhou T, Duffner P, Burger
PC, et al. Induction chemotherapy and conformal radiation therapy
with these brain tumors being treated with proton therapy. for very young children with nonmetastatic medulloblastoma:
Children’s Oncology Group study P9934. J Clin Oncol. 2012;
30(26):3181–6.
References 16. ICRU report 50. Dose specification for reporting external beam
therapy with photons and electrons. No. 9. Washington, DC:
International Commission on Radiation Units and Measurements;
1. Ries LAG, Smith MA, Gurney JG, Linet M, Tamra T, Young JL, 1993.
et al., editors. Cancer incidence and survival among children and 17. ICRU report 62. Dose specification for reporting external beam
adolescents: United States SEER Program 1975–1995. National therapy with photons and electrons. Washington, DC: International
Cancer Institute, SEER Program. NIH Publication No. 99-4649. Commission on Radiation Units and Measurements; 1999.
Bethesda, MD: NIH; 1999. 18. Leung W, Rose SR, Merchant TE. Neuroendocrine complications
2. Boëthius J, Ulfarsson E, Rähn T, Lippitz B. Gamma knife radiosur- of cancer therapy. In: Schwartz CL, Hobbie WL, Constine LS,
gery for pilocytic astrocytomas. J Neurosurg. 2002;97(5 Suppl): Ruccione KS, editors. Survivors of childhood and adolescent
677–80. cancer, a multidisciplinary approach. 2nd ed. Heidelberg: Springer;
3. Kano H, Kondziolka D, Niranjan J, Flickinger J, Lunsford 2005. p. 51–80.
L. Stereotactic radiosurgery for pilocytic astrocytomas. Part 2: out- 19. Shaw EG, Tatter SB, Lesser GJ, Ellis TL, Stanton CA, Stieber
comes in pediatric patients. J Neurooncol. 2009;95(2):219–29. VW. Current controversies in the radiotherapeutic management of
4. Kida Y, Kobayashi T, Mori Y. Gamma knife radiosurgery for low- adult low-grade glioma. Semin Oncol. 2004;31(5):653–8.
grade astrocytomas: results of long-term follow up. J Neurosurg. 20. Merchant TE, Mulhern RK, Krasin MJ, Kun LE, Williams T, Li C,
2000;93 Suppl 3:42–6. et al. Preliminary results from a phase II trial of conformal radiation
5. Hallemeier CL, Pollock BE, Schomberg PJ, Link MJ, Brown PD, therapy and evaluation of radiation-related CNS effects for pediat-
Stafford SL. Stereotactic radiosurgery for recurrent or unresectable ric patients with localized ependymoma. J Clin Oncol.
pilocytic astrocytoma. Int J Radiat Oncol Biol Phys. 2012;83(1): 2004;22(15):3156–62.
107–12. 21. Garvin Jr JH, Selch MT, Holmes E, Berger MS, Finlay JL, Flannery
6. Packer RJ, Goldwein J, Nicholson HS, Vezina LG, Allen JC, Ris A, et al. Phase II study of pre-irradiation chemotherapy for child-
MD, et al. Treatment of children with medulloblastomas with hood intracranial ependymoma. Children’s Cancer Group protocol
reduced-dose craniospinal radiation therapy and adjuvant chemo- 9942: a report from the Children’s Oncology Group. Pediatr Blood
therapy: a Children’s Cancer Group Study. J Clin Oncol. 1999; Cancer. 2012;59(7):1183–9.
17(7):2127–36. 22. Netson KL, Conklin HM, Wu S, Xiong X, Merchant TE. A 5-year
7. Walter AW, Mulhern RK, Gajjar A, Heidelman RL, Reardon D, investigation of children’s adaptive functioning following confor-
Sanford RA. Survival and neurodevelopmental outcome of young mal radiation therapy for localized ependymoma. Int J Radiat
children with medulloblastoma at St Jude Children’s Research Oncol Biol Phys. 2012;84(1):217–223.e1.
Hospital. J Clin Oncol. 1999;17(12):3720–8. 23. Merchant TE, Kiehna EN, Li C, Xiong X, Mulhern RK. Radiation
8. Ris MD, Packer R, Goldwein J, Jones-Wallace D, Boyett dosimetry predicts IQ after conformal radiation therapy in pediatric
JM. Intellectual outcome after reduced-dose radiation therapy plus patients with localized ependymoma. Int J Radiat Oncol Biol Phys.
adjuvant chemotherapy for medulloblastoma: a Children’s Cancer 2005;63(5):1546–54.
Group Study. J Clin Oncol. 2001;19(15):3470–6. 24. Sharma S, Conklin HM, Xiong X, Merchant TE. The effect of cer-
9. Merchant TE, Kun LE, Krasin MJ, Wallace D, Chintagumpala ebellum radiation dosimetry on cognitive outcomes in children with
MM, et al. Multi-institution prospective trial of reduced-dose cra- infratentorial ependymoma. Int J Radiat Oncol Biol Phys. 2011;
niospinal irradiation (23.4 Gy) followed by conformal posterior 81(2):S114–5.
fossa (36 Gy) and primary site irradiation (55.8 Gy) and dose- 25. Merchant TE, Goloubeva O, Pritchard D, Gabeer MW, Xiong X,
intensive chemotherapy for average-risk medulloblastoma. Int J Danish RK, et al. Radiation dose-volume effects on growth hor-
Radiat Oncol Biol Phys. 2008;70(3):782–7. mone secretion. Int J Radiat Oncol Biol Phys. 2002;52(5):
10. Merchant T, Palmer S, Lukose R, Wu S, Xiong S, Gajjar A. Critical 1264–70.
combinations of radiation dose and volume predict IQ and aca- 26. Merchant TE, William T, Smith JM, Rose SR, Danish RK, et al.
demic achievement after craniospinal irradiation in children. Int J Preirradiation endocrinopathies in pediatric brain tumor patients
Radiat Oncol Biol Phys. 2010;78(3):s17. determined by dynamic tests of endocrine function. Int J Radiat
11. Merchant TE. Proton beam therapy in pediatric oncology. Cancer Oncol Biol Phys. 2002;54(1):45–50.
J. 2009;15(4):298–305. 27. Merchant TE, Li C, Xiong X, Kun LE, Boop FA, Sanford
12. Yock T, Yep BY, Pulsifer MP, Ebb D, MacDonald SM, Marcus KC, RA. Conformal radiotherapy after surgery for paediatric ependy-
et al. Results from a prospective trial of proton radiotherapy for moma: a prospective study. Lancet Oncol. 2009;10(3):258–66.
Pediatric Brain Tumors:
Viewpoint—Chemotherapy 33
Nathan J. Robison
survival is quite favorable, and spontaneous malignant

Introduction transformation exceedingly rare [7]. Disease- and treatment-
related morbidity are therefore of primary concern. Histori-
Recent decades have seen significant advances in the treatment cally, radiation therapy has been considered definitive
of childhood brain tumors [1]. Despite these, the outcome for therapy for non-resectable disease [8, 9]. However, as data
many tumors remains extremely poor, while long-term survi- continues to accrue regarding the long-term radiation-
vors often face significant long-term sequelae. For many related morbidities in this population, post-radiation disease
better-prognosis pediatric brain tumors, a major aim of progression, and the probability of very-long-term survival
recent clinical innovation has been to decrease late effects, without radiation, significant doubt has been cast on the
particularly the neurocognitive effects consequent to irradi- appropriateness of radiation in many cases.
ating eloquent areas of the developing brain [2]. Various che- Low-grade gliomas, pilocytic astrocytoma especially, are
motherapeutic approaches have been used to allow reduction usually responsive to a number of relatively low-intensity
in radiation dose and field, or to replace irradiation entirely. chemotherapeutic regimens [1]. A randomized clinical trial
Stereotactic radiosurgery (SRS) has the potential to address comparing carboplatin and vincristine (CV) with thioguanine,
some of these problems as well: SRS allows treatment of procarbazine, lomustine, and vincristine (TPCV) showed
focal areas of disease with high-dose radiation, with signi- similar outcomes with both regimens; event-free and overall
ficantly reduced risk of injury to the surrounding brain. survival rates for the entire cohort were 45 and 86 %, respec-
Anecdotal experience lends support to the safe use of SRS in tively, at 5 years [10]. The high rate of eventual disease
combination with other modalities as a radiation-limiting progression does not necessarily reflect a poor long-term
strategy [3, 4]. The risk of radiation necrosis and subsequent prognosis, as response rate to salvage therapy is high, even
injury to eloquent areas of the brain must be considered [5, 6]. with multiply recurrent disease, as the high long-term overall
This chapter discusses the relative roles and limitations of survival indicates [11]. Chemotherapy should generally be
SRS and chemotherapy in the treatment of specific pediatric considered standard front-line therapy for non-resectable
brain tumors. tumors requiring treatment as well as for recurrent disease.
Besides CV and TPCV, regimens that have been used with
success in the treatment of newly diagnosed or recurrent
Low-Grade Glioma low-grade glioma include cisplatin and etoposide [12], temo-
zolomide [13], monthly carboplatin [14], or weekly vinblas-
Low-grade gliomas include pilocytic astrocytoma, which tine [15], among others.
is the most common brain tumor in children, as well as a SRS in the treatment of low-grade gliomas has been used
heterogeneous collection of rarer histologic entities [1]. anecdotally and in several small series with generally favor-
Complete resection, where feasible, is usually curative. able results [6, 16–22]. A retrospective series of 50 children
However, even for non-resectable tumors, very long-term who underwent SRS for unresectable disease showed a
10-year progression-free survival (PFS) of 89 % for patients
N.J. Robison, M.D. (*) treated at the time of diagnosis and 45 % for patients treated
Children’s Center for Cancer and Blood Diseases, Children’s at time of recurrence/progression [23]. This is similar to out-
Hospital Los Angeles, University of Southern California Keck come seen after treatment with chemotherapy. Adverse radi-
School of Medicine, 4650 W. Sunset Blvd, MS #54,
ation effects are reported to 10–35 % of LGG patients treated
Los Angeles, CA 90027, USA
e-mail: nrobison@chla.usc.edu with SRS [18, 23]. The risk of second malignancy after SRS
440 N.J. Robison
is thought to be reduced, but is not yet well-defined [24]. demonstrate any role for adjuvant chemotherapy [40].
Efficacy of SRS in tumors progressive after prior fraction- The diffusely infiltrative nature of DIPG precludes SRS as a
ated radiotherapy appears to be limited [22]. viable option.
Patients with neurofibromatosis type 1 (NF1) have a high
incidence of low-grade glioma, especially optic pathway gli-
oma. Low-grade gliomas typically follow an indolent course in Ependymoma
this population [25]. Radiation therapy is generally contraindi-
cated in patients with NF1, due to extremely high risk of Ependymomas account for about 9 % of pediatric intracra-
second malignancy and radiation-induced vascular disease nial tumors [41]. They occur most commonly in the posterior
[25–28]. Observation alone is appropriate for most asymptom- fossa, and less commonly in the cranial hemispheres or the
atic patients, even in the setting of radiographic progression spinal cord. Prognosis is highly dependent on the degree of
[29, 30]. NF1 patients are also at increased risk of secondary surgical resection, with long-term PFS rates of 60–77 % for
leukemia after treatment with alkylating agents such as cyclo- patients with completely resected vs. 9–34 % for those with
phosphamide or nitrosureas [31]. For this reason, treatment incompletely resected tumors [42–45]. Tumor grade (ana-
with non-alkylating chemotherapy is recommended for NF1 plastic vs. non-anaplastic) may also have prognostic implica-
patients with symptomatic low-grade gliomas. tions [44–46]. Best outcome for localized tumors has been
achieved with maximal surgical resection, followed by
involved field radiotherapy [45]. For completely resected grade
High-Grade Glioma II supratentorial ependymoma, observation alone without
adjuvant radiotherapy may be appropriate [47, 48]. Historically,
High-grade gliomas include grade III astrocytoma (anaplastic), the role of adjuvant chemotherapy in the management of
grade IV astrocytoma (glioblastoma), and other rarer entities this disease has been unclear [42, 43, 49]. Attempts to use
such as anaplastic oligodendroglioma and oligoastrocytoma. dose-intensive chemotherapy instead of radiation have been
Although relatively rare in children, high-grade gliomas unsuccessful, with PFS rates of 9–26 % [50–52]. However,
nonetheless contribute substantially to pediatric cancer- ependymomas are known to be chemotherapy-responsive in
related mortality. Because long-term survival is the excep- the short term, and combination chemotherapy may have a role
tion, late effects concerns have not played an important role either in delaying definitive radiation therapy in young chil-
in guiding therapeutic strategies. dren [39, 53] or as post-surgical, pre-radiation therapy for
The mainstay of therapy for high-grade glioma is maxi- patients with incompletely resected tumors [41]. The use of
mal resection where feasible, followed by involved field SRS in conjunction with fractionated radiotherapy to treat
radiation. High-grade gliomas are relatively chemotherapy- small residual non-resectable ependymoma has been proposed,
resistant. Adjuvant chemotherapy, typically temozolomide with anecdotal reports of success [54, 55]. SRS may also have
monotherapy, concurrent with and following radiation, has a role in the management of recurrent ependymoma. A retro-
been shown to have benefit in adults with glioblastoma [32] spective review of 26 patients treated with SRS for ependy-
and may have benefit in children [33]. Temozolomide moma recurrence showed a 3-year PFS of 66 %, and a 3-year
appears to be active only in tumors characterized by MGMT local control rate of 72 % [56]. In another series of 21 patients
methylation [34]. The antiangiogenic agent bevacizumab is treated with SRS for recurrent ependymoma, the local control
effective in the management of recurrent disease in adults; rate was similarly high, but the distant failure rate was 80 % at
initial pediatric experience is less promising [35, 36]. The 3 years [57].
potential role of SRS in the management of high-grade
glioma includes treatment of very small tumors [20] and
treatment of larger tumors as an adjunct to fractionated Embryonal Tumors
radiotherapy [17]. SRS also may have a role in palliative
local tumor control at recurrence [5, 37, 38]. Medulloblastoma is the most common histologically malig-
Glioblastoma in infants has unique clinical features and is nant brain tumor of childhood [58]. Standard therapy in most
often responsive to chemotherapy. Sustained responses have cases includes maximal surgical resection, cranial spinal
been described in a substantial portion of infants treated with irradiation with boost to tumor bed, and chemotherapy [59].
surgery and chemotherapy alone, without radiation [39]. The addition of chemotherapy before, after, and/or during
Diffuse intrinsic pontine glioma (DIPG) is a uniquely radiation has led to increased survival for high-risk (meta-
chemotherapy-resistant disease, with an abysmal prognosis. static, incompletely resected, and/or anaplastic) disease
Although radiation has palliative benefit, the disease remains and allowed dose-reduction of craniospinal irradiation for
universally fatal, with a median overall survival of 11 months standard-risk disease. With this approach, long-term PFS
from diagnosis. Over 200 clinical trials have failed to exceeds 80 % for patients with standard risk tumors and
33 Pediatric Brain Tumors: Viewpoint—Chemotherapy 441
approaches 70 % for patients with high-risk disease [58].

For infants and young children, dose-intensive chemother- Craniopharyngioma
apy regimens without radiation have been used. Survival
outcomes for young patients treated without radiation are Craniopharyngiomas are believed to arise from ectodermal
quite favorable for those with desmoplastic histology, but nests within the pituitary stalk [70]. Although histologically
more guarded for those with classic and anaplastic histology benign, they are often associated with substantial morbidity,
[60–62]. both disease-related and iatrogenic. Standard therapy includes
Anecdotal experience supports the uses of SRS in medul- either surgical resection alone, or partial resection with
loblastoma in certain settings [20]. In theory, SRS has utility involved field radiotherapy. Successful use of SRS has been
either in the treatment of concerning lesions persistent after reported: in a series of 100 pediatric and adult patients treated
definitive therapy, or in the palliative or adjuvant manage- with SRS for postoperative residual or recurrent craniopha-
ment of small tumors at recurrence. The marked propensity ryngioma, the 10-year PFS was 60 % [71]. Another similar
of embryonal tumors towards distant recurrence and lepto- series of 46 patients with craniopharyngioma showed a 5-year
meningeal dissemination limits the utility of SRS in most PFS of 68 % [72]. Risk of visual compromise from optic
cases [20]. nerve irradiation is an important consideration [73]. There is
Pineoblastoma and primitive neuroectodermal tumor of currently no established role for chemotherapy in the treat-
the central nervous system are less common embryonal ment of craniopharyngioma. However, systemic interferon
tumors occurring outside of the posterior fossa. Treatment is alpha-2b may have activity in this disease [74].
similar to that of medulloblastoma, but with a less favorable
prognosis. Atypical teratoid rhabdoid tumor is a more
recently described embryonal tumor, initially thought to Conclusion
have an extremely poor prognosis. Long-term survival
exceeding 40 % has been reported in AT/RT patients treated SRS has a role in the management of pediatric brain tumors.
with an intensive anthracycline-based chemotherapy regi- The propensity of many pediatric neural malignancies either
men and fractionated radiotherapy [63]. to disseminate along CSF channels or to infiltrate surround-
ing brain limits the utility of SRS as monotherapy. Its most
promising role is as an adjuvant to surgery, chemotherapy,
Germ Cell Tumors and/or conventional fractionated radiation in the management
of localized residual disease. With notable exceptions, most
Germ cell tumors of the central nervous system are classi- data on the use of SRS in children is from small retrospective
fied either as pure germinoma or as nongerminomatous series. Further prospective analysis is needed.
germ cell tumor (NGGCT), the latter typically representing
mixed-histology tumors. Standard treatment for CNS ger-
minoma includes radiation, either of the entire craniospinal References
axis or limited to the ventricular fields. Platinum-based
chemotherapy regimens have been used as an adjuvant 1. Pollack IF. Multidisciplinary management of childhood brain
modality to reduce radiation dose and/or volume, while tumors: a review of outcomes, recent advances, and challenges.
J Neurosurg Pediatr. 2011;8(2):135–48.
still maintaining excellent long-term survival [64]. 2. Turner CD, Rey-Casserly C, Liptak CC, Chordas C. Late effects of
Attempts to treat germinoma with chemotherapy alone therapy for pediatric brain tumor survivors. J Child Neurol. 2009;
have been less successful [65]. NGGCT has historically 24(11):1455–63.
had a poorer prognosis, although recent trials have shown 3. Guruangan S, Dunkel IJ, Goldman S, Garvin JH, Rosenblum M,
Boyett JM, et al. Myeloablative chemotherapy with autologous
significantly improved outcome with the use of dose-inten- bone marrow rescue in young children with recurrent malignant
sive pre-radiation platinum-based chemotherapy [66]. brain tumors. J Clin Oncol. 1998;16(7):2486–93.
High-dose myeloablative chemotherapy with autologous 4. Heideman RL, Douglass EC, Krance RA, Fontanesi J, Langston
stem cell rescue has been used with success as a salvage JA, Sanford RA, et al. High-dose chemotherapy and autologous
bone marrow rescue followed by interstitial and external-beam
regimen in patients with recurrent germ cell tumors [67]. radiotherapy in newly diagnosed pediatric malignant gliomas.
SRS has been used anecdotally in patients with germinoma, J Clin Oncol. 1993;11(8):1458–65.
either in combination with fractionated radiotherapy for 5. Hall WA, Djalilian HR, Sperduto PW, Cho KH, Gerbi BJ, Gibbons
primary management of small tumors, or in combination JP, et al. Stereotactic radiosurgery for recurrent malignant gliomas.
J Clin Oncol. 1995;13(7):1642–8.
with dose-intensive myeloablative chemotherapy with 6. Baumann GS, Wara WM, Larson DA, Sneed PK, Gutin PH,
stem cell rescue and conventional radiotherapy for relapsed Ciricillo SF, et al. Gamma knife radiosurgery in children. Pediatr
disease [68, 69]. Neurosurg. 1996;24(4):193–201.
442 N.J. Robison
7. Parsa CF, Givrad S. Juvenile pilocytic astrocytomas do not undergo 25. Guillamo J-S, Créange A, Kalifa C, Grill J, Rodriguez D, Doz F,
spontaneous malignant transformation: grounds for designation as et al. Prognostic factors of CNS tumours in neurofibromatosis 1
hamartomas. Br J Ophthalmol. 2008;92(1):40–6. (NF1): a retrospective study of 104 patients. Brain. 2003;126(Pt 1):
8. Marcus KJ, Goumnerova L, Billett AL, Lavally B, Scott RM, 152–60.
Bishop K, et al. Stereotactic radiotherapy for localized low-grade 26. Kestle JR, Hoffman HJ, Mock AR. Moyamoya phenomenon after
gliomas in children: final results of a prospective trial. Int J Radiat radiation for optic glioma. J Neurosurg. 1993;79(1):32–5.
Oncol Biol Phys. 2005;61(2):374–9. 27. Evans DGR, Birch JM, Ramsden RT, Sharif S, Baser ME. Malignant
9. Perilongo G. Considerations on the role of chemotherapy and mod- transformation and new primary tumours after therapeutic radiation
ern radiotherapy in the treatment of childhood low grade glioma. for benign disease: substantial risks in certain tumour prone syn-
J Neurooncol. 2005;75(3):301–7. dromes. J Med Genet. 2006;43(4):289–94.
10. Ater JL, Zhou T, Holmes E, Mazewski CM, Booth TN, Freyer DR, 28. Sharif S, Ferner R, Birch JM, Gillespie JE, Gattamaneni HR, Baser
et al. Randomized study of two chemotherapy regimens for treat- ME, et al. Second primary tumors in neurofibromatosis 1 patients
ment of low-grade glioma in young children: a report from the treated for optic glioma: substantial risks after radiotherapy. J Clin
Children’s Oncology Group. J Clin Oncol. 2012;30(21):2641–7. Oncol. 2006;24(16):2570–5.
Available from: http://www.ncbi.nlm.nih.gov/pubmed/22665535 29. Ullrich NJ, Raja AI, Irons MB, Kieran MW, Goumnerova
11. Scheinemann K, Bartels U, Tsangaris E, Hawkins C, Huang A, L. Brainstem lesions in neurofibromatosis type 1. Neurosurgery.
Dirks P, et al. Feasibility and efficacy of repeated chemotherapy for 2007;61(4):762–6; discussion 766–7.
progressive pediatric low-grade gliomas. Pediatr Blood Cancer. 30. Jahraus CD, Tarbell NJ. Optic pathway gliomas. Pediatr Blood
2011;57(1):84–8. Cancer. 2006;46(5):586–96.
12. Sardi I, Bresci C, Schiavello E, Biassoni V, Fratoni V, Cardellicchio 31. Davies SM. Therapy-related leukemia associated with alkylating
S, et al. Successful treatment with a low-dose cisplatin-etoposide agents. Med Pediatr Oncol. 2001;36(5):536–40.
regimen for patients with diencephalic syndrome. J Neurooncol. 32. Stupp R, Mason WP, van den Bent MJ, Weller M, Fisher B, Taphoorn
2012;109(2):375–83. Available from: http://www.ncbi.nlm.nih. MJB, et al. Radiotherapy plus concomitant and adjuvant temozolo-
gov/pubmed/22717669 mide for glioblastoma. N Engl J Med. 2005;352(10): 987–96.
13. Gururangan S, Fisher MJ, Allen JC, Herndon 2nd JE, Quinn JA, 33. Cohen KJ, Pollack IF, Zhou T, Buxton A, Holmes EJ, Burger PC,
Reardon DA, et al. Temozolomide in children with progressive low- et al. Temozolomide in the treatment of high-grade gliomas in chil-
grade glioma. Neuro Oncol. 2007;9(2):161–8. dren: a report from the Children’s Oncology Group. Neuro Oncol.
14. Gururangan S, Cavazos CM, Ashley D, Herndon 2nd JE, Bruggers 2011;13(3):317–23.
CS, Moghrabi A, et al. Phase II study of carboplatin in children 34. Donson AM, Addo-Yobo SO, Handler MH, Gore L, Foreman
with progressive low-grade gliomas. J Clin Oncol. 2002;20(13): NK. MGMT promoter methylation correlates with survival benefit
2951–8. and sensitivity to temozolomide in pediatric glioblastoma. Pediatr
15. Bouffet E, Jakacki R, Goldman S, Hargrave D, Hawkins C, Shroff Blood Cancer. 2007;48(4):403–7.
M, et al. Phase II study of weekly vinblastine in recurrent or refrac- 35. Gururangan S, Chi SN, Young Poussaint T, Onar-Thomas A,
tory pediatric low-grade glioma. J Clin Oncol. 2012;30(12): Gilbertson RJ, Vajapeyam S, et al. Lack of efficacy of bevacizumab
1358–63. plus irinotecan in children with recurrent malignant glioma and dif-
16. Somaza SC, Kondziolka D, Lunsford LD, Flickinger JC, Bissonette fuse brainstem glioma: a Pediatric Brain Tumor Consortium study.
DJ, Albright AL. Early outcomes after stereotactic radiosurgery for J Clin Oncol. 2010;28(18):3069–75.
growing pilocytic astrocytomas in children. Pediatr Neurosurg. 36. Parekh C, Jubran R, Erdreich-Epstein A, Panigrahy A, Bluml S,
1996;25(3):109–15. Finlay J, et al. Treatment of children with recurrent high grade glio-
17. Grabb PA, Lunsford LD, Albright AL, Kondziolka D, Flickinger mas with a bevacizumab containing regimen. J Neurooncol. 2011;
JC. Stereotactic radiosurgery for glial neoplasms of childhood. 103(3):673–80.
Neurosurgery. 1996;38(4):696–701; discussion 701–2. 37. Shrieve DC, Alexander 3rd E, Wen PY, Fine HA, Kooy HM, Black
18. Kida Y, Kobayashi T, Mori Y. Gamma knife radiosurgery for low- PM, et al. Comparison of stereotactic radiosurgery and brachyther-
grade astrocytomas: results of long-term follow up. J Neurosurg. apy in the treatment of recurrent glioblastoma multiforme.
2000;93 Suppl 3:42–6. Neurosurgery. 1995;36(2):275–82; discussion 282–4.
19. Boëthius J, Ulfarsson E, Rähn T, Lippittz B. Gamma knife radiosur- 38. Song KS, Phi JH, Cho B-K, Wang K-C, Lee JY, Kim DG, et al.
gery for pilocytic astrocytomas. J Neurosurg. 2002;97(5 Suppl): Long-term outcomes in children with glioblastoma. J Neurosurg
677–80. Pediatr. 2010;6(2):145–9.
20. Giller CA, Berger BD, Pistenmaa DA, Sklar F, Weprin B, Shapiro 39. Duffner PK, Horowitz ME, Krischer JP, Friedman HS, Burger PC,
K, et al. Robotically guided radiosurgery for children. Pediatr Cohen ME, et al. Postoperative chemotherapy and delayed radia-
Blood Cancer. 2005;45(3):304–10. tion in children less than three years of age with malignant brain
21. Lizarraga KJ, Gorgulho A, Lee SP, Rauscher G, Selch MT, DeSalles tumors. N Engl J Med. 1993;328(24):1725–31.
AAF. Stereotactic radiation therapy for progressive residual pilo- 40. Khatua S, Moore KR, Vats TS, Kestle JRW. Diffuse intrinsic pon-
cytic astrocytomas. J Neurooncol. 2012;109(1):129–35. tine glioma-current status and future strategies. Childs Nerv Syst.
22. Hallemeier CL, Pollock BE, Schomberg PJ, Link MJ, Brown PD, 2011;27(9):1391–7.
Stafford SL. Stereotactic radiosurgery for recurrent or unresectable 41. Garvin Jr JH, Selch MT, Holmes E, Berger MS, Finlay JL, Flannery
pilocytic astrocytoma. Int J Radiat Oncol Biol Phys. 2012;83(1): A, et al. Phase II study of pre-irradiation chemotherapy for child-
107–12. hood intracranial ependymoma. Children’s Cancer Group protocol
23. Kano H, Niranjan A, Kondziolka D, Flickinger JC, Pollack IF, 9942: a report from the Children’s Oncology Group. Pediatr Blood
Jakacki RI, et al. Stereotactic radiosurgery for pilocytic astrocyto- Cancer. 2012;59(7):1183–9.
mas part 2: outcomes in pediatric patients. J Neurooncol. 2009; 42. Pollack IF, Gerszten PC, Martinez AJ, Lo KH, Shultz B, Albright
95(2):219–29. AL, et al. Intracranial ependymomas of childhood: long-term out-
24. Balasubramaniam A, Shannon P, Hodaie M, Laperriere N, Michaels come and prognostic factors. Neurosurgery. 1995;37(4):655–66;
H, Guha A. Glioblastoma multiforme after stereotactic radiother- discussion 666–7.
apy for acoustic neuroma: case report and review of the literature. 43. Horn B, Heideman R, Geyer R, Pollack I, Packer R, Goldwein J,
Neuro Oncol. 2007;9(4):447–53. et al. A multi-institutional retrospective study of intracranial
33 Pediatric Brain Tumors: Viewpoint—Chemotherapy 443
ependymoma in children: identification of risk factors. J Pediatr 59. Padovani L, André N, Gentet JC, Muracciole X. Medulloblastoma.
Hematol Oncol. 1999;21(3):203–11. Eur J Cancer. 2011;47 Suppl 3:S338.
44. Oya N, Shibamoto Y, Nagata Y, Negoro Y, Hiraoka M. Postoperative 60. Chi SN, Gardner SL, Levy AS, Knopp EA, Miller DC, Wisoff JH,
radiotherapy for intracranial ependymoma: analysis of prognostic et al. Feasibility and response to induction chemotherapy intensi-
factors and patterns of failure. J Neurooncol. 2002;56(1):87–94. fied with high-dose methotrexate for young children with newly
45. Merchant TE, Li C, Xiong X, Kun LE, Boop FA, Sanford diagnosed high-risk disseminated medulloblastoma. J Clin Oncol.
RA. Conformal radiotherapy after surgery for paediatric ependy- 2004;22(24):4881–7.
moma: a prospective study. Lancet Oncol. 2009;10(3):258–66. 61. Dhall G, Grodman H, Ji L, Sands S, Gardner S, Dunkel IJ, et al.
46. Agaoglu FY, Ayan I, Dizdar Y, Kebudi R, Gorgun O, Darendeliler Outcome of children less than three years old at diagnosis with non-
E. Ependymal tumors in childhood. Pediatr Blood Cancer. 2005; metastatic medulloblastoma treated with chemotherapy on the
45(3):298–303. “Head Start” I and II protocols. Pediatr Blood Cancer. 2008;
47. Ghia AJ, Mahajan A, Allen PK, Armstrong TS, Lang Jr FF, Gilbert 50(6):1169–75.
MR, et al. Supratentorial gross-totally resected non-anaplastic 62. Rutkowski S, Gerber NU, von Hoff K, Gnekow A, Bode U, Graf N,
ependymoma: population based patterns of care and outcomes et al. Treatment of early childhood medulloblastoma by postopera-
analysis. J Neurooncol. 2013;115(3):513–20. tive chemotherapy and deferred radiotherapy. Neuro Oncol. 2009;
48. Aizer AA, Ancukiewicz M, Nguyen PL, Macdonald SM, Yock TI, 11(2):201–10.
Tarbell NJ, et al. Natural history and role of radiation in patients with 63. Chi SN, Zimmerman MA, Yao X, Cohen KJ, Burger P, Biegel JA,
supratentorial and infratentorial WHO grade II ependymomas: results et al. Intensive multimodality treatment for children with newly
from a population-based study. J Neurooncol. 2013;115(3): 411–9. diagnosed CNS atypical teratoid rhabdoid tumor. J Clin Oncol.
49. Evans AE, Anderson JR, Lefkowitz-Boudreaux IB, Finlay JL. 2009;27(3):385–9.
Adjuvant chemotherapy of childhood posterior fossa ependymoma: 64. Kretschmar C, Kleinberg L, Greenberg M, Burger P, Holmes E,
cranio-spinal irradiation with or without adjuvant CCNU, vincris- Wharam M. Pre-radiation chemotherapy with response-based radi-
tine, and prednisone: a Childrens Cancer Group study. Med Pediatr ation therapy in children with central nervous system germ cell
Oncol. 1996;27(1):8–14. tumors: a report from the Children’s Oncology Group. Pediatr
50. Geyer JR, Zeltzer PM, Boyett JM, Rorke LB, Stanley P, Albright Blood Cancer. 2007;48(3):285–91.
AL, et al. Survival of infants with primitive neuroectodermal 65. Da Silva NS, Cappellano AM, Diez B, Cavalheiro S, Gardner S,
tumors or malignant ependymomas of the CNS treated with eight Wisoff J, et al. Primary chemotherapy for intracranial germ cell
drugs in 1 day: a report from the Childrens Cancer Group. J Clin tumors: results of the third international CNS germ cell tumor
Oncol. 1994;12(8):1607–15. study. Pediatr Blood Cancer. 2010;54(3):377–83.
51. Zacharoulis S, Levy A, Chi SN, Gardner S, Rosenblum M, Miller 66. Yoo KH, Lee SH, Lee J, Sung KW, Jung HL, Koo HH, et al.
DC, et al. Outcome for young children newly diagnosed with epen- Improved outcome of central nervous system germ cell tumors:
dymoma, treated with intensive induction chemotherapy followed implications for the role of risk-adapted intensive chemotherapy.
by myeloablative chemotherapy and autologous stem cell rescue. J Korean Med Sci. 2010;25(3):458–65.
Pediatr Blood Cancer. 2007;49(1):34–40. 67. Bouffet E. The role of myeloablative chemotherapy with autolo-
52. Massimino M, Gandola L, Barra S, Giangaspero F, Casali C, gous hematopoietic cell rescue in central nervous system germ cell
Potepan P, et al. Infant ependymoma in a 10-year AIEOP tumors. Pediatr Blood Cancer. 2010;54(4):644–6.
(Associazione Italiana Ematologia Oncologia Pediatrica) experi- 68. Zissiadis Y, Dutton S, Kieran M, Goumnerova L, Scott RM,
ence with omitted or deferred radiotherapy. Int J Radiat Oncol Biol Kooy HM, et al. Stereotactic radiotherapy for pediatric intracranial
Phys. 2011;80(3):807–14. germ cell tumors. Int J Radiat Oncol Biol Phys. 2001;51(1):
53. Grundy RG, Wilne SA, Weston CL, Robinson K, Lashford LS, 108–12.
Ironside J, et al. Primary postoperative chemotherapy without radio- 69. Malone K, Croke J, Malone C, Malone S. Successful salvage using
therapy for intracranial ependymoma in children: the UKCCSG/ combined radiation and ABMT for patients with recurrent CNS
SIOP prospective study. Lancet Oncol. 2007;8(8):696–705. NGGCT following failed initial transplant. BMJ Case Rep. 2012.
54. Aggarwal R, Yeung D, Kumar P, Muhlbauer M, Kun LE. Efficacy 70. Tavangar SM, Larijani B, Mahta A, Hosseini SMA, Mehrazine M,
and feasibility of stereotactic radiosurgery in the primary manage- Bandarian F. Craniopharyngioma: a clinicopathological study of
ment of unfavorable pediatric ependymoma. Radiother Oncol. 141 cases. Endocr Pathol. 2004;15(4):339–44.
1997;43(3):269–73. 71. Kobayashi T, Mori Y, Tsugawa T, Hashizume C, Takahashi
55. Vaidya K, Smee R, Williams JR. Prognostic factors and treatment H. Prognostic factors for tumor recurrence after gamma knife
options for paediatric ependymomas. J Clin Neurosci. 2012;19(9): radiosurgery of partially resected and recurrent craniopharyngio-
1228–35. mas. Nagoya J Med Sci. 2012;74(1–2):141–7.
56. Stauder MC, Ni Laack N, Ahmed KA, Link MJ, Schomberg PJ, 72. Niranjan A, Kano H, Mathieu D, Kondziolka D, Flickinger JC,
Pollock BE. Stereotactic radiosurgery for patients with recurrent Lunsford LD. Radiosurgery for craniopharyngioma. Int J Radiat
intracranial ependymomas. J Neurooncol. 2012;108(3):507–12. Oncol Biol Phys. 2010;78(1):64–71.
57. Kano H, Yang H, Kondziolka D, Niranjan A, Arai Y, Flickinger JC, 73. Hasegawa T, Kobayashi T, Kida Y. Tolerance of the optic apparatus
et al. Stereotactic radiosurgery for pediatric recurrent intracranial in single-fraction irradiation using stereotactic radiosurgery: evalu-
ependymomas. J Neurosurg Pediatr. 2010;6(5):417–23. ation in 100 patients with craniopharyngioma. Neurosurgery.
58. Gajjar A, Packer RJ, Foreman NK, Cohen K, Haas-Kogan D, 2010;66(4):688–94; discussion 694–5.
Merchant TE. Children’s Oncology Group’s 2013 blueprint for 74. Yeung JT, Pollack IF, Panigrahy A, Jakacki RI. Pegylated interferon-
research: central nervous system tumors. Pediatr Blood Cancer. α-2b for children with recurrent craniopharyngioma. J Neurosurg
2013;60(6):1022–6. Pediatr. 2012;10(6):498–503.
Pineal Region Tumors
34
Gregory P. Lekovic and Andrew G. Shetter
children, for whom whole brain irradiation carries significant

Introduction and Background morbidity. Gamma knife radiosurgery (GKRS) offers a simi-
lar advantage in providing highly conformal targeting with
The pineal gland is situated at the geographic center of the maximization of delivered dose to the therapeutic target and
brain between the posterior and habenular commissures. Its concomitant minimization of dose exposure to neighboring
central location, as well as its solitary nature in an organ com- structures. Although the role of radiosurgery in the manage-
prised of paired structures, led early anatomists and physiolo- ment of pineal region tumors is still evolving, it is safe to say
gists to impart to it singular importance (e.g., famously that radiosurgery plays an important role in the multimodal-
asserting that the pineal was the seat of the human soul). ity management of these lesions.
As cited by Baumgartner and Edwards [1], the first
description of a lesion in the pineal region is attributed to
Virchow in 1865. By the early twentieth century, neurosurgi- Anatomy
cal pioneers had devised approaches to the pineal region,
variants of which are still in use today, including the trans- The pineal region is broadly defined as the region lying
callosal approach of Dandy and Fedor Krause’s supracere- between the splenium of the corpus callosum and tela cho-
bellar infratentorial approach. These early experiences with roidea dorsally, and the quadrigeminal plate and tectum ven-
surgery for pineal regions were highly morbid, a fact that is trally, and between the posterior third ventricle rostrally, and
not surprising given the technological limitations of the time cerebellar vermis caudally. The pineal gland itself is situated
[2]. At the same time, the radiosensitivity of germinomas, between the habenula dorsally and posterior commissure
the most common pineal region tumor, was beginning to be anterior-inferiorly (Fig. 34.1).
appreciated. Gradually, surgical approaches were given up in The pineal region is notable in that it is a region rich in
lieu of empiric treatment with radiotherapy. With the advent vasculature, especially the deep cerebral veins. The vascular
of the era of microneurosurgery, interest in the surgical man- supply to the pineal gland proper is predominantly from the
agement of pineal region tumors remerged, and empiric medial posterior choroidal arteries, with contributions from
treatments with radiotherapy are now obsolete. the lateral posterior choroidal and quadrigeminal arteries as
Most recently, the development of chemotherapy and well [3]. However, the pineal is surrounded by the deep
radiosurgery has added additional variables to the treatment venous drainage of the brain on all sides—the velum inter-
algorithm of these tumors. Chemotherapy has been used suc- positum, containing the internal cerebral veins, lies rostral
cessfully to lower radiation doses used to treat patients with and dorsal to the pineal gland (to which it is often attached to
germinoma. This is especially important in the treatment of pineal region tumors by thick arachnoid adhesions) [4], the
basal veins of Rosenthal are located at either flank, and the
G.P. Lekovic, M.D., Ph.D. vein of Galen lies posterior and superior to it. For this reason,
House Clinic, Good Samaritan Gama Knife Radiosurgery Center, stereotactic biopsy of the pineal region is viewed with trepi-
2100 West Third Street, Los Angeles, CA 90057, USA dation by some.
A.G. Shetter, M.D. (*) The cellular constituents of the pineal region include
Section of Functional Stereotactic Neurosurgery, Division pinealocytes, astrocytes, and sympathetic nerves. The sym-
of Neurological Surgery, Barrow Neurological Institute, St.
pathetic innervation comes from the superior cervical gan-
Joseph’s Hospital and Medical Center,
350 W. Thomas Road, Phoenix, AZ 85013, USA glion. The physiological role of the pineal is thought to be
e-mail: neuropub@dignitiyhealth.org concerned with the regulation of circadian rhythms.
446 G.P. Lekovic and A.G. Shetter
In Western populations, GCTs account for 0.3–0.5 % of

all primary intracranial neoplasms, but comprise approxi-
mately 3.0 % of those encountered in children. In contrast, in
Asia, GCTs comprise at least 2.0 %, and up to 9–15 %, of all
intracranial neoplasms and pediatric neoplasms, respec-
tively. In Japanese and Korean populations, 80 % of pineal
region tumors in patients aged 15–35 years of age are germi-
noma [6]. In contrast, out of 370 French patients undergoing
stereotactic biopsy of pineal region tumors, only 51 % of
patients under the age of 30 had radiosensitive tumors
(germinoma + pineoblastoma) [7].
The most common presentation of pineal region tumors is
from symptoms attributable to local mass effect, principally
increased intracranial pressure related to aqueductal com-
pression and/or to mass effect on the quadrigeminal plate
(e.g., Parinaud’s syndrome). The growth pattern of pineal
region tumors tends to reflect the underlying histology, with
benign, well-encapsulated tumors filling the posterior third
Fig. 34.1 Midsagittal section demonstrating the anatomy of the pineal ventricle, quadrigeminal cistern, and displacing the anterior
region. Labeled structures include the pineal body (PB), the subforni- cerebellum [4]. Malignant tumors are more likely to dif-
ceal organ (SFO), choroid plexus (CP), area postrema (AP), neurohy- fusely infiltrate the midbrain, thalamus, or other neighboring
pophysis (NH), median eminence (ME), and organum vasculorum of structures.
the lamina terminalis (OVLT). Used with permission from Barrow
Neurological Institute Importantly, there is tremendous variation in the natural
history of pineal region tumors depending on their histology.
The management of tumors of this region, therefore, is
Ontologically and embryologically the pineal is related to extremely dependent on an accurate diagnosis. In most cases,
photoreceptor organs. Indeed, this role persists in certain this necessitates a tissue diagnosis, whether through stereo-
reptiles where the pineal is known as the “parietal eye.” tactic, endoscopic, or open biopsy. This issue will be dis-
Because of this shared evolutionary background, pineoblas- cussed in greater detail below.
tomas are sometimes seen in association with retinoblastoma
(so-called “trilateral retinoblastoma”).
Pathology
Pineal Region Tumor Presentation Germ Cell Tumors

and Natural History
Pure germinoma, the most common central nervous system
Although the physiological role of the pineal in humans is GCT, consists of large undifferentiated cells arranged in
still somewhat obscure, it is an important locus of pathology. monomorphous sheets. Nuclei are round and typically cen-
Broadly speaking, there are four categories of lesions occur- trally positioned amidst abundant clear cytoplasm. Mitoses
ring in the pineal region: germinoma (including pure germi- are common and necrosis uncommon. Some tumors may
noma, teratoma, and nongerminomatous germ cell tumors, incite an inflammatory response that is evident on histology
NGCTs), pineal parenchymal tumors (pineocytoma and by the presence of either many lymphocytes or occasionally
pineoblastoma), glial tumors, and miscellaneous tumors fibrous tissue. Immunohistochemical positivity for placental
(including metastases, meningioma, and ependymoma). alkaline phosphatase is typical.
Although pineal region tumors are relatively rare
(accounting for less than 1 % of all intracranial neoplasms),
they are much more common in Asia [5]. It is well known Nongerminomatous Germ Cell Tumors
that there is a higher incidence of germ cell tumors (GCTs)
in Asian populations than in North American or European The group of tumors that together comprise the nongermino-
populations. GCTs typically account for the majority of matous germ cell tumors (NGGCTs) include teratoma
pineal region tumors, with pineal parenchymal tumors sec- (mature and immature), yolk sac tumors, embryonal carci-
ond and miscellaneous tumors third. noma, and choriocarcinoma. Only teratoma is commonly
34 Pineal Region Tumors 447
encountered as a pure tumor type, with most NGGCTs hav- Because many tumors of the pineal region are histologically
ing regional variations in histology suggestive of multiple heterogeneous (especially the NGGCTs), the ability of the sur-
tumor subtypes. geon to avoid sampling errors in obtaining tissue for histopath-
By definition, teratomas recapitulate somatic elements from ological examination is paramount. In general, this need for
endodermal, mesodermal, and ectodermal lines. Differentiated accurate histological diagnosis of disparate regions of tumor
teratomas include well-formed, fully differentiated elements has been commonly cited as support for open or endoscopic
such as rests of skin, hair, teeth, or cysts of mucosal epithelia. biopsies, during which regions of tumor appearing grossly dis-
Immature teratomas are much more common, and contain similar can be biopsied separately.
incompletely differentiated tissue. Occasionally, a carcinoma
or sarcoma can arise from within a teratoma (so-called tera-
toma with malignant transformation). Imaging
Yolk sac tumors are characterized by a reticular epithe-
lium set in a myxoid matrix; classically these may form Magnetic resonance imaging (MRI) with and without gado-
papillae known as Schiller–Duvall bodies. Embryonal carci- linium contrast enhancement is the diagnostic imaging
noma is composed of sheets of large cells with prominent modality of choice. Pineocytomas are typically hypointense
nucleoli, many mitoses, and abundant cytoplasm. Necrosis is on T1-weighted images, and hyperintense on T2-weighted
common. Trophoblastic elements, such as syncytiotropho- sequences. Pineoblastomas are heterogeneous, and can be
blastic cells (which may be seen scattered with other CNS either hypo- to isointense on T1-weighted MRI. GCTs are
GCTs) indicate tumor trophoblastic differentiation and typically well circumscribed, iso- to hyperdense relative to
hence choriocarcinoma. grey matter, and intensely enhancing. However, no certain
The immunohistochemistry of GCTs is helpful in differ- method exists for differentiating pineal region tumors on the
entiating tumor types. Typically, cell membranes and/or basis of imaging studies alone.
cytoplasm are positive for placental alkaline phosphatase in Tumors of the pineal region enhance variably, as does the
germinoma; beta-HCG and human placental lactogen in cho- normal pineal gland due to the absence of a blood–brain bar-
riocarcinoma, and alpha-fetoprotein in yolk sac tumor and rier. Incidental calcification of the pineal is common,
teratoma. although such calcifications are also associated with “brain
sand” seen in pineocytomas. The diagnosis of GCT is sug-
gested if the neoplasm appears to surround or engulf the nor-
Pineal Parenchymal Tumor mal pineal gland calcifications, whereas pineal parenchymal
tumors are said to cause an “explosion” of pineal calcifica-
Pineocytomas (aka, pinealocytoma) are slow growing neo- tions to the periphery of the lesion as the mass expands. On
plasms of the pineal gland composed of small uniform sized the other hand, hemorrhage (pineal apoplexy) is suggestive
cells. Characteristically, the cells exhibit cytoplasmic proof choriocarcinoma. In general, tumors arising from the col-
cesses that are said to resemble club-like endings. Unlike licular plate are more likely to be of glial origin [1], and the
pineoblastomas, there is some preservation of the lobular presence of fat signal is characteristic of lipoma, mature tera-
architecture of the normal pineal gland. toma, or dermoid tumors. The presence of mature teratoma-
Pineoblastomas are members of the family of small blue tous elements such as hair or teeth may also be evident on
cell primitive neuroectodermal tumors, similar to medullo- imaging studies.
blastoma or retinoblastoma. Like these tumors, they are
highly cellular neoplasms with a high nucleus to cytoplasm
ratio, characterized by the presence of either Homer–Wright Management of Pineal Region Tumors
or Flexner–Wintersteiner rosettes, the latter being a mark of
retinoblastic differentiation. True to the oncologic common- Because of the varied natural history of pineal region tumors
ality between the pineal and photoreceptor organs, pineal and the lack of diagnostic specificity of imaging studies, the
parenchymal tumors stain positively for interphotoreceptor importance of obtaining a tissue diagnosis cannot be over-
retinoid binding protein. emphasized. Therefore, a biopsy, whether stereotactic, endo-
True intermediate tumor classification is relatively rare, scopic, or open (see discussion below) should be obtained as
with only about 8 % of pineal parenchymal tumors belong- the initial step in the management of tumors of this region in
ing to this category (WHO). These tumors are highly cellular all cases, except where cerebrospinal fluid (CSF) is positive
with numerous mitoses and nuclear atypia, but without for markers of malignant germinoma (i.e., BHCG or alpha-
pineocytomatous rosettes. Even more rarely, a pineal paren- fetoprotein), in the case of new metastatic lesions (where the
chymal tumor may have within it areas of both pineocytoma- tissue diagnosis is already known), or perhaps in patients
tous and pineoblastomatous differentiation. who are too frail to undergo even a biopsy. As discussed in
more detail below, there is no role for empiric radiation ther-

apy in the management of these tumors, at least not in occi- The Role of Surgery
dental populations.
Whether it is better to perform an open biopsy, coinci- Pineal region tumors can be successfully resected with
dent with attempted gross total resection or cytoreductive microsurgical techniques at tolerable levels of morbidity and
surgery, or to perform a minimally invasive (i.e., stereotactic mortality. Approaches to the pineal region include both
or endoscopic biopsy) is controversial. Proponents of open supratentorial (interhemispheric transcallosal and occipital
biopsy cite the relative safety and low morbidity of transtentorial) and infratentorial (supracerebellar infratento-
approaches to the pineal region given modern instrumenta- rial approach, Fig. 34.2). Bruce and Stein [12] described
tion and microsurgical technique. Because of the histopath- their experience with 160 pineal region tumors. They
ologic heterogeneity common to tumors of this region, the achieved a gross total resection in 46/53 benign pineal region
ability to minimize sampling errors with open biopsy may tumors, with overall mortality and morbidity of 4 % and 3 %,
yield more accurate specimens for diagnosis. Benign tumors respectively. Their results are comparable to other major sur-
may be completely respectable; therefore, an open approach gical series for whom mortality rates range from 2 to 11 %
offers the potential for cure of these lesions. Moreover, even and serious morbidity from 3 to nearly 30 % (for a recent
malignant tumors and radiation sensitive tumors may bene- review, see [13]).
fit from cytoreduction. The role of surgery for maximal debulking and/or resec-
On the other hand, minimally invasive approaches avoid tion is best established for benign lesions, where gross total
the risks of craniotomy, and have high diagnostic yield. resection may be curative. Conversely, these are also the
Because of the pineal region’s proximity to the deep cere- lesions best suited to radiosurgical treatment. The argument
bral veins, some authors argue that stereotactic biopsy of for surgery for benign pineal region tumors is that gross total
this region is a relatively higher risk. However, Regis et al. resection is often possible, and that this may afford a perma-
[7] reviewed 370 stereotactic biopsies of the pineal region nent cure and obviate the need for CSF diversion.
from 15 Fr neurosurgical centers. They reported a 1.3 % Whether surgical resection or debulking plays a signifi-
mortality and 0.8 % severe neurological morbidity attrib- cant role in the management of malignant tumors is much
utable to stereotactic biopsy. This was believed not to more hotly debated. Aggressive malignancies, especially
reflect an increased risk of death or neurologic injury over those that on preoperative imaging are seen to invade the
other intracranial sites. Overall, diagnostic tissue was brainstem or other neighboring structures, are unlikely to
obtained in 94 % of patients. However, 2.3 % of biopsied be amenable to complete resection. Nevertheless, some
patients underwent repeat biopsy or subsequent open sur- studies have shown benefit of cytoreduction even in these
gery and were found to have been initially misdiagnosed. cases, although these data have not reached statistical
Sampling errors were felt to contribute to about half of significance [14, 15]
these cases (i.e., about 1 % diagnostic error due to sam-
pling errors).
The endoscopic approach to the pineal region for the pur-
poses of biopsy was first described by Fukushima [8, 9]. The
advantages of an endoscopic approach include direct visual-
ization of the biopsy site which could lessen sampling errors
and lead to greater diagnostic accuracy [10]. In addition,
because many patients with pineal region tumors either pres-
ent with, or are at high risk of hydrocephalus, a CSF diver-
sionary procedure may be required. An endoscopic biopsy
affords the possibility of simultaneously performing an
endoscopic third ventriculostomy at the same operative sit-
ting. Oi et al. reported their experience with 20 consecutive
patients followed prospectively who underwent endoscopic
biopsy in the initial management of their pineal region
tumors [11]. Importantly, they reported the endoscopic
detection of spread of tumor not visualized in preoperative
neuroimaging. This suggests that endoscopy may have ben- Fig. 34.2 Surgical approaches to the pineal region. Midsagittal section
of the human brain demonstrating the surgical corridors to the pineal
efit as a diagnostic tool per se. Nevertheless, the proper role
region, the transcallosal, occipital transtentorial, and the supracerebel-
of the endoscope in the management of pineal region lesions lar infratentorial. Used with permission from Barrow Neurological
remains controversial. Institute
Finally, some centers advocate “second-look” surgery if Furthermore, 7 of 71 patients treated died of chemotherapy
imaging abnormalities persist after initial treatment of related toxicity.
NGGCTs with chemotherapy and radiation [13]. The major- In contrast, Buckner et al. [19] reported the results of a
ity of these surgeries find either necrosis or, in the case of Phase II trial of primary chemotherapy followed by reduced-
mixed NGGCTs, rests of mature teratoma. dose radiation for the treatment of CNS GCT in which the
partial brain dose was reduced to 30 from 54 Gy. All patients
were alive without progression at 51-month mean follow-up.
Role of Radiation One patient relapsed distally (spinal cord) and was salvaged
with spinal irradiation. Thus, the prevailing paradigm at the
It has been known for decades that radiation is curative for present time is to use chemotherapy as an adjuvant therapy in
intracranial germinoma [16], with 5-year survival rates of up combination with reduced-dose radiation. Spinal irradiation
to 95 % reported for pure germinoma and up to 76 % for should be reserved for those cases presenting with dissemi-
NGGCTs. Indeed, radiation remains the mainstay of treat- nated disease or in the case of relapse.
ment of most pineal region tumors, whether as primary treat-
ment modality or as adjuvant therapy. Nevertheless, radiation
has significant morbidity. Efforts at reducing radiation expo- The Role of Radiosurgery
sure have included avoiding whole brain and prophylactic
craniospinal irradiation, the use of chemotherapy, and The role of radiosurgery in the management of pineal region
radiosurgery. tumors remains controversial. Traditionally, pineal region
Traditionally, germinoma was treated with craniospinal tumors have either been approached surgically with the goal
irradiation, given estimates of spinal cord metastasis or seed- of gross total resection and cure (if benign), or with conven-
ing after surgery range of 13–40 %. However, as the sequelae tional radiation if the tumor is of a histology known to be
of prophylactic craniospinal irradiation have become more radiosensitive and/or malignant. Radiosurgery, on the other
appreciated, efforts have been made to delay or avoid spinal hand, has the potential to be used either as a “boost” modal-
irradiation, and to limit cranial irradiation from whole brain ity in conjunction with conventional radiotherapy, or even as
to local field. The effectiveness of partial brain fields an alternative to radiotherapy and/or surgery altogether.
comprising tumor plus a 2-cm margin was shown initially by Table 34.1 summarizes the results of reports in the English
Dattoli and Newall in 1990 [17]. In this retrospective study, language literature, including three from our institution, on
12 patients were treated with partial brain irradiation, con- the use of stereotactic radiosurgery in the treatment of pineal
sisting of either fields comprising tumor plus a 2-cm margin region tumors. Several factors, however, make it difficult to
(n = 10), or fields comprising the ventricular system with a draw easy comparisons or conclusions from these studies,
boost to the tumor (n = 2). Nine out of ten patients treated including relatively small sample sizes and because of the
with partial brain irradiation were complete responders; heterogeneity of diagnoses included within individual series.
however, one patient relapsed and ultimately succumbed to Backlund et al. [20] first reported the treatment of two
the disease. This patient, however, also received a reduced cases of pineocytoma with GKRS in 1974. The patients were
dose of less than 40 Gy. Thus, the authors concluded that treated with peak dose of 50 Gy. At 13 and 36 months fol-
whole neuraxis radiation could be safely avoided in the low-up neither patient had displayed evidence of tumor pro-
majority of patients, provided that sufficient tumoricidal gression. Subach et al. [21] included eight patients with
doses of radiation were delivered. pineocytoma (out of 14), with three tumors demonstrating a
Because gonadal GCTs respond well to chemotherapy, complete response to GKRS, three a partial response, and
chemotherapy has been advocated as a means to further two no change. GKRS mean marginal doses for all tumor
reduce the amount of radiation necessary to obtain control of types treated were 15.4 Gy, and all tumors were treated to the
the tumor. In this regard, the rationale for chemotherapy 50 % isodose line. Hasegawa et al. [22], however, reported a
anticipates the argument made for radiosurgery. Balmaceda distant recurrence of pineocytoma after GKRS, although
et al. [18] enrolled 71 patients with GCTs (including both 100 % local control was achieved. A complete response of
pure and NGGCTs) in an international cooperative study pineoblastoma to GKRS was demonstrated by Manera et al.
comprising 31 institutions in 6 countries, and demonstrated [23] in one patient and a partial response in another.
that chemotherapy could not be used alone for the treatment However, Hasegawa et al. [22] experienced three distant
of GCTs. They employed a high-dose regimen of carbopla- failures following GKRS for pineoblastoma treated with a
tin, etoposide, and bleomycin. Although 41 of the 71 patients mean marginal dose of 15.3 Gy to the 50 % isodose line.
were successfully managed with chemotherapy alone initially, Kano et al. retrospectively reviewed 13 patients with pineo-
35 patients showed evidence of recurrence or progression. cytoma treated with a median marginal dose of 15 Gy [24].
450
Table 34.1 Summary of pineal region tumors in literature

Mean
follow-up
Reference No. and diagnosis Biopsy (months) Prior treatment Radiological results Dosimetry Progression Complication
Backlund et al. 2 pineocytoma 2 SB 13 and 36 None Regression of tumor GKRS 50 Gy None None
(1974) [20]
Dempsey and 4 meningioma, 2 4 SB, 5 OB 20.7 4 primary, 5 OR, 6 partial or complete GKRS mean None 3 new
Lunsford anaplastic (3–32) 4 with previous responses, 3 no marginal dose of neurological
(1992) [32] astrocytoma, 1 WBRT, 1 change (all 16.6 Gy (10–2- Gy) deficits
ependymoma, 1 chemotherapy meningioma) at 50–60 % isodose
craniopharyngioma, line
1 pineocytoma
Manera et al. 1 pineocytoma, 1 7 SB, 4 NO 12.3 1 OR 3 complete responses GKRS mean None None
(1996) [23] astrocytoma, 2 (2–34) (1pinealoblastoma, 2 marginal
germinoma, 2 germinoma), 5 partial, dose = 9.3 Gy
pinealoblastoma, 3 2 no change, 1 (6–20), 40–50 %
meningioma insufficient follow-up isodose line
Subach et al. 8 pineocytoma, 9 SB, 2 OB, Mean 1 55 Gy EBRT, 1 4 complete responses GKRS marginal 1 (embryonal 1 new Parinaud’s,
(1998) [21] 2pineoblastoma, 2 3 NO imaging subtotal OR, 2 (3 pineocytoma and 1 doses 15.4 (12–20), carcinoma) 3 deaths (only 1
germinoma, 2 follow-up chemotherapy pineoblastoma), 6 all treated to 50 % attributed to
NGGCT 21 partial (3 isodose line mean progression)
pineocytoma, 1 volume 6.4 cm3
pineoblastoma, 2
germinoma)
Kobayashi et al. 8 germinoma, 4 11 NO N/A Germinoma: 8 8 complete response, GKRS marginal 8 7 deaths
(2001) [26] STGC, 13 chemotherapy, 5 overall response dose = 16.8 Gy for attributable to
malignant GCT, 3 OR malignant rate = 73.3 %, For germinoma, 13.4 Gy progression
pineocytoma, 2 GCT: 10 OR, germinoma and for malignant GCT,
pineoblastoma, 2 chemotherapy pineocytoma 100 % 17.5 Gy for PPT and
unknown 10, EBRT 8 control rate others
Hasegawa et al. 10 pineocytoma, 2 10 SB, 4 OB 52 4 OR, 4 EBRT (2 4 complete response, GKRS mean 4 distant treatment 2 new deficits, 5
(2002) [22]a mixed histology, 4 as boost and 2 8 partial, 2 no change, marginal failure (1 pineocytoma, deaths (4
pineoblastomas after recurrence) 2 no imaging dose = 15.3 Gy at 3 pineoblastomas) attributable to
follow-up, local 50 % isodose line progression)
control = 100 %
Hasegawa et al. 4 NGGCT 1 SB, 2 OB, 25 2 OR 2 partial, 1 no change, 12–16 Gy marginal 1 (death) None
(2003) [28]a 1 CSF 1 progression dose to 50 %
isodose line
G.P. Lekovic and A.G. Shetter
34
Deshmukh 5 pineocytoma 5 OR 14.6 5 OR 4/5 partial response GKRS marginal None None
et al. (2004) 100 % local control dose 14–16 Gy
[33]
Casentini et al. 6 germinoma 5 yes 27.6 “Inverse boost” Complete response Varian LINAC None 1 periprocedural
(1990) [25] (1–58) paradigm with (marginal dose death (unrelated)
extended field 10–11 Gy, peak
radiotherapy: doses 10–12.5)
mean 31 Gy,
range 24–36)
Pineal Region Tumors
Kano et al. 13 pineocytoma, 5 20 with OB 54.1 6 OR, 3 EBRT, 3 26 % complete, 47 % GKRS mean 6 deaths, 5-year No permanent
(2009) [24] pineoblastoma, 2 or SB chemotherapy, 2 partial, 11 % “stable,” marginal dose 15 Gy progression-free morbidity
mixed histology EBRT and 11 % local survival 89 %
chemotherapy progression
Mori et al. 38 germ cell tumor, Biopsy in 13 33.5 13 OR, 20 EBRT, NR Mean marginal dose Progression-free 1 optic
(2009) [27] 9 PPT, 2 unknown germ cell 27 chemotherapy 15.5 Gy survival at 10 years neuropathy
tumors, OB 68 % for germ cell
in 9 PPT tumors, 67 % for PPTs
Wilson et al. 14 pineocytoma 12 OB, 2 EB 53 12 OR, 5 with 5 complete resection, Mean marginal dose Progression in 3 None for GKRS
(2012) [30] gross total 9 residual tumors of 14.6 Gy with patients with subtotal
excision postoperatively GKRS in 5 patients excision, none in
patients with gross
total excision or GKRS
Yianni et al. 11 PPT, 2 germ cell 16 SB, 62.5 17 OR, 10 EBRT, NR Mean marginal dose
8 deaths, progression- None
(2012) [31] tumors, 2 papillary 11 EB, 17 4 chemotherapy 18 Gy free survival 27 %
epithelial tumors, 9 OB at 5 years
gliomas, 20 with no
histologic diagnosis
Used with permission from Barrow Neurological Institute
GKRS Gamma knife radiosurgery, NR not reported, NGGCT nongerminomatous germ cell tumor, SB stereotactic biopsy, OB open biopsy (craniotomy), EB endoscopic biopsy, STGC germinoma
with syncytiotrophoblastic giant cells, OR surgery, EBRT external beam radiation therapy, PPT pineal parenchymal tumor
a
Patients partially included in previous reports
451
Overall survival after GKRS was 92.3 % at 5 years, with All patients were treated using Leksell Gamma Plan treat-
23 % of tumors treated demonstrating a complete radio- ment planning software, with a mean marginal dose of
graphic response. 14.06 Gy (range 12–18 Gy). All doses were prescribed to the
Although germinoma is highly curable by external beam 50 % isodose line. Seven to 27 isocenters (mean 11.4) were
irradiation, efforts to reduce the dose of EBRT radiation have used to treat a mean target volume of 7.42 cm3 (range
gained momentum because of concerns of the toxicity of 1.2–32.5 cm3).
even partial brain irradiation (especially in children). In Fourteen patients were treated with GKS as either the pri-
1990, Casentini et al. [25] reported their results using Varian mary radiation modality or as a salvage treatment for a recur-
LINAC radiosurgery as an “inverse boost” paradigm for the rence after conventional EBRT failed. Two patients, one with
treatment of germinoma, with all six patients so treated dem- an NGGCT and another with an anaplastic astrocytoma,
onstrating a complete response. Kobayashi et al. [26] were administered a 15- and 12-Gy GKS boost to the pineal
included eight patients with the diagnosis of germinoma who region after receiving reduced-dose craniospinal radiation
were treated with chemotherapy prior to GKRS. They therapy (3,600 cGy).
obtained a 100 % control rate for these tumors with a mar- There were no complications attributable to GKRS. There
ginal dose of 16.8 Gy. Mori et al. [27] treated 18 germinomas were three mortalities after GKRS. One patient died 6 days
with GKRS, the largest case series to date. Marginal doses after radiosurgery (this patient was excluded from further
ranged from 9.9 to 25.7 Gy, and all but one patient received analysis), one died 2 months after radiosurgery, and the third
fractionated external beams radiation therapy as well. The died after developing widespread metastatic disease. These
progression-free survival rate at 5 years was 63 %. The sin- latter two patients demonstrated local control of tumor,
gle patient who received GKRS as the initial therapy experi- despite of their clinical progression. Upon latest follow-up
enced CSF dissemination at 15 months. imaging, local control was established in 100 % of the
Increasingly, GKRS is being reported in the treatment patients.
of NGGCTs. The number of reported cases is now well Wilson et al. examined a subset of 15 patients with histo-
over 50. However, due to the heterogeneity of diagnoses in logically proven pineocytomas treated at the BNI over a
the reported series it is difficult to draw comparisons. 12-year interval [30]. The mean clinical and radiographic
Nevertheless, it is clear that NGGCTs are less responsive follow-ups were 44 and 53 months, respectively. Gross total
to GKRS than are germinoma or pineocytoma. Mori et al. tumor excisions were achieved in five patients with no known
[27] used adjuvant GKRS for 16 NGGCTs with marginal recurrences, although two of those patients were lost to
doses similar to those for GCTs. The progression-free sur- radiographic follow-up. Of the nine patients who underwent
vival rate at 5 years was 37 % [27]. Neither of the two subtotal tumor excisions or endoscopic biopsy, three received
NGGCTs reported by Subach et al. [21] demonstrated a initial GKRS and six were observed. None of the patients
response to GKRS. Hasegawa et al. [28] treated four treated with “up-front” radiosurgery recurred. The remaining
NGGCTs with between 12 and 16 Gy at the 50 % isodose six patients demonstrated recurrence at varying intervals.
line. At a mean follow-up of 25 months, one tumor pro- Four received subsequent radiosurgery and developed no
gressed, one demonstrated no change, and two tumors par- further signs of tumor progression. The authors concluded
tially responded. Kobayashi et al. [26] reported the results that adjuvant GKRS is an effective treatment for pineocyto-
of 13 malignant GCTs treated with GKRS. Follow-up data mas when complete surgical excision is not feasible, and that
was obtained in 12 patients, of whom 3 demonstrated a radiosurgery can be successful in controlling recurrent dis-
complete response, 3 a partial response, and 6 had pro- ease. Their data also suggests that subtotal pineocytoma
gressed. Five patients were deceased at a mean follow-up excision plus radiosurgery may yield results equivalent to
of 12.6 months. gross total surgical excision, although greater patient num-
bers and larger follow-ups will be needed before this finding
can be stated with certainty.
Barrow Neurological Institute GKRS
Experience
Conclusions
We recently reported our experience with GKRS for pineal
region tumors in 17 patients with non-metastatic tumors of the Although tumors in this region are relatively rare, the pineal
pineal region [29]. Diagnoses included pineocytoma (n = 8), as region is an important site of a wide range of pathological
well as choroid plexus papilloma, neurocytoma, anaplastic processes. Because of the wide variation in natural history of
astrocytoma, PNET, low-grade astrocytoma, pineoblastoma, these tumors depending on this histological heterogeneity,
NGGCT, malignant teratoma, and pineal parenchymal tumor obtaining a tissue diagnosis prior to instituting definitive
of intermediate differentiation (n = 1 each). therapy is paramount. Exceptions to this rule may be in met-
astatic disease (where the presumptive tissue diagnosis is sible. Moreover, the sensitivity of pineal region tumors to
already known) and in patients that are truly too poor surgi- radiation tends to weigh against surgical aggressiveness.
cal candidates. In our experience, open craniotomy with Nevertheless, because surgery does offer at least the poten-
attempted resection or at least debulking of tumor with tial for cure, we believe radiosurgery may be reserved in
biopsy has been the favored approach to obtaining a diagno- these cases for poor surgical candidates or for the treatment
sis. Endoscopic biopsy potentially spares the patient the risk of residual or recurrent disease. For malignant lesions, the
of craniotomy if the tumor is found to be radiosensitive. In role of radiosurgery is perhaps less well defined. Many
addition, endoscopy does afford the theoretical advantage of malignant lesions are surgically incurable, and likely
being able to perform an endoscopic third ventriculostomy at treatable with radiotherapy and chemotherapy alone.
the same surgery as that for obtaining a biopsy; however, in Nevertheless, like surgery, radiotherapy has significant mor-
our series all patients’ third ventriculostomies eventually bidity that may be ameliorated by using radiosurgery as
failed, ultimately requiring ventriculoperitoneal shunting. either a boost modality or as an alternative treatment alto-
Other centers have reported good results using stereotactic gether. The rationale for radiosurgery for radiation sensitive
biopsy to obtain a histologic diagnosis [7, 31]. tumors is therefore analogous to that of chemotherapy (i.e.,
The role of radiosurgery in the management of pineal to avoid, delay, or at a minimum to reduce the dose of, con-
region tumors remains controversial. We have found it help- ventional external beam radiation therapy). However, enthu-
ful to dichotomize our approach to the appropriateness of siasm for the role of radiosurgery in malignant disease must
pineal region tumors based on the grade of tumor involved. be tempered, in our experience, by the fact that we have
For benign lesions, i.e., those with an indolent natural his- found that Gamma knife as a primary radiation modality in
tory, that are relatively radioresistant, and curable with gross the treatment of malignant disease may lead to distal failure.
total surgical resection, radiosurgery competes with conven- Of course, in this case the possibility of salvage with either
tional surgery in the treatment options. Despite advances in repeat radiosurgery or radiotherapy remains. In such case,
microsurgical approaches to the pineal region, surgical radiosurgery may still be of value in delaying the exposure
extirpation of tumors occurring in this region is often impos- to whole brain external beam radiation therapy.
Case Examples
Patient #1 for supracerebellar infratentorial approach for biopsy and

A 12-year-old boy presenting with progressive headache, tumor debulking 2 days later. Despite third ventriculos-
nausea, and vomiting for 5 days underwent computed tomy, the patient developed recurrent symptoms of hydro-
tomography (CT) of the head revealing a pineal region cephalus and eventually required ventriculoperitoneal
mass with hydrocephalus at an outside institution and was shunting. Pathological examination was consistent with a
subsequently transferred to the our institution for defini- malignant mixed GCT.
tive care. The patient had a prior history of strabismus but After surgery, the patient was begun on induction che-
was otherwise previously healthy. On admission exami- motherapy with a 3-day course of VePesid, carboplatin,
nation, the patient was awake, alert, and fully oriented. He cytotaxin, and bleomycin followed by high-dose Neupogen
complained of diplopia but his extraocular movements and Procrit. He returned for subsequent cycles of chemo-
were intact to examination except for some divergence of therapy every 3 weeks for a total of four cycles. Repeat
gaze looking upwards. There were no other cranial nerve imaging demonstrated persistent disease.
findings, and motor examination revealed 5/5 motor The patient then underwent craniospinal irradiation
strength throughout without drift. Imaging revealed a with planned Gamma knife boost to the pineal region. The
pineal region tumor with hydrocephalus. Serum markers brain was treated with shaped-opposed 6 MV photon
were positive for an elevated alpha-fetoprotein of 213, beams with dose calculated at midplane. The brain was
although CSF markers were absent. treated to 3,600 cGy in 20 fractions. The spine was treated
The patient underwent stereotactic wand-guided endo- with two separate fields, one including the cervical and
scopic biopsy and third ventriculostomy. However, endo- thoracic spines primarily, and the other comprising the
scopic biopsy material was nondiagnostic, and so the lumbar region. Both fields were treated to a depth of
patient subsequently underwent a suboccipital craniotomy 4.5 cm to 3,600 cGy.
(continued)
(continued)
Fig. 34.3 A 12-year-old boy presented with hydrocephalus secondary to pineal region NGGCT. Axial (a) and sagittal (b) gadolinium-
enhanced MRI demonstrating complete of response of pineal region NGGCT 6 years after GKRS boost to craniospinal irradiation
The patient tolerated both chemotherapy and cranio- Follow up imaging studies demonstrated the complete
spinal irradiation well and was felt to be an excellent can- response of the tumor to treatment at 1 year of follow-up.
didate for GKRS boost to the pineal. Two days after the Six years after GKRS, the patient’s MRI scan continues to
completion of his course of craniospinal irradiation, the show no evidence of tumor (Fig. 34.3). He is performing
patient underwent GKRS. The prescription dose was in high school at his grade level, and he has been off treat-
15 Gy prescribed at the 50 % isodose line. The target vol- ment for 5 years, although he continues to require pitu-
ume was 1.2 cm3, and was covered in a single matrix with itary replacement therapy.
two 4-mm collimator isocenters and 6 8 mm isocenters.
Patient #2 5,580 cGy to this location. In addition, the patient under-

A 20-year-old man presented with 3 weeks of persistent went GKRS to the residual tumor, with a prescription dose
headache, dizziness, and diplopia. CT demonstrated an of 14 at the 50 % isodose line. The defined target volume
enhancing pineal region mass. The patient underwent a was 23.00 cm3, and 22.00 cm3 were covered within the
suboccipital craniotomy for supracerebellar infratentorial dose matrix with a combination of three 8-mm collimator
approach for resection of the mass. Pathological examina- isocenters and twelve 18-mm collimator isocenters.
tion was consistent with pineoblastoma. The patient sub- The patient tolerated craniospinal irradiation and
sequently underwent ventriculoperitoneal shunting. GKRS well. There were no treatment-related complica-
The patient was begun on emergent whole brain radio- tions. The patient has been followed with serial imaging
therapy, and an initial response was found after seven which at 4.5 years post GKRS demonstrate stable tumor
treatments. The patient was therefore administered cranio- with loss of central enhancement and without evidence of
spinal irradiation, with 3,600 cGy prescribed to the brain growth (Fig. 34.4). Clinically, the patient has a KPS of 100
and spine with an IMRT boost to the pineal, for a total of and has returned to work full-time.
(continued)
(continued)
Fig. 34.4 A 20-year-old man presented with pineoblastoma, treated follow-up imaging (b) demonstrates stable tumor size with a promi-
with GKRS boost after craniospinal irradiation. Sagittal gadolin- nent loss of central enhancement
ium-enhanced MRI at the time of GKRS (a), compared to that on
Patient #3 ebellar infratentorial approach. In addition, she eventually

A 52-year-old woman, a nursing home resident with required ventriculoperitoneal shunting after failure of her
severe psychiatric disease, was found to have a diminish- third ventriculostomy.
ing level of consciousness and incontinence of bowel and After surgery, the patient was evaluated for GKRS to
bladder. She denied headache and her neurological exam residual tumor. The defined target volume was 7.33 cm3.
was nonfocal. Imaging studies revealed a large (>5 cm) The prescription dose was 14 Gy to the 50 % isodose line.
contrast enhancing lesion in the pineal region. The patient The target dose volume histogram volume was 7.22 cm3;
underwent endoscopic third ventriculostomy and biopsy. the treatment plan included five 8 mm and five 14-mm
Pathological examination demonstrated a low-grade collimator isocenters. The patient tolerated the treatment
appearing lesion with preserved pineal glandular architec- well and there were no treatment-related complications.
ture; the MIB labeling index was less than 0.1 %. Because After treatment, the patient returned to her care facility
of the benign pathological appearance of the tumor, the in her usual state of health. On 25-month follow-up, MR
patient was taken back to the operating room for resection imaging demonstrated slight shrinkage of the tumor
of the tumor via a suboccipital craniotomy and supracer- (Fig. 34.5).
(continued)
Fig. 34.5 A 52-year-old woman presented with a large (>5 cm) after treatment, axial MRI with contrast through the pineal
pineocytoma. The patient underwent subtotal resection with demonstrates stable appearance of the lesion (b)
treatment of 7.3 cm3 of residual tumor with GKRS (a). Two years
11. Oi S, Shibata M, Tominaga J, et al. Efficacy of neuroendoscopic

References procedures in minimally invasive preferential management of pineal
region tumors: a prospective study. J Neurosurg. 2000;93:245–53.
12. Bruce JN, Stein BM. Surgical management of pineal region tumors.
1. Baumgartner JE, Edwards MS. Pineal tumors. Neurosurg Clin N Acta Neurochir (Wien). 1995;134:130–5.
Am. 1992;3:853–62. 13. Bruce JN, Ogden AT. Surgical strategies for treating patients with
2. Bruce JN. Pineal tumors Historical perspective. In: Winn HR, edi- pineal region tumors. J Neurooncol. 2004;69:221–36.
tor. Youman’s neurologica surgery. Philadelphia: Elsevier; 2005. 14. Matsutani M, Sano K, Takakura K, et al. Primary intracranial germ
p. 1–6. cell tumors: a clinical analysis of 153 histologically verified cases.
3. Yamamoto I, Kageyama N. Microsurgical anatomy of the pineal J Neurosurg. 1997;86:446–55.
region. J Neurosurg. 1980;53:205–21.
15. Weiner HL, Finlay JL. Surgery in the management of primary intra-
4. Stein BM. Supracerebellar-infratentorial approach to pineal tumors.
cranial germ cell tumors. Childs Nerv Syst. 1999;15:770–3.
Surg Neurol. 1979;11:331–7.
16. Sung DI, Harisiadis L, Chang CH. Midline pineal tumors and
5. Knierim DS, Yamada S. Pineal tumors and associated lesions: the
suprasellar germinomas: highly curable by irradiation. Radiology.
effect of ethnicity on tumor type and treatment. Pediatr Neurosurg.
1978;128:745–51.
2003;38:307–23.
17. Dattoli MJ, Newall J. Radiation therapy for intracranial germi-
6. Oi S, Matsumoto S. Controversy pertaining to therapeutic modali-
noma: the case for limited volume treatment. Int J Radiat Oncol
ties for tumors of the pineal region: a worldwide survey of different
Biol Phys. 1990;19:429–33.
patient populations. Childs Nerv Syst. 1992;8:332–6.
7. Regis J, Bouillot P, Rouby-Volot F, et al. Pineal region tumors and 18. Balmaceda C, Heller G, Rosenblum M, et al. Chemotherapy with-
the role of stereotactic biopsy: review of the mortality, morbidity, out irradiation—a novel approach for newly diagnosed CNS germ
and diagnostic rates in 370 cases. Neurosurgery. 1996;39:907–12. cell tumors: results of an international cooperative trial. The First
8. Fukushima T. Endoscopic biopsy of intraventricular tumors with International Central Nervous System Germ Cell Tumor Study.
the use of a ventriculofiberscope. Neurosurgery. 1978;2:110–3. J Clin Oncol. 1996;14:2908–15.
9. Fukushima T, Ishijima B, Hirakawa K, et al. Ventriculofiberscope: 19. Buckner JC, Peethambaram PP, Smithson WA, et al. Phase II trial
a new technique for endoscopic diagnosis and operation. Technical of primary chemotherapy followed by reduced-dose radiation for
note J Neurosurg. 1973;38:251–6. CNS germ cell tumors. J Clin Oncol. 1999;17:933–40.
10. Robinson S, Cohen AR. The role of neuroendoscopy in the treat- 20. Backlund EO, Rahn T, Sarby B. Treatment of pinealomas by stereo-
ment of pineal region tumors. Surg Neurol. 1997;48:360–5. taxic radiation surgery. Acta Radiol Ther Phys Biol. 1974;13:368–76.
21. Subach BR, Lunsford LD, Kondziolka D. Stereotactic radiosur- 27. Mori Y, Kobayashi T, Hasegawa T, et al. Stereotactic radiosurgery
gery in the treatment of pineal region tumors. Prog Neurol Surg. for pineal and related tumors. Prog Neurol Surg. 2009;23:106–18.
1998;14:175–94. 28. Hasegawa T, Kondziolka D, Hadjipanayis CG, et al. Stereotactic
22. Hasegawa T, Kondziolka D, Hadjipanayis CG, et al. The role of radiosurgery for CNS nongerminomatous germ cell tumors. Report
radiosurgery for the treatment of pineal parenchymal tumors. of four cases. Pediatr Neurosurg. 2003;38:329–33.
Neurosurgery. 2002;51:880–9. 29. Lekovic GP, Gonzalez LF, Shetter AG, et al. Role of Gamma Knife
23. Manera L, Regis J, Chinot O, et al. Pineal region tumors: the role of surgery in the management of pineal region tumors. Neurosurg
stereotactic radiosurgery. Stereotact Funct Neurosurg. 1996;66 Focus. 2007;23:E12.
Suppl 1:164–73. 30. Wilson DA, Awad AW, Brachman D, et al. Long-term radiosurgical
24. Kano H, Niranjan A, Kondziolka D, et al. Role of stereotactic control of subtotally resected adult pineocytomas. J Neurosurg.
radiosurgery in the management of pineal parenchymal tumors. 2012;117:212–7.
Prog Neurol Surg. 2009;23:44–58. 31. Yianni J, Rowe J, Khandanpour N, et al. Stereotactic radiosurgery
25. Casentini L, Colombo F, Pozza F, et al. Combined radiosurgery and for pineal tumours. Br J Neurosurg. 2012;26:361–6.
external radiotherapy of intracranial germinomas. Surg Neurol. 32. Dempsey PK, Lunsford LD. Stereotactic radiosurgery for pineal
1990;34:79–86. region tumors. Neurosurg Clin N Am. 1992;3:245–53.
26. Kobayashi T, Kida Y, Mori Y. Stereotactic gamma radiosurgery for 33. Deshmukh VR, Smith KA, Rekate HL, et al. Diagnosis and man-
pineal and related tumors. J Neurooncol. 2001;54:301–9. agement of pineocytomas. Neurosurgery. 2004;55:349–55.
Pineal Region Tumors: Viewpoint—
Surgery 35
Adam M. Sonabend, Alfred Ogden, and Jeffrey N. Bruce
The management strategy for achieving control of tumor

Introduction growth is strongly dependent on histology. While benign
tumors are often cured by surgery, germinomas, on the
Since the seminal articles that described safe and effective other extreme, are exquisitely radiosensitive and best man-
approaches to the pineal region in the early 1970s [1, 2], aged with radiation therapy [22]. Most malignant pineal
numerous surgical series have established open surgery as a region tumors benefit from a combination of surgical resec-
critical tool in the treatment of pineal region lesions tion followed by chemotherapy and/or radiation. In rare
(Table 35.1) [3–15]. Although the difficulties and perils of instances of mixed tumors containing both benign and
pineal surgery were considered prohibitive during the malignant elements, surgical resection may be necessary to
nascent years of neurosurgery [16–18], the advent of the remove benign elements after prior chemotherapy/radiation
operating microscope, microneurosurgical techniques, mag- has eradicated the malignant portion. Open surgery also
netic resonance imaging (MRI), and the development of neu- facilitates comprehensive tissue sampling reducing the risk
rologic intensive care as a specialty have all helped to develop of sampling error given the rarity, diversity, and similar
pineal region surgery within a reasonable degree of safety radiographic appearance of different of pineal tumor his-
(Table 35.1). tologies (Fig. 35.1).
Open surgery is the treatment of choice for benign tumors. Stereotactic radiosurgery has an unproven role as an
Depending on specific histology, open surgery has a variable adjunct or even an alternative to open surgery. The great
role in the treatment of malignant tumors and tumors with variety of tumors that occur in the pineal region and the
mixed elements, either as a means of cytoreduction prior to need to direct adjuvant therapy based on histopathologic
chemotherapy and/or radiation [15, 19, 20] or as the best diagnosis mandate that, with few exceptions, treatment,
option to eradicate benign elements in mixed tumors after including stereotactic radiosurgery, should not be pursued
chemotherapy/radiation [7, 21]. Additionally, through imme- without a tissue diagnosis. Stereotactic biopsy followed by
diate removal of obstructing lesions, open surgery can obvi- stereotactic radiosurgery may prove to be a reasonable
ate the need for permanent cerebrospinal fluid (CSF) alternative to open surgery in selected cases, but this strat-
diversion. Open surgery also offers comprehensive tissue egy is currently most appropriate for patients with signifi-
sampling for a group of tumors that can be of mixed histology cant medical contraindications to open surgery. Stereotactic
and are notoriously difficult to diagnose. radiosurgery may also have an additional role as an alterna-
tive to whole-brain radiation in the treatment of select
radiosensitive tumors or in the adjuvant setting following
surgical resection. Retrospective series of patients with
A.M. Sonabend, M.D.
Department of Neurological Surgery, Bartoli Brain Tumor
pineal region tumors treated with stereotactic radiosurgery
Research Laboratory, Columbia University Medical Center, report variable outcomes [23, 24]. The interpretation of
New York Presbyterian Hospital, 710 W 168th Street, New York, treatment efficacy in these studies is complicated as the
NY 10032, USA patient populations include different or unknown histology,
A. Ogden, M.D. • J.N. Bruce, M.D. (*) and nonuniformity of additional treatments including crani-
Department of Neurological Surgery, Bartoli Brain Tumor otomy and radiation therapy. Predictors of poor outcome
Research Laboratory, Columbia University Medical Center,
New York Presbyterian Hospital, 710 W 168th Street, New York,
following stereotactic radiosurgery include radiographic
NY 10032, USA signs of tumor necrosis, prior radiation treatment and high
e-mail: jnb2@columbia.edu grade histology [23].
460 A.M. Sonabend et al.
Table 35.1 Results of large microsurgical series for pineal region tumors
Percent
of patients Major Permanent
No. of with GTR Mortality morbidity minor morbidity
Authors Year cases Approach Patient population Pathology (%) (%) (%) (%)
Hoffman et al. [8] 1983 61 TCIH/ITSC Pediatric All NA 20a NA NA
Neuwelt et al. [9] 1985 13 OTT Adult/pediatric All 60 0 0 20
Lapras et al. [10] 1987 86 TCIH/OTT Adult/pediatric All 65 5.8† 5.8† 28
Edwards et al. [11] 1988 36 TT/OTT/ITSC Pediatric All NA 0 3.3 3.3
Pluchino et al. [12] 1989 40 ITSC Adult/pediatric All 25 5 NA NA
Luo SQ et al. [13] 1989 64 OTT Adult/pediatric All 21 10 NA
Vaquero et al. [14] 1992 29 TCIH/ITSC/OTT Adult/pediatric All NA 11 NA NA
Herrmann et al. [87] 1992 49 IHTC/ITSC Adult/pediatric All NA 8 NA NA
Bruce and Stein [6] 1995 160 ITSC/TC/OTT Adult/pediatric All 45 4 3 19
Chandy et al. [88] 1998 48 ITSC/OTT Adult/pediatric “Benign 55 0 NA NA
lesions”
Kang et al. [89] 1998 16 OTT/ITSC/TCIH Adult/pediatric All 37.5 0 0 19
Shin et al. [90] 1998 21 OTT Pediatric/adult All 54.5 0 0 5
Konovalov et al. [3] 2003 201 OTT (54 %) Adult/pediatric All 58 10‡ NA >20
ITSC (34 %)
Bruce [91] 2004 81 ITSC/TCIH/OTT Adult/pediatric All 47 1 2 NA
Hernesniemi et al. [4] 2008 119 ITSC (93 %)/ Adult/pediatric All 88 0 1 4.9
OTT (7 %)
TCIH transcallosal interhemispheric, ITSC infratentorial supracerebellar, OTT occipital transtentorial, TT transcortical transventricular
a
All except one mortality prior to 1975
Surgical Management: An Overview
Initial management of a pineal region mass includes evalua-

tion for hydrocephalus, which is present in most patients.
Although ventriculoperitoneal shunting is acceptable, endo-
scopic third ventriculostomy is preferred, achieving the same
end without exposing a patient to the potential problems of
shunt malfunction, shunt infection, and abdominal seeding
of a malignancy. After CSF diversion, a procedure to obtain
tissue is indicated, either via a stereotactic, an endoscopic, or
an open approach. Either diverting procedure provides a con-
venient opportunity to assay CSF for malignant germ cell
tumor markers (α[alpha]-feto-protein or β[beta]-HCG) and
cytology, which can in very rare cases obviate the need for
tissue sampling. If these markers are elevated in either CSF
or serum, then by definition a malignant nongerminomatous
germ cell tumor (NGGCT) is present, and chemotherapy and
radiation therapy can proceed without the need for histologic
confirmation.
Stereotactic Biopsy
Although safe, stereotactic biopsy in the pineal region requires

more care and planning than biopsies of most other areas of
Fig. 35.1 Similar imaging characteristics of pineal lesions with differ- the brain. An anterolateral approach is favored with a pre-cor-
ent histologies. T1-weighted MRIs, precontrast (left panels) and post- onal entry point just behind the nexus of the superior temporal
contrast (right panels), of three patients with pineal region tumors: (a)
ependymoma, (b) pineocytoma and (c) germinoma line and the hairline (Fig. 35.2). Using image guidance, the
35 Pineal Region Tumors: Viewpoint—Surgery 461
Fig. 35.2 Eighteen year-old male presented with headache, nausea and tomy, a stereotaxic needle biopsy was obtained. Stereotaxic needle tra-
vomiting. MRI revealed a pineal lesion (a), with dissemination into the jectory with a pre-coronal and slightly lateral entry point was chosen to
third ventricle (a, red arrow). Following endoscopic third ventriculos- avoid the ventricular system, and eloquent cortical territory (b)
trajectory can be adjusted to enter a gyrus at the pial surface, Endoscopic Biopsy
avoid the lateral ventricle, and steer clear of any obvious vas-
culature that may be distorted into the path of the biopsy nee- Some authors have advocated obtaining tissue specimens
dle by the tumor itself [25]. endoscopically during the same operation as the third ven-
Close proximity of the target to the deep venous system triculostomy. Although a potentially elegant one-step solution
and a lack of adjacent tissue turgor should theoretically to CSF diversion and tissue diagnosis, such a strategy neces-
increase the likelihood of a significant biopsy-induced hem- sarily biases tissue sampling towards one quadrant of the
orrhage. In fact, such hemorrhages appear to be mildly more tumor capsule. Also, unless a flexible endoscope is used,
common in the pineal region, but they are rarely of any clini- endoscopic biopsy and third ventriculostomy must be per-
cal significance [26, 27]. Some series have suggested a formed through two separate burr holes and as such offers no
higher morbidity associated with pineal region biopsies less “invasiveness” than stereotactic biopsy and third ventric-
compared with routine biopsies, whereas others do not. ulostomy. Recent studies have demonstrated the feasibility of
In any case, overall complication rates are low and are both biopsy and third ventriculostomy through a single burr
almost always a function of transient exacerbations hole positioned midway between the two trajectories [33].
of existing symptoms [25, 28–31]. A rare, but nonethe-
less reported complication of stereotaxic biopsy is meta-
static seeding of a pineoblastoma along the biopsy Craniotomy for Open Resection
tract [32]. The major drawback to biopsies is the limited
tissue sampling that makes definitive diagnosis difficult Although diagnostic biopsies are a suitable first option in
when dealing with these widely diverse and often some cases, an open approach is generally preferred. The
heterogeneous lesions. advantages of an open resection are numerous, including
Fig. 35.3 Fifty two year-old female presented with nausea and vomit- (a). Following endoscopic third ventriculostomy, the patient underwent
ing over the course of few weeks. MRI revealed a pineal region a craniotomy through occipital transtentorial approach (trajectory
lesion with no involvement of the third ventricle but with inferior shown with red arrow in a), and postoperative imaging showed gross
extension dorsal to the brainstem, and obstructive hydrocephalus total resection (b)
generous tissue sampling, potential obviation of a shunt, and laterally away from the natural reach achieved with a midline
the ability to proceed with a radical resection if indicated trajectory from the infratentorial supracerebellar (ITSC)
after intraoperative pathology consultation [34]. The relative approach. The supratentorial approaches include the occipi-
advantages of pursuing radical resection depend on an accu- tal transtentorial (OTT) (Fig. 35.3), the transcallosal inter-
rate histopathologic diagnosis and must be weighed against hemispheric (TCIH), and the rarely used
rates of morbidity and mortality from radical surgery. transcortical-transventricular (TT) approach which is limited
Complications from open surgery can be serious and even to the setting of ventricular dilatation and tumor extension
catastrophic typically resulting from postoperative hema- into the lateral ventricle [10, 34, 35, 38, 39]. Some of the
toma, venous infarction from vein sacrifice, thalamic injury surgical morbidity can be related to deficits specific to a
from dissection of the anterior tumor capsule, and visual given approach. One of the major inconveniences of the
deficits if the occipital transtentorial approach is used. supratentorial approaches is that important deep venous
However, when complete tumor removal can be achieved, structures such as the vein of Galen and the internal cerebral
the risk of postoperative hemorrhage into a subtotally veins lay between the surgeon and the lesion. The presence
resected tumor is reduced. Minor complications include of these veins in the field increases the risk of venous infarct,
infection and exacerbation of existing ataxia or Parinaud and might obstruct the operative field limiting.
phenomena. Pineal region surgery is certainly not trivial, but Infratentorial access to the pineal region is obtained
within the past 25 years, major morbidity and mortality from though the ITSC corridor (Fig. 35.4). With the patient prefer-
published surgical series have improved dramatically and ably in the sitting position, the superior surface of the cere-
over the past quarter century has dropped to 0–2 % [35] bellum is dissected from the tentorium and the precentral
(Table 35.1). cerebellar vein is divided, allowing the cerebellum to fall by
A craniotomy is performed when cytoreductive radical gravity clearing an unobstructed pathway to the pineal region
resection is the goal, and as mentioned above, this surgery [35, 40]. With the sitting position, gravity minimizes retrac-
provides significantly more reliable material for histological tion, and prevents blood from pooling in the microsurgical
diagnosis. There are a series of approaches to access the field. Additionally, the deep venous structures lie superior to
pineal region, and the selection of a particular one depends the tumor mass, facilitating resection of the lesion. This
on factors such as the surgeon’s experience, and anatomical approach is favorable for removing tumor extending into the
factors. In general, the approaches can be divided into supra- third ventricle and velum interpositum. Disadvantages of the
tentorial and infratentorial [35–37]. approach include limited access to lateral and supratentorial
Supratentorial approaches provide a wider lateral expo- areas. Also, the sitting position carries a risk of air embolism,
sure, and are preferred when the tumor mass extends signifi- pneumocephalus and subdural hematoma, although proper
cantly into the supratentorial compartment, or has grown precautions can minimize the risk [35, 41].
Fig. 35.4 Twenty five year-old female presented with Parinaud’s syn- with an infratentorial supracerebellar approach was performed for
drome and headache. MRI revealed a pineal region lesion with exten- tumor resection (trajectory shown with red arrow in a), and postopera-
sion into the posterior portion of the third ventricle (a). A craniotomy tive imaging showed gross total resection (b)
Treatment Results by Tumor Histology pineocytomas, and rare pilocytic astrocytomas. In each case,
radical resection is the standard of care if it can be performed
The field of oncology is predicated on the belief that any within a reasonable degree of safety. Gross total resection
treatment modality must be evaluated on the basis of efficacy provides the best chance for a cure or extended remission but
in treating groups of patients bound by a common diagnosis. this goal must be weighed against the risks associated with
This diagnosis is almost always made on the basis of tissue radical resection. Surgical series demonstrate good outcomes
histology. Pineal region tumors can arise from any of the myr- over a range of pathologies, although there is a need in the
iad cell types that normally occur in and around the pineal literature for surgical outcomes to be analyzed according to
gland, as well as from the range of ectopic tissues that can be specific histology. Only a few small surgical series are dedi-
trapped near the pineal gland during embryogenesis. Because cated to benign pineal lesions of a single histological type
few places in the body can match such cellular diversity and [42–44].
biopsies of pineal lesions have been demonstrated to be safe,
histologic diagnosis of pineal region tumors is mandatory
unless tumor markers are positive. MRI is excellent at defin- Glial Tumors
ing the relationship between tumor and adjacent anatomic
structures but is not able to distinguish between tumors of Glial neoplasms encountered in the pineal region can present
different histologies (Fig. 35.1). in three different categories: (a) brainstem astrocytomas,
Tumors in the pineal region can be categorized as benign usually tectal lesions that extend rostrally, (b) “true” pineal
or malignant. Alternatively, they can be classified into one of astrocytomas that arise from the supporting astrocytes of the
four broad histologic categories: (1) pineal cell; (2) germ pineal gland itself, and (c) ependymomas associated with
cell; (3) glial cell; and (4) miscellaneous. Although surgical the third ventricle. Brainstem astrocytomas that grow into the
results significantly influence the outcome for these patients, pineal region are solid and, although they may be “low
individual prognosis is heavily dependent upon tumor type. grade,” are generally invasive. These are not amenable to
The rarity of individual histopathologic types within the aggressive resection and require irradiation after biopsy [36,
pineal region makes it difficult to generate sufficient num- 45]. The “true” pineal astrocytomas are often cystic, encap-
bers of patients treated to evaluate individual therapies, how- sulated, and resemble pilocytic astrocytomas on histology
ever some generalizations can be made. [46]. These can be completely resected and have excellent
long-term results. Anaplastic gliomas and glioblastomas
have also been encountered in the pineal region [3, 4].
Benign Pineal Region Tumors Surgical outcomes from ependymomas are binary, depend-
ing on the degree of anaplasia. Pineal region ependymomas
Benign tumors account one third of the masses found with low cellularity and few mitoses have excellent long-
in pineal region [6]. These are composed mainly of term outcomes although they may recur more readily than
well-differentiated ependymomas, meningiomas, teratomas, ependymomas associated with the lateral ventricles [47].
Pineal Parenchymal Tumors showing local tumor control of 30 % at 2 years from diagno-
sis [24, 54–57], likely reflecting their malignant nature and
Arising from the melatonin-producing cells of the pineal potential to spread through CSF. The relative utility of radio-
gland, pineal parenchymal tumors exist along a histopatho- surgery cannot be ascertained based on a few published cases
logic continuum from benign and indolent pineocytomas to except to say that radiosurgery does not appear to be a “magic
malignant and aggressive pineoblastomas. Tumors of inter- bullet” for these lesions, and it does not address the meta-
mediate grade are referred to as mixed pineal parenchymal static potential of these tumors. Ultimately, prognosis for
tumors, and various classification schemes that are tied to pineoblastoma is most predicted by age of presentation and
prognosis have been proposed [42, 48–52]. Making an accu- disease dissemination at the time of diagnosis. Although far
rate pathologic diagnosis is difficult even with generous tis- from comprehensive, evidence suggests that resection has a
sue sampling. The rarity of these tumors makes the beneficial impact on survival.
aggregation of sufficient numbers of patients for an instruc- Intermediate-grade pineal parenchymal tumors behave in
tive clinical trial difficult. Further complicating matters is the an unpredictable fashion. Treatment successes and failures
variable behavior of tumors with similar histopathologies have been reported from a range of treatment strategies, and
depending upon age of presentation. there is little helpful data to direct clinical decisions [19, 51,
Amidst the paucity of data to direct clinical decisions, 52]. Although patients with pineoblastoma are routinely irra-
certain general treatment guidelines are apparent. The goal diated after surgery, it is unclear whether this is universally
for well-differentiated pineocytomas should be a cure, at required in cases of intermediate-grade pineal parenchymal
least in adults. There is some evidence that this standard is tumors. Recently, some immunohistochemical markers have
unrealistic in pediatric cases where pineocytomas may been shown to be of prognostic value [68]. Clinical studies
behave more aggressively [53]. that tie adjuvant therapy to tumor grade and outcome are
The reported low recurrence rates with gross total resec- required with long-term outcome as tumors can recur more
tion of pineocytomas (overall survival approximately than 5 years from diagnosis [19].
100 % at 40 months median follow-up) has established a On the whole, pineal parenchymal tumors present a for-
new standard of care for these low grade tumors [3, 4, 42]. midable challenge to clinicians. Treatment optimization will
Outcomes following radiosurgical treatment of pineocy- only be possible with systematic and expert pathologic diag-
toma have been favorable as well [54–56], although the few nosis that is not restricted by availability of tissue. Ideally,
reported adult cases had only had a short follow-up [56] this tissue should be sought via open surgery as radical resec-
and included at least one adult treatment failure resulting in tion loosely correlates with long-term survival.
death from CNS metastasis [55]. One pediatric pineocy-
toma radiosurgical series reported four of seven treatment
failures resulting in death during a follow-up of 3 months to Papillary Tumor of the Pineal Region
4 years [57], although recently, more favorable results have
been reported [23]. Papillary tumor of the pineal region (PTPR) was first
At the other end of the neoplastic spectrum are pineoblas- described in 2003 [69], and introduced as a distinct tumor
tomas. Although pineoblastomas often appear identical to entity in the 2007 WHO classification of brain tumors. The
pineocytomas on MRI, they are histologically indistinguish- cells of origin for these lesions are presumed to be special-
able from primitive neuroectodermal tumors (PNETs) and ized ependyma of the subcommissural organ [69–71]. Due to
behave in a similar clinical fashion. Like PNETs, they tend to their recent identification as a separate entity, and their rarity,
be more aggressive in children than in adults [58], and within not much is known about their clinical behavior. Fevre-
the pediatric population they are increasingly aggressive Montange et al.’s retrospective study of 31 patients is the
with decreasing age of presentation [59–66]. Although some largest published series of these tumors and reported an age
indication of increased survival exists in both adults [19] and range from 5 to 66 years (median age, 29 years) and slight
children [20] undergoing open surgery, no statistically rigor- female predominance [15]. Gross total resection was per-
ous study has examined the impact of extent of resection formed on 21 or 31 patients, and in 15, adjuvant radiotherapy
on survival or disease progression. In the absence of such was given. Despite aggressive treatment, the majority of
studies, it seems logical to follow the standard of care for patients had recurrences. Overall, the 5-year survival was
medulloblastoma (pinealoblastoma, shares phonotypical 73 % with progression free survival of 27 % [15]. Details on
characteristics of PNET) for which significant survival ben- their histological and immunohistochemical profile, as well
efits are apparent after reduction of tumor mass under 1.5 cm3 as reliable methods for distinguishing between papillary
[67]. The few published cases of stereotactic radiosurgery tumors, choroid plexus papillomas, and metastatic carcino-
for pineoblastoma report poor results with the most recent mas have been reported [15, 69, 70, 72, 73]
Germ Cell Tumors stereotaxic radiotherapy for germinomas has been performed,
with local tumor control rates at 3 and 5 years of 82 %, and
Germ cell tumors are considered and studied in two separate 72 %, respectively although 62 % germinomas with STGCs
groups, germinomas and NGGCTs. The latter group consists had 5-year control rate of 62 % [24]. Some authors propose
of endodermal sinus tumors, choriocarcinomas, embryonal the use of stereotactic radiosurgery as an effective adjunct for
carcinomas, mature teratomas, and immature teratomas. The pure germinomas to reduce cumulative radiation doses [54].
role of open surgery for these lesions is perhaps as well This hypothesis could be ethically tested in the pediatric
defined as for any pineal region tumors. population.
Germinomas are the most common of pineal region NGGCTs have historically had a much worse prognosis
tumors, especially in adolescent boys and young men. In than germinomas. Because the individual types are so rare and
Japan and Korea, for reasons that are not understood, germ they are frequently of mixed histology, they have been lumped
cell tumors in general and germinomas in particular are more together in retrospective analyses and clinical trials. These are
prevalent than in any other part of the world including neigh- the only pineal region tumors that should be treated without a
boring China [62]. Because they do not secrete a specific tissue diagnosis as elevated levels of markers in the serum and/
tumor marker and they cannot be distinguished radiographi- or CSF are pathognomonic for specific histopathologies.
cally from other types of tumors that call for different treat- Although open surgery in a pre-adjuvant therapy cytoreductive
ment paradigms, diagnosis should be made with tissue role has been examined with variable results [81, 83, 84], the
confirmation regardless of age of presentation. best results seem to occur when radiation and/or chemotherapy
Cytoreduction from open surgery has not been shown to is followed by “second look” surgery when a persistent radio-
improve the excellent outcomes with radiation alone [63] and graphic lesion exists [7, 81, 82, 85]. Using this approach, only
thus the overwhelming majority of germinomas are diagnosed residual teratomatoma elements or scar tissue has been found,
by open or stereotactic biopsy and treated with whole-brain and 5-year survival rates have improved dramatically to >90 %.
radiation. Historically, the relatively high prevalence of ger- Stereotactic radiosurgery has an undefined role for
minomas, their exquisite radiosensitivity, and the perceived NGGCTs. One report [86] of four patients who received
dangers of pineal region biopsy all contributed to the practice radiosurgery along with fractionated radiation and chemo-
of up-front radiation without tissue diagnosis for all pineal therapy showed tumor regression in three patients after a
region tumors. This approach was still commonly pursued in follow-up of 2 years.
the Far East as recently as 1992 [64, 65]. Although it is diffi-
cult to take issue with cure rates that today are over 90 % [66,
74–77], improvements in the treatment of intracranial germi- Conclusion
noma can still be made. A small percentage of patients fail
radiation and suffer CSF dissemination that is ultimately fatal. Despite sharing a common anatomic location and similar
Treatment failure is much more likely when syncytiotropho- imaging characteristics, pineal region tumors are extremely
blastic giant cells (STGCs) are mixed within the usual histo- heterogeneous with respect to histopathology, natural his-
logic features, and failure rates within this histologic subtype tory, and response to therapy. Except for cases of marker-
have been reported as high as 40 % [78–80]. Because these secreting germ cell tumors, a tissue diagnosis is required to
are the rare germinomas that secrete β[beta]-hCG to produce direct therapy. In the absence of elevated markers, in most
mildly elevated serum levels, their diagnosis should be cases, open surgery is the best initial therapeutic option.
straightforward with adequate biopsy tissue and standard Benign lesions can be cured with gross total resection, and in
assays of serum markers. As such, they should be considered experienced hands gross total resection can be achieved with
in a separate category from “pure germinomas” and the sub- low complication rates. Stereotactic biopsy followed by ste-
ject of separate clinical trials to reexamine all treatment reotactic radiosurgery is a potential alternative that is most
modalities, including radiosurgery and surgical cytoreduc- appropriate in patients with tumor seeding or medical con-
tion. A study of stereotactic radiosurgery showed 72 and 62 % traindications to open surgery. The most effective treatments
local tumor control at 3 and 5 years from diagnosis [24]. for malignant lesions will likely require a combination of
Because many patients with intracranial germinoma have modalities that will be directed by histopathologic diagnosis.
long-term remission, the extended sequelae of whole-brain The merit of this approach is apparent in the example of
radiation, particularly in the pediatric population, are another NGGCTs whose prognosis has vastly improved after numer-
major concern. Strides in chemotherapy have been made in ous clinical trials that were predicated on cohorts of patients
trials composed of children too young to receive full doses of with specific diagnoses. Further progress will depend on
whole-brain radiation [81, 82]. Radiosurgery is a theoreti- consistent histologic confirmation of these rare tumors rather
cally attractive alternative because of relatively limited radia- than empiric up-front conventional radiation or stereotactic
tion exposure to adjacent brain tissue. Over the last decade, radiosurgery as some authors have proposed [57].
20. Reddy AT, Janss AJ, Phillips PC, Weiss HL, Packer RJ. Outcome for
References children with supratentorial primitive neuroectodermal tumors treated
with surgery, radiation, and chemotherapy. Cancer. 2000;88(9):
2189–93. PubMed PMID: 10813733. Epub 2000/05/17. eng.
1. Stein BM. The infratentorial supracerebellar approach to pineal 21. Weiner HL, Finlay JL. Surgery in the management of primary intra-
lesions. J Neurosurg. 1971;35(2):197–202. PubMed PMID: cranial germ cell tumors. Childs Nerv Syst. 1999;15(11–12):770–3.
5315191. Epub 1971/08/01. eng. PubMed PMID: 10603021. Epub 1999/12/22. eng.
2. Jamieson KG. Excision of pineal tumors. J Neurosurg. 22. Dattoli MJ, Newall J. Radiation therapy for intracranial germi-
1971;35(5):550–3. PubMed PMID: 5120004. Epub 1971/11/01. eng. noma: the case for limited volume treatment. Int J Radiat Oncol
3. Konovalov AN, Pitskhelauri DI. Principles of treatment of the Biol Phys. 1990;19(2):429–33. PubMed PMID: 2394620. Epub
pineal region tumors. Surg Neurol. 2003;59(4):250–68. PubMed 1990/08/01. eng.
PMID: 12748006. Epub 2003/05/16. eng. 23. Yianni J, Rowe J, Khandanpour N, Nagy G, Hoggard N, Radatz M,
4. Hernesniemi J, Romani R, Albayrak BS, Lehto H, Dashti R, Ramsey et al. Stereotactic radiosurgery for pineal tumours. Br J Neurosurg.
3rd C, et al. Microsurgical management of pineal region lesions: per- 2012;26(3):361–6. PubMed PMID: 22168969. Epub 2011/12/16. eng.
sonal experience with 119 patients. Surg Neurol. 2008;70(6):576–83. 24. Mori Y, Kobayashi T, Hasegawa T, Yoshida K, Kida Y. Stereotactic
PubMed PMID: 19055952. Epub 2008/12/06. eng. radiosurgery for pineal and related tumors. Prog Neurol Surg.
5. Stern WE, Batzdorf U, Rich JR. Challenges of surgical excision of 2009;23:106–18. PubMed PMID: 19329865. Epub 2009/03/31. eng.
tumors in the pineal region. Bull Los Angeles Neurol Soc. 25. Zacharia BE, Bruce JN. Stereotactic biopsy considerations for
1971;36(4):105–18. PubMed PMID: 5132521. Epub 1971/10/01. eng. pineal tumors. Neurosurg Clin N Am. 2011;22(3):359–66, viii.
6. Bruce JN, Stein BM. Surgical management of pineal region tumors. PubMed PMID: 21801984.
Acta Neurochir (Wien). 1995;134(3–4):130–5. PubMed PMID: 26. Sawin PD, Hitchon PW, Follett KA, Torner JC. Computed imaging-
8748771. Epub 1995/01/01. eng. assisted stereotactic brain biopsy: a risk analysis of 225 consecutive
7. Weiner HL, Lichtenbaum RA, Wisoff JH, Snow RB, Souweidane cases. Surg Neurol. 1998;49(6):640–9. PubMed PMID: 9637625.
MM, Bruce JN, et al. Delayed surgical resection of central nervous Epub 1998/06/24. eng.
system germ cell tumors. Neurosurgery. 2002;50(4):727–33. 27. Field M, Witham TF, Flickinger JC, Kondziolka D, Lunsford
PubMed PMID: 11904022. discussion 33–4; Epub 2002/03/21. eng. LD. Comprehensive assessment of hemorrhage risks and outcomes
8. Hoffman HJ, Yoshida M, Becker LE, Hendrick EB, Humphreys after stereotactic brain biopsy. J Neurosurg. 2001;94(4):545–51.
RP. Pineal region tumors in childhood. Experience at the Hospital PubMed PMID: 11302651. Epub 2001/04/17. eng.
for Sick Children. 1983. Pediatr Neurosurg. 1994;21(1):91–103. 28. Kreth FW, Schatz CR, Pagenstecher A, Faist M, Volk B, Ostertag
9. Neuwelt EA. An update on the surgical treatment of malignant CB. Stereotactic management of lesions of the pineal region.
pineal region tumors. Clin Neurosurg. 1985;32:397–428. PubMed Neurosurgery. 1996;39(2):280–9. PubMed PMID: 8832665. dis-
PMID: 2415282. Epub 1985/01/01. eng. cussion 9–91; Epub 1996/08/01. eng.
10. Lapras C, Patet JD, Mottolese C, Lapras Jr C. Direct surgery for pineal 29. Regis J, Bouillot P, Rouby-Volot F, Figarella-Branger D, Dufour H,
tumors: occipital-transtentorial approach. Prog Exp Tumor Res. Peragut JC. Pineal region tumors and the role of stereotactic biopsy:
1987;30:268–80. PubMed PMID: 3628812. Epub 1987/01/01. eng. review of the mortality, morbidity, and diagnostic rates in 370
11. Edwards MS, Hudgins RJ, Wilson CB, Levin VA, Wara WM. Pineal cases. Neurosurgery. 1996;39(5):907–12. PubMed PMID: 8905744.
region tumors in children. J Neurosurg. 1988;68(5):689–97. discussion 12–4; Epub 1996/11/01. eng.
PubMed PMID: 2451717. Epub 1988/05/01. eng. 30. Dempsey PK, Kondziolka D, Lunsford LD. Stereotactic diagnosis
12. Pluchino F, Broggi G, Fornari M, Franzini A, Solero CL, Allegranza and treatment of pineal region tumours and vascular malforma-
A. Surgical approach to pineal tumours. Acta Neurochir (Wien). tions. Acta Neurochir (Wien). 1992;116(1):14–22. PubMed PMID:
1989;96(1–2):26–31. PubMed PMID: 2929390. Epub 1989/01/01. eng. 1615765. Epub 1992/01/01. eng.
13. Luo SQ, Li DZ, Zhang MZ, Wang ZC. Occipital transtentorial 31. Apuzzo ML, Chandrasoma PT, Cohen D, Zee CS, Zelman
approach for removal of pineal region tumors: report of 64 consecu- V. Computed imaging stereotaxy: experience and perspective related
tive cases. Surg Neurol. 1989;32(1):36–9. PubMed PMID: to 500 procedures applied to brain masses. Neurosurgery.
2734686. Epub 1989/07/01. eng. 1987;20(6):930–7. PubMed PMID: 3302751. Epub 1987/06/01. eng.
14. Vaquero J, Ramiro J, Martinez R, Bravo G. Neurosurgical experi- 32. Rosenfeld JV, Murphy MA, Chow CW. Implantation metastasis of
ence with tumours of the pineal region at Clinica Puerta de Hierro. pineoblastoma after stereotactic biopsy. J Neurosurg. 1990;73:
Acta Neurochir (Wien). 1992;116(1):23–32. PubMed PMID: 287–90.
1319669. Epub 1992/01/01. eng. 33. Robinson S, Cohen AR. The role of neuroendoscopy in the treat-
15. Fevre-Montange M, Hasselblatt M, Figarella-Branger D, Chauveinc ment of pineal region tumors. Surg Neurol. 1997;48(4):360–5.
L, Champier J, Saint-Pierre G, et al. Prognosis and histopathologic PubMed PMID: 9315134. discussion 5–7; Epub 1997/10/07. eng.
features in papillary tumors of the pineal region: a retrospective mul- 34. Kennedy BC, Bruce JN. Surgical approaches to the pineal region.
ticenter study of 31 cases. J Neuropathol Exp Neurol. 2006; Neurosurg Clin N Am. 2011;22(3):367–80, viii. PubMed PMID:
65(10):1004–11. PubMed PMID: 17021405. Epub 2006/10/06. eng. 21801985.
16. Dandy W. Operative experience in cases of pineal tumor. Arch 35. Bruce JN, Ogden AT. Surgical strategies for treating patients with
Surg. 1936;33:19–46. pineal region tumors. J Neurooncol. 2004;69(1–3):221–36.
17. Horrax G. Treatment of tumors of the pineal body. Experience in a PubMed PMID: 15527093.
series of twenty-two cases. Arch Neurol Psychiatry. 1950;64: 36. Stein BM, Bruce JN. Surgical management of pineal region tumors
227–42. (honored guest lecture). Clin Neurosurg. 1992;39:509–32. PubMed
18. Wagenen V. A surgical approach for the removal of certain pineal PMID: 1458755. Epub 1992/01/01. eng.
tumors. Surg Gynecol Obstet. 1931;37:216–20. 37. Reid WS, Clark WK. Comparison of the infratentorial and transten-
19. Lutterbach J, Fauchon F, Schild SE, Chang SM, Pagenstecher A, torial approaches to the pineal region. Neurosurgery. 1978;3(1):1–
Volk B, et al. Malignant pineal parenchymal tumors in adult 8. PubMed PMID: 683486.
patients: patterns of care and prognostic factors. Neurosurgery. 38. Bruce JN. Posterior third ventricular tumors. In: Kaye AH, Black
2002;51(1):44–55; discussion -6. PubMed PMID: 12182434. Epub PM, editors. Operative neurosurgery, vol. 1. London: Churchill
2002/08/17. eng. Livingstone; 2000. p. 769–75.
39. Van Wagenen WP. A surgical approach for the removal of certain 57. Raco A, Raimondi AJ, D’Alonzo A, Esposito V, Valentino
pineal tumors. Surg Gynecol Obstet. 1931;53:216–20. V. Radiosurgery in the management of pediatric brain tumors.
40. Bruce JN, Stein BM. Supracerebellar approach to pineal region Childs Nerv Syst. 2000;16(5):287–95. PubMed PMID: 10883372.
neoplasms. In: Schmidek HH, editor. Operative Neurosurgical Epub 2000/07/07. eng.
Techniques. 1. 4th ed. Philadelphia: W.B. Saunders Company; 58. Chang SM, Lillis-Hearne PK, Larson DA, Wara WM, Bollen AW,
2000. p. 908–15. Prados MD. Pineoblastoma in adults. Neurosurgery.
41. Bruce JN, Stein BM. Complications of surgery for pineal region 1995;37(3):383–90. PubMed PMID: 7501100. discussion 90–1;
tumors. In: Post KD, Friedman ED, McCormick PC, editors. Epub 1995/09/01. eng.
Postoperative Complications in Intracranial Neurosurgery. 59. Duffner PK, Cohen ME, Sanford RA, Horowitz ME, Krischer JP,
New York: Thieme Medical Publishers; 1993. p. 74–86. Burger PC, et al. Lack of efficacy of postoperative chemotherapy
42. Vaquero J, Ramiro J, Martinez R, Coca S, Bravo and delayed radiation in very young children with pineoblastoma.
G. Clinicopathological experience with pineocytomas: report of Pediatric Oncology Group. Med Pediatr Oncol. 1995;25(1):38–44.
five surgically treated cases. Neurosurgery. 1990;27(4):612–8. PubMed PMID: 7753001. Epub 1995/07/01. eng.
PubMed PMID: 2234367. discussion 8–9; Epub 1990/10/01. eng. 60. Abay 2nd EO, Laws Jr ER, Grado GL, Bruckman JE, Forbes GS,
43. Konovalov AN, Spallone A, Pitzkhelauri DI. Meningioma of the Gomez MR, et al. Pineal tumors in children and adolescents. Treatment
pineal region: a surgical series of 10 cases. J Neurosurg. by CSF shunting and radiotherapy. J Neurosurg. 1981;55(6):889–95.
1996;85(4):586–90. PubMed PMID: 8814160. Epub 1996/10/01. eng. PubMed PMID: 7299463. Epub 1981/12/01. eng.
44. Fevre Montange M, Vasiljevic A, Bergemer Fouquet AM, Bernier 61. Jakacki RI, Zeltzer PM, Boyett JM, Albright AL, Allen JC, Geyer
M, Champier J, Chretien F, et al. Histopathologic and ultrastructural JR, et al. Survival and prognostic factors following radiation and/or
features and claudin expression in papillary tumors of the pineal chemotherapy for primitive neuroectodermal tumors of the pineal
region: a multicenter analysis. Am J Surg Pathol. 2012;36(6): region in infants and children: a report of the Childrens Cancer
916–28. PubMed PMID: 22588068. Epub 2012/05/17. eng. Group. J Clin Oncol. 1995;13(6):1377–83. PubMed PMID:
45. Barnett DW, Olson JJ, Thomas WG, Hunter SB. Low-grade astrocy- 7751882. Epub 1995/06/01. eng.
tomas arising from the pineal gland. Surg Neurol. 1995;43(1):70–5. 62. Oi S, Matsuzawa K, Choi JU, Kim DS, Kang JK, Cho BK. Identical
PubMed PMID: 7701429. discussion 5–6; Epub 1995/01/01. eng. characteristics of the patient populations with pineal region tumors
46. DeGirolami U, Armbrustmacher VW. Juvenile pilocytic astrocy- in Japan and in Korea and therapeutic modalities. Childs Nerv Syst.
toma of the pineal region: report of a case. Cancer. 1982;50(6):1185– 1998;14(1–2):36–40. PubMed PMID: 9548339. Epub 1998/04/21.
8. PubMed PMID: 7104961. Epub 1982/09/15. eng. eng.
47. Schwartz TH, Kim S, Glick RS, Bagiella E, Balmaceda C, Fetell 63. Sawamura Y, de Tribolet N, Ishii N, Abe H. Management of pri-
MR, et al. Supratentorial ependymomas in adult patients. mary intracranial germinomas: diagnostic surgery or radical resec-
Neurosurgery. 1999;44(4):721–31. PubMed PMID: 10201296. tion? J Neurosurg. 1997;87(2):262–6. PubMed PMID: 9254091.
Epub 1999/04/14. eng. Epub 1997/08/01. eng.
48. Borit A, Blackwood W, Mair WG. The separation of pineocytoma 64. Oi S. Recent advances and racial differences in therapeutic strategy
from pineoblastoma. Cancer. 1980;45(6):1408–18. PubMed PMID: to the pineal region tumor. Childs Nerv Syst. 1998;14(1–2):33–5.
6986979. Epub 1980/03/15. eng. PubMed PMID: 9548338. Epub 1998/04/21. eng.
49. Cho BK, Wang KC, Nam DH, Kim DG, Jung HW, Kim HJ, et al. 65. Oi S, Matsumoto S. Controversy pertaining to therapeutic modali-
Pineal tumors: experience with 48 cases over 10 years. Childs Nerv ties for tumors of the pineal region: a worldwide survey of different
Syst. 1998;14(1–2):53–8. PubMed PMID: 9548342. Epub patient populations. Childs Nerv Syst. 1992;8(6):332–6. PubMed
1998/04/21. eng. PMID: 1394280. Epub 1992/09/01. eng.
50. Coca S, Vaquero J, Escandon J, Moreno M, Peralba J, Rodriguez 66. Jenkin RD, Simpson WJ, Keen CW. Pineal and suprasellar germi-
J. Immunohistochemical characterization of pineocytomas. Clin nomas. Results of radiation treatment. J Neurosurg. 1978;48(1):99–
Neuropathol. 1992;11(6):298–303. PubMed PMID: 1473313. Epub 107. PubMed PMID: 619029. Epub 1978/01/01. eng.
1992/11/01. eng. 67. Zeltzer PM, Boyett JM, Finlay JL, Albright AL, Rorke LB, Milstein
51. Fauchon F, Jouvet A, Paquis P, Saint-Pierre G, Mottolese C, Ben JM, et al. Metastasis stage, adjuvant treatment, and residual tumor
Hassel M, et al. Parenchymal pineal tumors: a clinicopathological are prognostic factors for medulloblastoma in children: conclusions
study of 76 cases. Int J Radiat Oncol Biol Phys. 2000;46(4):959– from the Children’s Cancer Group 921 randomized phase III study.
68. PubMed PMID: 10705018. Epub 2000/03/08. eng. J Clin Oncol. 1999;17(3):832–45. PubMed PMID: 10071274. Epub
52. Jouvet A, Saint-Pierre G, Fauchon F, Privat K, Bouffet E, Ruchoux 1999/03/10. eng.
MM, et al. Pineal parenchymal tumors: a correlation of histological 68. Arivazhagan A, Anandh B, Santosh V, Chandramouli BA. Pineal
features with prognosis in 66 cases. Brain Pathol. 2000;10(1):49– parenchymal tumors–utility of immunohistochemical markers in
60. PubMed PMID: 10668895. Epub 2000/02/11. eng. prognostication. Clin Neuropathol. 2008;27(5):325–33. PubMed
53. D’Andrea AD, Packer RJ, Rorke LB, Bilaniuk LT, Sutton LN, PMID: 18808064. Epub 2008/09/24. eng.
Bruce DA, et al. Pineocytomas of childhood. A reappraisal of natu- 69. Jouvet A, Fauchon F, Liberski P, Saint-Pierre G, Didier-Bazes M,
ral history and response to therapy. Cancer. 1987;59(7):1353–7. Heitzmann A, et al. Papillary tumor of the pineal region. Am J Surg
PubMed PMID: 3815306. Epub 1987/04/01. eng. Pathol. 2003;27(4):505–12. PubMed PMID: 12657936.
54. Kobayashi T, Kida Y, Mori Y. Stereotactic gamma radiosurgery for 70. Hasselblatt M, Blumcke I, Jeibmann A, Rickert CH, Jouvet A, van
pineal and related tumors. J Neurooncol. 2001;54(3):301–9. de Nes JA, et al. Immunohistochemical profile and chromosomal
PubMed PMID: 11767295. Epub 2002/01/05. eng. imbalances in papillary tumours of the pineal region. Neuropathol
55. Hasegawa T, Kondziolka D, Hadjipanayis CG, Flickinger JC, Appl Neurobiol. 2006;32(3):278–83. PubMed PMID: 16640646.
Lunsford LD. The role of radiosurgery for the treatment of pineal 71. Shibahara J, Todo T, Morita A, Mori H, Aoki S, Fukayama
parenchymal tumors. Neurosurgery. 2002;51(4):880–9. PubMed M. Papillary neuroepithelial tumor of the pineal region. A case
PMID: 12234394. Epub 2002/09/18. eng. report. Acta Neuropathol. 2004;108(4):337–40. PubMed PMID:
56. Manera L, Regis J, Chinot O, Porcheron D, Levrier O, Farnarier P, 15221340.
et al. Pineal region tumors: the role of stereotactic radiosurgery. 72. Kasper M, Terpe HJ, Perry G. Age-dependent pattern of intermedi-
Stereotact Funct Neurosurg. 1996;66 Suppl 1:164–73. PubMed ate filament protein expression in the human pineal gland.
PMID: 9032858. Epub 1996/01/01. eng. J Hirnforsch. 1990;31(2):215–21. PubMed PMID: 1694197.
73. Montecinos HA, Richter H, Caprile T, Rodriguez EM. Synthesis of germ cell tumors by administering neoadjuvant chemotherapy and
transthyretin by the ependymal cells of the subcommissural organ. radiotherapy before excision of residual tumors. J Neurosurg.
Cell Tissue Res. 2005;320(3):487–99. PubMed PMID: 15846516. 2003;99(1):106–14. PubMed PMID: 12854751. Epub 2003/07/12. eng.
74. Rich TA, Cassady JR, Strand RD, Winston KR. Radiation therapy 83. Matsutani M, Sano K, Takakura K, Fujimaki T, Nakamura O, Funata
for pineal and suprasellar germ cell tumors. Cancer. 1985;55(5):932– N, et al. Primary intracranial germ cell tumors: a clinical analysis of
40. PubMed PMID: 2981605. Epub 1985/03/01. eng. 153 histologically verified cases. J Neurosurg. 1997;86(3):446–55.
75. Shirato H, Nishio M, Sawamura Y, Myohjin M, Kitahara T, PubMed PMID: 9046301. Epub 1997/03/01. eng.
Nishioka T, et al. Analysis of long-term treatment of intracranial 84. Robertson PL, DaRosso RC, Allen JC. Improved prognosis of
germinoma. Int J Radiat Oncol Biol Phys. 1997;37(3):511–5. intracranial non-germinoma germ cell tumors with multimodality
PubMed PMID: 9112446. Epub 1997/02/01. eng. therapy. J Neurooncol. 1997;32(1):71–80. PubMed PMID:
76. Sung DI, Harisliadis L, Chang CH. Midline pineal tumors and supra- 9049865. Epub 1997/03/01. eng.
sellar germinomas: highly curable by irradiation. Radiology. 85. Friedman JA, Lynch JJ, Buckner JC, Scheithauer BW, Raffel
1978;128(3):745–51. PubMed PMID: 674649. Epub 1978/09/01. eng. C. Management of malignant pineal germ cell tumors with residual
77. Wara WM, Fellows CF, Sheline GE, Wilson CB, Townsend mature teratoma. Neurosurgery. 2001;48(3):518–22. PubMed
JJ. Radiation therapy for pineal tumors and suprasellar germino- PMID: 11270541. discussion 22–3; Epub 2001/03/29. eng.
mas. Radiology. 1977;124(1):221–3. PubMed PMID: 405710. 86. Hasegawa T, Kondziolka D, Hadjipanayis CG, Flickinger JC,
Epub 1977/07/01. eng. Lunsford LD. Stereotactic radiosurgery for CNS nongerminoma-
78. Uematsu Y, Tsuura Y, Miyamoto K, Itakura T, Hayashi S, Komai tous germ cell tumors. Report of four cases. Pediatr Neurosurg.
N. The recurrence of primary intracranial germinomas. Special ref- 2003;38(6):329–33. PubMed PMID: 12759512. Epub 2003/05/22.
erence to germinoma with STGC (syncytiotrophoblastic giant cell). eng.
J Neurooncol. 1992;13(3):247–56. 87. Herrmann HD, Winkler D, Westphal M. Treatment of tumours of
79. Utsuki S, Kawano N, Oka H, Tanaka T, Suwa T, Fujii K. Cerebral the pineal region and posterior part of the third ventricle. Acta
germinoma with syncytiotrophoblastic giant cells: feasibility of Neurochir (Wien). 1992;116(2–4):137–46. PubMed PMID:
predicting prognosis using the serum hCG level. Acta Neurochir 1502947. Epub 1992/01/01. eng.
(Wien). 1999;141(9):975–7. PubMed PMID: 10526079. discussion 88. Chandy MJ, Damaraju SC. Benign tumours of the pineal region: a
7–8; Epub 1999/10/20. eng. prospective study from 1983 to 1997. Br J Neurosurg.
80. Utsuki S, Oka H, Tanaka S, Tanizaki Y, Fujii K. Long-term outcome 1998;12(3):228–33. PubMed PMID: 11013685. Epub 2000/10/03.
of intracranial germinoma with hCG elevation in cerebrospinal fluid eng.
but not in serum. Acta Neurochir (Wien). 2002;144(11):1151–4. 89. Kang JK, Jeun SS, Hong YK, Park CK, Son BC, Lee IW, et al.
PubMed PMID: 12434171. discussion 4–5; Epub 2002/11/16. eng. Experience with pineal region tumors. Childs Nerv Syst.
81. Balmaceda C, Heller G, Rosenblum M, Diez B, Villablanca JG, 1998;14(1–2):63–8. PubMed PMID: 9548344. Epub 1998/04/21.
Kellie S, et al. Chemotherapy without irradiation–a novel approach eng.
for newly diagnosed CNS germ cell tumors: results of an interna- 90. Shin HJ, Cho BK, Jung HW, Wang KC. Pediatric pineal tumors:
tional cooperative trial. The First International Central Nervous need for a direct surgical approach and complications of the occipi-
System Germ Cell Tumor Study. J Clin Oncol. 1996;14(11):2908– tal transtentorial approach. Childs Nerv Syst. 1998;14(4–5):174–8.
15. PubMed PMID: 8918487. Epub 1996/11/01. eng. PubMed PMID: 9660118. Epub 1998/07/11. eng.
82. Kochi M, Itoyama Y, Shiraishi S, Kitamura I, Marubayashi T, Ushio 91. Bruce J. Youman’s neurological surgery: pineal tumors. 5th ed.
Y. Successful treatment of intracranial nongerminomatous malignant Philadelphia: WB Saunders; 2004.
Pineal Tumors: Viewpoint—
Fractionated Radiation Therapy 36
William G. Rule and Steven E. Schild
that form sheets, cords, and gland-like arrangements. In rare

Introduction instances, they contain plate-like miniature embryos or
embryoid bodies. Yolk sac tumors are epithelial tumors that
An estimated 22,910 primary brain/CNS tumors will be diag- contain compact sheets, ribbons, cords, or papillae. Schiller–
nosed in the United States in 2012. Of these, approximately Duval bodies are diagnostic of yolk sac tumors. They contain
13,700 will result in death [1]. Pineal tumors make up only tufted epithelium-covered vessels projecting into clear spaces.
0.4–1 % of all brain tumors and demonstrate an overall male In addition to Schiller–Duval bodies, yolk sac tumors often
to female predominance of 3 to 1 [2, 3]. The pineal gland have a loose-knit vitelline pattern. Choriocarcinomas contain
region is unusual given the wide variety of primary tumors a bilaminar arrangement of both syncytiotrophoblasts and
that can arise from this site [4–9]. Benign cysts can also occur cytotrophoblasts. Combinations of any of these tumor types
in this site, as can ordinary brain tumors such as astrocytomas are referred to as mixed GCTs. One specific mixed GCT, the
or meningiomas. However, these are not as common as the teratocarcinoma, contains elements of both embryonal carci-
germ cell tumors (GCTs), which make up approximately two noma and teratoma [4–9].
thirds of the malignancies of the pineal gland. Some GCTs produce markers such as AFP (α[alpha]-
It is assumed that germ cells migrate during embryogene- fetoprotein) or β[beta]-HCG (β[beta]-human chorionic
sis and can lodge at the ectopic pineal region site, becoming gonadotropin). Yolk sac tumors (endodermal sinus tumors)
malignant GCTs later in life. The most common of these are can produce AFP, choriocarcinomas can produce β-HCG,
the germinomas, which are histologically indistinguishable and embryonal tumors can produce both markers. Pure ger-
from testicular seminomas and ovarian dysgerminomas. minomas can produce β-HCG but usually not in very high
Nongerminomatous germ cell tumors (NGGCTs) also occur levels. These markers are generally found in higher concen-
in pure and mixed forms. Histologic varieties of nongermino- tration in the cerebrospinal fluid (CSF) than in the serum and
matous GCTs include teratomas, embryonal carcinomas, yolk can aid in identifying the tumor type, as well as monitoring
sac tumors (endodermal sinus tumors), choriocarcinomas, or the effects of therapy.
combinations referred to as mixed GCTs. Teratomas typically Tumors may also arise in the pineocytes forming a group
include cell types derived from all three germ cell layers: of tumors called pineal parenchymal tumors (PPTs), which
endoderm, mesoderm, and ectoderm. Mature teratomas con- compose 15–30 % of all pineal tumors. These can be classi-
tain fully differentiated cells. Immature teratomas consist of fied into four groups: pineocytomas, mixed PPTs, PPTs with
cells and tissues resembling those of the developing fetus. By intermediate differentiation, and pineoblastomas [4, 6, 8, 9].
convention, tumors in which mature and immature compo- Pineocytomas are well-circumscribed masses that compress
nents coexist are classified as immature. Other teratomas with surrounding structures. They are composed of mature-
malignant epithelial or mesenchymal elements are considered appearing cells arranged in sheets or ill-defined irregular
to be teratomas with malignant transformation. Embryonal lobules. Fibrillary processes with club-like endings project
carcinomas contain large, primitive-appearing epithelial cells from these cells, which take up silver stains. Pineocytomatous
rosettes are formed around collections of these fibrillary
processes. Pineoblastomas are grossly invasive and micro-
W.G. Rule, M.D. (*) • S.E. Schild scopically highly cellular with small, mitotically active,
Department of Radiation Oncology, Mayo Clinic College
poorly differentiated cells with scant cytoplasm arranged
of Medicine, Mayo Clinic Arizone, 5777 East Mayo Boulevard,
Phoenix, AZ 85054, USA in pattern-less sheets similar to other primitive neuro-
e-mail: rule.william@mayo.edu ectodermal tumors (PNETs) and small blue cell tumors.
470 W.G. Rule and S.E. Schild
They resemble retinoblastomas and medulloblastomas. Table 36.1 Probability of spinal seeding at diagnosis according to
Flexner–Wintersteiner rosette formation is a prominent tumor type
feature of pineoblastomas. These are rosettes of cells form- No. with spine seeding/
ing around small lumens. Pineoblastomas occurring in Histologic types total no. of patients (%)
patients with bilateral retinoblastomas are called trilateral 1. Mature and immature teratoma 0/20 (0 %)
retinoblastomas. 2. Mixed NGGCT 1/26 (4 %)a
3. Other NGGCT 0/11 (0 %)
Symptoms from pineal masses are related to the anatomy
4. Germinoma 2/48 (4 %)b
of the region. They often obstruct the sylvian aqueduct caus-
5. Pineocytoma 0/9 (0 %)
ing hydrocephalus and can result in increased intracranial
6. Pineoblastoma, PPTID, mixed PPT 4/21 (19 %)c
pressure, which can lead to headaches, nausea, vomiting, cog- Total 7/135 (5 %)
nitive dysfunction, and incontinence. Additionally, they can a
This patient had a mixed germ cell tumor, composed of immature tera-
cause pressure on the superior colliculus (midbrain) causing toma and germinoma, and had positive CSF cytology
Parinaud syndrome or dorsal midbrain syndrome with a triad b
Of these two patients with evidence of seeding at diagnosis, one had a
of signs consisting of vertical gaze palsy, light-near dissocia- positive CSF cytology and the other had radiographic evidence of spi-
tion of the pupils, and convergence retraction nystagmus. nal metastases
c
Of the four patients with evidence of seeding at diagnosis, two had
A major scientific advance took place in 1973: computed pineoblastomas (one with clinical evidence of spinal cord compression
tomography (CT) was first used in major US medical cen- and one with radiographic evidence of spinal seeding) and two had pineal
ters. This tool made the diagnosis and localization of pineal parenchymal tumors with intermediate differentiation (one with positive
tumors easier, leading to a resurgence of surgical interven- CSF cytology and one with radiographic evidence of spinal seeding)
Note: All seven patients with evidence of spinal seeding at diagnosis
tion with greater safety. Further improvements took place received craniospinal irradiation and of these only one (with a pineoblas-
with the introduction of magnetic resonance imaging (MRI), toma) had a subsequent treatment failure at the primary site and the spine
which aided in localization as well as staging of the entire
central nervous system (CNS).
Prior to 1973, the operative risks of a biopsy or resection The GCTs included 48 germinomas, 26 mixed GCTs, 11
were substantial, with mortality rates as high as 50 % [10– mature teratomas, 9 immature teratomas, 6 malignant terato-
12]. At that time, many experts recommended operative inter- mas (with carcinomatous or sarcomatous components), 2
vention be reserved for CSF shunting in patients with yolk sac tumors, and 3 choriocarcinomas. At the time of
hydrocephalus and for the treatment of tumors that pro- diagnosis, 7 (5 %) of the 135 patients were found to have
gressed after radiotherapy. Conservative therapy, including evidence of spinal seeding (Table 36.1). Spinal seeding most
CSF shunting and radiotherapy, resulted in a generally favor- commonly occurred in patients with pineoblastomas.
able 5-year survival rate of approximately 70 % [10–12]. One hundred patients underwent radiotherapy (Table 36.2).
However, therapy can only be customized to the specific Regions treated included the craniospinal axis in 35 patients,
tumor type if the exact tumor histology is known. Additionally, the whole brain in 26 patients, and the partial brain in the
obtaining histologic confirmation, prior to therapy, allows remaining 39 patients. Doses delivered to the primary tumor
one to avoid irradiating benign lesions such as cysts [13, 14]. bed ranged from 4,500 to 6,480 cGy (median dose, 4,132 cGy)
Over the past several decades, advances in CT imaging, MRI, in 150- to 200-cGy fractions. Eleven patients treated during
stereotactic techniques, operating microscopes, neurosurgi- the earlier years of this study received total doses of less than
cal techniques, and postsurgical care have dramatically 3,000 cGy. When treated, the whole brain received 2,000–
decreased the morbidity and mortality previously associated 5,040 cGy and the spine received 2,000–3,800 cGy.
with biopsy or resection of pineal region tumors [15]. Chemotherapy was administered to 35 patients as a compo-
We evaluated a multi-institutional cohort of 135 patients nent of initial therapy (Table 36.2). These patients received
with histologically verified pineal region tumors [8]. Thirty- various combinations of the following agents: BCNU,
five females and 100 males with ages ranging from 3 days to CCNU, cyclophosphamide, prednisone, vincristine, vinblas-
77 years and a median age of 17 years were included in this tine, cisplatin, etoposide, dactinomycin, bleomycin, chloram-
study. This series was not pure in that primary sites included bucil, thiotepa, procarbazine, and melphalan.
the pineal gland in 81 patients, the suprasellar region in 31 Patients were followed for 0.25–37.3 years or until death
patients, and the cerebrum in 23 patients. Gross total resec- (median follow-up, 5.3 years). The survival rate for the entire
tions were performed in 26 patients, subtotal resections were group of patients was 62 % at 5 years. Survival was worse for
performed in 49 patients, and biopsies were performed in 60 patients treated during the early years of this study. Patients
patients. There were two groups of neoplasms included in diagnosed before 1973 had a 5-year survival rate of 34 %
this series: PPTs and GCTs. The PPTs included 15 pineo- compared with 66 % for those patients diagnosed more
blastomas (PB), 2 mixed PPTs (mixed PPT), 4 PPTs with recently (p = 0.0002). Twelve of the patients who were
intermediate differentiation (PPTID), and 9 pineocytomas. operated on between 1936 and 1950 died in the immediate
36 Pineal Tumors: Viewpoint—Fractionated Radiation Therapy 471
Table 36.2 Summary of therapya Therapy

Histologic type RT No RT Chemotherapy No chemotherapy
1. Mature teratoma 2 5 1 6
2. Immature teratoma 2 7 0 9
3. Mixed NGGCT 19 6 17 8
4. Other pure NGGCT 5 4 3 6
5. Germinoma 48 0 8 40
6. Pineocytoma 6 1 0 7
7. PB, PPTID, mixed PPT 18 0 6 12
Surgical procedure
Histologic type Biopsy Subtotal resection Gross total resection
1. Mature teratoma 0 3 4
2. Immature teratoma 3 2 4
3. Mixed NGGCT 3 12 10
4. Other pure NGGCT 4 3 2
5. Germinoma 29 16 3
6. Pineocytoma 3 4 0
7. PB, PPTID, mixed PPT 9 6 3
RT radiotherapy, NGGCT nongerminomatous germ cell tumor, PB pineoblastoma, PPTID pineal parenchymal
tumor with intermediate differentiation, mixed PPT mixed pineal parenchymal tumor
a
Excludes 12 of the 135 patients who died postoperatively during the early years of the study
postoperative period. These patients were excluded from radiotherapy in patients with mature and immature teratomas
the remainder of the analysis because their deaths were could not be assessed. Patients with NGGCTs, other than
more likely a result of operative complications than of mature and immature teratomas, who received radiotherapy
progressive disease. had a 3-year survival rate of 46 % compared with 11 % for
Tumor histology was evaluated for its relationship to sur- patients who received no radiotherapy (p = 0.0089). There
vival. The 5-year patient survival rate was 86 % for those was an association between the dose of radiation adminis-
with mature teratomas, 86 % with pineocytomas, 80 % with tered and survival in patients with NGGCTs other than
germinomas, 67 % with immature teratomas, 49 % with mature and immature teratomas. Patients who received
PPTs other than pineocytomas (PB, PPTIDs, and mixed ≤5,000 cGy had a 3-year survival rate of 21 % compared
PPTs), 38 % with mixed GCTs, and 17 % with the other pure with 55 % for patients receiving higher doses (p = 0.02). The
NGGCTs (p = 0.0001). Age, sex, and tumor location were dose of radiation administered correlated with survival rates
not associated with survival. in patients with PPTs. Patients with PPTs receiving doses of
The extent of resection was evaluated for its effect on ≤5,000 cGy had a 3-year survival rate of 56 %, and patients
patient survival. NGGCTs were the only tumors for which who received higher doses had a 3-year survival rate of 94 %
survival was associated with extent of tumor resection. The (p = 0.03). There was also an association between the dose of
3-year survival rate was 0 for patients having a biopsy, 36 % radiation administered and survival in patients with germino-
for patients having a subtotal resection, and 73 % for patients mas. Those patients with germinomas receiving doses of
having a gross total resection (p = 0.0002). ≤4,400 cGy had a 5-year survival rate of 70 % compared
The relationship between the administration of chemo- with 92 % for patients who received higher doses (p = 0.04).
therapy (as a component of initial therapy) and survival was There was no association between patient survival and radia-
evaluated. Patients who received chemotherapy had a 5-year tion field arrangement (partial-brain irradiation, whole-brain
survival rate of 45 % compared with 65 % for those patients irradiation, or craniospinal axis irradiation).
who did not (p = 0.37). However, the administration of che- Spinal seeding was the predominant pattern of distant
motherapy was associated with improved survival rates in failure. The outcome for patients with evidence of spinal
patients with NGGCTs other than mature and immature tera- seeding at the time of diagnosis is summarized in the foot-
tomas. Patients with NGGCTs who received chemotherapy notes of Table 36.1. The risk of spinal failure relative to both
had a 3-year survival rate of 56 % compared with 8 % for tumor type and radiotherapy field arrangement for patients
patients who did not receive chemotherapy (p = 0.0001). without evidence of spinal seeding at diagnosis is shown in
Most of the patients (100 of 123) received radiotherapy. Table 36.3. Patients with PPTs (other than pineocytomas)
However, only four patients with either mature or immature and GCTs (other than mature and immature teratomas) had
teratomas received radiotherapy. Therefore, the effect of the greatest risk of spinal failure.
Table 36.3 Probability of spinal failure in patients without evidence of spinal seeding at diagnosis
Histologic types No. with spine failure/total no. of patients (%)a
1. Mature and immature teratoma 0/16 (0 %)
2. Mixed NGGCTb 1/24 (4 %)
3. Other NGGCTb 3/9 (33 %)
4. Germinoma 8/46 (17 %)
5. Pineocytoma 0/7 (0 %)
6. Pineoblastoma, PPTID, mixed PPT 8/14 (57 %)
Total 20/116 (17 %)
Probability of spinal failure according to the radiotherapy fields in patients without evidence of spinal seeding at
diagnosisc (including only seeding malignancies) (%)
Histologic types PB WB CSPRT Total
1. NGGCTd 2/4 (50 %) 0/8 (0 %) 1/11 (9 %) 3/23 (13 %)
2. Germinoma 7/25 (28 %) 1/11 (9 %) 0/10 (0 %) 8/46 (17 %)
3. PPTe 2/3 (66 %) 2/5 (40 %) 4/6 (66 %) 8/14 (57 %)
Total 11/32 (34 %) 3/24 (13 %) 5/27 (19 %) 19/83 (23 %)
PB partial-brain fields, WB whole-brain fields, CSPRT craniospinal irradiation, NGGCT nongerminomatous germ cell
tumor, PPT pineal parenchymal tumor, RT radiotherapy
a
Some of these patients received no RT
b
Of the NGGCT patients with spinal failure, three of these patients had teratomas with malignant transformation and one
had a yolk sac tumor. Not all of the patients with NGGCTs received RT
c
This included only patients with potentially seeding tumors (NGGCTs excluding mature and immature teratomas, pineal
parenchymal tumors excluding pineocytomas, and germinomas) who received RT. One patient with a NGGCT received
no RT and later had a spinal failure
d
NGGCTs excluding mature and immature teratomas
e
Excluding pineocytomas
Only one patient developed distant metastases outside the Caution should be exercised when obtaining CSF from
central nervous system. This individual had a mixed GCT patients with increased intracranial pressure, because hernia-
with a component of malignant teratoma. After placement of tion of the brain may occur if pressure is rapidly reduced.
a ventriculo-peritoneal shunt for the treatment of tumor- The treatment of pineal and other primary CNS GCTs
induced hydrocephalus, he developed peritoneal metastases requires a precise knowledge of tumor histology. Obtaining a
and died. histologic diagnosis is important because optimal therapy for
the various tumor types differs significantly. Improvements
in surgical technique and postoperative care have decreased
Conclusion the morbidity once associated with resection or biopsy of
these tumors. Popovic and Kelly reported on 34 patients with
Pineal region tumors are rare and include a large variety of pineal lesions who underwent 66 stereotactic procedures at
lesions. Many studies included patients whose tumor types the Mayo Clinic (including 37 biopsies and 10 resections)
were not histologically verified because of the risks associ- [15]. Diagnostic tissue was obtained in 33 of the 34 patients.
ated with operative intervention [10, 11, 16–18]. As a result, No mortality or permanent morbidity was observed. Because
these series may have included patients with benign pro- the survival of patients with NGGCTs appears dependent
cesses such as cysts and vascular lesions. Prognosis is clearly upon the extent of resection, total resection should be
dependent upon tumor histology. This finding has been attempted in all cases where it can be safely performed.
reported by many investigators [19–32], with patients with Obtaining tissue for diagnosis is recommended in all patients
pineocytomas, germinomas, and mature teratomas demon- with pineal region tumors whenever possible.
strating the most favorable survival rates. Platinum-based multiagent chemotherapy has dramati-
Jennings et al. found that survival was dependent upon the cally improved the outcome in patients with NGGCTs [11].
extent of tumor, which emphasizes the importance of careful Our series revealed that a significant increase in survival was
staging [25]. The staging workup should include a careful his- associated with the administration of chemotherapy to
tory and physical examination, MRI of the brain and spine, patients with NGGCTs other than mature and immature tera-
spinal fluid cytology, complete blood cell count, chemistry tomas. Many other authors have recommended the use of
panel, baseline ophthalmologic exam, and tumor marker stud- platinum-based multiagent chemotherapy for the treatment
ies (AFP and β[beta]-HCG from both the CSF and serum). of NGGCTs of the brain [20, 21, 25, 26, 32–36].
Radiotherapy should be administered to all patients with Given the good response rates of germinomas to chemo-
pineal region tumors other than possibly the mature and therapy, attempts have been made to use chemotherapy alone
immature teratomas for which little data exist regarding in order to avoid radiotherapy. However, the chemotherapy-
radiotherapy. Careful consideration should be given to the only approach has been shown to yield unacceptably high
dose, because survival in patients with NGGCTs (other than tumor recurrence rates [45–47]. In an attempt to strike an
mature and immature teratomas), PPTs, and germinomas is ideal balance between disease-related outcomes and long-
dependent upon the dose of radiation administered to the pri- term treatment-related sequelae, current work is aimed at
mary tumor. The extent and dose of radiation should be tai- exploring the best combination of pre-radiotherapeutic che-
lored to the particular tumor type as outlined below. motherapy with reduced dose radiation (in the disseminated
Patients with PPTs (other than pineocytomas) have a high and non-disseminated settings). When possible, every effort
risk of spinal failure, and craniospinal axis radiotherapy should be made to enroll patients with germinomas in cur-
would appear reasonable (particularly in pineoblastomas). rent cooperative group clinical trials.
However, patients with PPTs of intermediate differentiation Based on our series, patients with NGGCTs (other than
with no spinal metastasis can be given whole-brain irradia- immature and mature teratomas) but no spinal seeding can
tion because the risk of spinal seeding appears no greater be treated with either whole-brain or craniospinal irradiation.
after whole-brain irradiation than after craniospinal irradia- Spinal failure occurred in two of the four patients treated
tion (Table 36.3). In general, pineoblastomas are treated with partial brain fields. There was a low risk of spinal fail-
similarly to medulloblastomas, with craniospinal axis irradi- ure with either whole-brain or craniospinal irradiation
ation with a subsequent boost, followed by adjuvant chemo- (Table 36.3). In other studies, whole-ventricular-irradiation
therapy. The primary tumor should receive 5,040–5,400 cGy, has shown some promise as a potentially reasonable alterna-
and spinal metastases should receive 5,040 cGy in 180 cGy tive in selected patients [48–50]. As discussed above, it
fractions. Prophylactic therapy can include the delivery of has been shown that NGGCT patients also experience
3,000–4,500 cGy to uninvolved high-risk areas. Patients with improved survival with the administration of chemotherapy.
pineocytomas can be irradiated to the local tumor alone. Chemotherapeutic regimens typically include cisplatin (or
None of the pure pineocytomas we studied metastasized. carboplatin) and etoposide, with some regimens also adding
It appears that the radiation dose (>5,000 cGy versus cyclophosphamide or ifosfamide. The potential benefits of
≤5,000 cGy) delivered to the primary tumor was the most prophylactic spinal irradiation must be carefully weighed
significant factor affecting survival in patients with PPTs. against potential toxicity in patients requiring platinum-
Patients with germinomas but no spinal seeding still have a based multiagent chemotherapy. Craniospinal irradiation
modest risk of spinal failure. Historically, children with non- and systemic chemotherapy should be considered for patients
disseminated germinomas were treated with craniospinal irra- who present with spinal seeding. The primary tumor should
diation (3,600 cGy) followed by a boost to the primary tumor receive 5,040–5,400 cGy and spinal metastases should
(total dose of 5,000–5,400 cGy). Lower doses of craniospinal receive 5,040 cGy in 180 cGy fractions. Prophylactic ther-
irradiation and lower boost doses have demonstrated good apy can include the delivery of 2,400–3,600 cGy to unin-
efficacy in retrospective as well as prospective settings [37– volved high-risk areas. Second-look surgery should be
39]. Replacing craniospinal irradiation with whole-brain or considered for patients with persistent masses/residual tumor
whole-ventricular-irradiation has also been shown to yield after a course of chemotherapy and radiation. As with germi-
good results, with low spinal failure rates [5, 37, 40–43]. In nomas, enrollment in clinical trials should be strongly con-
our study, no patient who received craniospinal irradiation had sidered for NGGCT patients.
a spine failure and only 1 of 11 (9 %) patients treated with Immature and mature teratomas should be resected, if this
whole-brain irradiation failed (Table 36.3). While the rate of can be safely achieved. We cannot make definitive recom-
spinal failure in patients who received partial-brain fields was mendations regarding the efficacy of radiotherapy or chemo-
substantial in our study, multiple other series have demon- therapy for patients with immature and mature teratomas,
strated very good outcomes with whole-ventricular-irradiation because of the small number of patients in our series treated
[37, 38, 41, 42, 44, 45]. The standard radiation therapy alone with these modalities. If progression occurs, further surgical
arm in the Children’s Oncology Group (COG) ACNS0232 intervention, radiotherapy, and/or chemotherapy should be
protocol was structured to deliver 2,400 cGy in 150 cGy frac- considered.
tions to whole-ventricular fields, with a 2,100 cGy boost in Table 36.4 lists the relative advantages of stereotactic
150 cGy fractions subsequently delivered to regions of gross radiosurgery (SRS) versus fractionated external beam irradi-
disease (total tumor dose of 4,500 cGy). Patients with spinal ation. Stereotactic radiotherapy and radiosurgery have poten-
seeding were to receive 2,400 cGy in 150 cGy fractions to the tial roles in the treatment of pineal tumors. These techniques
craniospinal axis, followed by a 2,100 cGy boost in 150 cGy would have the greatest potential as a single treatment modal-
fractions to regions of gross disease. ity in non-seeding tumors. The advantages for radiosurgery
Table 36.4 Advantages of stereotactic radiosurgery (SRS) versus fractionated external beam radiotherapy (FRT)
Advantage Disadvantage
Effectively address those at ↑ risk of CSF seeding FRT SRS
Most effectively minimizes RT dose to surrounding normal brain SRS FRT
More radiobiologically effective in the treatment of small, non-seeding tumors SRS FRT
Marrow suppression in patients who in some cases receive chemotherapy SRS FRT
May be an effective alternative to resection in histologies in which resection is believed to improve survival SRS FRT
Requires less patient time SRS FRT
Proven long-term efficacy FRT SRS
Germinomas FRT SRS
Nongerminomas FRT SRS
Pineocytomas SRS FRT
Other PPTs FRT SRS
Note: Ultimately, one might consider the following paradigm for future investigation in seeding pineal tumors (all those other than pineocytomas,
mature teratomas, and immature teratomas): Biopsy first (and shunting if needed), followed by FRT followed by a specialized boost (SRS, IMRT,
heavy ions …), and chemotherapy when appropriate. In non-seeding tumors, one might consider: Biopsy first (and shunting if needed) followed
by a specialized boost technique. The major rationale for this approach is the opinion that complete resection of a pineal tumor incurs a substan-
tially greater overall risk to the patient than does SRS. Additionally, the specialized boost would provide greater local control than FRT alone in
unresected tumors
in these cases are a shortened course compared with fraction- unknown. Pineoblastomas, variants of supratentorial PNETs,
ated external beam irradiation, and, more importantly, a are, in general, treated similarly to medulloblastomas. For all
reduction in radiation exposure to surrounding normal struc- other pineal tumors with seeding potential, radiosurgery or
tures. The latter is especially advantageous in pediatric stereotactic radiotherapy could be used as a boost to gross
patients. Pineocytomas and teratomas (both mature and disease after other therapy. Hasegawa et al. also reported on
immature) represent tumors in which this approach would be a small series of four patients treated for nongerminomatous
most applicable as they do not generally seed the CSF. There GCTs of the brain [35]. These patients received a Gamma
are examples in the literature where pineocytomas have been Knife boost after other therapies, and three of four were alive
reported to seed, but this may represent confusion regarding without disease at last follow-up. Although this is a small
the classification of PPTs, which should be subdivided into series, it does show potential and should lead to further
four groups [4, 6, 8]. The risks of radiosurgery are likely sub- investigation. The specialized radiation techniques might
stantially less than resection. Regarding PPTs, the University also be considered as salvage therapeutic maneuvers when
of Pittsburgh experience with Gamma Knife radiosurgery other treatments have failed.
would support these contentions. Hasegawa et al. reported
that the 5-year survival of ten patients with pineocytomas
was 90 % utilizing Gamma Knife radiosurgery [34]. In con- References
trast, the results of Gamma Knife treatment with pineoblas-
tomas were poor, with death occurring in four of six patients. 1. Siegel R, Naishadham D, Jemal A. Cancer statistics, 2012. CA
Cancer J Clin. 2012;62(1):10–29 [Comparative Study].
Complications including upward gaze paralysis occurred in
2. Zulch K. [Some characteristics of brain tumors related to the age,
2 of 16 (12.5 %) and was likely due to the dose delivered to site and sex of the tumor patients]. Zentralbl Chir. 1965;90(23):
the superior colliculus, which lies in direct contact with 890–8.
pineal tumors. Pineocytomas are round, circumscribed 3. Al-Hussaini M, Sultan I, Abuirmileh N, Jaradat I, Qaddoumi
I. Pineal gland tumors: experience from the SEER database. J
tumors that are generally not large at diagnosis because they
Neurooncol. 2009;94(3):351–8 [Research Support, N.I.H.,
cause symptoms as soon as pressure is exerted on the sylvian Extramural Research Support, Non-U.S. Gov’t].
aqueduct or the superior colliculus. However, PPTs with 4. Burger PC, Scheithauer BW. Pineal tumors. In: Tumors of the cen-
intermediate differentiation and mixed PPTs are often dif- tral nervous system: atlas of tumor pathology. Bethesda: Armed
Forces Institute of Pathology; 1994. p. 227–37.
fuse, invasive, and seeding. These should be treated with
5. Haddock MG, Schild SE, Scheithauer BW, Schomberg
either whole-brain radiation therapy followed by a boost to PJ. Radiation therapy for histologically confirmed primary central
the tumor in cases without seeding or craniospinal radiation nervous system germinoma. Int J Radiat Oncol Biol Phys.
therapy in cases with seeding followed by a boost to areas of 1997;38(5):915–23.
6. Mena H, Nakazato Y, Jouvet A, Scheithauer BW. Pineal parenchy-
gross disease. The options for the boost include external
mal tumors. In: Kleihues P, Cavanee WK, editors. World Health
beam irradiation, stereotactic radiotherapy, or radiosurgery. Organization classification of tumours. Pathology and genetics.
Which boost technique option offers the best outcome is Tumors of the nervous system. Lyon: IARC Press; 2000. p. 115–21.
7. Schild SE, Haddock MG, Scheithauer BW, Marks LB, Norman 29. Sano K, Matsutani M, Seto T. So-called intracranial germ cell
MG, Burger PC, et al. Nongerminomatous germ cell tumors of the tumours: personal experiences and a theory of their pathogenesis.
brain. Int J Radiat Oncol Biol Phys. 1996;36(3):557–63. Neurol Res. 1989;11(2):118–26 [Review].
8. Schild SE, Scheithauer BW, Haddock MG, Wong WW, Lyons MK, 30. Takakura K. Intracranial germ cell tumors. Clin Neurosurg.
Marks LB, et al. Histologically confirmed pineal tumors and other 1985;32:429–44 [Case Reports Comparative Study Research
germ cell tumors of the brain. Cancer. 1996;78(12):2564–71. Support, Non-U.S. Gov’t Review].
9. Schild SE, Scheithauer BW, Schomberg PJ, Hook CC, Kelly PJ, 31. Wolden SL, Wara WM, Larson DA, Prados MD, Edwards MS,
Frick L, et al. Pineal parenchymal tumors. Clinical, pathologic, and Sneed PK. Radiation therapy for primary intracranial germ-cell
therapeutic aspects. Cancer. 1993;72(3):870–80 [Review]. tumors. Int J Radiat Oncol Biol Phys. 1995;32(4):943–9.
10. Abay 2nd EO, Laws Jr ER, Grado GL, Bruckman JE, Forbes GS, 32. Yoshida J, Sugita K, Kobayashi T, Takakura K, Shitara N, Matsutani
Gomez MR, et al. Pineal tumors in children and adolescents. M, et al. Prognosis of intracranial germ cell tumours: effectiveness
Treatment by CSF shunting and radiotherapy. J Neurosurg. of chemotherapy with cisplatin and etoposide (CDDP and VP-16).
1981;55(6):889–95 [Comparative Study]. Acta Neurochir (Wien). 1993;120(3–4):111–7.
11. Donat JF, Okazaki H, Gomez MR, Reagan TJ, Baker Jr HL, Laws 33. Gobel U, Bamberg M, Engert J, Gnekow AK, Haas HJ, Jurgens H,
Jr ER. Pineal tumors. A 53-year experience. Arch Neurol. et al. [Treatment of non-testicular germ cell tumors in children and
1978;35(11):736–40. adolescents with BEP and VIP: initial results of the MAKEI 89
12. Marsh WR, Laws Jr ER. Shunting and irradiation of pineal tumors. therapy study]. Klin Padiatr. 1991;203(4):236–45 [Clinical Trial
Clin Neurosurg. 1985;32:384–96 [Review]. Multicenter Study].
13. Fain JS, Tomlinson FH, Scheithauer BW, Parisi JE, Fletcher GP, 34. Hasegawa T, Kondziolka D, Hadjipanayis CG, Flickinger JC,
Kelly PJ, et al. Symptomatic glial cysts of the pineal gland. Lunsford LD. The role of radiosurgery for the treatment of pineal
J Neurosurg. 1994;80(3):454–60. parenchymal tumors. Neurosurgery. 2002;51(4):880–9.
14. Fleege MA, Miller GM, Fletcher GP, Fain JS, Scheithauer 35. Hasegawa T, Kondziolka D, Hadjipanayis CG, Flickinger JC,
BW. Benign glial cysts of the pineal gland: unusual imaging char- Lunsford LD. Stereotactic radiosurgery for CNS nongerminoma-
acteristics with histologic correlation. AJNR Am J Neuroradiol. tous germ cell tumors. Report of four cases. Pediatr Neurosurg.
1994;15(1):161–6. 2003;38(6):329–33 [Case Reports].
15. Popovic EA, Kelly PJ. Stereotactic procedures for lesions of the pineal 36. Herrmann HD, Westphal M, Winkler K, Laas RW, Schulte
region. Mayo Clin Proc. 1993;68(10):965–70 [Case Reports Review]. FJ. Treatment of nongerminomatous germ-cell tumors of the pineal
16. Linstadt D, Wara WM, Edwards MS, Hudgins RJ, Sheline region. Neurosurgery. 1994;34(3):524–9. discussion 9.
GE. Radiotherapy of primary intracranial germinomas: the case 37. Bamberg M, Kortmann RD, Calaminus G, Becker G, Meisner C,
against routine craniospinal irradiation. Int J Radiat Oncol Biol Harms D, et al. Radiation therapy for intracranial germinoma: results
Phys. 1988;15(2):291–7 [Research Support, Non-U.S. Gov’t]. of the German cooperative prospective trials MAKEI 83/86/89. J Clin
17. Rich TA, Cassady JR, Strand RD, Winston KR. Radiation therapy for Oncol. 1999;17(8):2585–92 [Clinical Trial Multicenter Study].
pineal and suprasellar germ cell tumors. Cancer. 1985;55(5):932–40. 38. O’Neil S, Ji L, Buranahirun C, Azoff J, Dhall G, Khatua S, et al.
18. Wara WM, Jenkin RD, Evans A, Ertel I, Hittle R, Ortega J, et al. Neurocognitive outcomes in pediatric and adolescent patients with
Tumors of the pineal and suprasellar region: Childrens Cancer central nervous system germinoma treated with a strategy of che-
Study Group treatment results 1960–1975: a report from Childrens motherapy followed by reduced-dose and volume irradiation.
Cancer Study Group. Cancer. 1979;43(2):698–701 [Research Pediatr Blood Cancer. 2011;57(4):669–73.
Support, U.S. Gov’t, P.H.S.]. 39. Tseng CK, Tsang NM, Jaing TH, Liau CT, Wu CT, Lin KL. Outcome
19. Bjornsson J, Scheithauer BW, Okazaki H, Leech RW. Intracranial of central nervous system germinoma treatment by chemoradiation.
germ cell tumors: pathobiological and immunohistochemical J Pediatr Hematol Oncol. 2011;33(4):e138–42 [Clinical Trial].
aspects of 70 cases. J Neuropathol Exp Neurol. 1985;44(1):32–46. 40. Dattoli MJ, Newall J. Radiation therapy for intracranial germi-
20. Bruce J, Fetell M, Stein B. Incidence of spinal metastases in patients noma: the case for limited volume treatment. Int J Radiat Oncol
with malignant pineal region tumors: avoidance of prophylactic Biol Phys. 1990;19(2):429–33.
cranial irradiation (meeting abstract). J Neurosurg. 1990;72:354A. 41. Aoyama H, Shirato H, Kakuto Y, Inakoshi H, Nishio M, Yoshida H,
21. Bruce J, Stein B, Balmaceda C. Management of pineal region germ et al. Pathologically-proven intracranial germinoma treated with
cell tumors: results and long-term follow-up (meeting abstract). radiation therapy. Radiother Oncol. 1998;47(2):201–5 [Multicenter
Proc Ann Meet Am Soc Clin Oncol. 1994;13:A505. Study].
22. Dearnaley DP, A’Hern RP, Whittaker S, Bloom HJ. Pineal and CNS 42. Shikama N, Ogawa K, Tanaka S, Toita T, Nakamura K, Uno T, et al.
germ cell tumors: Royal Marsden Hospital experience 1962–1987. Lack of benefit of spinal irradiation in the primary treatment of
Int J Radiat Oncol Biol Phys. 1990;18(4):773–81. intracranial germinoma: a multiinstitutional, retrospective review
23. Fuller BG, Kapp DS, Cox R. Radiation therapy of pineal region of 180 patients. Cancer. 2005;104(1):126–34 [Evaluation Studies
tumors: 25 new cases and a review of 208 previously reported Multicenter Study Research Support, Non-U.S. Gov’t].
cases. Int J Radiat Oncol Biol Phys. 1994;28(1):229–45. 43. Yen SH, Chen YW, Huang PI, Wong TT, Ho DM, Chang KP, et al.
24. Hoffman HJ, Otsubo H, Hendrick EB, Humphreys RP, Drake JM, Optimal treatment for intracranial germinoma: can we lower radia-
Becker LE, et al. Intracranial germ-cell tumors in children. tion dose without chemotherapy? Int J Radiat Oncol Biol Phys.
J Neurosurg. 1991;74(4):545–51. 2010;77(4):980–7 [Evaluation Studies Research Support,
25. Jennings MT, Gelman R, Hochberg F. Intracranial germ-cell Non-U.S. Gov’t].
tumors: natural history and pathogenesis. J Neurosurg. 1985;63(2): 44. Shirato H, Nishio M, Sawamura Y, Myohjin M, Kitahara T,
155–67 [Review]. Nishioka T, et al. Analysis of long-term treatment of intracranial
26. Linggood RM, Chapman PH. Pineal tumors. J Neurooncol. germinoma. Int J Radiat Oncol Biol Phys. 1997;37(3):511–5
1992;12(1):85–91. [Research Support, Non-U.S. Gov’t].
27. Mineura K, Sasajima T, Sakamoto T, Kowada M. Gan No Rinsho. 45. Kellie SJ, Boyce H, Dunkel IJ, Diez B, Rosenblum M, Brualdi L,
1990;36(14):2399–403 [Results of the treatment of intracranial et al. Intensive cisplatin and cyclophosphamide-based chemother-
germ cell tumors]. apy without radiotherapy for intracranial germinomas: failure of a
28. Sano K. Pineal region tumors: problems in pathology and treatment. primary chemotherapy approach. Pediatr Blood Cancer. 2004;43(2):
Clin Neurosurg. 1983;30:59–91. 126–33 [Clinical Trial].
46. Balmaceda C, Heller G, Rosenblum M, Diez B, Villablanca JG, 48. Calaminus G, Frappaz D, Kortmann RD, et al. Localized and
Kellie S, et al. Chemotherapy without irradiation—a novel approach metastatic nongerminoma treated according to the SIOP CNS GCT
for newly diagnosed CNS germ cell tumors: results of an interna- 96 protocol: update on risk profiles and outcome. J Neurooncol.
tional cooperative trial. The First International Central Nervous 2008;10:418.
System Germ Cell Tumor Study. J Clin Oncol. 1996;14(11):2908– 49. Ogawa K, Toita T, Nakamura K, Uno T, Onishi H, Itami J, et al.
15 [Clinical Trial Controlled Clinical Trial Multicenter Study Treatment and prognosis of patients with intracranial nongermino-
Research Support, Non-U.S. Gov’t]. matous malignant germ cell tumors: a multiinstitutional retrospec-
47. da Silva NS, Cappellano AM, Diez B, Cavalheiro S, Gardner S, tive analysis of 41 patients. Cancer. 2003;98(2):369–76.
Wisoff J, et al. Primary chemotherapy for intracranial germ cell 50. Matsutani M. Treatment of intracranial germ cell tumors: the sec-
tumors: results of the third international CNS germ cell tumor ond phase II study of Japanese GCT Study Group. J Neurooncol.
study. Pediatr Blood Cancer. 2010;54(3):377–83 [Clinical Trial]. 2008;10:420.
Pineal Region Tumors:
Viewpoint—Chemotherapy 37
Christopher Dardis and Roy A. Patchell
writing, the NIH lists over 50 studies including patients

Introduction affected by one of these tumors, the majority of which include
a chemotherapeutic element. Overall, regimens tend to be
Tumors involving the pineal gland are uncommon, accounting similar in adults and children, reflecting a belief in the simi-
for 4–11 % of intracranial tumors in children and 0.4–1 % of larity of tumor biology across the lifespan. Likewise, given
such tumors in adults [1]. As noted in preceding chapters, the their rarity, the tendency has been to lump rather than split in
conventional classification system divides them into four retrospective reviews and in designing effective regimens.
broad histologic types, and treatment is generally based
on this type (rather than anatomic location). These types are:
(1) germ cell, (2) pineal cell, (3) glial cell, and (4) miscella- Germ Cell Tumors
neous. Speaking generally, chemotherapy has an adjunctive
role in most of these tumors. Primary GCTs account for approximately half of pineal
Regarding glial tumors, little will be said here as the che- tumors. Central nervous system (CNS) GCTs mostly arise in
motherapeutic management is not different for those arising the midline: 80 % or more arise in structures near the third
in the pineal as opposed to elsewhere; interested readers are ventricle, with the region of the pineal gland being the most
referred to Chap. 20 in the current volume. The following common site of origin. Multifocal disease is common. CNS
may also be seen here: meningiomas, choroid plexus papil- GCTs are seen most commonly in childhood and young
lomas, neuronal cell tumors, ependymomas, lipomas, and adulthood. These tumors can be divided into germinomas
metastases. Again, the reader is referred to the appropriate (aka seminomas, also the most common type of testicular
chapters. cancer) and nongerminomatous GCTs, which include tera-
As noted elsewhere in the current volume, biopsy of toma (from Lt. teratos = deformity, consisting of tissues
pineal tumors is strongly recommended prior to the start derived from all three embryonal layers), embryonal cell
of therapy. In occasional cases, magnetic resonance imaging carcinoma (or stem cell, arising from a portion of cells with-
(MRI) of the brain with cerebrospinal fluid (CSF) sampling out which embryogenesis cannot continue), yolk sac tumor
may be sufficient to establish a diagnosis of certain hormone- (from O.E. geolu = yellow), and choriocarcinoma (from Gk.
secreting germ cell tumors (GCTs). (β[beta]HCG for chor- khorion = membrane enclosing the fetus).
iocarcinoma, α[alpha]FP for yolk sac, and embroyonal
carcinoma) An MRI of the entire spine is recommended to
look for “drop metastases” in the case of pineoblastoma. Germinomas
Recent years have seen substantially increased interest in
the role of chemotherapy for these conditions. At the time of For germinomas, conventional is to aim for subtotal resec-
tion where possible, with adjuvant radiation. As they are par-
C. Dardis, M.D. ticularly sensitive to radiation, this remains the cornerstone
Barrow Neurological Clinics, 500 W. Thomas Rd., of treatment, typically involving ventricular irradiation with
Suite 300, Phoenix, AZ 85013, USA a local boost [2]. Five-year survival in a series of 49 germi-
R.A. Patchell, M.D., F.A.A.N. (*) nomas was reported as 88 ± 5 % for such an approach. At one
National Brain Tumor Center, Capital Institute for Neurosciences, time complete craniospinal irradiation was employed pro-
Capital Health Medical Center-Hopewell, Two Capital Way, phylactically; however, given the rates of adverse sequelae
Suite 456, Pennington, NJ 08534, USA
e-mail: rpatchell@capitalhealth.org with this approach, surveillance is now preferred [3].
478 C. Dardis and R.A. Patchell
Generally these tumors are also sensitive to chemotherapy From this experience, it appears that aggressive chemo-
and the following agents have all been used with some therapy alone can induce durable complete remissions in only
success: alkylating agents (cisplatin, carboplatin, cyclophos- approximately 50 % of patients. Given the superior results
phamide, ifosphamide) as well as etoposide, vinblastine, with radiation for germinomas, the use of chemotherapy with-
adrimycin, and bleomycin. A platinum-based agent with out radiation cannot be considered routine for these patients.
etoposide is most commonly employed. Given how sensitive
they are to carboplatin in particular, it was hoped at one point
that this single agent may be able to replace radiation com- Non-germinomatous GCTs
pletely, although this no longer is a widely-advocated
approach [4]. Mature teratoma, like germinoma, tends to have a favorable
Neo-adjuvant chemotherapy with a platinum-based prognosis and remission may be achieved with resection
regime has been advocated by a number of centers, although alone. However, given the generally unfavorable prognosis
it remains to be seen whether this is of additional benefit by of the other tumors in this class, chemotherapy is not
comparison with the regimen proposed above [5]. A phase II controversial.
study by Allen et al. in 1994 showed complete response rates Historically non-germinomatous GCTs have been ana-
in 7 of 11 pediatric patients receiving carboplatin (150 mg/ lyzed together rather than separately and so relatively little is
m2 weekly, 2–4 cycles) [6]. known about their relative response to chemotherapy, i.e.,
In cases with dissemination there is clearly a role for whether regimes should include the particular histology in
craniospinal irradiation and for adjuvant chemotherapy. their design. Of the group, choriocarcinomas tend to have a
Likewise, in cases of local or systemic relapse, adjuvant che- significantly worse prognosis than the others.
motherapy is generally recommended. Given the relatively A consensus has emerged that adjuvant chemotherapy
low incidence of any of these complications, decisions are should generally include a platinum-based agent. A recent
typically based on small retrospective studies. Plantinum and review advises first line adjuvant treatment with POMB-
etoposide-based regimens are commonly employed, although ACE (cisPlantin, vincristine (Oncovorin), methotrexate,
bleomycin and cyclophosphamide have also been used [7]. bleomycin, actiomycin-d, cyclophosphamide, etoposide) or
Cyclophosphamide with re-irradiaiton had previously been EMA/CO (etoposide, methotrexate actinomycin-d, folinic
suggested as optimal monotherapy for recurrence, with acid, cyclophosphamide, vincrisite (Oncovorin)) [11].
response rates of up to 5 months [8]. Other proposed adjuvant regimes include cisplatin, ifos-
Recent trials have also explored the possibility of reducing phamide, and etoposide [12], PVB (cisplatin, vinblastine,
radiation dosing by giving adjuvant chemotherapy. One study bleomycin) and PIV (cisplatin, ifosphamide, bleomycin) [13].
using cisplatin and etoposide (up to six cycles) following Earlier approaches with neo-adjuvant cisplantin and etopo-
whole brain or ventricular irradiation in 16 pediatric patients side followed by adjuvant carboplatin, etoposide, bleomycin,
has shown promising results [9]. and vinblastine also confirmed improvement with this
It bears noting that there have been attempts to eliminate multimodality approach [14]. These are typically selected on
radiation therapy altogether from the treatment regimen. The the basis of tolerability and employed over 4–6 cycles. All of
largest of these was initiated by the international CNS GCT the platinum-based regimens have improved prognosis for
study group. these tumors. However, given that this tends to remain rather
This was a trial with a common protocol for all GCTs dismal (3-year survival rates reported as 30 % for adults,
[10]. Of 71 patients enrolled, 45 had germinomas. Induction 60 % for those age under 16), clearly there remains room for
chemotherapy consisted of four cycles of carboplatin, etopo- improvement in these regimes [13]. In all cases, following
side, and bleomycin chemotherapy, a regimen adapted from hormonal markers, typically in serum, it is advised to assess
systemic GCT experience. (Cisplatin is considered more response to treatment.
effective than carboplatin for systemic disease.) Those For recurrence, further chemotherapy is advocated, typi-
achieving complete remission (CR) were given an additional cally platinum-based and more aggressive than before,
four cycles. An intensified regimen was offered to those not although a consensus has yet to emerge on an optimal
achieving complete remission. Of the germinomas, 37 of 45 regimen.
achieved complete response. However, the relapse-free sur- Metastases have been reported with these tumors, most
vival was disappointing, with approximately 50 % of patients commonly when mixed (i.e., a non-germinomatous element
with each pathologic subtype experiencing relapse or pro- is present). These can occur as a result of transit through a
gression. Twenty eight of 55 patients in complete remission VP-shunt or independently—presumably hematogenously.
without irradiation subsequently developed tumor recur- If well localized, these may be treated with irradiation, typi-
rence at a median time of 18 months from diagnosis. cally with a platinum and etoposide combination.
37 Pineal Region Tumors: Viewpoint—Chemotherapy 479
The frustratingly high relapse rate and the initial management consists of total resection and postoperative
refractoriness of nongerminoma GCTs to chemotherapy has radiation, with 1- and 5-year survival rates of 91 % and 88 %,
led to the use of high-dose chemotherapy (HDC) and autolo- respectively [17]. No significant improvement in survival
gous stem cell transplant for either relapse disease or to con- was found in the addition of radiation to subtotal resection,
solidate partially responsive disease. This strategy is based although this is likely to remain standard practice in most
on experience with systemic GCTs, as well as the increasing cases; chemotherapy was not considered further. It was sug-
experience with other high-risk primary brain tumors, such gested as far back as 2004 that platinum-based chemother-
as medulloblastoma. One of the largest such experiences was apy could have a supportive role to play, although this has
published by Modak et al. [15], a group based in New York not been explored further with clinical trials [18]. Sensitivity
and one of the leaders in the field of high-dose therapy and to nimustine (ACNU) has also been reported [19].
stem cell rescue for brain tumors. Like the CNS GCT trial Certain variants are more aggressive and, given the poor
above, it included all germ-cell tumors, in this case 21 patients prognosis, should generally be offered chemotherapy. The
with CNS GCTs who had either relapsed or progressed, papillary variant is particularly aggressive [20]. Meningeal
despite receiving chemotherapy and/or radiation therapy. They metastases also tend to have a poorer prognosis and ACNU
received high-dose thiotepa with another agent, typically eto- and vincristine has been suggested as a reasonable approach
poside, often with carboplatin. Resection after chemotherapy for such cases [21].
was undertaken when practical. Posttransplant radiation ther-
apy was added focally or to the entire neural axis whenever
feasible. Four of 12 patients with non-germinomatous GCTs Pineoblastomas and Pineal Parenchymal
achieved CR with a range of 24–55 months at time of publica- Tumors of Intermediate Differentiation
tion. One was alive with residual disease at 51 months (7 of 9
patients with recurrent germinomas survived disease-free, Pineoblastomas (PBs) and pineal parenchymal tumors of
with a range of 6–87 months, median 48). intermediate differentiation (PPTIDs) are very rare in adults.
These encouraging results suggest that long-term remis- To date, no prospective studies analyzing different chemo-
sions can be obtained even in a group of patients initially therapy regimens are available. A recent multicenter review
refractory to chemotherapy. It remains to be seen whether indicated the poorer outcomes of these entities compared
adopting this as the approach at time of diagnosis will offer with pineocytomas [22].
additional benefits over current regimens. Primary or secondary failure (relapse) rates were about
50 % at 10 years, with the worse outcomes seen among
patients who present with disseminated disease or who have
Teratomas large-volume residual disease. PB histology compared with
PPTID also had worse outcomes.
Given their relative rarity, a consensus on the role of adjuvant Conventional approaches to adjuvant chemotherapy for
chemotherapy for immature teratomas has yet to emerge. One these tumors typically involves use regimens similar to those
large retrospective series showed a good response to chemo- used for medulloblastoma (which pineoblastoma resembles
therapy, typically platinum-based, when signs of recurrence histologically), such as cisplatin, carboplatin, etoposide,
developed [16]. This did not, however, apply to those in the cyclophosphamide, and vincristine [23].
highly malignant subgroup, for whom no salvage therapies Given the poor prognosis treatment with HDC and autolo-
were effective. gous stem-cell rescue (ASCR) in patients with newly diag-
nosed PB remains a possible alternative [24]. In the largest
trial to date on this subject (2004), 12 patients (6 children
Pineal Parenchymal Tumors and 6 adults) underwent biopsy or complete resection (5 of
12) of primary tumor at the time of diagnosis.
These tumors arise from the pineocyte. They are classified by All patients received induction chemotherapy with one of
the WHO into pineocytoma (grade I), pineal parenchymal tumor three regimens.
of intermediate differentiation (grade II-III), and pineoblastoma Stem cells were harvested from bone marrow and/or
(grade IV). They are considered to represent a spectrum. peripheral blood after the first or second cycle of induction
chemotherapy after using granulocyte colony-stimulating
factor (G-CSF).
Pineocytomas All patients, except two infants, received craniospinal
irradiation at a median dose of 36 Gy with a boost to the
Chemotherapy is far from standard of care for these tumors. pineal region boost.
A review in 2010 looking at outcomes for all English- Eleven patients received HDC with cyclophosphamide
language cases of pineocytoma concluded that optimal 50 mg/kg daily for 4 days followed by melphalan 60 mg/m2/day
480 C. Dardis and R.A. Patchell
for 3 days or busulfan 1 mg/kg every 6 h for 16 total doses over 5. Edwards MS, Hudgins RJ, Wilson CB, Levin VA, Wara WM, et al.
4 days followed by melphalan 60 mg/m2/day for 3 days. Pineal region tumors in children. J Neurosurg. 1988;68(5):689–97.
6. Allen JC, DaRosso RC, Donahue B, Nirenberg A. A phase II trial
After HDC and ASCR, nine patients were disease-free of preirradiation carboplatin in newly diagnosed germinoma of the
at a range of 28–125 months (median 62), including three central nervous system. Cancer. 1994;74(3):940–4.
of four with metastatic disease at presentation and the two 7. Salzman KL, Rojiani AM, Buatti J, Quisling RG, Marcus Jr RB,
infants who did not receive radiotherapy. Two of eight Maria BL, et al. Primary intracranial germ cell tumors: clinicopath-
ologic review of 32 cases. Pediatr Pathol Lab Med. 1997;17(5):
patients with localized disease and one of four patients with 713–27.
metastatic disease developed progression and died. 8. Allen JC, Bosl G, Walker R. Chemotherapy trials in recurrent
This outcome in this small sample of patients using HDC primary intracranial germ cell tumors. J Neurooncol. 1985;3(2):
after induction chemotherapy and radiation are superior to 147–52.
9. Tseng CK, Tsang NM, Jaing TH, Liau CT, Wu CT, Lin KL, et al.
less-aggressive regimens. Outcome of central nervous system germinoma treatment by
Subsequent trials have tended to confirm these initial chemoradiation. J Pediatr Hematol Oncol. 2011;33(4):138–42.
results, although the numbers with pineoblastoma were gen- 10. Balmaceda C, Heller G, Rosenblum M, Diez B, Villablanca JG,
erally small and tended to be grouped with high-grade recur- Kellie S, et al. for the First International Central Nervous System
Germ Cell Tumor Study. Chemotherapy without irradiation—a
rent CNS tumors [25–27]. novel approach for newly diagnosed CNS germ cell tumors: results
of an international cooperative trial. J Clin Oncol. 1996;14:
2908–15.
Conclusion 11. Kyritsis AP. Management of primary intracranial germ cell tumors.
J Neurooncol. 2010;96(2):143–9.
12. Aoyama H, Shirato H, Ikeda J, Fujieda K, Miyasaka K, Sawamura
Chemotherapy has an established role to play in GCTs, par- Y, et al. Induction chemotherapy followed by low-dose involved-
ticularly non-germinomatous and pineoblastomas. At present, field radiotherapy for intracranial germ cell tumors. J Clin Oncol.
the agents in use are relatively nonspecific. Given their prog- 2002;20(3):857–65.
13. Khafaga Y, El Weshi A, Nazmy M, Hassounah M, Alshail E,
nosis, future developments will likely focus on higher doses Moussa E, et al. Intracranial germ cell tumors: a single-institution
of these agents, with stem-cell rescue as required, or on more experience. Ann Saudi Med. 2012;32(4):359–65.
targeted biological agents, possibly based on molecular char- 14. Robertson PL, DaRosso RC, Allen JC. Improved prognosis of
acteristics of the tumor. New modalities such as stereotactic intracranial non-germinoma germ cell tumors with multimodality
therapy. J Neurooncol. 1997;32(1):71–80.
radiosurgery may improve the response rates to adjuvant 15. Modak S, Gardner S, Dunkel IJ, Balmaceda C, Rosenblum MK,
treatment, although it does not appear likely to replace it Miller DC, et al. Thiotepa-based high-dose chemotherapy with
entirely for now. When classifying outcomes, the issue of autologous stem-cell rescue in patients with recurrent or progres-
“splitting” vs. “lumping” remains to be clarified, particularly sive CNS germ cell tumors. J Clin Oncol. 2004;22:1934–43.
16. Sawamura Y, Kato T, Ikeda J, Murata J, Tada M, Shirato H, et al.
in relation to the response to chemotherapy in the pediatric Teratomas of the central nervous system: treatment considerations
and adult forms of these tumors and also in variable responses based on 34 cases. J Neurosurg. 1998;89(5):728–37.
to subtypes of non-germinomatous GCT. 17. Clark AJ, Sughrue ME, Ivan ME, Aranda D, Rutkowski MJ, Kane
Treatment decisions should be made in a multi- AJ, et al. Factors influencing overall survival rates for patients with
pineocytoma. J Neurooncol. 2010;100(2):255–60.
disciplinary context and that given their rarity, centralization, 18. Jackson AS, Plowman PN. Pineal parenchymal tumours:
co-ordination of services, and detailed reporting of outcomes I. Pineocytoma: a tumour responsive to platinum-based chemother-
are key goals to facilitate future improvements in care. apy. Clin Oncol (R Coll Radiol). 2004;16(4):238–43.
19. Sakoda K, Uozumi T, Kawamoto K, Fujioka Y, Hasada J, Hatayama
T, et al. Responses of pineocytoma to radiation therapy and chemo-
therapy—report of two cases. Neurol Med Chir (Tokyo). 1989;
References 29(9):825–9.
20. Marcol W, Kotulska K, Grajkowska W, Gołka D, Właszczuk P,
1. Hasegawa T, Kondziolka D, Hadjipanayis CG, Flickinger JC, Drogosiewicz M, et al. Papillary pineocytoma in child: a case
Lunsford LD, et al. The role of radiosurgery for the treatment of report. Biomed Pap Med Fac Univ Palacky Olomouc Czech Repub.
pineal parenchymal tumors. Neurosurgery. 2002;51:880–9. 2007;151(1):121–3.
2. Haas-Kogan DA, Missett BT, Wara WM, Donaldson SS, Lamborn 21. Nakatsukasa M, Toya S, Otani M, Yoshida K, Ishida Y, Ogawa Y,
KR, Prados MD, et al. Radiation therapy for intracranial germ cell et al. Cystic pineocytoma successfully treated with synchronized
tumors. Int J Radiat Oncol Biol Phys. 2003;56:511–8. chemoradiotherapy. Case report. Neurol Med Chir (Tokyo). 1989;
3. Ishida Y, Sakamoto N, Kamibeppu K, Kakee N, Iwai T, Ozono S, 29(4):333–7.
et al. Late effects and quality of life of childhood cancer survivors: 22. Lutterbach J, Fauchon F, Schild SF, Chang SM, Pagenstecher A,
Part 2. Impact of radiotherapy. Int J Hematol. 2010;92(1):95–104. Volk B, et al. Malignant pineal parenchyma tumors in adult patients:
4. Paximadis P, Hallock A, Bhambhani K, Chu R, Sood S, Wang Z, patterns of care and prognostic factors. Neurosurgery. 2002;51:
et al. Patterns of failure in patients with primary intracranial germi- 44–56.
noma treated with neoadjuvant chemotherapy and radiotherapy. 23. Tate MC, Rutkowski MJ, Parsa AT. Contemporary management of
Pediatr Neurol. 2012;47(3):162–6. pineoblastoma. Neurosurg Clin N Am. 2011;22(3):409–12.
37 Pineal Region Tumors: Viewpoint—Chemotherapy 481
24. Gururangan S, McLaughlin C, Quinn J, Rich J, Reardon D, Halperin 26. Shih CS, Hale GA, Gronewold L, Tong X, Laningham FH, Gilger
EC, et al. High-dose chemotherapy with autologous stem-cell EA, et al. High-dose chemotherapy with autologous stem cell
rescue in children and adults with newly diagnosed pineoblastomas. rescue for children with recurrent malignant brain tumors. Cancer.
J Clin Oncol. 2003;27:2187–91. 2008;112(6):1345–53.
25. Rosenfeld A, Kletzel M, Duerst R, Jacobsohn D, Haut P, Weinstein 27. Gill P, Litzow M, Buckner J, Arndt C, Moynihan T, Christianson T,
J, et al. A phase II prospective study of sequential myeloablative et al. High-dose chemotherapy with autologous stem cell transplan-
chemotherapy with hematopoietic stem cell rescue for the treat- tation in adults with recurrent embryonal tumors of the central
ment of selected high risk and recurrent central nervous system nervous system. Cancer. 2008;112(8):1805–11.
tumors. J Neurooncol. 2010;97(2):247–55.
Skull Base Tumors
38
Reinhart A.J. Sweeney and Matthias Guckenberger
patient function and quality of life, as well as prevention of

Introduction new neurologic deficits. Very few complications occur after
stereotactic radiosurgery, such as perifocal edema, delayed
This chapter will focus on the rarer skull base tumors not intratumoral hemorrhage, or radionecrosis requiring neuro-
discussed in detail in the respective meningioma, pituitary, surgical intervention [1, 2].
and schwannoma chapters. In general, single-fraction stereotactic radiosurgery is
These skull base tumors can be divided into those that are used for small localized lesions under 3 cm in diameter by
unique to a particular region and those that can occur through- using external stereotactic techniques such as Gamma Knife
out the skull base. These tumors may derive from the bone, which requires the applications of an invasive stereotactic
paranasal sinuses, nasopharynx, inner ear, dura, cranial frame attached to the skull with four pins. The same tech-
nerves, or brain. They may be primary tumors invading local nique has also been adapted to linac-based systems and
structures, or metastatic disease. They can also be divided charged-particle therapy such as protons. However, in the
into benign and malignant lesions. last decade, noninvasive alternatives have emerged, using
Unfortunately, even benign lesions can become life noninvasive fixation, most commonly thermoplastic masks
threatening if critically located, rendering them unresectable or a vacuum-mouthpiece [3–5]. This allows temporal separa-
or otherwise not amenable to effective local therapy. tion of the imaging, planning, and treatment sessions as well
Tumors of the skull base pose the greatest challenge to as the option of fractionation should this be required. The
neurosurgeons as they tend to grow around and invade criti- inherent repositioning inaccuracy and intrafraction motion
cal structures such as cranial nerves and vessels, hindering of noninvasive fixation systems are fully compensated by
complete resections as the risk of neurologic deficits and image guidance and, in many cases even six degrees of free-
morbidity is high. Especially here, the advantages of mod- dom treatment couches [6]. As of late, it has become possi-
ern, high-precision radiosurgery with its high rate of local ble to fractionate treatment with the Gamma Knife as well,
control with minimal risk due to optimal sparing of these using the abovementioned vacuum-mouthpiece [5, 7].
critical structures become an attractive alternative. Primary Linac-based radiosurgery can be performed with various
radiosurgery can also be an alternative for inoperable patients techniques, such as the traditional arc-based approach with
or a palliative alternative for those with a poor prognosis. For circular collimators [8], static beams [9]. Dynamic Circular
these, symptom relief such as tumor-related facial pain is Arc (DCA) [10] and intensity-modulated techniques (IMRT/
common. In general however, the goals of radiosurgery are IMSRS) [11]. When circular collimators are used for irregu-
long-term prevention of tumor growth, maintenance of lar treatment volumes (Gamma Knife, but also so equipped
Linacs), multiple isocenters are usually required. These
result in overlapping beams which in turn cause substantial
dose inhomogeneity within the treatment volume. This inho-
R.A.J. Sweeney, M.D., B.Sc. (*) mogeneity may damage organs at risk (OAR), in cases where
Department of Radiation Oncology, MVZ Immed, Klinik Bad these OARs are located within the treated volume. This can
Trissl, Bad-Trissi Str. 76, Oberaudorf, Germany
e-mail: reinhart.sweeney@klinik-bad-trissl.de be of special relevance at the base of skull where cranial
nerves abound [12]. The introduction of (micro-) multileaf
M. Guckenberger, M.D.
Department of Radiation Oncology, University Würzburg, collimation in linac radiosurgery improved conformity and
Josef-Schneider Street 11, 97080 Würzburg, Germany homogeneity [13].
484 R.A.J. Sweeney and M. Guckenberger
Table 38.1 Advantages of different radiosurgical techniques; a literature review

First Author (year) Methods compared Patients (N) PTV (cm3) Results
Bourland et al. (‘94) [9] Noncoplanar circular arc vs. NA NA Smaller high-dose volume and
conformal fixed beam better homogeneity for fixed
beam
Hamilton et al. (‘95) [99] Noncoplanar circular arc vs. 1 3.5 Smaller high-dose volume and
beam
Shiu et al. (‘97) [100] Noncoplanar circular arc vs. 2 4.5–9 Smaller high-dose volume and
beam
Kubo et al. (‘97) [101] Noncoplanar circular arc vs. 11 0.4–17.6 Lower doses to normal tissues
conformal fixed beam and shorter planning time for
fixed beam
Kramer et al. (‘98) [102] Noncoplanar circular arc vs. 1 NA Better CI, lower max. Dose,
IMSRT better HI, but larger penumbra
in IMSRT
Yu et al. (‘99) [14] Noncoplanar circular arc vs. 3 9.6–36.7 IMSRT: better CI, lower dose
IMSRT vs. conformal fixed beam to normal brain than others
Cardinale et al. (‘98) [103] Noncoplanar circular arc vs. 3 11.5 Depending upon shape: IMSRT
conformal fixed beam vs. IMSRT better conformity and OAR-
sparing (irregular) or arcs better
(ellipsoid)
Benedict et al. (‘01) [104] Noncoplanar circular arc vs. 4 2.3–3.5 IMSRS: lower dose to OAR,
conformal fixed beam-IMSRT lower volume normal brain
receiving >50 % of prescr. Dose
Leavitt et al. (‘01) [105] DCA vs. IMSRS/IMSRT 3 8.7–55 Lower dose to OAR in IMSRS/
IMSRT
Yu et al. (‘02) [10] 3D conformal fixed beam vs. 50 Large volumes, no SRS DCA: best conformity, lower
DCA vs. IMSRT dose to normal tissues than
conformal beam but not IMSRT
Perks et al. (‘03) [13] Gamma Knife vs. Conformal 8 0.3–10.6 Gamma Knife: best conformity,
Fixed Beam vs. DCA fixed beam and arc : better
homogeneity
Baumert et al. (‘03) [11] DCA vs. IMSRS 10 15–43 IMSRS: better coverage and
lower OAR dose, but higher
low-dose regions in the normal
brain
Ernst-Stecken et al. (‘05) [106] DCA vs. IMSRS 6 1.1–10.4 IMSRS: lower volume of
normal brain receiving >90 %
of PD but higher integral dose.
RTOG criteria best met by DCA
NA not available, CI conformity index, HI homogeneity index, IMSRT intensity-modulated stereotactic radiotherapy, IMSRS intensity-modulated
stereotactic radiosurgery, OAR organ at risk, DCA dynamic conformal arc
Table 38.1 summarizes the evolution of techniques with

their inherent advantages and disadvantages in the form of a Dose Limitations at the Skull Base
literature review; initially, static beams showed superior con-
formality compared to circular collimator arc treatment [14]. Radiosurgery in the region of the skull base poses special
Perks et al. showed that more beams don’t necessarily trans- challenges, as many critical structures converge there,
late into better normal brain sparing as four to six noncopla- most notably the optic apparatus and cranial nerves.
nar beams yielded similar results as plans using up to 30 Table 38.2 summarizes the current data from literature,
fields [15]. Then followed the next level of conformality in which can assist in determining the feasibility of radiosur-
the form of DCA therapy, where the beam is continually gery itself or a given plan. We cannot emphasize enough
MLC-shaped during gantry rotation [16]. Most recently, that these data are to be used with utmost caution as, with
IMSRS has been shown to offer superior treatment volume exception of the RTOG 90–05 data, they are not the result
(PTV) coverage and lower OAR doses for irregular and con- of dose escalation or randomized trials but stem originate
cave targets [11]. mostly from retrospective clinical outcome data, with not
38 Skull Base Tumors 485
Table 38.2 Summary of radiosurgical tolerance doses at the base of skull

Organ # fx Vol. cm3 Vol. % Vol. limit (Gy) Max. limit (Gy)
Major vessels 1 0.035–10 31–37 37
3 5–10 21–39 45
4 1–10 35–43 49
5 10 47 53
Pituitary gland 1 NA
Brainstem 1 1 10 15
3 1 18 23
5 1 26 31
5 100 20
Brain 5 100 20
Chiasm 1 8–15 10 Gy safe → 77 % chance
of RON above 15 Gy
3 0.2 15
5 0.2 100 20 25
Cranial nerves 1 20–30 Fifth cranial nerve 20 Gy max
Cochlea 1 12
3 20
5 27.5
Lens 1 2–3
2 3–6
3 3–7
5 3–7
Retina and lacrimal gland 1 5
2 5–10
3 5–15
5 5–15
Neurovascular bundle 5 20–50 38
Optic nerve 1 8–15
2 10
3 0.2 15 19.5
5 0.03–0.5 12.5–25 25–30
Area postrema 6.2
Adapted from Grimm J, LaCouture T, Croce R, Yeo I, Zhu Y, Xue J. Dose tolerance limits and dose volume histogram evaluation for stereotactic
body radiotherapy. Journal of Applied Clinical Medical Physics/American College of Medical Physics. 2011; 12(2): 3368; (open source journal)
only limited long-term outcome data but also small patient or less [20] and similar results have been reported by others
collectives. [21]. Considering an effective dose of 13–16 Gy to achieve
From these data, one can surmise that the most critical local control of a given tumor and a recommended dose of
structures are the optic pathways. The risk of clinically sig- 8 Gy as the maximum for the optic chiasm this means that in
nificant radiation optic neuropathy for patients receiving clinical practice a distance between tumor margin and optic
SRS for skull base tumors is 1–2 % following doses to optic apparatus should be at least of 2–3 mm to avoid visual
chiasm below 10 Gy and this percentage may significantly deterioration.
increase for higher doses [17, 18]. Leber et al. reviewed 50 The auditory apparatus is second in line with 12–15 Gy
patients having SRS for benign skull base tumors in which SRS tolerance using a single fraction or 18 Gy in three frac-
the optic nerves or chiasm were exposed to 4.5 Gy or more. tions. Then follows the trigeminal nerve and finally the motor
For patients receiving 10–15 Gy and greater than 15 Gy, the cranial nerves [2, 4, 6, 7, 9–12] which have rarely been
risk of radiation-induced optic neuropathy was 26.7 % and reported having a deficit using doses under 16 Gy.
77.8 %, respectively, however no optic neuropathy was Considering these data and published risks of optic neu-
observed when a dose less than 10 Gy was delivered to the ropathy for conventionally fractionated radiotherapy,
optic apparatus [19]. Stafford et al. found that the risk of α/β[alpha/beta] ratios in the range of 0–1 seem reasonable
developing a clinically significant optic neuropathy was for estimating radiosurgery dose equivalents for optic and
1.1 % for patients receiving a maximum point dose of 12 Gy cranial nerves.
tumors, local control rates of 90–100 % have been reported.

Re-irradiation at the Skull Base On the other end of the spectrum, local control rates for chor-
domas range from 50 to 70 %.
This often represents a relatively high-risk treatment, as criti- On the following pages, the radiosurgical options for
cal tolerance doses have usually been applied in the first primary malignant and benign skull base tumors will be
course of treatment. Historically, most radiation oncologists discussed.
have refused re-irradiation due to concern about the risks of
late central nervous system toxicity, especially radionecro-
sis, which may appear several months to years after treat- Angiofibroma
ment. Re-irradiation of brain tumors has recently attracted
more interest as our understanding of the tolerance of the Disease Pathophysiology
brain to radiation evolves, and developments in radiation
technology and imaging make highly accurate targeting of Juvenile angiofibroma is one of the most common benign
biologically relevant tumor volumes possible. Prospective nasal tumors affecting males between 9 and 19 years of age
data addressing this approach is however lacking. accounting for 0.05 % of all head and neck tumors [26]. In
Obviously, the applicable doses depend to large extent on the USA it is the most common Head and Neck tumor of
time since initial treatment, treatment volume, fractionation adolescence [27]. The tumor originates from the broad area
used and location as well as individual factors such as histol- of the posterolateral wall of the nasal cavity in the region of
ogy, location, and imminent danger to the patient, so that the sphenopalatine foramen [28]. The etiology is still unclear,
individual recommendations are well beyond the scope of however recent electron microscope studies suggest it is
this book. rather a vascular malformation, possibly associated with
However, it seems reasonable to apply the same rationale incomplete regression of the first brachial artery, than a
to re-irradiation in the base of skull regions as is applicable tumor [29]. They often act in a malignant manner by eroding
to the brain itself; own and other published clinical experi- into the surrounding sinuses developing an aggressive growth
ences have yet to describe any major untoward effects of pattern. Intracranial extension is noted in 10–20 % of cases.
such re-irradiation if certain radiobiological principles are Different staging systems based on tumor extension have
followed: been proposed [30].
Although available data come mainly from animal spinal Typical clinical symptoms are frequent epistaxis, nasal
models, the pathogenesis of radiation toxicity and recovery obstructions, and rhinorrhea. Chronic rhinosinusitis, swell-
potential in the brain is assumed to be similar with a similar, ing of the cheek, alteration of olfaction are possible; unilat-
low α/β[alpha/beta] [22]. Animal spinal cord models suggest eral otitis media may result by eustachian tube blockage. By
that significant recovery follows irradiation; conservative eroding into the cranial fossa, diplopia may occur as well as
estimates being up to 50 % recovery within 1–2 years after symptomatic pressure of the chiasm and optic nerves.
initial exposure [23, 24]. An increasing body of evidence is
available (mainly for re-irradiation of gliomas) resulting in a
solid clinical rationale should re-irradiation be required. But Treatment Options
there may also be a case to be made for repeat radiosurgery
in the base of skull region, i.e. for acoustic neuromas [25]. After CT/MRT and bilateral angiography for staging and
In summary, re-irradiation at the skull base remains a determination of blood supply, treatment of choice in
complex procedure and should be left to centers with exten- patients with primary and recurrent juvenile nasopharyn-
sive experience herein. With developments in molecular- geal angiofibroma is surgical resection as sole treatment of
targeted therapy, further exploration of the role of early-stage tumor when gross total resection can be
re-irradiation on its own or in combination with novel agents achieved. Preoperative embolization is recommended by
is needed. most authors to reduce intraoperative blood loss [31]. In the
last decade, endoscopic resections have evolved, providing
reduction of complications and intraoperative bleeding and
Clinical Entities thus an alternative to open surgery for early to intermediate
stage angiofibromas [30]. Complications in advanced-stage
Skull base tumors are relatively rare. Approximately 0.1 % angiofibromas after surgery include intraoperative blood
of all intracranial tumors are chondrosarcomas and also loss requiring transfusions, neuralgia, hearing loss, and
approximately 0.1 % of all intracranial tumors are chordo- ophthalmoplegia [32]. Surgical contraindications include
mas. For benign tumors of the skull base such as glomus unresectable intracranial involvement.
After primary resection of extracranial angiofibromas, Table 38.3 Staging of esthesioneuroblastoma according to Kadish et al.
cure rates of nearly 100 % can be achieved compared with Stage Characteristic
results in patients with intracranial lesions where the cure A Confined to the nasal cavity
rates are approximately 70 %. B Confined to the nasal cavity and one or more
In case of incomplete resection, or in advanced-stage paranasal sinuses
lesions, a combination of surgery followed by radiotherapy C Extending beyond the nasal cavity or paranasal
sinuses, including involvement of the orbit,
is indicated due to the high recurrence rate in these patients.
base of skull or intracranial cavity, cervical
Advanced-stage disease with cranial base involvement and lymph nodes or distant metastatic sides
intracranial extension often allows only subtotal resection of From Kadish S, Goodman M, Wang CC. Olfactory neuroblastoma.
the tumor. A clinical analysis of 17 cases. Cancer. 1976; 37(3): 1571–6; used with
External beam irradiation has been shown to be a useful permission
adjunct to therapy in patients with unresectable recurrent dis-
ease. Gamma Knife and linac-based radiosurgery with a
dose of 20 Gy to the tumor margin (55 % isodose line) is an Rationale for Treatment and Alternatives
effective way to deliver high-dose radiation to incompletely
resected angiofibromas [32] as they represent slow-growing Because of the rarity of esthesioneuroblastoma and its wide
and late-responding tissues. Therefore, a radiobiological variety of clinical behavior, there is no definitive consensus
advantage to radiosurgery may be given. Radiosurgery is regarding the optimal treatment. For small, low-grade tumors
regarded as a reasonable strategy in small-volume and local- confined to the ethmoids, surgery alone appears to be an
ized angiofibromas. adequate method. Patients with locally advanced disease or
For larger angiofibromas, fractionated conformal radiation high-grade tumors should receive aggressive treatment with
therapy with total doses of 36–46 Gy is also effective and combined modalities such as surgery, radiation therapy, and
may reduce the risk of late effects such as cranial nerve defi- chemotherapy [38]. Most authors recommend en bloc resec-
cits, bone and soft tissue necrosis. Another treatment tech- tion, combined with radiation therapy [39, 40]. Local failure
nique described in the literature is the use of intensity-modulated rates of 44 % in low-grade and 60 % in high-grade tumors
radiation therapy (IMRT) in three cases [33]. The applied and metastatic rates of 25 % in low-grade and 47 % in high-
tumor dose varied from 34 to 45 Gy. In all three cases, a grade tumors are described [41]. A significantly lower recur-
reduction of tumor size occurred without significant toxicity. rence rate with overall 5- and 10-year survival rates of 81 and
Especially in children, inhibition of facial bone growth and 54.5 % in patients with response to neoadjuvant radiotherapy
second malignancy are severe possible side effects and must combined with chemotherapy has been reported [42]. At
be considered in treatment decisions [34–36]. recurrence, either surgical and/or radiosurgical retreatment
can lead to long lasting remissions in 42 % of patients [43,
44]. Thus close follow-up is recommended.
Esthesioneuroblastoma
Disease Pathophysiology Radiosurgery for Esthesioneuroblastoma
Esthesioneuroblastomas are rare tumors originating from the Very few studies report on radiosurgical treatment in the pri-
olfactory epithelium of the upper nasal cavity [36]. The sex mary situation; the limiting factors are usually the radiosen-
distribution of esthesioneuroblastomas is uniform. The olfac- sitivity of cranial nerves as discussed in Table 38.1.
tory nerves perforate the groove in the ethmoid bone in the Walch et al. [45] reported on three patients with olfactory
cribriform plate and continue into the subarachnoid spaces. neuroblastoma treated with a combination of endoscopic sur-
Therefore, a high incidence of intracranial extension results. gery and Gamma Knife. Stereotactic radiosurgery was per-
This highly dedifferentiated tumor occurs in all periods of formed within the first 3 months of surgery. The maximum
life with a bimodal peak at the second and sixth decades. diameter of the tumors was approximately 24.3 mm and the
The two most common clinical signs of esthesioneuroblastoma marginal dose to the tumor varied from 16 to 34 Gy; 1–5 iso-
constitute unilateral nasal obstruction and epistaxis. Other centers were used. Radiation-induced side effects were nasal
clinical symptoms include headache, swelling of the cheek, discharge and crusts. One patient developed bilateral frontal
blurred vision, and dental pain. Esthesioneuroblastoma can chronic sinusitis, and in a second, endoscopic operation was
metastasize to regional lymph nodes, primarily of the neck necessary.
[37] lung, or bones. According to the WHO classification An Austrian group described the combined treatment
system, the terms olfactory neuroblastoma and olfactory of endoscopic surgery and radiosurgery for olfactory
neurogenic tumors are used. The Kadish staging classifica- neuroblastoma [46]. Median marginal doses ranged from
tion is shown in Table 38.3. 15 to 34 Gy at a marginal isodose between 45 and 85 %.
The maximum tumor volume treated with radiosurgery was within the third ventricle in 55–85 % of the patients [49].
approximately 20 cm3. The median follow-up period was 58 Compression of the pituitary and hypothalamic region can
months. Observed radiosurgical side effects were mild and produce antidiuretic hormone and growth hormone deficiency
transient, such as cephalea and dizziness. No changes in or obesity in children. Diabetes insipidus is present in approx-
mental status were observed. No new pathology of the optic imately 10 % of the patients. Visual field defects and decreased
pathway was described during follow-up. vision due to compression of the optic chiasm and optic path-
IMRT is recommended for larger and more complex- ways are the initial symptoms in approximately 40–60 % of
shaped Esthesioneuroblastoma [40]. However also here, the these patients [50, 51] (Table 38.4).
required conventionally fractionated doses of 50–60 Gy
(postoperatively) or 65–70 Gy required for inoperable cases,
the challenge remains similar [47, 48]. Rationale for Treatment and Alternatives
The main treatment modality for craniopharyngiomas is

Craniopharyngioma surgery. Microsurgery allows complete tumor removal in
49–100 % of the patients with low morbidity and operative
Disease Pathophysiology mortality [52–54]. After radical resection, 10-year
progression-free survival rates between 60 and 93 % are
Craniopharyngiomas are benign tumors located at the base of reported [55, 56]. Treatment modalities include complete
the skull next to the pituitary gland. Differentiation resection of the tumor with radiation therapy at the time of
between craniopharyngioma and pituitary can therefore recurrence or subtotal resection followed by radiotherapy.
sometimes be difficult on CT or magnetic resonance (MR) The probability of complete tumor resection decreases
scans. Approximately 5–10 % of primary brain tumors are with increasing tumor volume. Because of the proximity to
craniopharyngiomas. They typically occur in childhood as critical normal structures and the relatively high associa-
well as in the sixth to eighth decades [29]. Histopathologically, tion of radical surgery with visual loss and impaired hor-
craniopharyngiomas are benign tumors arising from squa- mone function requiring replacement therapy, many
mous cell remnants of the Rathke pouch during embryo- authors recommend less radical surgery (partial resection,
genesis at the junction of the pituitary stalk and pituitary. biopsy, and aspiration of cystic contents) followed by radi-
Craniopharyngiomas present as a suprasellar lesion, fre- ation therapy or radiosurgery. With this strategy, local con-
quently partially calcified and usually including an intrasellar trol rates of 70–83 % after 10 years are reported [57, 58]
component. These tumors are often composed of solid and and assumed to be similar to complete surgical resection of
cholesterol-rich cystic components. Cystic or solid compo- the tumor [59]. Treatment-related toxicities after subtotal
nents of this tumor extension may occur laterally into the resection followed by radiotherapy include impairment of
middle or into the posterior cranial fossa. Symptoms relate to hormone function. Impairment of vision is reported for
compression effects of the tumor due to its vicinity to pitu- less than 10 % of all patients treated with the combination
itary gland, chiasm, optic nerves, and hypothalamic region. of subtotal resection and irradiation compared with up to
Locally, these tumors can produce signs and symptoms of 20 % after complete tumor resection [60]. Other side
increased intracranial pressure such as headache, drowsiness, effects such as radionecrosis, radiation-induced malignan-
or vomiting at the time of diagnosis and are due to hydro- cies, vascular morbidity, and cognitive decline occur less
cephalus by obstruction of the foramen Monro by tumor parts frequently [60, 61].
Table 38.4 Review of the literature for radiosurgery of craniopharyngioma

Mean peripheral
Author N Median follow-up (range) dose (range) Local control (%)a Morbidity
Chung et al. [64] 31 33 months (5–69)b 12.2 Gy (9.5–16) 87.2 Visual field deficit (1 patient)
Mokry [65] 23 23 months (6–57) 10.8 Gy (8–15) 78.2 None
Ulfarsson et al. [67] 21 42 months (6–348)b 30 Gy (20–50)c 36.4 Visual field deficit (8 patients)
Kobayashi et al. [68] 98 66 months (6–148) 11.5 Gy (NA) 79.5 Visual/endocrine (6 %)
Iwata [62] 44 40 (12–92) 13–25 Gy (1–5 85 Hypopituitarism (1 patient)
fractions)
a
Crude local control rates
b
Median values
c
Given in maximum dose
The major goal of radiotherapy treatment strategies is deterioration related to radiosurgery. In a Swedish study,
sparing of critical normal structures. Radiosurgery as well as 21 patients were treated with Gamma Knife radiosurgery.
intracavitary irradiation with stereotactically applied β[beta]- They found a statistically significant difference between
emitting radioisotopes maximize normal tissue sparing. The tumor progression and applied dose; a higher progression
cystic nature of craniopharyngioma has led to trials of intra- rate was found in patients treated with less than 6 Gy to the
cystic applications of β[beta]-emitting radioisotopes such as margin than in patients treated with a dose higher than 6 Gy.
yttrium-90 or phosphorus-32. The use of radiosurgery has Four of these patients developed pituitary dysfunction [67].
been reported in patients with minimal residual or recurrent In the literature, parenchymal injuries of the brain or second
disease. However, for patients with larger target volumes, malignancies caused by radiotherapy are estimated to be
tumors immediately abutting the optic apparatus and multi- less than 1–2 % [68].
ple cystic configured lesions, fractionated stereotactic radio-
therapy should be preferred as excellent local control with
minimal morbidity can be realized [62, 63]. Chordomas, Chondromas,
and Chondrosarcomas
Treatment Planning and Results Disease Pathophysiology

for Radiosurgery
Chondromas are rare benign tumors arising at the base of the
The target volume for craniopharyngiomas is narrowly skull, especially in the area close to the pituitary gland. It is
defined to the tumor volume, including solid and cystic com- a very-slow-growing tumor and might be present for a long
ponents. In cases with cyst aspiration or subtotal resection, it time before causing any symptoms. Chondromas are com-
is important to cover the complete cyst wall. This technique posed of cartilage formed by the meninges and is usually
can be used for selected patients with smaller tumors (<2 cm) attached to the dura mater. Surgical intervention might be the
not abutting critical structures such as the chiasm and the treatment of primary choice because of their usually well-
brain stem. Median doses to the margin of the tumor range defined margins.
from 9 to 16 Gy. Chung et al. recommend a margin dose of Chordomas are relatively rare, slow-growing, primary
12 Gy to induce satisfactory tumor response [64]. The main bone tumors arising from embryonic remnants of the noto-
restriction with radiosurgery treatment is the tolerance dose chord (chorda dorsalis) at the two extreme ends of the verte-
of the neighboring visual pathway. The dose to the optic bral axis. They are most often diagnosed in the second or
nerves and the chiasm should be kept below 8 Gy with third decade of life, more common in males (2:1) and com-
single-dose techniques to avoid damage to these structures. prise less than 1 % of intracranial tumors [69, 70]. Twenty
Stereotactic radiosurgery has been used to treat small resid- five to forty percent of chordomas occur in the spheno-
ual or recurrent tumors after surgical intervention. occipital or skull base region. The clivus is the most common
Mokry et al. [65] treated 23 patients with Gamma Knife site. Chordomas are locally more aggressive with a poorer
radiosurgery for craniopharyngioma and found no relevant outcome compared with chondrosarcomas.
morbidity. Ten patients had additional therapy with intracys- Chondrosarcomas are malignant tumors composed of
tic bleomycin before radiosurgery. Tumor progression was cartilage-producing cells encountered in the skull base. Two
observed in 5 of 23 patients. They conclude that the best histologic variants of chordoma have been described. The
results might be obtained in monocystic tumors amenable to first is chondroid chordoma, a typical chordoma that also
stereotactic drainage and intracystic bleomycin treatment. contains areas resembling low-grade hyaline chondrosar-
The Cologne group of Kickingrieder et al. summarizes their coma. The second variant is dedifferentiated chordoma,
results on 53 patients with cystic craniopharyngiomas treated which contains areas of typical chordoma mixed with com-
with stereotactically applied colloidal β-emitting radioactive ponents that resemble high-grade or poorly differentiated
sources. They concede few but notable severe side effects spindle cell sarcoma. Chondrosarcomas rarely metastasize,
(hemiparesis and third nerve palsy) as well as suboptimal are slow-growing but often invade local structures. Prognostic
progression-free survival (79.4 ± 6.1, 72.4 ± 6.8, and factors that most influence choice of treatment are location,
45.6 ± 8.7 % at 12, 24, and 60 months, respectively) [66]. local tumor extension, and surgical resectability.
After an average follow-up of 36 months, Chung et al. Approximately 0.1 % of all intracranial tumors are chon-
reported a tumor control rate of 87 % for 31 patients treated drosarcomas. This locally invasive tumor is a malignant vari-
with Gamma Knife radiosurgery and a prescribed dose to ant of a benign chondroma arising from bone and is composed
the tumor margin from 9.5 to 16 Gy [64]. One patient devel- of cartilage.
oped a mildly restricted visual field. None of the patients Chondrosarcomas are mostly located in the sphenoid
showed additional endocrinologic impairment or neurologic bone or clivus. Chondrosarcomas are also more common in
Table 38.5 Radiosurgical literature on chordoma and chondrosarcoma

Median
Median follow-up peripheral dose
Author N (range) in Gy (range) Local control* (%) Morbidity Comments
Krishnan et al. [107] 4.8 years (0.8–11.4) 15 (10–20)
Chordoma 25 32‡ 34 % (all with
combined EBRT)
Chondrosarcoma 4 100
Feigl et al. [82] 17 months (6–36)† 17 (14–18)†
Chordoma 3 33 Cranial nerve deficits,
headaches, diplopia
Chondrosarcoma 10 100
Pamir et al. [89] 23.3 months (NA)† NA 29 NA
Chordoma 7
Chang et al. [108]§ 4 years (1–9)† 19.4 (18–24)†
Chordoma 10 80‡ None
Hauptman et al. [85] 5 4.5 years 15.5 (to 90 % 60 Cranial neuropathy,
isodose line) visual deficits
Koga et al. [84] 14 65 months (12–167) 15 (10–20) Local relapse in all cases
of Radiosurgery after
fractionated xRT, Marginal
doses of >16 Gy crucial
Chordoma 10 40.5 (12–167) months 13.7 (10–20) 15 1 transient visual
deficit
Chondrosarcoma 4 20 (45–145) months 15.5 (12–20) 100 1 transient visual
deficit
Martin et al. [109] 7.7 years (2–17) 16.5 (10–25)
Chordoma 18 88 months 62.9‡ 1 transient effect
Chondrosarcoma 10 88 months 80‡
Hasegawa et al. [110] 59 months (1–172) 14 (9–20) 80‡ 1 worsening of facial Under 20 mL volume sig.
numbness better LC, at least 15 Gy
marginal dose required
Chordoma 30 0.72
Chondrosarcoma 7
males than in females (1.5:1). They can be classified into and mortality. Furthermore, it is difficult to treat skull base
three grades (I–III). Lower-grade tumors are less aggressive chordomas and chondrosarcomas with radiotherapy alone
and act clinically similar to chordomas. Historically, skull because of the large tumor size, the extent of infiltrated tis-
base chordoma and chondrosarcoma were often pooled sues, and because of dose limitations imposed by the sensi-
together in reported series due to the rarity of these tumors; tivity of adjacent critical normal tissues such as the brain
however, recently published studies have shown important stem and cranial nerves [73]. Frequently used fractionated
differences with respect to diagnosis, treatment, and progno- doses are 55–66 Gy. A clear dose-response relationship with
sis, strongly suggesting a more aggressive therapy for chor- improved outcomes after doses exceeding 60 Gy have been
domas than chondrosarcomas [71] (Table 38.5). shown in chordomas [74]. Similar effects have been shown
with radiosurgery with better results after marginal doses
>15 Gy [73]. Because of the slow proliferative nature of
Rationale for Treatment and Alternatives chordomas, high linear energy transfer may be useful. To
yield good results with respect to local tumor control, high-
Treatment of choice is total surgical resection, if feasible, dose radiation therapy with photons combined with proton
followed by radiation therapy. The best results in the treat- beam boost [75, 76] or particle therapy alone [77, 78], are
ment of chordomas have been obtained by complete surgi- reported. Heavily charged particles such as protons or car-
cal resection followed by high doses of proton irradiation bon ions may be superior because of their finite range in
[72]. Complete resection of these tumors at the skull base is tissues. This radiation technique offers an excellent chance
challenging and associated with higher rates of morbidity of cure with acceptable radiation-induced toxicity. However,
IMRT has also shown its effectiveness [79], so that the final Chordomas
verdict may ultimately require long-term results from a ran-
domized trial.
Skull base approaches
Radiosurgery for Chordomas Surgical intervention

and Chondrosarcomas
No residual tumors Residual tumors
There is a strong argument for radiosurgical treatment of
chordomas and chondrosarcomas to reduce the incidence of
brain or bone necrosis in smaller circumscribed post-surgical
tumor-remnants. In literature, lesions with a diameter of less Follow-up <30 CC >30 CC
than 30 mm were treated with 17–20 Gy to the tumor margin
[73, 80, 81].
Kondziolka et al. [73] reported on four patients with chor- Radiosurgery as Radiosurgery Consider
doma and two patients with chondrosarcoma treated with soon as recurrence Re-operation
radiosurgery. All tumors were less than 30 mm in diameter detected
and were treated with 20 Gy to the tumor margin. During a
Fig. 38.1 Radiosurgical management algorithm for skull base chordomas
mean follow-up of 22 months (range, 8–36 months), no in- according to Pamir et al. [83]. Should re-operation not be possible,
field progression was seen. Three patients showed improve- high-dose particle beam therapy or IMRT should be recommended
ment of preexisting neurologic deficits. The other three
patients remained in stable neurologic condition. Serial fol-
low-up imaging studies showed tumor volume reduction in Glomus Jugulare Tumors/Chemodectoma/
two patients, whereas the other four patients showed stable Paraganglioma
tumor size. One patient showed tumor progression outside
the irradiated tumor volume. In the study of Feigl et al., 13 Disease Pathophysiology
patients with chordoma and chondrosarcoma were treated
with Gamma Knife radiosurgery after maximal tumor resec- Glomus jugulare tumors are rare, radiosensitive tumors of
tion [82]. The mean treated tumor volume was 9.7 cm3 (range the skull base/neck regions, slow-growing, hypervascular,
1.4–20.3 cm3). The mean treatment dose was 17 Gy and the and histologically benign. They comprise 0.6 % of all tumors
mean marginal isodose was 52 %. After a mean follow-up of and are closely associated with the sympathetic system, aris-
17 months, only one recurrence of disease was seen at the ing from the paraganglia of the chemoreceptor system. Thus,
margin of the radiation field. Pamir et al. reported on 26 skull chemodectoma and paraganglioma are frequently used syn-
base chordomas with a mean follow-up period of 48.5 months onyms. Due of their tendency to invade and compress adja-
receiving multimodality treatment with various combina- cent tissues, local problems can evolve.
tions of conventional surgery, skull base surgical techniques, Typical clinical symptoms are gradual hearing loss, uni-
and Gamma Knife surgery [83]. The mean follow-up after lateral pulsatile tinnitus, or imbalance.
Gamma Knife treatment was 23.3 months. They recommend Familial as well as multilocal/bilateral occurrence is pos-
Gamma Knife radiosurgery immediately after initial surgical sible and should be excluded by MRI of the skull base and
intervention if the tumor volume is less than 30 cm3. This neck (Fig. 38.2).
suggested treatment algorithm of Pamir and colleagues is These tumors can show metabolic activity on PET/SPECT
shown in Fig. 38.1. which may change disease management [85]. Malignant
Interesting results were published by the Tokyo group by transformation rates are rare (2–5 %) with metastatic spread
Koga et al. [84] on ten chordoma and four chondrosarcoma to lung, liver, and bone [86, 87]. Because of their slow
patients with tumor volumes at radiosurgery ranging from growth, up to 10 years of follow-up is necessary to establish
3.4 to 55.5 cm3. Four chordoma patients received radiosur- a cure rate for these lesions [88].
gery with lower marginal doses (12 Gy) after 60 Gy fraction-
ated radiotherapy. All four patients recurred locally. One
chondrosarcoma patient receiving 12 Gy marginal dose Rationale for Treatment and Alternatives
recurred locally. Five-year PFS rates for patients with higher
and lower marginal doses were 80 % and 14 % respectively The treatment options for glomus jugulare tumors include sur-
(p = 0.0005). The authors conclude that a marginal dose of at gery, radiosurgery, radiotherapy, and endovascular occlusion
least 16 Gy is required for local control. of feeding vessels either in combination or alone [89–91].
Fig. 38.2 This right sided incidental jugular glomus tumor was sided glomus tumor at the level of the carotid bifurcation.
detected on 99mTc-octreotide SPECT imaging (bottom in fusion with Retrospectively, it is quite obvious in MRI
MRI) after being missed on both CT (top) and MRI (middle) for a left
Embolization alone does not prevent further tumor progression are based on the patients’ comorbidities. Surgical resection of
[92]. Surgical resection is the only treatment option that glomus jugulare tumors carries a high complication rate,
can offer immediate and complete tumor elimination. due to their high vascularity and the involvement of critical
Contraindications to surgical intervention of skull base tumors vascular and neuronal structures, which included stroke,
cranial nerve injury with 8–40 %, and an overall mortality rate of the tumor. The steep dose gradient achievable with radio-
of 5–13 % [93]. Compared with radiosurgery, fractionated ste- surgery minimizes the irradiation dose to surrounding nor-
reotactic radiotherapy may reduce the risk of radiation-induced mal tissue. Whereas diagnostic imaging techniques have
side effects providing additional radiobiologic sparing and been much improved within recent years, they cannot reli-
should be recommended for larger tumors or those, which ably separate tumor from adjacent cranial nerve when target-
cannot be precisely defined by imaging. ing radiosurgery treatment. Because of the close proximity
of glomus tumors to cranial nerves, permanent cranial nerve
deficits are possible side effects of radiosurgery. Most pub-
Patient Selection for Radiosurgery lished studies reported only transient dizziness or occasional
and Treatment Planning Details hearing loss (Table 38.6).
Jordan et al. [82] reported on eight patients treated with
For radiosurgical techniques, the prescribed dose to the Gamma Knife radiosurgery between 1990 and 1998. The
tumor margin ranges from 12 to 25 Gy with typical doses mean tumor volume of these patients was 9.8 cm3 (range,
greater than 20 Gy for small- to medium-sized tumors but 17.3–4.3 cm3). The mean applied marginal dose was 16.3 Gy
lower doses to larger tumors due to an increased risk of (range, 12–20 Gy). None of these patients developed delayed
radiation-induced side effects [82, 83, 94]. Median tumor cranial neuropathy or tumor progression during a mean fol-
volumes of glomus tumors should be less than 10 cm3 low-up of 27 months.
because of the possible increased risk of radiation-induced Lim et al. [94] reported on ten patients treated with
cranial nerve deficits due to the radiosensitivity of cranial radiosurgery, whereas four patients were treated with pri-
nerves. On the other hand and in analogy to excellent control mary radiosurgery due to several comorbidities. The other
rates after only 45–50 Gy in fractionated radiotherapy [95] five had prior surgeries for their tumor. Tumor size ranged
12 Gy single fraction or 25 in 5 Gy fractions [96] likely suf- from 1.2 to 3.6 cm at the largest diameter with an average
fice, especially in critical locations. of 2.4 cm. Six patients were treated with a frame-based
linac system and four with the CyberKnife. Prescribed
dose to the 80 % isodose ranged from 16 to 25 Gy to the
Review of the Literature tumor margin. After a median follow-up of 21.5 months,
nine patients had no change of tumor size, whereas one
For well-defined and noninfiltrating glomus jugulare tumors, patient showed tumor regression. Nine patients had stable
stereotactic radiotherapy should be particularly beneficial. neurologic symptoms and only one patient experienced
They usually present in a small size due to their proximity to transient ipsilateral tongue weakness and hearing loss
cranial nerves whose dysfunctions often herald the presence (Fig.38.3).
Table 38.6 Side effects of radiosurgery

Median peripheral
Author N Median follow-up (range) dose (range) Local controla Morbidity Comments
Jordan et al. [82] 8 27 months (9.7–102)b 16.3 Gy (12–20)b 1 Acute vertigo (1 patient)
Foote et al. [89] 25 37 months (11–118) 15 Gy (12–18) 1 Late vertigo (1 patient) 8 decreased in size
Eustacchio et al. 19 7 years (1.5–10) 14 Gy (12–20) 0.95 None
[111]
Maarouf et al. [83] 12 4 years (0.8–9) 15 Gy (11–20) 1 Moderate facial palsy (1 8 decreased in size
patient)
Liscak et al. [88] 52 24 months (4–70) 16.5 Gy (10–30) 1 Tinnitus (2 patients) Tumor size
decreased in 40 %
Gottfried et al. 142 39.4 monthsb 0.98 8.5 % morbidity
c
[112]
Pollock et al. [113] 42 44 (6–149 months) 14.9b( 0.98 15 % new deficits (4× hearing
loss, 2× vocal cord paralysis,
temporary imbalance/vertigo ×1
Hurmuz et al. [96] 14 39 months (7–60) 25 (18–30) (in 1 None Tumor regression
1–5 fractions) in 6
a
Crude local control rates
b
Mean values
c
Literature review
Fig. 38.3 Treatment plan of a patient presenting with hearing deficit, vertigo and vocal cord paresis. MR imaging showed a left sided glomus
tumor treated radiosurgically with 17 Gy marginal dose (70 % isodose). After 24 months the tumor has shrunk, no new toxicity (courtesy of Dr.
A. Muacevic, CyberKnife Centre Munich)
After an observation period of up to 6.7 years, Saringer neurologic toxicity was identified in all 25 patients. Only one
and co-workers [90] reported on 13 patients with glomus patient experienced clinically significant vertigo 8.5 months
tumors treated with Gamma Knife radiosurgery. Three of after treatment. They observed no new or progressive neu-
these patients showed tumor size reduction, whereas ten had ropathy of cranial nerves V–XII.
stable tumor volume. Clinical symptoms remained Recently, even clearer evidence for radiosurgery’s role
unchanged in six, and six patients showed an improvement has been published in the form of a meta-analysis of 869
of clinical symptoms. One patient failed the follow-up. Two patients with glomus tumors. Subtotal resection (STR), gross
patients developed transient cranial nerve complications total resection (GTR), STR + postoperative radiosurgery and
(worsening of preexisting swallowing disorders and tempo- radiosurgery alone were compared in terms of local control
rary facial nerve palsy 1 and 12 months after radiosurgery, and cranial neuropathy. Local control after radiosurgery was
respectively). significantly (p < 0.01) higher than the other groups. The
The Mayo Clinic experiences were published by Foote GTR patient group had higher rates of deficits than those
et al. in 2002 [97]. They described treatment outcome for a who underwent SRS alone. The authors concluded that
total of 25 patients and long-term results for a cohort of 9 higher rates of morbidity are not associated with improved
patients with a median largest diameter of 3.3 cm. No acute local control rates, compared to radiosurgery [98].
high precision radiotherapy; first phantom and clinical results.

Conclusion Radiat Oncol. 2010;5:42.
7. Sayer FT, Sherman JH, Yen CP, Schlesinger DJ, Kersh R, Sheehan
JP. Initial experience with the eXtend System: a relocatable frame
It can be generalized in the base of skull region as well, that system for multiple-session gamma knife radiosurgery. World
when radiosurgery is limited to lesions that are 3 cm or less, Neurosurg. 2011;75(5–6):665–72.
dose fall-off is sharp and only a small amount of normal 8. Colombo F, Benedetti A, Pozza F, Zanardo A, Avanzo RC,
Chierego G, et al. Stereotactic radiosurgery utilizing a linear
brain tissue receives a high radiation dose. At larger volumes, accelerator. Appl Neurophysiol. 1985;48(1–6):133–45.
the radiation fall-off into the surrounding normal tissue is not 9. Bourland JD, McCollough KP. Static field conformal stereotactic
as steep and the risk of delayed radiation-induced side effects radiosurgery: physical techniques. Int J Radiat Oncol Biol Phys.
increases. Still, many patients with skull base tumors larger 1994;28(2):471–9.
10. Yu CX, Li XA, Ma L, Chen D, Naqvi S, Shepard D, et al. Clinical
than 30 mm can undergo radiosurgery safely because these implementation of intensity-modulated arc therapy. Int J Radiat
tumors are often in contact with the brain for only a portion Oncol Biol Phys. 2002;53(2):453–63.
of their surface, and therefore, radiation fall-off occurs in the 11. Baumert BG, Norton IA, Davis JB. Intensity-modulated stereotac-
bone, the sinuses, or the infratemporal or cervical regions. tic radiotherapy vs stereotactic conformal radiotherapy for the
treatment of meningioma located predominantly in the skull base.
For larger lesions and those located next to the optic path- Int J Radiat Oncol Biol Phys. 2003;57(2):580–92.
way, fractionated radiation therapy should be recommended 12. Plowman PN. Post-radiation sensorineuronal hearing loss. Int J
due to its radiobiological advantages. Radiat Oncol Biol Phys. 2002;52(3):589–91.
However, in order to maximize radiosurgical benefits and 13. Perks JR, St George EJ, El Hamri K, Blackburn P, Plowman
PN. Stereotactic radiosurgery XVI: isodosimetric comparison of
minimize its dangers, all highly precise radiation techniques photon stereotactic radiosurgery techniques (gamma knife vs.
must be preceded by the integration of optimal imaging (see micromultileaf collimator linear accelerator) for acoustic neu-
Chap. 2) into the treatment planning process. This is espe- roma—and potential clinical importance. Int J Radiat Oncol Biol
cially challenging and critical in the base of skull region. Phys. 2003;57(5):1450–9.
14. Yu C, Luxton G, Jozsef G, Apuzzo ML, Petrovich Z. Dosimetric
Whether chordomas and chondrosarcomas are better comparison of three photon radiosurgery techniques for an elon-
treated with particles such as protons and/or ions is awaiting gated ellipsoid target. Int J Radiat Oncol Biol Phys. 1999;45(3):
clarification, ideally in the form of randomized studies. 817–26.
Radiosurgery should be the considered treatment of choice 15. Perks JR, Jalali R, Cosgrove VP, Adams EJ, Shepherd SF,
Warrington AP, et al. Optimization of stereotactically-guided con-
for smaller and well-defined glomus tumors due to minimal formal treatment planning of sellar and parasellar tumors, based
morbidity compared to surgery as well as shorter treatment on normal brain dose volume histograms. Int J Radiat Oncol Biol
times and better normal tissue sparing compared to fraction- Phys. 1999;45(2):507–13.
ated radiotherapy. 16. Leavitt DD, Gibbs Jr FA, Heilbrun MP, Moeller JH, Takach Jr
GA. Dynamic field shaping to optimize stereotactic radiosurgery.
Int J Radiat Oncol Biol Phys. 1991;21(5):1247–55.
17. Kurita H, Sasaki T, Kawamoto S, Taniguchi M, Terahara A, Tago
References M, et al. Role of radiosurgery in the management of cavernous
sinus meningiomas. Acta Neurol Scand. 1997;96(5):297–304.
1. Pirzkall A, Debus J, Lohr F, Fuss M, Rhein B, Engenhart-Cabillic 18. Colombo F, Casentini L, Cavedon C, Scalchi P, Cora S, Francescon
R, et al. Radiosurgery alone or in combination with whole-brain P. Cyberknife radiosurgery for benign meningiomas: short-term
radiotherapy for brain metastases. J Clin Oncol. results in 199 patients. Neurosurgery. 2009;64(2 Suppl):A7–13.
1998;16(11):3563–9. 19. Leber KA, Bergloff J, Pendl G. Dose-response tolerance of the
2. Shaw E, Scott C, Souhami L, Dinapoli R, Kline R, Loeffler J, et al. visual pathways and cranial nerves of the cavernous sinus to ste-
Single dose radiosurgical treatment of recurrent previously irradi- reotactic radiosurgery. J Neurosurg. 1998;88(1):43–50.
ated primary brain tumors and brain metastases: final report of 20. Stafford SL, Pollock BE, Leavitt JA, Foote RL, Brown PD, Link
RTOG protocol 90-05. Int J Radiat Oncol Biol Phys. 2000;47(2): MJ, et al. A study on the radiation tolerance of the optic nerves
291–8. and chiasm after stereotactic radiosurgery. Int J Radiat Oncol Biol
3. Ramakrishna N, Rosca F, Friesen S, Tezcanli E, Zygmanszki P, Phys. 2003;55(5):1177–81.
Hacker F. A clinical comparison of patient setup and intra-fraction 21. Morita A, Coffey RJ, Foote RL, Schiff D, Gorman D. Risk of
motion using frame-based radiosurgery versus a frameless image- injury to cranial nerves after gamma knife radiosurgery for skull
guided radiosurgery system for intracranial lesions. Radiother base meningiomas: experience in 88 patients. J Neurosurg.
Oncol. 2010;95(1):109–15. 1999;90(1):42–9.
4. Guckenberger M, Roesch J, Baier K, Sweeney RA, Flentje 22. Withers HR. Biologic basis for altered fractionation schemes.
M. Dosimetric consequences of translational and rotational errors Cancer. 1985;55(9 Suppl):2086–95.
in frame-less image-guided radiosurgery. Radiat Oncol. 2012;7:63. 23. Ang KK, Price RE, Stephens LC, Jiang GL, Feng Y, Schultheiss
5. Ruschin M, Nayebi N, Carlsson P, Brown K, Tamerou M, Li W, TE, et al. The tolerance of primate spinal cord to re-irradiation. Int
et al. Performance of a novel repositioning head frame for gamma J Radiat Oncol Biol Phys. 1993;25(3):459–64.
knife perfexion and image-guided linac-based intracranial stereo- 24. Wong CS, Hao Y. Long-term recovery kinetics of radiation damage
tactic radiotherapy. Int J Radiat Oncol Biol Phys. 2010;78(1): in rat spinal cord. Int J Radiat Oncol Biol Phys. 1997;37(1):171–9.
306–13. 25. Kano H, Kondziolka D, Niranjan A, Flannery TJ, Flickinger JC,
6. Wilbert J, Guckenberger M, Polat B, Sauer O, Vogele M, Flentje M, Lunsford LD. Repeat stereotactic radiosurgery for acoustic neuro-
et al. Semi-robotic 6 degree of freedom positioning for intracranial mas. Int J Radiat Oncol Biol Phys. 2010;76(2):520–7.
26. Lund VJ, Stammberger H, Nicolai P, Castelnuovo P, Beal T, 46. Unger F, Haselsberger K, Walch C, Stammberger H, Papaefthymiou
Beham A, et al. European position paper on endoscopic manage- G. Combined endoscopic surgery and radiosurgery as treatment
ment of tumours of the nose, paranasal sinuses and skull base. modality for olfactory neuroblastoma (esthesioneuroblastoma).
Rhinol Suppl. 2010;22:1–143. Acta Neurochir (Wien). 2005;147(6):595–601; discussion 2.
27. Chandler JR, Goulding R, Moskowitz L, Quencer 47. Chao KS, Kaplan C, Simpson JR, Haughey B, Spector GJ,
RM. Nasopharyngeal angiofibromas: staging and management. Sessions DG, et al. Esthesioneuroblastoma: the impact of treat-
Ann Otol Rhinol Laryngol. 1984;93(4 Pt 1):322–9. ment modality. Head Neck. 2001;23(9):749–57.
28. Neel 3rd HB, Whicker JH, Devine KD, Weiland LH. Juvenile 48. Duprez F, Madani I, Morbee L, Bonte K, Deron P, Domjan V,
angiofibroma. Review of 120 cases. Am J Surg. 1973;126(4): et al. IMRT for sinonasal tumors minimizes severe late ocular tox-
547–56. icity and preserves disease control and survival. Int J Radiat Oncol
29. Schick B, Plinkert PK, Prescher A. [Aetiology of angiofibromas: Biol Phys. 2012;83(1):252–9.
reflection on their specific vascular component]. Laryngorhinoo- 49. Rutka JT, Hoffman HJ, Drake JM, Humphreys RP. Suprasellar and
tologie. 2002;81(4):280–4. sellar tumors in childhood and adolescence. Neurosurg Clin N
30. Nicolai P, Schreiber A, Bolzoni Villaret A. Juvenile angiofibroma: Am. 1992;3(4):803–20.
evolution of management. Int J Pediatr. 2012;2012:412545. 50. Fisher PG, Jenab J, Gopldthwaite PT, Tihan T, Wharam MD, Foer
31. Midilli R, Karci B, Akyildiz S. Juvenile nasopharyngeal angiofi- DR, et al. Outcomes and failure patterns in childhood craniopha-
broma: analysis of 42 cases and important aspects of endoscopic ryngiomas. Childs Nerv Syst. 1998;14(10):558–63.
approach. Int J Pediatr Otorhinolaryngol. 2009;73(3):401–8. 51. Abrams LS, Repka MX. Visual outcome of craniopharyngioma in
32. Jafek BW, Krekorian EA, Kirsch WM, Wood RP. Juvenile naso- children. J Pediatr Ophthalmol Strabismus. 1997;34(4):223–8.
pharyngeal angiofibroma: management of intracranial extension. 52. Fahlbusch R, Honegger J, Paulus W, Huk W, Buchfelder
Head Neck Surg. 1979;2(2):119–28. M. Surgical treatment of craniopharyngiomas: experience with
33. Kuppersmith RB, Teh BS, Donovan DT, Mai WY, Chiu JK, Woo 168 patients. J Neurosurg. 1999;90(2):237–50.
SY, et al. The use of intensity modulated radiotherapy for the treat- 53. Samii M, Bini W. Surgical treatment of craniopharyngiomas.
ment of extensive and recurrent juvenile angiofibroma. Int J Zentralbl Neurochir. 1991;52(1):17–23.
Pediatr Otorhinolaryngol. 2000;52(3):261–8. 54. Yasargil MG, Curcic M, Kis M, Siegenthaler G, Teddy PJ, Roth
34. Cummings BJ, Blend R, Keane T, Fitzpatrick P, Beale F, Clark R, P. Total removal of craniopharyngiomas. Approaches and long-
et al. Primary radiation therapy for juvenile nasopharyngeal term results in 144 patients. J Neurosurg. 1990;73(1):3–11.
angiofibroma. Laryngoscope. 1984;94(12 Pt 1):1599–605. 55. Hoffman HJ, De Silva M, Humphreys RP, Drake JM, Smith ML,
35. Reddy KA, Mendenhall WM, Amdur RJ, Stringer SP, Cassisi Blaser SI. Aggressive surgical management of craniopharyngio-
NJ. Long-term results of radiation therapy for juvenile nasopha- mas in children. J Neurosurg. 1992;76(1):47–52.
ryngeal angiofibroma. Am J Otolaryngol. 2001;22(3):172–5. 56. Kalapurakal JA, Goldman S, Hsieh YC, Tomita T, Marymont
36. Takeda A, Shigematsu N, Suzuki S, Fujii M, Kawata T, Kawaguchi O, MH. Clinical outcome in children with recurrent craniopharyngi-
et al. Late retinal complications of radiation therapy for nasal and oma after primary surgery. Cancer J. 2000;6(6):388–93.
paranasal malignancies: relationship between irradiated-dose area 57. Becker G, Kortmann RD, Skalej M, Bamberg M. The role of
and severity. Int J Radiat Oncol Biol Phys. 1999;44(3):599–605. radiotherapy in the treatment of craniopharyngioma—indications,
37. Zanation AM, Ferlito A, Rinaldo A, Gore MR, Lund VJ, results, side effects. Front Radiat Ther Oncol. 1999;33:100–13.
McKinney KA, et al. When, how and why to treat the neck in 58. Jose CC, Rajan B, Ashley S, Marsh H, Brada M. Radiotherapy for
patients with esthesioneuroblastoma: a review. Eur Arch the treatment of recurrent craniopharyngioma. Clin Oncol (R Coll
Otorhinolaryngol. 2010;267(11):1667–71. Radiol). 1992;4(5):287–9.
38. Weiden PL, Yarington Jr CT, Richardson RG. Olfactory neuro- 59. Lin LL, El Naqa I, Leonard JR, Park TS, Hollander AS, Michalski
blastoma. Chemotherapy and radiotherapy for extensive disease. JM, et al. Long-term outcome in children treated for craniopha-
Arch Otolaryngol. 1984;110(11):759–60. ryngioma with and without radiotherapy. J Neurosurg Pediatr.
39. Dulguerov P, Calcaterra T. Esthesioneuroblastoma: the UCLA 2008;1(2):126–30.
experience 1970-1990. Laryngoscope. 1992;102(8):843–9. 60. Sanford RA. Craniopharyngioma: results of survey of the
40. Zabel A, Thilmann C, Milker-Zabel S, Schlegel W, Zuna I, American Society of Pediatric Neurosurgery. Pediatr Neurosurg.
Wannenmacher M, et al. The role of stereotactically guided con- 1994;21 Suppl 1:39–43.
formal radiotherapy for local tumor control of esthesioneuroblas- 61. Rajan B, Ashley S, Gorman C, Jose CC, Horwich A, Bloom HJ,
toma. Strahlenther Onkol. 2002;178(4):187–91. et al. Craniopharyngioma—a long-term results following limited
41. Morita A, Ebersold MJ, Olsen KD, Foote RL, Lewis JE, Quast surgery and radiotherapy. Radiother Oncol. 1993;26(1):1–10.
LM. Esthesioneuroblastoma: prognosis and management. 62. Iwata H, Tatewaki K, Inoue M, Yokota N, Baba Y, Nomura R,
Neurosurgery. 1993;32(5):706–14. discussion 14-5. et al. Single and hypofractionated stereotactic radiotherapy with
42. Polin RS, Sheehan JP, Chenelle AG, Munoz E, Larner J, Phillips CyberKnife for craniopharyngioma. J Neurooncol. 2012;106(3):
CD, et al. The role of preoperative adjuvant treatment in the man- 571–7.
agement of esthesioneuroblastoma: the University of Virginia 63. Schulz-Ertner D, Frank C, Herfarth KK, Rhein B, Wannenmacher
experience. Neurosurgery. 1998;42(5):1029–37. M, Debus J. Fractionated stereotactic radiotherapy for craniopha-
43. Gore MR, Zanation AM. Salvage treatment of local recurrence in ryngiomas. Int J Radiat Oncol Biol Phys. 2002;54(4):1114–20.
esthesioneuroblastoma: a meta-analysis. Skull Base. 2011;21(1): 64. Chung WY, Pan DH, Shiau CY, Guo WY, Wang LW. Gamma
1–6. knife radiosurgery for craniopharyngiomas. J Neurosurg. 2000;93
44. Dinca EB, Radatz MW, Rowe J, Kemeny AA. Gamma Knife(R) Suppl 3:47–56.
radiosurgery for recurrent intracranial olfactory neuroblastoma 65. Mokry M. Craniopharyngiomas: A six year experience with
(esthesioneuroblastoma): a case report. J Med Case Reports. Gamma Knife radiosurgery. Stereotact Funct Neurosurg. 1999;72
2012;6(1):240. Suppl 1:140–9.
45. Walch C, Stammberger H, Anderhuber W, Unger F, Kole W, 66. Kickingereder P, Maarouf M, El Majdoub F, Fuetsch M, Lehrke R,
Feichtinger K. The minimally invasive approach to olfactory Wirths J, et al. Intracavitary brachytherapy using stereotactically
neuroblastoma: combined endoscopic and stereotactic treat- applied phosphorus-32 colloid for treatment of cystic craniopha-
ment. Laryngoscope. 2000;110(4):635–40. ryngiomas in 53 patients. J Neurooncol. 2012;109(2):365–74.
67. Ulfarsson E, Lindquist C, Roberts M, Rahn T, Lindquist M, 86. Foote RL, Coffey RJ, Gorman DA, Earle JD, Schomberg PJ, Kline
Thoren M, et al. Gamma knife radiosurgery for craniopharyngio- RW, et al. Stereotactic radiosurgery for glomus jugulare tumors: a
mas: long-term results in the first Swedish patients. J Neurosurg. preliminary report. Int J Radiat Oncol Biol Phys. 1997;38(3):491–5.
2002;97(5 Suppl):613–22. 87. Saringer W, Khayal H, Ertl A, Schoeggl A, Kitz K. Efficiency of
68. Kobayashi T, Kida Y, Mori Y, Hasegawa T. Long-term results of gamma knife radiosurgery in the treatment of glomus jugulare
gamma knife surgery for the treatment of craniopharyngioma in tumors. Min Invasive Neurosurg. 2001;44(3):141–6.
98 consecutive cases. J Neurosurg. 2005;103(6 Suppl):482–8. 88. Woods CI, Strasnick B, Jackson CG. Surgery for glomus tumors:
69. al-Mefty O, Borba LA. Skull base chordomas: a management the Otology Group experience. Laryngoscope. 1993; 103(11 Pt 2
challenge. J Neurosurg. 1997;86(2):182–9. Suppl 60):65–70.
70. Gay E, Sekhar LN, Rubinstein E, Wright DC, Sen C, Janecka IP, 89. Pauw BK, Makek MS, Fisch U, Valavanis A. Preoperative embo-
et al. Chordomas and chondrosarcomas of the cranial base: results lization of paragangliomas (glomus tumors) of the head and neck:
and follow-up of 60 patients. Neurosurgery. 1995;36(5):887–96. histopathologic and clinical features. Skull Base Surg. 1993;3(1):
discussion 96-7. 37–44.
71. Cho YH, Kim JH, Khang SK, Lee JK, Kim CJ. Chordomas and 90. Chretien PB, Engelman K, Hoye RC, Geelhoed GW. Surgical
chondrosarcomas of the skull base: comparative analysis of clini- management of intravascular glomus jugulare tumor. Am J Surg.
cal results in 30 patients. Neurosurg Rev. 2008;31(1):35–43; dis- 1971;122(6):740–3.
cussion 43. 91. Jordan JA, Roland PS, McManus C, Weiner RL, Giller CA.
72. Hug EB, Loredo LN, Slater JD, DeVries A, Grove RI, Schaefer Stereotactic radiosurgery for glomus jugulare tumors. Laryngo-
RA, et al. Proton radiation therapy for chordomas and chondrosar- scope. 2000;110(1):35–8.
comas of the skull base. J Neurosurg. 1999;91(3):432–9. 92. Maarouf M, Voges J, Landwehr P, Bramer R, Treuer H, Kocher M,
73. Kondziolka D, Lunsford LD, Flickinger JC. The role of radiosur- et al. Stereotactic linear accelerater-based radiosurgery for the
gery in the management of chordoma and chondrosarcoma of the treatment of patients with glomus jugulare tumors. Cancer.
cranial base. Neurosurgery. 1991;29(1):38–45; discussion 6. 2003;97(4):1093–8.
74. Pearlman AW, Friedman M. Radical radiation therapy of chordoma. 93. Lim M, Gibbs IC, Adler Jr JR, Martin DP, Chang SD. The efficacy
Am J Roentgenol Radium Ther Nucl Med. 1970;108(2):332–41. of linear accelerator stereotactic radiosurgery in treating glomus
75. Munzenrider JE, Liebsch NJ. Proton therapy for tumors of the jugulare tumors. Technol Cancer Res Treat. 2003;2(3):261–5.
skull base. Strahlenther Onkol. 1999;175 Suppl 2:57–63. 94. Springate SC, Weichselbaum RR. Radiation or surgery for chemo-
76. Noel G, Habrand JL, Mammar H, Pontvert D, Haie-Meder C, dectoma of the temporal bone: a review of local control and com-
Hasboun D, et al. Combination of photon and proton radiation plications. Head Neck. 1990;12(4):303–7.
therapy for chordomas and chondrosarcomas of the skull base: the 95. Hurmuz P, Cengiz M, Ozyigit G, Yazici G, Akyol F, Yildiz F, et al.
Centre de Protontherapie D'Orsay experience. Int J Radiat Oncol Robotic stereotactic radiosurgery in patients with unresectable
Biol Phys. 2001;51(2):392–8. glomus jugulare tumors. Technol Cancer Res Treat. 2013;12(4):
77. Schulz-Ertner D, Nikoghosyan A, Thilmann C, Haberer T, Jakel 275–84.
O, Karger C, et al. Results of carbon ion radiotherapy in 152 96. Foote RL, Pollock BE, Gorman DA, Schomberg PJ, Stafford SL,
patients. Int J Radiat Oncol Biol Phys. 2004;58(2):631–40. Link MJ, et al. Glomus jugulare tumor: tumor control and compli-
78. Hug EB, Sweeney RA, Nurre PM, Holloway KC, Slater JD, cations after stereotactic radiosurgery. Head Neck. 2002;24(4):
Munzenrider JE. Proton radiotherapy in management of pediatric 332–8. discussion 8-9.
base of skull tumors. Int J Radiat Oncol Biol Phys. 2002;52(4):1017. 97. Ivan ME, Sughrue ME, Clark AJ, Kane AJ, Aranda D, Barani IJ,
79. Debus J, Schulz-Ertner D, Schad L, Essig M, Rhein B, Thillmann et al. A meta-analysis of tumor control rates and treatment-related
CO, et al. Stereotactic fractionated radiotherapy for chordomas morbidity for patients with glomus jugulare tumors. J Neurosurg.
and chondrosarcomas of the skull base. Int J Radiat Oncol Biol 2011;114(5):1299–305.
Phys. 2000;47(3):591. 98. Hamilton RJ, Kuchnir FT, Sweeney P, Rubin SJ, Dujovny M,
80. Feigl GC, Bundschuh O, Gharabaghi A, Safavi-Abassi S, El Pelizzari CA, et al. Comparison of static conformal field with mul-
Shawarby A, Samii M, et al. Evaluation of a new concept for the tiple noncoplanar arc techniques for stereotactic radiosurgery or
management of skull base chordomas and chondrosarcomas. J stereotactic radiotherapy. Int J Radiat Oncol Biol Phys. 1995;33(5):
Neurosurg. 2005;102(Suppl):165–70. 1221–8.
81. Pamir MN, Kilic T, Ture U, Ozek MM. Multimodality manage- 99. Shiu AS, Kooy HM, Ewton JR, Tung SS, Wong J, Antes K, et al.
ment of 26 skull-base chordomas with 4-year mean follow-up: Comparison of miniature multileaf collimation (MMLC) with cir-
experience at a single institution. Acta Neurochir (Wien). cular collimation for stereotactic treatment. Int J Radiat Oncol
2004;146(4):343–54; discussion 54. Biol Phys. 1997;37(3):679–88.
82. Koga T, Shin M, Saito N. Treatment with high marginal dose is 100. Kubo HD, Pappas CT, Wilder RB. A comparison of arc-based and
mandatory to achieve long-term control of skull base chordomas static mini-multileaf collimator-based radiosurgery treatment
and chondrosarcomas by means of stereotactic radiosurgery. J plans. Radiother Oncol. 1997;45(1):89–93.
Neurooncol. 2010;98(2):233–8. 101. Kramer BA, Wazer DE, Engler MJ, Tsai JS, Ling MN. Dosimetric
83. Konefal JB, Pilepich MV, Spector GJ, Perez CA. Radiation therapy comparison of stereotactic radiosurgery to intensity modulated
in the treatment of chemodectomas. Laryngoscope. 1987;97(11): radiotherapy. Radiat Oncol Investig. 1998;6(1):18–25.
1331–5. 102. Cardinale RM, Benedict SH, Wu Q, Zwicker RD, Gaballa HE,
84. Naswa N, Kumar A, Sharma P, Bal C, Malhotra A, Kumar Mohan R. A comparison of three stereotactic radiotherapy tech-
R. Imaging carotid body chemodectomas with (6)(8)Ga-DOTA- niques; ARCS vs. noncoplanar fixed fields vs. intensity modula-
NOC PET-CT. Br J Radiol. 2012;85(1016):1140–5. tion. Int J Radiat Oncol Biol Phys. 1998;42(2):431–6.
85. Liscak R, Vladyka V, Wowra B, Kemeny A, Forster D, Burzaco 103. Benedict SH, Cardinale RM, Wu Q, Zwicker RD, Broaddus WC,
JA, et al. Gamma Knife radiosurgery of the glomus jugulare Mohan R. Intensity-modulated stereotactic radiosurgery using
tumour—early multicentre experience. Acta Neurochir (Wien). dynamic micro-multileaf collimation. Int J Radiat Oncol Biol
1999;141(11):1141–6. Phys. 2001;50(3):751–8.
104. Leavitt DD, Watson G, Tobler M, Williams G, Gaffney DK, for chordomas and chondrosarcomas of the skull base: focus on
Shrieve DC. Intensity-modulated radiosurgery/radiotherapy using complications. Int J Radiat Oncol Biol Phys. 2012;83(2):
a micromultileaf collimator. Med Dosim. 2001;26(2):143–50. 542–51.
105. Ernst-Stecken A, Lambrecht U, Ganslandt O, Mueller R, 109. Martin JJ, Niranjan A, Kondziolka D, Flickinger JC, Lozanne KA,
Fahlbusch R, Sauer R, et al. Radiosurgery of small skull-base Lunsford LD. Radiosurgery for chordomas and chondrosarcomas
lesions. No advantage for intensity-modulated stereotactic radio- of the skull base. J Neurosurg. 2007;107(4):758–64.
surgery versus conformal arc technique. Strahlenther Onkol. 110. Hasegawa T, Ishii D, Kida Y, Yoshimoto M, Koike J, Iizuka
2005;181(5):336–44. H. Gamma Knife surgery for skull base chordomas and chondro-
106. Krishnan S, Foote RL, Brown PD, Pollock BE, Link MJ, Garces sarcomas. J Neurosurg. 2007;107(4):752–7.
YI. Radiosurgery for cranial base chordomas and chondrosarco- 111. Eustacchio S, Trummer M, Unger F, Schrottner O, Sutter B, Pendl
mas. Neurosurgery. 2005;56(4):777–84; discussion 84. G. The role of Gamma Knife radiosurgery in the management of
107. Chang SD, Martin DP, Lee E, Adler Jr JR. Stereotactic radiosur- glomus jugular tumours. Acta Neurochir Suppl. 2002;84:91–7.
gery and hypofractionated stereotactic radiotherapy for residual or 112. Gottfried ON, Liu JK, Couldwell WT. Comparison of radiosur-
recurrent cranial base and cervical chordomas. Neurosurg Focus. gery and conventional surgery for the treatment of glomus jugu-
2001;10(3):E5. lare tumors. Neurosurg Focus. 2004;17(2):E4.
108. Hauptman JS, Barkhoudarian G, Safaee M, Gorgulho A, Tenn S, 113. Pollock BE. Stereotactic radiosurgery in patients with glomus
Agazaryan N, et al. Challenges in linear accelerator radiotherapy jugulare tumors. Neurosurg Focus. 2004;17(2):E10.
Skull Base Tumors: Viewpoint—Surgery
39
Richard F. Schmidt, Smruti K. Patel, Robert W. Jyung,
Jean Anderson Eloy, and James K. Liu
of the cranial nerves and brain stem. Surgical resection

Introduction offers direct decompression of neurologic structures in
order to restore functionality or limit symptom progression.
Tumors of the skull base pose unique challenges to the However, due to the complex location and pathology of
neurosurgeon. These tumors tend to extend throughout intra- skull base lesions, many of these tumors are treated with a
cranial and extracranial compartments and may invade or multidisciplinary approach catered toward the needs of the
compress surrounding neurovascular structures, causing sig- individual patient. Stereotactic radiosurgery (SRS) has
nificant neurological deficits and making complete surgical evolved as a therapeutic option in many circumstances.
resection difficult. However, new technologies and advances While the goals of surgery are tumor resection and decom-
in surgical techniques have made many of these tumors more pression of neurovascular structures, the primary goal of
surgically accessible, increasing the extent of tumor resec- SRS is tumor control. As such, this treatment can be used
tion and offering the potential for a surgical cure. As a result either as a primary intervention for small, asymptomatic
of developments in frameless stereotactic image guidance, benign tumors or as an adjuvant therapy to surgery in order
electrophysiological monitoring, modern microneurosurgi- to optimize tumor control. Due to the risk of radiation-
cal skull base techniques, and neuroendoscopy, neurosur- induced complications, this treatment modality is typically
geons now have the ability to access and remove a number of limited to treating lesions less than 30 mm, where dose fall-
these lesions with minimal manipulation and damage to sur- offs can be steep enough to limit the effect on surrounding
rounding neurovascular structures, ultimately leading to structures. This is a particular concern for lesions near the
significant improvements in patient outcomes [1]. optic nerve and brainstem, as radiation can cause drastic
The primary goal of surgical intervention in symptomatic complications at these sensitive locations.
patients with skull base tumors is decompression of neuro- There is currently much controversy over the role of sur-
vascular structures. Many patients often present with sig- gery vs. radiosurgery for the primary management of a num-
nificant neurological deficits as a result of tumor compression ber of tumors at the skull base. Currently, the treatment
decision largely depends on a combination of factors such
as tumor location, tumor biology, involvement of adjacent
R.F. Schmidt, M.D. • S.K. Patel, M.D. neurovascular structures, overall clinical status of the
Department of Neurological Surgery, Rutgers New Jersey
patient, availability of the technology, and the training of the
Medical School, Newark, NJ, USA
neurosurgeon. Many benign lesions can be cured by surgical
R.W. Jyung, M.D.
resection. Moreover, surgical resection is favorable to
Department of Otolaryngology-Head and Neck Surgery, Rutgers
New Jersey Medical School, Newark, NJ, USA decompress larger symptomatic lesions because SRS alone
may not offer sufficient symptomatic improvement. For
J.A. Eloy, M.D.
Departments of Otolaryngology-Head and Neck Surgery and malignant lesions or benign lesions with highly infiltrative
Neurological Surgery, Center for Skull Base and Pituitary Surgery, properties, surgery can offer both symptomatic palliation
Rutgers New Jersey Medical School, Newark, NJ, USA and cytoreduction as part of a multimodality treatment plan.
J.K. Liu, M.D., Ph.D. (*) Nevertheless, surgery and SRS are both important tools in
Departments of Neurological Surgery and Otolaryngology-Head the armamentarium of the treatment team, and can ulti-
and Neck Surgery, Center for Skull Base and Pituitary Surgery,
mately complement each other to achieve the optimal out-
Neurological Institute of New Jersey, Rutgers New Jersey Medical
School, 90 Bergen Street, Suite 8100, Newark, NJ 07103, USA come for the patient. For example, a petroclival meningioma
e-mail: james.liu.md@rutgers.edu with cavernous sinus invasion might be best treated with a
500 R.F. Schmidt et al.
subtotal resection to decompress the brainstem followed by (MRI) to identify the pathology and its relationship to
postoperative SRS for tumor control of the residual tumor in important structures, computed tomography (CT) to evalu-
the cavernous sinus. ate for tumor calcification or bony changes of the skull base,
This chapter will discuss the surgical considerations for and vascular imaging such as CT angiography/venography
tumors of the skull base, both as an individual treatment and (CTA/CTV), MR angiography/venography (MRA/MRV),
as part of a multimodality treatment plan. Due to the wide or digital subtraction angiography (DSA) to determine the
variety of pathological and anatomical considerations for surrounding vascular anatomy and vascular supply of the
various skull base lesions, the surgical management is highly tumor and venous drainage patterns. Recently, intraopera-
variable across tumor types. As such, the specific surgical tive electrophysiologic neuromonitoring and frameless ste-
considerations for a number of benign and malignant tumors reotactic navigation have also proven to be useful adjuncts.
will be discussed, including trigeminal schwannomas, glo- Ultimately, the best surgical approach, extent of resection,
mus jugulare tumors, chordomas, chondrosarcomas, esthe- and use of adjuvant therapies should be thoroughly consid-
sioneuroblastomas, craniopharyngiomas, and skull base ered in the context of the individual patient’s clinical status
meningiomas. Since these tumors fundamentally differ on a and surgical goals.
number of levels, discussion of their radiographic appear-
ance and pathological findings are out of the scope of this
chapter. Therefore, the discussion will focus primarily on Trigeminal Schwannomas
their clinical and anatomical presentation as well as the indi-
cations and outcomes of surgical intervention. Trigeminal schwannomas are relatively rare tumors,
accounting for approximately 0.07–0.5 % of all intracranial
tumors and 0.8–10 % of all intracranial schwannomas [2–6].
Surgical Considerations The emergence of microsurgical techniques and skull base
approaches has significantly improved the postoperative
While the tumors discussed in this chapter vary greatly, there outcomes for patients with these tumors. Since the late
are a number of considerations that are universally applied 1990s, rates of gross total resection (GTR) have consis-
when addressing the surgical management of the skull base. tently ranged from 60 to 100 %, with the majority of large
Primarily, the specific goals of the surgery should be dis- case series reporting rates greater than 70 %. Furthermore,
cussed with the individual patient based on their age and recurrence rates have consistently decreased, occurring in
unique clinical presentation; there is no “one-size-fits-all” less than 19 % of patients who undergo surgery as a pri-
method of treatment and the risk-to-benefit ratio of surgery mary treatment [2–5, 7–10]. Currently, surgery remains the
varies among individual patients. An asymptomatic, elderly standard treatment for most of these tumors, as GTR is
patient with multiple comorbidities presenting with a small, often curative.
benign tumor may not benefit from surgical resection because Trigeminal schwannomas typically arise from the trigem-
of the relatively greater risk of morbidity associated with the inal nerve root, the Gasserian ganglion or one of the three
procedure and postoperative recovery. On the other hand, a peripheral branches (V1–V3), making the clinical presenta-
young patient with severe facial pain due to compression of tion, radiographic findings, and treatment of these tumors
the trigeminal nerve by a large tumor would likely benefit highly variable [3]. They usually occur during the third or
from surgical decompression and complete resection, if fourth decade of life but may arise earlier in patients with
safely possible. Furthermore, tumor adherence to surround- neurofibromatosis type II [4, 6]. The most common present-
ing neurovascular structures may prohibit safe complete ing symptom is altered facial sensation occurring in 50–91 %
resection and is therefore an important consideration when of patients, which may range from pain to hypesthesias and/
determining the extent of surgical resection. A tumor encas- or paresthesias [3–6]. It has been suggested that facial pain is
ing the basilar artery might be best managed by achieving a typically associated with tumors arising from the Gasserian
subtotal resection to avoid complications of vascular com- ganglion, whereas pain is frequently absent in tumors arising
promise, opting for postoperative adjuvant radiation therapy from the nerve root [2].
for further tumor control. For the purposes of treatment by surgical resection,
Because of the variability seen in tumor extension and tumors are typically classified based on their location. The
infiltration, it is also important that the surgeon be equipped most commonly used classification system was developed by
with as much information as possible both before and dur- Jefferson [11], grouping tumors into four distinct classes
ing the surgery. Preoperative imaging is essential for based on location. Type A tumors are located primarily in the
determining the appropriate surgical approach and minimiz- middle fossa and represent 25–55 % of cases. Type B tumors
ing the manipulation of critical structures. A typical preop- are mainly situated in the posterior fossa, representing
erative work-up may include magnetic resonance imaging 11–50 % of cases. Type C tumors span both the middle and
39 Skull Base Tumors: Viewpoint—Surgery 501
posterior fossa, commonly referred to as “dumbbell tumors,” vention in small- to medium-sized tumors [12–15]. However,
and represent 14–40 % of tumors. Type D tumors are charac- it has been suggested that larger tumors (≥15 cm3) are more
terized by extradural extension and are the least commonly likely to progress despite SRS [13]. Given these results and
reported, comprising 6–13 % of tumors [2, 5, 11]. recommendations, surgery continues to remain the method
Recently, skull base approaches have replaced conven- of choice for most tumors, especially large, symptomatic
tional approaches to these tumors as a result of superior lesions.
long-term outcomes [3]. The advantages of these approaches
over conventional intradural approaches include better tumor
exposure, less brain retraction, and more complete removal Glomus Jugulare Tumors
[3]. A number of approaches have been recommended for
accessing type A tumors, including frontotemporal transsyl- Glomus jugulare tumors are an extremely rare group of
vian, zygomatic infratemporal, and orbitozygomatic infra- tumors representing 0.03 % of all intracranial tumors with
temporal approaches, with some studies advocating an an incidence of approximately 1 per 1.3 million population
additional anterior petrosectomy for tumors with minimal [16]. These tumors arise from the neuroendocrine chief cells
extension into the posterior fossa [2, 3, 5, 8, 10]. Type B of the paraganglia, which are located in the adventitia of the
tumors are typically treated using a lateral suboccipital retro- dome of the jugular bulb. Typically, these tumors are char-
sigmoid approach, similar to that used for vestibular schwan- acterized as benign, slow growing lesions that cause com-
nomas [2, 5]. Type C tumors tend to be the most difficult to pression of adjacent bone, cranial nerves, and blood vessels
resect because they occupy both the posterior and the middle by local infiltration and spread. Only about 1–5 % of glomus
fossa. As a result, numerous methods for accessing and jugulare tumors are malignant and exhibit metastatic spread
removing these tumors have been proposed, with little con- [17]. Due to their difficult anatomic location at the skull
sensus on which is most appropriate. These approaches base, hypervascularity, and advanced stage at diagnosis,
include a subtemporal, extradural-intradural approach with glomus jugulare tumors pose a formidable challenge for
anterior petrosectomy [7, 8], combined subtemporal presig- neurosurgeons (Fig. 39.1). However, recent advances in
moid/retromastoid approach [2], and a zygomatic middle neuroimaging, microsurgical skull base techniques, and
fossa approach, which avoids tentorial incision or anterior radiation therapy have enhanced and improved the manage-
petrosectomy by expanding Meckel’s cave [9]. Type D ment of these tumors. Currently, multiple modalities of
tumors typically extend down the mandibular or maxillary treatment for glomus jugulare tumors exist, including pri-
divisions of the trigeminal nerve through the foramen ovale mary modes of treatment such as microsurgical resection
or foramen rotundum, respectively. These tumors usually via various skull base approaches and SRS or adjunctive
require more extensive approaches to access the pterygopal- treatments such as preoperative vascular embolization and
atine or infratemporal fossa. Thus, the orbitozygomatic conventional external beam radiotherapy.
infratemporal and zygomatic infratemporal approaches are Microsurgical resection remains the treatment of choice
most commonly used [2, 3]. An extradural subtemporal- for glomus jugulare tumors [18, 19]. Early attempts at surgi-
infratemporal approach through a preauricular-infratemporal cal resection of these tumors were fraught with poor tumor
incision has also been proposed, as well as minimally inva- control, high rates of recurrence, and significant morbidity
sive endoscopic endonasal/transmaxillary approaches [5]. and mortality. As a result, over time, several approaches for
Tumors with cavernous sinus involvement are relatively microsurgical resection of glomus jugulare tumors have
common and are typically treated with less aggressive resec- been described, either as single or staged operations. These
tion, focusing instead on maximal preservation of vascular approaches include mastoid-neck [20], mastoid-neck with
structures and surrounding cranial nerves rather than com- limited facial nerve mobilization [20], infratemporal fossa
plete tumor resection [5]. type A and B [20–27], posterior fossa [23, 24], subtemporal-
While surgery is the treatment of choice for most of these infratemporal [25], retrosigmoid [25], extreme lateral trans-
tumors, especially in cases of large tumors or those causing condylar [25], posterolateral [28], transmastoid-transcervical
significant symptoms, SRS is emerging as a potential treat- [22], combined infratentorial and posterior fossa [29],
ment option for many of these lesions. It is well known that petro-occipital trans-sigmoid [30], combined transmastoid
SRS is very useful in patients with incomplete surgical resec- retro- and infralabyrinthine transjugular transcondylar
tion, tumor recurrence, or those who would not be able to transtubercular high cervical [31], infralabyrinthine retrofa-
tolerate prolonged surgical procedures [4, 5, 12]. Recent cial [32], and various combined approaches.
studies suggest that SRS may be useful in patients who are The primary goal of the microsurgical management of
poor candidates for surgery, have tumor that is tightly adher- glomus jugulare tumors is curative GTR. Recent studies
ent to vital structures, such as the brainstem or posterior have repeatedly shown excellent rates of GTR of these
portion of the cavernous sinus [3], or even as a primary inter- lesions and surgical control rates have consistently remained
Fig. 39.1 (a, b) Preoperative T1-weighted post-gadolinium MRI (a axial temporal fossa approach was performed with gross total resection of the
view, b coronal view) demonstrating a glomus jugulare tumor in the right tumor. (c, d) Postoperative T1-weighted post-gadolinium MRI (c axial
jugular foramen (white arrow). An extended far lateral transjugular infra- view, d coronal view) shows complete resection of the tumor
high in numerous studies [32–34]. Surgical control rate (also ture review by Suarez et al. [33], 226 patients with glomus
referred to as tumor control rate) is defined as complete jugulare tumors demonstrated a surgical control rate of
tumor elimination with no evidence of residual or recurrent 93.3 %.
disease throughout the follow-up period, including patients Despite the relatively high success rate, postoperative
who have multiple surgeries. Recently, many studies have complications, such as cranial nerve deficits, are significant
been published that have corroborated this evidence. A sys- causes of morbidity in patients with glomus jugulare tumors
tematic review of seven microsurgery case series with a total that are treated with microsurgery. Due to their location near
of 374 patients showed 88.2 % GTR of tumors during the the cerebellopontine angle, CN VII, VIII, and the lower cra-
initial surgery and a surgical control rate of 92.1 % [17, 20, nial nerves are primarily affected during these operations.
25, 27, 29, 30, 35]. A meta-analysis by Ivan et al. [34] The most common postoperative presentations related to cra-
showed that GTR was achieved in 351 (81 %) of 433 patients nial nerve dysfunction reported in the literature include facial
during the initial surgery for glomus jugulare tumors with an palsy, hearing loss, vocal cord paralysis resulting in hoarseness,
86 % surgical control rate. In a retrospective study by Borba dysphagia, and decreased gag reflex resulting in aspiration.
et al. [32] of 34 patients who underwent microsurgery for In most cases, these cranial neuropathies were transient.
glomus jugulare tumors, GTR was achieved in 91 % with a Thus, it is clear that cranial nerve preservation during resec-
surgical control rate of 94.2 % [32]. In a more recent litera- tion of glomus jugulare tumors is a key factor in reduction of
postoperative morbidity in these patients. In addition, pre- ously progress and invade surrounding structures, ulti-
vention of postoperative CSF leak is another important chal- mately resulting in significant disability and death [45].
lenge that is encountered during open microsurgical Clival chordomas typically present with cranial nerve
treatment of glomus jugulare tumors. CSF leak has been palsy secondary to tumor compression [44], and involvement
reported in more than half the cases presented in the litera- of the sella turcica by these tumors may also result in endo-
ture [20–24, 26, 29, 32, 34, 36]. crinopathies [46]. Less commonly, they may present acutely
More recently, the use of Gamma Knife, linear accelera- with epistaxis or intracranial hemorrhage [47, 48]. The main-
tor (LINAC), and CyberKnife SRS has been increasing for stay of treatment for skull base chordomas is surgical resec-
both primary and secondary treatments of these tumors tion followed by postoperative adjuvant radiotherapy [49].
largely due to the advantage of avoiding high rates of mor- However, due to their proximity to critical neurovascular
bidity. These tumors can serve as ideal candidates for treat- structures, the balance between obtaining maximal tumor
ment with SRS because they are well demarcated on MRI resection and limiting significant postoperative morbidity
and rarely invade the brain, allowing a steep dose decrease at from extensive resection has remained a topic of significant
the margin [37]. A recent meta-analysis of 19 studies describ- debate [50]. A recent meta-analysis of 23 observational stud-
ing the outcomes of patients with glomus jugulare tumors ies conducted by Di Maio et al. [49] suggested that radical
who were treated with SRS as a primary treatment exhibited resection was the most promising method of treatment for
an average tumor control rate of 97 % with minimal morbid- improving the progression-free survival and overall survival
ity [38]. Thus, SRS should be considered as a primary treat- in patients with skull base chordomas. They found
ment in patients who cannot tolerate surgery (i.e., advanced progression-free survival rates of 85 % and 47 % with com-
age and/or medical comorbidities), patients who have under- plete resection compared to 50 % and 25 % with incomplete
gone contralateral jugular foramen surgery, and for smaller resection at 5- and 10-year intervals, respectively.
tumors that do not exhibit mass effect. SRS is an important Furthermore, patients treated with complete resection had
adjuvant therapy for residual or recurrent tumor that is not overall survival rates of 95 % at both 5 and 10 years, whereas
amenable to surgical resection. patients treated with incomplete resection had overall sur-
The method of treatment of glomus jugulare tumors must vival rates of 71 % and 53 % at 5 and 10 years, respectively.
be individualized and tailored based on the tumor size, loca- Using a Kaplan–Meier analysis, they found a 3.83-fold
tion, extent of intracranial extension, as well as the patient’s increase in the risk of recurrence and a 5.85-fold increase in
overall surgical risk. Early and aggressive surgical resection the risk of death at 5 years for patients treated with incom-
is ultimately favored for small tumors in younger patients in plete resection when compared to those treated with com-
order to offer the patient the best chance for preservation of plete resection.
neurological function and prevention of future deterioration Currently, a number of surgical approaches exist for
given that the patient is an appropriate surgical candidate accessing clival chordomas depending on the size and loca-
[18, 21–23, 32, 36]. Tumors with significant intracranial tion of the tumor. Staged procedures using more than one
extension exhibiting mass effect or brainstem compression approach have also been advocated by some authors for
and those that are malignant glomus jugulare tumors should treatment of exceedingly complex lesions [51, 52]. In a study
be removed surgically. of 71 patients, Sen et al. [42] described their rationale for
approaching these tumors using midline anterior approaches,
including extended subfrontal, transoral, transmaxillary,
Chordomas transmandibular and more recently, endoscopic endonasal
approaches, for central extradural tumors [42]. For tumors
Chordomas are a rare neoplasm derived from undifferenti- extending into the cavernous sinus, surrounding the internal
ated notochordal remnants in the axial skeleton, accounting carotid artery, or involving the vertebrobasilar system and
for 1–4 % of all bone malignancies and 17.5 % of malig- the brainstem, they recommend using a lateral skull base
nancies of the axial skeleton [39, 40]. Chordomas of the approach, suggesting that it allows for better control of the
skull base typically arise from the clivus (Fig. 39.2) and vessels and better separation of the brainstem–tumor inter-
have an incidence of approximately one per two million face [42]. Recently, the endoscopic endonasal approach has
population, representing approximately 25–39 % of all become a favorable approach among many neurosurgeons
chordomas [41–43]. While these tumors are histologically [42, 53]. The utility of this approach in treating chordomas is
low grade, they are highly infiltrative, associated with a largely due to the fact that these tumors are typically midline
large tumor burden at diagnosis, and exhibit high rates of and remain extradural, offering a low risk of CSF leak [53].
recurrence, making them very difficult tumors to manage In an analysis of 37 studies including 766 patients, Komotor
clinically [44]. Without intervention, these tumors insidi- et al. [53] found that the endoscopic endonasal approach
offered significant advantages over open microsurgical
Fig. 39.2 (a, b) Preoperative sagittal T2 (a) and T1-weighted post- formed resulting in near-total resection of the tumor. (c, d) Postoperative
gadolinium (a) MRI demonstrating a clival chordoma with brainstem sagittal T2 (c) and T1-weighted post-gadolinium (d) MRI shows
compression. An endoscopic endonasal transclival approach was per- decompression of the brainstem and basilar artery
approaches. These include significantly higher rates of gross been suggested to have some utility in cases of gross total
total resection (61.0 % vs. 48.1 %), fewer cranial nerve defi- resection [49, 51, 54]. Proton-beam radiotherapy is currently
cits (1.3 % vs. 24.2 %), lower rates of meningitis (0.9 % vs. the most widely accepted form of adjuvant treatment for
5.9 %), less overall mortality (4.7 % vs. 21.6 %), and fewer skull base chordomas, with studies suggesting that it pro-
cases of tumor recurrence (16.9 % vs. 40.0 %). However, vides improved progression-free survival when compared to
their conclusions were limited by selection bias and a shorter conventional radiotherapy [55]. However, there is a signifi-
length of follow-up for endoscopically treated tumors. cant dearth of data comparing proton-beam radiotherapy
Nevertheless, they recommend that the endoscopic endona- with other modalities, such as SRS. Furthermore, multiple
sal route is at least as efficacious as existing microsurgical studies have suggested that there is no difference between
techniques and suggest that it be highly considered for the types of adjuvant radiotherapy that are offered [49, 56].
patients with small midline tumors (<4 cm diameter or Ultimately, we recommend that all patients with skull base
<35 cm3 volume) with minimal lateral extension or carotid chordomas receive aggressive surgical resection as the pri-
artery involvement [53]. mary treatment and be considered for adjuvant postoperative
Adjuvant postoperative radiation therapy is a widely radiation therapy or SRS for residual disease, with the type
accepted adjunct therapy for the management of residual of radiotherapy left to the discretion of the treatment team
tumor following incomplete surgical resection, and has also based on the available options.
injury to other surrounding structures is minimal [66]. In an

Chondrosarcoma analysis of tumor recurrence and outcomes of 560 patients
with cranial chondrosarcomas conducted by Bloch et al. [58,
Chondrosarcomas of the skull base account for approxi- 62], they found that patients treated with GTR had similar
mately 0.15 % of all intracranial tumors and 6 % of all skull rates of recurrence-free survival compared with patients who
base neoplasms [57, 58]. It is postulated that these tumors received incomplete resections followed by adjuvant radia-
arise de novo from chondrocytes within the endochondral tion therapy. Furthermore, they found that recurrence rates
cartilage remnants of the petrous portion of the temporal were 44 % vs. 9 % and 5-year mortality rates were 25 % vs.
bone and areas of the petro-occipital, spheno-occipital, and 9 % for patients treated with surgery alone and those treated
spheno-petrosal synchondroses; however, other authors sug- with surgery followed by radiation therapy, respectively.
gest that some of these tumors may arise as a result of the Ultimately, they suggest that less aggressive resections may
metaplasia of mature fibroblasts [58–61]. Chondrosarcomas be appropriate in many situations in order to minimize the
tend to present in the fourth and fifth decades and are associ- potential morbidity from surgery, especially in low-grade
ated with Ollier’s disease, Maffucci syndrome, Paget’s dis- and conventional-type tumors.
ease, and osteochondromas [58]. They are typically divided The surgical approach to skull base chondrosarcomas var-
into four groups based on histological appearance, including ies greatly depending on the location of the tumor and the
conventional, dedifferentiated, clear cell, and mesenchymal degree of involvement of surrounding structures. In patients
subtypes. Most skull base chondrosarcomas fall under the with large complex tumors, a staged method using multiple
conventional or mesenchymal classifications, with conven- surgical approaches to access various surgical corridors may
tional subtypes being far more common. Fortunately, con- be warranted [52]. Wanebo et al. [66] illustrated a variety of
ventional subtypes also have a better prognosis than approaches determined by tumor location. For tumors of the
mesenchymal subtypes, with lower rates of both recurrence paraclinoid region and upper half of the clivus, they recom-
(16 % vs. 63 %) and 5-year mortality (6 % vs. 54 %) [58, 62]. mend employing a pterional approach with the addition of an
Additionally, these tumors are histologically stratified into a orbitozygomatic osteotomy, as this provides a more exten-
three-tiered grading system. Grade 1 tumors have the least sive basal exposure [66–68]. For tumors of the lower half of
malignant properties and lowest rates of recurrence, with the clivus, they advocate use of a retrosigmoid approach. If
better long-term survival, while Grade 2 and Grade 3 tumors there is widespread soft-tissue involvement or lateral exten-
possess progressively more malignant histological character- sion of tumor, they elect for a multidisciplinary strategy
istics and are associated with worse outcomes [62, 63]. It is including the addition of transfrontal and transmaxillary
important to recognize these factors when determining the approaches. Since these tumors frequently appear in the mid-
appropriate treatment plan, as high-grade and non- line, numerous authors have also advocated for the use of
conventional lesions may require more aggressive treatment endoscopic endonasal approaches [69–71]. Proponents of
to obtain adequate tumor control. Furthermore, it is impor- this approach suggest that the endoscope offers the benefits
tant to distinguish these tumors from chordomas, as chon- of minimal access and wider panoramic visualization of
drosarcomas are associated with a lower rate of recurrence deep neural structures [70, 71].
and better long-term survival [64–66]. This distinction can It is important to note that these tumors frequently involve
be made by staining for markers such as epithelial membrane the cavernous sinus and internal carotid artery, and may
antigen and cytokeratin, which will stain positive in the case extend intradurally toward the brainstem, hypothalamus, and
of chordoma, but negative for chondrosarcomas [52]. internal capsule [52]. Careful resection of the tumor off of
Chondrosarcomas tend to compress or invade surround- these elements is crucial in avoiding patient morbidity.
ing neural structures as they progressively enlarge. Most Furthermore, chondrosarcomas have been shown to spread
commonly, patients will present with diplopia and headache along the subperiosteal spaces and through the bone marrow.
secondary to tumor enlargement and compression of cranial Therefore, if total resection is desired, extensive bone drill-
nerves III, IV, and VI [52, 58, 66]. Less frequently, the tumor ing should be performed along the tumor margins until
may also affect cranial nerves II, V, and VIII, resulting in bleeding marrow is reached [52]. Ultimately, the choice of
visual loss, trigeminal pain or numbness, and hearing loss surgical approach and extent of resection should be deter-
[52, 66]. Unfortunately, in many cases, complete surgical mined by the location of the tumor, the involvement of sur-
resection is not attainable due to the extensive involvement rounding structures, the patient’s clinical status, and the
of surrounding critical structures, such as the cavernous operator’s experience and comfort with various techniques.
sinus and internal carotid engulfment [66]. Therefore, Radiation therapy has been shown to give a distinct
decompression of neural structures, blood vessels, and the advantage when performed in combination with surgery,
brainstem by tumor debulking is typically the primary goal, regardless of the degree of resection [58, 62]. However, there
with additional resection being attempted only if the risk of is scarce comparison between the different types of radiation
therapies available. Furthermore, most studies are limited in adjuvant) chemotherapy followed by craniofacial resection
that they include both chordomas and chondrosarcomas, and postoperative radiation therapy should be considered for
even though chondrosarcomas have been shown to be rela- Kadish C (tumors with skull base or intracranial extension)
tively more susceptible to radiation treatment [52, 63, 72, and higher-grade lesions (tumor with neck or distant metas-
73]. Radiation therapy as a primary treatment has been tases) [88, 89].
shown to be efficacious in certain situations; however, the Over the past 2 decades, various advances in skull base
outcomes are inferior to those tumors treated with both sur- techniques have been made to incorporate both endoscopic-
gery and postoperative radiotherapy [58, 62]. As such, we assisted craniofacial resection and more recently, purely
suggest that all patients with chondrosarcomas of the skull endoscopic endonasal resection of these tumors in addition
base should be treated with surgical resection as the primary to standard craniofacial surgical resection [90]. The intro-
method of treatment followed by postoperative adjuvant duction of endoscopy has transformed traditional open skull
radiation therapy. base surgery and confers various advantages over conven-
tional craniofacial resection. These advantages include: (1)
no retraction of the frontal lobes, (2) better visualization in
Esthesioneuroblastomas difficult areas, (3) use of natural orifices (nasal corridor) to
avoid visible incisional scar, (4) reduction in recovery time,
Esthesioneuroblastomas, also known as olfactory neuroblas- and (5) less overall morbidity and mortality [91–95].
tomas, are rare, malignant neoplasms of neural crest origin However, longer follow-up is necessary in order to better
that arise from the olfactory epithelium of the upper nasal assess the recurrence-free survival rates and prognosis in
cavity and anterior skull base [74–76]. These tumors may patients with esthesioneuroblastoma resected via the purely
invade locally and spread to the paranasal sinuses, nasal cav- endoscopic endonasal route. In our experience, careful
ity, and other surrounding structures, typically causing nasal patient selection is important when considering the purely
obstruction and epistaxis (Fig. 39.3) [77, 78]. They may also endoscopic approach. We generally reserve the endoscopic
extend superiorly across the cribriform plate intracranially to approach for sinonasal tumors with minimal intracranial
involve the brain or seed the cerebrospinal fluid. Metastasis extension, and use a combined transbasal bifrontal approach
may occur via the lymphatic or hematogenous route in and endoscopic endonasal approach (combined cranionasal)
17–48 % patients, and is most commonly seen in the cervical for sinonasal tumors that have more extensive intracranial
lymph nodes and more distant sites, including the lungs, involvement. Additionally, the role of SRS in management
brain, and bone [79, 80]. Due to the rarity of these tumors of esthesioneuroblastomas has not been very well defined
and the heterogeneity of treatment modalities described in and further studies are required with longer follow-up in
the literature, a standardized management protocol has not order to assess whether this treatment modality is efficacious
been established for esthesioneuroblastomas [80]. [86, 96–98].
Craniofacial surgical resection has largely remained the
preferred method of treatment and usually requires the
removal of the cribriform plate in order to prevent perineural Craniopharyngiomas
extension. Surgical resection is typically followed by adju-
vant radiotherapy, and in some cases, chemotherapy in order Craniopharyngiomas are benign, epithelial-squamous, extra-
to achieve better long-term results [80–84]. The decision by axial tumors that arise along the path of the craniopharyngeal
which the modality of therapy is chosen for treatment is duct or Rathke’s pouch, and are designated as WHO Grade 1
largely driven by the characteristics of the tumor and primar- tumors [99, 100]. They have an overall incidence of 0.5–2
ily based on Kadish staging. According to most available per 100,000 [101] and represent approximately 5 % of all
studies, Kadish A (tumors limited to the nasal cavity) and intracranial tumors and 5–10 % of all pediatric intracranial
Kadish B tumors (tumors that extend to paranasal sinuses) tumors (Fig. 39.4) [102, 103]. These tumors typically exhibit
are optimally treated with craniofacial resection followed by a bimodal distribution, with peak incidences at 5–14 years
postoperative radiotherapy for greater disease-free survival. and 50–74 years [101, 104], with childhood craniopharyn-
Postoperative radiation therapy, in patients with Kadish A giomas accounting for 30–50 % of all cases [101, 105]. They
and B tumors, has shown to improve recurrence-free survival are classified into two histopathological subtypes, adamanti-
at 5 years when compared to patients treated with surgery nomatous craniopharyngiomas, which are associated with a
alone [85]. Overall survival rates in patients treated with sur- mutation of the β-catenin gene and predominantly occur in
gical resection followed by postoperative radiation therapy the first 2 decades, and papillary craniopharyngiomas, which
have been reported to be up to 100 % [86] and 5-year disease- typically occur in adults [106]. Some authors have suggested
free survival rates up to 85.7 % [87]. Recent literature also lower recurrence rates in those with papillary histology, but
suggests that multimodality therapy with preoperative (neo- others have reported no difference in outcome or recurrence
Fig. 39.3 (a, b) Preoperative T1-weighted post-gadolinium MRI of the tumor with reconstruction of the resultant skull base defect. (c, d)
(a sagittal view, b coronal view) demonstrating an esthesioneuroblas- Postoperative T1-weighted MRI (c sagittal, d coronal) shows gross
toma with intracranial extension through the cribriform plate. An endo- total resection of the tumor. Packing is visualized in the nasal cavity
scopic endonasal transcribriform approach was performed for removal bolstering the skull base repair
between the two subtypes [106, 107]. Grossly, they are typi- system has been developed based on and the degree of
cally characterized as cystic, solid, and mixed, and many hypothalamic involvement, with Grade 0 tumors having no
tumors exhibit extensive calcification. involvement, Grade 1 exhibiting compression of the hypo-
While these tumors are classified as benign, they can thalamus, and Grade 2 involving the hypothalamus.
sometimes exhibit tumor adherence to critical structures in Furthermore, the clinical presentation of craniopharyngio-
the parasellar region, including the optic apparatus, pituitary mas varies by the anatomical location of the lesion with rela-
stalk, pituitary gland, hypothalamus, third ventricle, anterior tion to the optic chiasm. Prechiasmatic lesions typically
cerebral artery complex, and surrounding perforating involve the optic apparatus, resulting in decreased visual
arteries. Hypothalamic involvement is of particular concern, acuity and visual field defects, usually presenting as bitemporal
as it has been shown to be significantly related to pre- and hemianopsia in up to 49 % of cases [100, 111, 112].
postoperative morbidity, including behavioral disturbances, Retrochiasmatic lesions commonly present with hydroceph-
obesity, cognitive decline, and an overall decrease in quality alus and signs of increased intracranial pressure, including
of life [108–110]. Because of this, a separate classification headache, nausea, vomiting, papilledema, and horizontal
Fig. 39.4 (a, b) Preoperative T1-weighted post-gadolinium-enhanced (c, d) Postoperative T1-weighted post-gadolinium-enhanced MR
MRI (a sagittal, b coronal) demonstrating a pediatric craniopharyngi- images (c sagittal, d coronal) show gross total resection of the tumor
oma resulting in visual loss with optic compression. An endoscopic with decompression of the optic apparatus
endonasal transplanum transtuberculum approach was performed.
double vision [100, 113]. Intrasellar lesions typically present being put into question, especially in cases where GTR
with headache and endocrinopathies, including polyuria, might result in significant morbidity [109, 110]. Some
polydipsia, growth retardation, developmental disturbances authors suggest that significant hypothalamic involvement
during puberty in children, hypogonadism in adults, and sig- (i.e., Grade 2) is a strong indicator to undergo subtotal
nificant weight gain [100, 114]. However, despite these resection followed by radiation therapy as opposed to GTR
trends associated with anatomical location, many patients [110, 117, 119]. Additionally, some recommend minimal
will present with a combination of these clinical features. resection, such as cyst decompression followed by radiation
Obtaining GTR has traditionally been the primary goal therapy in elderly patients with multiple comorbidities, or in
of surgical management due to the relatively high rates of patients with stalk infiltration but an intact hypothalamic–
recurrence. It has been shown that tumor recurrence is pres- pituitary axis [117]. Nevertheless, a GTR should be
ent in approximately 10–30 % of cases treated by GTR and attempted whenever safely possible.
up to 50 % in cases treated without GTR [104, 106, 109, Initially, transcranial routes, including pterional, orbito-
111, 112, 115–118]. Furthermore, recurrent tumors are zygomatic, subfrontal, transcallosal, and bifrontal
associated with worse morbidity, mortality, and decreased approaches, were the method of choice for accessing these
long-term survival. However, this paradigm is currently lesions [103, 104, 106, 109, 111, 112, 115, 116, 119–121].
While these approaches can expose a wide variety of ana- invaded and hyperostotic bone and surrounding dura, if
tomical locations, they typically require significant brain safely possible [129, 130]. In many cases, only subtotal
retraction and manipulation of neurovascular structures in resection can be achieved, and in these cases postoperative
order to obtain optimal visualization and surgical resection radiosurgery or radiotherapy is often required for tumor
[117]. The extended transsphenoidal microsurgical approach, control [131].
utilizing an operative microscope through a nasal speculum, Over the past 2 decades, it has been well documented that
provides a more direct route to the sellar and suprasellar extent of resection of skull base meningiomas correlates with
regions [103, 117, 122, 123]. Recently, endoscopic endona- long-term tumor recurrence [132–136]. Therefore, the key
sal approach (transplanum transtuberculum) has emerged as goals of the skull base meningioma treatment paradigm
a viable treatment option for resection of these tumors [117, include: (1) maximal resection of tumor with minimal neuro-
124]. The use of an endoscope has enabled surgeons to logical deficit, (2) preservation of functional outcome, and
obtain a wider field of view, allowing minimally invasive (3) prevention of tumor recurrence [129]. Various approaches
resection of these tumors with minimal manipulation of sur- of microsurgical resection have been described for skull base
rounding structures. In general, studies have shown similar meningiomas including infratemporal fossa [137], retrosig-
rates of GTR among the different approaches, ranging from moid [138], and extended approaches to the jugular foramen;
9 to 90 % for transcranial approaches, 12–90 % for micro- the supraorbital [139], pterional [140], and orbitozygomatic
scopic transsphenoidal approaches, and 29–86 % for endo- [141] approaches for tuberculum sellae meningiomas; com-
scopic endonasal transsphenoidal approaches [104, 106, bined subtemporal-suboccipital approach, the preauricular-
109, 111, 112, 115–117, 119, 120, 122, 123]. Ultimately, the subtemporal-infratemporal approach [142], the posterior
decision of which operative approach to use largely depends transpetrosal [143], anterior transpetrosal-transtentorial
on the specific goals of the patient as well as the training and [144], and combined supratentorial and infratentorial
technical preference of the neurosurgeon. approaches [145, 146] for sphenoid wing, petroclival, and
As previously mentioned, radiation treatment has become tentorial meningiomas; the transoral approach to the lower
a commonly used therapy in certain cases of craniopharyn- clivus and upper cervical spine [43]; and the extreme lateral
giomas. Rates of tumor control and mortality have been transcondylar approach [147–149] for ventral foramen mag-
shown to be comparable in cases where subtotal resection num meningiomas at the craniocervical junction. More
with radiation therapy was chosen over GTR or when radia- recently, endoscopic endonasal techniques have also been
tion therapy was chosen for the treatment of recurrent tumors described and reported for certain midline skull base menin-
[118, 125, 126]. However, studies addressing the use of SRS giomas such as planum sphenoidale, olfactory groove, tuber-
are currently very limited. Currently, SRS is used for small, culum sellae, and clival meningiomas [150–157].
solid craniopharyngiomas located 3–5 mm away from the Among the most challenging meningiomas to treat surgi-
optic nerve or chiasm [127] or in cases of recurrent or resid- cally include cavernous sinus, tuberculum sellae, and petro-
ual tumors following subtotal or GTR [128]. In general, sur- clival meningiomas. Cavernous sinus meningiomas typically
gical resection remains the treatment of choice for present with visual and ocular motor deficits such as ptosis,
craniopharyngiomas, with GTR being the ultimate goal diplopia, anisocoria, and ophthalmoplegia, in addition to tri-
when obtainable. However, more studies are needed to com- geminal nerve dysfunction [158]. Cavernous sinus meningi-
pare the long-term outcomes of various combinations of sur- omas are inherently difficult tumors to resect because of their
gical techniques and radiotherapy interventions in order to involvement of the cavernous sinus, internal carotid artery,
establish the best possible course for patient care. and cranial nerves III–VI. The treatment strategies for cav-
ernous sinus meningiomas have evolved over the past decade
[159–163]. In many cases, subtotal resection and decom-
Skull Base Meningiomas pression of the cavernous sinus and optic nerve can be con-
sidered. In these instances, postoperative radiotherapy or
Skull base meningiomas are typically difficult lesions to SRS for residual tumor may be indicated if the meningioma
resect because of their location and close proximity to vital displays atypical features on histology or if it is found to be
neurovascular structures. These tumors include, but are not malignant [129]. Long-term outcomes of Gamma Knife SRS
limited to, orbital, optic nerve sheath, olfactory groove, for patients with small- or medium-sized cavernous sinus
tuberculum sellae, cavernous sinus, sphenoid wing, cerebel- tumors have shown good results with respect to progression-
lopontine angle, petroclival, and tentorial meningiomas. free survival (90 and 75 %) and functional outcomes (93 and
The other inherent difficulty in treating these lesions is that 91 %) at 5 and 10 years, respectively [164]. Tumor progres-
skull base meningiomas tend to differ in volume, shape, sion was more likely found to occur from margins that
location, and clinical symptomatology. Optimal manage- remain outside of the treatment volume [164]. Furthermore,
ment of these tumors involves GTR with removal of the tumors that are asymptomatic and radiographically stable
Fig. 39.5 (a, b) Preoperative T1-weighted post-gadolinium MRI (a approach resulting in gross total resection of the tumor. (c, d)
axial, b coronal) demonstrating a left petroclival meningioma with Postoperative T1-weighted post-gadolinium MRI (c axial, d coronal)
brainstem compression. The patient underwent a combined petrosal shows complete removal of the tumor with brainstem decompression
over time are followed up at appropriate intervals without Petroclival meningiomas are typically treated on a case-
intervention. specific basis and numerous aspects of management are con-
Petroclival meningiomas (Fig. 39.5) are rare and make up sidered, including the size and location of tumor, choice of
only about 3–10 % of all posterior fossa meningiomas [165]. surgical approach, the postoperative evaluation of results, and
These tumors are particularly complex because of their deep treatment of recurrence [129, 138, 144, 145, 169–171]. For
location and often intimate relationship with vessels, nerves, smaller tumors in symptomatic patients, GTR, if feasible,
and the brainstem. In addition, the natural history of these should be the goal of surgery. On the other hand, small and
tumors may be unpredictable and frequently involves pro- asymptomatic meningiomas, particularly in elderly patients,
gressive growth and neurological deterioration [166, 167]. may be observed and followed up with serial neurological
The neurologic presentation of petroclival meningiomas var- exams and imaging. Larger petroclival meningiomas may
ies, though some of the more common symptoms include require combined surgical approaches to allow adequate
headache, gait disturbances, cranial nerve deficits, and other tumor exposure and debulking [145, 146, 171]. In cases
symptoms related to brainstem compression [166, 168]. involving large or giant petroclival meningiomas, it is impor-
tant to consider the patient’s postoperative quality of life giomas is dural involvement that extends laterally over the
[135]. In some cases, settling for a subtotal resection with optic canal and anterior clinoid process. The risk of CSF
decompression of adjacent structures is reasonable, with or leakage is also higher but has been reduced since the advent
without postoperative radiosurgery, particularly in cases that of the pedicled nasoseptal flap reconstruction [182, 183].
involve cavernous sinus invasion, pial invasion, and tumor More recently, SRS has been studied both as primary
adherence to cranial nerves and vessels [166]. treatment for skull base meningiomas that are not readily
Tuberculum sellae meningiomas make up 5–10 % of all amenable to surgery and as an adjuvant therapy in cases
intracranial meningiomas and pose a unique surgical chal- where GTR was not achieved. These studies have shown
lenge. They typically arise from the tuberculum sellae, chias- that this treatment modality provides acceptable rates of
matic sulcus, limbus sphenoidale, and diaphragma sellae tumor control (upwards of 85 % in most studies), while pre-
[172]. These tumors commonly cause displacement of the serving neurological function in many patients and avoiding
optic chiasm and optic nerves and often have optic canal the mortality and morbidity associated with an open surgi-
involvement. The most common clinical manifestation of cal procedure [131, 184]. Of note, neurological preservation
tuberculum sellae meningiomas is progressive visual loss, or improvement was more commonly observed in patients
classically known as “the chiasmal syndrome,” a term first who received Gamma Knife treatment for lesions that origi-
coined by Harvey Cushing in 1930 to describe the hallmark nated in the petroclival, parasellar, and cerebellopontine
presentation of these lesions [173]. The ideal treatment angle locations [184]. Though the results of these studies
involves decompression of the optic apparatus by achieving are certainly promising in the treatment of these tumors, the
a Simpson Grade I removal (i.e., surgical GTR of the tumor, decision-making protocol for skull base meningiomas
its dural attachments, and associated hyperostotic bone) should always include surgery as the primary tier of man-
while preserving visual function, the integrity of critical agement, if it can be safely achieved, given that this is the
adjacent neurovascular structures, and endocrine function only option that can achieve tumor reduction, decompres-
[154, 173]. In cases where tumor extends into the optic canal, sion of compromised neural structures, and provide poten-
early bony decompression of the optic canal (including ante- tial oncologic cure.
rior clinoidectomy, optic canal unroofing, and incising falci-
form ligament) has been advocated during the removal of
tuberculum sellae meningiomas and appears to have a strong References
influence on postoperative visual recovery [154, 173–177].
This decompression involves extradural, and in some cases, 1. Nanda A, Jawahar A, Sathyanarayana S (2003) Microsurgery for
intradural removal of the anterior clinoid process as well as potential radiosurgical skull base lesions: a retrospective analysis
and comparison of results. Skull Base 13(3):131–138
the unroofing of the optic canal followed by incision of the 2. Samii M, Migliori MM, Tatagiba M, Babu R (1995) Surgical treat-
falciform ligament and dural sheath of the involved optic ment of trigeminal schwannomas. J Neurosurg 82(5):711–718
nerve. In addition to early decompression of the optic nerve, 3. Zhang L, Yang Y, Xu S, Wang J, Liu Y, Zhu S (2009) Trigeminal
this particular maneuver allows exploration of the optic canal schwannomas: a report of 42 cases and review of the relevant sur-
gical approaches. Clin Neurol Neurosurg 111(3):261–269
for additional tumor since this can be a frequent site of recur- 4. Fukaya R, Yoshida K, Ohira T, Kawase T (2010) Trigeminal
rence [154, 173, 174, 177–179]. schwannomas: experience with 57 cases and a review of the litera-
Classically, skull base meningiomas have been removed ture. Neurosurg Rev 34(2):159–171
through standard, transcranial approaches. However, in more 5. Guthikonda B, Theodosopoulos PV, van Loveren H, Tew JM Jr,
Pensak ML (2008) Evolution in the assessment and management
recent years, with the introduction of extended transsphenoi- of trigeminal schwannoma. Laryngoscope 118(2):195–203
dal microsurgical approaches and expanded endoscopic 6. Sharma BS, Ahmad FU, Chandra PS, Mahapatra AK (2008)
endonasal techniques, the optimal surgical approach for Trigeminal schwannomas: experience with 68 cases. J Clin
these tumors has been much debated [180]. Although there is Neurosci 15(7):738–743
7. Taha JM, Tew JM Jr, van Loveren HR, Keller JT, el-Kalliny M
no firm consensus as to which method is superior for removal (1995) Comparison of conventional and skull base surgical
of these tumors, most authors agree that each surgical approaches for the excision of trigeminal neurinomas. J Neurosurg
approach has its own advantages and disadvantages [181]. 82(5):719–725
Ultimately, the route eventually chosen depends upon the 8. Yoshida K, Kawase T (1999) Trigeminal neurinomas extending
into multiple fossae: surgical methods and review of the literature.
preference of the surgeon, the precise location and size of J Neurosurg 91(2):202–211
tumor in relation to surrounding structures, and individual 9. Al-Mefty O, Ayoubi S, Gaber E (2002) Trigeminal schwannomas:
patient factors. In either scenario, the goal of oncologic removal of dumbbell-shaped tumors through the expanded Meckel
treatment remains the same: a Simpson Grade I tumor cave and outcomes of cranial nerve function. J Neurosurg
96(3):453–463
removal, with adequate optic nerve decompression and pres- 10. Goel A, Muzumdar D, Raman C (2003) Trigeminal neuroma:
ervation of the visual function. The major limitation for analysis of surgical experience with 73 cases. Neurosurgery
endoscopic endonasal resection of tuberculum sellae menin- 52(4):783–790; discussion 90
11. Jefferson G (1953) The trigeminal neurinomas with some remarks for resection of glomus jugulare tumors. Neurosurgery 59(1 Suppl
on malignant invasion of the gasserian ganglion. Clin Neurosurg 1):ONS115–ONS125; discussion ONS-25
1:11–54 32. Borba LAB, Araujo JC, de Oliveira JG, Filho MG, Moro MS,
12. Kano H, Niranjan A, Kondziolka D, Flickinger JC, Dade LL Tirapelli LF et al (2010) Surgical management of glomus jugulare
(2009) Stereotactic radiosurgery for trigeminal schwannoma: tumors: a proposal for approach selection based on tumor relation-
tumor control and functional preservation Clinical article. J ships with the facial nerve. J Neurosurg 112(1):88–98
Neurosurg 110(3):553–558 33. Suarez C, Rodrigo JP, Bodeker CC, Llorente JL, Silver CE, Jansen
13. Hasegawa T, Kida Y, Yoshimoto M, Koike J (2007) Trigeminal JC et al (2013) Jugular and vagal paragangliomas: systematic
schwannomas: results of gamma knife surgery in 37 cases. study of management with surgery and radiotherapy. Head Neck
J Neurosurg 106(1):18–23 35(8):1195–1204
14. Sheehan J, Yen CP, Arkha Y, Schlesinger D, Steiner L (2007) 34. Ivan ME, Sughrue ME, Clark AJ, Kane AJ, Aranda D, Barani IJ
Gamma knife surgery for trigeminal schwannoma. J Neurosurg et al (2011) A meta-analysis of tumor control rates and treatment-
106(5):839–845 related morbidity for patients with glomus jugulare tumors.
15. Peker S, Bayrakli F, Kilic T, Pamir MN (2007) Gamma-knife J Neurosurg 114(5):1299–1305
radiosurgery in the treatment of trigeminal schwannomas. Acta 35. Jackson CG, McGrew BM, Forest JA, Netterville JL, Hampf CF,
Neurochir (Wien) 149(11):1133–1137; discussion 7 Glasscock ME 3rd (2001) Lateral skull base surgery for glomus
16. Al-Mefty O, Fox JL, Rifai A, Smith RR (1987) A combined infra- tumors: long-term control. Otol Neurotol 22(3):377–382
temporal and posterior fossa approach for the removal of giant glo- 36. Karaman E, Yilmaz M, Isildak H, Hacizade Y, Korkut N,
mus tumors and chondrosarcomas. Surg Neurol 28(6):423–431 Devranoglu I et al (2010) Management of jugular paragangliomas
17. Gottfried ON, Liu JK, Couldwell WT (2004) Comparison of in otolaryngology practice. J Craniofac Surg 21(1):117–120
radiosurgery and conventional surgery for the treatment of glomus 37. Saringer W, Khayal H, Ertl A, Schoeggl A, Kitz K (2001)
jugulare tumors. Neurosurg Focus 17(2):E4 Efficiency of gamma knife radiosurgery in the treatment of glo-
18. Fayad JN, Keles B, Brackmann DE (2010) Jugular foramen mus jugulare tumors. Minim Invasive Neurosurg 44(3):141–146
tumors: clinical characteristics and treatment outcomes. Otol 38. Guss ZD, Batra S, Limb CJ, Li G, Sughrue ME, Redmond K et al
Neurotol 31(2):299–305 (2011) Radiosurgery of glomus jugulare tumors: a meta-analysis.
19. Semaan MT, Megerian CA (2008) Current assessment and man- Int J Radiat Oncol Biol Phys 81(4):e497–e502
agement of glomus tumors. Curr Opin Otolaryngol Head Neck 39. Healey JH, Lane JM (1989) Chordoma: a critical review of diag-
Surg 16(5):420–426 nosis and treatment. Orthop Clin North Am 20(3):417–426
20. Green JD Jr, Brackmann DE, Nguyen CD, Arriaga MA, Telischi FF, 40. Watkins L, Khudados ES, Kaleoglu M, Revesz T, Sacares P,
De la Cruz A (1994) Surgical management of previously untreated Crockard HA (1993) Skull base chordomas: a review of 38
glomus jugulare tumors. Laryngoscope 104(8 Pt 1):917–921 patients, 1958–88. Br J Neurosurg 7(3):241–248
21. Gstoettner W, Matula C, Hamzavi J, Kornfehl J, Czerny C (1999) 41. McMaster ML, Goldstein AM, Bromley CM, Ishibe N, Parry DM
Long-term results of different treatment modalities in 37 patients (2001) Chordoma: incidence and survival patterns in the United
with glomus jugulare tumors. Eur Arch Otorhinolaryngol States, 1973–1995. Cancer Causes Control 12(1):1–11
256(7):351–355 42. Sen C, Triana AI, Berglind N, Godbold J, Shrivastava RK (2010)
22. Gjuric M, Rudiger Wolf S, Wigand ME, Weidenbecher M (1996) Clival chordomas: clinical management, results, and complica-
Cranial nerve and hearing function after combined-approach sur- tions in 71 patients. J Neurosurg 113(5):1059–1071
gery for glomus jugulare tumors. Ann Otol Rhinol Laryngol 43. Crockard HA (1985) The transoral approach to the base of the brain
105(12):949–954 and upper cervical cord. Ann R Coll Surg Engl 67(5):321–325
23. Anand VK, Leonetti JP, al-Mefty O (1993) Neurovascular consid- 44. Walcott BP, Nahed BV, Mohyeldin A, Coumans JV, Kahle KT,
erations in surgery of glomus tumors with intracranial extensions. Ferreira MJ (2012) Chordoma: current concepts, management,
Laryngoscope 103(7):722–728 and future directions. Lancet Oncol 13(2):e69–e76
24. Huy PT, Kania R, Duet M, Dessard-Diana B, Mazeron JJ, 45. Cummings BJ, Hodson DI, Bush RS (1983) Chordoma: the results
Benhamed R (2009) Evolving concepts in the management of of megavoltage radiation therapy. Int J Radiat Oncol Biol Phys
jugular paraganglioma: a comparison of radiotherapy and surgery 9(5):633–642
in 88 cases. Skull Base 19(1):83–91 46. Stark AM, Mehdorn HM (2003) Images in clinical medicine.
25. Patel SJ, Sekhar LN, Cass SP, Hirsch BE (1994) Combined Chondroid clival chordoma. N Engl J Med 349(10):e10
approaches for resection of extensive glomus jugulare tumors. 47. Kitai R, Yoshida K, Kubota T, Sato K, Handa Y, Kasahara K et al
A review of 12 cases. J Neurosurg 80(6):1026–1038 (2005) Clival chordoma manifesting as nasal bleeding. A case
26. Sanna M, Jain Y, De Donato G, Rohit, Lauda L, Taibah A (2004) report. Neuroradiology 47(5):368–371
Management of jugular paragangliomas: the Gruppo Otologico 48. Levi AD, Kucharczyk W, Lang AP, Schutz H (1991) Clival chor-
experience. Otol Neurotol 25(5):797–804 doma presenting with acute brain stem hemorrhage. Can J Neurol
27. Whitfield PC, Grey P, Hardy DG, Moffat DA (1996) The surgical Sci 18(4):515–518
management of patients with glomus tumours of the skull base. Br 49. Di Maio S, Temkin N, Ramanathan D, Sekhar LN (2011) Current
J Neurosurg 10(4):343–350 comprehensive management of cranial base chordomas: 10-year
28. Watkins LD, Mendoza N, Cheesman AD, Symon L (1994) meta-analysis of observational studies. J Neurosurg 115(6):
Glomus jugulare tumours: a review of 61 cases. Acta Neurochir 1094–1105
(Wien) 130(1–4):66–70 50. Lanzino G, Dumont AS, Lopes MB, Laws ER Jr (2001) Skull base
29. Al-Mefty O, Teixeira A (2002) Complex tumors of the glomus chordomas: overview of disease, management options, and out-
jugulare: criteria, treatment, and outcome. J Neurosurg 97(6): come. Neurosurg Focus 10(3):E12
1356–1366 [Case Reports] 51. al-Mefty O, Borba LA (1997) Skull base chordomas: a manage-
30. Pareschi R, Righini S, Destito D, Raucci AF, Colombo S (2003) ment challenge. J Neurosurg 86(2):182–189
Surgery of Glomus Jugulare Tumors. Skull Base 13(3):149–157 52. Tzortzidis F, Elahi F, Wright D, Natarajan SK, Sekhar LN (2006)
31. Liu JK, Sameshima T, Gottfried ON, Couldwell WT, Fukushima Patient outcome at long-term follow-up after aggressive microsur-
T (2006) The combined transmastoid retro- and infralabyrinthine gical resection of cranial base chordomas. Neurosurgery
transjugular transcondylar transtubercular high cervical approach 59(2):230–237; discussion 7
53. Komotar RJ, Starke RM, Raper DM, Anand VK, Schwartz TH 73. Iyer A, Kano H, Kondziolka D, Liu X, Niranjan A, Flickinger JC
(2011) The endoscope-assisted ventral approach compared with et al (2012) Stereotactic radiosurgery for intracranial chondrosar-
open microscope-assisted surgery for clival chordomas. World coma. J Neurooncol 108(3):535–542
Neurosurg 76(3–4):318–327; discussion 259–262 74. Dulguerov P, Allal AS, Calcaterra TC (2001) Esthesioneuroblastoma:
54. Jian BJ, Bloch OG, Yang I, Han SJ, Aranda D, Tihan T et al (2010) a meta-analysis and review. Lancet Oncol 2(11):683–690
Adjuvant radiation therapy and chondroid chordoma subtype are 75. Smee RI, Broadley K, Williams JR, Meagher NS, Bridger GP
associated with a lower tumor recurrence rate of cranial chor- (2011) Retained role of surgery for olfactory neuroblastoma. Head
doma. J Neurooncol 98(1):101–108 Neck 33(10):1486–1492
55. Amichetti M, Cianchetti M, Amelio D, Enrici RM, Minniti G 76. Lund VJ, Howard D, Wei W, Spittle M (2003) Olfactory neuroblas-
(2009) Proton therapy in chordoma of the base of the skull: a sys- toma: past, present, and future? Laryngoscope 113(3):502–507
tematic review. Neurosurg Rev 32(4):403–416 77. Theilgaard SA, Buchwald C, Ingeholm P, Kornum Larsen S,
56. Brada M, Pijls-Johannesma M, De Ruysscher D (2007) Proton Eriksen JG, Sand HH (2003) Esthesioneuroblastoma: a Danish
therapy in clinical practice: current clinical evidence. J Clin Oncol demographic study of 40 patients registered between 1978 and
25(8):965–970 2000. Acta Otolaryngol 123(3):433–439
57. Cianfriglia F, Pompili A, Occhipinti E (1978) Intracranial malig- 78. Thompson LD (2009) Olfactory neuroblastoma. Head Neck
nant cartilaginous tumours. Report of two cases and review of lit- Pathol 3(3):252–259
erature. Acta Neurochir (Wien) 45(1–2):163–175 79. Levine PA, Gallagher R, Cantrell RW (1999) Esthesioneuroblastoma:
58. Bloch OG, Jian BJ, Yang I, Han SJ, Aranda D, Ahn BJ et al (2009) reflections of a 21-year experience. Laryngoscope 109(10):
A systematic review of intracranial chondrosarcoma and survival. 1539–1543
J Clin Neurosci 16(12):1547–1551 80. Oskouian RJ Jr, Jane JA Sr, Dumont AS, Sheehan JM, Laurent JJ,
59. Lau DP, Wharton SB, Antoun NM, Bottrill ID, Moffat DA (1997) Levine PA (2002) Esthesioneuroblastoma: clinical presentation,
Chondrosarcoma of the petrous apex. Dilemmas in diagnosis and radiological, and pathological features, treatment, review of the
treatment. J Laryngol Otol 111(4):368–371 literature, and the University of Virginia experience. Neurosurg
60. Coltrera MD, Googe PB, Harrist TJ, Hyams VJ, Schiller AL, Focus 12(5):e4
Goodman ML (1986) Chondrosarcoma of the temporal bone. 81. Cantrell RW, Ghorayeb BY, Fitz-Hugh GS (1977)
Diagnosis and treatment of 13 cases and review of the literature. Esthesioneuroblastoma: diagnosis and treatment. Ann Otol Rhinol
Cancer 58(12):2689–2696 Laryngol 86(6 Pt 1):760–765
61. Seidman MD, Nichols RD, Raju UB, Mehta B, Levy 82. Shah JP, Kraus DH, Bilsky MH, Gutin PH, Harrison LH, Strong EW
HG. Extracranial skull base chondrosarcoma. Ear Nose Throat (1997) Craniofacial resection for malignant tumors involving the ante-
J. 1989;68(8):626–32, 35. rior skull base. Arch Otolaryngol Head Neck Surg 123(12):
62. Bloch OG, Jian BJ, Yang I, Han SJ, Aranda D, Ahn BJ et al (2010) 1312–1317
Cranial chondrosarcoma and recurrence. Skull Base 20(3): 83. Bilsky MH, Kraus DH, Strong EW, Harrison LB, Gutin PH,
149–156 Shah JP (1997) Extended anterior craniofacial resection for
63. Neff B, Sataloff RT, Storey L, Hawkshaw M, Spiegel JR (2002) intracranial extension of malignant tumors. Am J Surg 174(5):
Chondrosarcoma of the skull base. Laryngoscope 112(1): 565–568
134–139 84. Eden BV, Debo RF, Larner JM, Kelly MD, Levine PA, Stewart FM
64. Gay E, Sekhar LN, Rubinstein E, Wright DC, Sen C, Janecka IP et al (1994) Esthesioneuroblastoma. Long-term outcome and pat-
et al (1995) Chordomas and chondrosarcomas of the cranial base: terns of failure—the University of Virginia experience. Cancer
results and follow-up of 60 patients. Neurosurgery 36(5):887– 73(10):2556–2562
896; discussion 96–97 85. Aboziada MA, Eisbruch A (2010) Esthesioneuroblastoma: the
65. Radner H, Katenkamp D, Reifenberger G, Dekert M, Pietsch T, role of post-operative irradiation after complete surgical resection.
Wiestler OD (2001) New developments in the pathology of skull J Egypt Natl Canc Inst 22(2):143–148
base tumors. Virchows Arch 438(4):321–335 86. Ward PD, Heth JA, Thompson BG, Thompson BG, Marentette LJ
66. Wanebo JE, Bristol RE, Porter RR, Coons SW, Spetzler RF (2006) (2009) Esthesioneuroblastoma: results and outcomes of a single
Management of cranial base chondrosarcomas. Neurosurgery institution’s experience. Skull Base 19(2):133–140
58(2):249–255; discussion 55 87. Nichols AC, Chan AW, Curry WT, Barker FG, Deschler DG, Lin
67. Gonzalez LF, Crawford NR, Horgan MA, Deshmukh P, Zabramski DT (2008) Esthesioneuroblastoma: the Massachusetts eye and ear
JM, Spetzler RF (2002) Working area and angle of attack in three infirmary and Massachusetts general hospital experience with cra-
cranial base approaches: pterional, orbitozygomatic, and maxil- niofacial resection, proton beam radiation, and chemotherapy.
lary extension of the orbitozygomatic approach. Neurosurgery Skull Base 18(5):327–337
50(3):550–555; discussion 5–7 88. Porter AB, Bernold DM, Giannini C, Foote RL, Link MJ, Olsen
68. Honeybul S, Neil-Dwyer G, Lees PD, Evans BT, Lang DA (1996) KD et al (2008) Retrospective review of adjuvant chemotherapy
The orbitozygomatic infratemporal fossa approach: a quantitative for esthesioneuroblastoma. J Neurooncol 90(2):201–204
anatomical study. Acta Neurochir (Wien) 138(3):255–264 89. McLean JN, Nunley SR, Klass C, Moore C, Muller S, Johnstone
69. Qiu QH, Liang MZ, Liu H, Chen SH, Zhang HB, Zhang QH PA (2007) Combined modality therapy of esthesioneuroblastoma.
(2010) [Nasal endoscopic surgical treatment for chondrosarcoma Otolaryngol Head Neck Surg 136(6):998–1002
of paranasal sinus and the skull base]. Zhonghua Er Bi Yan Hou 90. Soler ZM, Smith TL (2012) Endoscopic versus open craniofacial
Tou Jing Wai Ke Za Zhi 45(7):551–554 resection of esthesioneuroblastoma: what is the evidence?
70. Zanation AM, Snyderman CH, Carrau RL, Gardner PA, Prevedello Laryngoscope 122(2):244–245
DM, Kassam AB (2009) Endoscopic endonasal surgery for 91. Monteiro EM, Lopes MG, Santos ER, Diniz CV, Albuquerque
petrous apex lesions. Laryngoscope 119(1):19–25 AS, Monteiro AP et al (2011) Endoscopic treatment of esthesio-
71. Zhang Q, Kong F, Yan B, Ni Z, Liu H (2008) Endoscopic endona- neuroblastoma. Braz J Otorhinolaryngol 77(2):171–177
sal surgery for clival chordoma and chondrosarcoma. ORL J 92. Kodama S, Kawano T, Suzuki M (2009) Endoscopic transnasal
Otorhinolaryngol Relat Spec 70(2):124–129 resection of ectopic esthesioneuroblastoma in the pterygopalatine
72. Munzenrider JE, Liebsch NJ (1999) Proton therapy for tumors of fossa: technical case report. Neurosurgery 65(6 Suppl):E112–
the skull base. Strahlenther Onkol 175(Suppl 2):57–63 E113; discussion E3
93. Folbe A, Herzallah I, Duvvuri U, Bublik M, Sargi Z, Snyderman 113. Karavitaki N, Wass JA. Craniopharyngiomas. Endocrinol Metab
CH et al (2009) Endoscopic endonasal resection of esthesioneu- Clin North Am. 2008;37(1):173–93, ix-x.
roblastoma: a multicenter study. Am J Rhinol Allergy 114. Müller HL (2013) Childhood craniopharyngioma. Pituitary
23(1):91–94 16(1):56–67
94. Gallia GL, Reh DD, Salmasi V, Blitz AM, Koch W, Ishii M (2011) 115. Van Effenterre R, Boch AL (2002) Craniopharyngioma in adults
Endonasal endoscopic resection of esthesioneuroblastoma: the and children: a study of 122 surgical cases. J Neurosurg 97(1):3–11
Johns Hopkins Hospital experience and review of the literature. 116. Minamida Y, Mikami T, Hashi K, Houkin K (2005) Surgical man-
Neurosurg Rev 34(4):465–475 agement of the recurrence and regrowth of craniopharyngiomas.
95. Casiano RR, Numa WA, Falquez AM (2001) Endoscopic resec- J Neurosurg 103(2):224–232
tion of esthesioneuroblastoma. Am J Rhinol 15(4):271–279 117. Leng LZ, Greenfield JP, Souweidane MM, Anand VK, Schwartz
96. Unger F, Haselsberger K, Walch C, Stammberger H, Papaefthymiou TH (2012) Endoscopic, endonasal resection of craniopharyngio-
G (2005) Combined endoscopic surgery and radiosurgery as treat- mas: analysis of outcome including extent of resection, cerebro-
ment modality for olfactory neuroblastoma (esthesioneuroblas- spinal fluid leak, return to preoperative productivity, and body
toma). Acta Neurochir (Wien) 147(6):595–601; discussion 2 mass index. Neurosurgery 70(1):110–123; discussion 23–24
97. Unger F, Walch C, Stammberger H, Papaefthymiou G, 118. Lin LL, El Naqa I, Leonard JR, Park TS, Hollander AS, Michalski
Haselsberger K, Pendl G (2001) Olfactory neuroblastoma (esthe- JM et al (2008) Long-term outcome in children treated for cranio-
sioneuroblastoma): report of six cases treated by a novel combina- pharyngioma with and without radiotherapy. J Neurosurg Pediatr
tion of endoscopic surgery and radiosurgery. Minim Invasive 1(2):126–130
Neurosurg 44(2):79–84 119. De Vile CJ, Grant DB, Kendall BE, Neville BG, Stanhope R,
98. Walch C, Stammberger H, Unger F, Anderhuber W (2000) A new Watkins KE et al (1996) Management of childhood craniopharyn-
therapy concept in esthesioneuroblastoma. Laryngorhinootologie gioma: can the morbidity of radical surgery be predicted? J
79(12):743–748 Neurosurg 85(1):73–81
99. Prabhu VC, Brown HG (2005) The pathogenesis of craniopharyn- 120. Matson DD, Crigler JF Jr (1969) Management of craniopharyn-
giomas. Childs Nerv Syst 21(8–9):622–627 gioma in childhood. J Neurosurg 30(4):377–390
100. Stamm AC, Vellutini E, Balsalobre L. Craniopharyngioma. 121. Liu JK, Christiano LD, Gupta G, Carmel PW (2010) Surgical
Otolaryngol Clin North Am. 2011;44(4):937–52, viii. nuances for removal of retrochiasmatic craniopharyngiomas via
101. Bunin GR, Surawicz TS, Witman PA, Preston-Martin S, Davis F, the transbasal subfrontal translamina terminalis approach.
Bruner JM (1998) The descriptive epidemiology of craniopharyn- Neurosurg Focus 28(4):E6
gioma. J Neurosurg 89(4):547–551 122. Maira G, Anile C, Albanese A, Cabezas D, Pardi F, Vignati A
102. Caldarelli M, Massimi L, Tamburrini G, Cappa M, Di Rocco C (2004) The role of transsphenoidal surgery in the treatment of cra-
(2005) Long-term results of the surgical treatment of craniopha- niopharyngiomas. J Neurosurg 100(3):445–451
ryngioma: the experience at the Policlinico Gemelli, Catholic 123. Couldwell WT, Weiss MH, Rabb C, Liu JK, Apfelbaum RI,
University, Rome. Childs Nerv Syst 21(8–9):747–757 Fukushima T (2004) Variations on the standard transsphenoidal
103. Laws ER Jr (1980) Transsphenoidal microsurgery in the manage- approach to the sellar region, with emphasis on the extended
ment of craniopharyngioma. J Neurosurg 52(5):661–666 approaches and parasellar approaches: surgical experience in 105
104. Fahlbusch R, Honegger J, Paulus W, Huk W, Buchfelder M (1999) cases. Neurosurgery 55(3):539–547; discussion 47–50
Surgical treatment of craniopharyngiomas: experience with 168 124. Liu JK, Eloy JA. Endoscopic endonasal transplanum transtubercu-
patients. J Neurosurg 90(2):237–250 lum approach for resection of retrochiasmatic craniopharyngioma.
105. Nielsen EH, Feldt-Rasmussen U, Poulsgaard L, Kristensen LO, J Neurosurg. 2012;32 Suppl:E2.
Astrup J, Jorgensen JO et al (2011) Incidence of craniopharyngioma 125. Pemberton LS, Dougal M, Magee B, Gattamaneni HR (2005)
in Denmark (n = 189) and estimated world incidence of craniopha- Experience of external beam radiotherapy given adjuvantly or at
ryngioma in children and adults. J Neurooncol 104(3):755–763 relapse following surgery for craniopharyngioma. Radiother
106. Weiner HL, Wisoff JH, Rosenberg ME, Kupersmith MJ, Cohen H, Oncol 77(1):99–104
Zagzag D et al (1994) Craniopharyngiomas: a clinicopathological 126. Yang I, Sughrue ME, Rutkowski MJ, Kaur R, Ivan ME, Aranda D
analysis of factors predictive of recurrence and functional out- et al (2010) Craniopharyngioma: a comparison of tumor control
come. Neurosurgery 35(6):1001–1010; discussion 10–11 with various treatment strategies. Neurosurg Focus 28(4):E5
107. Adamson TE, Wiestler OD, Kleihues P, Yasargil MG (1990) 127. Veeravagu A, Lee M, Jiang B, Chang SD (2010) The role of radio-
Correlation of clinical and pathological features in surgically surgery in the treatment of craniopharyngiomas. Neurosurg Focus
treated craniopharyngiomas. J Neurosurg 73(1):12–17 28(4):E11
108. Muller HL, Bruhnken G, Emser A, Faldum A, Etavard-Gorris N, 128. Lee M, Kalani MY, Cheshier S, Gibbs IC, Adler JR, Chang SD
Gebhardt U et al (2005) Longitudinal study on quality of life in (2008) Radiation therapy and CyberKnife radiosurgery in the
102 survivors of childhood craniopharyngioma. Childs Nerv Syst management of craniopharyngiomas. Neurosurg Focus 24(5):E4
21(11):975–980 129. Di Maio S, Ramanathan D, Garcia-Lopez R, Rocha MH, Guerrero
109. Puget S, Garnett M, Wray A, Grill J, Habrand JL, Bodaert N et al FP, Ferreira M Jr et al (2012) Evolution and future of skull base
(2007) Pediatric craniopharyngiomas: classification and treatment surgery: the paradigm of skull base meningiomas. World
according to the degree of hypothalamic involvement. J Neurosurg Neurosurg 78(3–4):260–275
106(1 Suppl):3–12 130. Komotar RJ, Starke RM, Raper DM, Anand VK, Schwartz TH
110. Merchant TE, Kiehna EN, Sanford RA, Mulhern RK, Thompson (2012) Endoscopic endonasal versus open transcranial resection
SJ, Wilson MW et al (2002) Craniopharyngioma: the St. Jude of anterior midline skull base meningiomas. World Neurosurg
Children’s Research Hospital experience 1984–2001. Int J Radiat 77(5–6):713–724
Oncol Biol Phys 53(3):533–542 131. Ichinose T, Goto T, Ishibashi K, Takami T, Ohata K (2010) The
111. Hoffman HJ, De Silva M, Humphreys RP, Drake JM, Smith ML, role of radical microsurgical resection in multimodal treatment for
Blaser SI (1992) Aggressive surgical management of craniopha- skull base meningioma. J Neurosurg 113(5):1072–1078
ryngiomas in children. J Neurosurg 76(1):47–52 132. Couldwell WT, Fukushima T, Giannotta SL, Weiss MH (1996)
112. Baskin DS, Wilson CB (1986) Surgical management of cranio- Petroclival meningiomas: surgical experience in 109 cases. J
pharyngiomas. A review of 74 cases. J Neurosurg 65(1):22–27 Neurosurg 84(1):20–28
133. Jung HW, Yoo H, Paek SH, Choi KS (2000) Long-term outcome and with optic canal involvement using the endoscopic endonasal
growth rate of subtotally resected petroclival meningiomas: experi- extended transsphenoidal transplanum transtuberculum approach.
ence with 38 cases. Neurosurgery 46(3):567–574; discussion 74–75 Neurosurg Focus 30(5):E2
134. Kano T, Kawase T, Horiguchi T, Yoshida K (2010) Meningiomas 155. Aydin S, Cavallo LM, Messina A, Dal Fabbro M, Cappabianca P,
of the ventral foramen magnum and lower clivus: factors influenc- Barlas O et al (2007) The endoscopic endonasal trans-sphenoidal
ing surgical morbidity, the extent of tumour resection, and tumour approach to the sellar and suprasellar area. Anatomic study.
recurrence. Acta Neurochir (Wien) 152(1):79–86; discussion 86 J Neurosurg Sci 51(3):129–138
135. Natarajan SK, Sekhar LN, Schessel D, Morita A (2007) Petroclival 156. Cavallo LM, de Divitiis O, Aydin S, Messina A, Esposito F,
meningiomas: multimodality treatment and outcomes at long- Iaconetta G et al (2007) Extended endoscopic endonasal transs-
term follow-up. Neurosurgery 60(6):965–979; discussion 79–81 phenoidal approach to the suprasellar area: anatomic consider-
136. Obeid F, Al-Mefty O (2003) Recurrence of olfactory groove ations—part 1. Neurosurgery 61(3 Suppl):24–33; discussion 4
meningiomas. Neurosurgery 53(3):534–542; discussion 42–43 157. de Divitiis E, Cavallo LM, Cappabianca P, Esposito F (2007) Extended
137. Fisch U (1978) Infratemporal fossa approach to tumours of the endoscopic endonasal transsphenoidal approach for the removal of
temporal bone and base of the skull. J Laryngol Otol suprasellar tumors: Part 2. Neurosurgery 60(1):46–58; discussion 9
92(11):949–967 158. Heth JA, Al-Mefty O (2003) Cavernous sinus meningiomas.
138. Samii M, Tatagiba M, Carvalho GA (1999) Resection of large pet- Neurosurg Focus 14(6):e3
roclival meningiomas by the simple retrosigmoid route. J Clin 159. Walsh MT, Couldwell WT (2009) Management options for cav-
Neurosci 6(1):27–30 ernous sinus meningiomas. J Neurooncol 92(3):307–316
139. Jane JA, Park TS, Pobereskin LH, Winn HR, Butler AB (1982) 160. Long DM (2001) The treatment of meningiomas in the region of
The supraorbital approach: technical note. Neurosurgery the cavernous sinus. Childs Nerv Syst 17(3):168–172
11(4):537–542 161. Pichierri A, Santoro A, Raco A, Paolini S, Cantore G, Delfini R
140. Yasargil MG, Reichman MV, Kubik S (1987) Preservation of the (2009) Cavernous sinus meningiomas: retrospective analysis and
frontotemporal branch of the facial nerve using the interfascial proposal of a treatment algorithm. Neurosurgery 64(6):1090–
temporalis flap for pterional craniotomy. Technical article. 1099; discussion 9–101
J Neurosurg 67(3):463–466 162. De Jesus O, Sekhar LN, Parikh HK, Wright DC, Wagner DP
141. Hakuba A, Liu S, Nishimura S (1986) The orbitozygomatic infra- (1996) Long-term follow-up of patients with meningiomas involv-
temporal approach: a new surgical technique. Surg Neurol ing the cavernous sinus: recurrence, progression, and quality of
26(3):271–276 life. Neurosurgery 39(5):915–919; discussion 9–20
142. Sekhar LN, Schramm VL Jr, Jones NF (1987) Subtemporal- 163. Sekhar LN, Patel S, Cusimano M, Wright DC, Sen CN, Bank WO
preauricular infratemporal fossa approach to large lateral and pos- (1996) Surgical treatment of meningiomas involving the cavern-
terior cranial base neoplasms. J Neurosurg 67(4):488–499 ous sinus: evolving ideas based on a ten year experience. Acta
143. Al-Mefty O, Smith RR (1988) Surgery of tumors invading the Neurochir Suppl 65:58–62
cavernous sinus. Surg Neurol 30(5):370–381 164. Hasegawa T, Kida Y, Yoshimoto M, Koike J, Iizuka H, Ishii D
144. Kawase T, Shiobara R, Toya S (1991) Anterior transpetrosal- (2007) Long-term outcomes of Gamma Knife surgery for cavern-
transtentorial approach for sphenopetroclival meningiomas: surgi- ous sinus meningioma. J Neurosurg 107(4):745–751
cal method and results in 10 patients. Neurosurgery 28(6):869–875; 165. Ware ML, Pravdenkova S, Erkmen K, Al-Mefty O (2009)
discussion 75–76 Petroclival and upper clival meningiomas I: an overview of surgi-
145. Samii M, Gerganov V, Giordano M, Samii A (2010) Two step cal approaches. In: Lee JH (ed) Meningiomas: diagnosis, treat-
approach for surgical removal of petroclival meningiomas with ment, and outcome. Springer, London
large supratentorial extension. Neurosurg Rev 34(2):173–179 166. Nanda A, Javalkar V, Banerjee AD (2011) Petroclival meningio-
146. Cho CW, Al-Mefty O (2002) Combined petrosal approach to pet- mas: study on outcomes, complications and recurrence rates.
roclival meningiomas. Neurosurgery 51(3):708–716; discussion J Neurosurg 114(5):1268–1277
16–18 167. Van Havenbergh T, Carvalho G, Tatagiba M, Plets C, Samii M
147. Babu RP, Sekhar LN, Wright DC (1994) Extreme lateral transcon- (2003) Natural history of petroclival meningiomas. Neurosurgery
dylar approach: technical improvements and lessons learned. 52(1):55–62; discussion 4
J Neurosurg 81(1):49–59 168. Al-Mefty O, Sekhar LN, Sen C, van Loveren HR (2001) Petroclival
148. Sekhar LN, Babu RP, Wright DC (1994) Surgical resection of cra- meningioma: case history and responses. Skull Base 11(2):143–148
nial base meningiomas. Neurosurg Clin N Am 5(2):299–330 169. Tatagiba M, Samii M, Matthies C, Vorkapic P (1996) Management
149. Sen CN, Sekhar LN (1990) An extreme lateral approach to intra- of petroclival meningiomas: a critical analysis of surgical treat-
dural lesions of the cervical spine and foramen magnum. ment. Acta Neurochir Suppl 65:92–94
Neurosurgery 27(2):197–204 170. Ammirati M, Samii M (1992) Presigmoid sinus approach to petro-
150. Laufer I, Anand VK, Schwartz TH (2007) Endoscopic, endonasal clival meningiomas. Skull Base Surg 2(3):124–128
extended transsphenoidal, transplanum transtuberculum approach 171. Samii M, Gerganov VM (2011) Petroclival meningiomas: quo
for resection of suprasellar lesions. J Neurosurg 106(3):400–406 vadis? World Neurosurg 75(3–4):424
151. Schwartz TH, Anand VK (2007) The endoscopic endonasal trans- 172. Al-Mefty O (1991) Meningiomas. Raven Press, New York
sphenoidal approach to the suprasellar cistern. Clin Neurosurg 173. Mahmoud M, Nader R, Al-Mefty O. Optic canal involvement in
54:226–235 tuberculum sellae meningiomas: influence on approach, recur-
152. Gardner PA, Kassam AB, Thomas A, Snyderman CH, Carrau RL, rence, and visual recovery. Neurosurgery. 2010;67(3 Suppl
Mintz AH et al (2008) Endoscopic endonasal resection of anterior Operative):ons108–18; discussion ons18–19.
cranial base meningiomas. Neurosurgery 63(1):36–52; discussion 4 174. Nozaki K, Kikuta K, Takagi Y, Mineharu Y, Takahashi JA,
153. Liu JK, Christiano LD, Patel SK, Tubbs RS, Eloy JA (2011) Hashimoto N (2008) Effect of early optic canal unroofing on the
Surgical nuances for removal of olfactory groove meningiomas outcome of visual functions in surgery for meningiomas of the
using the endoscopic endonasal transcribriform approach. tuberculum sellae and planum sphenoidale. Neurosurgery
Neurosurg Focus 30(5):E3 62(4):839–844; discussion 44–46
154. Liu JK, Christiano LD, Patel SK, Tubbs RS, Eloy JA (2011) 175. Attia M, Kandasamy J, Jakimovski D, Bedrosian J, Alimi M, Lee
Surgical nuances for removal of tuberculum sellae meningiomas DL, et al. The importance and timing of optic canal exploration
and decompression during endoscopic endonasal resection of route? A series of 51 consecutive cases. Neurosurgery 62(3):556–
tuberculum sella and planum sphenoidale meningiomas. 563; discussion 63
Neurosurgery. 2012;71(1 Suppl Operative):ons58–67. 181. Bowers CA, Altay T, Couldwell WT (2011) Surgical decision-
176. Schick U, Hassler W (2005) Surgical management of tuberculum making strategies in tuberculum sellae meningioma resection.
sellae meningiomas: involvement of the optic canal and visual Neurosurg Focus 30(5):E1
outcome. J Neurol Neurosurg Psychiatry 76(7):977–983 182. Kassam AB, Thomas A, Carrau RL, Snyderman CH, Vescan A,
177. Otani N, Muroi C, Yano H, Khan N, Pangalu A, Yonekawa Y (2006) Prevedello D et al (2008) Endoscopic reconstruction of the cranial
Surgical management of tuberculum sellae meningioma: role of selec- base using a pedicled nasoseptal flap. Neurosurgery 63(1 Suppl
tive extradural anterior clinoidectomy. Br J Neurosurg 20(3):129–138 1):ONS44–ONS52; discussion ONS3
178. Sade B, Lee JH (2009) High incidence of optic canal involvement 183. Hadad G, Bassagasteguy L, Carrau RL, Mataza JC, Kassam
in tuberculum sellae meningiomas: rationale for aggressive skull A, Snyderman CH et al (2006) A novel reconstructive
base approach. Surg Neurol 72(2):118–123; discussion 23 technique after endoscopic expanded endonasal approaches:
179. Chai Y, Yamazaki H, Kondo A, Oshitari T, Yamamoto S (2012) Case vascular pedicle nasoseptal flap. Laryngoscope 116(10):
of acute optic nerve compression caused by tuberculum sellae menin- 1882–1886
gioma with optic canal involvement. Clin Ophthalmol 6:661–666 184. Starke RM, Williams BJ, Hiles C, Nguyen JH, Elsharkawy MY,
180. de Divitiis E, Esposito F, Cappabianca P, Cavallo LM, de Divitiis Sheehan JP (2012) Gamma knife surgery for skull base meningio-
O (2008) Tuberculum sellae meningiomas: high route or low mas. J Neurosurg 116(3):588–597
Skull Base Tumors:
Viewpoint—Fractionated Radiotherapy 40
or Stereotactic Radiotherapy
René-O. Mirimanoff and Laura Negretti
Introduction Chordomas and Chondrosarcomas
Skull base tumors account for up to a third of all intracranial Chordomas: General Features
tumors. The most common ones are represented by menin-
gioma, pituitary adenoma, acoustic nerve schwannoma, and Chordomas are rare tumors: according to the SEER data, the
craniopharyngioma; less frequent types comprise chordoma incidence of these tumors is around 0.08 per 100,000 with a
and chondrosarcoma, PGL or chemodectoma, and very occa- peak incidence between 50 and 60 years of age; they repre-
sionally rare bone tumors like bone lymphoma, giant cell sent only 0.1–0.2 % of all primary intracranial neoplasms
tumors, and carcinoma of the skull base can be found. Some, [2]. Chordomas are of neuroectodermal origin and are pre-
like the meningioma, can involve almost any part of the ante- sumed to develop from notochordal remnants (Fig. 40.1).
rior, middle, or posterior fossa, whereas others, like the PGL, The discovery of a gene duplication in the transcription
arise from well-defined anatomic structures. Surgery still factor T gene (brachyury) in familial chordoma seems to
remains the cornerstone in the management of most skull support this hypothesis [3]. There are several histopathologic
base tumors [1]. However, in spite of a better understanding subcategories, with or without predominance of particular
of the natural behavior and biology of these skull base tumors, cell types. Those include the physaliferous type, the epitheli-
continuous progress in anesthesiology, and the refinements in oid and the chondroid cell types [4].
surgical techniques, a complete surgical removal is extremely Immunohistochemistry shows that almost all tumors stain
difficult in many instances or made at a price of considerable positively for the epithelial membrane antigen, cytokeratins,
neurologic deficits. This is why noninvasive treatments, such S-100 proteins, and are negative for HMB-45 and desmin
as high-precision radiotherapy with intensity-modulated radi- [4]. One third of all chordomas arise from the skull base, the
ation therapy (IMRT), radiosurgery (RS), stereotactic frac- remainder from the sacrum and the spine. Although they do
tionated radiotherapy (STRT), and particle therapy, play an grow slowly, and rarely metastasize, they are quite often
increasing role in the multidisciplinary management of most lethal due to their local progression. They commonly involve
skull base tumors. We will focus here on chordoma and chon- vital neural or bony structures (Fig. 40.2), thus compromis-
drosarcoma on the one hand, and on the PGL category on the ing the effectiveness of surgical or radiation therapy [5].
other, with particular emphasis on their treatment by fraction- Intracranial chordomas typically arise from the clivus and
ated external beam radiotherapy, particle therapy, and STRT. can invade the dura, extend in any direction, for example
toward the foramen magnum, the petrous bone, the cranio-
spinal junction or compress the brain stem, or infiltrate ante-
R.-O. Mirimanoff, M.D. (*) riorly the cavernous sinus [6].
Radiation Oncology Department,
CHUV, University Hospital Lausanne, Lausanne, Switzerland
Clinique de La Source, Vinet Avenue 30, Chondrosarcomas: General Features
1004 Lausanne, Switzerland
e-mail: rene-olivier.mirimanoff@chuv.ch
Chondrosarcomas are also rare tumors that can arise from
L. Negretti, M.D.
the skull base. Like chordomas, they represent 0.15 % of
Radiation Oncology Department, CHUV, University Hospital
Lausanne, Clinica Luganese, via Moncucco 10, 6900 Lugano, all intracranial tumors [7] although many series suggest
Ticino, Switzerland that they are even less common than chordomas [8–10].
518 R.-O. Mirimanoff and L. Negretti
Fig. 40.1 Pathology slide of a chordoma Fig. 40.3 Pathology slide of a skull base chondrosarcoma
On immunohistochemistry, almost all chondrosarcomas

stain for S-100 protein, but none for keratin and less than
10 % for epithelial membrane antigens [18]. Thus, tumor
markers are helpful adjuncts to differentiate between chor-
doma and chondrosarcoma: chordomas express cytokeratin
and epithelial membrane antigens, whereas chondrosarco-
mas lack the former and rarely stain for the latter. Like chor-
domas, chondrosarcomas tend to invade local structures.
Distant metastases are uncommon; Hassounah et al. in their
literature review found five cases of metastases out of 50
reported patients [19].
Standard Treatment of Chordomas

and Chondrosarcomas
Although they have a good number of features in common

and their overall management is relatively similar, chordoma
and chondrosarcoma of the skull base present with important
Fig. 40.2 MRI of a skull base chordoma differences in histopathology, biological behavior, and out-
come [6, 18, 20, 21]. For example, in an important series of
109 patients, recurrence and death rates were much higher in
Most chondrosarcomas arise de novo although they can patients with chordomas than in patients with chondrosarco-
occur in Ollier’s disease, Paget’s disease, or in osteochondromas [6]. Two excellent papers were recently published on
mas. It is thought that they originate from primitive mesen- chordomas [20] and chondrosarcomas [21], which review
chymal cells or embryonal remnants of the cartilaginous the management and outcome of these tumors. There is no
matrix of the cranium [11] (Fig. 40.3). They are sometimes question that surgery remains the cornerstone of treatment.
mistakenly diagnosed as chordoma, but it is quite important The surgeon’s goal should be, whenever possible, to carry
to distinguish them from the latter as they have a better prog- out an en bloc resection with a gross total removal [6, 22].
nosis [6, 8, 12–16]. Several subtypes have been proposed Major advances in surgical techniques and microsurgery in
that include the hyaline, the myxoid, and mixed hyaline- particular have allowed neurosurgeons to perform more mac-
myxoid; the mesenchymal and dedifferentiated subtypes are roscopically complete resections, with clearly improved out-
less common and disclose a more aggressive behavior, comes [22]. However en bloc resections are rarely possible
whereas the clear cell subtype is extremely rare [17, 18]. with clival chordoma and thus a maximally safe aggressive
40 Skull Base Tumors: Viewpoint—Fractionated Radiotherapy or Stereotactic Radiotherapy 519
surgery with neurological preservation should be the goal chondrosarcomas, followed by male chordomas and female
[20]. With extensive surgery, Crockard et al. have reported chordomas, with 94 %, 65 %, and 42 % local control rates,
5-year survival rates of 77 % and 93 %, respectively, for respectively [14]. At Loma Linda University, 58 patients
chordoma and chondrosarcoma [15, 16]. Gay et al. obtained with chordomas and chondrosarcomas received proton beam
a 67 % total or near total resection rate, and for these an 84 % radiotherapy with a mean dose of 70 CGE, with a local con-
recurrence-free rate was achieved [8]. trol rate of 92 % for chondrosarcomas and 78 % for chordo-
In spite of these improvements in surgery, it appears that mas [13]. An update of the French series at Orsay on 100
a genuine curative resection with clear margins is not often patients confirmed their previous results with a combination
realistic, and ultimately a majority of patients will recur and of photons and proton beams (201 MV with a median dose of
die from their disease if no adjuvant radiation is adminis- 67 CGE). In this report, the 2- and 4-year local control rates
tered. For example, in a series of 13 operated patients with were 86 % and 54 %, respectively [9]. Patients with chordo-
chondrosarcomas from the Netherlands, only three received mas treated with proton beam therapy at Tsukuba University
postoperative photon (2) or proton (1) radiotherapy, and the received a median dose of 72 Gy with 5-year local control
recurrence-free survival was only 43 % at 5 years [11]. In a and local-specific survivals of 42 % and 72 %, respectively
systematic literature review of intracranial chondrosarcoma, [27]. At the Paul Scherrer Institute, Switzerland, 64 patients
Bloch et al. found that the 5-year mortality rate was 25 % with skull base chordomas (42) and chondrosarcomas (22)
without and 9 % with postoperative radiotherapy [21]. were treated with the spot-scanning-based proton radiother-
Therefore, adjuvant, high-dose postoperative radiotherapy technique with a median dose of 75.5 Gy for chordomas
apy has been considered and used for many years both for and 68.4 Gy for chondrosarcomas [28]. Actuarial 5-year
chordomas and chondrosarcomas. In spite of old claims that control rates were 81 % for chordomas and 94 % for chon-
chondrosarcomas and chordomas are radioresistant tumors drosarcomas; the 5-year freedom from high-grade toxicity
[23], conventional external beam radiotherapy was shown to was 94 % [28]. With carbon ion radiotherapy, the experience
provide useful and prolonged palliation in overt residual dis- from Heidelberg and GSI in Darmstadt on 96 chordoma
ease [5]. It was however suggested for a long time that these patients disclosed a 70 % local control and 88 % survival at
tumors, in order to be controlled, need to receive relatively 5 years [29]. The same group obtained a 90 % 4-year local
high-dose radiotherapy and that the total dose should be control and 98 % 5-year survival in a series of 54 patients
increased to 70–80 Gy [24–27]. Currently, with high- with chondrosarcomas [30]. In the experience with proton
precision external beam photon radiotherapy, the dose to the beam and carbon ion radiotherapy, late grade 3–4 toxicities
PTV can be escalated to 65–70 Gy, with local control rates were generally low, given the high radiotherapy dose deliv-
comparable to those obtained with particle therapy [26]. ered, and were reported to be between 4.5 and 7 % [13, 25,
Because of the presence of neighboring critical structures 28–31]. In conclusion, high-precision, high-dose particle
like the brain stem or the temporal lobe, with older tech- radiotherapy is followed by an excellent local control for
niques the total dose was limited to 50–60 Gy, which was chondrosarcoma and a reasonably good local control for
clearly insufficient to control even microscopic residual dis- chordoma, with a low toxicity. Postoperative proton beam
ease in the long term. Charged particles such as protons, radiotherapy is often considered now as the standard treat-
helium, or neons or carbon ions are well suited for extremely ment for these tumors. We will examine next if high-precision
precise localization of radiation and permit an increase of the and high-dose STRT can be considered as alternative treat-
total dose from 15 to 35 % compared with conventional ments to proton beam therapy.
X-rays [10]. In their experience with charged particle irradia-
tion of tumors of the skull base, Castro et al. have treated 53
chordomas and 27 chondrosarcomas, from 1977 to 1992, Fractionated Stereotactic Radiotherapy
with a mean dose of 65 Gray equivalent (GyE) (range, 60–85 of Chordomas and Chondrosarcomas
GyE) [10]. Five-year local control was 63 % for chordomas
and 78 % for chondrosarcomas [10]. Since then, a number of Patient Selection
patients were treated with particle therapy, generally proton As previously emphasized, the standard management of
beam therapy in the United States, France, Japan, Switzerland, chordomas and chondrosarcomas should include a maximal
and Germany, with protocols using high doses of irradiation. surgical resection, followed by postoperative high-precision,
At the Massachusetts General Hospital (MGH), where high-dose radiotherapy, if one wants to maximize the chances
chordomas and chondrosarcomas were treated with 160 MV of local control [6, 15, 16, 20–30].
proton beams, with a median dose of 69 CGE (Cobalt-Gray- The indications for STRT are essentially the same as those
Equivalent), it was found that chordomas had a 31 % rate of for RS (see chapter on RS), except that STRT combines the
failure of which 95 % were local [24, 25]. It was also precision of stereotactic positioning with the radiobiological
shown that the 10-year local control was the highest with advantage of fractionation, especially in large tumors [12].
Fig. 40.4 Treatment plan for

STRT of a case of chordoma
(a) sagittal (b) axial
Table 40.1 Treatment outcomes, chordomas

Median
No. of prescribed 5-year local 5-year Median
Series Year patients Technique dose (Gy) control (%) survival (%) follow-up Comments
Krishnan et al. [33] 2005 25 GKRS 15 55 4.8 years EBRT also used in a
proportion of patients
Martin et al. [32] 2007 18 GKRS 16 63 63 7.7 years EBRT also used in a
proportion of patients
Hasegawa et al. [36] 2007 27 GKRS 14 72 84 6 years
Dassoulas et al. [38] 2009 15 GKRS 12.7 42 80 5.8 years LC improved to 50 %
after 2nd GKRS
Kano et al. [39] 2011 71 GKRS 15 66 74 5 years NAGKC (6 centers)
Henderson et al.[40] 2009 18 CyberKnife 35 59 82 4 years 35 Gy in 5 sessions
Debus et al. [12] 2000 37 Linac STRT 66.6 50 7682 19 months 1.8 Gy per fraction
Bugoci et al. [37] 2012 12 Linac STRT 66 38 76 3.5 years 2 Gy per fraction
Thus, in tumors larger than 30 mm in diameter, there is an They added a 2-mm safety margin for the planning target
advantage of using STRT rather than RS, at least in theory. In volume (PTV). The median prescribed dose at isocenter was
the series from Heidelberg, where STRT was used for chordo- 66.6 Gy for chordomas and 64.3 Gy for chondrosarcomas,
mas and chondrosarcomas of the skull base, the median target with a median daily dose of 1.8 Gy [12]. More recently,
volume was 56 cm3 (range, 17–215 cm3) for chordomas and Boguci et al. have used image-guided intensity-modulated
102 cm3 (range, 24–237 cm3) for chondrosarcomas [12]. This fractionated STRT, with a median dose of 66.6 Gy prescribed
is in sharp contrast with the mean or median volumes described at the 90 % isodose, so that the dose at the isocenter was 74 Gy
in the RS series, which ranged from 4.6 cm3 (range, 0.98– [37]. Figure 40.4a, b shows an example of linac-based STRT
10.8 cm3) to 14.6 cm3 (range, 2.9–52 cm3) [32–36]. treatment plan at our institution (CHUV) for a 59-year-old
patient who was not operated due to important comorbidities.
Treatment Techniques The patient, with a skull base chordoma, received a total dose
Planning techniques are similar to those used for any RS or of 63 at 1.8 Gy per fraction, using a micro-multileaf collima-
STRT of other intracranial lesions. Immobilization systems tor with six fixed beams from a 6-MV linear accelerator.
including various stereotactic head frames (RS) or relocatable
head masks (STRT) are used for imaging studies under ste- Treatment Outcomes
reotactic guidance and for treatment. All patients should Because of the rare occurrence of chordomas and chondrosar-
undergo high-resolution computed tomography (CT) and if comas, the published experience with either RS or STRT is
not contra-indicated magnetic resonance imaging (MRI), still limited. Table 40.1 summarizes some recent series with
with image fusion for treatment planning. In both cases, ultra- RS and STRT for chordomas and Table 40.2 the experience
thin slices are required. With regard to fractionated STRT, the with chondrosarcomas. As can be seen, these series of Gamma
Heidelberg group has defined the clinical target volume Knife radiosurgery (GKRS) and STRT have gathered a lim-
(CTV) as the visible tumor on CT and MRI and the potential ited number of cases of chordomas [12, 32, 33, 36–40] and
residual tumor, taking preoperative imaging into account [12]. very few cases of chondrosarcomas [12, 32–34, 36].
Table 40.2 Treatment outcomes, chondrosarcomas

Median
No. of prescribed 5-year local 5-year Median
Series Year patients Technique dose (Gy) control survival follow-up Comments
Krishnan et al. [33] 2005 4 GKRS 15 100 % 4.8 years
Feigl et al. [34] 2005 10 GKRS 17 100 % NA 17 months One patient had 2 GKRS
Martin et al. [32] 2007 10 GKRS 15 88 % (6 years) 88 % 7.7 years 1.8 Gy per fraction
Hasegawa et al. [36] 2007 7 GKRS 14 80 % 83 % 6 years
Debus et al. [12] 2000 8 Linac STRT 64.9 100 % 100 % 19 months
Overall, these results confirm the known differences in

outcome between chordomas and chondrosarcomas. Results
on local controls and survivals are fairly similar to those
obtained after postoperative proton beam radiotherapy. For
chordomas, the 5-year local control rates varied between 38
and 72 %, whereas for chondrosarcomas, it was between 80
and 100 %. The rather wide variations between the different
series are probably due to differences in surgical tech-
niques, tumor size, and in the patients’ populations, some
reports having included both primary treatments and treat-
ments for recurrence. The experience at the University of
Heidelberg and at the Kaiser Permanente in Los Angeles
with STRT showed encouraging local control rates, which
were comparable to those obtained with RS [12, 37].
It should also be re-emphasized that patients treated by
STRT had on average a much larger residual tumor volume
compared with those who received GKRS (vide supra). Fig. 40.5 Pathology slide of a case of paraganglioma
STRT yields the same excellent local control rates for chon-
drosarcomas as with RS, but admittedly the number of
patients was quite low [12]. Paragangliomas
Functional outcome after RS was reported in the rather
large North America Gamma Knife consortium: of 57 General Features
patients with prior neurological deficits, 17 improved, 18 PGLs are neuroendocrine tumors originating from the auto-
were unchanged, and 16 deteriorated [39]. nomic nervous system derived from neural crest cells.
Radiation-induced complications are generally moderate; Approximately 80–85 % of these tumors are located in the
however, cranial nerve dysfunction, pituitary insufficiency, and adrenal medulla and called pheochromocytomas (PCCs),
occasionally brain necroses were reported [32, 33]. Similarly, whereas 15–20 % grow in the extra-adrenal chromaffin tissue
the complication rates after STRT are quite low: from the 45 and are referred to as secreting paraganglioma (sPGLs).
patients treated with STRT in Heidelberg, only one presented These tumors are rare and occur in 2–8 per 1,000,000, with a
with symptomatic infarction of the pons [12]. peak incidence in the third to the fourth decade of life, with
an almost equal distribution between female and male
patients; familial tumors occur at an earlier age [41]. They
Conclusion are generally slow growing, highly vascular, with a high
intracellular liquid content and with frequent intratumoral
In conclusion, RS, either with GKRS or linac RS, and STRT cysts [42–45] (Fig. 40.5). Depending on their origin, PGLs
appear to be safe treatments as postoperative complement to can be divided into two groups: sympathetic and parasympa-
surgery or for recurrent chordomas and chondrosarcomas of thetic tumors. PCCs fall into the former category. Sympathetic
the base of the skull. It is thus reasonable to state that, in the PGLs can be found anywhere along the sympathetic ganglia,
absence of a nearby proton beam therapy facility, fraction- in the thorax, abdomen, and pelvis, and they tend to hyperse-
ated STRT represents a possible, safe and cheaper alternative crete catecholamines, like adrenal PCC. In contrast, para-
treatments for these skull base tumors. sympathetic PGLs are principally located in the head and
NF1 for Neurofibromatosis Type 1, and RET for Multiple

Endocrine Neoplasia Type 2 (MEN2); others comprise the
succinate dehydrogenase (SDH) complex subunit genes
(SDHA, SDHB, SDHC, SDHD), the SDH complex cofac-
tors (SDHAF2), and two newly recognized susceptibility
genes, TMEM127 and MAX [57]. SDHB mutations are usu-
ally linked to a higher morbidity and mortality than muta-
tions in the other SDH genes [58]. A recent meta-analysis of
studies concerning SDHB-mutated patients has emphasized
that 31 % of their tumors were malignant [59]. Notably,
VHL- and SDH-mutated PGLs are associated with angio-
genesis, hypoxia, and reduced oxidative response [60] while
the group containing RET- and NF1-mutated tumors is
linked to an abnormal activation of kinase-signaling path-
ways, such as RAS/RAF/MAPK and PI3K/AKT/mTOR [61,
62]. Also TMEM127- and MAX-mutated tumors have been
associated with the activation of mTOR-signaling pathway
Fig. 40.6 MRI of a case of paraganglioma [63, 64]. The recommended screening test for initial evalua-
tion of PGLs is the measurement of plasma-free metaneph-
rines or urine-deconjugated differential metanephrines [65].
neck region (HNPGLs) and base of the skull, and in their Metanephrines have a higher sensitivity, ranging around
majority are nonsecretory. They can be multicentric, espe- 98–99 %, in comparison to plasma or urine catecholamines
cially in familial cases [45]. In some instances, they can and vanilmandelic acid [66, 67]. The biochemical phenotype
reach very large sizes and are referred to as giant tumors does not allow to differentiate malignant from benign PGLs.
[45]. One quarter of PGLs are malignant, defined by the However, the presence of predominantly noradrenaline-
presence of chromaffin tissue in sites where it is normally producing PGLs and high levels of plasma dopamine or its
absent, such as lymph nodes, bone, liver, and lung [41]. metabolite methoxytyramine may suggest malignancy [66, 68].
Unfortunately, there are currently no consistent methods to Plasma chromogranin A (CgA), which is a protein stored and
differentiate benign from malignant forms, and malignancy cosecreted with catecholamines, is frequently increased in
is revealed by distant metastases [46]. The clinical behavior functioning and non-functioning PGLs [69]. CgA shows a
of PGLs is generally indolent with long time intervals sensitivity of 83–89 %, and very high plasma levels of CgA
between the first symptom and the diagnosis [42, 45]. In case are generally related to malignancy [66]. Computed tomog-
of sPGLs, hypertension is the main symptom, which can be raphy (CT) and magnetic resonance (MRI) are major tools
continuous, intermittent, or paroxysmal. HNPGLs are char- for the first imaging approach in patients with PGLs. MRI
acterized by a mass effect or the infiltration of adjacent shows a good accuracy, with a sensitivity of 90–100 % and
anatomical structures such as the temporal bone, the middle specificity of 50–100 %, especially for the detection of extra-
ear, the clivus, the jugular vein, the internal carotid artery, the adrenal disease [70]. Ultrasound is used for the detection of
cavernous sinus, the hypoglossal canal, and cranial nerves HNPGLs, whereas for other disease sites its role is limited
V–XII [43, 44, 47–52] (Fig. 40.6). A palpable neck mass, as [70, 71]. Concerning functional imaging, 131I- or 123I-
well as dysphagia, pain, tinnitus, or cranial nerve palsies can metaiodobenzylguanidine (MIBG) scintigraphy has been used
also occur [52]. The several classifications take into account as the first-line nuclear medicine technique but 123I-MIBG is
tumor location, extent, and size. The Fisch classification is superior to 131I-MIBG, with a sensitivity of 83–100 % and a
one of the most commonly used [50]. The clinical behavior specificity of 95–100 % [70]. Regarding PET imaging, soma-
of malignant PGLs may also be associated with systemic tostatin analogues labelled with gallium-68 can be used. This
symptoms, including fatigue, weight loss, and anorexia, or tracer seems to be superior to 18F-labelled fluorodeoxyglu-
clinical manifestations due to metastatic disease, such as cose (18F-FDG) in identifying malignant sPGLs [72].
bone metastases. Metastases can occur at the outset or after
many years. Population-based studies have established that
approximately one third of patients having sporadic PGL Standard Treatment for Paragangliomas
have a germline mutation in a known susceptibility gene
[53–56]. Ten known susceptibility genes for PGL have been The treatment of choice of PGL is a matter of controversy.
identified to date: three genes for three distinct cancer sus- Surgery has an important role but it is by far not the only option.
ceptibility syndromes, VHL for von Hippel–Lindau disease, With recent imaging tools, innovative surgical techniques, and
cranial nerve monitoring, results have improved regarding the controls with exclusive or postoperative radiotherapy [74, 75].
possibility of complete removal, and immediate complications Powell et al. have treated 46 patients with glomus jugulare or
have decreased. However, like for chordomas and chondrosar- glomus tympanicum tumors with doses between 45 and 50 Gy
comas of the skull base, a genuine total resection without and a 75 % actuarial control was achieved at 25 years [75].
sacrifice of cranial nerves is rarely possible. Therefore, postop- Notably, in these radiotherapy series, between 18 and 30 %
erative cranial nerve damage can be quite substantial, with were previously, and sometimes heavily pretreated by surgery.
potential impairment of nerves VII and IX–XII [45, 48, 49]. Contrarily to older reports, complication rates were low with
For example, in one series, preoperative deficit of cranial nerves the most recent radiotherapy techniques [78–80]. Previously, it
X–XII was between 20 and 30 % and postoperative deficits was thought that these tumors were “radioresistant” because
increased to 25–50 %, although facial function recovery was after irradiation, they shrink but rarely disappear completely on
observed in 95 % of patients [49]. The rate of complete surgical follow-up [78]. Therefore, local control after radiotherapy
removal is quite variable and is claimed to be between 40 and should be defined as the absence of clinical and radiological
96 % [48, 49]. In general, data regarding resection rates and the progression. Springate et al. have reviewed the outcome of
actual occurrence of complications are difficult to interpret due PGL after various treatments [81]. They noted that local con-
to the heterogeneity and limited size of most series. There are trol after surgery, surgery combined with radiotherapy, or
only few data on local control and long-term survival after sur- radiotherapy alone was 86 %, 90 %, and 92 %, respectively.
gery. Some surgeons have claimed that in their hands, the In the same review, treatment-related morbidity after surgical
recurrence rate was quite low. Nora et al. have reported on 59 excision was frequent, whereas late complications were rare
carotid body tumors, of which only 3 (6 %) presented with a after radiotherapy [81]. This was more recently confirmed by
recurrence after surgery, and only one developed metastatic Huy et al. who compared in a retrospective and comparative
disease [73]. Pareschi et al. in their series of 37 glomus jugulare analysis radiotherapy and surgery in 88 cases of jugular PGLs,
tumors, of which 96 % had a complete resection, found no in terms of function and tumor control [82]. Forty-seven
relapse, but the mean follow-up was only 4.9 years [48]. patients with type C or D jugular PGLs underwent surgery after
Amongst 28 patients with the so-called complex glomus jugu- endovascular embolization between 1984 and 1998, and were
lare tumors, Al-Mefty and Teixeira were able to perform a followed up on average during 66 months. Forty-one patients
complete resection in 24 patients and observed altogether two with type C jugular PGLs were treated by external beam con-
recurrences [45]. However, in other studies, local control was formal radiotherapy between 1998 and 2003 with a total mean
much lower, with 70–100 % recurrence rates after surgery alone dose of 45 Gy (range, 44–50 Gy); the mean follow-up was 50
[74, 75]. Because these tumors tend to recur after many years, months [82]. This study demonstrated that with radiotherapy, a
the actual failure rates are underestimated, especially with short 96 % local control was achieved, whereas with surgery it was
follow-ups. In addition many patients are lost to follow-up [45, 86 % [82]. After surgery, postoperative complications included
48]. Long-term mortality due to recurrence after surgery is not dysphagia, aspiration, and facial paralysis. Patients treated by
so trivial and was reported to be between 5 and 13 % [76, 77]. radiotherapy developed only minor sequelae. In a British retro-
Like for chordoma and chondrosarcoma, some authors have spective study, 21 patients with PGL of the head and neck
claimed that PGLs are radioresistant and that radiotherapy was region received fractionated external beam radiotherapy with a
associated with long-term complications [45]. These assertions median dose of 50 Gy in 30 fractions [83]. With a median fol-
were based on old studies in which ancient orthovoltage tech- low-up of 55 months, a 92 % 5-year local control rate was
niques were used. Indeed, with very conformal linac-based observed [83]. Another retrospective study, with a median
techniques, fractionated external beam radiotherapy represents follow-up of 9 years, demonstrated that treatment of PGL by
an excellent alternative, or sometimes a complementary treat- external beam radiotherapy yielded a 5-year local control rate
ment to surgery. Cole et al. have treated 32 tumors of the glo- of 96 % and a 10- and 15-year local control rate of 90 % [84].
mus jugulare or glomus vagale with megavoltage units with
doses of 45 Gy in 5 weeks [78]. Their very long-term results
have yielded a local control rate of 94 % at 10 years [78]. Fractionated Stereotactic Radiotherapy
Konefal et al. have treated 26 patients with 45–50 Gy; amongst of Paraganglioma
these, 15 of 16 glomus tympanicum and four of six glomus
jugulare tumors achieved a long-term control, with a mean Patient Selection
follow-up period of 10.5 years [79]. In a large series, 80 PGLs Since external beam radiotherapy is an effective alternative
of the temporal bone, carotid body, or glomus vagale were to surgery, newer, high-precision, high-dose radiotherapy
treated by radiotherapy alone (72) or postoperatively (8) [80]. techniques such as STRS represent a logical further option.
Local control was 94 % and only 5 local recurrences were With STRS, a smaller volume of normal tissue exposed to
observed between 2.6 and 18.8 years, of which two could be the effect of radiation can be expected [47]. The criteria for
salvaged by surgery [80]. Other groups could obtain good local patient selection for RS or STRS for PGL are comparable to
Table 40.3 Treatment outcomes, paraganglioma

Median
No. of marginal Local Median
Series Year patients Technique dose (Gy) control Survival (%) follow-up Neurologic status
Eustacchio et al. [43] 1999 13 GKRS 13.5 PR 4/10 11/13 46 months Improved in 5/13
SD 6/10 Unchanged in 5/13
Stroke in 1/13
Liscak et al. [42] 1999 66 GKRS 16.6 PR 15/47 NS 24 months Improved in 15/52
SD 28/47 Unchanged in 34/52
Worsened in 3/52
Jordan et al. [86] 2000 8 GKRS 16.3 PR 4/7 NS 27 months Improved in 4/8
NC 3/7 Unchanged in 3/8
Worsened in 1/8
Foote et al. [47] 2002 25 GKRS 15 PR 8/25 80 % 5 years 37 months Improved in 15/25
Unchanged in
NC 17/25 10/25
Feigenberg et al. [85] 2002 5 Linac RS 15 PR or SD 3/5 100 % 27 months NS
PRO 2/5
Zabel et al. [44] 2004 22 Linac STRT 57.6 PR 7/22 89.5 % 5 and 10 5.7 years Improved in 13/22
years
Feigl et al. [87] 2006 12 GKRS 17 SD 13/22 100 % 33 months Unchanged in 7/22
Ganz et al. [88] 2009 14 GKRS 13.6 PRO 2/22 100 % 28 months Worsened in 2/22
Genç et al. [89] 2010 18 GKRS 15.6 90 % LC 5 and 17/18 41.5 months 1 hearing loss
10 years
Lee et al. [90] 2011 14 GKRS 13.7 100 % 100 % 40 months Improved in 12/14
Lieberson et al. [92] 2012 41 Linac STRT 20 100 % 100 % 4.8 years Improved in 10/18
94 % Unchanged in 7/18
100 % Worsened in 1/18
93 % cranial nerve
preservation
1 persistent vertigo
those used for external beam radiotherapy. RS and STRT can defined the CTV as the macroscopic tumor on MRI plus a
be used either as primary treatment, at progression after one 10-mm margin along the involved vessels [44]. They added a
or multiple surgeries, after embolization, postoperatively, or 2-mm safety margin for the PTV. The median total dose was
even after previous external beam radiotherapy [42–44, 47, 57.6 Gy at a median daily dose of 1.8 Gy [44].
85]. The indications for STRT are essentially the same as
those for RS, except that STRT combines the precision of Treatment Outcomes
stereotactic positioning with the radiobiological advantage Because of the rare occurrence of PGL, the published experi-
of fractionation, especially in large tumors [12, 44]. In the ence with either RS or STRT is still rather limited. Table 40.3
Heidelberg experience with STRT for PGL, the median tar- [42, 43, 47, 85, 86] summarizes the available data with RS
get volume was 71.8 cm3 (range, 10.5–212 cm3) [44], which and STRT. Altogether, each RS studies have gathered a lim-
is in sharp contrast with the median volumes described in the ited number of cases mainly by GKRS [42, 43, 47, 86–90]
RS series, which ranged from 5.7 cm to 10.8 cm3 (range 0.5– and less by linac-based RS [85] (Table 40.3). The results dis-
27 cm3) [42, 43, 47, 85, 86]. closed a good to excellent local control, especially in the
most recent series. Overall, response rates included 41 %
Treatment Techniques of partial responses, 57 % of stable lesions, and only 3 % of
Pretreatment imaging includes high-resolution CT, MRI, and progressions. When reported, 5-year actuarial local control
preferably both and wherever needed angiography [43, 47]. rates varied between 90 and 100 %. The neurological status
For treatment planning, CT and MRI fusion are required, was improved in 40 %, stable in 55 %, and worse in 5 % of
especially with linac-based RS/STRT. In both cases, ultra- cases. Information on long-term survivals is more limited,
thin slices are required. Head frames (RS) or relocatable but it appears that when described, the 5-year overall sur-
head masks (STRT) are used for imaging under stereotactic vival was between 80 and 100 %. Chen et al. have conducted
guidance and treatment. For STRT, the Heidelberg group a systematic review evaluating GKRS for PGLs: their review
which included also their own cases consisted in a pooled

analysis from 11 selected studies and confirmed the excellent References
(90.5 %) success rate after GKRS treatment [91]. The
Heidelberg series of STRT for PGL also discloses encourag- 1. Prabhu S, Demonte F. Treatment of skull base tumors. Curr Opin
Oncol. 2003;15:209–12.
ing results, with a 32 % rate of partial response, 59 % of
2. Mc Master ML, Goldstein AM, Bromley CM, et al. Chordoma,
stable disease, and 9 % of progression [44]. The apparently incidence and survival patterns in the United States. Cancer Causes
slightly worse rate of progression is likely to be due on one Control. 2001;12:1–12.
hand to the longer follow-up than that of the RS series and on 3. Yang XR, Ng D, Alcorta DA, et al. T (brachyury) gene duplication
confers major susceptibility to familial chordoma. Nat Genet.
the other hand to tumors with much larger volumes (vide
2009;41:1176–8.
supra). The neurologic improvement was also quite good, 4. Sell M, Sampaolo S, Di Lorio G, et al. Chordomas: a histological
with 59 % of patients who improved, 32 % who remained and immunohistochemical study of cases with or without recurrent
stable, and 9 % who deteriorated [44]. Altogether, the tumors. Clin Neuropathol. 2004;23:277–85.
5. Catton C, O’Sullivan B, Bell R, et al. Chordoma: long-term follow-
Heidelberg data are the most reliable in terms of length of
up after radical photon irradiation. Radiother Oncol. 1995;41:
follow-up and adequate actuarial reporting of local control 67–70.
and survival. The experience at Stanford on 41 lesions in 36 6. Almfety K, Pravdenkova S, Coll BO, et al. Chordoma and
patients confirms also that STRT results in an excellent Chondrosarcoma: similar but quite different, skull base tumors.
Cancer. 2007;110:2457–67.
100 % local control with very few complications [92]. As far
7. Cianfriglia F, Pompili A, Occhipiati E. Intercranial malignant carti-
as complications of RS are concerned, it is difficult to distin- laginous tumors. Report of 2 cases and review of literature. Act
guish between neurologic worsening due to treatment and Neurochir. 1978;45:163–75.
due to recurrence. However, the rate of reported complica- 8. Gay E, Sekhar LN, Rubinstein E, et al. Chordomas and chondrosar-
comas of the cranial base: results and follow-up of 60 patients.
tions appears to be low. Eustacchio et al. had no single
Neurosurgery. 1995;36:887–96.
case of complications in their 13 cases [43]. In the multi- 9. Noel G, Feuvret L, Calugaru V, et al. Chordoma of the base of the
institutional report of Liscak et al., three cases of neurologic skull and upper cervical spine. One hundred patients irradiated by a
deficits out of 52 patients are described, 2 of which were 3 D conformal technique combining photons and proton beams.
Acta Oncol. 2005;44:700–8.
permanent: facial nerve impairment and deafness in one case
10. Castro JR, Linstadt DE, Bahary JP, et al. Experience in charged
and facial nerve impairment and vertigo in another [42]. Two particle irradiation of tumors of the skull base: 1977–1992. Int J
further cases of inner ear inflammatory complications are Radiat Oncol Biol Phys. 1994;29:647–55.
also noted [42]. From the Mayo Clinic experience, only one 11. Korten AGC, ter Berg HJW, Spincemaille GH, et al. Intracranial
chondrosarcoma: review of the literature and report of 15 cases.
case of temporary vertigo is described [47]. Again, this low
J Neurol Neurosurg Psychiatry. 1998;65:88–92.
rate of complications should be interpreted with caution as 12. Debus J, Schulz-Ertner D, Schad L, et al. Stereotactic fractionated
the average median follow-up of the RS studies is altogether radiotherapy for chordomas and chondrosarcomas of the skull base.
not very long. Out of the 22 patients who benefited from Int J Radiat Oncol Biol Phys. 2000;47:591–6.
13. Hug EB, Loredo LN, Slater JD, et al. Proton radiation therapy for
STRT, 3 experienced temporary xerostomia and 2 temporary
chordomas and chondrosarcomas of the skull base. J Neurosurg.
taste impairment, 4 had middle-ear effusion, but none of 1999;91:432–9.
these had more than a grade 2 CTC reaction, and all recuper- 14. Munzenrider JE, Liebsch NJ. Proton therapy for tumors of the skull
ated; no late adverse reactions from STRT were observed at base. Strahlenther Onkol. 1999;S2(Suppl):57–63.
15. Crockard HA, Steel T, Plowman N, et al. A multidisciplinary
follow-up [44].
team approach to skull base chordomas. J Neurosurg. 2001;95:
175–83.
16. Crockard HA, Cheeseman A, Steel T, et al. A multidisciplinary
Conclusion team approach to skull base chondrosarcomas. J Neurosurg. 2001;
95:184–9.
17. Koch BB, Karnell LH, Hoffman HT, et al. National Cancer data-
In conclusion, fractionated external beam radiotherapy, RS, base report on chondrosarcoma of the head and neck. Head Neck.
and STRT appear to be safe and efficient therapeutic modali- 2000;22:408–25.
ties in the primary treatment or at progression for PGL. Yet 18. Roseberg AE, Nielsen GP, Keel SB, et al. Chondrosarcoma of the
base of the skull: a clinicopathologic study of 200 cases with
available follow-ups remain limited for most RS series. With
emphasis on its distinction from chordoma. Am J Surg Pathol.
regard to STRT, the results also appear to be quite good, with 1999;23:1370–8.
longer time of evaluation, knowing also that on average, 19. Hassounah M, Al-Mefty O, Akhtar M, et al. Primary cranial and
tumors were larger. External beam radiotherapy, RS, and intracranial chondrosarcoma. A survey Acta Neurochir. 1985;78:
123–32.
STRT deserve continuous experience, but it is reasonable to
20. Walcott B, Nahed BV, Mohyeldin A, et al. Chordoma: current
state that they represent possible and safe alternative treat- concepts, management and future directions. Lancet Oncol. 2012;
ments for these skull base tumors. 13:69–76.
21. Bloch OG, Jian BJ, Yiang I, et al. A systemic review of intracranial 42. Liscak R, Vladyka V, Wowra B, et al. Gamma-knife radiosurgery of
chondrosarcoma and survival. J Clin Neurosciences. 2009;16: the glomus jugulare tumors—early multicentre experience. Acta
1547–51. Neurochir. 1999;141:1141–6.
22. Tzortzidis F, Elali F, Wright D, et al. Patient outcome and long-term 43. Eustacchio S, Leber K, Trummer M, et al. Gamma-knife radio-
follow-up after aggressive microsurgical resection of cranial base surgery for glomus jugulare tumors. Acta Neurochir. 1999;141:
chordoma. Neurosurgery. 2006;59:230–7. 811–8.
23. Krayenbuhl H, Yasargil MG. Cranial chordomas. Prog Neurol 44. Zabel A, Milker-Zabel S, Huber P, et al. Fractionated stereotactic
Surg. 1975;6:380–434. conformal radiotherapy in the management of large chemodecto-
24. Suit HD, Goitein M, Munzenreider J, et al. Definitive radiation mas of the skull base. Int J Radiat Oncol Biol Phys. 2004;
therapy for chordoma and chondrosarcoma of the base of skull and 58:1445–50.
cervical spine. J Neurosurg. 1982;56:377–85. 45. Al-Mefty O, Teixeira A. Complex tumors of the glomus jugulare:
25. Fagundes MA, Hug EB, Liebsch NJ, et al. Radiation therapy for criteria, treatment and outcome. J Neurosurg. 2002;97:1256–65.
chordomas of the base of skull and cervical spine: patterns of 46. Fassnacht M, Kreissl MC, Weismann D, et al. New targets and
failure and outcome after relapse. Int J Radiat Oncol Biol Phys. therapeutic approaches for endocrine malignancies. Pharmacol
1995;15:579–84. Ther. 2009;123(1):117–41.
26. Potluri S, Jefferies SJ, Jena K, et al. Residual post-operative tumor 47. Foote RL, Pollok BE, Gorman DA, et al. Glomus jugulare tumors:
volume predicts outcome after high-dose radiotherapy for chor- tumor control and complications after stereotactic radiosurgery.
doma and chondrosarcoma of the skull base and spine. Clin Oncol. Head Neck. 2002;24:332–9.
2011;23:199–208. 48. Pareschi R, Righini S, Destifi D, et al. Surgery of glomus tumors.
27. Igaki H, Tokuuye K, Okumura T, et al. Clinical results of proton Skull Base. 2003;13:149–57.
beam therapy for skull base chordoma. Int J Radiat Oncol Biol 49. Green JD, Brackmann DE, Nguyen CD, et al. Surgical management
Phys. 2004;60:1120–6. of previously untreated glomus jugulare tumors. Laryngoscope.
28. Ares C, Hug EB, Lomax AJ, et al. Effectiveness and safety of spot- 1994;104:917–21.
scanning proton radiation therapy for chordomas and chondrosar- 50. Fisch U, Fagan P, Valavanis A. The infratemporal fossa approach
comas of the skull base : first long-term report. Int J Radiat Oncol for the lateral skull base. Otolaryngol Clin North Am. 1984;
Biol Phys. 2009;75:1111–8. 17:513–22.
29. Schulz-Ertner D, Karger CP, Feuerhake A, et al. Effectiveness of 51. Parenti G, Zampetti B, Rapizzi E, et al. Updated and new perspec-
carbon ion radiotherapy in the treatment of skull-base chordoma. tives on diagnosis, prognosis, and therapy of malignant pheochro-
Int J Radiat Oncol Biol Phys. 2007;68:449–57. mocytoma/paraganglioma. J Oncol 2012 [Epub ahead]
30. Schultz-Ertner D, Nikoghosyan A, Hof H, et al. Carbon ion radio- 52. Erickson D, Kudva YC, Ebersold MJ, et al. Benign paraganglio-
therapy of skull-base chondrosarcoma. Int J Radiat Oncol Biol mas: clinical presentation and treatment outcomes in 236 patients.
Phys. 2007;67:171–7. J of Clin Endocrinol and Metab. 2001;86:5210–6.
31. Ricken S, Habemehl D, Nikoghosyan A, et al. Assessment of early 53. Neumann HP, Bausch B, McWhinney SR, et al. Germ-line muta-
toxicity and response in patients treated with proton and carbon ion tions in nonsyndromic pheochromocytoma. N Engl J Med. 2002;
therapy at the Heidelberg ion therapy center using the raster-scanning 346:1459–66.
technique. Int J Radiat Oncol Biol Phys. 2011;81:793–801. 54. Bausch B, McWhinney SR. Germ-line mutations in nonsyndromic
32. Martin JJ, Niranjan A, Kondziolka D, et al. Radiosurgery for chor- pheochromocytoma. N Engl J Med. 2002;346:1459–66.
domas and chondrosarcomas of the skull base. J Neurosurg. 2007; 55. Mannelli M, Castellano M, Schiavi F, et al. Clinically guided
107:758–69. genetic screening in a large cohort of Italian patients with pheo-
33. Krishnan S, Foote RL, Brown PD, et al. Radiosurgery for cranial base chromocytomas and/or functional or nonfunctional paraganglio-
chordomas and chondrosarcomas. Neurosurgery. 2005;56:777–83. mas. J Clin Endocrinol Metab. 2009;94:1541–7.
34. Feigl GC, Bundschuh O, Gharabaghi A, et al. Evaluation of a new 56. Neumann HP, Erlic Z, Boedeker CC, et al. Clinical predictors for
concept for the management of skull base chordomas and chondro- germline mutations in head and neck paraganglioma patients: cost
sarcomas. J Neurosurg. 2005;S102(Suppl):165–70. reduction strategy in genetic diagnostic process as fall-out. Cancer
35. Kocher M, Voges J, Staar S, et al. Linear accelerator radiosurgery Res. 2009;69:3650–6.
for recurrent malignant tumors of the skull base. Am J Clin Oncol. 57. Fishbein L, Nathanson KL. Pheochromocytoma and paragangli-
1998;21:18–22. oma: understanding the complexities of the genetic background.
36. Hasegawa T, Ishii D, Kida Y, et al. Gamma- knife surgery for skull- Cancer Genetics. 2012;205:1–11.
base chordomas and chondrosarcomas. J Neurosurg. 2007;107: 58. Gimenez-Roqueplo AP, Favier J, Rustin P, et al. Mutations in the
752–7. SDHB gene are associated with extra-adrenal and/or malignant
37. Bugoci DM, Girvigian MR, Chen JCI, et al. Photon-based fraction- phaeochromocytomas. Cancer Res. 2003;63:5615–21.
ated stereotactic radiotherapy for post-operative treatment of skull 59. Welander J. Söderkvist, Gimm O. Genetics and clinical characteris-
base chordomas. Am J Clin Oncol 2012; [Epub ahead] tics of hereditary pheochromocytomas and paragangliomas. Endocr
38. Dassoulas K, Schlesinger D, Yen CP, et al. The role of Gamma- Rel Cancer. 2011;18:253–76.
knife surgery in the treatment of skull base chordomas. J Neuro- 60. Favier J, Brière J, Burnichon N, et al. The Warburg effect is geneti-
oncol. 2009;94:243–8. cally determined in inherited pheochromocytomas. PLoS One,
39. Kano H, Iqbal F, Sheehan J, et al. Stereotactic radiosurgery for 2009;4; 9, e7094,
chordoma: a report from the North American Gamma Knife 61. Califano D, Rizzo C, D’Alessio A, et al. Signaling through Ras is
Consortium. Neurosurgery. 2011;68:379–89. essential for ret oncogene-induced cell differentiation in PC12
40. Henderson FC, Mc Cool K, Seigle J, et al. Treatment of chordomas cells. J Biol Chem. 2000;275:9297–305.
with Cyberknife: Geortown University experience and treatment 62. Johannessen CM, Reczek EE, James MF, et al. The NF1 tumor
recommendations. Neurosurgery 2009; 64:(2 S) A 44–53 suppressor critically regulates TSC2 and mTOR. Proc Natl Acad
41. De Lellis RA, Lloyd RV, Heitz PU, et al. World Health Organization Sci U S A. 2005;102:8573–8.
classification of tumours. Pathology and genetics of tumours of 63. Qin Y, Yao L, King EE, et al. Germline mutations in TMEM127
endocrine organs. In: DeLellis RA, Lloyd RV, Heitz PU, Eng C, confer susceptibility to pheochromocytoma. Nat Genet. 2010;42:
editors. Lyon,:IARC Press; 2004 229–33.
64. Comino-Méndez I, Gracia-Aznarez FJ, Schiavi F, et al. Exome 79. Konefal JB, Pilepich MV, Spector GJ, et al. Radiation therapy in the
sequencing identifies MAX mutations as a cause of hereditary treatment of chemodectomas. Laryngoscope. 1987;97:1331–5.
pheochromocytoma. Nat Genet. 2011;43:663–7. 80. Hinerman RW, Mendenhall WM, Amdar RJ, et al. Definitive radio-
65. Lenders JWM, Pacak K, Walther MM, et al. Biochemical diagnosis therapy in the management of chemodectoma arising in the temporal
of pheochromocytoma: which test is best? JAMA. 2002;287: bone, carotid body and glomus vagale. Head and Neck. 2001;
1427–34. 23:363–71.
66. Grossman A, Pacak K, Sawka A, et al. Biochemical diagnosis and 81. Springate SC, Haraf D, Weichselbaum RR. Temporal bone chemo-
localization of pheochromocytoma: can we reach a consensus? Ann dectomas—comparing surgery and radiation therapy. Oncology.
New York Acad Sci. 2006;1073:332–47. 1991;5:131–7.
67. Davidson FD. Phaeochromocytoma with normal urinary catechol- 82. Huy PT, Kania R, Duet M, et al. Evolving concepts in the manage-
amines: the potential value of urinary free metadrenalines. Ann ment of jugular paraganglioma : a comparison of radiotherapy and
Clin Biochem. 2002;39:557–66. surgery in 88 cases. Skull Base. 2009;19(1):83–91.
68. Van der Harst E, De Herder WW, De Krijger RR, et al. The value of 83. Lightowlers S, Benedict S, Jefferies SJ, et al. Excellent local con-
plasma markers for the clinical behaviour of phaeochromocytomas. trol of paraganglioma in the head and neck with fractionated radio-
Eur J Endocr. 2002;147:85–94. therapy. Clin Oncol. 2010;22:382–9.
69. Guignat L, Bidart JM, Nocera M, et al. Chromogranin A and the 84. Chino JP, Sampson JH, Tucci DL, et al. Paraganglioma of the head
α-subunit of glycoprotein hormones in medullary thyroid carci- and neck: Long-term local control with radiotherapy. Am J Clin
noma and phaeochromocytoma. Br J Cancer. 2001;84:808–12. Oncol. 2009;32:304–7.
70. Ilias I, Pacak K. Current approaches and recommended algorithm 85. Feigenberg SJ, Mendenhall WM, Hinerman RW, et al. Radiosurgery
for the diagnostic localization of pheochromocytoma. Journal of for paraganglioma of the temporal bone. Head Neck. 2002;24:
Clin Endocr Metab. 2004;89:479–91. 384–9.
71. Blake M A, Kalra M K, Maher M M, et al. Pheochromocytoma: an 86. Jordan J, Roland PS, McManus C, et al. Stereotactic radiosurgery
imaging chameleon. Radiographics 2004; 24 (S) S87-S99 for glomus jugulare tumors. Laryngoscope. 2000;110:35–8.
72. Buchmann I, Henze M, Engelbrecht S, et al. Comparison of 87. Feigl GC, Horstmann GA. Intracranial glomus jugulare tumors:
68
Ga-DOTATOC PET and 111In-DTPAOC (Octreoscan) SPECT in volume reduction with Gamma Knife surgery. J Neurosurg. 2006;
patients with neuroendocrine tumours. Eur J Nuc Med Molec Imag. 105:161–7.
2007;34:1617–26. 88. Ganz JC, Abdelkarim K. Glomus jugulare tumors: certain clinical
73. Nora JD, Hallet JW, O’Brien PC, et al. Surgical resection of carotid and radiological aspects observed following Gamma Knife radio-
body tumors: long-term survival, recurrence and metastasis. Mayo surgery. Acta Neurochir. 2009;151:423–6.
Clin Proc. 1988;63:348–52. 89. Genç A, Bicer A, Abacioglu U, et al. Gamma Knife radiosurgery
74. Reddy EK, Mansfield CM, Hartman GV. Chemodectoma of glomus for the treatment of glomus jugulare tumors. J Neurooncol. 2010;
jugulare. Cancer. 1983;52:337–40. 97:101–8.
75. Powell S, Peters N, Harmer C, et al. Chemodectoma of the head and 90. Lee CC, Pan DHC, Wu JC, et al. Gamma knife radiosurgery for
neck: results of treatment in 84 patients. Int J Radiat Oncol Biol glomus jugulare and tympanicum. Stereotact Funct Neurosurg.
Phys. 1992;22:919–24. 2011;89:291–8.
76. Spector GJ, Sobol S. Surgery for glomus tumors at the skull base. 91. Chen PG, Nguyen JH, Payne SC, et al. Treatment of glomus jugu-
Otolaryngol Head Neck Surg. 1980;88:524–30. lare tumors with gamma knife radiosurgery. Laryngoscope. 2010;
77. Rosenwasser H. Long-term results of therapy of glomus jugulare 120:1856–62.
tumors. Arch Otolaryngol. 1973;97:49–54. 92. Lieberson RE, Adler JR, Soltys SG, et al. Stereotactic radiosurgery
78. Cole JM, Beiler D. Long-term results of treatment for glomus jugu- as the primary treatment for new and recurrent paraganglioma: is
lare and glomus vagale tumors with radiotherapy. Laryngoscope. open surgical resection still the treatment of choice? World
1994;104:1461–5. Neurosurg. 2012;77:745–61.
Head and Neck Tumors
41
Daniel T.T. Chua
radiosensitivity, and high incidence observed in some countries.

Introduction In this chapter, the clinical results of using radiosurgery in
the management of head and neck tumors are reviewed and
The technique of stereotactic radiosurgery has proved to be discussed, with emphasis on the management of NPC.
effective in the treatment of a variety of intracranial tumors,
both benign and malignant, and abnormal conditions such as
vascular malformations. Because of the proximity and the Nasopharyngeal Carcinoma
frequent involvement of the base of the skull, head and neck
tumors as a group have also been explored as targets for Local Control After Radiotherapy
extracranial radiosurgery. Gamma Knife and gantry-based
linear accelerator have been used to treat head and neck can- Radiotherapy is the mainstay of treatment for all stages of
cers using either a rigid or relocatable frame. Gamma Knife NPC without distant metastases. The largest series on treat-
however has a more limited role for treating lesions below ment outcome of NPC was reported by Lee et al. based on
skull base. Advances in radiation delivery and image-guided 5,037 patients treated at a single institution during the period
radiotherapy have lead to the development of new techniques 1976–1985 [1]. The 10-year local control and survival rates
and systems capable of performing stereotactic radiotherapy were 61 % and 43 %, respectively. In the past decade,
with improved dose conformity and comparable precision. advances in both imaging and radiotherapy techniques have
CyberKnife (Accuray Incorporated, Sunnyvale, CA, USA) is had significant impact on the management of NPC. As a
a frameless robotic radiosurgery system that uses image- result, the outlook of patients with NPC treated in recent
guided radiotherapy from a lightweight 6-MV linear accel- years, even for those presenting with a more advanced stage,
erator mounted on a robotic arm. The system can deliver is likely to have improved significantly. This is supported by
isocentric or non-isocentric beams from 120 different direc- a recent report on the outcome of patients with newly diag-
tions with six degrees of freedom. Real-time image guidance nosed NPC treated in the modern era. In that report, based on
system is used to ensure accurate localization and tracking of the results of 2,687 patients treated at all public oncology
target. Because malignant head and neck tumors are usually centers in Hong Kong during the period 1996–2000, the
highly aggressive and infiltrative, radiosurgery alone is sel- observed 5-year local control and survival rates were 85 %
dom appropriate as the primary treatment modality for newly and 75 %, respectively [2]. The 5-year local control rates for
diagnosed disease. Radiosurgery, however, may offer signifi- T1–2 and T3–4 diseases were 88 % and 79 %, respectively.
cant benefits when used as a salvage treatment for recurrent These patients were mostly treated by radiotherapy alone,
disease or as a boost treatment after conventional radiother- and only 23 % had additional chemotherapy. The results
apy. Among all the head and neck cancers, nasopharyngeal showed significant improvement when compared with those
carcinoma (NPC) has attracted more attention with regard to reported in old series. Despite the improved results, about
the application of radiosurgery due to its unique location 11.7 % of patients still developed local failures, and 8.2 %
near the skull base, pattern of growth and invasion, relative had isolated local failures without regional or distant failures.
In order to further improve local tumor control, combined
chemoradiotherapy has been used in advanced-stage disease,
D.T.T. Chua, M.D., F.R.C.R., F.H.K.C.R. (*) which showed improvement in local control compared with
Department of Radiotherapy, Hong Kong Sanatorium & Hospital, radiotherapy alone [3–6]. Another way to improve local
Happy Valley, Hong Kong
e-mail: danielchua@hksh.com control is by escalating the dose delivered to primary tumor
530 D.T.T. Chua
based on the established dose–volume–control relationships Results of Conventional Salvage Treatments

in NPC [7, 8]. For early-stage disease, dose escalation can be In salvaging local failures of NPC, a better outcome was
achieved by brachytherapy using intracavitary intubation. usually observed in patients treated for persistent disease
For advanced-stage disease, dose escalation is best deliv- after primary radiotherapy then recurrent disease. Brachy-
ered by either conformal or stereotactic radiotherapy. therapy in the form of intracavitary intubation or gold grain
implantation can be employed in salvaging early-stage
local failure. Our institution has been using gold grain
Conventional Salvage Treatment Options implantation as salvage treatment for NPC since 1986, and
for Local Failures and Selection Criteria the reported 5-year local control rates were 87 %, 63 %,
and 23 % for persistent disease, first recurrence, and second
Conventional Salvage Treatment Options recurrence, respectively [9]. Using high-dose-rate intracav-
Aggressive salvage treatment should be considered and itary intubation, Leung et al. reported an excellent 5-year
offered to patients with local failures of NPC whenever pos- local relapse-free survival rate of 85 % in 87 NPC patients
sible because a significant proportion of them can still with persistent local disease after radiotherapy [10]. The
achieve long-term survival after successful treatment. Many same group also reported a lower control rate when intra-
salvage treatments are available, and the choice of method cavitary brachytherapy alone was used in salvaging local
depends on several factors including extent of disease, site recurrence in a small group of patients (n = 8), with a 3-year
of involvement of disease, any synchronous nodal relapse, local relapse-free survival rate of 42 %. Surgical resection
cumulative radiation dose already received by the patient, of persistent or recurrent NPC has also been attempted
the patient’s general condition and preference, and expertise using various approaches, with reported control rates of
available. In general, salvage treatments for NPC can be 31–52 % [11–14], but is not commonly practiced because
classified into three types: surgery using different approaches of limited experience in most centers. For patients requiring
such as maxillary swing or transmandibular resection; external reirradiation usually due to extensive disease not
brachytherapy using intracavitary intubation, mold applica- amenable to brachytherapy or surgery, treatment results
tion, or gold grain implantation; and external reirradiation remained poor especially in advanced T stage. The reported
using various techniques including stereotactic radiosurgery 5-year survival rates after external reirradiation ranged
and radiation. from 7.6 to 36 % with the use of conventional two-
dimensional treatment planning and radiotherapy [15–17]
Patient Selection Criteria for Conventional and 12.4 % in a mixed cohort of patients treated either with
Salvage Treatment conventional two-dimensional radiotherapy or three-
A team approach involving both radiation oncologist and dimensional conformal radiotherapy [18]. A high incidence
head and neck surgeon is recommended in formulating a of late complication was commonly observed after external
treatment plan for local failures of NPC. All patients should reirradiation, the majority being neurologic damage and
be restaged by physical examination, nasopharyngoscopy soft tissue fibrosis. In our experience, patients with rT1–2
and biopsy, and computed tomography (CT) and/or magnetic NPC treated by conventional two-dimensional external
resonance imaging (MRI) of nasopharynx and neck. Patients reirradiation alone had a 5-year survival rate of about 57 %,
with disease confined to nasopharynx (rT1) are suitable can- but late neuroendocrine complications were common [17].
didates for either surgery or brachytherapy. For brachyther- The combination of intracavitary brachytherapy and exter-
apy, patients with unilateral small-volume disease are best nal radiotherapy has also been employed for recurrent NPC
treated by gold grain implantation, whereas those with more with the advantage of sparing more normal tissues while
diffuse or bilateral disease should be treated by intracavitary achieving a high total dose to the target. In one series by
intubation. Surgery should be considered in patients with Lee et al., combined intracavitary brachytherapy and
bulky disease, especially those already receiving a high external radiotherapy yielded a superior 5-year local con-
cumulative radiation dose to nasopharynx. Patients with lim- trol rate of 45 % compared with 32 % by external radio-
ited extension to nasal fossa, parapharyngeal space, and oro- therapy and 29 % by brachytherapy [19]. The outcome of
pharynx (rT2) may also be surgical candidates although most patients with local failures of NPC may improve with the
would require external reirradiation. For patients with skull use of intensity-modulated radiation treatment, which can
base involvement (rT3) or intracranial extension of disease achieve better dose distribution while sparing dose-limiting
(rT4), external reirradiation is usually the only option avail- critical structures [20]. Using intensity-modulated radiation
able, although reirradiation with curative intent is often treatment for high-dose (68–70 Gy) reirradiation of NPC,
difficult due to the large numbers of critical structures in the Lu et al. reported 100 % locoregional control rate in
vicinity of the target that were already irradiated to a high 49 patients, although the longest follow-up was only about
dose during primary radiotherapy. 1 year [21].
41 Head and Neck Tumors 531
Fig. 41.1 (a) Baseline CT of a patient with recurrent T1 nasopharyn- ered by single isocenter. The prescribed dose to target periphery was
geal carcinoma showing mass at left lateral wall of nasopharynx 12.5 Gy. (c) Follow-up CT taken 12 months after radiosurgery showing
(arrow). (b) Planning CT showing isodose distribution of target cov- complete resolution of tumor
Radiosurgery for Local Failures of NPC:

The Queen Mary Hospital Experience Case Study 41.2
(Case Study 41.1 and Case Study 41.2) A 45-year-old man was diagnosed with T3N0 NPC in
1997. He received a course of external radiotherapy
with a dose of 68 Gy in 34 fractions followed by a
Case Study 41.1
booster dose of 10 Gy in four fractions to the right
A 75-year-old man was diagnosed with T1N0 NPC in parapharyngeal space. He was found to have local
1985. He underwent external radiotherapy with a dose recurrence involving the right side of the nasophar-
of 61 Gy delivered in 26 fractions using a fraction dose ynx extending to the nasal fossa about 1 year after
of 2.5–3 Gy. The biological equivalent dose to 2 Gy treatment (Fig. 41.2a). Biopsies from the nasal fossa
(using α/β (alpha/beta) of 10) was 65 Gy. He was inci- and the right posterior wall both showed undifferenti-
dentally found to have local recurrence in 2000, after a ated carcinoma, but biopsies from the left side of the
long latency period of 15 years. Assessment showed nasopharynx were negative. The recurrent stage was
a small tumor involving the left lateral wall of the naso- rT2N0, which was not suitable for brachytherapy or
pharynx, and biopsy showed undifferentiated carci- surgery. The patient was treated with stereotactic
noma, hence rT1 disease (Fig. 41.1a). Biopsies from radiosurgery using two isocenters with collimator
posterior and right lateral walls were negative. Surgery sizes of 25 and 27.5 mm. Because of the larger bulk
was offered but refused by the patient, and the site of of tumor, a higher dose of 14 Gy was delivered to
lesion was considered not suitable for brachytherapy 80 % isodose line (Fig. 41.2b). Endoscopy and biop-
because of involvement of the cushion. Stereotactic sies after radiosurgery showed no evidence of resid-
radiosurgery using a single isocenter with 27.5-mm ual disease. Follow-up CT was performed at 6 months
collimator was performed, with a dose of 12.5 Gy deliv- after treatment, which showed complete resolution of
ered to 80 % isodose line (Fig. 41.1b). In this patient, tumor (Fig. 41.2c). He was found to have asymptom-
both imaging and endoscopy showed that tumor was atic right temporal lobe necrosis 15 months after
lateralized and confined to the lateral wall; hence, only radiosurgery. The complication was thought to be
part of the nasopharynx was treated. Follow-up CT scan related more to the first course of radiotherapy, which
performed at 3 and 12 months after radiosurgery showed delivered a high dose to the right temporal lobe to
complete resolution of tumor (Fig. 41.1c). Endoscopy ensure adequate coverage of tumor extent. He was
and biopsy after radiosurgery were also negative. The later found to have lung metastases and died of
patient has now been followed up for more than 4 years uncontrolled systemic disease 24 months after
with no evidence of relapse or major complications. radiosurgery.
All patients had previously received radical radiotherapy for

Patient Population undifferentiated type of NPC with or without concurrent/
Between March 1996 and February 2005, 48 patients adjuvant chemotherapy. Although the disease in some
received radiosurgery as a salvage treatment for local patients that relapsed after a long latency period may actu-
failures of NPC at Queen Mary Hospital, Hong Kong. ally represent a second primary in the nasopharynx, most
532 D.T.T. Chua
Fig. 41.2 (a) Baseline CT of a patient with recurrent T2 nasopharyn- of target covered by two isocenters. The prescribed dose to target
geal carcinoma showing mass lesion at right side of nasopharynx periphery was 14 Gy. (c) Follow-up CT taken 6 months after radiosur-
extending to nasal fossa. (b) Planning CT showing isodose distribution gery showing complete resolution of tumor
patients with local failures had residual tumor after prior bulky tumor or those with relapse at the cushion. Patients
treatment. We arbitrarily classified these residual tumors into with disease amenable to surgery or brachytherapy who
persistent and recurrent disease based on the time interval refused these treatments or were considered to be medically
from primary radiotherapy: persistent disease was defined as contraindicated for these procedures were also offered
local failure diagnosed within 6 months of completion of pri- radiosurgery.
mary radiotherapy, and recurrent disease was defined as that
occurring beyond 6 months. Using these definitions, half of Target Localization
the patients were treated for persistent disease and the other Tumor extent was assessed by imaging and endoscopic
half for recurrent disease. None of these patients received examination before radiosurgery. For imaging study, axial
additional treatment after radiosurgery unless progression or contrast CT with a slice thickness of 2.5–3 mm was per-
further failure was documented. Table 41.1 summarizes patient formed in all patients, supplemented by axial contrast MRI
characteristics of the whole group treated by radiosurgery: with a slice thickness of 3 mm in 73 % of patients. PET-CT
those with persistent disease and those with recurrent disease. was not performed in this patient group for target localiza-
Prior to radiosurgery, all patients were restaged by physi- tion, although we are currently also evaluating its role in
cal examination, nasopharyngoscopy and biopsy, and CT of salvaging local failures of NPC. Target volume was defined
nasopharynx and neck. Patients with disease confined to as any abnormal soft tissue mass and/or contrast-enhancing
nasopharynx or with limited extension to nasal fossa, areas as shown in axial imaging plus a margin of about
parapharyngeal space, and oropharynx were first evaluated 2–3 mm. Endoscopic examination was always performed
jointly by a surgeon and a radiation oncologist for brachy- irrespective of imaging findings to assess the mucosal extent
therapy or surgery, and radiosurgery was offered only to of tumor, including any spread to nasal cavity or oropharynx.
those patients whose disease was deemed not amenable In our experience, endoscopic examination is the most reli-
to these salvage treatments. At our institution, brachytherapy able means to access the mucosal extent of disease and
for persistent or recurrent NPC was usually performed using should not be omitted even when imaging showed that only
gold grain implantation. This technique was applicable only part of the nasopharynx was involved. During endoscopy,
to patients with a well-defined and relatively small-volume multiple biopsies were taken from both sides of the naso-
tumor confined to the nasopharynx. Patients with extensive pharynx to assess the macroscopic as well as microscopic
mucosal recurrence involving a large area were not suitable for extent of tumor. In patients just completing primary radio-
gold grain implantation. In addition, patients with relapse at therapy, biopsies were routinely taken from bilateral roofs,
the cushion near the opening of eustachian tube were also lateral walls, and posterior walls. In patients with tumor
not suitable for the procedure because gold grain could not apparently localized to one part of the nasopharyngeal
be directly implanted into the relatively thin mucosa overly- mucosa, mapping of tumor extent was performed by taking
ing the cartilaginous cushion. In general, patients with small- multiple biopsies from grossly uninvolved mucosa surround-
volume mucosal tumor were usually treated by gold grain ing the tumor. The results of mapping were then used to
implantation, with surgery reserved for those with more guide the extent of target coverage for radiosurgery.
Table 41.1 Characteristics of patients treated with radiosurgery for local failures of nasopharyngeal carcinoma at Queen
Mary Hospital, Hong Kong
Persistent disease (n = 24) Recurrent disease (n = 24) All (n = 48)
Gender
Male (n) 17 18 35
Female (n) 7 6 13
Age
<50 (n) 14 14 28
≥50 (n) 10 10 20
Median (years) 45 47 46
Range (years) 32–86 35–84 32–86
rT classification
rT1 (n) 18 9 27
rT2 (n) 2 4 6
rT3 (n) 4 7 11
rT4 (n) 0 4 4
Prior local failures
Yes (n) 0 9 9
No (n) 24 15 39
Previous salvage Rx
Surgery (n) 0 4 4
ERT (n) 0 7 7
Radiosurgery (n) 0 1 1
Cumulative ERT dose
65–68 Gy (n) 22 16 38
70–76 Gy (n) 2 1 3
125–130 Gy (n) 0 7 7
Interval from end of first ERT
≤6 months (n) 24 0 24
>6 months to 1 year (n) 0 4 4
>1–2 years (n) 0 3 3
>2–3 years (n) 0 5 5
>3 years (n) 0 12 12
Median (months) 4 36 8
Range (months) 3–6 9–197 3–197
ERT external radiotherapy
Radiosurgery Planning and Treatment adopted the approach of anchoring the head ring anteriorly to
Radiosurgery was performed using the commercial XKnife the zygomas to ensure complete coverage of the nasopharynx,
system (Radionics, Burlington, MA) to deliver multiple we did not find this to be necessary for most patients. The
non-coplanar arcs of photon to the target with a modified target volume was localized as described above. For rT1
6-MV linear accelerator (Clinac 600C; Varian, Milpitas, tumor, target confined to one side of nasopharynx was usually
CA). Head immobilization and target localization were per- covered using one isocenter, whereas target involving both
formed with the Brown–Roberts–Wells head frame and sides was covered using one or two isocenters. For rT2–4
stereotaxic system (Radionics). The head frame was ancho- tumors, target was usually covered by one or two isocenters.
red by placing two anterior head pins to the forehead and We found it seldom necessary to use more than two isocenters.
two to the occiput. The target volume was treated using 3–5 arcs of beams with a
The head frame was not placed in a strict horizontal posi- degree of 90–210°. The following dose limits were set to the
tion but with the posterior side tilted downward to ensure that critical structures: 5 Gy to brain stem, 4 Gy to optic appara-
the localizer ring would include the whole nasopharynx. tus, and 5 Gy to temporal lobes. An additional dose limit of
Using this approach, it was possible to treat the entire naso- 8 Gy to internal carotid artery was later set to reduce the risk
pharynx and adjacent soft tissues, although special attention of hemorrhage. Most patients received a dose of 12.5 Gy
should be given to those with disease extending down to delivered to the 80 % isodose line. Table 41.2 summarizes the
the junction of the oropharynx. Although some investigators radiosurgery treatment.
534 D.T.T. Chua
Table 41.2 Summary of radiosurgery treatment parameters of patients treated at Queen Mary Hospital, Hong Kong
Parameters Persistent disease Recurrent disease All
Number of isocenters
One (n) 20 19 39
Two (n) 4 4 8
Three (n) 0 1 1
Size of collimator
≤20 mm (n) 4 1 5
>20–30 mm (n) 19 16 35
>30 mm (n) 1 7 8
Median (mm) 25 30 25
Range (mm) 12.5–35 15–37.5 12.5–37.5
Target volume
<10 cm3 (n) 23 15 38
≥10 cm3 (n) 1 9 10
Median (cm3) 4.0 5.7 4.6
Range (cm3) 1.3–14.5 2.7–30.7 1.3–30.7
Prescribed dose to 80 % isodose line
<12 Gy (n) 3 5 8
12–13 Gy (n) 17 13 30
>13 Gy (n) 4 6 10
Median (Gy) 12.5 12.5 12.5
Range (Gy) 11.1–18 8–14 8–18
Dose to isocenter
<15 Gy (n) 4 5 9
15–16 Gy (n) 16 13 29
>16 Gy (n) 4 6 10
Median (Gy) 15.6 15.6 15.6
Range (Gy) 13.9–22.5 10–17.5 10–22.5
Tumor Control After Radiosurgery survival rates were 69 % for rT1 disease, 40 % for rT2–3
Thirty-seven (77 %) patients achieved complete regression disease, and 0 % for rT4 disease (Fig. 41.5). Patients with
of tumor after radiosurgery. All patients (100 %) treated for bulky tumor also had a poorer control rate after radiosurgery:
persistent disease achieved complete regression of disease 3-year local relapse-free survival rate was 61 % for target
compared with 13 (54 %) patients for recurrent disease. volume <10 cm3 compared with 20 % for target volume
Local failure after radiosurgery occurred in 22 (46 %) ≥10 cm3 (Fig. 41.6).
patients, regional failure in 5 (10 %) patients, and distant
metastasis in 7 (15 %) patients. For those patients who failed Complications
locally after radiosurgery, two received brachytherapy with Radiosurgery was well tolerated, and we did not observe any
intubation and both were salvaged, four had nasopharyngec- acute complications. In assessing late effects after treatment,
tomy and three were salvaged, and five had external radio- it was not possible to attribute the cause of complication to
therapy and two were salvaged. One patient with local failure radiosurgery or external radiotherapy. Some patients also
outside the radiosurgery-treated volume was salvaged by received another course of external radiotherapy either before
second radiosurgery. Three-year local relapse-free and over- or after radiosurgery, and they would expect to have a
all survival rates for all patients were 52 % and 66 %, respec- relatively high incidence of late complications even without
tively. Patients treated for persistent disease had better radiosurgery. Table 41.3 summarizes the late complications
outcome than those treated for recurrent disease: 3-year local observed in our patients, which included all observed compli-
relapse-free survival rate was 76 % in the former and 29 % in cations (other than xerostomia) irrespective of whether they
the latter (Fig. 41.3), and the corresponding 3-year overall were thought to be mainly caused by radiosurgery or not.
survival rates were 81 and 54 % (Fig. 41.4). Patients with Overall, 31 % of patients developed one or more late compli-
advanced T classification at the time of relapse had a poorer cations. The percentage decreased to 24 % if patients with two
control rate after radiosurgery: 3-year local relapse-free courses of radiotherapy before radiosurgery were excluded,
1.0 1.0
Local Relapse-free Survival Probability

Local Relapse-free Survival Probability
0.9 0.9
Persistent disease 0.8
0.8
0.7 0.7
0.6 0.6 rT1
0.5 0.5
0.4 0.4
Recurrent disease 0.3 rT2-3

0.3
0.2
0.2
rT4
p = 0.0004 0.1
0.1 p = 0.0042
0.0
0.0
0 12 24 36 48 60 72 84 96 108 120
0 12 24 36 48 60 72 84 96 108 120
Months after Radiosurgery Months after Radiosurgery
Fig. 41.3 Comparison of local relapse-free survival curves in patients Fig. 41.5 Comparison of local relapse-free survival curves according
with persistent and recurrent nasopharyngeal carcinoma treated by to recurrent T classification after stereotactic radiosurgery for local fail-
stereotactic radiosurgery at Queen Mary Hospital ures of nasopharyngeal carcinoma at Queen Mary Hospital
1.0 1.0
Local relapse-free survival probability
0.9 0.9
0.8 0.8
Overall Survival Probability
Persistent disease
0.7 0.7
0.6 0.6
0.5 0.5 Tumor volume < 10 cc
0.4 0.4
0.3 Recurrent disease 0.3

0.2 0.2 Tumor volume > or = 10 cc
0.1 p = 0.035 0.1 p = 0.0041
0.0 0.0
0 12 24 36 48 60 72 84 96 108 120 0 12 24 36 48 60 72 84 96 108 120
Months after Radiosurgery Months after radiosurgery
Fig. 41.4 Comparison of overall survival curves in patients with per- Fig. 41.6 Comparison of local relapse-free survival curves according
sistent and recurrent nasopharyngeal carcinoma treated by stereotactic to tumor volume after stereotactic radiosurgery for local failures of
radiosurgery at Queen Mary Hospital nasopharyngeal carcinoma at Queen Mary Hospital
Table 41.3 Late complications after radiosurgery for locally recurrent nasopharyngeal carcinoma at Queen Mary Hospital, Hong Kong
Complication One course of ERT (n = 36) Two courses of ERT (n = 12) One to two courses of ERT (n = 48)
Cranial neuropathy 2 5 7
Temporal lobe necrosis 4 3 7
Hypopituitarism 1 1 2
Carotid artery aneurysm 1 1 2
Trismus 0 1 1
None 29 4 33
ERT external beam radiotherapy, n number of patients
536 D.T.T. Chua
and it further decreased to 19 % if those with a second course median of 14 Gy. Significant regression of tumor was noted
of radiotherapy before or after radiosurgery were excluded. in five patients and limited regression in another three. Chang
Patients with more advanced tumor tend to have a higher risk et al. reported 15 patients with locally recurrent NPC who
of developing late complications. The percentages of patients received external reirradiation followed by radiosurgery
who developed late complications were 19 % in rT1 disease, using a dose ranging from 8 to 15 Gy [18] and noted a 3-year
33 % in rT2 disease, 46 % in rT3 disease, and 75 % in rT4 survival rate of 52 %. Pai et al. reported 36 patients with
disease. Common late complications observed were cranial recurrent NPC also treated with external reirradiation fol-
neuropathy and temporal lobe necrosis, both occurring in lowed by radiosurgery boost [28]. The radiosurgery dose to
about 15 % of patients. Two patients developed severe epi- target periphery ranged from 8 to 20 Gy with a median
staxis, and both were found to have internal carotid artery of 12 Gy. A 3-year local control rate of 58 % was achieved.
aneurysm located at cavernous sinus and retropharyngeal In these reports, it is uncertain whether radiosurgery was
space. Stenting was performed in both patients, which suc- used to boost the entire or partial tumor volume, with the
cessfully controlled the bleeding, and there were no treatment- intention to further escalate the total dose or to selectively
related deaths. boost the volume that was underdosed during external
reirradiation.
Radiosurgery Treatment Results for Local Fractionated Stereotactic Radiation

Failures of NPC in Other Series Fractionated stereotactic radiation instead of radiosurgery
has also been employed as a salvage treatment for locally
Radiosurgery Alone recurrent NPC. Mitsuhashi et al. treated three patients with
Firlik et al. first reported the use of Gamma Knife radiosur- rT1 NPC using stereotactic radiation at a dose of 50–64 Gy,
gery in a patient with recurrent NPC and noted complete and all three patients achieved complete response and
regression of tumor after delivering a dose of 20 Gy to 50 % remained free of local disease at 4–61 months [29]. The
isodose line [22]. Miller et al. reported another three patients report by Mitsuhashi et al. also included another patient with
with NPC that were also treated by Gamma Knife: one mucoepidermoid carcinoma of the nasopharynx treated by
patient was noted to have symptomatic improvement before stereotactic radiation after previous two courses of external
disease progression, and one patient had not responded radiotherapy, but the treatment was complicated by rupture
to the treatment [23]. Buatti et al. treated three patients of the internal carotid artery resulting in patient death. Using
with recurrent NPC by Linac radiosurgery using a dose of hypofractionated stereotactic radiation at a dose of 24 Gy in
12.5 Gy to 80 % isodose line [24]. Two patients also received 2–4 fractions, Orecchia et al. reported a less satisfactory out-
reirradiation, one before and the other after radiosurgery. come in 13 patients with locally recurrent NPC, with a 3-year
Of these two patients, one remained disease-free 1 year after survival rate of 31 % [30]. Ahn et al. treated 12 patients with
treatment, whereas the other had local recurrence 6 months recurrent NPC by stereotactic radiation using a median dose
after treatment. The third patient received radiosurgery of 54 Gy and reported a 2-year local control rate of 92 %
alone for recurrent disease and had neurologic deterioration [31]. Yau et al. also reported improved local control with
of uncertain etiology 6 months after the treatment. Kocher fractionated stereotactic radiation using a median dose of
et al. treated five patients with recurrent NPC at the skull 15 Gy at 6–8 Gy per fractions in patients with persistent local
base using radiosurgery with a dose of 15–24 Gy [25]. disease of NPC when compared with brachytherapy of 20 Gy
Only one patient’s tumor was controlled after treatment. [32]. Xiao et al. treated 50 patients with persistent or recur-
Three patients developed late complications, including two rent NPC with a dose ranging from 14 to 35 Gy using a frac-
with fatal internal carotid artery hemorrhage. Cmelak et al. tion dose of 5–15 Gy [33]. Of the 31 evaluable patients with
reported using Linac radiosurgery in treating 12 recurrent persistent disease, 94 % had complete response with a 1-year
NPC lesions in nine patients [26]. The dose delivered ranged disease-free survival rate of 47 %. Eighteen patients, most of
from 15 to 20 Gy with a median of 18 Gy. With a median them with rT3–4 tumor, were treated for recurrent disease.
follow-up of 17 months, the crude local control was 53 % The complete response rate was 56 % and 1-year disease-
(7/12), and only one patient developed radiation-induced free survival rate was 47 %. In Xiao’s series, however, 16 %
cranial neuropathy. of patients treated by fractionated radiation developed fatal
hemorrhage, probably due to the relatively high cumulative
Radiosurgery as a Boost After External dose delivered and large fractional dose used in some
Reirradiation patients. The largest series of fractionated stereotactic radia-
Chen et al. reported the outcome of 11 patients with rT3–4 tion in salvaging local failures was reported by Wu et al. who
NPC after conformal radiotherapy and Linac radiosurgery treated 90 patients with local failure of NPC following external
[27]. The radiosurgery dose ranged from 10 to 19 Gy with a beam radiotherapy [34]. Thirty-four patients had persistent
disease and received stereotactic radiotherapy using a included cranial neuropathy in four, radiological temporal
Linac-based gantry system with a median dose of 18 Gy in lobe necrosis in three, and retinopathy in one. In a subse-
three fractions. Fifty-six patients had recurrent disease and quent follow-up report that included 33 patients treated
they received a median of 48 Gy in six fractions. Three-year by Linac-based gantry system and 49 patients treated by
local failure-free rates in those treated for persistent and CyberKnife system, excellent local control was again noted
recurrent disease were 89.4 % and 75.1 %, respectively. The [38]. The 5-year local relapse-free rate was 98 %. Late
corresponding 3-year survival rates were 80.7 % and 45.9 %, complications occurred in 18 % of patients and included
respectively. Late complications occurred in 18.9 % of pati- radiological temporal lobe necrosis in ten, retinopathy in
ents, including two fatal hemorrhages. Multivariate analysis three, and carotid aneurysm in one. The high incidence of
showed that recurrent disease and large tumor volume were brain necrosis in their series was related to the high number
independent factors that predicted poor survival. In order to of patients who received boost treatment for T4 tumors with
study the relative efficacy of single versus fractionated intracranial extension, illustrating the higher risk of com-
stereotactic radiation in salvaging local failures of NPC, we plication and importance of dose conformity in treating this
conducted a matched-cohort study using our own series and group of patients. In these series, radiosurgery was deliv-
that of Wu et al. [35]. Forty-eight pairs of patients matched ered after treatment using a radical dose of external radio-
for failure type, recurrent T stage, and tumor volume were therapy with the aim of improving local control by dose
selected for comparison. Patients received a median dose of escalation. Ahn et al. used stereotactic radiotherapy instead
12.5 Gy in single fraction or 34 Gy in 2–6 fractions. Local as a boost treatment, with a dose ranging from 8 to 40 Gy
control rate at 3 years was better in those treated by multiple [31]. The dose delivered by external radiotherapy ranged
(83 %) than single fraction (51 %). Incidence of severe late from 36 to 61.2 Gy. The reported 4-year local control and
complications was higher with single (33 %) than multiple survival rates were 89 % and 75 %, respectively. One patient
(21 %) fractions, including brain necrosis (16 % versus 12 %) developed mucosal necrosis after the treatment. The
and bleeding (5 % versus 2 %). The differences reflect the approach by Ahn et al. was to reduce the dose delivered
better outcome as a result of a higher dose that can be safely by conventional radiotherapy and substitute it by stereotac-
delivered using fractionated stereotactic radiation. tic radiotherapy thereby reducing the incidence of late
complications.
Radiosurgery as a Boost Treatment After

Radiotherapy for Newly Diagnosed NPC Prognostic Scoring System for Prediction
of Treatment Outcome After Radiosurgery
Radiosurgery has also been employed as a planned boost for NPC
treatment after radiotherapy for newly diagnosed NPC. The
treatment is similar in principle to the use of intracavitary To better define patient group with local failures of NPC that
intubation or parapharyngeal boost commonly given after may benefit from radiosurgery, we have designed a prognos-
conventional two-dimensional radiotherapy. Cmelak et al. tic scoring system based on five adverse factors including
treated 11 patients using radiosurgery boost with a median age >45, time interval >6 months from primary radiotherapy,
dose of 12 Gy (7–16 Gy) after a dose of 64.8–70 Gy was recurrent T4 disease, prior local failure, and tumor volume of
delivered by external radiotherapy [26]. The reported control 10 cc or above [39]. Prognostic score was calculated using a
rate was 91 % without any major late complications. formula based on the presence or absence of these factors
Investigators at Stanford University employed single fraction with a differential weighting: score = 0.22 × age (0,1) + 0.27 x
of radiosurgery as boost treatment using Linac-based gantry time interval >6 months (0,1) + 0.05 × rT4 disease
system following standard course of chemoradiotherapy in (0,1) + 0.28 × tumor volume ≥10 cc (0,1) + prior local recur-
patients with NPC and reported an excellent local control. rence (0,1). Patients were then classified as good (score: 0),
Tate et al. first reported 23 patients treated with a median intermediate (score: >0–0.5), or poor (score: >0.5) prognostic
radiosurgery boost dose of 12 Gy (7–15 Gy) after external group according to the scores. The 5-year local failure-free
radiotherapy; the latter delivered a dose of 64.8–70 Gy to the rates in patients with good, intermediate, and poor prognos-
primary [36]. All patients had their local tumors controlled tic groups after radiosurgery were 100 %, 42.5 %, and 9.6 %,
after the treatment although 35 % subsequently developed respectively. The corresponding 5-year overall survival rates
regional or distant failures. Le et al. then reported a follow- were 100 %, 5.1 %, and 0 %, respectively. The prognostic
up series of 45 patients treated with the same median dose, scoring system was subsequently validated using published
and the local control rate was 100 % at 31 months [37]. Late reports that contained sufficient clinical information for esti-
complications however occurred in 17.8 % of patients and mation of the prognostic scores [40].
538 D.T.T. Chua
Table 41.4 Summary of reported treatment outcomes after stereotactic radiosurgery/radiotherapy for local failures of NPC
Local control and Late complications
Authors No. of patients rT classification Treatment SRS/SRT dose survival rates crude rate: all (fatal)
Chua et al. 48 rT1–2: 69 % SRS 8–18 Gy LC: 52 % (3 years) 31 % (0 %)
(current report)
rT3–4: 31 % OS: 66 % (3 years)
Cmelak et al [26] 9 Not reported SRS 15–20 Gy LC: 58 % (crude) 11 % (0 %)
Kocher et al. [25] 5 rT3–4: 100 % SRS 15–24 Gy LC: 20 % (crude) 60 % (40 %)
Pai et al. [28] 36 rT1–2: 64 % ER + SRS 8–20 Gy LC: 58 % (3 years) 22 % (0 %)
rT3–4: 36 % OS: 54 % (3 years)
Chang et al. [18] 15 rT1–2: 67 % ER + SRS 8–15 Gy 52 % (3-year survival) Not separately reported
rT3–4: 33 %
Chen et al. [27] 11 rT3–4: 100 % ER + SRS 10–19 Gy LC: 45 % (crude) Not reported
OS: 37 % (2 years)
Xiao et al. [33] 50 rT1–2: 38 % SRT 14–35 Gy CR: 76 % 34 % (16 %)
rT3–4: 62 % OS:
P 78 % (2 years)
R 42 % (2 years)
Wu et al. [34] 90 rT1–2: 57 % SRT P: 10–24 Gy LC: 89 % (3 years) 19 % (2 %)
Orecchia et al. [30] 13 rT3–4: 43 % SRT R: 20–49 Gy OS: 81 % (3 years) 0 % (0 %)
rT1–2: 23 % 24 Gy LC: 75 % (3 years)
OS: 46 % (3 years)
LC:31 % (crude)
rT3–4: 77 % OS: 31 % (3 years)
Ahn et al. [31] 12 Not reported SRT 45–65 Gy LC: 92 % (2 years) Not reported (8 %)
OS: 60 % (2 years)
SRS stereotactic radiosurgery, SRT stereotactic radiotherapy, ER external beam reirradiation, LC local control, OS overall survival, CR complete
response, P persistent disease, R recurrent disease
Summary and Recommendations or radiosurgery, which may lead to underestimation of

complication risks of radiosurgery. Although most series
Table 41.4 summarizes the outcome after radiosurgery for reported a relatively low risk of late complications, hemor-
local failures of NPC as reported in different series. Based on rhage remains the most severe form of complication after
these results, there is strong evidence indicating that radio- radiosurgery with a possible fatal outcome. Although hemor-
surgery is an effective salvage treatment for local failures of rhage could also occur as a complication of tumor progres-
NPC. There is, however, no data comparing the relative effi- sion, most cases that developed after radiosurgery were
cacy and complication risks of radiosurgery with other sal- probably due to radiation damage to the carotid artery as a
vage options for NPC. In practice, selection of treatment result of high cumulative dose. The complication is probably
modalities depends mainly on extent of disease and expertise not related to rapid shrinkage of tumor as suggested by
available. For rT1–2 disease, treatment outcome after radio- Kocher et al. [25], because hemorrhage is rare after chemo-
surgery appears to be comparable with that of brachytherapy therapy and brachytherapy, both of which can induce rapid
and surgery. For rT3–4 disease, radiosurgery also seems to tumor shrinkage. To minimize the risk of hemorrhage, it is
yield similar results with conventional two-dimensional reir- advised to use single fraction radiosurgery only in the
radiation. The advent of three-dimensional conformal radio- absence of direct tumor encasement of carotid artery; other-
therapy and intensity-modulated radiation treatment appears wise the patient should be treated by stereotactic or fraction-
to improve the outcome of recurrent NPC, particularly in ated radiotherapy using a small fraction dose. A proposed
those with more advanced tumor. Hence, patients with rT3–4 decision tree for salvaging local failures of NPC is outlined
disease should be treated by external beam reirradiation in Fig. 41.7. The role of radiosurgery as a boost treatment
using these new techniques with radiosurgery reserved as a after radical radiotherapy in newly diagnosed NPC is more
boost treatment or for further recurrence. In assessing late difficult to define, although its impact is likely to be
complications after radiosurgery, it was always difficult if not small with the advent of modern conformal radiotherapy,
impossible to attribute them to either external radiotherapy chemotherapy, and intensity-modulated radiation therapy.
Yes Yes Brachytherapy (gold grain implantation

Tumor confined Unilateral
or intracavitary intubation) or SRS or
to nasopharynx? involvement?
nasopharyngectomy
No No
SRS
rT3/4 disease? Yes Yes Yes

(skull base or Persistent Encasement of SRT or IMRT
intracranial disease? carotid artery
invasion) or tumor
volume? SRS
10 cc? No
No
No
Encasement of Yes
SRT or IMRT
carotid artery or
tumor volume?
10 cc? SRS
No
Fig. 41.7 A proposed decision tree for salvaging local failure of nasopharyngeal carcinoma. SRS stereotactic radiosurgery, SRT stereotactic radio-
therapy, IMRT intensity-modulated radiation therapy
refusal for conventional treatments. Six patients were treated

Other Head and Neck Tumors for squamous cell carcinoma of the head and neck, two for
basal cell carcinoma, one for bilateral paragangliomas, and
Radiosurgery has been employed in the treatment of head one for Hurtle cell carcinoma of thyroid. Two lesions were
and neck tumors other than NPC. A wide range of tumors, treated with single fraction of 18 Gy, whereas the remaining
both benign and malignant, has been treated using either ste- lesions received 30–48 Gy in 5–6 fractions. Complete
reotactic radiosurgery or radiotherapy. Most treated cases response rate was 64 %. The 1- and 2-year local control rates
were recurrent tumors after prior surgery and/or radiother- were 83.3 % and 66.7 %, respectively. Median overall sur-
apy. The more commonly treated histologic type includes vival was 28.7 months. The only severe late complication
squamous cell carcinoma, adenoid cystic carcinoma, and was observed in the patient treated for paraganglioma who
chordoma. Most lesions were located at the skull base, mak- developed severe and persistent facial pain following
ing them ideal targets for radiosurgery. For lesions located treatment.
below the skull base, fixation and immobilization are not as
rigid as for intracranial target, which should be taken into
consideration during localization of target and adjacent criti- Stereotactic Radiotherapy as Salvage
cal structures. The use of image-guided radiotherapy and Treatment for Recurrent Tumors
system capable of delivering radiation with high-dose con-
formity such as CyberKnife may be more suitable and is Stereotactic Radiation Delivered by Gamma
increasingly being used in treating head and neck cancers. Knife or Linac-Based Gantry System
Cmelak et al. [26] treated 36 recurrent head and neck tumors
in 29 patients by radiosurgery using a median dose of 20 Gy
Stereotactic Radiotherapy as Primary (range: 7–35 Gy). Fifteen patients were treated for squamous
Treatment for Newly Diagnosed Tumors cell carcinoma, six for adenocarcinoma, four for malignant
meningioma, three for adenoid cystic carcinoma, and three
Use of stereotactic radiotherapy as primary treatment of for melanoma. Local control rate was 72 % but late compli-
newly diagnosed head and neck tumors was reported in a cations including CSF leakage, trismus, and cranial neuropa-
small series by Siddiqui et al. [41]. Eleven lesions in ten thy occurred in four patients after treatment. Miller et al. [23]
patients were treated by stereotactic radiation using a Linac treated 32 new or recurrent head and neck tumors in 29
gantry-based system (BrainLab, Inc., Heimstetten, Germany). patients using radiosurgery with or without external radio-
Rationale for using stereotactic radiation as primary treat- therapy. Twelve patients were treated for adenoid cystic
ment included comorbidities that precluded patients from carcinoma, eight for squamous cell carcinoma, and eight for
receiving surgery or conventional radiation and patient chordoma. The dose delivered by radiosurgery was in the
540 D.T.T. Chua
range of 12–20 Gy (median, 15 Gy). Local control rates were cell carcinomas by CyberKnife system using a median dose
83 % for adenoid cystic carcinoma and 100 % for chordoma. of 30 Gy in 3–8 fractions. The authors reported an overall
Late complications occurred in four patients and included survival rate of 50 % at 2 years, but 28.6 % of patients
memory loss, hypopituitarism, CSF leakage, brain edema, developed severe late complications including two deaths
and optic neuropathy. Kocher et al. [25] treated four patients due to massive hemorrhage in the pharynx. Roh et al. [46]
with recurrent sarcoma using radiosurgery and achieved treated 44 sites of recurrent head and neck tumors in 36
good tumor control in three. They also treated three patients patients using CyberKnife with a median dose of 30 Gy
with recurrent squamous cell or adenoid cystic carcinoma (range: 18–40 Gy) in 3–5 fractions to the 65–85 % isodose
but achieved local control in only one patient. The radiosur- line and reported a 42.9 % complete response rate. One- and
gery dose used was in the range of 9–20 Gy. One patient 2-year local recurrence-free survival rates were 61 % and
developed brain necrosis and cranial neuropathy after radio- 52.2 %, respectively. The corresponding overall survival
surgery. Ryu et al. [42] treated five patients with recurrent rates were 52.1 % and 30.9 %, respectively. Late complica-
squamous cell or mucoepidermoid carcinoma by radiosur- tions were observed in three patients (8.6 %) including one
gery with a dose of 12–18 Gy and achieved complete soft tissue necrosis, one mandibular bone necrosis, and one
response in two patients. They also treated five recurrent skull base necrosis. One patient died due to late complica-
tumors (squamous cell, adenoid cystic, and basal cell carci- tions. Voynov et al. [47] at the University of Pittsburg treated
noma) and three metastatic cervical nodes by stereotactic 22 patients with recurrent head and neck tumors by
radiotherapy using a dose ranging from 30 to 36 Gy in six CyberKnife using a median radiosurgery dose of 24 Gy
fractions and noted three complete responses in the former with various fractionation schemes. The reported 2-year
group and one in the latter group. Elshaikh et al. [43] treated local control rate was 26 % and overall survival rate was
12 patients with recurrent chemodectoma (seven with first 22 %. They did not observe any significant late complica-
recurrence after surgery and five with second recurrence tions initially after a median follow-up of 19 months. Based
after surgery) using Gamma Knife radiosurgery at a dose of on an expanded cohort from the same institution, Rwigema
13–16 Gy and achieved 83 % control rate without any major et al. later [48] reported the outcome in 96 patients with
late complications. Using stereotactic radiotherapy, Ahn locally recurrent head and neck tumors treated by Cyber
et al. [31] treated three patients with steroid refractory orbital Knife or Trilogy system and observed a higher locoregional
pseudotumor using a dose of 20 Gy in ten fractions and noted control rate with the use of higher radiotherapy dose espe-
good response in all patients. Siddiqui et al. treated 29 lesions cially in patients with large tumor volume. The 1-, 2-, and
in 21 patients with recurrent head and neck tumors using a 3-year locoregional control rates for patients who received
Linac gantry-based system [41]. Ten lesions received single doses of 40–50 Gy were 69.4 %, 57.8 %, and 41.1 %,
fraction of 16–18 Gy, and 19 lesions received 30–48 Gy in respectively. The corresponding rates for patients who
6–8 fractions. The 1- and 2-year local control rates were received doses of 15–36 Gy were 51.9, 31.7, and 15.9 %.
60.6 % and 40.4 %, respectively. Median overall survival The 1- and 2-year overall survival rates were 58.9 % and
was 6.7 months. Three patients developed fistula, and one 28.4 %, respectively. Grade 3 late complications occurred in
patient developed mucosal ulceration and dysphagia after three patients including two with dysphagia and one with
treatment soft tissue fibrosis. There were no grade 4 or 5 complica-
tions, and no relationship between rates of complications
Stereotactic Radiation Delivered and prescribed doses was found. Unger et al. [49] at
by CyberKnife System Georgetown University reported the outcome of 65 patients
Kawaguchi et al. [44] treated 22 patients with advanced, with recurrent head and neck tumors treated by CyberKnife
recurrent head and neck tumor using CyberKnife. Eight with a median dose of 30 Gy in 2–5 fractions. The cohort
patients had synchronous regional nodal metastases. All included patients treated for radical (58 %) or palliative
patients had prior radical radiotherapy. Marginal radiosur- (42 %, those with distant metastases) intent. About half
gery doses were 29–42 Gy delivered in 2–5 fractions. All of the patients also received concurrent chemotherapy.
patients also received oral chemotherapy using S-1 after Complete responses were achieved in 54 % of patients, with
radiosurgery. At a median follow-up of 24 months, 9 patients a 2-year locoregional rate of 30 % and overall survival rate
(64.3 %) without regional metastases remained in complete of 41 %. Multivariate analysis showed that higher total
remission as compared with only one patient (12.5 %) in dose, prior surgical resection, and nasopharynx site were
those with regional metastases. Actuarial survival rates at associated with better locoregional control. Severe late radi-
2 years for patients with and without regional nodal metasta- ation complications occurred in 11 % of patients including
ses were 12.5 % and 78.6 %, respectively. Kodani et al. [45] two arterial bleeding requiring embolization and one
treated 21 patients with recurrent head and neck squamous treatment-related death.
Summary and Recommendations 11. Fee Jr WE, Roberson Jr JB, Goffinet DR. Long-term survival
after surgical resection for recurrent nasopharyngeal cancer after
radiotherapy failure. Arch Otolaryngol Head Neck Surg. 1991;
Because of the heterogeneous nature of the patient population 117:1233–6.
treated in these series, it is not possible to conclude the efficacy 12. Morton RP, Liavaag PG, McLean M, Freeman JL. Transcervico-
and indication of radiosurgery for specific type and site of head mandibulo-palatal approach for surgical salvage of recurrent naso-
and neck tumors. Most series treated patients with recurrent pharyngeal cancer. Head Neck. 1996;18:352–8.
13. King WW, Ku PK, Mok CO, Teo PM. Nasopharyngectomy in the
squamous cell carcinoma of oropharynx and hypopharynx and treatment of recurrent nasopharyngeal carcinoma: a twelve-year
reported satisfactory local control rate. Experiences gathered experience. Head Neck. 2000;22:215–22.
from treatment of NPC may also be relevant and applicable to 14. Wei WI. Salvage surgery for recurrent primary nasopharyngeal
radiosurgery treatment of other head and neck tumors. Currently carcinoma. Crit Rev Oncol Hematol. 2000;33:91–8.
15. Teo PM, Kwan WH, Chan AT, Lee WY, King WW, Mok CO. How
it seems reasonable to recommend stereotactic radiosurgery or successful is high dose (≥60 Gy) reirradiation using mainly exter-
radiotherapy as salvage treatment in patients with recurrent nal beams in salvaging local failures of nasopharyngeal carcinoma?
head and neck tumors not amenable to further surgery if the site Int J Radiat Oncol Biol Phys. 1998;40:897–913.
and extent of the lesions are a suitable target of radiosurgery. 16. Öksüz DÇ, Meral G, Uzel Ö, Çağatay P, Turkan S. Reirradiation
for locally recurrent nasopharyngeal carcinoma: treatment results
and prognostic factors. Int J Radiat Oncol Biol Phys. 2004;60:
388–94.
17. Chua DT, Sham JS, Kwong DL, Wei WI, Au GK, Choy D. Locally
References recurrent nasopharyngeal carcinoma: treatment results for patients
with computed tomography assessment. Int J Radiat Oncol Biol
1. Lee AW, Poon YF, Foo W, Law SC, Cheung FK, Chan DK, et al. Phys. 1998;41:379–86.
Retrospective analysis of 5037 patients with nasopharyngeal carci- 18. Chang JT, See LC, Liao CT, Ng SH, Wang CH, Chen IH, et al.
noma treated during 1976–1985; overall survival and patterns of Locally recurrent nasopharyngeal carcinoma. Radiother Oncol.
failure. Int J Radiat Oncol Biol Phys. 1992;23:261–70. 2000;54:135–42.
2. Lee AW, Sze WM, Au JS, Leung SF, Leung TW, Chua DT, et al. 19. Lee AW, Foo W, Law SC, Poon YF, Sze WM. O SK et al.
Treatment results for nasopharyngeal carcinoma in the modern era: Reirradiation for recurrent nasopharyngeal carcinoma: factors
the Hong Kong experience. Int J Radiat Oncol Biol Phys. 2005; affecting the therapeutic ratio and ways for improvement. Int J
61:1107–16. Radiat Oncol Biol Phys. 1997;38:43–52.
3. Al-Sarraf M, LeBlanc M, Shanker Giri PG, Fu KK, Cooper J, 20. Hsiung CY, Yorke ED, Chui CS, Hunt MA, Ling CC, Huang EY,
Vuong T, et al. Chemoradiotherapy versus radiotherapy in patients et al. Intensity-modulated radiotherapy versus conventional three-
with advanced nasopharyngeal cancer: phase III randomized inter- dimensional conformal radiotherapy for boost or salvage treatment
group study 0099. J Clin Oncol. 1998;16:1310–7. of nasopharyngeal carcinoma. Int J Radiat Oncol Biol Phys. 2002;
4. Chan AT, Teo PM, Ngan RK, Leung TW, Lau WH, Zee B, et al. 53:638–47.
Concurrent chemotherapy-radiotherapy compared with radiother- 21. Lu TX, Mai WY, Teh BS, Zhao C, Han F, Huang Y, et al. Initial
apy alone in loco-regionally advanced nasopharyngeal carcinoma: experience using intensity-modulated radiotherapy for recurrent
progression-free survival analysis of a phase III randomized trial. nasopharyngeal carcinoma. Int J Radiat Oncol Biol Phys. 2004;
J Clin Oncol. 2002;20:2038–44. 58:682–7.
5. Lin JC, Jan JS, Hsu CY, Liang WM, Jiang RS, Wang WY. Phase III 22. Firlik KS, Kondziolka D, Lunsford LD, Janecka IP, Flickinger
study of concurrent chemoradiotherapy versus radiotherapy alone JC. Radiosurgery for recurrent cranial base cancer arising from the
for advanced nasopharyngeal carcinoma: positive effect on overall head and neck. Head Neck. 1996;18:160–6.
and progression-free survival. J Clin Oncol. 2003;21:631–7. 23. Miller RC, Foote RL, Coffey RJ, Gorman DA, Earle JD, Schomberg
6. Chua DT, Ma J, Sham JST, Mai HQ, Choy DT, Hong MH, et al. PJ, et al. The role of stereotactic radiosurgery in the treatment of
Long-term survival after cisplatin-based induction chemotherapy malignant skull base tumors. Int J Radiat Oncol Biol Phys. 1997;
and radiotherapy for nasopharyngeal carcinoma: a pooled data 39:977–81.
analysis of two phase III trials. J Clin Oncol. 2005;23:1118–24. 24. Buatti JM, Friedman WA, Bova FJ, Mendenhall WM. Linac radio-
7. Chua DT, Sham JS, Kwong DL, Tai KS, Wu PM, Lo M, et al. surgery for locally recurrent nasopharyngeal carcinoma: rationale
Volumetric analysis of tumor extent in nasopharyngeal carcinoma and technique. Head Neck. 1995;17:14–9.
and correlation with treatment outcome. Int J Radiat Oncol Biol 25. Kocher M, Voges J, Staar S, Treuer H, Sturm V, Mueller RP. Linear
Phys. 1997;39:711–9. accelerator radiosurgery for recurrent malignant tumors of the skull
8. Wu PM, Chua DT, Sham JS, Leung L, Kwong DL, Lo M, et al. base. Am J Clin Oncol. 1998;21:18–22.
Tumor control probability of nasopharyngeal carcinoma: a com- 26. Cmelak AJ, Cox RS, Adler JR, Fee Jr WE, Goffinet
parison of different mathematical models. Int J Radiat Oncol Biol DR. Radiosurgery for skull base malignancies and nasopharyngeal
Phys. 1997;37:913–20. carcinoma. Int J Radiat Oncol Biol Phys. 1997;37:997–1003.
9. Kwong DL, Wei WI, Cheng AC, Choy DT, Lo AT, Wu PM, et al. 27. Chen HJ, Leung SW, Su CY. Linear accelerator based radiosurgery
Long term results of radioactive gold grain implantation for the as a salvage treatment for skull base and intracranial invasion of
treatment of persistent and recurrent nasopharyngeal carcinoma. recurrent nasopharyngeal carcinoma. Am J Clin Oncol. 2001;
Cancer. 2001;91:1105–13. 24:255–8.
10. Leung TW, Tung SY, Sze WK, Sze WM, Wong VY. O SK. Salvage 28. Pai PC, Chuang CC, Wei KC, Tsang NM, Tseng CK, Chang
brachytherapy for patients with locally persistent nasopharyngeal CN. Stereotactic radiosurgery for locally recurrent nasopharyngeal
carcinoma. Int J Radiat Oncol Biol Phys. 2000;47:405–12. carcinoma. Head Neck. 2002;24:748–53.
542 D.T.T. Chua
29. Mitsuhashi N, Sakurai H, Katano S, Kurosaki H, Hasegawa M, nasopharyngeal carcinoma. Int J Radiat Oncol Biol Phys. 2006;
Akimoto T, et al. Stereotactic radiotherapy for locally recurrent 66:1415–21.
nasopharyngeal carcinoma. Laryngoscope. 1999;109:805–9. 40. Chua DT, Hung KN, Lee V, Ng SC, Tsang J. Validation of a prognos-
30. Orecchia R, Redda MGR, Regona R, Nassisi D, Jereczek-Fossa B, tic scoring system for locally recurrent nasopharyngeal carcinoma
Zurrida S, et al. Results of hypofractionated stereotactic re-irradiation treated by stereotactic radiosurgery. BMC Cancer. 2009;9:131.
on 13 locally recurrent nasopharyngeal carcinoma. Radiother Oncol. 41. Siddiqui F, Patel M, Khan M, McLean S, Dragovic J, Jin JY, et al.
1999;53:23–8. Stereotactic body radiation therapy for primary, recurrent, and met-
31. Ahn YC, Lee KC, Kim DY, Huh SJ, Yeo IH, Lim DH, et al. astatic tumors in the head-and-neck region. Int J Radiat Oncol Biol
Fractionated stereotactic radiation therapy for extracranial head and Phys. 2009;74:1047–53.
neck tumors. Int J Radiat Oncol Biol Phys. 2000;48:501–5. 42. Ryu S, Khan M, Yin FF, Concus A, Ajlouni M, Benninger MS,
32. Yau TK, Sze WM, Lee WM, Yeung MW, Leung KC, Hung WM, et al. Image-guided radiosurgery of head and neck cancers.
et al. Effectiveness of brachytherapy and fractionated stereotactic Otolaryngol Head Neck Surg. 2004;130:690–7.
radiotherapy boost for persistent nasopharyngeal carcinoma. Head 43. Elshaikh MA, Mahmoud-Ahmed AS, Kinney SE, Wood BG, Lee
Neck. 2004;26:1024–30. JH, Barnett GH, et al. Recurrent head-and-neck chemodectomas: a
33. Xiao JP, Xu GZ, Miao YJ. Fractionated stereotactic radiosurgery comparison of surgical and radiotherapeutic results. Int J Radiat
for 50 patients with recurrent or residual nasopharyngeal carci- Oncol Biol Phys. 2002;52:953–6.
noma. Int J Radiat Oncol Biol Phys. 2001;51:164–70. 44. Kawaguchi K, Sato K, Horie A, Iketani S, Yamada H, Nakatani Y,
34. Wu SX, Chua DT, Deng M, Zhao C, Li FY, Sham JS, et al. Outcome et al. Stereotactic radiosurgery may contribute to overall survival
of fractionated stereotactic radiotherapy for 90 patients with locally for patients with recurrent head and neck carcinoma. Radiat Oncol.
persistent and recurrent nasopharyngeal carcinoma. Int J Radiat 2010;5:51.
Oncol Biol Phys. 2007;69:761–9. 45. Kodani N, Yamazaki H, Tsubokura T, Shiomi H, Kobayashi K,
35. Chua D, Wu S, Lee V, Tsang J. Comparison of single versus Nishimura T, et al. Stereotactic body radiation therapy for head and
fractionated dose of stereotactic radiotherapy for salvaging local neck tumors: disease control and morbidity outcomes. J Radiat Res.
failures of nasopharyngeal carcinoma: a matched-cohort analysis. 2011;52:24–31.
Head Neck Oncol. 2009;1:13. 46. Roh KW, Jang JS, Kim MS, Sun DI, Kim BS, Jung SL, et al.
36. Tate DJ, Adler JR, Chang SD, Marquez S, Eulau SM, Fee WE, Fractionated stereotactic radiotherapy as reirradiation for locally
et al. Stereotactic radiosurgical boost following radiotherapy in pri- recurrent head and neck cancer. Int J Radiat Oncol Biol Phys.
mary nasopharyngeal carcinoma: impact on local control. Int J 2009;74:1348–455.
Radiat Oncol Biol Phys. 1999;45:915–21. 47. Voynov G, Heron DE, Burton S, Grandis J, Quinn A, Ferris R, et al.
37. Le QT, Tate D, Koong A, Gibbs IC, Chang SD, Adler JR, et al. Frameless stereotactic radiosurgery for recurrent head and neck
Improved local control with stereotactic radiosurgical boost in carcinoma. Technol Cancer Res Treat. 2006;5:529–35.
patients with nasopharyngeal carcinoma. Int J Radiat Oncol Biol 48. Rwigema JC, Heron DE, Ferris RL, Andrade RS, Gibson MK,
Phys. 2003;56:1046–54. Yang Y, et al. The impact of tumor volume and radiotherapy dose
38. Hara W, Loo Jr BW, Goffinet DR, Chang SD, Adler JR, Pinto HA, on outcome in previously irradiated recurrent squamous cell carci-
et al. Excellent local control with stereotactic radiotherapy noma of the head and neck treated with stereotactic body radiation
boost after external beam radiotherapy in patients with nasophary- therapy. Am J Clin Oncol. 2011;34:372–9.
ngeal carcinoma. Int J Radiat Oncol Biol Phys. 2008;71: 49. Unger KR, Lominska CE, Deeken JF, Davidson BJ, Newkirk KA,
393–400. Gagnon GJ, et al. Fractionated stereotactic radiosurgery for reirra-
39. Chua DT, Sham JS, Hung KN, Leung L, Au GK. Predictive factors diation of head-and-neck cancer. Int J Radiat Oncol Biol Phys.
of tumor control and survival after radiosurgery for local failures of 2010;77:1411–9.
Head and Neck Tumors:
Scharukh Jalisi
In this article the authors randomized patients with T2–T4

Introduction piriform sinus cancer to either total laryngectomy with partial
pharyngectomy and neck dissection or to three doses of
Seventy-five thousand new cases of head and neck cancer are induction chemotherapy, followed by concomitant chemora-
reported in the United States each year. Incidence is rising in diation in responders (nonresponders were placed in surgical
thyroid (up 52 %), bone (43 %), soft tissues (20 %), salivary arm). This study showed no difference in local, regional, and
(20 %), tongue (16 %), tonsil (12 %), and nose (12 %). 5-year disease-free survival between the two groups.
Incidence is falling in lip (down 58 %), hypopharynx (35 %), Unfortunately this study excluded any advanced case requir-
cervical esophagus (32 %), oropharyngeal mucosa (26 %), ing plastic surgery repair and did not have any data on func-
and larynx (26 %). There were 30,000 deaths from head and tionality of the larynx.
neck cancer in 2001 [1]. The most common risk factors are Since then multiple articles have been published compar-
alcohol and tobacco, but more recently there is interest and ing the efficacy of radiation with or without chemotherapy
detection of human papillomavirus (most commonly HPV and have supplanted the induction chemotherapy paradigm
16) as an etiologic factor, particularly in the nonsmoker. set by the VA laryngeal study with concomitant chemoradia-
The National Comprehensive Cancer Network® (NCCN®), tion therapy (CRT) protocols. The Radiation Therapy
a not-for-profit alliance of 21 of the world’s leading cancer Oncology Group trial (RTOG) 9111 was instrumental in
centers was formed and helps produce guidelines to treat can- demonstrating improved survival with concurrent CRT [11].
cers by different subsites. The treatment for head and neck The problem again has been that this randomized trial did
cancer has been either surgery or radiation for early stage not have a surgical arm. Similarly it seems that nonsurgical
disease and a combined modality treatment (chemotherapy, trials are being extrapolated onto treating head and neck can-
radiation, and surgery) for advanced disease cancer. cers without adequate comparisons to the gold standard sur-
gery. In the example of larynx cancer, the National Cancer
Data Base (NCDB) demonstrates a decrease in survival in
Treatment Options recent years that parallels the observed trend of increasing
use of nonoperative treatment and is not due to an increased
There has been an increase in the utilization of nonsurgical incidence of advanced-stage disease [4, 6].
therapies for head and neck cancer since the publication of Despite advances in surgery and CRT, local/regional
the Veterans Affairs Laryngeal Study Group results in 1991 recurrence rates were around 30 %, distant metastases devel-
[2–9]. This study showed that 66 % of patients could preserve oped in 25 %, and 5-year survival was around 40 %.
their larynx without changing their survival if undergoing Furthermore, in the presence of high-risk features such as
nonsurgical therapy versus total laryngectomy. The only other extracapsular extension, positive surgical margins, and oth-
randomized trial in head and neck cancer treatment compar- ers, recurrence rates of up to 61 % had been observed, and
ing surgery to nonsurgical treatment is by Lefebvre et al. [10]. 5-year survival dipped to around 30 % [11]. For this reason
the RTOG 9501 [12] trial was designed to see the effect of
S. Jalisi, M.D., M.A., F.A.C.S. (*) adding cisplatin-based chemotherapy to postoperative radia-
Division of Head and Neck Surgical Oncology and Skull Base tion therapy. Two hundred and thirty-one patients were ran-
Surgery, Department of Otolaryngology and Neurological
domly assigned to receive radiotherapy alone (60–66 Gy in
Sciences, Boston Medical Center, Boston University School
of Medicine, Boston, MA, USA 30–33 fractions over a period of 6–6.6 weeks) and 228
e-mail: scharukh.jalisi@bmc.org patients to receive the identical treatment plus concurrent
544 S. Jalisi
cisplatin (100 mg per square meter of body surface area oropharyngeal cancers, and endoscopic and robotic techniques
intravenously on days 1, 22, and 43). The rate of local and for partial laryngectomies. The goal has been to eliminate
regional control was significantly higher in the combined- the cancer via targeted surgical technique and apply CRT
therapy group than in the group given radiotherapy alone for the high-risk patients.
(hazard ratio for local or regional recurrence, 0.61; 95 % The advantage of primary surgery has been the addition
confidence interval, 0.41–0.91; P = 0.01). The estimated of pathological data that can help guide adjuvant therapy. For
2-year rate of local and regional control was 82 % in the example, if an oral cancer is removed and the neck dissected
combined-therapy group, as compared with 72 % in the out, we can get valuable data about perineural and lymphatic
radiotherapy group. Disease-free survival was significantly invasion of the primary tumor, extracapsular spread of the
longer in the combined-therapy group than in the radiother- lymph nodes, and the number of lymph nodes. This data can
apy group (hazard ratio for disease or death, 0.78; 95 % con- then help formulate the need or not of adjuvant radiation
fidence interval, 0.61–0.99; P = 0.04), but overall survival therapy with or without chemotherapy. Moreover, per the
was not [13]. At the same time another study from the NCCN guidelines [16], radiation dose can be deintensified in
European Organization for Research and Treatment of the postoperative period. For oropharyngeal cancers the rec-
Cancer Trial 22931 was published. In this study after under- ommended definitive primary radiation dose is 66–74 Gy to
going surgery with curative intent, 167 patients were ran- the primary site and involved neck nodal stations. Uninvolved
domly assigned to receive radiotherapy alone (66 Gy over a nodal stations can receive 44–64 Gy. On the other hand once
period of 6 1/2 weeks) and 167 to receive the same radio- the cancer has been primarily extirpated with neck node dis-
therapy regimen combined with 100 mg of cisplatin per section, the radiation dose to the primary and involved nodes
square meter of body surface area on days 1, 22, and 43 of can be reduced to 60–66 Gy. Interest in deintensification of
the radiotherapy regimen. The rate of progression-free sur- radiotherapy is gaining momentum due to the detrimental
vival was significantly higher in the combined-therapy group impact on swallowing and hence quality of life with increas-
than in the group (47 % versus 36 %) given radiotherapy ing radiation dose. A recent article [17] noted that a dose of
alone (P = 0.04). The overall survival was also higher in com- >50 Gy to the base of the tongue and a mean dose of >51 Gy
bined modality group (53 % versus 40 %) (P = 0.02). The to the superior pharyngeal constrictor, middle pharyngeal
locoregional relapse rate was lower in combined group as constrictor, and inferior pharyngeal constrictor all correctly
well (18 % versus 31 %) (P = 0.007) [14]. In both studies the identified penetration and aspiration of fluids at 6 months
complication rates and toxicity rates were higher in the com- post treatment. Mean dose to the glottic/supraglottic larynx
bined modality group. The high-risk patients were consid- of >48 Gy, maximum dose to the upper esophageal sphincter
ered as those with positive margins and extracapsular spread of >60 Gy, and mean dose to the esophagus of >17 Gy
on lymph nodes only. also correctly identified penetration and aspiration of fluids
The toxicities in both the abovementioned trials ranged at 6 months. Higher doses of radiation resulted in higher
from mucositis, dysphagia, and ultimately gastrostomy tube dysfunction.
dependence [15]. In addition we see the rendering of the lar- Human papillomavirus (HPV) type 16 has been demon-
ynx to be nonfunctional with the patient suffering from aspi- strated as a cause of oropharyngeal cancer that confers a good
ration, voice changes, and ultimately requiring additional prognosis on the patient. In a study comparing accelerated
surgery. Data suggests that complications from salvage laryn- fraction versus standard fraction radiation therapy for stage
geal surgery after CRT have complication rate as high as 3 and 4 oropharyngeal cancer (RTOG 0129), it was demon-
59 % with the most common complication being pharyngo- strated that 3-year overall survival in HPV-positive patients
cutaneous fistula [13]. This morbidity may require additional was 82.4 % versus 57.1 % in HPV-negative patients [18].
surgery including free tissue transplant reconstruction. This study has helped define HPV-positive cancers as a
Over the years many cancers were automatically referred different entity. The new dilemma that head and neck oncol-
for CRT treatment, but better surgical extirpative and recon- ogists now face is whether there can be deintensification of
structive techniques are making many unresectable tumors treatment for these favorable cancers.
resectable with good quality of life and oncological control. Transoral robotic surgery (TORS), transoral laser surgery,
and endoscopic laser surgery have enabled the management
of oral cavity, oropharynx, and larynx surgery via minimally
Evolution of Surgery invasive techniques that allow for faster return to swallowing
without traditional neck incisions. Currently there are studies
Over the years many subsites in the head and neck can be under way to consider the impact of TORS on HPV-positive
safely operated on with minimally invasive techniques. cancers. TORS is considered as a treatment deintensification
The last decade has seen the expansion of endoscopic skull modality for oropharyngeal cancers that allows for targeted
base resection techniques, transoral robotic surgery for therapy of cancer and along with convention neck nodal
42 Head and Neck Tumors: Viewpoint—Surgery 545
surgery may in the future obviate the need for adjuvant CRT They were then re-irradiated with CyberKnife radiosurgery
in appropriately selected patients in the future. (SRS) with three fractions of 10 or 13 Gy. The authors
Endoscopic skull base techniques have similarly enabled changed to a fractionation schedule of five fractions of 5 or
us to remove skull base tumors without the need for cranioto- 8 Gy [31]. Fifteen patients (42.9 %) showed complete
mies. Local vascularized reconstruction techniques that can tumor regression and 13 (37.1 %) partial regression with
be accomplished endoscopically have allowed for excellent median local control rates and overall survival rates after 1
watertight closure of the skull base [19]. year of 61 % and 52.1 %, respectively. Late adverse events
Furthermore, there has been development of accountable were observed in three patients: two grade 4 and one grade
care organizations (ACO) to link provider reimbursements to 5 toxicity.
quality metrics and reduce overall total cost of care. ACOs Unger et al. reported on SRS re-irradiation in 65 head and
aim to place a higher degree of financial responsibility on the neck cancer patients [32]. Initial median RT dose was 67 Gy
providers, with hopes of improving care management and (32–120 Gy) and median treatment interval between initial
limiting unnecessary expenditures [20]. There is also a trend course and SRS 26 months. Treatment sites were oropharynx
towards discussion of “regionalization” of cancer care that (n = 13), hypopharynx (n = 8), nasopharynx (n = 7), paranasal
may result in better outcomes for patients treated in high vol- sinus (n = 7), neck (n = 7), and others (n = 23). Patients were
ume hospitals [21–23]. treated with SRS re-irradiation either definitively (n = 38) or
with palliative intent due to metastatic or untreated local
disease (n = 27). Nine patients had complete macroscopic
Stereotactic Radiosurgery resection before SRS, and 33 patients were treated with
simultaneous chemoradiotherapy. Most patients were treated
Stereotactic radiosurgery (SRS) has not been shown to be of with 5 × 6 Gy (2–5 × 4–12 Gy), but dose and fractionation
benefit as a primary treatment modality for any head and were individualized by the treating physician, and total dose
neck cancer. Additionally therapy options for locoregional was commonly reduced after complete macroscopic resec-
recurrences in previously irradiated head and neck patients tion. For 56 patients evaluable for response, complete, par-
are limited. Salvage surgery if clinically feasible is the stan- tial, and no response were observed in 30 patients (54 %), 15
dard treatment for recurrences [24, 25], but with extensive patients (27 %), and 11 patients (20 %), respectively. Median
locoregional spread, surgery alone is not sufficient. Those overall survival was 12 months. The overall survival (OS)
patients who are considered inoperable often receive pallia- rate was 41 % for definitively treated patients and the locore-
tive chemotherapy with low response rates ranging from 10 gional control rate 30 % in these patients. The authors
to 40 % [26, 27]. Salama et al. [28] reported the long-term reported that nasopharynx, surgical resection, and higher
outcome of concurrent treatment with chemotherapy and re- total dose were significantly correlated with improved
irradiation of patients suffering from recurrent or second pri- locoregional control while nonsquamous histology and sur-
mary squamous cell carcinoma of the head and neck gical resection with improved OS. One patient died of
(SCCHN). One hundred and fifteen patients were treated unspecified causes 2 weeks after re-irradiation, considered
with locoregional control, and free of distant metastasis rate treatment related. Grade 4 late toxicities were observed in six
at 3 years was 51 % and 61 %, respectively. Re-treatment patients (9 %) including soft tissue necrosis, arterial bleed-
was highly toxic with 19 toxic events, five of them due to ing, fistula formation, and severe dysphagia. No association
carotid hemorrhages. The authors came to the conclusion of higher SRS dose or higher cumulative dose with severe
that re-irradiation of recurrent head and neck cancer should late complications was seen, though the outcome was not
be limited to clinical trials. significant due to the small number of cases.
As conventional reradiation therapy has high toxicity, To increase efficacy of SRS re-irradiation, Heron et al.
SRS in the re-irradiation of head and neck cancer patients investigated combinations of SRS with cetuximab [33].
has been pursued in the last few years [29]. Thirty-five patients were treated with SRS re-irradiation
Voynov et al. reviewed 22 patients re-treated with SRS in alone or with additional weekly cetuximab during SRS,
recurrent previously irradiated SCCHN [30]. Using the respectively. Toxicities were comparable in both patient
CyberKnife, patients were irradiated with a median single cohorts, and an overall survival benefit was seen for the
fraction dose of 5 Gy for a median of five fractions. cetuximab arm.
Re-treatment with SRS was considered to be safe as no grade Vargo et al. recently reported on SRS in recurrent, non-
4 or 5 toxicities or late toxicities occurred. Local control rate squamous cell cancers of the head and neck [34]. Thirty-four
was 26 % at 2 years, with a 22 % 2-year survival. In 2009 an patients were re-irradiated with a median SRS dose of 40 Gy
analysis was published by Roh et al. of 36 patients with recur- in five fractions. The toxicity and quality of life were fol-
rent head and neck cancer who had received a full-dose radia- lowed prospectively with a median follow-up of 10 months.
tion of 66–70.2 Gy in the first radiation therapy (RT) course. The authors found promising local control rates of 77 % and
546 S. Jalisi
59 % after 6 months and 12 months, respectively. For tumors 8. The Department of Veterans Affairs Laryngeal Cancer Study
<25 mL they reported a significant improvement of local Group. Induction chemotherapy plus radiation compared with sur-
gery plus radiation in patients with advanced laryngeal cancer. N
control compared to larger tumors (>25 mL). Acute and late Engl J Med. 1991;324:1685–90.
grade 3 toxicities occurred in 15 % and 6 % of the patients, 9. Gourin CG, Johnson JJ. A Contemporary Review of Indications for
respectively, and no grade 4 or 5 toxicities were observed. Primary Surgical Care of Patients With Squamous Cell Carcinoma
In conclusion, SRS in the re-irradiation for head and neck of the Head and Neck. Laryngoscope. 2009;119:2124–34.
10. Lefebvre JL, Chevalier D, Luboinski B, Kirkpatrick A, Collette L,
tumors has demonstrated to be a promising salvage therapy Sahmoud T. Larynx preservation in pyriform sinus cancer: prelimi-
modality with encouraging local control rates and justifiable nary results of European organization for research and treatment of
toxicities. Severe late adverse events (grade 4–5 toxicities) cancer phase III trial. J Natl Cancer Inst. 1996;88:890–9.
have been reported in some studies but are less frequent than 11. Forastiere AA, Goepfert H, Maor M, Pajak TF, Weber R, Morrison
W, Glisson B, Trotti A, Ridge JA, Chao C, Peters G, Lee DJ, Leaf
in patients re-treated with conventional techniques. Very A, Ensley J, Cooper J. Concurrent chemotherapy and radiotherapy
high single fraction doses of 10–13 Gy or higher should be for organ preservation in advanced laryngeal cancer. N Engl J Med.
avoided. There have been no studies comparing SRS to con- 2003;349:2091–8.
ventional IMRT re-irradiation for salvage therapy. 12. Cooper JS, Pajak TF, Forastiere AA, Jacobs J, Campbell BH,
Saxman SB, Kish JA, Kim HE, Cmelak AJ, Rotman M, Machtay
M, Ensley JF, Chao KS, Schultz CJ, Lee N, Lee N, Fu KK. Radiation
Therapy Oncology Group 9501/Intergroup. Postoperative concur-
Conclusion rent radiotherapy and chemotherapy for high-risk squamous-cell
carcinoma of the head and neck. N Engl J Med. 2004;350(19):
1937–44.
Surgery is an integral part of the treatment of head and neck 13. Weber RS, Berkey BA, Forastiere A, Cooper J, Maor M, Goepfert
cancers and is still the gold standard treatment. Over the H, Morrison W, Glisson B, Trotti A, Ridge JA, Chao KS, Peters G,
years several studies have demonstrated the superiority of Lee DJ, Leaf A, Ensley J. Outcome of salvage total laryngectomy
concurrent CRT to radiation therapy alone, but there is a following organ preservation therapy: the Radiation Therapy
Oncology Group Trial 91–11. Arch Otolaryngol Head Neck Surg.
dearth of randomized controlled trials comparing the effi- 2003;129:44–9.
cacy of CRT to surgery. Newer surgical modalities enable us 14. Bernier J, Domenge C, Ozsahin M, Matuszewska K, Lefèbvre JL,
to provide targeted therapies with excellent functional and Greiner RH, Giralt J, Maingon P, Rolland F, Bolla M, Cognetti F,
cosmetic outcomes. There is a growing understanding of the Bourhis J, Kirkpatrick A, van Glabbeke M, European Organization
for Research and Treatment of Cancer Trial 22931. Postoperative
impact of radiation dose on swallowing outcomes, and there irradiation with or without concomitant chemotherapy for locally
is a new focus on treatment deintensification especially in advanced head and neck cancer. N Engl J Med. 2004;350(19):
HPV-positive patients. SRS is currently useful in the setting 1945–52.
of re-irradiation amongst inoperable patients with recurrent 15. Machtay M, Moughan J, Trotti A, Garden AS, Weber RS, Cooper
JS, Forastiere A, Ang KK. Factors associated with severe late toxic-
head and neck cancer. ity after concurrent chemoradiation for locally advanced head and
neck cancer: an RTOG analysis. J Clin Oncol. 2008;26:3582–9.
16. National Clinical Practice Guidelines in Oncology (NCCN
Guidelines). Head and Neck Cancers. Version 1.2012. 2012. Cited
References 2013 Mar 3. http://www.nccn.org/professionals/physician_gls/pdf/
head-and-neck.pdf
1. Davies L, Welch HG. Epidemiology of head and neck cancer in the 17. Cartmill B, Cornwell P, Ward E, Davidson W, Nund R, Bettington
United States. Otolaryngol Head Neck Surg. 2006;135(3):451–7. C, Rahbari RM, Poulsen M, Porceddu S. Emerging understanding
2. Hoffman HT, Karnell LH, Funk GF, Robinson RA, Menck HR. The of dosimetric factors impacting on dysphagia and nutrition follow-
National Cancer Data Base report on cancer of the head and neck. ing radiotherapy for oropharyngeal cancer. Head Neck.
Arch Otolaryngol Head Neck Surg. 1998;124:951–62. 2012;35(8):1211–9. doi:10.1002/hed.23040.
3. Cooper JS, Porter K, Mallin K, Hoffman HT, Weber RS, Ang KK, 18. Ang KK, Harris J, Wheeler R, Weber R, Rosenthal DI, Nguyen-Tân
Gay EG, Langer CJ. National Cancer Database report on cancer of PF, Westra WH, Chung CH, Jordan RC, Lu C, Kim H, Axelrod R,
the head and neck: 10-Year update. Head Neck. 2009;31:748–58. Silverman CC, Redmond KP, Gillison ML. Human papillomavirus
4. Hoffman HT, Porter K, Karnell LH, Cooper JS, Weber RS, Langer and survival with oropharyngeal cancer. N Engl J Med. 2010;
CJ, Ang KK, Gay G, Stewart A, Robinson RA. Laryngeal cancer in 363:24–35.
the United States: changes in demographics, patterns of care, and 19. Jalisi S, O’Gara B, Toshkezi G, Chin L. Local vascularized flap
survival. Laryngoscope. 2006;116 suppl 111:1–13. reconstruction of the skull base-clinical outcomes and analysis.
5. Shah JP, Karnell LH, Hoffman HT, Ariyan S, Brown GS, Fee WE, World Neurosurg. 2012. doi:10.1016/j.wneu.2012.11.073. pii:
Glass AG, Goepfert H, Ossoff RH, Fremgen A. Patterns of care for S1878-8750(12)01382-4.
cancer of the larynx in the United States. Arch Otolaryngol Head 20. Centers for Medicare & Medicaid Services [Internet] [updated
Neck Surg. 1997;123:475–83. 2012 Apr 5, cited 2013 Feb 17] Accountable Care Organizations
6. Chen AY, Schrag N, Hao Y, Flanders WD, Kepner J, Stewart A, (ACO). https://www.cms.gov/Medicare/Medicare-Fee-for-Service-
Ward E. Changes in treatment of advanced laryngeal cancer 1985– Payment/ACO/index.html?redirect=/ACO
2001. Otolaryngol Head Neck Surg. 2006;135:831–7. 21. Weber RS, Lewis CM, Eastman SD, Hanna EY, Akiwumi O, Hessel
7. Chen AY, Schrag N, Hao Y, Stewart A, Ward E. Changes in treat- AC, Lai SY, Kian L, Kupferman ME, Roberts DB. Quality and per-
ment of advanced oropharyngeal cancer, 1985–2001. Laryngoscope. formance indicators in an academic department of head and neck
2007;117:16–21. surgery. Arch Otolaryngol Head Neck Surg. 2010;136:1212–8.
42 Head and Neck Tumors: Viewpoint—Surgery 547
22. Gourin CG, Forastiere AA, Sanguineti G, Koch WM, Marur S, apy and reirradiation for recurrent and second primary head-and-
Bristow RE. Impact of surgeon and hospital volume on short term neck squamous cell carcinoma. Int J Radiat Oncol Biol Phys.
outcomes and cost of laryngeal cancer surgical care. Laryngoscope. 2006;64(2):382–91. doi:10.1016/j.ijrobp.2005.07.005.
2011;121:85–90. 29. Mantel F, Flentje M, Guckenberger M. Stereotactic body radiation
23. Jalisi S, Bearelly S, Abdillahi A, Truong MT. Outcomes in head therapy in the re-irradiation situation–a review. Radiat Oncol.
and neck oncologic surgery in academic medical centers in the 2013;8:7. doi:10.1186/1748-717X-8-7.
United States. Laryngoscope. 2013;123(3):689–98. doi:10.1002/ 30. Voynov G, Heron DE, Burton S, Grandis J, Quinn A, Ferris R,
lary.23835. Ozhasoglu C, Vogel W, Johnson J. Frameless stereotactic radiosur-
24. McLaughlin MP, Parsons JT, Fein DA, Stringer SP, Cassisi NJ, gery for recurrent head and neck carcinoma. Technol Cancer Res
Mendenhall WM, Million RR. Salvage surgery after radiotherapy Treat. 2006;5(5):529–35.
failure in T1-T2 squamous cell carcinoma of the glottic larynx. 31. Roh KW, Jang JS, Kim MS, Sun DI, Kim BS, Jung SL, Kang JH,
Head Neck. 1996;18(3):229–35. doi:10.1002/(SICI)1097- Yoo EJ, Yoon SC, Jang HS, et al. Fractionated stereotactic radio-
0347(199605/06)18:3<229::AID-HED4>3.0.CO;2-1. therapy as reirradiation for locally recurrent head and neck cancer.
25. Wong LY, Wei WI, Lam LK, Yuen AP. Salvage of recurrent head Int J Radiat Oncol Biol Phys. 2009;74(5):1348–55. doi:10.1016/j.
and neck squamous cell carcinoma after primary curative surgery. ijrobp.2008.10.013.
Head Neck. 2003;25(11):953–9. doi:10.1002/hed.10310. 32. Unger KR, Lominska CE, Deeken JF, Davidson BJ, Newkirk KA,
26. Forastiere AA, Metch B, Schuller DE, Ensley JF, Hutchins LF, Gagnon GJ, Hwang J, Slack RS, Noone AM, Harter
Triozzi P, Kish JA, McClure S, VonFeldt E, Williamson SK, et al. KW. Fractionated stereotactic radiosurgery for reirradiation of
Randomized comparison of cisplatin plus fluorouracil and carbo- head-and-neck cancer. Int J Radiat Oncol Biol Phys.
platin plus fluorouracil versus methotrexate in advanced squamous- 2010;77(5):1411–9. doi:10.1016/j.ijrobp.2009.06.070.
cell carcinoma of the head and neck: a Southwest Oncology Group 33. Heron DE, Rwigema JC, Gibson MK, Burton SA, Quinn AE,
study. J Clin Oncol. 1992;10(8):1245–51. Ferris RL. Concurrent cetuximab with stereotactic body radiother-
27. Schrijvers D, Johnson J, Jiminez U, Gore M, Kosmidis P, Szpirglas apy for recurrent squamous cell carcinoma of the head and neck: a
H, Robbins K, Oliveira J, Lewensohn R, Schuller J, et al. Phase III single institution matched case–control study. Am J Clin Oncol.
trial of modulation of cisplatin/fluorouracil chemotherapy by inter- 2011;34(2):165–72.
feron alfa-2b in patients with recurrent or metastatic head and neck 34. Vargo JA, Wegner RE, Heron DE, Ferris RL, Rwigema JC, Quinn
cancer. Head and Neck Interferon Cooperative Study Group. J Clin A, Gigliotti P, Ohr J, Kubicek GJ, Burton S. Stereotactic body radi-
Oncol. 1998;16(3):1054–9. ation therapy for locally recurrent, previously irradiated nonsqua-
28. Salama JK, Vokes EE, Chmura SJ, Milano MT, Kao J, Stenson KM, mous cell cancers of the head and neck. Head Neck.
Witt ME, Haraf DJ. Long-term outcome of concurrent chemother- 2012;34(8):1153–61. doi:10.1002/hed.21889.
Head and Neck Tumors:
Viewpoint—Fractionated Radiation 43
Therapy and Chemotherapy
Michael Rutenberg and Mohan Suntharalingam
risk stratification, and incorporating novel biologic targeted

Introduction therapies into treatment regimens [4–8]. As we continue to
improve disease control and organ preservation, there is
Squamous cell carcinomas of the head and neck (HNSCC) increasing emphasis on reducing treatment-related morbidity
represent approximately 5 % of cancers diagnosed in the and improving patient quality of life.
United States each year. More than 40,000 patients will be
diagnosed this year, and worldwide this number will exceed
500,000 [1]. Although patients presenting with early-stage Locoregional Control and Overall Survival
disease are highly curable with either radiation therapy or
surgery alone, those with locally advanced disease have his- The link between locoregional control and its effect on
torically had poorer prognoses and limited treatment options. overall survival has long been appreciated in head and neck
While most of these cancers remain clinically limited to the cancers. Historically, the overall survival of patients with
head and neck region, locally advanced tumors present sig- locally advanced disease has largely been determined by the
nificant therapeutic challenges. They are often characterized ability to achieve local control. Radiosensitizing systemic
by invasion into surrounding anatomic structures such as chemotherapy has been delivered with definitive doses
muscle, bone, nerves, or blood vessels and may present with of radiotherapy to improve the outcomes for these patients.
clinically detectable lymph node metastases. Surgical resec- The integration of chemotherapy into the overall treatment
tion with adjuvant radiotherapy offers a chance of cure for scheme has helped improve locoregional control and
some patients; however, this can come at a significant cost in increased overall survival rates, while achieving organ pres-
terms of function and cosmesis [2, 3]. Patients with advanced ervation and maintaining organ function. A review of the
head and neck cancers tend to have poor long-term survival concurrent chemotherapy and radiation literature reveals a
and a significant number ultimately die of locoregional dis- consistent improvement in local control associated with com-
ease progression. bined modality therapy for patients presenting with advanced
The management of patients diagnosed with locally disease when compared to radiotherapy alone [6, 9–12]. While
advanced squamous cell carcinomas of the head and neck has these findings are encouraging, a critical review reveals room
undergone a major paradigm shift during the past several for improvement. Efforts to improve locoregional control
decades. The standard of monotherapy using either surgery or with reduced toxicity provide the rationale for ongoing
radiation has been replaced by a multidisciplinary and com- clinical trials.
bined modality approach [4–6]. For many patients with Despite the fact that combination chemotherapy and radi-
locally advanced disease, the use of systemic therapy and ation trials have been conducted for the past 4 decades, the
radiation offers the opportunity for cure while achieving absolute survival benefit associated with combined modality
organ preservation. In recent decades, progress has been made therapy was historically poorly defined. In order to better
in understanding optimal radiotherapy delivery schedules, define the potential advantage of cytotoxic therapy, four
defining effective chemotherapy combinations, improving large meta-analyses were performed comparing chemother-
apy (delivered as neoadjuvant, concurrent, or adjuvant) plus
local therapy vs. local therapy alone [11, 13–15]. The primary
M. Rutenberg • M. Suntharalingam (*) end point of these studies was overall survival.
Department of Radiation Oncology, University of Maryland, These four reports included three studies that were literature
22 South Greene Street, Baltimore, MD 21201, USA
e-mail: msuntha@unm.edu based [11, 13, 14] and one that used individual patient data [15].
550 M. Rutenberg and M. Suntharalingam
The total number of cases included in each analysis ranged of 84 %, 72 %, and 67 %, respectively (p < 0.01). In a recent
from 4,292 (from 28 trials) to as high as 10,741 (63 trials). The update of this trial with 10-year follow-up, these data con-
median follow-up for all patients ranged from 2 to 6.8 years. tinue to show a benefit for concurrent chemoradiotherapy
Importantly, all four meta-analyses reached similar conclusions in locoregional control and larynx preservation, but no differ-
despite significant differences in the databases analyzed. ence in laryngectomy-free survival or overall survival [18].
The results from all of the studies confirmed a small, but In 2004, the French Head and Neck Oncology and
reproducible survival benefit in favor of the addition of che- Radiotherapy Group published the results of GORTEC
motherapy. The magnitude of this benefit ranged from an 94–01, a phase III trial comparing radiotherapy alone to con-
absolute survival advantage of 2.8–6.5 %. The original meta- current chemoradiation (carboplatin and fluorouracil) for
analysis by Pignon et al. was recently updated to include 87 stage III or IV oropharyngeal carcinomas [19]. They showed
trials and 16,485 patients. With updated data, they observed a statistically significant benefit for 5-year disease-specific
a hazard ratio for death of 0.88 and a 5-year absolute survival survival and locoregional control and a strong trend towards
benefit of 4.5 % with the addition of chemotherapy to locore- improvement in overall survival with concomitant chemora-
gional treatment over locoregional treatment alone [12]. diotherapy [19].
Each of these meta-analyses noted the largest contribution of Perhaps no other head and neck subsite has made more
systemic therapy to overall survival came with the concomi- progress with the addition of chemotherapy to radiotherapy
tant use of chemotherapy with locoregional treatment than carcinomas involving the nasopharynx. In 1998,
[12–15]. Al-Sarraf and colleagues published the results of Intergroup
00–99, a phase III randomized trial including patients with
stage III and IV squamous cell carcinomas of the nasophar-
Combined Chemotherapy and Radiation ynx [10]. This trial compared standard radiation (70 Gy in 7
weeks) to concurrent chemoradiation with cisplatin (100 mg/m2)
The VA Larynx trial, originally published in 1991, was a and three cycles of adjuvant cisplatin (80 mg/m2) and fluoro-
landmark study demonstrating equivalent survival between uracil (1,000 mg/m2). The study resulted in early closure due
larynx preserving induction chemotherapy (cisplatin and to demonstration of a survival benefit in the combined
5FU) followed by radiotherapy (66–76 Gy) vs. laryngec- chemotherapy and radiation arm at the first interim analysis.
tomy with postoperative radiotherapy (50 Gy) for laryngeal One hundred forty-seven patients were evaluable for the pri-
cancers [16]. A subsequent EORTC study evaluated patients mary end points of PFS and overall survival and revealed
with hypopharynx cancers and similarly showed effective 3-year PFS of 24 % for radiotherapy alone vs. 69 % in the
organ preservation using chemotherapy followed by radio- chemoradiation arm (p < 0.001). The 3-year overall survival
therapy compared to total laryngectomy and adjuvant rate was 47 % vs. 78 % (p = 0.005) in favor of combined
radiotherapy, with no difference between arms in progres- modality treatment [10].
sion-free survival (PFS) or overall survival [17]. The 5-year This study was among the first individual randomized tri-
functional intact larynx rate in the EORTC study was 35 % als to demonstrate an overall survival benefit associated with
in the chemotherapy and radiation arm. Following these the use of concurrent chemoradiation in any head and neck
trials, the Radiation Therapy Oncology Group (RTOG) con- cancer population. These results helped define the current
ducted a phase III randomized trial evaluating the optimal standard of care for patients with locally advanced squamous
sequencing of chemotherapy and radiation in organ preserv- cell carcinomas of the nasopharynx. Multiple institutional
ing treatment for locally advanced head and neck cancers. trials have now documented local control rates between 80
RTOG 91–11 evaluated sequential chemotherapy and radio- and 90 % for these patients when combination chemotherapy
therapy vs. concurrent chemoradiotherapy vs. radiation alone and modern radiotherapy are utilized [20–22]. Several meta-
in patients with stage III/IV laryngeal cancer [6]. The induc- analyses have confirmed the benefits in local control and
tion chemotherapy arm consisted of cisplatin (100 mg/m2) survival with combined chemotherapy and radiotherapy over
and 5-fluorouracil (1,000 mg/m2) followed by 70 Gy. The radiation alone [23–25]. They also indicate the largest bene-
concurrent arm included cisplatin (100 mg/m2) delivered fits in survival and local control are achieved with concurrent
days 1, 22, and 43 of RT (70 Gy). The radiation-only arm chemotherapy and radiation compared to induction or adju-
received 70 Gy in 2 Gy/fraction. With a primary end point of vant chemotherapy. With the use of concurrent chemoradia-
2-year larynx preservation rate, this trial convincingly dem- tion, long-term follow-up of these patients has shown that a
onstrated a benefit to concurrent chemoradiation compared majority of cancer-related deaths are a result of metastatic
to sequential chemotherapy with radiation or radiation alone, disease without evidence of local recurrence. As a result,
with 2-year locoregional control rates of 80 %, 64 %, and current efforts focus on improvements in reducing metastatic
56 %, respectively (p = 0.004), and larynx preservation rates disease and treatment-related morbidity [26–28].
43 Head and Neck Tumors: Viewpoint—Fractionated Radiation Therapy and Chemotherapy 551
Additional support for the use of concurrent chemother- patients who responded to induction [31]. The TPF arm
apy and radiation as definitive treatment in head and neck showed a benefit over the PF arm in overall response rate
cancers comes from the previously noted meta-analysis by (80 % vs. 59.2 %, p = 0.002) and 3-year larynx preservation
Pignon et al. [12]. Eighty-seven randomized trials including (70.3 % vs. 57.5 %, p = 0.03). No differences between arms
16,485 patients were evaluated and showed a significant were seen in the 3-year disease-free interval or overall survival.
overall survival benefit for patients treated curatively with There is much debate regarding the use of induction
concomitant chemotherapy. Concomitant chemotherapy chemotherapy followed by concurrent chemoradiotherapy
resulted in a hazard ratio of death of 0.81 (95 % CI, 0.78; compared to upfront chemoradiotherapy. Proponents of
0.86) and an absolute survival benefit of 6.5 % at 5 years. No induction chemotherapy point to the high overall response
statistically significant survival benefit was seen in this anal- rates of induction chemotherapy and the benefit of post-
ysis for induction or adjuvant chemotherapy. Importantly, induction chemotherapy patient reevaluation for possible
the survival benefit observed for concomitant chemotherapy risk-based treatment modifications. They also argue that
was seen for all subsites analyzed, including oral cavity, oro- TAX 323, TAX 324, and GORTEC 2000–1 suggest potential
pharynx, larynx, and hypopharynx [12]. survival benefits when compared to concurrent chemoradio-
therapy trials.
Opponents of induction chemotherapy argue that the pub-
Induction Chemotherapy Prior to Definitive lished induction trials fail to compare induction chemother-
Radiotherapy apy to the gold standard of upfront chemoradiation and
instead only show a benefit of one induction regimen over
Recent studies comparing induction chemotherapy regimens another. Additionally, they caution that toxicity associated
followed by radiotherapy have reinvigorated interest in with induction chemotherapy may interfere with or prevent
induction chemotherapy as a viable approach in organ pres- the delivery of chemoradiation as the curative treatment for
ervation therapy. The TAX 323 trial randomized 358 patients these patients. Notably, approximately 30 % of patients in
with unresectable stage III/IV, non-metastatic, SCC of the each arm of TAX 323 and TAX 324 did not complete their
head and neck to induction chemotherapy with either cispla- planned radiotherapy [29, 30]. Critics of induction chemo-
tin and fluorouracil (PF) or docetaxel, cisplatin, and fluoro- therapy note that the promise of reduced distant failures has
uracil (TPF), each followed by radiotherapy to 66–74 Gy not been convincingly shown in the induction chemotherapy
[29]. Median overall survival (18.8 months vs. 14.5 months, trials to date.
p = 0.02) and PFS benefits (11 months vs. 8.2 months, The DeCIDE and PARADIGM trials are two recently
p = 0.007) were seen in the TPF arm compared to the PF completed phase III trials which compare induction
arm. Importantly, most treatment failures were locoregional chemotherapy followed by concurrent chemoradiation with
(>80 % in each arm). upfront chemoradiation. The DeCIDE trial included patients
In a trial similar to TAX 323, TAX 324 randomized 501 with non-metastatic HNSCC with N2/N3 disease, random-
patients with stage III/IV head and neck cancers to three ized to accelerated hyperfractionated radiotherapy with con-
cycles of induction chemotherapy using docetaxel, cisplatin, current docetaxel, fluorouracil, and hydroxyurea or induction
and fluorouracil followed by concurrent chemoradiation chemotherapy with two cycles of TPF followed by the same
with carboplatin to doses of 70–74 Gy or induction cisplatin chemoradiation. The initial report with 3-year results shows
and fluorouracil followed by the same course of concurrent no suggestion of a difference in the primary end point
chemoradiation [30]. A major difference between this study of overall survival (73 % vs. 75 %, p = 0.70) [32]. The
and TAX 323 was the addition of concurrent weekly carbo- PARADIGM trial compared patients with locally advanced
platin (AUC 1.5) with radiotherapy. Median overall survival head and neck cancers receiving radiotherapy with concur-
was significantly improved in the TPF arm compared to the rent cisplatin vs. three cycles of induction TPF followed by
PF arm, 71 months and 30 months (p = 0.006), respectively. concurrent radiotherapy with carboplatin or docetaxel [33].
PFS was 36 months in the TPF arm compared to 13 months Because this study closed early due to slow accrual, it has
in the PF arm (p = 0.004). Similar to TAX 323, over 80 % of inadequate power to appropriately evaluate the primary end
treatment failures were locoregional [30]. point of survival. The data as reported fail to show a survival
GORTEC 2000–1 was a French phase III study compar- benefit to induction chemotherapy. With a median follow-up
ing larynx preservation in patients with locally advanced of 49 months, the 3-year overall survival and PFS were 73 %
squamous cell carcinoma of the larynx or hypopharynx vs. 78 % (p = 0.77) and 67 % and 73 % (p = 0.55) for the
randomized to either induction TPF or PF followed by radio- induction chemotherapy arm and chemoradiation arms,
therapy with or without concomitant chemotherapy for respectively [33].
to concurrent radiotherapy with cisplatin or cetuximab. With

Targeted Therapies as Part of a Combined the early successes of cetuximab, other EGFR targeting
Modality Strategy agents are under investigation for treatment of head and neck
cancers, including EGFR small molecule inhibitors (e.g.,
The limitations of most systemic cytotoxic agents are often Gefitinib, Erlotinib).
defined by the nonspecific toxicities of normal tissues. Vascular endothelial growth factor-A (VEGF) is a secreted
To improve the therapeutic ratio, investigators have focused protein that is active in angiogenesis, largely mediated
efforts on biologic agents targeting cellular proteins. through binding of cell surface receptor tyrosine kinases.
Epidermal growth factor receptor, a member of the ErbB Increased expression of VEGF has been associated with
family of receptor tyrosine kinases, is one of four receptors worse outcomes in head and neck squamous cell carcinoma
important in cellular proliferation, differentiation, and sur- and overexpression has been observed in high numbers of
vival [17]. The EGFR pathway controls cell-cycle events that nasopharyngeal cancers. Bevacizumab is a monoclonal anti-
directly affect survival. It is estimated that 80–90 % of head body that targets VEGF and blocks its signaling. Lee et al.
and neck carcinomas overexpress EGFR and its ligand recently reported the results of RTOG 0615, a phase II study
TGF-α [alpha] [34]. Its expression has been associated with combining radiotherapy with concurrent and adjuvant cispl-
a poor outcome in HNSCC [34]. Anti-EGFR therapy pro- atin and bevacizumab in nasopharyngeal cancer [38]. Early
vides a powerful target for future combined modality treat- results are promising for safety and efficacy.
ment strategies. As our understanding of tumorigenesis and tumor biology
Cetuximab is an anti-EGFR receptor that targets the continues to improve, new cellular proteins and pathways
extracellular ligand binding domain. In 2006, it became will become targets of focused therapy. Small molecule
the first FDA-approved EGFR targeting agent for locally inhibitors and antibodies will undoubtedly assume increas-
advanced head and neck squamous cell carcinoma. This fol- ing roles in the multimodality treatment of head and neck
lowed from a phase III, randomized trial evaluating the addi- malignancies, with expected benefits in increased response
tion of cetuximab to radiation for patients with squamous rates and decreased treatment morbidity.
cell carcinomas of the head and neck. Bonner et al. reported
a significant improvement in locoregional control (hazard
ratio 0.68, p = 0.005), progression-free survival (3-year PFS Human Papillomavirus in Head and Neck
42 % vs. 31 %, p = 0.04), and overall survival (median sur- Squamous Cell Carcinoma: Treatment
vival 49 months vs. 29.3 months, p = 0.03) with the addition Implications
of cetuximab [7]. Notably, these improvements were accom-
plished without a significant increase in any grade 3 or greater In the late 1990s, a link between human papillomavirus
acute reactions (other than acneform rash and infusion reac- (HPV) and the development of a subset of squamous cell
tions). The recent update of this trial showed a continued carcinomas in the head and neck began to emerge [39, 40].
benefit with the addition of cetuximab to radiotherapy, with HPV is a sexually transmitted DNA virus with hundreds of
5-year overall survival of 45.6 % in the cetuximab arm com- subtypes. Several subtypes (e.g., HPV6 and -11) are known
pared to 36.4 % in the radiotherapy-only arm (p = 0.018) [35]. to cause benign papillomas or warts, while other oncogenic
Several phase III trials are currently investigating subtypes (HPV16, -18, -31, -33), already implicated in the
additional roles of cetuximab in head and neck cancers. An development of cervical cancer, have been implicated in the
overall survival benefit with the addition of cetuximab to development of HNSCC [39, 41, 42]. Gillison et al. at Johns
platinum-based chemotherapy has already been shown for Hopkins found 25 % of HNSCC contained HPV DNA, with
recurrent or metastatic head and neck cancers [36]. Following high-risk oncogenic subtype HPV16 accounting for 90 % of
from these data, combination systemic therapy and radio- the infections [41]. This study found an association between
therapy are being evaluated. RTOG 05–22 is a recently com- HPV-infected tumors, oropharyngeal location, and poorly
pleted phase III trial of accelerated radiotherapy with differentiated cancers. HPV-positive oropharyngeal cancers
concomitant cisplatin with or without cetuximab for stage were less likely among moderate to heavy drinkers and
III/IV HNSCC. The initial results of this study have been smokers, had lower rates of TP53 mutations, and had
presented and show no benefit to the addition of cetuximab improved disease-specific survival compared to HPV-
to cisplatin and radiotherapy for disease-free or overall negative cancers. When patients with HPV-positive tumors
survival [37]. RTOG 09–20 is an ongoing phase III trial ran- (from all head and neck subsites) were evaluated, a 59 %
domizing patients with locally advanced head and neck reduction in risk of death compared to HPV-negative patients
cancers to postoperative radiotherapy with or without con- was seen. A subsequent case-control study was performed
current cetuximab. RTOG 10–16 is a phase III study random- with 100 patients with newly diagnosed oropharyngeal
izing patients with locally advanced HPV-positive HNSCC squamous cell carcinoma compared to 200 control patients
without cancer [42]. This study showed a strong association HNSCC, studies evaluating treatment de-intensification in an
between HPV16-positive SCC of the oropharynx and high- effort to reduce treatment morbidity are underway. Conversely,
risk sexual activities, including increasing numbers of new approaches to improve the relatively poor outcomes of
lifetime vaginal-sex and oral-sex partners, lack of condom HPV-negative HNSCC are ongoing.
use, and early age at first intercourse. The authors showed
an increased risk of HPV-related oropharyngeal cancer in
patients without a history of alcohol or tobacco use and Intensity-Modulated Radiation Therapy
found no evidence of an additive effect between HPV infec-
tion and alcohol/tobacco in the development of oropharyn- The treatment of head and neck cancers presents particular
geal cancer. challenges in radiation delivery due to the high concentration
Several studies evaluating outcomes by HPV status have of critical, radiosensitive tissues in and around the treatment
now confirmed a favorable prognosis for HPV-related fields, including the spinal cord, major and minor salivary
HNSCC treated with chemotherapy and radiation compared glands, lacrimal glands, retina, optic nerves, and the mandi-
to HPV-negative tumors. Fakhry et al. prospectively evalu- ble. The long-term morbidity of tumoricidal doses of radia-
ated patients enrolled in ECOG 2399, a phase II trial using tion delivered to this region can be significant. Despite its
induction chemotherapy followed by chemoradiation in development for treatment of prostate cancer in the mid-
patients with stage III/IV laryngeal or oropharyngeal cancers 1990s, the benefits of intensity-modulated radiation therapy
[43]. Their data showed improved overall response rates for (IMRT) were quickly appreciated in the field of head and
induction chemotherapy (82 % vs. 55 %, p = .01) and chemo- neck radiation oncology, which has become one of its main
radiation (84 % vs. 57 %, p = .007) in HPV-positive tumors beneficiaries and proponents. The benefits of IMRT in head
and improved 2-year survival for HPV-positive tumors com- and neck radiotherapy include the potential for improved
pared to HPV-negative (95 % vs. 62 %, p = 0.005). After tumor control with delivery of high dose radiation, while
adjusting for known risk factors, patients with HPV-positive reducing dose to nearby normal tissues with sharp dose fall-
cancers had lower risks for progression and death from any off. An expected benefit of IMRT for head and neck cancers
cause compared to HPV-negative patients. is improved salivary gland sparing with decreased xerosto-
RTOG 0129 is a phase III trial comparing concurrent mia and improved quality of life.
chemoradiation with either standard-fractionated radiation Early single-institution retrospective studies using IMRT
or accelerated-fractionated radiotherapy in stage III/IV in head and neck cancers showed good tumor control rates
HNSCC [8]. In a post hoc subset analysis including more and an ability to improve normal tissue sparing [20, 21, 45–
than 300 oropharynx cancer patients stratified by HPV sta- 47]. The RTOG conducted a phase II, multi-institutional trial
tus, Ang et al. showed improved 3-year overall survival evaluating hypofractionated IMRT for early-stage oropha-
(82.4 % vs. 57.1 %), PFS (73.7 % vs. 43.4 %), and local- ryngeal cancer. RTOG 02–22 included T1–2, N0–1, M0 SCC
regional control (86.4 % vs. 64.9 %, p < 0.001) for patients in of the oropharynx treated to 66 Gy to gross disease and
the HPV-positive subgroup. They found HPV status to be the 54 Gy to at-risk areas using a dose-painting technique [48].
most important determinant for overall survival. This study With central quality assurance for treatment planning evalu-
also established p16 upregulation in tumor cells as a clinical ation, this study helped guide the use of IMRT for head and
surrogate for HPV-related tumors. P16 is an inhibitor of neck cancers into routine use. The disease control rates and
cyclin-dependent kinases (CDK) and is responsible for slow- xerostomia rates compared favorably with historical coop-
ing cell-cycle progression via inhibition of CDK phosphory- erative group studies. Importantly, RTOG 02–22 showed a
lation of pRB [8]. HPV-oncoprotein E7 activity induces p16 significant association between major protocol underdosing
expression via pRB inactivation. RTOG 01–29 demonstrated violations and local disease recurrence [48].
the correlation between the presence of HPV DNA and p16 RTOG 02–25 was a multi-institutional phase II pros-
expression in tumor cells. Stratification using p16 expression pective trial evaluating the use of intensity-modulated radio-
was found to be consistent with HPV-based stratification for therapy with or without chemotherapy for all stages of
overall survival and disease-free survival [8]. Because of its nasopharyngeal cancer [26]. The results showed 2-year PFS
high sensitivity, ease of analysis, wide availability, and and overall survival rates of 72.7 % and 80.2 %, respectively.
strong correlation with both transcriptionally active HPV Using IMRT, the rate of grade 2 xerostomia was 14 % and
infection and patient outcomes, p16 is often used in lieu of over one-third of patients reported no xerostomia.
direct HPV evaluation in HNSCC [44]. Several randomized studies comparing IMRT and 2D or
HPV status is now a major focus in research and the clinical 3D radiotherapy for HNSCC with the aim of reducing xero-
management of HNSCC. Stratification schemes in clinical tri- stomia have been conducted [49, 50]. The results so far have
als have incorporated this important prognostic factor. Because been encouraging, with improved salivary flow rates and
of the favorable prognosis in patients with HPV-positive reduced observer reported xerostomia using IMRT.
A meta-analysis with individual patient data from 15

Altered Fractionation Schedules randomized trials including over 6,500 patients compared
conventionally fractionated radiotherapy with altered frac-
Modifications in radiotherapy fractionation schedules allow tionation (hyperfractionated and accelerated fractionation)
clinicians to take advantage of fundamental differences in to determine if altered fractionation improved overall sur-
the biology between tumor cells and normal tissue, specifi- vival. The combined altered fractionation group showed a
cally differences in growth rates and genomic repair mecha- 5-year absolute survival benefit of 3.4 % compared to stan-
nisms. The ultimate goal of altered fractionation is to increase dard fractionation (39.7 % vs. 36.3 %, p = 0.003). When indi-
tumor control while decreasing chronic toxicity. The basis vidual altered fractionation schedules were compared to
for altered fractionation lies in the dissociation between standard fractionation, hyperfractionation showed the largest
acute and late effects, where tumor control and acute toxicity survival benefit (8.2 % at 5 years) with accelerated fraction-
are increased with decreasing treatment time and increasing ation showing a smaller, but significant benefit (2 % at
total dose, and less related to fraction size. On the other hand, 5 years) [58].
late effects are related to fraction size and total dose and less With clear local control benefits of altered fractionation
dependent on overall treatment time [51]. radiotherapy over conventionally fractionated RT alone, the
Hyperfractionated radiotherapy is characterized by an opportunity for further improvements with the addition of
increase in the total dose and the number of fractions, a concurrent chemotherapy was evaluated in several trials.
decrease in the dose per fraction, and maintenance of the Brizel and colleagues reported the results of a single-
overall treatment time [51–53]. In comparison, accelerated institution randomized trial comparing hyperfractionated
fractionation reduces overall treatment time, while the num- radiotherapy delivered at 1.25 Gy twice daily to a total dose
ber of fractions, total dose, and fraction size remains rela- of 75 Gy vs. 70 Gy plus the addition of concurrent cisplatin
tively unchanged. In either of these altered fractionation (12 mg/m2/day) and 5-fluorouracil followed by two addi-
schedules, acute toxicity is expected to increase as a result of tional cycles of chemotherapy at the completion of radio-
increased total dose (hyperfractionation) or decreased over- therapy [9]. One hundred twenty-two patients with locally
all treatment time (accelerated fractionation). advanced HNSCC were randomized. The complete response
Initially conceived prior to the era of combined chemo- rate was 73 % for the RT-alone group vs. 88 % for those
therapy and radiation, altered fractionation has been shown receiving CT + RT (p = 0.52). With a median follow-up of 41
in several large multi-institutional randomized trials to months, locoregional control was 44 % in the radiotherapy
improve local control in HNSCC when compared to conven- arm vs. 70 % with combined modality treatment (p = 0.01).
tionally fractionated RT alone (DAHANCA 6&7, RTOG However, the improvement in local control did not translate
90–03, EORTC 22791) [54–56]. RTOG 90–03 randomized into a statistically significant overall survival advantage (RT
1,113 patients with locally advanced HNSCC to one of the alone 34 % vs. CT + RT 55 %, p = 0.07) [9].
following four arms: (1) Standard-fractionated radiotherapy Jeremic et al. conducted a randomized trial comparing
(70 Gy in 2 Gy/fraction), (2) Hyperfractionated radiotherapy hyperfractionated radiotherapy alone (77 Gy in 1.1 Gy BID)
(81.6 Gy in 1.2 Gy/fraction BID), (3) Accelerated split- vs. the same radiotherapy delivered in conjunction with
course fractionation (67.2 Gy in 1.6 Gy/fraction BID with a daily, low-dose cisplatin (6 mg/m2/day) in 130 patients with
2-week break after 38.4 Gy), or (4) Accelerated fractionation non-metastatic, stage III/IV HNSCC [59]. With a median
with concomitant boost (72 Gy in 1.8 Gy/fraction qdaily to follow-up of 79 months, statistically significant improve-
32.4 Gy then BID with 1.8 Gy/fx qAM and 1.5 Gy/fx qPM ments were reported in the combined modality arm for all
for the final 12 days). A recent update with a median follow- end points evaluated. The complete response rates were sig-
up time of 14 years for surviving patients showed 5-year nificantly higher for the combined modality group (75 %) vs.
local control rates significantly improved in the hyperfrac- those receiving RT alone (45 %) (p = 0.002). Five-year overall
tionation arm compared to standard fractionation. At the ini- survival was 46 % for CT + RT vs. 25 % for the RT arm
tial publication with 2-year results, accelerated fractionation (p = 0.041). The locoregional progression-free survival rates
with concomitant boost also showed a significant improve- were 50 % and 36 % (p = 0.041), respectively, while distant
ment in local control over standard fractionation; however, metastasis-free survival was 86 % and 57 % (p = 0.0013).
this was lost with longer follow-up [55]. The update also Notably, radiation-induced toxicities were similar in both
showed a significant disease-free survival benefit to hyper- groups (including grade III/IV xerostomia and subcutaneous
fractionation (p = 0.046) and a strong trend towards improved fibrosis), whereas high-grade hematologic toxicities were
disease-free survival in the accelerated fractionation with more commonly seen in the combined modality group [59].
concomitant boost arm (p = 0.052) and the accelerated split- RTOG 0129 was a multi-institutional prospective random-
course arm (p = 0.056). No statistically significant differ- ized trial evaluating concurrent chemoradiotherapy using either
ences were seen in severe late toxicities at 5 years [57]. accelerated-fractionation radiotherapy or standard-fractionation
radiotherapy in 721 patients with locally advanced HNSCC potential of delivering second courses of radiotherapy
[8]. Accelerated fractionation radiotherapy consisted of 72 Gy with concurrent chemotherapy (Table 43.2). Reirradiation
in 42 fractions with a concomitant boost delivered using twice- with concurrent chemotherapy for locoregionally recurrent
daily radiotherapy for the final 12 days of treatment (1.8 Gy/fx HNSCC has been evaluated as definitive treatment and for
qAM and 1.5 Gy qPM). Chemotherapy consisted of cisplatin postoperative therapy.
(100 mg/m2) q3weeks during radiotherapy. With a median Spencer and colleagues from the University of Alabama
follow-up of 4.8 years, there were no significant differences at Birmingham published the results of their phase I/II trial
between arms in severe acute toxicity, PFS, or overall of reirradiation with concurrent chemotherapy in unresect-
survival. able locoregionally recurrent HNSCC [64]. They showed the
GORTEC 99–02 is a three-arm multi-institutional phase ability to safely deliver 60 Gy (delivered at 1.5 Gy BID) to
III trial evaluating altered fractionation in non-metastatic previously irradiated tissues with concurrent 5-fluorouracil
stage III/IV HNSCC [60]. Eight hundred forty patients were and hydroxyurea. All therapy was delivered on a week-on,
randomized to one of the following arms: (1) conventional week-off schedule. The response rate of 74 % was encourag-
chemoradiotherapy (70 Gy in 2 Gy/fraction with concurrent ing and median survival and 2-year overall survival was
carboplatin/fluorouracil), (2) accelerated radiotherapy- 10.5 months and 20 %, respectively. Importantly, this was
chemotherapy (70 Gy in 2.3 Gy/fraction with concurrent car- one of the first series to accurately define the risks associated
boplatin/fluorouracil), or (3) very accelerated radiotherapy with reirradiation, with 4 of 35 patients developing severe
alone (64.8 Gy in 1.8 Gy/fraction twice daily). With a median late toxicities (two radiation myelitis, two esophageal stric-
follow-up of 5.2 years, no difference in 3-year PFS was tures) [64]. Their results led to a multi-institutional phase II
seen for accelerated radiotherapy-chemotherapy compared trial through the RTOG. RTOG 96–10 enrolled 86 patients
to either conventional chemoradiotherapy or very acceler- with unresectable recurrent or second primaries within a pre-
ated radiotherapy. There was no benefit to the accelerated viously irradiated field. Patients received concurrent chemo-
radiotherapy-chemotherapy arm for 3-year locoregional con- radiation to 60 Gy in 1.5 Gy BID with concurrent 5-FU and
trol compared to conventional chemoradiotherapy. However, hydroxyurea delivered on a week-on, week-off schedule
the very accelerated radiotherapy arm had significantly [65]. The final report of RTOG 96–10 showed 2- and 5-year
worse 3-year locoregional control and PFS compared to con- survival of 15.2 % and 3.8 %, respectively. Acute grade 4
ventional chemoradiotherapy (50.1 % vs. 58.3 %, p = 0.045, and 5 toxicities occurred in 17.7 % and 7.6 %, respectively.
and 32.2 % vs. 37.6 %, p = 0.041, respectively). Very accelerated Late grade 4 and 5 toxicities occurred in 19.4 % and 3.0 %,
radiotherapy also had significantly worse grade 3–4 toxicity respectively [66]. RTOG 99–11, a multi-institutional phase
compared to both chemoradiation arms [60]. II trial, further evaluated the safety and efficacy of concur-
With the results of RTOG 0129 and GORTEC 99–02 rent chemotherapy and reirradiation (see Table 43.2). This
showing no benefits for altered-fractionation radiotherapy study treated patients with recurrent HNSCC after prior
when combined with chemotherapy, the enthusiasm for this radiotherapy with chemoradiotherapy with cisplatin and
approach has dwindled. Concurrent chemotherapy with stan- paclitaxel using accelerated radiotherapy to a total dose of
dard fraction radiotherapy remains the standard of care for 60 Gy in 1.5 Gy BID (4–6 h between fractions). They docu-
organ preserving treatment for locally advanced HNSCC. mented 2-year progression-free and overall survival rates of
Concurrent chemotherapy and external beam radiother- 16 % and 26 %, respectively [61].
apy have had a dramatic impact in improving local control Reirradiation with concurrent chemotherapy has also been
and in many circumstances survival for patients with locally evaluated by multiple investigators in the postoperative setting
advanced squamous cell carcinomas of the head and neck. for in-field recurrence or second primary HNSCC [67–70].
Table 43.1 details the outcomes of multiple phase III trials A prospective, randomized trial showed significantly improved
that have documented these benefits over radiotherapy alone. locoregional control and disease-free survival for patients
treated with postoperative chemotherapy and reirradiation vs.
observation [67]. Long-term survival is achievable in this
Reirradiation for Locoregional Recurrence patient population; however, treatment toxicity rates in these
studies were not trivial.
The prognosis for patients with locally recurrent HNSCC is Reirradiation of head and neck malignancies requires
dismal, with 50–60 % of patients dying as a direct con- special consideration for treatment planning and delivery in
sequence of local failure and disease progression [61]. order to maximize tumor killing and minimize treatment-
Historically, these patients were managed with systemic ther- related morbidity. Tumor location, extent of recurrence, prior
apy alone, with documented overall response rates between radiation details, and time interval between radiation courses
20 and 40 % and median survivals of 5–8 months [62, 63]. are important factors. Because local normal tissues have
Against this backdrop, investigators began examining the received prior high dose radiotherapy (often approaching
556
Table 43.1 Local control and survival with concurrent chemotherapy and radiation
Author Site Total dose Dose per fraction Chemo LC/*LRC DFS/PFS OS
Brizel et al. [1] OC, OP, HYP, LAR, PNS 70 Gy w/chemo 1.25 Gy BID Cisplatin, 5-FU weeks 1,6 *3 years 70 % 5 years 61 % 3 years 55 %
75 GyGy w/o chemo (2 chemo cycles post-CRT)
Jeremic et al. [2] OC, OP, NP, HYP, LAR 77 Gy 1.1 Gy BID Daily bolus cisplatin *5 years 50 % 5 years 46 % 5 years 46 %
Al-Sarraf et al. [3] OC, OP, HYP, NP, 66–73.8 Gy 1.8–2 Gy daily Cisplatin q3weeks 4 years 43 % 4 years 34 %
mSinus, LAR
Aldelstein et al. [4] OC, OP, LAR, HYP 66–72 Gy 1.8–2 Gy daily Cisplatin, 5-FU q3weeks 5 years 77 % 5 years 50 %
Wendt et al. [5] OC, OP, HYP, LAR 70.2 Gy 1.8 Gy BID (with 2 breaks) Cisplatin, 5-FU, leucovorin *3 years 36 % 3 years 48 %
q3weeks
Forastiere et al. (RTOG LAR 70 Gy 2 Gy daily Cisplatin q3weeks 5 years 71 % 5 years 38 % 5 years 55 %
91–11) [6, 7] 10 years 69 % 10 years 22 % 10 years 28 %
Denis et al. (GORTEC) [8] OP 70 Gy 2 Gy daily Carboplatin, 5-FU q3weeks 5 years 48 % 5 years 27 % 5 years 22 %
5-FU 5-fluorouracil, OC oral cavity, OP oropharynx, HYP hypopharynx, LAR larynx, PNS paranasal sinus, mSinus maxillary sinus, BID twice daily, CRT chemoradiotherapy, LC local control,
LRC locoregional control, DFS disease-free survival, PFS progression-free survival, OS overall survival
M. Rutenberg and M. Suntharalingam
Table 43.2 Definitive reirradiation with concurrent chemotherapy for locoregional recurrence
Author Total dose Dose per fraction Chemo Tx delivery Grade 4/5 toxicity LC DFS/PFS OS
Haraf et al. (U. Chicago) [9] Median 50 Gy 1.8–2 Gy qdaily HU, 5-FU CRT given on 11 % Grade 5 5 years 20 % 5 years 14 % 5 years 15 %
or 1.5 Gy BID alternate weeks
Spencer et al. 40–60 Gy 2 Gy daily or HU, 5-FU CRT given on none 2 years 20 %
(UAB) [10] 1.2 Gy BID alternate weeks
RTOG 96–10 [11, 12] 60 Gy 1.5 Gy BID HU, 5-FU CRT given on 18 % Grade 4 1 year 62 % 2 years 15 %
alternate weeks 8 % Grade 5 5 years 4 %
RTOG 99–11 [13] 60 Gy 1.5 Gy BID Cisplatin, paclitaxel CRT given on 23 % Grade 4 (acute) 2 years 16 % 2 years 26 %
alternate weeks 7 % Grade 5
DeCrevoisier et al. (Gustave- 60–65 Gy 2 Gy qdaily or RT alone or various Variable 25 % Grade 4 2 years 11 % 2 years 21 %
Roussy) [14] 1.5 Gy BID chemo 3 % Grade 5 5 years 9 %
Sulman et al. (MDACC) [15] 60–70 Gy 2 Gy qdaily 66 % with cisplatin- Variable 32 % Grade 4 2 years 58 % 2 years 54 %
based chemo 0 % Grade 5
43 Head and Neck Tumors: Viewpoint—Fractionated Radiation Therapy and Chemotherapy
5-FU 5-fluorouracil, BID twice daily, HU hydroxyurea, CRT chemoradiotherapy

557
maximum tolerable doses), there is considerable risk for Recurrent nasopharyngeal squamous cell carcinoma became
severe morbidity and possible mortality. Furthermore, the first HNSCC managed with SRS. Early studies using
because these cancers have recurred after a prior full course either Gamma Knife or Linac-based SRS reirradiation for
of radiotherapy, the recurrent tumor cells may be more recurrent disease showed good response and local control
radioresistant, requiring treatment intensification. Altered rates with acceptable toxicity [73, 74, 81, 82]. These experi-
fractionation schemes are often employed to lower the risk of ences have now been expanded, with many institutions repor-
severe late toxicity while allowing treatment intensification. ting outcomes for salvage stereotactic radiosurgery for
A related consideration is the time interval between reirra- recurrent nasopharyngeal cancer. With the increased availabil-
diation and the initial radiotherapy. With a longer interval, ity of SBRT, this approach has become widely used for naso-
normal tissues have increased capacity for recovery from the pharyngeal recurrences and is being applied more broadly for
initial radiotherapy course. Most studies have required an other sites of inoperable recurrent HNSCC [83–85].
interval of at least 6 months from the completion of the initial Subsites of the head and neck are attractive for the use of
radiotherapy to the start of reirradiation [61, 65, 67, 68]. SRS/SBRT for reirradiation for several reasons. The dense
This time interval is also prognostic for survival [71]. concentration and radiosensitivity of tissues and organs and
Another important consideration for head and neck reirradia- their impact on patient quality of life make precision treat-
tion is target volume delineation. With an increased risk of ment and minimization of treatment of normal tissue impera-
treatment-related morbidity, there is an emphasis on reduc- tive. Additionally, this area has little or no respiratory-related
ing treatment volumes and minimizing elective target treat- movement, minimizing motion-based uncertainties (although
ment. Generally, planning treatment volumes have included swallowing related motion may require consideration).
gross disease with a limited expansion or only high-risk Additionally, the shortened course of SRS/SBRT is ideal for
areas for postoperative reirradiation [65, 67, 68, 72]. Because recurrent head and neck cancer patients who are frequently
of the emphasis on treatment volume reduction, stereotactic debilitated with poor performance status, precluding exten-
treatment can be particularly well suited for head and ded treatment courses and systemic therapy.
neck reirradiation. The accurate target delineation, effective Ryu et al. published a series from Henry Ford Hospital
patient immobilization, and rapid dose falloff that character- including 13 patients with recurrent head and neck carcino-
ize stereotactic treatment enable tumor treatment with mini- mas (62 % squamous cell carcinoma) treated with either
mal normal tissue coverage. Indeed, stereotactic approaches single-dose or fractionated stereotactic radiotherapy (FSRT)
for head and neck reirradiation have historically treated only to various head and neck sites [83]. SRS doses ranged from
areas of gross tumor volume [73, 74]. Recent reports using 12–18 Gy (five patients) and FSRT was treated with
stereotactic body radiotherapy (SBRT) for reirradiation have 30–36 Gy in five or six fractions delivered twice weekly
used variable expansions from gross tumor, ranging from 0 (eight patients). Dose was prescribed to the 90 % isodose
to 10 mm [75, 76]. Studies are currently underway to define line covering the entire gross tumor volume. Objective
optimal target coverage when utilizing stereotactic reirradia- responses were seen in 70 % of patients (46 % complete
tion for HNSCC [77]. Another feature of stereotactic treat- response). A follow-up report from the same group included
ment conducive to head and neck reirradiation is the large 21 patients treated for recurrent head and neck malignancies
fraction sizes, which provides a means of treatment intensifi- using SRS (13–18 Gy) or SBRT (36–48 Gy in 5–8 fractions)
cation to effectively treat recurrent cancer clonogens that are [84]. This study showed an overall response rate of 69 %
radioresistant after a prior course of standard fraction (31 % complete response). All treated patients treated in the
radiotherapy. oropharynx or larynx experienced RTOG grades 1 and 2
mucositis and esophagitis. Five percent (one patient) and
19 % of patients treated with reirradiation experienced
Stereotactic Therapy in Head and Neck RTOG grade 3 (swallowing difficulty) and grade 4 toxicities
Cancers (1 dysphagia and ulceration, 2 mucocutaneous fistulas, 1
mucocutaneous fistula and mandibular radionecrosis),
The use of stereotactic radiotherapy for head and neck cancers respectively. One-year tumor control was 60.6 % with a
originated with the treatment of base of skull and nasopha- median survival of 6.7 months [84].
ryngeal cancers [73, 78–81]. In the era preceding combined Roh and colleagues published their experience using
chemoradiation, the local failure rate for nasopharyngeal CyberKnife-based SBRT on 36 patients with recurrent head
squamous cell carcinomas approximated 30 %. With consid- and neck cancers with prior irradiation [85]. Median doses of
erable morbidity associated with progression in this area, 30 Gy (range 18–40 Gy) in 3–5 fractions were delivered to
attempts at local control for nasopharyngeal SCC included the 65–85 % isodose line on consecutive days. Their plan-
reirradiation with external beam radiotherapy, brachytherapy, ning target volumes (PTV) consisted of gross tumor volume
stereotactic radiosurgery (SRS), and combinations thereof. (GTV) with a 2–3 mm expansion. With a median follow-up
Table 43.3 SRS/SBRT reirradiation for head and neck cancers

Median dose Number of Dose per Fractionation Grade 3+ Grade 5 Median PFS Median OS
Author, institution Pt. No. (range) fractions fraction (Gy) schedule toxicity toxicity (months) (months)
Cengiz et al. [16] 46 30 Gy (18–35) 1–5 5–18 qdaily 30 % 16 % 10.5 12
Kodani et al. [17] 21 30 Gy (19.2–45) 3–7 qdaily 29 % 10 %
Roh et al. [18] 36 30 Gy (18–40) 3–5 5–13 qdaily 44 % 3% 16.2
Unger et al. [19]a 65 30 Gy (21–35) 2–5 4–12 qdaily 11 %c 1.5 % 5.7 12
Siddiqui et al. [20]b 44 1–8 6–18 QOD 16 % 0 6.7e
Vargo et al. [21]d 34 40 Gy (17–50) 1–5 6–20 QOD 21 % (G3 only) 0 11
Comet et al. [22] 40 36 Gy 6 6 QOD 10 % (G3 only) 0 8.8 13.6
5-FU 5-fluorouracil, QOD every other day
a
Includes metastatic and postoperative patients
b
Includes patients with primary and metastatic disease
c
Includes only grade 4/5
e
Recurrent, non-metastatic disease only
d
Includes patients with recurrent disease with metastases
time of 17.3 months, the overall response rate was 80 % this subject is provided in the preceding main chapter.
(42.9 % complete response) and local recurrence-free sur- As stereotactic treatment for reirradiation of recurrent head
vival was 52.1 % and 30.9 % at 1 and 2 years, respectively. and neck cancers becomes more widespread, it is important
Thirty-six percent (13 patients) experienced RTOG grade 3 to rigorously adhere to the basic principles that guide all
acute toxicity (mucositis, nausea, dermatitis), 8 % experi- SRS/SBRT treatments, including effective patient positioning/
enced RTOG grade 4 late toxicity (one bone necrosis, two immobilization, high quality imaging for disease evaluation
soft tissue necrosis), and one patient died of treatment- and treatment planning, highly conformal treatment plan-
related causes with soft-tissue and skull base necrosis [85]. ning, and accurate target localization. The same features of
Unger et al. published a series of 56 patients with unre- head and neck malignancies that make it attractive for SRS/
sectable, recurrent HNSCC treated at Georgetown University SBRT can be sources of increased patient morbidity and
with SBRT reirradiation [75]. Planning target volumes mortality.
included gross disease with a 2–10 mm expansion and stan-
dard dosing of 30 Gy in five fractions. Notably, many patients
in this series received concurrent systemic therapy with Conclusion
cetuximab or carboplatin. They observed an 81 % overall
response rate (54 % complete response). Two-year locore- Patients presenting with locally advanced head and neck
gional control and overall survival were 31 % and 40 %, cancers represent one of the most challenging patient groups
respectively. Median PFS was 5.7 months. Eleven percent in oncology. In recent decades, multimodality therapy, often
(6 patients) experienced grade 4 treatment-related toxicity employing combined chemotherapy and radiation, has become
and one patient suffered a treatment-related death two weeks the standard of care.
after completion [75]. Our current treatment standards for head and neck
A phase I dose escalation trial conducted at the University cancers result from advances in tumor and radiation biol-
of Pittsburgh enrolled twenty-five patients who received ogy, drug discovery and development, imaging technolo-
SBRT reirradiation for unresectable, locally recurrent gies, advances in radiation planning and delivery, and
HNSCC after previous irradiation [86]. Reirradiation was multidisciplinary collaboration. This has enabled the use of
delivered to the 80 % isodose line with at least 90 % gross stereotactic radiation therapy to become an effective tool in
tumor coverage with a five-fraction course over two weeks. the management of these malignancies. As we continue to
Patients started at the lowest-dose tier of 25 Gy in 5 Gy/ refine these treatments, the use of stereotactic radiation for
fraction and increased until the top-tier dose of 44 Gy was head and neck cancers will broaden. However, it is impera-
reached. No dose limiting toxicities occurred. Four patients tive that we proceed cautiously with the delivery of high
experienced acute grade 1/2 toxicities. No grade 3–5 toxici- doses per fraction to our targets. The proximity and sensi-
ties were reported [86]. tivity of the surrounding normal structures must be appreci-
Table 43.3 shows a selection of single-institution reports ated when tumors are delineated and dose constraints are
utilizing SBRT for head and neck reirradiation. defined. Patient safety must be central to our evaluation of
Investigations of optimal dose fractionation, treatment these technologies as we move forward. Ultimately, this
volumes, and use of systemic therapy with SRS/SBRT for will lead to improved local control, overall survival, and
head and neck malignancies are ongoing and more detail on patient quality of life.
18. Forastiere AA, Zhang Q, Weber RS, Maor MH, Goepfert H, Pajak
References TF, et al. Long-term results of RTOG 91–11: A comparison of three
nonsurgical treatment strategies to preserve the larynx in patients
with locally advanced larynx cancer. J Clin Oncol. 2013;31(7):
1. Siegel R, Naishadham D, Jemal A. Cancer statistics, 2013. CA 845–52.
Cancer J Clin. 2013;63(1):11–30. 19. Denis F, Garaud P, Bardet E, Alfonsi M, Sire C, Germain T, et al.
2. Bernier J, Domenge C, Ozsahin M, Matuszewska K, Lefebvre JL, Final results of the 94-01 French head and neck oncology and
Greiner RH, et al. Postoperative irradiation with or without con- radiotherapy group randomized trial comparing radiotherapy alone
comitant chemotherapy for locally advanced head and neck cancer. with concomitant radiochemotherapy in advanced-stage orophar-
N Engl J Med. 2004;350(19):1945–52. ynx carcinoma. J Clin Oncol. 2004;22(1):69–76.
3. Cooper JS, Pajak TF, Forastiere AA, Jacobs J, Campbell BH, 20. Lee N, Xia P, Fischbein NJ, Akazawa P, Akazawa C, Quivey
Saxman SB, et al. Postoperative concurrent radiotherapy and che- JM. Intensity-modulated radiation therapy for head-and-neck can-
motherapy for high-risk squamous-cell carcinoma of the head and cer: The UCSF experience focusing on target volume delineation.
neck. N Engl J Med. 2004;350(19):1937–44. Int J Radiat Oncol Biol Phys. 2003;57(1):49–60.
4. Adelstein DJ, Li Y, Adams GL, Wagner Jr H, Kish JA, Ensley JF, 21. Lee N, Puri DR, Blanco AI, Chao KS. Intensity-modulated radia-
et al. An intergroup phase III comparison of standard radiation tion therapy in head and neck cancers: An update. Head Neck.
therapy and two schedules of concurrent chemoradiotherapy in 2007;29(4):387–400.
patients with unresectable squamous cell head and neck cancer. 22. Wolden SL, Chen WC, Pfister DG, Kraus DH, Berry SL, Zelefsky
J Clin Oncol. 2003;21(1):92–8. MJ. Intensity-modulated radiation therapy (IMRT) for nasopharynx
5. Calais G, Le Floch O. Concomitant radiotherapy and chemotherapy cancer: Update of the memorial Sloan-Kettering experience. Int J
in the treatment of cancers of the upper respiratory and digestive Radiat Oncol Biol Phys. 2006;64(1):57–62.
tracts. Bull Cancer Radiother. 1996;83(4):321–9. 23. Baujat B, Audry H, Bourhis J, Chan AT, Onat H, Chua DT, et al.
6. Forastiere AA, Goepfert H, Maor M, Pajak TF, Weber R, Morrison Chemotherapy in locally advanced nasopharyngeal carcinoma: An
W, et al. Concurrent chemotherapy and radiotherapy for organ pres- individual patient data meta-analysis of eight randomized trials and
ervation in advanced laryngeal cancer. N Engl J Med. 2003;349(22): 1753 patients. Int J Radiat Oncol Biol Phys. 2006;64(1):47–56.
2091–8. 24. Huncharek M, Kupelnick B. Combined chemoradiation versus
7. Bonner JA, Harari PM, Giralt J, Azarnia N, Shin DM, Cohen RB, radiation therapy alone in locally advanced nasopharyngeal carci-
et al. Radiotherapy plus cetuximab for squamous-cell carcinoma of noma: Results of a meta-analysis of 1,528 patients from six
the head and neck. N Engl J Med. 2006;354(6):567–78. randomized trials. Am J Clin Oncol. 2002;25(3):219–23.
8. Ang KK, Harris J, Wheeler R, Weber R, Rosenthal DI, Nguyen-Tan 25. Langendijk JA, Leemans CR, Buter J, Berkhof J, Slotman BJ. The
PF, et al. Human papillomavirus and survival of patients with oro- additional value of chemotherapy to radiotherapy in locally
pharyngeal cancer. N Engl J Med. 2010;363(1):24–35. advanced nasopharyngeal carcinoma: A meta-analysis of the pub-
9. Brizel DM, Albers ME, Fisher SR, Scher RL, Richtsmeier WJ, lished literature. J Clin Oncol. 2004;22(22):4604–12.
Hars V, et al. Hyperfractionated irradiation with or without concur- 26. Lee N, Harris J, Garden AS, Straube W, Glisson B, Xia P, et al.
rent chemotherapy for locally advanced head and neck cancer. Intensity-modulated radiation therapy with or without chemother-
N Engl J Med. 1998;338(25):1798–804. apy for nasopharyngeal carcinoma: RTOG phase II trial 0225.
10. Al-Sarraf M, Pajak TF, Marcial VA, Mowry P, Cooper JS, Stetz J, J Clin Oncol. 2009;27(22):3684–90.
et al. Concurrent radiotherapy and chemotherapy with cisplatin 27. Fountzilas G, Ciuleanu E, Bobos M, Kalogera-Fountzila A,
in inoperable squamous cell carcinoma of the head and neck. An Eleftheraki AG, Karayannopoulou G, et al. Induction chemother-
RTOG study. Cancer. 1987;59(2):259–65. apy followed by concomitant radiotherapy and weekly cisplatin
11. Pignon JP, Bourhis J, Domenge C, Designe L. Chemotherapy added versus the same concomitant chemoradiotherapy in patients with
to locoregional treatment for head and neck squamous-cell carci- nasopharyngeal carcinoma: A randomized phase II study conducted
noma: three meta-analyses of updated individual data. MACH-NC by the Hellenic cooperative oncology group (HeCOG) with bio-
Collaborative Group. Meta-Analysis of Chemotherapy on Head marker evaluation. Ann Oncol. 2012;23(2):427–35.
and Neck Cancer. Lancet. 2000;355(9208):949–55. 28. Hui EP, Ma BB, Leung SF, King AD, Mo F, Kam MK, et al.
12. Pignon JP, le Maitre A, Maillard E, Bourhis J. MACH-NC Randomized phase II trial of concurrent cisplatin-radiotherapy with
Collaborative Group. Meta-analysis of chemotherapy in head and or without neoadjuvant docetaxel and cisplatin in advanced naso-
neck cancer (MACH-NC): An update on 93 randomised trials and pharyngeal carcinoma. J Clin Oncol. 2009;27(2):242–9.
17,346 patients. Radiother Oncol. 2009;92(1):4–14. 29. Vermorken JB, Remenar E, van Herpen C, Gorlia T, Mesia R,
13. Stell PM. Adjuvant chemotherapy in head and neck cancer. Semin Degardin M, et al. Cisplatin, fluorouracil, and docetaxel in unre-
Radiat Oncol. 1992;2(3):195–205. sectable head and neck cancer. N Engl J Med. 2007;357(17):
14. Munro AJ. An overview of randomised controlled trials of adjuvant 1695–704.
chemotherapy in head and neck cancer. Br J Cancer. 1995; 30. Posner MR, Hershock DM, Blajman CR, Mickiewicz E, Winquist
71(1):83–91. E, Gorbounova V, et al. Cisplatin and fluorouracil alone or with
15. Bourhis J, Pignon JP. Meta-analyses in head and neck squamous docetaxel in head and neck cancer. N Engl J Med. 2007;357(17):
cell carcinoma. What is the role of chemotherapy? Hematol Oncol 1705–15.
Clin North Am. 1999;13(4):769,75, vii. 31. Pointreau Y, Garaud P, Chapet S, Sire C, Tuchais C, Tortochaux J,
16. Induction chemotherapy plus radiation compared with surgery et al. Randomized trial of induction chemotherapy with cisplatin
plus radiation in patients with advanced laryngeal cancer. The and 5-fluorouracil with or without docetaxel for larynx preserva-
Department of Veterans Affairs Laryngeal Cancer Study Group. tion. J Natl Cancer Inst. 2009;101(7):498–506.
N Engl J Med. 1991;324(24):1685–90. 32. Cohen E, Karrison T, Kocherginsky M, Chao H, Agulnik M, Mittal
17. Lefebvre JL, Chevalier D, Luboinski B, Kirkpatrick A, Collette L, B, et al. Docetaxel based chemoradiotherapy plus or minus induc-
Sahmoud T. Larynx preservation in pyriform sinus cancer: tion chemotherapy to decrease events in head and neck cancer
Preliminary results of a European organization for research and (DeCIDE). J Clin Oncol. 2012;30(suppl; abstract 5500.
treatment of cancer phase III trial. EORTC head and neck cancer 33. Haddad R, Rabinowits G, Tishler R, Adkins D, Khuri F, Clark J,
cooperative group. J Natl Cancer Inst. 1996;88(13):890–9. et al. The PARADIGM study: A phase III study comparing sequen-
tial therapy (ST) to concurrent chemoradiotherapy (CRT) in locally 50. Nutting CM, Morden JP, Harrington KJ, Urbano TG, Bhide SA,
advanced head and neck cancer (LANHC). J Clin Oncol. 2012;30 Clark C, et al. Parotid-sparing intensity modulated versus conven-
(suppl; abstract 5501). tional radiotherapy in head and neck cancer (PARSPORT): A phase
34. Ang KK, Andratschke NH, Milas L. Epidermal growth factor 3 multicentre randomised controlled trial. Lancet Oncol. 2011;
receptor and response of head-and-neck carcinoma to therapy. Int J 12(2):127–36.
Radiat Oncol Biol Phys. 2004;58(3):959–65. 51. Mendenhall WM, Riggs CE, Amdur RJ, Hinerman RW, Villaret
35. Bonner JA, Harari PM, Giralt J, Cohen RB, Jones CU, Sur RK, DB. Altered fractionation and/or adjuvant chemotherapy in defini-
et al. Radiotherapy plus cetuximab for locoregionally advanced tive irradiation of squamous cell carcinoma of the head and neck.
head and neck cancer: 5-year survival data from a phase 3 ran- Laryngoscope. 2003;113(3):546–51.
domised trial, and relation between cetuximab-induced rash and 52. Mendenhall WM, Million RR, Stringer SP, Cassisi NJ. Squamous
survival. Lancet Oncol. 2010;11(1):21–8. cell carcinoma of the glottic larynx: A review emphasizing the uni-
36. Vermorken JB, Mesia R, Rivera F, Remenar E, Kawecki A, Rottey versity of Florida philosophy. South Med J. 1999;92(4):385–93.
S, et al. Platinum-based chemotherapy plus cetuximab in head and 53. Wang CC, Efird J, Nakfoor B, Martins P. Local control of T3 carci-
neck cancer. N Engl J Med. 2008;359(11):1116–27. nomas after accelerated fractionation: A look at the “gap”. Int J
37. Ang K, Zhang Q, Rosenthal D, Nguyen-Tan P, Sherman E, Weber Radiat Oncol Biol Phys. 1996;35(3):439–41.
R, et al. A randomized phase III trial (RTOG 0522) of concurrent 54. Overgaard J, Hansen HS, Specht L, Overgaard M, Grau C, Andersen
accelerated radiation plus cisplatin with or without cetuximab E, et al. Five compared with six fractions per week of conventional
for stage III-IV head and neck squamous cell carcinomas (HNC). radiotherapy of squamous-cell carcinoma of head and neck:
J Clin Oncol 2011;29(suppl; abstract 5500). DAHANCA 6 and 7 randomised controlled trial. Lancet. 2003;
38. Lee NY, Zhang Q, Pfister DG, Kim J, Garden AS, Mechalakos J, 362(9388):933–40.
et al. Addition of bevacizumab to standard chemoradiation for 55. Fu KK, Pajak TF, Trotti A, Jones CU, Spencer SA, Phillips TL,
locoregionally advanced nasopharyngeal carcinoma (RTOG 0615): et al. A radiation therapy oncology group (RTOG) phase III ran-
A phase 2 multi-institutional trial. Lancet Oncol. 2012;13(2): domized study to compare hyperfractionation and two variants of
172–80. accelerated fractionation to standard fractionation radiotherapy for
39. Schwartz SM, Daling JR, Doody DR, Wipf GC, Carter JJ, head and neck squamous cell carcinomas: First report of RTOG
Madeleine MM, et al. Oral cancer risk in relation to sexual history 9003. Int J Radiat Oncol Biol Phys. 2000;48(1):7–16.
and evidence of human papillomavirus infection. J Natl Cancer 56. Horiot JC, Le Fur R, N’Guyen T, Chenal C, Schraub S, Alfonsi S,
Inst. 1998;90(21):1626–36. et al. Hyperfractionation versus conventional fractionation in oro-
40. Gillison ML, Koch WM, Shah KV. Human papillomavirus in head pharyngeal carcinoma: Final analysis of a randomized trial of the
and neck squamous cell carcinoma: Are some head and neck EORTC cooperative group of radiotherapy. Radiother Oncol. 1992;
cancers a sexually transmitted disease? Curr Opin Oncol. 1999; 25(4):231–41.
11(3):191–9. 57. Beitler J, Zhang Q, Fu K, Trotti A, Spencer S, Jones C, et al. RTOG
41. Gillison ML, Koch WM, Capone RB, Spafford M, Westra WH, Wu 90-03: Final report. Int J Radiat Oncol Biol Phys. 2012;84(3S):S6.
L, et al. Evidence for a causal association between human papillo- 58. Bourhis J, Overgaard J, Audry H, Ang KK, Saunders M, Bernier J,
mavirus and a subset of head and neck cancers. J Natl Cancer Inst. et al. Hyperfractionated or accelerated radiotherapy in head and
2000;92(9):709–20. neck cancer: A meta-analysis. Lancet. 2006;368(9538):843.54.
42. D’Souza G, Kreimer AR, Viscidi R, Pawlita M, Fakhry C, Koch 59. Jeremic B, Shibamoto Y, Milicic B, Nikolic N, Dagovic A,
WM, et al. Case-control study of human papillomavirus and oro- Aleksandrovic J, et al. Hyperfractionated radiation therapy with or
pharyngeal cancer. N Engl J Med. 2007;356(19):1944–56. without concurrent low-dose daily cisplatin in locally advanced
43. Fakhry C, Westra WH, Li S, Cmelak A, Ridge JA, Pinto H, et al. squamous cell carcinoma of the head and neck: A prospective ran-
Improved survival of patients with human papillomavirus-positive domized trial. J Clin Oncol. 2000;18(7):1458–64.
head and neck squamous cell carcinoma in a prospective clinical 60. Bourhis J, Sire C, Graff P, Gregoire V, Maingon P, Calais G, et al.
trial. J Natl Cancer Inst. 2008;100(4):261–9. Concomitant chemoradiotherapy versus acceleration of radiother-
44. Lewis Jr JS. p16 immunohistochemistry as a standalone test for risk apy with or without concomitant chemotherapy in locally advanced
stratification in oropharyngeal squamous cell carcinoma. Head head and neck carcinoma (GORTEC 99–02): An open-label phase
Neck Pathol. 2012;6 Suppl 1:S75–82. 3 randomised trial. Lancet Oncol. 2012;13(2):145–53.
45. Eisbruch A, Marsh LH, Martel MK, Ship JA, Ten Haken R, Pu AT, 61. Langer CJ, Harris J, Horwitz EM, Nicolaou N, Kies M, Curran W,
et al. Comprehensive irradiation of head and neck cancer using con- et al. Phase II study of low-dose paclitaxel and cisplatin in combi-
formal multisegmental fields: Assessment of target coverage and nation with split-course concomitant twice-daily reirradiation in
noninvolved tissue sparing. Int J Radiat Oncol Biol Phys. 1998; recurrent squamous cell carcinoma of the head and neck: Results of
41(3):559–68. radiation therapy oncology group protocol 9911. J Clin Oncol.
46. Chao KS, Low DA, Perez CA, Purdy JA. Intensity-modulated 2007;25(30):4800–5.
radiation therapy in head and neck cancers: The Mallinckrodt expe- 62. Forastiere AA, Leong T, Rowinsky E, Murphy BA, Vlock DR,
rience. Int J Cancer. 2000;90(2):92–103. DeConti RC, et al. Phase III comparison of high-dose paclitaxel + cis-
47. Puri DR, Chou W, Lee N. Intensity-modulated radiation therapy in platin + granulocyte colony-stimulating factor versus low-dose pacli-
head and neck cancers: Dosimetric advantages and update of clini- taxel + cisplatin in advanced head and neck cancer: Eastern cooperative
cal results. Am J Clin Oncol. 2005;28(4):415–23. oncology group study E1393. J Clin Oncol. 2001; 19(4):1088–95.
48. Eisbruch A, Harris J, Garden AS, Chao CK, Straube W, Harari PM, 63. Gibson MK, Li Y, Murphy B, Hussain MH, DeConti RC, Ensley J,
et al. Multi-institutional trial of accelerated hypofractionated et al. Randomized phase III evaluation of cisplatin plus fluorouracil
intensity-modulated radiation therapy for early-stage oropharyn- versus cisplatin plus paclitaxel in advanced head and neck cancer
geal cancer (RTOG 00–22). Int J Radiat Oncol Biol Phys. 2010; (E1395): An intergroup trial of the eastern cooperative oncology
76(5):1333–8. group. J Clin Oncol. 2005;23(15):3562–7.
49. Kam MK, Leung SF, Zee B, Chau RM, Suen JJ, Mo F, et al. 64. Spencer SA, Wheeler RH, Peters GE, Beenken SW, Meredith RF,
Prospective randomized study of intensity-modulated radiotherapy Smith J, et al. Concomitant chemotherapy and reirradiation as man-
on salivary gland function in early-stage nasopharyngeal carcinoma agement for recurrent cancer of the head and neck. Am J Clin
patients. J Clin Oncol. 2007;25(31):4873–9. Oncol. 1999;22(1):1–5.
65. Spencer SA, Harris J, Wheeler RH, Machtay M, Schultz C, Spanos 76. Cengiz M, Ozyigit G, Yazici G, Dogan A, Yildiz F, Zorlu F, et al.
W, et al. RTOG 96–10: Reirradiation with concurrent hydroxyurea Salvage reirradiation with stereotactic body radiotherapy for locally
and 5-fluorouracil in patients with squamous cell cancer of the head recurrent head-and-neck tumors. Int J Radiat Oncol Biol Phys.
and neck. Int J Radiat Oncol Biol Phys. 2001;51(5):1299–304. 2011;81(1):104–9.
66. Spencer SA, Harris J, Wheeler RH, Machtay M, Schultz C, Spanos 77. Wang K, Heron DE, Clump DA, Flickinger JC, Kubicek GJ,
W, et al. Final report of RTOG 9610, a multi-institutional trial of reir- Rwigema JC, et al. Target delineation in stereotactic body radiation
radiation and chemotherapy for unresectable recurrent squamous cell therapy for recurrent head and neck cancer: A retrospective analysis
carcinoma of the head and neck. Head Neck. 2008;30(3):281–8. of the impact of margins and automated PET-CT segmentation.
67. Janot F, de Raucourt D, Benhamou E, Ferron C, Dolivet G, Radiother Oncol. 2013;106(1):90–5.
Bensadoun RJ, et al. Randomized trial of postoperative reirradia- 78. Foote RL, Pollock BE, Gorman DA, Schomberg PJ, Stafford SL,
tion combined with chemotherapy after salvage surgery compared Link MJ, et al. Glomus jugulare tumor: Tumor control and compli-
with salvage surgery alone in head and neck carcinoma. J Clin cations after stereotactic radiosurgery. Head Neck. 2002;24(4):332,8.
Oncol. 2008;26(34):5518–23. discussion 338–9.
68. De Crevoisier R, Bourhis J, Domenge C, Wibault P, Koscielny S, 79. Elshaikh MA, Mahmoud-Ahmed AS, Kinney SE, Wood BG, Lee
Lusinchi A, et al. Full-dose reirradiation for unresectable head and JH, Barnett GH, et al. Recurrent head-and-neck chemodectomas:
neck carcinoma: Experience at the Gustave-Roussy Institute in a a comparison of surgical and radiotherapeutic results. Int J Radiat
series of 169 patients. J Clin Oncol. 1998;16(11):3556–62. Oncol Biol Phys. 2002;52(4):953–6.
69. Salama JK, Vokes EE, Chmura SJ, Milano MT, Kao J, Stenson KM, 80. Ahn YC, Lee KC, Kim DY, Huh SJ, Yeo IH, Lim DH, et al.
et al. Long-term outcome of concurrent chemotherapy and reirra- Fractionated stereotactic radiation therapy for extracranial head and
diation for recurrent and second primary head-and-neck squamous neck tumors. Int J Radiat Oncol Biol Phys. 2000;48(2):501–5.
cell carcinoma. Int J Radiat Oncol Biol Phys. 2006;64(2):382–91. 81. Hwang JM, Fu KK, Phillips TL. Results and prognostic factors in
70. Machtay M, Rosenthal DI, Chalian AA, Lustig R, Hershock D, the retreatment of locally recurrent nasopharyngeal carcinoma. Int
Miller L, et al. Pilot study of postoperative reirradiation, chemo- J Radiat Oncol Biol Phys. 1998;41(5):1099–111.
therapy, and amifostine after surgical salvage for recurrent head- 82. Buatti JM, Friedman WA, Bova FJ, Mendenhall WM. Linac radio-
and-neck cancer. Int J Radiat Oncol Biol Phys. 2004;59(1):72–7. surgery for locally recurrent nasopharyngeal carcinoma: Rationale
71. Tanvetyanon T, Padhya T, McCaffrey J, Zhu W, Boulware D, and technique. Head Neck. 1995;17(1):14–9.
Deconti R, et al. Prognostic factors for survival after salvage reir- 83. Ryu S, Khan M, Yin FF, Concus A, Ajlouni M, Benninger MS,
radiation of head and neck cancer. J Clin Oncol. 2009;27(12): et al. Image-guided radiosurgery of head and neck cancers.
1983–91. Otolaryngol Head Neck Surg. 2004;130(6):690–7.
72. Langendijk JA, Kasperts N, Leemans CR, Doornaert P, Slotman 84. Siddiqui F, Patel M, Khan M, McLean S, Dragovic J, Jin JY, et al.
BJ. A phase II study of primary reirradiation in squamous cell car- Stereotactic body radiation therapy for primary, recurrent, and met-
cinoma of head and neck. Radiother Oncol. 2006;78(3):306–12. astatic tumors in the head-and-neck region. Int J Radiat Oncol Biol
73. Cmelak AJ, Cox RS, Adler JR, Fee Jr WE, Goffinet DR. Phys. 2009;74(4):1047–53.
Radiosurgery for skull base malignancies and nasopharyngeal car- 85. Roh KW, Jang JS, Kim MS, Sun DI, Kim BS, Jung SL, et al.
cinoma. Int J Radiat Oncol Biol Phys. 1997;37(5):997–1003. Fractionated stereotactic radiotherapy as reirradiation for locally
74. Kondziolka D, Lunsford LD. Stereotactic radiosurgery for squa- recurrent head and neck cancer. Int J Radiat Oncol Biol Phys.
mous cell carcinoma of the nasopharynx. Laryngoscope. 1991; 2009;74(5):1348–55.
101(5):519–22. 86. Heron DE, Ferris RL, Karamouzis M, Andrade RS, Deeb EL,
75. Unger KR, Lominska CE, Deeken JF, Davidson BJ, Newkirk KA, Burton S, et al. Stereotactic body radiotherapy for recurrent squa-
Gagnon GJ, et al. Fractionated stereotactic radiosurgery for reirra- mous cell carcinoma of the head and neck: Results of a phase I
diation of head-and-neck cancer. Int J Radiat Oncol Biol Phys. dose-escalation trial. Int J Radiat Oncol Biol Phys. 2009;75(5):
2010;77(5):1411–9. 1493–500.
Spinal Tumors and Radiosurgery
44
Evangelia Katsoulakis, Ilya Laufer, and Yoshiya Yamada
underwent prior resection (mean number of resections, 2.9).

Benign Tumors There was a total of 16 spinal lesions treated to a dose of
21 Gy (range, 20–25 GY) in 1–3 fractions (median, 2.0 frac-
Unlike with malignant tumors, where radiation had a long- tions) with a median follow-up of 33.5 months (range, 6–73
established role, radiation as a primary treatment modality months). The mean radiation dose to the periphery of tumor
for benign tumors is debatable because of the long life expec- was 23.4 Gy (range, 12–40 Gy) and the mean tumor volume
tancies of patients that are afflicted with these tumors. Late was 1.8 cm (range, 0.058–65.4 cm3). Of the tumors that
radiation toxicity can take many years to manifest and the exhibited neurologic symptoms, 88 % showed improvement
possibility of radiation-induced myelopathy must be thor- or complete resolution. The actuarial radiographic local con-
oughly considered when designing treatment management. trol rates for the spinal lesions at 36 months and 60 months
Our discussion for benign tumors is limited to vascular mal- were 92 % and 92 %, respectively [3]. Moreover, none of the
formations, namely, hemangioblastomas and arteriovenous patients treated with spinal SRS developed neurologic toxic-
malformations (AVMs). ity secondary to radiation [3].
In another series by Selch et al. [4], consisting solely of
spinal lesions, 20 tumors in 9 patients (5 of which had VHL)
Hemangioblastomas were treated with image-guided radiosurgery to a dose of
12 Gy in a single fraction. With a median follow-up of 51
Hemangioblastomas represent a group of benign, highly vas- months (range, 14–86 months), the 4-year local control rate
cular tumors that are typically pial-based. While the majority for the entire cohort was 90 %. It has been surmised that solid
of hemangioblastomas (~80 %) occur sporadically, they are or mural nodular components of hemangioblastomas tend to
often associated with an inherited syndrome, von Hipple– respond much more favorably to radiosurgery than cystic
Lindau (VHL) disease [1]. Historically, the standard of care components. Indeed, in this series, solid tumor control rates
for hemangioblastomas is complete microsurgical removal. were reportedly higher than cystic with a 4-year control rate
Surgical resection, however, may aggravate underlying neu- of 95 %. While the single-fraction 12 Gy dose was lower than
rological symptoms or incite new ones [2]. Additionally, other reported series, local control rates were excellent. None
many patients are also poor surgical candidates or have inop- of the patients developed neurotoxicity or myelopathy.
erable lesions, and radiosurgery provides an attractive treat- Daly et al. [5] reported on the highest maximum radiosur-
ment option. gical doses ever delivered to the spinal cord in their series of
In one of the largest retrospective series reported by Moss 25 spinal hemangioblastomas from 18 patients. Seventeen
et al., 82 hemangioblastomas involving either the spine or tumors were treated to a median dose of 20 Gy (range,
cranium in 28 patients underwent stereotactic radiosurgery 18–30 Gy) in a single fraction. The remainder of the lesions
(SRS). The majority of patients (89.2 %) had VHL and 93 % were treated to 18–25 Gy in 2–3 fractions. Thirty percent had
prior surgical resection. With a median follow-up of
E. Katsoulakis, M.D. • Y. Yamada, M.D., F.R.C.P.C. (*) 33.7 months (range, 6.6–84.0 months), the 3-year actuarial
Department of Radiation Oncology, Memorial Sloan Kettering
local control was 86 %. There was one late grade 2 neuro-
Cancer Center, 1275 York Avenue, New York, NY 10065, USA
e-mail: ekatsoulakis@gmail.com; yamadaj@mskcc.org logic toxicity, which manifested as a unilateral foot drop.
The 3-year rate of grade 2 or more toxicity was 4 %. While it
I. Laufer
Department of Neurosurgery, Memorial Sloan Kettering is difficult to discern the etiology of declining neurologic sta-
Cancer Center, 1275 York Avenue, New York, NY 10065, USA tus, and whether the cause is disease progression or radiation
564 E. Katsoulakis et al.
toxicity, patients received the highest radiosurgical dose ever median dose of 22 Gy (range, 15–27 Gy) had a 10-year total
reported to the spine. obliteration rate of 93 % determined by angiography or MR
Even though data continues to accumulate regarding the [6]. Smaller target volumes, smaller target maximum diam-
efficacy of radiosurgery for hemangioblastomas, appropriate eter, and larger dose correlated with greater rates of
spinal cord dose constraints are poorly defined in the radio- obliteration. The cumulative rates of AVM hemorrhage after
surgical literature and warrant further discussion and longer SRS at 5 and 10 years were 5.0 % and 6.1 %, respectively.
follow-up in the setting of benign disease. In the series by The corresponding rates of annual bleeding were 0.3 % and
Daly et al. [5], patients received a median spinal cord Dmax of 0.2 % for years 1–5 and 5–10, respectively. There are fewer
22.7 Gy (range, 17.8–30.9 Gy) in a single fraction, while the reports in the literature, however, for spinal AVMs.
multiple-fraction group had a median single-fraction biolog- The mainstay of therapy for type I and IV spinal AVMs,
ically equivalent dose (SFBED) spinal cord Dmax of 14.1 Gy3 which represent dural and perimedullary arteriovenous fistu-
(range, 12.3–19.4 Gy3). The majority of treatments (59 %) las, is microsurgical resection. Type II and III AVMs, how-
in the single-fraction group exceeded that of the Radiation ever, are located within the parenchyma of the spinal cord
Therapy Oncology Group (RTOG) spinal radiosurgery pro- and it is these lesions that are best suited for radiosurgery.
tocol 0631. While toxicity was limited in the study, prospec- Whereas type II spinal AVMs consist of a compact intramed-
tive studies with longer follow-up and larger patient numbers ullary nidus, type III spinal AVMs are composed of a diffuse
are required prior to drawing definitive conclusions with intramedullary nidus and often contain a paraspinal component.
respect to spinal cord tolerance. In addition, hemangioblas- Both types are not amenable to surgery. Sinclair et al. [7]
tomas are known to have a long time to progression and reported on the largest experience with spinal cord AVMs
exhibit quiescent phases that could partially account for the treated with radiosurgery in which 15 patients with intra-
excellent control rates. In a study of cranial hemangioblasto- medullary spinal cord AVMs (mean target volume, 2.36 mL;
mas treated with radiosurgery, median time to tumor pro- range, 0.79–5.23 mL) were treated to a dose of 20.5 Gy
gression was 5.9 years [2]. While radiosurgery is clearly (range, 15–25 Gy) in 3 sessions (range, 2–5 sessions). All
efficacious for the treatment of spinal hemangioblastomas, patients were followed with magnetic resonance imaging
much longer follow-up is necessary to definitively conclude (mean follow-up of 27.9 months; range, 3–59 months),
its equivalence to microsurgery. which revealed a reduction in the AVM volume. Only five
patients, however, underwent angiography as part of their
follow-up. Of these patients, only 1 had complete obliteration
Arteriovenous Malformation on angiography. There were no instances of rebleeding after
radiosurgery. Clinically, patient outcomes were improved or
Radiosurgery has long been acknowledged as a noteworthy remained stable in 92.3 %. The authors have updated this
alternative to microsurgical resection and embolization for series in abstract form and have successfully treated over 24
cerebral AVMs. Over 5000 patients have been successfully patients in total with similar results. Table 44.1 provides a
treated with high-dose radiation since the inception of SRS summary of IG-IMRT for benign tumors. The optimal dose
in the 1970s with obliteration rates of 80–85 % over a 2- to and long-term durability of radiosurgery for spinal AVMs
3-year period for lesions less than 2.5 cm. In a recent series have yet to be determined. These reports suggest that radio-
by Kano et al., 996 patients with grade 1 and 2 cerebral surgery will play an integral role in the management of
AVMs with a median tumor volume of 2.3 cm treated to a spinal AVMs with an intramedullary nidus.
Table 44.1 Spine radiosurgery for benign tumors

Total dose
Patients (range)/No. of Spinal cord Follow-up
Author (lesions) Histology fractions (range) dose (months) Local control Myelopathy
Moss et al. [3] 16 (16) Hemangioblastoma 21 Gy (20– NR 33.5 92 % 0%
25 Gy)/2 (1–3)
Selch et al. [4] 9 (20) Hemangioblastoma 12 Gy/1 Dmax 12 Gy 51 90 % 0%
and
V10 = 10 %
Daly et al. [5] 25 (18) Hemangioblastoma 20 Gy (18–30)/1 Dmax 22.7 Gy 33.7 86 % 4 % grade ≥2
18–25 Gy/(2–3) Dmax 14.1 Gy
Sinclair et al. [7] 15 AVM, type II 20.5 Gy NR 27.9 All patients had 0%
(15–25)/3 (2–5) radiographic
reduction of nidus
Long term follow-up in this patient population indicates excellent local control with minimal toxicity
44 Spinal Tumors and Radiosurgery 565
Henderson et al. reported on a series of 24 lesions, the

Malignant Tumors majority of which (61 %) involved the spine or sacrum. The
median tumor volume was 128 mL (range, 12.0–457.3 mL) to
Primary Tumors a median dose of 35 Gy (range, 24.0–40.0 Gy) in 5 fractions.
Median follow-up was 46 months (range, 7–65 months) and
Chordomas actuarial 5-year local control was 59.1 %. Toxicity profiles
Although quite rare, chordomas are the most common primary were similar to those reported in other series, and no patients
malignant tumor of the spine. Standard of care for these experienced grade 3 or 4 neurologic complications [13].
tumors is en bloc resection with wide margins. In the setting The authors currently employ image-guided high-dose
of an incomplete or intralesional resection, which often radiosurgery neoadjuvantly prior to surgical resection. In
arises in the presence of epidural disease, adjuvant radiation cases where gross total resection is achievable, a neoadju-
to a total dose of at least 65 Gy is indicated [8–10]. In a vant approach may facilitate intralesional resection. The
report by Terezakis et al., seven chordomas were treated with neoadjuvant approach was inspired by a case report in which
postoperative image-guided conventionally fractionated a patient with spinal chordoma underwent definitive single-
radiation therapy to a median dose of 6,600 cGy (range, fraction, high-dose, image-guided radiation to the tumor.
5,396–7,080 cGy) in 180–200 cGy per fraction. The median Unfortunately, 4 months after treatment the patient under-
planning target volume was 164 cm3 (range, 29–1,116 cm3). went L3 corpectomy due to symptoms of mechanical insta-
With a median follow-up of 17.4 months (range, 2.1– bility [14], which revealed near-complete necrosis of the
47.3 months), one patient developed in-field recurrence. resected chordoma. Results of our institutional experience
There were no patients who experienced myelopathy. with the neoadjuvant approach will be published shortly.
Proton radiation therapy is an established treatment option
for the treatment for sacral chordomas, and outcomes corre-
late with dose. In a study by Park et al. [11], 27 patients with Spinal Metastases
sacral chordomas (16 primary and 11 recurrent), most of
whom underwent surgical resection (78 %), received photon/ The prevalence of spinal metastases in North America is
proton radiation, and were followed for a mean of 8.8 years. over 100,000 patients each year. Over 40 % of cancer patients
The mean dose for primary tumors was 71 Gy (E) in 36 frac- will develop metastatic disease in the spine, the bulk of
tions versus 77 Gy (E) in 41 fractions for recurrent tumor. which is extradural. Highly effective and durable palliation
The mean diameter for primary tumors was 8.6 cm (range, is clinically significant, especially in the context of longer
2.8–19.3 cm) versus 5.7 cm (range, 2.7–9 cm) for recurrent life expectancies secondary to improved systemic agents
tumors. Local control rates for surgery and radiation were [15]. Conventionally fractionated radiation therapy has been
12/14 (86 %) for primary versus 1/7 (14 %) for recurrent an integral component of the treatment paradigm for spinal
chordomas. Among patients receiving definitive radiation, metastases; one or two open fields are used to deliver the
doses of 73 Gy (E) or more resulted in greater rates of local radiation dose bound by spinal cord tolerance [16]. The evo-
control [11]. The authors recommend 77.4 Gy (E) to areas of lution of image-guided technology and breakthroughs in
gross residual disease or unresected chordomas. Dose esca- treatment planning has revolutionized the management of
lation resulted in high rates of local control. Proton radiation spinal metastases [17–19]. In situations where the tumor is in
facilities are not readily accessible; however, this prompted close proximity to the spinal cord, safe dose escalation is
photon dose escalation with modern image-guided tech- possible via steep dose gradients. While these techniques
niques, enabling hypofractionation. have been used in the primary-irradiation, postoperative set-
In the study by Jiang et al. [12], both intracranial and ting, as well as re-irradiation, the optimal dose, patient-
extracranial chordomas were treated to a dose of 32.5 Gy selection, and delivery method has yet to be defined.
(range, 18–50 Gy) in 1–5 fractions. The average tumor vol-
ume was 16.1 cm3 (range, 2.4–45.9 cm3). With a reasonable
follow-up of 34 months (range, 2–131 months), overall Primary Irradiation: Single Fraction
tumor control for the whole cohort was achieved in 11 of 20
patients (55 %). Eight of those patients had a reduction in The ideal fractionation schema for palliation of bone
tumor size. Moreover, patients with spinal chordomas (seven metastases has been studied extensively [14, 20]. One of
in total) had improved symptoms and higher tumor control the most eminent trials, RTOG 9714, randomized patients
than intracranial lesions, 71 % versus 46 %. One patient with to 8 Gy × 1 versus 3 Gy × 10. The single-fraction, low-dose
no evidence of disease radiographically had collapsed at L3 arm resulted in higher rates of retreatment with radiation,
and required surgical decompression. 18 % versus 9 % [20]. In a contrasting meta-analysis of
over 3,260 patients, clinical outcome was unchanged with single-fraction, high-dose IG-IMRT was the primary
single-fraction, low-dose (median, 8 Gy) versus fractionated treatment modality. The mean tumor dose was 20 Gy
radiation (median, 4 Gy × 5) [14]. Palliation of pain, however, (range, 12–25 Gy), while mean follow-up was 21 months
favored single-fraction, low-dose radiation. The recently (range, 3–53 months). Long-term control and pain improve-
updated meta-analysis now includes over 5,000 patients and ment was excellent at 90 % and 86 %, respectively. There
findings show no significant difference in pain relief [21]. were no cases of myelopathy [28].
Particularly for spinal metastases, long-term outcomes of In a larger series of 93 patients with 103 spinal metastases
conventional fractionation or single-fraction, low-dose radi- treated with radiation as the primary treatment modality,
ation for the treatment of spinal metastases have been Yamada et al. reported excellent overall radiographic local
reported in a prospective study by Nagoya University. control of 90 % [29]. In this series, patients received single-
Approximately 101 patients with spinal tumors received fraction, high-dose IG-IMRT of 18–24 Gy (median, 24 Gy).
40 Gy in 20 fractions, were followed for more than 24 An example of isodose distribution for high-dose IG-IMRT
months (11 months for dying patients and 53 months for sur- is shown in Fig. 44.1. The median follow-up was 15 months
vivors), and reported 67 % pain relief [22]. Similarly, in a and all patients underwent close surveillance with magnetic
randomized study by Maranzano et al., metastatic cord com- resonance imaging at each scheduled follow-up visit to
pression received 8 Gy × 2 over one well versus a split course assess local control. The 2-year estimated cumulative inci-
of 5 Gy × 3 plus 3 Gy × 5. At 33 months follow-up, pain relief dence of local failure with death as a competing risk was
was similar for short- versus long-course patients, 56 % and 7.5 %, with median time to failure of 9 months. Tumor his-
59 %, respectively [23]. In general, the mean pain relief tology did not predict for local control on multivariate analy-
expected from conventional fractionated radiation is 70 % sis, while radiation dose was the only significant predictor.
(range, 56–100 %). The local control rates for low-dose (18–23 Gy, mean
Radiographic local control rates are similar to pain relief 20.8 Gy) versus high-dose (24 Gy) radiation were 80 % and
with conventionally fractionated radiation, about 77 %. One 95 %, respectively (p = 0.03). In all cases, death was due to
of the most critical prognosticators of recurrence is tumor systemic disease progression rather than local failure. The
histology. Favorable or radiosensitive histologies include median time to death after treatment was 10 months (range,
breast cancer, prostate cancer, lymphomas, seminomas, and 1–39 months), while median overall survival was 15 months.
myelomas. Unfavorable or radioresistant tumors include At the time of death, all but one patient had no evidence of
sarcomas, melanomas, renal cell carcinomas, gastrointestinal local failure.
cancers, and non-small cell lung cancer [24]. In the afore- Due to the observed dose response, the planned dose dis-
mentioned study by Maranzano, radiosensitive histologies tributions were retrospectively reviewed [30]. Dosing param-
had significantly improved median survival and duration of eters, including Dmin and D95, were examined for patients
neurologic motor improvement with spine radiation, 10 experiencing local failure and compared with the whole
versus 3 months (p = 0.001) and 11 versus 3 months cohort. The median Dmin (10.8 Gy) for patients experiencing
(p = 0.001), respectively [23]. Additionally, for the long- local failure was significantly lower than those patients expe-
term survivors of the Nagoya study, durable response at 53 riencing local control (Dmin, 15.7 Gy). Moreover, when Dmin
months was greater in the radiosensitive tumors (30 % ver- was <15 Gy, local failure rates were 16 %. However, when
sus 10 %) [22]. Dmin was >15 Gy, no failures were observed in any histology.
Successful outcomes for spinal metastases should be dif- It would appear that underdosing to even a small subvolume
ferentiated from osseous metastases and should include close of treatment target is clinically significant.
radiographic follow-up. Single-fraction, high-dose image-
guided intensity-modulated radiation therapy (IG-IMRT) is
proved to be highly effective in both pain and radiographic Adjuvant Postoperative: Single Fraction
control, irrespective of tumor histology. Pain improvement
is approximately 85 % and often immediate, occurring If the spinal cord is compromised by tumor, such as with
within days of treatment [19, 24]. In general, pain usually high-grade epidural compression or spinal instability,
decreases within weeks of treatment. Long-term radio- decompressive surgery is the standard of care, resulting in
graphic tumor-control rates are excellent at 90 % [25–27]. improved neurologic and functional outcomes [30]. Residual
Moreover, radioresistant tumor histologies that respond tumor at the dural margin is at risk for recurrence; thus, the
poorly to conventionally fractionated radiation such as renal majority of metastatic spinal tumors require adjuvant radia-
cell carcinoma or melanoma are potentially cured locally tion for microscopic or gross disease. High-dose, single-
with single-fraction, high-dose IG-IMRT. fraction IG-IMRT has demonstrated superior outcomes
Gerszten et al. reported on a mixed population of 500 compared with conventionally fractionated adjuvant radia-
patients with spinal metastases. In approximately 65 patients, tion for radioresistant histologies. Postsurgical intervention,
Fig. 44.1 Forty-eight-year-old man with radioresistant liposarcoma space. (b) Dose volume histogram reflecting the dosimetric advantages
and single metastasis at T12 treated with 2,400 cGy in a single fraction. of image-guided intensity-modulated radiation therapy and limited
Computed tomography myelogram is used to define the true cord. dose to the true cord
(a) Isodose distribution revealing a high-dose gradient in the epidural
the tumor environment is hypoxic as well as radioresistant skin dose is negligible due to steep dose gradients, IG-IMRT
and the 1-year local control rate is dismal at 20–30 % [31]. may be given early in the postoperative period.
Additionally, treatment is often delayed, with conventional In a small series of 18 patients which included highly radio-
radiation secondary to wound healing. Conversely, since sensitive histologies such as multiple myeloma, postoperative
single-fraction IG-IMRT was delivered with a mean dose of Just as in the definitive setting, a robust tumor ablative
11.4 Gy (range, 6–16 Gy). The results were outstanding, with dose is critical. In the MSKCC series, the salvage reirradia-
92 % of patients having stable or improved neurologic symp- tion dose was either 20 Gy (five 4 Gy fractions) or 30 Gy
toms at a median follow-up of 7 months [32]. (five 6 Gy fractions). The salvage dose was the only signifi-
In another series by Moulding et al., 21 patients received cant predictor of radiographic local failure (p = 0.04). The
adjuvant postoperative high-dose, single-fraction IG-IMRT incidence of radiographic local failure rates for 20 Gy and
to a higher median dose of 24 Gy (range, 18–24 Gy) [33]. 30 Gy was 45 % and 26 %, respectively. Moreover, the BED
In this series, the overwhelming majority of patients (95 %) for the high-dose 30 Gy and low-dose 20 Gy fractionations
had radioresistant histologies. Patients were grouped into are equivalent to 48 Gy10 and 28 Gy10, respectively. The 1.7-
high- and low-dose IG-IMRT groups of 24 Gy and 18 or fold BED increase correlated with a 1.7-fold decrease
21 Gy, respectively. The overall local control for the entire in local failure rate at 1 year. Dose, however, did not impact
cohort was 81 % at a median follow-up of 310 days. When survival; the median survival for the entire cohort was
stratified by dose group, the high-dose IG-IMRT group had a 13.6 months. Radiosensitive histology and time to reirradia-
lower 1-year estimated local failure risk than the low-dose tion were both associated with improved survival. Salvage
IG-IMRT group, 6.3 % versus 20 %. Overall survival was reirradiation was highly effective in palliating pain or neuro-
not significantly different among high-dose versus low-dose logic symptoms, with 85 % of symptomatic patients report-
IG-IMRT. ing no progression and 77 % experiencing mild or significant
Adjuvant IG-IMRT after surgical decompression and pain relief.
instrumentation is highly effective in ensuring local control The reirradiation prospective trial by Garg et al. reported
and sustaining neurologic recovery, especially for radioresis- similar results [40]. The freedom from neurologic deteriora-
tant histologies. tion at 1 and 3 years was 92 % and 81 %, respectively. While
not all 63 lesions were true in-field recurrences, the majority
(85 %) received 27 Gy in 3 fractions. In this series, tumor
Salvage Reirradiation Dose progression was not associated with tumor histology. Overall
and Hypofractionation survival correlated with prior radiation therapy dose received;
prior dose ≥35 Gy had a median survival of 33 months,
Tumor recurrence or progression after conventional radio- whereas prior dose <35 Gy had a median survival of 21
therapy at the original site is common, on the order of moths (p = 0.01). The correlation with prior radiation therapy
50–80 % at 2 years [22]. Furthermore, recurrences after a dose and survival, however, was most likely associated with
previous course of conventional radiation are more likely to whether radiation was initially given with a curative or a pal-
be radioresistant since they ultimately failed radiation and liative intent.
thus require robust radiation doses for successful salvage. In the study by Gerszten et al., 344 patients received sal-
Reirradiation with conventional means is limited by spinal vage reirradiation with a single fraction of 12–25 Gy [28].
cord tolerance and the concern for radiation-induced myelop- Prior radiation was either with 3 Gy × 10 fractions or
athy. In instances where additional conventional radiation or 2.5 Gy × 14 fractions. While long-term pain control was
surgery is not appropriate, IG-IMRT is a successful option. excellent (86 %) and independent of histology, radiographic
The effectiveness of IG-IMRT has been reported in many control was 88 % and varied with histology. For instance,
retrospective series with small patient numbers [34–37]. The radiosensitive breast carcinoma experienced 100 % local
largest series was reported by Memorial Sloan-Kettering control, whereas local control for radioresistant melanomas
Cancer Center, which examined 94 patients who underwent was 75 %. The relationship between dose and local control
reirradiation for true in-field recurrences [38]. IG-IMRT was was not reported. While the optimal dose and fractionation
delivered using 4–6 Gy per fraction and a total dose of of salvage reirradiation with IG-IMRT has yet to be defined,
20–30 Gy. The 1-year radiographic control was 66 %. In a safe delivery of high-dose radiation is feasible and provides
prospective phase 1–2 trial by M.D. Anderson Cancer Center, excellent rates of local control and symptom management.
63 previously irradiated paraspinal metastases received Table 44.2 provides a summary of salvage IG-IMRT after
6–9 Gy per fraction to a total of 27 or 30 Gy with a 76 % prior treatment with radiation or surgery.
radiographic control rate at 1 year [39, 40]. The University of
Pittsburgh conducted one of the largest prospective trials of
spinal metastases [28]. Over 500 mixed lesions were Brachy P32 Plaques
included, 68.8 % of whom received prior conventional radia-
tion. The patients were salvaged with high-dose single frac- While reirradiation using image-guided techniques is highly
tion radiation therapy to 12.5–25 Gy and long-term effective at tumor ablation and limiting dose to the cord,
radiographic control rates were excellent at 88 %. there are instances in which tumor extends into the spinal
Table 44.2 Spine radiosurgery as salvage for malignant tumors

Total dose (range)/
Patients No. of fractions Spinal cord Follow-up Local control
Authors (lesions) Prior treatment (range) dose (months) ( %) Myelopathy
Damast et al. [38] 94 (97) Radiation 20 Gy/5 Dmax 10 Gy 12.1 55 0%
30 Gy/5 Dmax 14 Gy 74
Garg et al. [40] 59 (63) Radiation 30 Gy/5 Dmean 10 Gy 17.6 76 3.3 % ≥grade 3
27 Gy/3 Dmean 9 Gy
Gerszten [28] 344 (344) Radiation 20 Gy NR 21 88 0%
(12–25.9 Gy)/1
Rock et al. [32] 18 (18) Surgery 11.4 Gy mean NR 7 94 NR
(6–16 Gy)/1
Moulding et al. [33] 21 (21) Surgery 24 Gy/1 Dmax 14 Gy 10.2 93.80 0%
18–22 Gy/1 81
In the setting of prior treatment with surgery or radiation, spinal cord sparing IG-IMRT provides good local control. Local control outcomes are
dose-dependent, with high-dose single-fraction treatment resulting in the highest reported rates
canal and impinges on the dura. Delivering a robust dose to

the dural margin, which is likely to harbor microscopic Conclusions
tumor involvement, with IG-IMRT alone can be challenging.
Brachytherapy has been used to circumvent spinal cord radi- Image-guided radiosurgery continues to widen the therapeutic
ation tolerance while delivering high doses to the dural mar- window for disease involving the spine. Spine radiosurgery is
gin. Many radioactive sources have been used including proof of the principle that the accurate delivery of tumoricidal
Iodine 125, Ytrium 90, Iridium 92, and P32. doses of radiation is feasible with the judicious use of image-
DeLaney et al. reported on their experience with three guided technology and highly conformal treatment planning.
plaque designs in the treatment of eight patients with para- Additionally, patients tolerate very high-dose radiation when
vertebral or vertebral tumors abutting the dura treated with a image-guided techniques and brachytherapy are utilized to min-
combination of conventional radiation (median dose, 54 Gy; imize the volume of normal spinal cord tissue that is irradiated.
range, 16.2–70.2 Gy), tumor resection, and intraoperative
brachytherapy to the dura. The authors support the use of Conflict of Interest Yoshiya Yamada is a member of the Speakers
rigid Ytrium 90 plaque for radiation safety and delivered Bureau of the Institute for Medical Education and has served as a con-
sultant for Varian Medical Systems.
7.5–15 Gy at the dural surface. The results were promising.
At a median follow-up of 24 months (range, 10–34), 6 of the
8 patients were locally controlled [41].
In the largest recent spine brachytherapy series to date, References
Folkert et al. [42] reported on 25 patients who received
prior conventional radiation and later developed recurrent 1. Resche F, Moisan JP, Mantoura J, et al. Haemangioblastoma, hae-
mangioblastomatosis, and von Hippel-Lindau disease. Adv Tech
primary and progressive metastatic disease requiring surgi-
Stand Neurosurg. 1993;20:197–304.
cal intervention due to dural involvement with an intraop- 2. Asthagiri AR, Mehta GU, Zach L, et al. Prospective evaluation of
erative flexible brachytherapy plaque. P32 plaque was radiosurgery for hemangioblastomas in von Hippel-Lindau disease.
prescribed as 10 Gy at 1 mm depth, resulting in a dural sur- Neuro Oncol. 2010;12(1):80–6.
3. Moss JM, Choi CY, Adler Jr JR, Soltys SG, Gibbs IC, Chang
face dose of 2,550 cGy and a maximum cord dose of
SD. Stereotactic radiosurgical treatment of cranial and spinal heman-
160 cGy. The majority of patients, 15 of 25 (60 %), received gioblastomas. Neurosurgery. 2009;65(1):79–85; discussion 85.
additional high-dose IG-IMRT (median dose, 2,700 cGy; 4. Selch MT, Tenn S, Agazaryan N, Lee SP, Gorgulho A, De Salles
range, 1,800–3,000 cGy). With a median follow-up of AA. Image-guided linear accelerator-based spinal radiosurgery for
hemangioblastoma. Surg Neurol Int. 2012;3:73.
7.3 months (range, 2.6–23.3 months), local control was
5. Daly ME, Choi CY, Gibbs IC, et al. Tolerance of the spinal cord to
82.4 % and time to failure was 3.2 months. There were no stereotactic radiosurgery: insights from hemangioblastomas. Int J
intraoperative or postoperative complications secondary to Radiat Oncol Biol Phys. 2011;80(1):213–20.
radiotherapy. While longer follow-up is warranted, the 6. Kano H, Lunsford LD, Flickinger JC, et al. Stereotactic radiosurgery
for arteriovenous malformations, Part 1: management of Spetzler-
future of brachytherapy in the management of spinal tumors
Martin Grade I and II arteriovenous malformations. J Neurosurg.
appears bright. 2012;116(1):11–20.
7. Sinclair J, Chang SD, Gibbs IC, Adler Jr JR. Multisession 25. Bilsky MH, Yamada Y, Yenice KM, et al. Intensity-modulated
CyberKnife radiosurgery for intramedullary spinal cord arteriove- stereotactic radiotherapy of paraspinal tumors: a preliminary report.
nous malformations. Neurosurgery. 2006;58(6):1081–9; discussion Neurosurgery. 2004;54(4):823–30; discussion 830–831.
1081–1089. 26. De Salles AA, Pedroso AG, Medin P, et al. Spinal lesions treated
8. Boriani S, Saravanja D, Yamada Y, Varga PP, Biagini R, Fisher CG. with Novalis shaped beam intensity-modulated radiosurgery and
Challenges of local recurrence and cure in low grade malignant stereotactic radiotherapy. J Neurosurg. 2004;101 Suppl 3:435–40.
tumors of the spine. Spine (Phila Pa 1976). 2009;34(22 Suppl): 27. Kim B, Soisson ET, Duma C, et al. Image-guided helical tomo-
S48–57. therapy for treatment of spine tumors. Clin Neurol Neurosurg.
9. Romero J, Cardenes H, la Torre A, et al. Chordoma: results of radi- 2008;110(4):357–62.
ation therapy in eighteen patients. Radiother Oncol. 1993;29(1): 28. Gerszten PC, Burton SA, Ozhasoglu C, Welch WC. Radiosurgery
27–32. for spinal metastases: clinical experience in 500 cases from a single
10. Terezakis SA, Lovelock DM, Bilsky MH, Hunt MA, Zatcky J, institution. Spine (Phila Pa 1976). 2007;32(2):193–9.
Yamada Y. Image-guided intensity-modulated photon radiotherapy 29. Yamada Y, Bilsky MH, Lovelock DM, et al. High-dose, single-
using multifractionated regimen to paraspinal chordomas and rare fraction image-guided intensity-modulated radiotherapy for meta-
sarcomas. Int J Radiat Oncol Biol Phys. 2007;69(5):1502–8. static spinal lesions. Int J Radiat Oncol Biol Phys. 2008;71(2):
11. Park L, Delaney TF, Liebsch NJ, et al. Sacral chordomas: Impact of 484–90.
high-dose proton/photon-beam radiation therapy combined with or 30. Lovelock DM, Zhang Z, Jackson A, et al. Correlation of local fail-
without surgery for primary versus recurrent tumor. Int J Radiat ure with measures of dose insufficiency in the high-dose single-
Oncol Biol Phys. 2006;65(5):1514–21. fraction treatment of bony metastases. Int J Radiat Oncol Biol Phys.
12. Jiang B, Veeravagu A, Lee M, et al. Management of intracranial 2009;77(4):1282–7.
and extracranial chordomas with CyberKnife stereotactic radiosur- 31. Maranzano E, Latini P. Effectiveness of radiation therapy without
gery. J Clin Neurosci. 2012;19(8):1101–6. surgery in metastatic spinal cord compression: final results from a
13. Henderson FC, McCool K, Seigle J, Jean W, Harter W, Gagnon GJ. prospective trial. Int J Radiat Oncol Biol Phys. 1995;32(4):
Treatment of chordomas with CyberKnife: Georgetown University 959–67.
experience and treatment recommendations. Neurosurgery. 2009; 32. Rock JP, Ryu S, Shukairy MS, et al. Postoperative radiosurgery for
64(2 Suppl):A44–53. malignant spinal tumors. Neurosurgery. 2006;58(5):891–8; discus-
14. Wu JS, Wong R, Johnston M, Bezjak A, Whelan T. Meta-analysis sion 891–898.
of dose-fractionation radiotherapy trials for the palliation of painful 33. Moulding HD, Elder JB, Lis E, et al. Local disease control after
bone metastases. Int J Radiat Oncol Biol Phys. 2003;55(3): decompressive surgery and adjuvant high-dose single-fraction radio-
594–605. surgery for spine metastases. J Neurosurg Spine. 2010;13(1):87–93.
15. Harel R, Angelov L. Spine metastases: current treatments and 34. Mahan SL, Ramsey CR, Scaperoth DD, Chase DJ, Byrne TE.
future directions. Eur J Cancer. 2010;46(15):2696–707. Evaluation of image-guided helical tomotherapy for the retreatment of
16. Rades D, Stalpers LJ, Veninga T, Hoskin PJ. Spinal reirradiation spinal metastasis. Int J Radiat Oncol Biol Phys. 2005;63(5):1576–83.
after short-course RT for metastatic spinal cord compression. Int J 35. Milker-Zabel S, Zabel A, Thilmann C, Schlegel W, Wannenmacher
Radiat Oncol Biol Phys. 2005;63(3):872–5. M, Debus J. Clinical results of retreatment of vertebral bone metas-
17. Gibbs IC, Kamnerdsupaphon P, Ryu MR, et al. Image-guided tases by stereotactic conformal radiotherapy and intensity-modulated
robotic radiosurgery for spinal metastases. Radiother Oncol. radiotherapy. Int J Radiat Oncol Biol Phys. 2003;55(1):162–7.
2007;82(2):185–90. 36. Sahgal A, Ames C, Chou D, et al. Stereotactic body radiotherapy is
18. Guckenberger M, Goebel J, Wilbert J, et al. Clinical outcome of effective salvage therapy for patients with prior radiation of spinal
dose-escalated image-guided radiotherapy for spinal metastases. metastases. Int J Radiat Oncol Biol Phys. 2009;74(3):723–31.
Int J Radiat Oncol Biol Phys. 2009;75(3):828–35. 37. Wright JL, Lovelock DM, Bilsky MH, Toner S, Zatcky J, Yamada
19. Ryu S, Jin R, Jin JY, et al. Pain control by image-guided radiosur- Y. Clinical outcomes after reirradiation of paraspinal tumors. Am J
gery for solitary spinal metastasis. J Pain Symptom Manage. Clin Oncol. 2006;29(5):495–502.
2008;35(3):292–8. 38. Damast S, Wright J, Bilsky M, et al. Impact of dose on local failure
20. Hartsell WF, Scott CB, Bruner DW, et al. Randomized trial of rates after image-guided reirradiation of recurrent paraspinal
short- versus long-course radiotherapy for palliation of painful metastases. Int J Radiat Oncol Biol Phys. 2010;81(3):819–26.
bone metastases. J Natl Cancer Inst. 2005;97(11):798–804. 39. Chang EL, Shiu AS, Mendel E, et al. Phase I/II study of stereotactic
21. Chow E, Harris K, Fan G, Tsao M, Sze WM. Palliative radiotherapy body radiotherapy for spinal metastasis and its pattern of failure.
trials for bone metastases: a systematic review. J Clin Oncol. J Neurosurg Spine. 2007;7(2):151–60.
2007;25(11):1423–36. 40. Garg AK, Wang XS, Shiu AS, et al. Prospective evaluation of spinal
22. Katagiri H, Takahashi M, Inagaki J, et al. Clinical results of nonsur- reirradiation by using stereotactic body radiation therapy: The
gical treatment for spinal metastases. Int J Radiat Oncol Biol Phys. University of Texas MD Anderson Cancer Center experience.
1998;42(5):1127–32. Cancer. 2011;117(15):3509–16.
23. Maranzano E, Bellavita R, Rossi R, et al. Short-course versus split- 41. DeLaney TF, Chen GT, Mauceri TC, et al. Intraoperative dural irra-
course radiotherapy in metastatic spinal cord compression: results diation by customized 192iridium and 90yttrium brachytherapy
of a phase III, randomized, multicenter trial. J Clin Oncol. plaques. Int J Radiat Oncol Biol Phys. 2003;57(1):239–45.
2005;23(15):3358–65. 42. Folkert MR, Bilsky MH, Cohen GN, et al. Intraoperative 32P high-
24. Gerszten PC, Mendel E, Yamada Y. Radiotherapy and radiosurgery dose rate brachytherapy of the dura for recurrent primary and meta-
for metastatic spine disease: what are the options, indications, and static intracranial and spinal tumors. Neurosurgery. 2012;71(5):
outcomes? Spine (Phila Pa 1976). 2009;34(22 Suppl):S78–92. 1003–11.
Spinal Tumors: Viewpoint—Surgery
45
Faiz U. Ahmad, Gabriel Zada, and Michael Y. Wang
which surgery remains the optimal treatment for patients with

Introduction spinal tumors [10–14]. Although surgery has not been widely
accepted as a primary treatment for metastatic spine disease
The majority of spine tumors encountered by clinicians are in the past, a shift in this ideology is currently underway,
metastatic, accounting for approximately 70 % of all spine with many physicians advocating more aggressive surgical
tumors [1]. Each year, an estimated 18,000 patients will be resection of metastatic spine lesions [3, 11, 15]. Recent
diagnosed with metastatic spine disease [2, 3]. Metastatic series have demonstrated that surgery can increase the over-
lesions most frequently originate from primary tumors in the all quality of life by reducing pain and maintaining neuro-
breast, lung, and prostate. Less frequently encountered neo- logic function as well as prolong survival in particular
plasms of the spine and spinal cord include primary lesions patients with metastatic spine disease [10, 12, 15–17].
such as meningiomas, schwannomas, osseous tumors of the Careful patient selection is the key to maximizing the poten-
spine, ependymomas, and gliomas. At the time of death, tial benefits of surgical intervention. This chapter will
approximately 70–90 % of all terminal cancer patients have describe the current role of surgery in the treatment of spinal
evidence of metastatic disease on postmortem examination tumors and the specific factors that make surgery the pre-
[4–6], with the spinal column being the most common site ferred treatment in a given situation.
for bony metastases [7]. Furthermore, it is likely that the
observed incidence of metastatic spine disease will continue
to increase over the next several decades, as diagnostic capa- Tumor Location
bilities improve [8, 9].
In the current era, a number of treatment modalities are Tumors of the spine are frequently characterized according
available for treating spinal tumors, including medical ther- to their location with respect to the spinal canal, meninges,
apy, chemotherapy, surgery, radiation therapy, and radiosur- and spinal cord. The overwhelming majority (95 %) of spine
gery. Quite frequently, combinations of these therapies are metastases are extradural, while the remaining are intradural
utilized to maximize clinical benefits and outcomes. Despite extramedullary (4 %) and intramedullary (0.5 %) [18].
the benefit of minimal invasiveness inherent to radiation Furthermore, metastases are most commonly found in the
therapy or radiosurgery, there are numerous situations in posterior half of the vertebral body, usually at the junction of
the vertebral body and pedicles due to the increased blood
supply to this area [19, 20]. They are infrequently confined to
F.U. Ahmad, M.D., M.Ch. the posterior spinal elements (i.e., lamina and spinous pro-
Department of Neurosurgery, Emory University, cess). The thoracic spine, with its numerous segments and a
80, Jesse Hill Dr SE, Atlanta, GA 30303, USA
e-mail: drfaizns@gmail.com narrower canal diameter relative to the spinal cord [21], is
the most common spinal region to present with neural com-
G. Zada, M.D.
Department of Neurosurgery, Keck School of Medicine, pression secondary to spinal metastases (70 % of patients),
University of Southern California, 1520 San Pablo Street, followed by the lumbar spine (20 %) and cervical spine
Suite 3800, Los Angeles, CA 90033, USA (10 %) [10, 22, 23]. The thoracolumbar junction and the T4
M.Y. Wang, M.D., F.A.C.S. (*) vertebral body, in particular, are the most common spinal
Department of Neurosurgery, Miller School of Medicine, levels to present with metastases [22, 23]. Unfortunately spi-
University of Miami, 1095 NW 14Tr, Lois Pope Center, nal metastases are located in multiple, noncontiguous levels
2nd floor, Miami, FL 33136, USA
e-mail: Mwang2@med.miami.edu in 10–38 % of cases at the time of presentation [24–26].
572 F.U. Ahmad et al.
Multiple surgical staging systems have been devised in surgery was also significantly higher in the surgery group
the past to assist in surgical planning based on anatomical (62 % versus 19 %, p = 0.01). Finally, patients treated with
criteria. The surgical considerations based on spinal level surgery retained the ability to walk for much longer than did
and relationship to the spinal cord will be discussed later in patients in the radiotherapy group (mean 122 versus 13 days,
this chapter. p = 0.001). The study was terminated prematurely secondary
to an overwhelmingly beneficial outcome observed in the
surgical group at interim analysis [33].
Deciding on a Treatment Modality Furthermore, a recent meta-analysis by Klimo and col-
leagues, reviewing 24 surgical series (999 patients) and 4
Following a diagnosis of metastatic spine disease, a decision large irradiation series (543 patients), demonstrated ambula-
of surgery, radiation, medical treatment, and/or palliative tory success rates in 85 % of surgical patients compared
care must be made. The patient’s expected survival time is a with 64 % of radiation patients. Surgical patients in this
crucial factor to consider in cancer patients, as many may be study were 1.3 times as likely to remain ambulatory after
terminal patients in which surgery may not provide an over- surgery and twice as likely to regain ambulation when lost
all quality of life benefit. As with any type of surgery, the preoperatively (rescue procedure) [34]. Thus, in general,
age, performance status, and comorbidities of the patient results from recent series of patients undergoing surgical
must be taken into account when considering operative man- intervention for metastatic spinal disease have been promising.
agement for metastatic spine disease. Comorbidities such as Several series have also reported that greater than 90 %
coagulopathy, anemia, malnutrition, liver disease, and renal of patients have benefited from surgery in terms of pain
disease are frequently encountered in cancer patients and relief as well as maintaining ambulatory status [11, 13, 15].
may complicate the surgery or postoperative course. The pri- Furthermore, survival benefits have been reported with
mary tumor type should be also factored into this decision, as extensive resections in patients with solitary bone lesions or
some primary tumors (such as renal and thyroid cancer) may oligometastatic disease, with some 5-year survival rates
require preoperative embolization [8]. being greater than 15 % [15, 35]. Despite recent data sup-
Radiation therapy was considered the mainstay of treat- porting surgical decompression as a primary treatment for
ment for metastatic spinal disease for several years, and sur- spinal metastases, radiation therapy remains a primary treat-
gery was considered to add little value or even deemed ment option, particularly for radiosensitive tumors such as
detrimental for these patients [27, 28]. This was primarily breast, lymphoma, multiple myeloma, small cell lung cancer,
because posterior laminectomy (the primary surgical option seminoma, neuroblastoma, and Ewing sarcoma. In addition,
used by neurosurgeons to treat cord compression) did not limited anticipated patient survival time, inability to tolerate
address the compression (which is usually anterior to the spi- a surgical procedure, greater than 24–48 h of total neuro-
nal cord) directly and often led to increased spinal instability logical deficit, or multilevel/diffuse spinal involvement
by removing the stable/intact posterior elements in the pres- remain factors favoring radiation therapy [36].
ence of an already diseased/weak anterior column. Therefore
multiple older studies in the past failed to demonstrate a
functional or survival benefit in patients undergoing surgery Guidelines for Surgical Decision Making
and irradiation as compared with radiation therapy alone
[16, 24, 29, 30]. Over the past decade, however, surgical In the past, various scoring systems have been developed to
decompression of the spinal canal has regained favor because guide physicians in evaluating the surgical potential of a
of more contemporary series utilizing modern circumferen- given patient based on multiple patient and tumor criteria
tial surgical approaches for decompression and stabilization. (see Table 45.1) [37–40]. The complexity of these algorithms
Many series have demonstrated greater than 90 % mainte- is indicative of how many factors must be taken into consid-
nance in ambulatory status and pain improvement following eration for a given scenario. Such frameworks for decision
decompression [13, 17, 31, 32]. making offer useful guidelines to assist physicians with sur-
Perhaps the most convincing evidence came from a recent gical decision making, but cannot encompass the entire spec-
study by Patchell et al. This prospective, multi-institutional trum of individual factors that must be considered for each
study randomized patients with metastatic spinal cord com- patient situation.
pression to surgical decompression plus radiotherapy or
radiotherapy alone. In the surgical decompression group,
42 out of 50 (84 %) of patients were able to walk following Goals of Therapy
surgery, compared with 29 of 51 patients (57 %) in the radio-
therapy group (p = 0.01). The number of nonambulatory In deciding on the optimal treatment for a patient, the goal of
patients who specifically regained the ability to walk following therapy remains a key factor. Goals for surgery in patients
45 Spinal Tumors: Viewpoint—Surgery 573
Table 45.1 Classification algorithms in surgical decision making for spinal metastases
Study Criteria Recommendations
Harrington et al. (1986) [37] Five categories: Based on category:
1. No neurologic dysfunction (1) and (2): nonoperative management
2. Involvement of bone without collapse or (3): left to judgment of surgeon
instability
3. Neurologic dysfunction without bony (4) and (5): surgery indicated
destruction
4. Vertebral collapse with secondary
mechanical pain
5. Vertebral collapse or instability with major
neurological deficits
Tokuhashi et al. (1990) [38] 0–2 points for each of the following categories: Based on total points (0–12):
1. Primary tumor type ≤5: nonoperative management
2. Neurological grade of the lesion 6–8: left to judgment of surgeon
3. Physical condition of the patient ≥9: surgery indicated
4. Number of lesions
5. Presence of visceral lesions
6. Spinal sites
Tomita et al. (2001) [39] Based on three categories: Based on total points (2–10):
1. Grade of malignancy: ≤3: wide/marginal excision (long-term local control)
– Slow growth, 1 point 4–5: marginal/intralesional excision (middle-term control)
– Moderate growth, 2 points 6–7: surgery for short term palliation
– Rapid growth, 4 points ≥8: nonoperative supportive care
2. Visceral metastases
– No metastasis, 0 points
– Treatable, 2 points
– Untreatable, 4 points
3. Bone metastases
– Solitary/isolated, 1 point
– Multiple, 2 points
Fourney et al. (2005) [40] “MAPS” assessment: Considerations by category:
1. Method of resection 1. En bloc spondylectomy, piecemeal excision, palliative
decompression
2. Anatomy of spinal disease 2. Tumor level, location, staging
3. Patient fitness level 3. Comorbidities, irradiation, patient age
4. Stabilization 4. Anterior, posterior, combined
with spinal tumors include: (1) neural decompression, (2) dysfunction, however, surgery offers less hope for improvement
biomechanical stabilization, (3) curative resection, (4) con- and radiotherapy is usually the mainstay of treatment [36].
trol of intractable pain, and (5) diagnosis. Patients with progressive neurologic deterioration that is
refractory to radiation therapy should also be considered for
Neural Decompression surgical decompression [36]. Cord compression caused by
The most important indication for surgery remains neural bone or disk fragments in the spinal canal that may result
decompression. Ambulatory status is a key prognostic marker from pathological fractures should also undergo primary
for terminal patients with spinal disease [26]. Nonambulatory surgical decompression instead of radiation therapy [41].
cancer patients have a mean survival time of only 45 days, Recent prospective randomized trials of surgical decompres-
due to multiple causes [13]. Maintaining neurologic function sion plus radiotherapy versus radiotherapy alone showed a
in patients with spinal metastases may therefore prolong sur- quality of life benefit in the surgical arm that was statistically
vival time in addition to improving quality of life. significant as well as a trend towards an increase in survival
Patients with rapidly progressive or far-advanced paraple- benefit in the surgical arm [42, 43]. These data represent high
gia, especially for less than 24–48 h, require emergent surgi- quality evidence favoring surgical decompression over tradi-
cal intervention [8]. In patients with complete spinal cord tional conservative radiotherapy in patients with spinal meta-
dysfunction for greater than 24 h or those with bowel/bladder static disease.
Stabilization [15, 52]. In fact, surgery is the only treatment that has demon-
Maintaining spinal stability is another indication for surgical strated a potential for cure of spinal metastases, as opposed to
intervention in patients with metastatic spinal disease. Stability local control of tumor available with radiation therapy [35].
has been defined by Panjabi et al. as allowing the degree of Procedures such as en bloc spondylectomy and vertebrec-
motion that prevents pain, neurological deficit, and abnormal tomy have become key developments in the ability to aggres-
spinal angulation [44]. Spinal instability in patients with sively treat and even potentially cure localized metastatic
metastases can result from related pathologic fractures, surgi- spinal disease. Patients with bony metastases as opposed to
cal intervention, or diffuse changes in bone density and visceral metastases, and those who present following a long
strength. Instability has been correlated with the following disease-free interval, are considered to have a low tumor bur-
findings in patients with spine tumors: two-column injury, den and may be optimal candidates for curative resection
vertebral body collapse greater than 50 %, kyphosis greater than [15, 35, 39, 52, 53]. Such patients may benefit from more
20–30°, or involvement of the same column in greater than two aggressive surgical management in terms of pain control and
levels [40, 45, 46]. Following most cases of surgical decom- quality of life and should be treated with the intent to cure
pression, especially in cancer patients, instrumentation is [35, 52]. The potential for a curative resection also depends
required to provide stability. In non-terminal patients, postop- on the source of the primary tumor. For instance, tumors of
erative stability is ideally achieved via a solid bone fusion fol- renal origin should be treated with the intent to cure [54].
lowing resection. In cancer patients, however, a bony fusion Furthermore, local malignancies such as chordomas, sarco-
may not be a realistic goal due to multiple potentially compli- mas, chondrosarcomas, Pancoast tumors, giant cell tumors,
cating factors. Solid bony fusion may be prevented by shorter and osteoblastomas should also be considered for curative en
patient survival times, anemia, malnutrition, or frequent bloc resection [13].
requirements for further chemotherapy, steroids, and radiation
therapy. Although the ability to study spinal fusion in cancer Pain Control
patients has been compromised by decreased follow-up times, Pain from metastatic lesions can be a result of inflammation
one study studying fusion rates in 25 patients undergoing ante- and mechanical forces, as well as direct nerve compression
rior decompression and interbody fusion with bone struts fol- or damage [31]. Radicular symptoms have been reported in
lowed by irradiation reported a relatively low pseudoarthrosis up to 90 % of patients with lumbar epidural cord compres-
rate of 16 % at 1 year following surgery [47]. sion, 79 % of patients with cervical disease, and 55 % of
Several options are currently available for spinal instru- patients with thoracic metastases [24]. Corticosteroids and
mentation in cancer patients. Rigid spinal fixation systems opioids are the primary medications available for pain man-
are often used to provide spinal stability [13, 48]. With ante- agement secondary to cord or root compression. Radiation
rior approaches, constructs are often created to replace the therapy remains the primary intervention reserved for
resected vertebral body. Cortical endplate bone integrity patients with terminal disease who are experiencing signifi-
must be preserved in order to stabilize anteriorly, however, cant pain. However, the beneficial effects of irradiation in
and loss of integrity may necessitate a posterior fusion or relieving pain may take up to 2 weeks and may not provide
even a combined anterior–posterior approach in cases of sig- adequate pain relief [55]. In cases where pain is refractory to
nificant instability [8]. In posterior approaches, 2–3 levels of radiation therapy, surgery remains a viable option. Many sur-
construct are generally used above and below the deficit gical procedures offer a significant benefit in pain relief, with
[13, 48]. Polymethylmethacrylate (PMMA) is frequently used recent series reporting pain relief success rates of greater
for reconstruction in conjunction with mechanical instru- than 90 % [10, 15, 56]. Percutaneous vertebroplasty and
mentation if the estimated patient survival time is estimated kyphoplasty are minimally invasive procedures that have
to be less than 6 months. PMMA may be inserted inside of a been shown to improve pain in patients who may not be can-
titanium cage or chest tube and can provide immediate sta- didates for more aggressive decompressive surgery. Several
bility [48, 49]. It has been shown to be a safe option for series demonstrate both short- and long-term pain relief rates
patients requiring subsequent radiation therapy [50] and is an of 85–100 % associated with percutaneous vertebroplasty
inexpensive, relatively easy to use, and safe alternative to and kyphoplasty [57–60].
bony fusion [51].
Diagnosis
Curative Resection and Local Control Since the advent of minimally invasive techniques for tumor
While a curative surgical resection of spinal metastases was biopsy of the spine, such as CT-guided and fluoroscopic-
not considered a reasonable goal in patients with metastatic guided biopsy, diagnosis is usually no longer a primary indi-
spinal disease for many years, recent series have demon- cation for surgical intervention. One study reported that a
strated that survival times can be significantly prolonged in tissue diagnosis was possible in 86 % of patients using
particular patients who have undergone surgical intervention CT-guided biopsy [61].
Timing of Combined Therapy for the primary treatment of spinal metastases. Class I evidence
from a case-control study comparing decompressive lami-
In spite of having recent Class I evidence showing the benefit nectomy plus radiation therapy with radiation therapy alone
of surgery in the primary treatment for spinal metastases, radi- did not show a significant functional benefit in ambulatory
ation therapy remains the primary modality for treatment at status following surgical intervention [30]. Class II and III
many medical institutions. Following a radiologic diagnosis evidence also failed to demonstrate significant functional
of spinal metastases, many patients are directly referred for benefits or improvement in pain [16, 24, 29, 65]. The inability
radiation therapy without a surgeon seeing the patient [62]. of posterior procedures to provide adequate decompression
The results of this are twofold. First, opting for radiation as a in any cases has been largely attributed to the majority of
primary measure may not provide the same survival benefit metastatic lesions being located anterior to the thecal sac.
for the patient as a wide marginal or radical excision could Retraction of the thecal sac would thus be necessary to com-
potentially provide. Second, patients who fail radiotherapy plete a gross total resection from a posterior approach in the
and ultimately require surgical decompression have signifi- majority of patients with metastases. In addition, the risk of
cantly higher chances of developing wound infections, espe- spinal instability is especially concerning following laminec-
cially in cases where irradiation was administered within 1 tomy in cancer patients with preexisting disease in the ante-
week prior to surgery [62]. One series of 85 patients undergo- rior and middle spinal columns [10, 32]. Decompressive
ing posterior decompression with or without instrumentation laminectomy is now primarily reserved for metastatic lesions
determined that patients with prior irradiation had a 32 % located within the posterior spinal elements or in cases where
chance of developing wound infections, compared with a patients cannot tolerate a more complex circumferential
12 % chance in the de novo surgery group. Multiple other approach. In contrast, posterior laminectomies remain the
studies have demonstrated similar results [34, 35, 43, 62, 63]. procedure of choice for the majority of intradural extramed-
For these reasons, many surgeons currently advocate consid- ullary and intramedullary primary (nonmetastatic) lesions.
ering de novo surgery as the primary intervention for symp- Some surgeons maintain that even patients with ventrally
tomatic metastatic spinal disease and have emphasized the located neoplasms can be adequately and safely decom-
spine surgeon’s involvement in the decision to irradiate patients. pressed with posterior laminectomy if internal fixation is used
in conjunction. This practice has been reported to realign the
spine as well as distract the vertebral bodies out of the spinal
Deciding on a Surgical Approach canal [66]. This procedure remains a useful option in patients
who are unable to undergo more extensive anterior approaches
Once surgery has been decided upon, choosing the correct or for patients with multilevel disease [67].
surgical approach is a key factor in providing the ideal degree
of exposure for resection and subsequent stabilization. This Newer Approaches for Spinal Cord
decision is based on the anatomical location of the spinal Decompression
metastases, the degree of surgical intervention a patient is Multiple studies have been published describing the efficacy
able to tolerate, and the ability of the surgeon to perform a of various circumferential approaches for pain relief and pres-
particular procedure safely. The Weinstein, Boriani, and ervation or improvement of neurologic function [10–12, 15,
Biagini staging system can be used to help guide surgeons in 32, 45, 68–70]. Overall, the proportion of ambulatory patients
choosing an approach for surgical decompression based on who retained ambulatory function postoperatively (defined as
the anatomical location of the tumor in regards to proximity the “success” of such a procedure) has ranged between 72 and
to the dura as well as anterior to posterior localization [64]. 98 %, but has been reported as lower when looking at only
This system maps tumors based on 12 radiating zones in a posterior approaches [71]. The proportions of nonambulatory
clockwise fashion, as well as five concentric layers. It can patients who regained ambulatory status postoperatively
assist in deciding on a type of excisional procedure, includ- (defined as a “rescue” procedure) have been less consistent,
ing en bloc spondylectomy, vertebrectomy, sagittal resec- ranging between 0 and 94 % in these series [71]. Finally,
tion, and posterior arch resection. surgical morbidity rates for circumferential decompressions
have ranged between 7 and 65 %, and mortality rates have
Decompressive Posterior Laminectomy ranged between 0 and 31 % [71]. The majority of series report
In the past, the primary surgical procedure utilized by neuro- mortality rates of less than 10 %.
surgeons to relieve compression secondary to spinal metasta- Approaches to the spine can be broadly classified as ante-
ses was the posterior laminectomy. This procedure was rior or posterior. Anterior approaches include traditional ante-
initially favorable because it could be performed by the rior cervical approaches, transthoracic, and transperitoneal
majority of neurosurgeons within relatively fast time frames. approaches. Posterior approaches include traditional laminec-
For several reasons, however, this procedure fell out of favor tomy and posterolateral approaches such as transpedicular,
costotransversectomy, and lateral extracavitary (LECA) The transpedicular approach can be used for disease in
approaches. All of these approaches require subsequent the dorsal and anterior elements of the vertebrae. The advan-
reconstruction and stabilization. tages of using this approach over an anterior approach
include early identification of the spinal canal, ability to
Anterior Approaches resect tumor in the posterior elements, and the ability to pro-
In general, an anterior approach is preferable for achieving vide long segment fixation [69, 75].
adequate decompression in patients who are able to tolerate A costotransversectomy allows resection of the lamina,
such an approach because spinal metastases tend to localize facet, transverse process, pedicle, rib head, and some of the
to the anterior spinal column. In general, series of anterior vertebral body to provide adequate exposure. Advantages of
procedures demonstrate superior rates of neurologic function this procedure include avoidance of a thoracotomy in patients
and survival times than do posterior procedures [10, 13, 66]. who may not be able to tolerate one, access to multilevel or
In addition to decompression, reconstruction of the vertebral discontinuous spine disease, and the ability to perform both
body following resection is also facilitated by an anterior anterior and posterior fixation. Disadvantages with this pro-
approach. cedure include a possibly limited or indirect exposure, pos-
In the cervical spine, reconstruction following an exten- sible entry into the lung space requiring chest tube placement,
sive resection is accomplished using bone autograft or and possible injury to the nerve roots or dura resulting in a
allograft, PMMA, or one of many titanium cages or spacers CSF fistula.
[72]. Stabilization is then usually achieved by way of ante- The LECA approach can be used for spinal decompres-
rior instrumentation, with plating being the most common sion and fixation in the thoracolumbar spine (T4–L3) and
technique in the cervical spine. In the thoracic spine, a trans- provides good lateral exposure to the vertebral column with-
thoracic approach can be utilized for anterior decompression out violation of the pleural or abdominal space or mobiliza-
and subsequent instrumentation. The side of approach is tion of the diaphragm [76]. Like the costotransversectomy,
based on the symmetry of disease, but the left side is often both anterior and posterior fixations are possible with the
preferable in order to avoid the liver and to more easily iden- LECA. It is commonly used in patients requiring near com-
tify the aorta. Rib resection, lung deflation, and chest tube plete resection of 1–3 vertebral bodies. Reconstruction of the
insertion are hallmarks of this procedure at the thoracic level vertebral body is then performed followed by posterior with
[41]. An anterior approach at the thoracolumbar junction or without anterior fixation [76].
(T11–L1) can be done via a thoracoabdominal approach. At
this level, a portion of the diaphragm requires mobilization Combined Anterior and Posterior Approaches
in order to achieve adequate exposure. At the lumbar level, While providing the benefit of excellent exposure to the rele-
dissection via the retroperitoneum is required, usually along vant elements of the spinal canal, anterior decompression also
the transversalis fascia. The psoas muscles and aorta are carries the potential drawback of decreased spinal stability.
mobilized to provide exposure to the lumbar vertebrae [41]. Several series report using anterior decompression with sub-
sequent posterior instrumentation to further stabilize the spi-
Posterolateral Approaches nal column [12, 32]. Particular indications for this combination
Lesions located in the posterior spinal elements should be approach include involvement of metastatic disease in all
approached in this fashion. Patients who may not be able to three spinal columns, significant instability, marked kyphosis,
tolerate a more extensive anterior approach, or those with involvement of more than one vertebral body, junctional
multilevel disease, may benefit from posterior decompres- site involvement, and prior laminectomy [12, 32]. One study
sion with stabilization. Finally, particular regional nuances including 110 patients undergoing spinal decompression
may contribute to the decision to approach posteriorly. The determined that approximately half (48 %) of all patients
disadvantages of a posterior approach include limited ante- undergoing decompressive surgery required a combined ante-
rior exposure with limited tumor resection and higher rates rior–posterior approach for added stability [12].
of wound complications [48].
The cervicothoracic junction (C7–T4) marks a transition
zone from the mobile, lordotic cervical spine to a rigid, En Bloc Spondylectomy
kyphotic thoracic spine. The hardware failure rates with sur-
gical decompression and subsequent instrumentation in this In recent years, aggressive surgical resection of spinal metas-
region are comparatively high, with two series reporting sta- tases has made curative resection of spinal metastatic lesions
bilization failure rates of 36 % [73, 74]. Le et al. advocate a a possible outcome. As mentioned, patients with solitary
posterior/posterolateral approach to the cervicothoracic bone metastases or oligometastatic disease can be treated
junction because of lower incidences of hardware failure. with intent to cure if the patient is a surgical candidate for an
aggressive surgical procedure. Patient selection is especially less than one-third of its original height, coagulopathies,
crucial when evaluating patients for an en bloc spondylec- inability to lie prone, and inability to perform an emergent
tomy, as this procedure tends to be long and complex with decompressive surgery if necessary [58, 59, 79].
relatively significant blood loss and a frequent requirement Vertebroplasty has been shown to provide its benefit by
for an open thoracotomy, significant reconstruction and sac- restoring strength and stiffness to the spine and preventing
rifice of neural and/or vascular structures [35]. Surgical plan- micro-motion [80, 81]. Additionally, PMMA has been shown
ning in terms of approach and degree of resection has been to destroy nerve endings that cause pain associated with
greatly emphasized, with some surgeons favoring the use of micromotion [58, 79, 82]. It has also been postulated that
a staging system such as the Boriani/Weinstein system in PMMA may also provide an antitumor effect secondary to
planning this type of procedure. thermal effects, cytotoxicity, and ischemia that prevents
Results from recent series of en bloc resections show tumor reinvasion [58, 82]. A biopsy can be done during the
promising results. One such study reported a mortality rate of same procedure prior to the injection of the filling agent.
less than 1 %, a morbidity rate of less than 10 %, and a mean In a series of 101 patients, Deramond reported an improve-
survival time of greater than 3 years with 48 % of patients ment in pain and quality of life in 80 % of patients [58, 82].
showing neurological improvement [35]. Pain improved Patients following vertebroplasty have experienced short-
following surgery in 95 % of patients following surgery, term (within 48 h) and long-term pain relief (at 3 months
with 76 % achieving total pain control. Another recent series postoperatively) rates of 85–97 % and 89–100 %, respec-
of patients undergoing en bloc spondylectomy demonstrated tively [57, 59]. At 6 months following vertebroplasty, pain
morbidity and mortality rates less than 1 % [15]. relief was still reported by 76 % of patients, the majority of
Once an en bloc spondylectomy or vertebrectomy has which developed either new metastases or epidural involve-
been performed, vertebral body replacement or reconstruc- ment of metastases [82].
tion is necessary to provide stabilization. There have been Despite its many benefits and a low degree of invasive-
many materials and implants used to replace the vertebral ness, vertebroplasty is not without complications. The over-
body following vertebrectomy that have been described in all complication rate has been reported as approximately
the literature, including allograft, bone cement, ceramic, 10 %, with the majority being short-term complications and
ceramic/glass, carbon fiber, and titanium spacers or cages only 1.7 % resulting in long-term complications [58].
[77]. Recently, expandable cages have been developed for Leakage of PMMA has been reported to occur in approxi-
replacing the vertebral body and disks. Such cages may offer mately 70–75 % of cases. In the majority this does not pres-
a benefit because they can be placed through smaller access ent a clinical problem [58, 78]. However significant clinical
site and expanded in situ to restore anterior column height. risks are increased with posterior cortical wall destruction or
epidural involvement. The major clinical risk of this proce-
Percutaneous Vertebroplasty and Kyphoplasty dure is radiculopathy secondary to PMMA leakage into the
Vertebroplasty is a relatively noninvasive surgical procedure neuroforamen or back pain caused by a local inflammatory
that has been increasingly utilized in recent years to treat reaction to PMMA [58, 81].
pain and instability in vertebral body collapse by injecting In summary, vertebroplasty and kyphoplasty are mini-
cement into the vertebral bodies through a posterior approach. mally invasive procedures that may be performed for selected
The procedure is often used for patients with severe, local- patients with metastatic spine disease in order to provide fast
ized mechanical back pain without epidural involvement pain relief and improved spinal stability. It is often the proce-
who may not be able to tolerate more aggressive procedures dure of choice as a palliating measure for patients with lim-
[78]. Kyphoplasty is a related procedure in which an inflat- ited survival or those who are otherwise unable to tolerate
able balloon is used initially to restore the loss of vertebral surgical intervention.
height and create a cavity for subsequent injection of cement.
Because vertebroplasty and kyphoplasty are relatively
quick percutaneous procedures, they can be used in cancer Conclusion and Considerations
patients with limited survival time and poor surgical poten-
tial [78]. It has been shown that vertebroplasty can be used In summary, surgical management remains a viable and
prior to the initiation of radiation therapy, without compro- increasingly efficacious option for the management of spinal
mising its efficacy [50]. Many series have advocated this tumors. The primary goals for surgery are neural decompres-
combination of vertebroplasty to confer stability followed by sion, spinal stabilization, curative resection, and pain man-
radiation therapy for its tumoricidal effects [58, 59, 79]. agement. Curative resection of solitary or localized metastatic
These procedures should not be used in patients with gross disease is now a potential goal with aggressive surgical meth-
spinal instability, cord compression, or epidural extension ods. Careful patient selection is the key factor for maximizing
that are able to undergo surgical intervention. Further contra- the potential benefits and avoiding the associated risks of
indications for these procedures include vertebral collapse to any treatment modality, be it surgery or radiotherapy.
Case 45.1
A 54-year-old female developed 1 week of progressive lower extremity weakness and attendant bowel and bladder
dysfunction. She was wheelchair-bound at the time of presentation, and a CT scan showed a pathologic fracture at T7
(Fig. 45.1a). Her past medical history was significant for invasive ductal carcinoma treated 10 years prior and in remission.
An MRI demonstrated this to be consistent with a metastatic lesion with a soft tissue component compressing upon the
spinal cord (Fig. 45.1b), and SPECT studies showed no evidence of other lesions (Fig. 45.1c). The patient underwent
an urgent T7 corpectomy via a transthoracic approach. The vertebral body was replaced by a PEEK vertebrectomy
spacer packed with rib autograft and supplemental plating was utilized (Fig. 45.1d, e). She went on to have a complete
neurologic recovery, ambulating without assistance 5 days after surgery.
Fig. 45.1 (a–e) Plain CT scan showing a pathologic fracture at T7 (a). MRI thoracic spine T2WI revealed the cord compression (b), and
SPECT studies showed no evidence of other lesions (c). The patient underwent an urgent T7 corpectomy via a transthoracic approach. The
vertebral body was replaced by a PEEK vertebrectomy spacer packed with rib autograft and supplemental plating was utilized. Postoperative
plain X-ray lateral (d) and AP view (e) showing the instrumentation
Case 45.2
A 54-year-old male with prostate cancer presented with worsening upper back pain, progressive inability to walk, and
signs of myelopathy for 10 days prior to presentation to the emergency room. He was found to have multiple spinal
lesions but only the lesion at T3 was causing spinal cord compression. MRI demonstrated diffusely abnormal signal
intensity in T3 body with extension to right pedicle and lamina and epidural enhancement with spinal cord compres-
sion (Fig. 45.2a–c). There was no signal change in the spinal cord itself. CT scan demonstrated multiple lytic and
blastic lesions in the thoracic spine with loss of height of T3 body and epidural breakthrough of soft tissue mass into
the spinal canal (Fig. 45.2d). He underwent T3 vertebrectomy with reconstruction of anterior spinal column with a
carbon fiber cage, along with supplemental posterior fusion from T2–T6 without complications (Fig. 45.2e).
Postoperatively his motor strength returned to normal and he was discharged home after a brief stay at the rehab center.
He subsequently underwent spinal radiation and hormonal therapy and was doing well 1 year after the surgery.
Fig. 45.2 (a–e) MRI demonstrating a diffusely abnormal signal intensity in T3 body with extension to right pedicle and lamina and epi-
dural enhancement with spinal cord compression on axial (a) and sagittal (b–d) images. There was no signal change in the spinal cord
itself. CT scan demonstrated multiple lytic and blastic lesions in the thoracic spine with loss of height of T3 body and epidural break-
through of soft tissue mass into the spinal canal (e). The patient underwent T3 vertebrectomy with reconstruction of anterior spinal column
with a carbon fiber cage, along with supplemental posterior fusion from T2–T6, as shown on postoperative AP X-ray (f)
25. O’Rourke T, George CB, Redmond 3rd J, Davidson H, Cornett P, Fill

WL, et al. Spinal computed tomography and computed tomographic
References metrizamide myelography in the early diagnosis of metastatic disease.
J Clin Oncol. 1986;4(4):576–83.
1. Posner JB. Back pain and epidural spinal cord compression. Med 26. Ruff RL, Lanska DJ. Epidural metastases in prospectively evalu-
Clin North Am. 1987;71(2):185–205. ated veterans with cancer and back pain. Cancer. 1989;63(11):
2. Perrin RG, Laxton AW. Metastatic spine disease: epidemiology, 2234–41.
pathophysiology, and evaluation of patients. Neurosurg Clin N Am. 27. Ippolito V, Micheletti E, Saccalani M, Barbera F, Tonoli S, Motta C.
2004;15(4):365–73. Radiotherapy and spinal brace: still first-choice treatment for ver-
3. Wu AS, Fourney DR. Evolution of treatment for metastatic spine tebral metastases from breast cancer. Chir Organi Mov. 1998;
disease. Neurosurg Clin N Am. 2004;15(4):401–11. 83(1–2):177–83.
4. Jaffe W. Tumors and tumorous conditions of the bones and joints. 28. Katagiri H, Takahashi M, Inagaki J, Kobayashi H, Sugiura H,
Philadelphia: Lea and Febiger; 1958. p. 589–618. Yamamura S, et al. Clinical results of nonsurgical treatment for spi-
5. Cobb 3rd CA, Leavens ME, Eckles N. Indications for nonoperative nal metastases. Int J Radiat Oncol Biol Phys. 1998;42(5):1127–32.
treatment of spinal cord compression due to breast cancer. 29. Constans JP, de Divitiis E, Donzelli R, Spaziante R, Meder JF, Haye
J Neurosurg. 1977;47(5):653–8. C. Spinal metastases with neurological manifestations. Review of
6. Wong DA, Fornasier VL, MacNab I. Spinal metastases: the obvious, 600 cases. J Neurosurg. 1983;59(1):111–8.
the occult, and the impostors. Spine. 1990;15(1):1–4. 30. Young RF, Post EM, King GA. Treatment of spinal epidural metas-
7. Bohm P, Huber J. The surgical treatment of bony metastases of the tases. Randomized prospective comparison of laminectomy and
spine and limbs. J Bone Joint Surg Br. 2002;84(4):521–9. radiotherapy. J Neurosurg. 1980;53(6):741–8.
8. Jacobs WB, Perrin RG. Evaluation and treatment of spinal metasta- 31. Weinstein SM, Walton O. Management of pain associated with spi-
ses: an overview. Neurosurg Focus. 2001;11(6):e10. nal tumor. Neurosurg Clin N Am. 2004;15(4):511–27.
9. Bailar 3rd JC, Gornik HL. Cancer undefeated. N Engl J Med. 32. Fourney DR, Abi-Said D, Rhines LD, Walsh GL, Lang FF,
1997;336(22):1569–74. McCutcheon IE, et al. Simultaneous anterior-posterior approach to
10. Gokaslan ZL, York JE, Walsh GL, McCutcheon IE, Lang FF, the thoracic and lumbar spine for the radical resection of tumors
Putnam Jr JB, et al. Transthoracic vertebrectomy for metastatic spi- followed by reconstruction and stabilization. J Neurosurg. 2001;
nal tumors. J Neurosurg. 1998;89(4):599–609. 94(2 Suppl):232–44.
11. Sundaresan N, Digiacinto GV, Hughes JE, Cafferty M, Vallejo 33. Patchell RTP, Regine W, Payne R, Saris S, Kryscio R, Mohiuddin
A. Treatment of neoplastic spinal cord compression: results of a M, Young B. Direct decompressive surgical resection in the treat-
prospective study. Neurosurgery. 1991;29(5):645–50. ment of spinal cord compression caused by metastatic cancer: a
12. Sundaresan N, Steinberger AA, Moore F, Sachdev VP, Krol G, Hough randomised trial. Lancet. 2005;366:643–8.
L, et al. Indications and results of combined anterior-posterior 34. Klimo Jr P, Thompson CJ, Kestle JR, Schmidt MH. A meta-analysis
approaches for spine tumor surgery. J Neurosurg. 1996;85(3):438–46. of surgery versus conventional radiotherapy for the treatment of
13. Vrionis FD, Small J. Surgical management of metastatic spinal metastatic spinal epidural disease. Neuro Oncol. 2005;7(1):64–76.
neoplasms. Neurosurg Focus. 2003;15(5):E12. 35. Yao KC, Boriani S, Gokaslan ZL, Sundaresan N. En bloc spondy-
14. Fourney DR, Abi-Said D, Lang FF, McCutcheon IE, Gokaslan ZL. lectomy for spinal metastases: a review of techniques. Neurosurg
Use of pedicle screw fixation in the management of malignant spi- Focus. 2003;15(5):E6.
nal disease: experience in 100 consecutive procedures. J Neurosurg. 36. Ryken TC, Eichholz KM, Gerszten PC, Welch WC, Gokaslan ZL,
2001;94(1 Suppl):25–37. Resnick DK. Evidence-based review of the surgical management of
15. Sundaresan N, Rothman A, Manhart K, Kelliher K. Surgery for vertebral column metastatic disease. Neurosurg Focus. 2003;
solitary metastases of the spine: rationale and results of treatment. 15(5):E11.
Spine. 2002;27(16):1802–6. 37. Harrington KD. Metastatic disease of the spine. J Bone Joint Surg
16. Sorensen S, Borgesen SE, Rohde K, Rasmusson B, Bach F, Boge- Am. 1986;68(7):1110–5.
Rasmussen T, et al. Metastatic epidural spinal cord compression. 38. Tokuhashi Y, Matsuzaki H, Toriyama S, Kawano H, Ohsaka
Results of treatment and survival. Cancer. 1990;65(7):1502–8. S. Scoring system for the preoperative evaluation of metastatic
17. Wai EK, Finkelstein JA, Tangente RP, Holden L, Chow E, Ford M, spine tumor prognosis. Spine. 1990;15(11):1110–3.
et al. Quality of life in surgical treatment of metastatic spine dis- 39. Tomita K, Kawahara N, Kobayashi T, Yoshida A, Murakami H,
ease. Spine. 2003;28(5):508–12. Akamaru T. Surgical strategy for spinal metastases. Spine.
18. Perrin RG, Livingston KE, Aarabi B. Intradural extramedullary spinal 2001;26(3):298–306.
metastasis. A report of 10 cases. J Neurosurg. 1982;56(6):835–7. 40. Fourney DR, Gokaslan ZL. Use of “MAPs” for determining the
19. Posner JB, Howieson J, Cvitkovic E. "Disappearing" spinal cord optimal surgical approach to metastatic disease of the thoracolum-
compression: oncolytic effect of glucocorticoids (and other chemo- bar spine: anterior, posterior, or combined. Invited submission from
therapeutic agents) on epidural metastases. Ann Neurol. 1977;2(5): the Joint Section Meeting on Disorders of the Spine and Peripheral
409–13. Nerves, March 2004. J Neurosurg Spine. 2005;2(1):40–9.
20. Byrne TN. Spinal cord compression from epidural metastases. 41. Fourney DR, Gokaslan ZL. Anterior approaches for thoracolumbar
N Engl J Med. 1992;327(9):614–9. metastatic spine tumors. Neurosurg Clin N Am. 2004;15(4):
21. Grant R, Papadopoulos SM, Greenberg HS. Metastatic epidural 443–51.
spinal cord compression. Neurol Clin. 1991;9(4):825–41. 42. Patchell R RW, Payne R, Saris S, Kryscio R, et al. A randomized
22. Nottebaert M, von Hochstetter AR, Exner GU, Schreiber A. clinical trial of direct decompressive surgical resection in the treat-
Metastatic carcinoma of the spine. A study of 92 cases. Int Orthop. ment of spinal cord compression caused by metastasis [abstract].
1987;11(4):345–8. Presented at the Annual ASCO Meeting, Chicago,. May/June. 2003.
23. Livingston KE, Perrin RG. The neurosurgical management of spinal 43. Regine WFYA, Payne R, Saris S, Kryscio RJ, et al. Metastatic spi-
metastases causing cord and cauda equina compression. J Neurosurg. nal cord compression: a randomized trial of direct decompressive
1978;49(6):839–43. surgical resection plus radiotherapy vs radiotherapy alone. Int J
24. Gilbert RW, Kim JH, Posner JB. Epidural spinal cord compression Radiat Oncol Biol Phys. 2003;57(2 Suppl):S125.
from metastatic tumor: diagnosis and treatment. Ann Neurol. 1978; 44. Panjabi MM, Thibodeau LL, Crisco 3rd JJ, White 3rd AA. What
3(1):40–51. constitutes spinal instability? Clin Neurosurg. 1988;34:313–39.
45. Kostuik JP, Errico TJ, Gleason TF, Errico CC. Spinal stabilization mass: its radiotherapeutic outcome on radiosensitivity. Int J Radiat
of vertebral column tumors. Spine. 1988;13(3):250–6. Oncol Biol Phys. 1993;27(5):1079–83.
46. Denis F. The three column spine and its significance in the classifi- 64. Boriani S, Weinstein JN, Biagini R. Primary bone tumors of the
cation of acute thoracolumbar spinal injuries. Spine. 1983;8(8): spine. Terminology and surgical staging. Spine. 1997;22(9):
817–31. 1036–44.
47. Emery SE, Hughes SS, Junglas WA, Herrington SJ, Pathria 65. Black P. Spinal metastasis: current status and recommended guide-
MN. The fate of anterior vertebral bone grafts in patients irradiated lines for management. Neurosurgery. 1979;5(6):726–46.
for neoplasm. Clin Orthop Relat Res. 1994;300:207–12. 66. Olerud C, Jonsson B. Surgical palliation of symptomatic spinal
48. Steinmetz MP, Mekhail A, Benzel EC. Management of metastatic metastases. Acta Orthop Scand. 1996;67(5):513–22.
tumors of the spine: strategies and operative indications. Neurosurg 67. Klimo Jr P, Schmidt MH. Surgical management of spinal metasta-
Focus. 2001;11(6):e2. ses. Oncologist. 2004;9(2):188–96.
49. Errico TJ, Cooper PR. A new method of thoracic and lumbar body 68. Harrington KD. Anterior decompression and stabilization of the
replacement for spinal tumors: technical note. Neurosurgery. spine as a treatment for vertebral collapse and spinal cord com-
1993;32(4):678–80; discussion 80–1. pression from metastatic malignancy. Clin Orthop Relat Res.
50. Murray JA, Bruels MC, Lindberg RD. Irradiation of polymethyl- 1988;233:177–97.
methacrylate. In vitro gamma radiation effect. J Bone Joint Surg 69. Bilsky MH, Boland P, Lis E, Raizer JJ, Healey JH. Single-stage
Am. 1974;56(2):311–2. posterolateral transpedicle approach for spondylectomy, epidural
51. Liu JK, Apfelbaum RI, Chiles 3rd BW, Schmidt MH. Cervical decompression, and circumferential fusion of spinal metastases.
spinal metastasis: anterior reconstruction and stabilization tech- Spine. 2000;25(17):2240–9; discussion 250.
niques after tumor resection. Neurosurg Focus. 2003;15(5):E2. 70. Wise JJ, Fischgrund JS, Herkowitz HN, Montgomery D, Kurz LT.
52. Singletary SE, Walsh G, Vauthey JN, Curley S, Sawaya R, Weber KL, Complication, survival rates, and risk factors of surgery for meta-
et al. A role for curative surgery in the treatment of selected patients static disease of the spine. Spine. 1999;24(18):1943–51.
with metastatic breast cancer. Oncologist. 2003;8(3):241–51. 71. Klimo Jr P, Kestle JR, Schmidt MH. Treatment of metastatic spinal
53. Hortobagyi GN. Can we cure limited metastatic breast cancer? epidural disease: a review of the literature. Neurosurg Focus.
J Clin Oncol. 2002;20(3):620–3. 2003;15(5):E1.
54. Boriani S, Biagini R, De Iure F, Bertoni F, Malaguti MC, Di Fiore M, 72. Liu JK, Apfelbaum RI, Schmidt MH. Surgical management of cer-
et al. En bloc resections of bone tumors of the thoracolumbar spine. vical spinal metastasis: anterior reconstruction and stabilization
A preliminary report on 29 patients. Spine. 1996;21(16):1927–31. techniques. Neurosurg Clin N Am. 2004;15(4):413–24.
55. Gilbert HA, Kagan AR, Nussbaum H, Rao AR, Satzman J, Chan P, 73. Le H, Balabhadra R, Park J, Kim D. Surgical treatment of tumors
et al. Evaluation of radiation therapy for bone metastases: pain relief involving the cervicothoracic junction. Neurosurg Focus. 2003;
and quality of life. AJR Am J Roentgenol. 1977;129(6):1095–6. 15(5):E3.
56. Hosono N, Yonenobu K, Fuji T, Ebara S, Yamashita K, Ono 74. Kraus DH, Huo J, Burt M. Surgical access to tumors of the cervico-
K. Vertebral body replacement with a ceramic prosthesis for meta- thoracic junction. Head Neck. 1995;17(2):131–6.
static spinal tumors. Spine. 1995;20(22):2454–62. 75. Cahill DW, Kumar R. Palliative subtotal vertebrectomy with ante-
57. Cortet B, Cotten A, Boutry N, Dewatre F, Flipo RM, Duquesnoy B, rior and posterior reconstruction via a single posterior approach.
et al. Percutaneous vertebroplasty in patients with osteolytic metas- J Neurosurg. 1999;90(1 Suppl):42–7.
tases or multiple myeloma. Rev Rhum Engl Ed. 1997;64(3): 76. Schmidt MH, Larson SJ, Maiman DJ. The lateral extracavitary
177–83. approach to the thoracic and lumbar spine. Neurosurg Clin N Am.
58. Deramond H, Depriester C, Galibert P, Le Gars D. Percutaneous 2004;15(4):437–41.
vertebroplasty with polymethylmethacrylate. Technique, indica- 77. Thongtrangan I, Balabhadra RS, Le H, Park J, Kim DH. Vertebral
tions, and results. Radiol Clin North Am. 1998;36(3):533–46. body replacement with an expandable cage for reconstruction after
59. Kaemmerlen P, Thiesse P, Bouvard H, Biron P, Mornex F, Jonas P. spinal tumor resection. Neurosurg Focus. 2003;15(5):E8.
[Percutaneous vertebroplasty in the treatment of metastases. 78. Pilitsis JG, Rengachary SS. The role of vertebroplaty inmetastatic
Technic and results]. J Radiol. 1989;70(10):557–62. Vertebroplastie spinal disease. Neurosurg Focus. 2001;11(6):e9.
percutanee dans le traitement des metastases. Technique et 79. Cotten A, Dewatre F, Cortet B, Assaker R, Leblond D, Duquesnoy
resultats. B, et al. Percutaneous vertebroplasty for osteolytic metastases and
60. Fourney DR, Schomer DF, Nader R, Chlan-Fourney J, Suki D, myeloma: effects of the percentage of lesion filling and the leakage
Ahrar K, et al. Percutaneous vertebroplasty and kyphoplasty for of methyl methacrylate at clinical follow-up. Radiology. 1996;
painful vertebral body fractures in cancer patients. J Neurosurg. 200(2):525–30.
2003;98(1 Suppl):21–30. 80. Liebschner MA, Rosenberg WS, Keaveny TM. Effects of bone
61. Ghelman B, Lospinuso MF, Levine DB, O’Leary PF, Burke SW. cement volume and distribution on vertebral stiffness after verte-
Percutaneous computed-tomography-guided biopsy of the thoracic broplasty. Spine. 2001;26(14):1547–54.
and lumbar spine. Spine. 1991;16(7):736–9. 81. Tohmeh AG, Mathis JM, Fenton DC, Levine AM, Belkoff
62. Ghogawala Z, Mansfield FL, Borges LF. Spinal radiation before SM. Biomechanical efficacy of unipedicular versus bipedicular ver-
surgical decompression adversely affects outcomes of surgery for tebroplasty for the management of osteoporotic compression frac-
symptomatic metastatic spinal cord compression. Spine. 2001; tures. Spine. 1999;24(17):1772–6.
26(7):818–24. 82. Weill A, Chiras J, Simon JM, Rose M, Sola-Martinez T, Enkaoua E.
63. Kim RY, Smith JW, Spencer SA, Meredith RF, Salter MM. Malignant Spinal metastases: indications for and results of percutaneous injec-
epidural spinal cord compression associated with a paravertebral tion of acrylic surgical cement. Radiology. 1996;199(1):241–7.
Spinal Metastases: Viewpoint—
Fractionated Radiation Therapy 46
Quynh-nhu Nguyen, Almon S. Shiu, and Eric L. Chang
spinal tumor while minimizing dose to the spinal cord.

Introduction Before SBRT to the spine can be considered a viable treat-
ment, clinical safety, which is predicated upon accurate and
Approximately 40 % of all cancer patients will develop precise treatment, should be demonstrated.
metastatic spinal disease [1]. Some patients may be satisfacto- While there is some ability to adjust or compensate for
rily treated with conventional radiotherapy for palliation of positional setup errors after hypofractionated SBRT, no such
pain and neurologic symptoms arising from spinal metastases. remedy exists to make corrections after single-session
For other patients, a conventional dose of 30–45 Gy may be SBRT. This statement has profound implications for new
inadequate to achieve durable control. Furthermore, reirradia- investigators choosing between fractionated and single-
tion of the spine with conventional means is rarely an option. session SBRT. Thus, at the inception of our SBRT program
The American Society for Therapeutic Radiology and at The University of Texas M. D. Anderson Cancer Center
Oncology (ASTRO) has described a new field of stereotactic (MDACC), we conservatively chose hypofractionated
body radiation therapy (SBRT) in conjunction with the SBRT. Our conservative philosophy recognizes the reality
American College of Radiology (ACR). SBRT is a “newly that a learning curve exists with any new procedure that must
emerging radiotherapy treatment method to deliver a high be climbed by the entire treatment team while meeting the
dose of radiation to the target, utilizing either a single dose or simultaneous requirement of safety. Single-fraction SBRT
small number of fractions with a high degree of precision should be deferred until the entire team has sufficient confi-
within the body. The ability to deliver a single or a few frac- dence in the entire procedure. The treatment plans represent-
tions of high-dose ionizing radiation with high targeting accu- ing SBRT to a typical T10 vertebral body target volume
racy and rapid dose falloff gradients encompassing tumors prescribed to 18 Gy in 1 fraction, 27 Gy in 3 fractions, and
within a patient provides the basis for the development of 30 Gy in 5 fractions while limiting the spinal cord to 8 Gy,
SBRT [2].” This perspective explores the respective merits of 9 Gy, and 10 Gy, respectively, are shown in Fig. 46.1.
single-dose and hypofractionated SBRT to the spine. The long-term patient-reported outcomes data collected
prospectively for hypofractionated SBRT at MDACC was
reported by Wang et al. [3] on 149 patients with mechani-
Fractionated Versus Single-Session SBRT cally stable spinal metastases. Fractionated SBRT was found
to be effective therapy for tumor control and symptomatic
SBRT to spinal tumors has the potential to expand on treat- improvement. Patients received a total dose of 27–30 Gy,
ment options currently available. Potential indications for delivered in 3 fractions. With a median follow-up of
SBRT to the spine include primary treatment for a single or 15.9 months, the progression-free survival after SBRT was
oligometastases in the spine, postoperative treatment, or sal- 80.5 % at 1 year and 72.4 % at 2 years. The number of
vage treatment after surgery or previous irradiation. SBRT patients reporting no pain from bone metastases increased
has advantages, such as the ability to give higher doses to the from 26 to 54 % 6 months after SBRT. In addition, there was
a statistically significant reduction in multiple symptoms
including disturbed sleep, drowsiness, sadness, fatigue, dis-
Q. Nguyen, M.D. • A.S. Shiu, Ph.D. • E.L. Chang, M.D. (*) tress, lack of appetite, nausea, and difficulty remembering.
Department of Radiation Oncology, Keck School of Medicine
On the other hand, single-fraction SBRT has the advan-
of USC, USC Norris Cancer Hospital, 1441 Eastlake Avenue,
Los Angeles, CA 90033, USA tage of convenience to the patient and treatment team and
e-mail: eric.chang@med.usc.edu is thus attractive, if and only if it can be delivered safely.
584 Q. Nguyen et al.
it may be prudent to request that the neurosurgeon resects the

disease in close proximity to the spinal cord so that a more
favorable anatomic configuration can be obtained for
SBRT. A comparison of modalities including single-fraction
and hypofractionated SBRT, conventional radiotherapy, and
surgery is shown in Fig. 46.1.
Published Clinical Results
At selected centers, SBRT programs are becoming more

mature, and there is recognition of the importance of cre-
dentialing system that is being put into place at many cen-
ters. Therefore, the opportunity to collect larger experiences
or long-term clinical data is now happening. Bilsky et al.
from Memorial Sloan Kettering reported on their initial
clinical experience in treating 16 paraspinal tumors with
intensity-modulated stereotactic radiotherapy. Metastatic
tumors received a dose of 20 Gy in 4–5 fractions while lim-
iting the spinal cord to a maximum dose of 6 Gy. Of the 15
patients who underwent radiographic follow-up, 13 demon-
strated either no interval growth or a reduction in tumor size
in a median follow-up period of 12 months (range, 2–23
months) [4]. Our group at The University of Texas MD
Anderson Cancer Center (MDACC) reported on 15 consec-
utive patients with metastatic spinal disease who underwent
75 treatments involving 90 isocenter setups on a phase I
clinical trial involving intensity-modulated, computed
tomography (CT) image-guided SBRT. Patients uniformly
received 30 Gy in 5 fractions while the spinal cord was con-
strained to a maximum dose of 10 Gy [5]. The procedure
Fig. 46.1 Comparison of treatment plans for 1-, 3-, and 5-fraction was technically feasible to perform in all patients, and no
SBRT neurologic toxicity was observed in any patient with a
median follow-up time of 9 months (range, 6–16 months).
The spinal cord tolerance may be better defined for single- Because SBRT for spinal metastases is usually given in the
fraction SBRT based on extrapolation of data from radiosur- context of palliation and avoiding spinal cord damage is
gery to the optic chiasm of approximately 8 Gy making it paramount, we initially chose to prescribe a hypofraction-
easier to achieve complete target volume coverage. This can ated dose of 30 Gy in 5 fractions on alternating days to
be appreciated by examining dose-volume histograms from allow for sufficient sublethal damage repair while conserva-
treatment plans representing 30 Gy in 5 fractions, 27 Gy in 3 tively limiting the spinal cord dose to 10 Gy. It is important
fractions, and 18 Gy in 1 fraction (Fig. 46.2). Furthermore, to note that treatment planning programs approximate doses
single-fraction SBRT follows the successful paradigm estab- to critical structures and are limited by the accuracy of beam
lished in the arena of intracranial radiosurgery and may lead data taken by ion chamber measurements, which is affected
to superior local control. Single-fraction SBRT may be pref- by the size or volume of the ion chamber, and whether leak-
erable for single tumors that are small in volume and ana- age between the leaves used in multileaf collimators and
tomically favorable in terms of greater distance from the leaf shape are taken into account. Recognizing that treat-
spinal cord. Fractionated SBRT may be preferable for spinal ment plans may over- or underestimate the actual doses
tumors with paraspinal components spanning multiple verte- delivered, it is prudent to conservatively limit doses to the
bral body levels or with large prevertebral components juxta- spinal cord. A dose of 10 Gy in 5 fractions is considered
posed to small bowel, because the biological dose to clinically insignificant. The biologic equivalence (BED2Gy)
surrounding normal tissues can be minimized to a greater of 30 Gy in 5 fractions is 40 Gy for early responding tissues
degree. For tumors surrounding the spinal cord, involving assuming an α/β of 10 Gy and 54–64 Gy assuming an α/β of
the epidural space, or causing compression of the thecal sac, 1.5–3 Gy for the spinal cord. With the increased confidence
46 Spinal Metastases: Viewpoint—Fractionated Radiation Therapy 585
Fig. 46.2 Single-fraction dose delivering 18 Gy to nonrenal cell spinal metastases (a) axial and (b) sagittal
of our safety and setup data demonstrating consistent daily space between disease and spinal cord. There were 63 non-
patient setups within 1 mm of isocenter [5], we have pro- cervical spinal metastases treated with single-fraction SBRT,
ceeded to shorten our treatments to a total dose of 27 Gy in all tumors received >16 Gy. Patients were stratified according
3 fractions on alternating days to allow for sufficient repair to histology (renal versus nonrenal). For nonrenal spinal
while limiting the spinal cord dose to 10 Gy. metastasis the mean GTV was prescribed 18 Gy, and for renal
Garg et al. reported the MDACC experience in delivering spinal metastasis the mean GTV received 24 Gy. For paraspi-
single-fraction SBRT to the spine in a prospective phase I/II nal tumors, nonrenal histology tumors were prescribed
study for previously unirradiated patients with sufficient 18 Gy to the GTV and 24 Gy to the GTV for renal tumors.
586 Q. Nguyen et al.
Table 46.1 Comparison of various spinal metastasis treatments

Single-fraction SBRT Hypofractionated SBRT Conventional radiotherapy Surgery
Safety Unforgiving. No Some compensation possible if 30 Gy in 10 fractions has Usual potential risks of
possibility for correction initial setup error made; spinal track record of safety in surgery including wound
after delivery of treatment; cord tolerance for treating a large number of infection, bleeding,
spinal cord tolerance hypofractionated radiotherapy patients worldwide damage to nerves, and
extrapolated from optic ill-defined; good approach for spinal cord
chiasm; may be less setting up new programs during
appropriate in previously learning curve; may be safer for
irradiated patients previously irradiated patients
Efficacy Theoretically may have Needs to be prospectively Mainstay for palliation of Rapid resolution of
greater ability than compared with single-fraction spinal metastases; limited neurologic symptoms
hypofractionated SBRT to SBRT and conventional by spinal cord tolerance due to spinal cord
overcome radioresistance radiotherapy so dose may be compression or pain
especially for renal cell inadequate; reirradiation caused by mechanical
carcinoma, melanoma, and usually not possible instability
sarcoma by getting over
shoulder of survival curve
Convenience Ultimate in convenience to Greater expenditure of resources Can start immediately, Postoperative recovery
patient and treatment required for labor-intensive easy to give, widely time, wound healing,
team, must be balanced procedure; may be required during available rehabilitation time to
against potential risks learning curve or for previously achieve ambulation
irradiated patients near cord
tolerance
Cost-effectiveness May be less costly than More costly due to multiple Cost-effective and Costly but definite
fractionated SBRT from a iterations of the same procedure; appropriate for majority indications for spinal
resource standpoint, and however, this cost may need to be of diffuse spinal cord compression and
therefore more cost- incurred initially by new programs metastases involving salvaging previously
effective if efficacy is to ensure patient safety until track multiple levels irradiated metastases
proved to be equivalent record for safety established
100
The 18-month actuarial rate of local tumor control for all T10GTV
patients was 88 %, and specifically for patients treated post- 1RX
Normalized volume (%)
80
operatively with SBRT, the local tumor control was 100 %. 3RX
Cord
The 18-month actuarial rate of freedom from neurologic 60 5RX
deterioration was 82 % [6] (Fig. 46.3).
There are also several publications documenting experi- 40 1RX
ence with single-fraction SBRT. Ryu et al. reported on pat- 3RX
terns of failure for 49 patients with 61 spinal metastases who 20
5RX
underwent single-dose radiosurgery (10–16 Gy) at Henry
Ford Hospital. Follow-up ranged from 6 to 24 months. 0
Complete and partial pain relief was achieved in 85 % of the 0 1000 2000 3000 4000
lesions treated, with relapse of pain at the treated site was Dose (cGy)
7 %, and there was progressive metastasis in the immediate
Fig. 46.3 Comparison of dose-volume histograms for 1-, 3-, and
adjacent spine in 5 % [7]. Benzil et al. treated 31 patients 5-fraction SBRT
who had 35 tumors with stereotactic radiosurgery to the
spine using the Novalis unit at New York Medical College.
Twenty-six tumors were spinal metastases and 9 were pri- Two patients experienced transient radiculitis, and one patient
mary tumors of the spine. In general, metastases not previ- had permanent neurologic deterioration due to pathologically
ously treated with radiation received 25 Gy in 10 fractions of confirmed radiation necrosis. The authors concluded that a bio-
external-beam radiation followed by a 6- to 8-Gy single- logical equivalent dose >60 Gy was associated with an
fraction boost. increased risk of radiculitis [8]. Gerszten et al. from the
Patients previously irradiated with 30 Gy using conven- University of Pittsburgh treated 115 consecutive patients with
tional radiation were treated with 10 Gy in 2 fractions. Rapid single-fraction radiosurgery using CyberKnife (45 cervical, 30
and significant pain relief was noted after treatment in 32 of thoracic, 36 lumbar, and 14 sacral). There were 17 benign and
34 treated tumors within 72 h and up to 3 months later. 108 metastatic tumors. Tumor volume ranged from 0.3 cm3 to
46 Spinal Metastases: Viewpoint—Fractionated Radiation Therapy 587
232 cm3 and tumor dose was 12–20 Gy prescribed to the 80 % integration of SBRT with systemic therapy, determination of
isodose line (mean, 14 Gy). No acute radiation toxicity or new spinal cord tolerance, and long-term sequelae of high doses
neurologic deficits occurred during the follow-up period of irradiation to the spine.
(range, 9–30 months; median, 18 months), and axial and radic-
ular pain improved in 74 of 79 patients who were symptomatic
before treatment. The authors concluded that the CyberKnife References
system was found to be feasible, safe, and effective [9].
DeSalles et al. from the University of California Los Angeles 1. Klimo P, Schmidt MH. Surgical management of spinal metastases.
Oncologist. 2004;9:188–96.
using the Novalis system treated 14 patients with 11 metasta-
2. Potters L, Steinberg M, Rose C, Potters L, Steinberg M, Rose C,
ses, 2 neurofibromas, and 1 meningioma. Twelve patients Timmerman R, Ryu S, Hevezi JM, Welsh J, Mehta M, Larson DA,
underwent previous conventional irradiation. A mean dose of Janjan NA. American Society for Therapeutic Radiology and
12 Gy (range, 8–21 Gy) was prescribed. With a mean follow- Oncology and American College of Radiology practice guideline
for the performance of stereotactic body radiation therapy. Int J
up of 6.1 months (range, 1–16 months), no complications were
Radiat Oncol Biol Phys. 2004;60:1026–32.
seen, and seven patients experienced significant pain relief. 3. Wang XS, Rhines LD, Shiu AS, Yang JN, Selek U, Gning I, Liu P,
Seven of 22 lesions that were treated progressed [10]. Allen PK, Azeem SS, Brown PD, Sharp HJ, Weksberg DC, Cleeland
Spinal SBRT continues to emerge as an effective thera- CS, Chang EL. Stereotactic body radiation therapy for management
of spinal metastases in patients without spinal cord compression: a
peutic approach for spinal metastases. However, treatment
phase 1-2 trial. Lancet Oncol. 2012;13(4):395–402.
planning and delivery remain individualized to specific 4. Bilsky MH, Yamada Y, Yenice KM, Lovelock M, Hunt M, Gutin
radiosurgery centers. The group at MDACC has developed a PH, Leibel SA. Intensity-modulated stereotactic radiotherapy of
generalizable class solution approach for spinal SBRT to paraspinal tumors: a preliminary report. Neurosurgery. 2004;54(4):
823–30.
provide optimal target coverage, reduce integral dose, and
5. Chang EL, Shiu AS, Lii MF, Rhines LD, Mendel E, Mahajan A,
maximize planning efficiency. This is of particular impor- Weinberg JS, Mathews LA, Brown BW, Maor MH, Cox JD. Phase
tance for treatment planners who do not have experience I clinical evaluation of near-simultaneous computed tomographic
with spinal SBRT planning. The class solution technique image-guided stereotactic body radiotherapy for spinal metastases.
Int J Radiat Oncol Biol Phys. 2004;59(5):1288–94.
generated plans that maintained target coverage and
6. Garg AK, Shiu AS, Yang J, Wang XS, Allen P, Brown BW,
improved conformality while reducing normal tissue integral Grossman P, Frija EK, McAleer MF, Azeem S, Brown PD, Rhines
dose for planners who had no prior experience with spinal LD, Chang EL. Phase 1/2 trial of single-session stereotactic body
SBRT. The class solution technique reduced treatment plan- radiotherapy for previously unirradiated spinal metastasis. Cancer.
2012;118(20):5069–77.
ning time by 30–60 % and MUs required by 20 % indepen-
7. Ryu S, Rock J, Rosenblum M, Kim JH. Patterns of failure after
dent of prior planning experience [11]. single-dose radiosurgery for spinal metastases. J Neurosurg. 2004;
Although this collective body of preliminary data is 101 Suppl 3:402–5.
encouraging, it should be cautioned that because the field is 8. Benzil DL, Saboori M, Mogilner AY, Rocchio R, Moorthy
CR. Safety and efficacy of stereotactic radiosurgery for tumors of
young, long-term complications are under investigation. It is
the spine. J Neurosurg. 2004;101 Suppl 3:413–8.
currently unknown to what extent high doses of radiation 9. Gerszten PC, Ozhasoglu C, Burton SA, Vogel WJ, Atkins BA,
will affect vertebral stability [12]. Vertebral compression Kalnicki S, Welch WC. CyberKnife frameless stereotactic radiosur-
fractures after spinal SBRT have been reported at two centers gery for spinal lesions: clinical experience in 125 cases.
Neurosurgery. 2004;55:89–98.
at Memorial Sloan Kettering and MDACC [13, 14] with
10. De Salles AA, Pedroso AG, Medin P, Agazaryan N, Solberg T,
rates ranging from 20 to 39 %. It also remains to be defined Cabatan-Awang C, Espinosa DM, Ford J, Selch MT. Spinal lesions
what is the threshold of spinal cord tolerance with hypofrac- treated with Novalis shaped beam intensity-modulated radiosur-
tionated or single-dose schedule of irradiation in the setting gery and stereotactic radiotherapy. J Neurosurg. 2004;101 Suppl
3:435–40.
of primary or reirradiation.
11. Weksberg DC, Palmer MB, Vu KN, Rebueno NC, Sharp HJ, Luo D,
Yang JN, Shiu AS, Rhines LD, McAleer MF, Brown PD, Chang
EL. Generalizable class solutions for treatment planning of spinal
stereotactic body radiation therapy. Int J Radiat Oncol Biol Phys.
Conclusion 2012;84(3):847–53.
12. Rock JP, Ryu S, Yin FF, Schreiber F, Abdulhak M. The evolving
Reports of the relative safety of fractionated and single- role of stereotactic radiosurgery and stereotactic radiation therapy
fraction SBRT to the spine have been published from several for patients with spine tumors. J Neurooncol. 2004;69:319–34.
13. Boehling NS, Grosshans DR, Allen PK, McAleer MF, Burton AW,
institutions. Larger experiences with long-term follow-up Azeem S, Rhines LD, Chang EL. Vertebral compression fracture
data are supporting the relative safety of high-dose SBRT to risk after stereotactic body radiotherapy for spinal metastases.
spinal tumors, although cases of spinal cord myelopathies J Neurosurg Spine. 2012;16:379–86.
are being reported. In addition, well-designed prospective 14. Rose PS, Laufer I, Boland PJ, Hanover A, Bilsky MH, Yamada J,
Lis E. Risk of fracture after single fraction image-guided intensity-
clinical investigations are needed to help address important modulated radiation therapy to spinal metastases. J Clin Oncol.
clinical issues such as optimal dose and fractionation, 2009;27:575–9.
Arteriovenous Malformation
Radiosurgery 47
Bruce E. Pollock
carry an increased annual hemorrhage risk or whether they

Introduction are just unlikely to develop other symptoms (seizures or
headaches) that would permit the diagnosis.
Intracranial arteriovenous malformations (AVM) are con-
gential lesions arising from abnormal blood vessel formation
[1–3]. Whereas normal embryogenesis results in the differ- AVM Management
entiation of primordial vascular channels into mature arter-
ies, veins, and capillaries, patients with AVMs develop direct The management options for patients diagnosed with intra-
arteriovenous shunts without the appropriate intervening cranial AVMs include observation, surgical resection, and
vascular beds. Recent large, prospective population-based stereotactic radiosurgery (SRS) (Fig. 47.1). Embolization of
studies have determined the incidence of newly diagnosed AVMs as primary treatment has recently been utilized [20, 21],
AVM patients to range from 1.12 to 1.34 per 100,000 person- but is more often performed in conjunction with either surgi-
years [4, 5]. The majority of patients become symptomatic in cal resection or SRS. Conventional fractionated radiation
the second through fourth decades of life with the most com- therapy has been utilized in the past but results in a low AVM
mon presentation being intracranial hemorrhage (ICH). cure rate [22]. Patients who present after a large ICH require
Patients may also have seizures or headaches, and the num- surgical evacuation to remove the mass effect associated
ber of incidentally discovered intracranial AVMs continues with the hematoma. In some instances, surgery is performed
to rise as more patients undergo magnetic resonance imaging before the patient has undergone cerebral angiography to
(MRI) of the head. Traditionally, the estimated annual risk of delineate the AVM morphology. Once the blood clot has
ICH from AVMs has ranged from 2 to 4 % [5–12] with a been removed and abnormal vessels discovered, most neuro-
combined annual morbidity and mortality of approximately surgeons recommend stopping surgery and performing angi-
1 % [11]. Because most AVM patients are diagnosed at a ography to determine a definitive plan for the residual nidus.
point when their life expectancy is long, the cumulative hem- After a recent hemorrhage, surgical resection is the preferred
orrhage risk is substantial. For example, a 30-year old person treatment for patients with accessible AVMs even if they do
carries approximately a 75 % lifetime chance of ICH. Factors not require immediate clot evacuation. The benefit of surgi-
associated with an increased risk of bleeding include prior cal resection compared to radiosurgery is the immediate
hemorrhage [9, 10, 12, 13], increasing age [14, 15], single or elimination of future hemorrhage risk. The Spetzler–Martin
deep draining veins [12, 13], associated arterial aneurysm grading system is the most frequently utilized scale to pre-
[16], and a diffuse AVM nidus [12]. Pediatric patients [17], dict outcomes after the surgical excision of cerebral AVMs
and patients with AVMs located in the basal ganglia, thala- [23]. This grading scale is based on AVM size, AVM loca-
mus, brain stem, or cerebellum [12, 18], and patients with tion, and the pattern of venous drainage, and has been shown
small AVMs [19] are more likely to present after a hemor- to predict patient outcomes after AVM resection in a number
rhage. It remains uncertain whether these patients actually of large AVM series [24–29].
As an alternative to surgical resection, SRS has been
shown to be a safe and effective method to manage patients
B.E. Pollock, M.D. (*) with cerebral AVMs. In 1972, Dr. Ladislau Steiner and col-
Department of Neurological Surgery and Radiation Oncology,
leagues from the Karolinska Institute recognized that single
Mayo Foundation, 200 First Street SW, Rochester,
MN 55905, USA fraction, high-dose irradiation caused the progressive oblit-
e-mail: pollock.bruce@mayo.edu eration of AVMs and subsequent cure from the risk of later
590 B.E. Pollock
Fig. 47.1 Treatment algorithm

for intracranial AVMs
hemorrhage [30]. In fact, because AVMs could be visualized Recently a number of factors have led some investigators
through angiography before the development of axial imag- to reexamine observation for patients discovered to have
ing, AVMs were a common indication treated in the early unruptured brain AVMs [39]. First, the annual rate of bleed-
radiosurgical series of Leksell. From 1968 until 1982, 204 of ing for patients with unruptured brain AVMs may be as low
the first 762 patients (27 %) having Gamma Knife radiosur- as 1 % [40, 41]. Second, the morbidity associated with AVM
gery had AVMs [31]. Concurrent with the work of Leksell bleeding may be less than previously thought [42, 43]. A
and Steiner, Kjellberg and Fabrikant were using heavy review of 119 AVM patients found that 47 % of patients suf-
charged particles instead of photons to irradiate AVMs fered no disability related to their first ICH and an additional
[32, 33], and other centers soon developed modified linear 37 % of patients remained independent in their activities of
accelerators (LINAC) to perform AVM radiosurgery [34–36]. daily living [42]. Third, neurologic deficits after surgical
Similar to the Swedish experience, these innovators noted resection of AVMs may be greater than previously described.
that focused radiation techniques could obliterate a high per- Hartmann et al. reported outcomes for 124 AVM patients fol-
centage of irradiated AVMs. lowed prospectively after surgical resection [44]. At last
Observation is also a recognized management strategy for follow-up, 38 % of patients had new postoperative neuro-
AVM patients especially if the AVM has not previously bled. logic deficits. Six percent of patients were disabled after sur-
The decision to treat an AVM is simply a comparison of the gery. Factors related to a decline in neurologic function were
estimated risk of intervention to the calculated risk if the female gender, AVM size, and deep venous drainage.
malformation is left untreated. If the risk of treatment is per- Likewise, Lawton et al. found that patients with unruptured
ceived to be less than the lifetime risk based on the natural AVMs were 2.3 times more likely to experience a decline in
history of untreated AVMs, then treatment is recommended. their modified Rankin Score (MRS) compared to patients
Prior to advent of modern imaging, improved microsurgical with ruptured AVMs [45]. A randomized trial of unruptured
techniques, and SRS, the majority of AVMs were not treated brain arteriovenous malformations (ARUBA) was designed
directly, and patients discovered to have AVMs were treated to compare the risk of observation versus prophylactic inter-
for their seizure disorder or headaches. Yet, once it was rec- vention for patients diagnosed with unruptured BAVM [46].
ognized that selected AVM patients could undergo treatment The primary endpoints of the ARUBA trial are the composite
directed at the AVM with acceptable morbidity, the trend risk of death or stroke and risk of death or clinical impair-
shifted so that surgical excision or radiosurgery was recom- ment, defined as a MRS ≥ 2. Although the design of ARUBA
mended for the majority of patients [37]. Despite technologi- has been criticized for a number of reasons including the
cal advances, it is recognized that the risk of treatment for intended follow-up period after randomization (planned,
patients with large AVMs is not insignificant, and observa- 5–10 years) [47], it was funded by the National Institute of
tion of Spetzler–Martin Grade IV and V AVMs is generally Neurological Disorders (U01 NS051483) and began enroll-
recommended unless the patient has hemorrhaged or is suf- ing patients in April 2007 at cerebrovascular centers around
fering from a progressive neurologic deficit [38]. the world.
47 Arteriovenous Malformation Radiosurgery 591
than four seizures prior to treatment. In the “low-seizure

AVM Radiosurgery group” with a follow-up longer than 2 years, 93 % were
seizure-free, 2 % improved but continued to have seizures,
Patient Selection and 5 % had worsening of their seizures. In patients with
more than four seizures preoperatively, seizure freedom
Proper patient selection is essential for successful AVM measured at 2 years of follow-up or longer was achieved in
radiosurgery. Once it has been determined that intervention 76 % of patients, 21 % were improved, and 3 % remained
is preferred over observation, several factors must be consid- unchanged. Overall, 83 % of patients remained seizure-free
ered when considering radiosurgery for AVM patients. Most at last follow-up in this study. Likewise, Han et al. reviewed
importantly, has the patient bled, and if so, how recently? 440 patients undergoing resection of supratentorial AVM
Patients with a recent ICH and a surgically accessible AVM [50]. Preoperative seizures were associated with prior AVM
typically are best managed with surgical resection. However, hemorrhage, male sex, and frontotemporal lesion location.
patients with a recent hemorrhage and a surgically inacces- At a mean follow-up of 21 months, 96 % of patients had a
sible AVM are generally good candidates for radiosurgery modified Engel class I outcome (seizure-free, 80 %; one
assuming that the AVM is not too large. Also, patients with a postoperative seizure, 16 %). We retrospectively reviewed 65
hemorrhage months or years earlier can certainly be consid- AVM patients with one or more seizures having SRS at our
ered for radiosurgery because they have passed the time center between 1990 and 1998 [52]. Twenty-six patients
when a re-hemorrhage is most likely to occur. In this situa- (51 %) were seizure-free (aura-free) after radiosurgery at
tion, a comparison of the chance of AVM elimination with- 3-year follow-up; 40 patients (78 %) had an excellent out-
out new deficits risk between surgical resection and come (non-disabling simple partial seizures only) at 3-year
radiosurgery should be undertaken. Standardized scales such follow-up. Factors associated with seizure-free or excellent
as the Spetzler–Martin Grade [23] and the radiosurgery- outcomes include a low seizure frequency score (<4) before
based arteriovenous malformation score (RBAS) [48, 49] radiosurgery, smaller AVM volume, and smaller AVM diam-
should be used to estimate the efficacy of surgical resection eter. Twenty-three patients had intractable partial epilepsy
and radiosurgery, respectively, for individual AVM patients. prior to treatment. Twelve (52 %) of 23 and 11 of 18 (61 %)
Of course, the success of any intervention must take into patients with medically intractable partial epilepsy had
account the physician’s experience with a particular group of excellent outcomes at years 1 and 3, respectively. Yang et al.
patients, rather relying on published reports from experi- analyzed 161 consecutive patients having SRS for unrup-
enced centers. tured AVM (mean follow-up, 90 months) [53]. Sixty-six of
Other considerations regarding treatment choice include 86 patients (77 %) were seizure-free after SRS. In this series,
the presence or absence of associated aneurysms, history of AVM obliteration correlated with a seizure-free outcome
seizures, and need for pretreatment embolization. Aneurysms after SRS (97 % versus 31 %). Although the number of
on feeding vessels of AVMs are likely caused by the stress on patients who are seizure-free appears higher in most studies
the arterial walls due to increased blood flow to the AVM. after AVM resection, selection bias in these anecdotal series
Patients with associated aneurysms and surgically accessible prevents any conclusion as to which approach correlates
AVMs can often have both operated in a single procedure, with improved seizure outcomes.
thereby immediately removing the hemorrhage risk. Patients
with associated aneurysms and surgically inaccessible AVMs
can be considered for pre-radiosurgical surgical clipping or Technique
endovascular treatment of their aneurysm, with radiosurgery
performed during a separate later procedure. Conversely, if The goal of SRS is to accurately deliver a high dose of radia-
the associated aneurysm is small, radiosurgery alone is gen- tion to an imaging-defined target. Although the definition of
erally sufficient because the aneurysm will likely either sta- SRS has been modified to include treatment in 1–5 fractions,
bilize or regress as the blood flow to the AVM decreases after the majority of patients with intracranial AVMs treated with
radiosurgery. SRS are treated in a single fraction. Thus, placement of a
Another consideration when discussing surgical resection stereotactic headframe is generally performed to ensure rigid
or radiosurgery for AVM patients is a history of seizures. fixation and minimize patient movement during imaging and
Approximately 15–20 % of AVM patients will present with radiation delivery. Headframe placement for adults is per-
a seizure [50, 51]. However, few patients will have medically formed under local anesthesia supplemented by a low dose
resistant epilepsy, and the use of standardized scales to deter- of benzodiazepine. Patients less than 16 years typically
mine seizure frequency is uncommon in AVM papers. require general anesthesia for radiosurgery. Once the stereo-
Piepgras et al. studied seizure outcomes after AVM surgery tactic headframe has been placed, most patients undergo a
[51]. Preoperative seizure burden was defined as less or more post-gadolinium MRI and biplanar cerebral angiography.
592 B.E. Pollock
Reliance on angiography alone for radiosurgical dose plan- Patients with residual AVM on follow-up angiography are
ning increases the chance of treating too much adjacent nor- evaluated for observation, repeat SRS, or surgical resection
mal brain tissue due to the often irregular shape of AVMs based on their age, clinical condition, and the AVM response
[54]. In addition, for AVMs located in the posterior fossa and from the first radiosurgical procedure.
lateral temporal regions, the chance of not including a por-
tion of the nidus in the prescription isodose volume (PIV) is
greater secondary to the difficulty in visualizing the AVM Obliteration After Radiosurgery
clearly on angiography. Over the past 15 years, we have
become increasingly confident in performing AVM SRS The primary goal of AVM radiosurgery is complete nidus
without a stereotactic angiogram and now perform approxi- obliteration to eliminate a patient’s risk of future hemorrhage.
mately 20 % of our cases based on MRI alone. Patients with Lindquist and Steiner in 1988 defined AVM obliteration as
small, compact, hemispheric AVMs with simple venous follows: “we have considered a result satisfactory only when
drainage are ideal candidates for using MRI alone for dose the arteriogram has shown a normal circulation time, com-
planning. Nonetheless, we continue to insist that patients plete absence of pathological vessels in the former nidus of
have a complete diagnostic angiogram, including appropri- the malformation, and the disappearance or normalization of
ate external carotid injections, before any decision is made draining veins from the area” [62]. Generally, AVM oblitera-
about the feasibility of SRS and to determine if the patient tion occurs between 1 and 5 years after radiosurgery.
has any associated aneurysms. Sequential histopathological changes after AVM radiosurgery
The goal of SRS planning is to create a conformal dose include early damage to the endothelial cells, followed by
plan that precisely covers the three-dimensional shape of the progressive thickening of the intimal layer secondary to pro-
nidus. Feeding arteries and draining veins are not included in liferation of smooth muscle cells which produce an extracel-
the dose plan if possible. Inclusion of these vessels will lular matrix containing Type IV collagen, then cellular
increase the volume covered, which may decrease the radia- degeneration and hyaline transformation [63]. The effect of
tion dose prescribed. Dose prescription must take into high-dose single fraction radiation has been studied in cell
account two conflicting considerations: the chance of oblit- cultures taken from patients undergoing AVM resection [64].
eration versus the chance of radiation-related complications. Within 5 days after irradiation, the proliferation index
Increasing radiation dose directly correlates with the chance decreased and remained decreased over the observation
of obliteration. Assuming that the radiation is well targeted, period. Other notable findings were immunohistochemical
the chance of AVM cure is approximately 70 %, 80 %, and evidence of apoptosis and transformation of fibroblastic cells
90 % for radiation doses of 16, 18, and 20 Gy, respectively into activated myofibroblasts. Histologic and electron micro-
[55, 56]. However, the likelihood of radiation-related com- scopic studies of seven AVMs resected after bleeding 10–52
plications after AVM radiosurgery increases at higher radia- months after radiosurgery revealed spindle cell proliferation
tion doses and larger AVM volumes. Early dose prescription in the connective tissue stroma and in the subendothelial
generally followed either Kjellberg’s 3 % isodose line [32] or region of irradiated vessels [65]. It was concluded that the
Flickinger’s integrated logistic formula [57] to predict the characteristics of the spindle cells were similar to myofibro-
probability of radiation-related complications. More recent blasts noted during wound healing, and these cells likely con-
studies have correlated the chance of radiation-related com- tributed to the occlusive process and final obliteration of
plications after AVM radiosurgery relates to some measure AVMs after radiosurgery.
of the radiation dose to the surrounding tissue [58, 59]. The most important factor associated with obliteration
Patients with AVMs in the thalamus, basal ganglia, and after AVM radiosurgery is the margin dose delivered to the
brainstem are more likely to develop neurologic deficits sec- nidus [55, 56, 66–69]. A number of models have been devel-
ondary to imaging changes noted on MRI [58]. oped to predict the chance of AVM cure after SRS. Karlsson
After the dose plan is reviewed by all members of the et al. reported the K index as a method to predict obliteration
radiosurgical team and found to be acceptable, the patient after AVM radiosurgery [56]. Based on 945 AVM patients
undergoes radiation delivery. Following radiation delivery, having radiosurgery from 1970 until 1990, they found a log-
the patient is typically discharged home the day of the proce- arithmic relationship between minimum dose and AVM
dure. Immediate complications are rare. Some patients com- obliteration increasing to a maximum of 87 %. A higher
plain of pin site discomfort, neck pain, or headache, but average dose also shortened the latency to AVM obliteration.
these are usually temporary and can be managed with over- The correlation of obliteration rate to the product minimum
the-counter medications. Follow-up after radiosurgery dose (AVM volume)1/3 was termed the K index. The oblitera-
typically consists of clinical examination and MRI at yearly tion rate increased linearly with the K index up to a value of
intervals for the first 5 years after SRS. If MRI suggests that approximately 27, and for higher K values, the obliteration
the AVM has gone on to complete obliteration, then follow- rate had a constant value of approximately 80 %. For the
up angiography is recommended to confirm cure [60, 61]. group of patients receiving an AVM margin dose of at least
25 Gy (n = 273), the obliteration rate at 2 years was 80 %. If

obliterations that occurred beyond 2 years are included, the
obliteration rate increased to 85 %. Schwartz et al. developed
the obliteration prediction index (OPI) as a method to predict
success or failure for individual AVM patients [70].
Analyzing a total of 426 patients having either Gamma Knife
or LINAC-based SRS, a relationship was noted between the
calculated OPI (AVM margin dose, Gy/lesion diameter, cm)
and AVM obliteration. Although both the K index and the
OPI correlate with AVM elimination, neither takes into
account the likelihood of producing radiation-related com-
plications for a given radiation dose.
Perhaps more instructive than studies analyzing factors
relating to AVM obliteration has been a number of papers
that have reviewed patients with incompletely obliterated
AVMs after SRS [71–73]. The most common reasons for
incomplete nidus obliteration are targeting errors, recanali-
zation of a portion of the AVM that was previously emboli-
zed, reexpansion of nidus after hemorrhage, and low radiation
dose. These studies have demonstrated the need for complete
nidus coverage at the time of SRS. The routine use of CT or Fig. 47.2 Neuroimaging studies of a 26-year-old woman with an inci-
MRI in conjunction with angiography has been the most dentally discovered left basal ganglia AVM. Lateral (Left, a) and
important change to reduce the chance of “missing” a por- anterior-posterior (Middle, b) left internal carotid angiograms showing
AVM on the day of SRS. (Right, c) Computed tomography performed 6
tion of the AVM with the PIV [54]. It is also critical that all months after SRS showing both intraparenchymal and intraventricular
possible feeding arteries are injected at the time of SRS, hemorrhage
including the external carotid arteries for peripherally located
AVMs. Certainly part of the problem in AVM SRS is defin-
ing the nidus accurately. Buis et al. had six independent cli- The issue of AVM bleeding during this latency interval has
nicians contour the nidus of AVM patients based off digital been extensively studied. Early papers on AVM radiosurgery
subtraction angiography (DSA) [74]. They noted significant suggested that the risk of bleeding initially increased before
interobserver variation when outlining the nidus and con- AVM obliteration occurred [35, 78]. Colombo et al. reported
cluded that this may contribute to failure in some AVM 180 patients having LINAC-based AVM radiosurgery [35].
radiosurgical cases. Yu et al. compared dose plans based on a Twenty-seven patients with large AVMs underwent partial
combination of angiography and MRI to those derived from treatment; the mean follow-up in this series was 43 months.
MRI alone [75]. They concluded that MRI-based dose plan- Fifteen patients bled after radiosurgery. In totally irradiated
ning without angiography should be limited to patients with cases, the bleeding risk decreased from 4.8 % in the first 6
smaller AVMs and compact niduses. A study from the months after SRS to 0 % starting from 1 year after SRS. Patients
University of Florida has suggested that AVM morphology is with partially irradiated AVMs had bleeding 4–10 % over the
an important factor associated with obliteration [76]. first 2 years, then no bleeding thereafter. A review of 65
Specifically, they noted patients with a diffuse nidus struc- patients with Spetzler–Martin Grade I–II AVMs found that
ture and associated neovascularity were at a higher risk of five patients (8 %) sustained an ICH after radiosurgery [78].
incomplete nidus obliteration when compared to patients Although the annual hemorrhage rate was 3.7 % over the
with compact AVMs. High-flow AVMs have also been noted latency interval, all the observed hemorrhages occurred within
to have a lower rate of obliteration compared to low-flow the first 8 months after the procedure. Despite these early
lesions. In a study of 139 patients, patients with angioarchi- papers suggesting an increased risk of bleeding after radiosur-
tectural features consistent with high-flow achieved had a gery, larger and more detailed analysis of this question has
complete obliteration rate less than 40 % [77]. confirmed that the risk of bleeding is either unchanged or
decreased following AVM SRS [79–83]. Maruyama et al. per-
formed a retrospective observational study of 500 AVM
Post-radiosurgical Hemorrhage patients having radiosurgery [83]. Comparing the risk of
bleeding before and after radiosurgery, they found a 54 %
The primary drawback of AVM SRS when compared to surgi- reduction in the bleeding risk during the latency interval. The
cal resection is that the patient remains at risk for hemorrhage risk reduction was greatest among patients who presented
until the AVM has gone onto complete obliteration (Fig. 47.2). with a hemorrhage. Likewise, Karlsson et al. analyzed the
594 B.E. Pollock
Fig. 47.3 Long-TR MRI of a 10-year-old boy who presented with showing the nidus to be smaller and adjacent edema. The patient con-
headaches. (Left, a) MRI before SRS showing the right-sided AVM tinued to have headaches, but was neurologically intact. (Right, c) MRI
involving the sensorimotor cortex. (Middle, b) MRI 9 months after 26 months after SRS showing near-complete resolution of the edema
large AVM experience at the Karolinska Institute and found induced imaging changes and dose distribution parameters
that some measure of protection occurred as early as 6 months in 73 patients having AVM radiosurgery [86]. The AVM tar-
after radiosurgery for patients receiving an AVM margin dose get volume correlated significantly with the development of
of 25 Gy [82]. The presence of nidal or associated aneurysms edema and breakdown of the blood–brain barrier. They con-
has also been associated with an increased risk of ICH after cluded that each measure studied (including V12 and
AVM radiosurgery [80, 81]. Following angiographic oblitera- Dmean20) yielded similar results, and no parameter was
tion, the chance of hemorrhage is markedly reduced, but cases favored over the others. Yet, although such imaging changes
of ICH after nidus elimination have been reported [84, 85]. are frequently noted within the first year after AVM radiosur-
Risk factors for postobliteration bleeding include young age gery, most are temporary and the majority of patients remain
and persistent enhancement of the irradiated nidus. asymptomatic (Fig. 47.3). It is believed that many “radiation-
related” imaging changes noted after AVM SRS are related
to alterations in the regional blood flow in the brain adjacent
Radiation-Related Complications to the AVM as the nidus obliterates and blood is being
directed into vasculature that has lost the capability for auto-
Advances in imaging and radiosurgical devices have signifi- regulation. Early closure of draining veins before occlusion
cantly improved patient outcomes after SRS. Nonetheless, of the nidus may also be a factor that contributes to the devel-
the tissue adjacent to a radiosurgical target does receive a opment of edema after SRS (Fig. 47.4) [87, 88]. Patients
dose of radiation. Assuming that the dose plan conforms to with AVMs in the thalamus, basal ganglia, and brainstem are
the three-dimensional shape of the target, the major factors more likely to develop neurologic deficits secondary to
that determine the amount of radiation delivered to nearby imaging changes noted on MRI [58, 67, 68, 89].
structures are the PIV and the prescribed radiation dose. As Although most imaging changes detected on MRI after
the PIV increases, the fall off of radiation becomes less steep, AVM SRS are temporary and resolve by 2 years, a small per-
resulting in more radiation to the adjacent brain. It is this centage of patients will continue to demonstrate imaging
general principle that typically limits the size of intracranial changes consistent with radiation necrosis (Fig. 47.5). MRI
radiosurgical targets to 3 cm or less in average diameter. findings consistent with radiation necrosis include persistent
Recognition of the relationship between the overall radiation enhancement at the irradiated site beyond 2 years with associ-
exposure and the chance of radiation-related complications ated edema and mass effect. Again, the chance that a patient
has resulted in studies correlating imaging changes on MRI will have symptoms related to radiation necrosis relates pri-
to measures of radiation such as the 12 Gy volume (V12) marily to the location of the AVM. In addition to radiation
[58] and the mean dose received by 20 cm3 surrounding the necrosis, other late complications have been noted after AVM
maximum radiation point (Dmean20) [59]. Areas of radiosurgery including diffuse white matter changes [90], cyst
increased signal on long-TR MRI are noted in approximately formation [91–93] (Fig. 47.6), and stenosis of major intracra-
30–50 % of AVM patients after SRS. The chance of develop- nial vessels [94]. The most devastating complication after any
ing these imaging changes increases at larger V12 or radiation-based procedure is a radiation-induced neoplasm.
Dmean20. Levegrun et al. studied the correlation of radiation- To date, only a few cases of radiation-induced tumors have
Fig. 47.4 MRI of a 49-year-old woman with an incidentally discov- (Middle, b) and T1-weighted without gadolinium (Right, c) MRI at that
ered left frontal AVM. (Left, a) Long-TR image before SRS. Eighteen time showed the nidus to be decreased in size with adjacent edema.
months after SRS the patient suffered a generalized seizure. Long-TR Note acute thrombus within nidus and the draining veins
Fig. 47.5 Post-gadolinium (left) and long-TR (right) MRI performed 3

years after SRS of a left sylvian fissure AVM demonstrating persistent
enhancement and edema consistent with radiation necrosis
been reported after AVM radiosurgery [95]. Although the

actual incidence of this complication will not be known for
years, it is estimated that the risk of a radiation-induced tumor Fig. 47.6 MRI of a 35-year-old woman who underwent LINAC-based
after radiosurgery will be significantly less than fractionated SRS 7 years earlier for an incidentally discovered left frontal AVM. The
patient developed increasing headaches, seizures, and syncope. (Top,
radiation therapy [96]. left, a, and right, b) Post-gadolinium and long-TR images showing cyst
formation, edema, and mass effect. Angiography revealed complete
obliteration. (Bottom, left, a, and right, b) Post-gadolinium and long-
Repeat AVM Radiosurgery TR images 2 months after surgical resection of the obliterated AVM
showing resolution of edema and mass effect
Patients with residual nidus after AVM radiosurgery remain at

risk for ICH. A number of studies have analyzed the results of obliteration (predicted number of obliterations, 65). Fourteen
repeat AVM radiosurgery [72, 97–99]. Karlsson et al. patients (13 %) had radiation-related complications after a
reviewed 112 patients and tested whether models developed second radiosurgical procedure. They concluded that the
after one radiosurgery also predicted obliteration and compli- obliteration rate after repeat radiosurgery is similar to pri-
cation rates after repeat AVM radiosurgery [97]. Sixty-two mary procedures, but the complication rate increases with the
of 101 patients with angiographic follow-up had complete overall amount of radiation given. Two recent studies from
596 B.E. Pollock
the University of Florida [99] and the University of Pittsburgh safe method to treat large AVMs not amenable to standard
[72] confirmed that complete obliteration is achieved in surgical or SRS treatment.
approximately 70 % of patients after repeat SRS. Patients Conventional fractionated radiation therapy has been
with smaller volume AVMs at the time of repeat SRS had used to treat patients with large AVMs [22, 105]. Redekop
higher rates of obliteration in both of these series. Kano et al. et al. reported 15 patients with inoperable AVMs ranging
found that the incidence of radiation-related complications from 1.5 to 6.5 cm in diameter who underwent conventional
increased from 5 % after initial SRS to 10 % after repeat SRS radiation therapy between 1955 and 1985 [105]. Fourteen of
[72], whereas Stahl et al. noted no change in the risk of radi- 15 patients received a total radiation dose of 40 Gy or more
ation-related complications between the first and second pro- in 15–28 fractions. At a mean follow-up of 8.1 years, the
cedures [99]. Overall, repeat SRS should be considered a safe annual hemorrhage rate was 3.3 %, and later angiography
and effective option for patients with incomplete AVM oblit- confirmed AVM obliteration in 2 of 12 patients (17 %).
eration after their initial radiosurgical procedure. Karlsson et al. reviewed the outcomes of 28 AVM patients
undergoing fractionated radiation therapy between 1980 and
1985 [22]. The median volume treated was 78 cm3. Using a
Radiosurgery of Large AVMs fractionation scheme of 42 Gy in 12 fractions, only two
patients (8 %) demonstrated angiographic cure. The annual
Although AVM obliteration is significantly related to higher hemorrhage rate after radiation therapy was 6 %. Although
radiation doses, dose prescription must also take into account the information on low dose fractionation for AVM patients
the likelihood of radiation-related complications. Studies on is limited, it appears that little protection is provided against
the dose–volume relationship of AVM post-radiosurgical future bleeding after this technique.
radiation-related complications have demonstrated a high More recently, a number of centers have examined the
morbidity for large AVMs [58, 59, 86]. Miyawaki et al. efficacy of hypofractionated radiation schedules using higher
reported that following linear accelerator-based radiosurgery radiation doses per fraction to treat patients with intracranial
of AVMs greater than 14 cm3 receiving 16 Gy or more, the AVMs [106–109]. Lindvall et al. used several fractionation
incidence of post-radiosurgical abnormalities on long-TR schemes (generally 30–35 Gy/five fractions) to treat 36
MRI was 72 % [100]. The incidence of radiation necrosis for AVMs patients [108]. Two-year follow-up angiography
these patients was 22 %. Consequently, single-fraction SRS showed that 48 % of the patients were cured; angiographic
is generally performed only for patients with AVMs of an obliteration rose to 76 % on 5-year angiography. Although
average diameter of 3 cm or less (approximately 14 cm3). the median AVM volume was only 8.5 cm3 for the entire
Alternative strategies that utilize radiation in the manage- group, obliteration for patients with AVMs larger than 10 cm3
ment of large AVMs include embolization followed by SRS, (n = 10) was 70 % at last follow-up. Veznedaroglu and col-
fractionated radiation techniques, and staged-volume SRS. leagues from Jefferson Hospital managed 30 patients with
Planned embolization and SRS has been used for many years large AVMs from 1995 to 1998 using a combination of pre-
to manage patients with large AVMs as an alternative to sur- radiation embolization and stereotactic radiation therapy
gical resection for these difficult lesions [101–104]. As (SRT) [109]. Patients received either 42 Gy in six fractions
opposed to embolization prior to resection where flow reduc- (n = 7) or 30 Gy in six fractions (n = 23). Angiographic oblit-
tion is the goal, the goal of pre-radiosurgical embolization is eration was noted on 5-year follow-up angiography in 83 %
permanent volume reduction. Gobin et al. published the of patients in the 42 Gy group (five of six patients) compared
results in 96 patients undergoing acrylate embolization fol- to only 22 % of patients in the 30 Gy group (4 of 18 patients).
lowed by radiosurgery [102]. Complete AVM obliteration However, the overall morbidity in the higher dose group was
was documented in 53 of 90 evaluable patients (59 %). 43 % (14 % permanent deficits). Chang et al. treated 33 AVM
Sixteen patients (13 %) had complications from the emboli- patients with SRT using a treatment regimen of 25–35 Gy in
zation procedures, and two patients died of intracerebral four daily fractions [106]. Fifteen of the 33 patients (45 %)
hemorrhages prior to having radiosurgery. Recanalization of had AVMs larger than 2.5 cm in diameter. The 5- and 6-year
the AVM was seen in 14 % of patients. Blackburn et al. actuarial obliteration rates were 61 % and 71 %, respectively.
described the management of 21 patients with large AVMs One patient (3 %) developed radiation necrosis after
(>3cm) using combined embolization and SRS from 1994 to SRT. Hattangadi et al. reported 59 patients having planned
2006 [101]. Eight complications occurred in 43 embolization two-fraction stereotactic proton beam SRS from 1991 to
procedures (24 %). The majority of complications were tem- 2009 [107]. The median AVM volume was 23 cm3; the most
porary and no patient had a permanent major neurologic defi- common dose prescription was 16 Gy radiobiologic equiva-
cit. Following SRS, 16 of 19 patients (84 %) had obliteration lent in two fractions. At a median follow-up of 56 months,
confirmed by either angiography (n = 13), MR angiography nine patients (15 %) had total obliteration. Five patients
(n = 2), or CT angiography (n = 1). They concluded that (9 %) had permanent complications after proton-beam ther-
endovascular therapy followed by SRS was an effective and apy, none of which were severe.
Recognition of the limitations of the various management technique to permit a better comparison with other available
options available to patients with large AVMs has encour- treatment options for this difficult patient group.
aged some radiosurgical centers to begin staged-volume
radiosurgery for these patients [110–112]. Volume staging of
large AVMs into multiple radiosurgical sessions separated by Radiosurgery-Based AVM Grading System
several months allows a higher radiation dose to be delivered
to the entire AVM volume while minimizing the radiation The Spetzler–Martin grading system has become widely
exposure to the adjacent brain. We performed a dosimetric accepted as an accurate method to predict patient outcomes
study comparing staged-volume AVM radiosurgery and after surgical resection of AVMs [23]. Comprising three
hypothetical single-session procedures for our first ten components (AVM size, location, and pattern of venous
patients [112]. Staged-volume radiosurgery decreased the drainage), this system has been validated prospectively [25]
V12 by an average of 11.1 %, and the non-AVM V12 was and by numerous cerebrovascular centers of excellence
reduced by an average of 27.2 %. Huang et al. reported the [26–29]. Although some authors have noted discrepancies
outcomes for 18 patients with large AVMs (>15cm3) under- between AVM grade and patient outcomes, especially in
going staged-volume SRS over a 13-year interval [110]. The regard to Grade III AVMs [24, 113, 114], the general consen-
median radiation dose at each stage was 15 Gy. The 5-year sus supports this grading scale as practical and reliable.
rate of complete obliteration and hemorrhage after SRS was Unfortunately, the Spetzler–Martin classification is insensi-
29 % and 31 %, respectively. One patient (6 %) had a tive to changes in AVM volume for lesions less than 3 cm,
radiation-related neurologic deficit. Kano et al. treated 47 and AVMs located in the basal ganglia, thalamus, or brain-
patients with staged-volume SRS between 1992 and 2006 stem are considered at equal risk for radiation-related com-
[111]. The median total volume was 22.0 cm3; the median plications as AVMs located in-or-near critical cortical
AVM margin dose was 16 Gy. The 5-year actuarial oblitera- locations such as the sensorimotor cortex. For example, a
tion rate was 28 %. Patients treated with doses ≥ 17 Gy had a 1 cm diameter AVM has a volume < 1 cm3, whereas a 3 cm
5-year obliteration rate of 62 %. Radiation-related complica- diameter AVM has an approximate volume of 14 cm3. The
tions occurred in 13 % of patients. By minimizing the radia- expected obliteration rates for these AVMs should be approx-
tion exposure to the adjacent brain, the incidence of imately 90 % and 50 %, respectively. Yet, both would be
radiation-related complications has been reduced after considered small (<3 cm) in the Spetzler–Martin system
staged-volume SRS for large AVMs compared to single- (Fig. 47.7). Consequently, a valid instrument capable of
fraction SRS [100]. More data are need on this relatively new accurately predicting outcomes after AVM SRS is necessary
Fig. 47.7 Lateral cerebral angiograms of two patients with Spetzler– and was found to have a 4 cm left frontal AVM (Spetzler–Martin Grade
Martin Grade I–II AVMs that have significantly different risks from SRS. II). The patient underwent staged-volume SRS (margin dose, 15 Gy); the
(Left, a) Nineteen-year-old woman with headaches and was found to modified RBAS was 2.14. From Pollock BE, Niranjan A, Kano H,
have a 1.2 cm right frontal AVM (Spetzler–Martin Grade I). The patient Lumsford LD, eds. Gama Knife Radiosurgery for Brain Vascular
underwent single-session SRS (margin dose, 25 Gy); the modified RBAS Malformations. Prog Neurol Surg. Basel, Karger, 2013; 27:1-9; used
was 0.48. (Right, b) Sixteen-year-old boy who had tonic–clonic seizure with permission from S. Karger AG, Basel
598 B.E. Pollock
to adequately compare the expected results of microsurgery Table 47.1 Radiosurgery-based AVM grading system
and SRS for individual AVM patients. Factor Coefficient
Successful AVM radiosurgery results in complete nidus AVM volume (cm3) 0.1
obliteration without new or worsened neurologic deficits. Patient age (years) 0.02
Karlsson et al. reported the K index as a method to predict AVM location 0.5
obliteration after AVM SRS [56]. In a similar fashion, Cerebral hemisphere, corpus callosum, cerebellum = 0
Schwartz et al. proposed the OPI as a means to estimate the Basal ganglia, thalamus, brainstem = 1
chance of AVM obliteration for individual patients [70]. AVM score = (0.1) (AVM volume) + (0.02) (patient age) + (0.5)
(AVM location)
Although both correlate with AVM obliteration after SRS,
neither takes into account the chance of radiation-related
complications. Also, the K index and OPI are based on the
radiation dosimetry used (AVM margin dose) at the time of preoperative decision making between surgical and radiosur-
treatment and not on patient and AVM characteristics alone. gical management for individual AVM patients.
Therefore, our center in collaboration with the University of Advances in endovascular techniques, microsurgical
Pittsburgh developed a radiosurgery-based AVM grading resection, and SRS have all contributed to improving
system that accounts for these shortcomings and predicted outcomes for AVM patients. To provide the best care for
the chance of successful, single-session AVM radiosurgery AVM patients, we should utilize every tool available to indi-
based solely on patient and AVM variables [48]. vidualize their management taking into account the patient’s
The RBAS was developed based on the multivariate anal- age, presentation, nidus size and morphology, AVM location,
ysis of 220 patients having SRS between 1987 and 1991 at and patient preference.
the University of Pittsburgh. The dependent variable in all
analyses was excellent patient outcomes (complete AVM
obliteration without new neurologic deficit). The grading Illustrative Case
scale was then tested on a separate set of 136 AVM patients
treated between 1990 and 1996 at the Mayo Clinic. Overall, A 10-year-old boy complained of headaches that continued
121 of 220 (55 %) of the Pittsburgh patients had excellent to increase despite medical therapy for migraines. In addi-
outcomes. Multivariate analysis found five variables related tion, he developed episodes of numbness and tingling involv-
to excellent patient outcomes: AVM volume, patient age, ing his left hand and arm that also were increasing in
AVM location, prior embolization, and number of draining frequency and intensity. Of note, the boy was left handed. An
veins. Regression analysis modeling permitted removal of MRI of his head revealed a large AVM located in the tempo-
two significant variables (prior embolization, number of ral, frontal, and parietal lobes with extension into the lateral
draining veins) and resulting in an equation to predict patient portion of the thalamus (Fig. 47.9). The AVM measured
outcomes after AVM radiosurgery. Seventy-nine of 136 Mayo 4.5 cm × 4.2 cm × 3.5 cm. There was no evidence of either
Clinic patients (58 %) had excellent outcomes. Testing of the recent or remote hemorrhage. An electroencephalogram
RBAS on the Mayo patients showed the AVM score predicted showed diffuse slowing in the region of the AVM, but no
patient outcomes after radiosurgery. All patients with an distinct epileptiform activity. A cerebral angiogram showed
AVM score ≤ 1.0 had an excellent outcome compared to only the AVM had arterial supply from the right anterior, middle,
39 % of patients with an AVM score > 2.0. Consequently, and posterior cerebral arteries. The AVM had both superficial
despite significant differences in preoperative patient charac- and deep venous drainage. There were no intranidal or feed-
teristics and dose prescription guidelines at the two centers, ing vessel aneurysms and no evidence of venous outflow
the RBAS strongly correlated with patient outcomes after restriction. The case was reviewed at our combined cerebro-
single-session radiosurgery for both patient groups. Since vascular conference and the options of observation, surgical
that time, the RBAS has been modified using location as a resection, and radiosurgery were discussed. It was felt that
two-tiered variable (basal ganglia, thalamus or brainstem ver- the AVM was becoming symptomatic, most likely from a
sus other) rather than a three-tiered variable [49] (Table 47.1). “steal phenomenon,” so treatment of the AVM was recom-
In side-by-side testing with the original RBAS, the modified mended. The risk of neurologic injury from surgical resec-
RBAS performed equally well in predicting both AVM oblit- tion of this Spetzler–Martin Grade IV AVM in the patient’s
eration without new neurologic deficits or decline in MRS dominant hemisphere was considered high, especially since
after radiosurgery (Fig. 47.8). The RBAS has been validated the AVM was unruptured. Due to the large size (estimated
by a number of centers performing not only the Gamma volume, 30–35 cm3) of the AVM, the RBAS was estimated to
Knife SRS but also LINAC-based SRS, CyberKnife SRS, as exceed 3.5. Therefore, radiosurgery of the entire lesion in a
well as proton beam SRS [89, 107, 115–124]. The RBAS per- single procedure was also considered a poor option. After
mits an accurate estimation of outcomes from SRS to aid in discussing the options of planned embolization followed by
Fig. 47.8 Graph showing

percentage of AVM patients who
achieve an excellent outcome
(nidus obliteration without new
neurologic deficits) (solid line) or
decline in modified Rankin Score
(dashed line) based on the
modified RBAS. From Pollock
BE, Niranjan A, Kano H,
Lumsford LD, eds. Gama Knife
Radiosurgery for Brain Vascular
Malformations. Prog Neurol
Surg. Basel, Karger, 2013;
27:1-9; used with permission
from S. Karger AG, Basel
Fig. 47.10 Lateral (left, a) and anterior–posterior (right, b) right

carotid angiograms showing the dose plan for the first SRS covering the
medial portion of the AVM. Injection of the vertebral artery filled an
additional posterior component of the AVM not visualized on the
carotid injection, but which was included in the dose plan
Fig. 47.9 Axial post-gadolinium MRI of a 10-year-old boy with a

large right-sided AVM located in the temporal, frontal, and parietal
lobes with extension into the lateral portion of the thalamus
SRS or staged-volume SRS, the patient’s family consented

to staged-volume radiosurgery.
The procedures were performed under general anesthesia
using stereotactic MRI and angiography for dose planning.
The first SRS covered the medial portion of the AVM
(Fig. 47.10). The dose plan consisted of 11 isocenters of
radiation to cover a volume of 15.5 cm3. The margin dose
was 15 Gy and the maximum dose was 30 Gy. The patient
Fig. 47.11 Lateral (left, a) and anterior–posterior (right, b) right
returned 4 months later and underwent a second SRS to carotid angiograms showing the dose plan for the second SRS covering
cover the lateral portion of the AVM (Fig. 47.11). The dose the lateral portion of the AVM
600 B.E. Pollock
Fig. 47.12 Lateral right carotid

angiogram at the time of repeat
SRS (52 months after completion
of staged-volume SRS). The
nidus has decreased in size from
>30 to <1 cm3
plan consisted of 13 isocenters of radiation to cover a volume 5. Stapf C, Mast H, Sciacca RR, Berenstein A, Nelson PK, Gobin
of 17.0 cm3. The margin dose was 15 Gy and the maximum YP, Pile-Spellman J, Mohr JP. The New York Islands AVM Study:
design, study progress, and initial results. Stroke. 2003;34:29–33.
dose was 30 Gy. The patient had a reduction in his headaches 6. Brown Jr RD, Wiebers DO, Forbes G, O’Fallon WM, Piepgras DG,
within several months of staged-volume SRS and remained Marsh WR, Maciunas RJ. The natural history of unruptured intra-
neurologically intact. Yearly MRI showed the AVM to be sig- cranial arteriovenous malformations. J Neurosurg. 1988;68:352–7.
nificantly smaller but there appeared to be persistent filling 7. Crawford PM, West CR, Chadwick DW, Shaw MDM.
Arteriovenous malformations of the brain: natural history in unop-
of an early draining vein. Fifty-two months after completion erated patients. J Neurol Neurosurg Psychiatry. 1986;49:1–10.
of staged-volume SRS, the patient underwent a stereotactic 8. Fults D, Kelly Jr DL. Natural history of arteriovenous of the brain:
angiogram which confirmed near-total obliteration of the a clinical study. Neurosurgery. 1984;15:658–62.
AVM (Fig. 47.12). The AVM volume at the time of repeat 9. Graf CJ, Perret GB, Torner JC. Bleeding from cerebral arteriove-
nous malformations as part of their natural history. J Neurosurg.
SRS was 0.8 cm3. The margin dose was 18 Gy and the maxi- 1983;58:331–7.
mum dose was 36 Gy. Three years after his last SRS proce- 10. Itoyama Y, Uemura S, Ushio Y, Kuratsu J, Nonaka N, Wada H,
dure the patient remains well and has completed high school. Sano Y, Fukumura A, Yoshida K, Yano T. Natural course of unop-
Follow-up angiography is planned in several years to deter- erated intracranial arteriovenous malformations: study of 50
cases. J Neurosurg. 1989;71:805–9.
mine the obliteration status of his AVM. 11. Ondra SK, Troupp H, George ED, Schwab K. The natural history
of symptomatic arteriovenous malformations of the brain: a
24-year follow-up assessment. J Neurosurg. 1990;73:387–91.
12. Pollock BE, Flickinger JC, Lunsford LD, Bissonette DJ,
References Kondziolka D. Factors that predict the bleeding risk of cerebral
arteriovenous malformations. Stroke. 1996;27:1–6.
1. McCormick WE. Pathology of vascular malformations of 13. Mast H, Young WL, Koennecke HC, Sciacca RR, Osipov A, Pile-
the brain. In: Wilson CB, Stein B, editors. Intracranial arteriove- Spellman J, Hacein-Bey L, Duong H, Stein BM, Mohr JP. Risk of
nous malformations. Baltimore: Williams and Wilkins; 1984. spontaneous hemorrhage after diagnosis of cerebral arteriovenous
p. 44–63. malformation. Lancet. 1997;350:1065–8.
2. Padget DH. The cranial venous system in man in reference to 14. Karlsson B, Lindquist C, Steiner L. The effect of gamma knife
development, adult configuration, and relation to arteries. Am J surgery on the risk of rupture prior to AVM obliteration. Minim
Anat. 1956;98:307–55. Invasive Neurosurg. 1996;39:21–7.
3. Takashima S, Becker LE. Neuropathology of cerebral arteriove- 15. Stapf C, Khaw AV, Sciacca RR, Hofmeister C, Schumacher HC,
nous malformations in children. J Neurol Neurosurg Psychiatry. Pile-Spellman J, Mast H, Mohr JP, Hartmann A. Effect of age on
1980;43:380–5. clinical and morphological characteristics in patients with brain
4. Al-Shahi R, Bhattacharya JJ, Currie DG, Papanastassiou V, Ritchie arteriovenous malformations. Stroke. 2003;34:2664–9.
V, Roberts RC, Sellar RJ, Warlow CP, SIVMS Collaborators. 16. Stapf C, Mohr JP, Pile-Spellman J, Sciacca RR, Hartmann A,
Prospective, population-based detection of intracranial vascular Schumacher HC, Mast H. Concurrent arterial aneurysms in brain
malformations in adults: the Scottish Intracranial Vascular arteriovenous malformations with haemorrhagic presentation.
Malformation Study (SIVMS). Stroke. 2003;34:1163–9. J Neurol Neurosurg Psychiatry. 2002;73:294–8.
17. Kondziolka D, Humphreys RP, Hoffman HJ, Hendrick EB, Drake 38. Han PP, Ponce FA, Spetzler RF. Intention-to-treat analysis of
JM. Arteriovenous malformations of the brain in children: a forty Spetzler-Martin grades IV and V arteriovenous malformations:
year experience. Can J Neurol Sci. 1992;19:40–5. natural history and treatment paradigm. J Neurosurg. 2003;98:3–7.
18. Khaw AV, Mohr JP, Sciacca RR, Schumacher HC, Hartmann A, 39. Choi JH, Mohr JP. Brain arteriovenous malformations in adults.
Pile-Spellman J, Mast H, Stapf C. Association of infratentorial Lancet Neurol. 2005;4:299–308.
brain arteriovenous malformations with hemorrhage at initial pre- 40. Halim XA, Johnston SC, Singh V, McCulloch CE, Bennett JP,
sentation. Stroke. 2004;35:660–3. Achrol AS, Sidney S, Young WL. Longitudinal risk of intracranial
19. Spetzler RF, Hargraves RW, McCormick PW, Zabramski JM, hemorrhage in patients with arteriovenous malformations of the
Flom RA, Zimmerman RS. Relationship of perfusion pressure and brain within a defined population. Stroke. 2004;35:1697–702.
size to risk of hemorrhage from arteriovenous malformations. 41. Stapf C, Mast H, Sciacca RR, Choi JH, Khaw AV, Connolly ES,
J Neurosurg. 1992;76:918–23. Pile-Spellman J, Mohr JP. Predictors of hemorrhage in patients
20. Kessler I, Riva R, Ruggiero M, Manisour M, Al-Khawaldeh M, with untreated brain arteriovenous malformation. Neurology.
Mounayer C. Successful transvenous embolization of brain arte- 2006;66:1350–5.
riovenous malformations using onyx in five consecutive patients. 42. Hartmann A, Mast H, Mohr JP, Koennecke HC, Osipov A, Pile-
Neurosurgery. 2011;69:184–93. Spellman J, Duong DH, Young WL. Morbidity of intracranial
21. van Rooij WJ, Jacobs S, Sluzewski M, van der Pol B, Beute GN, hemorrhage in patients with cerebral arteriovenous malformation.
Sprengers ME. Curative embolization of brain arteriovenous mal- Stroke. 1998;29:931–4.
formations with onyx: patient selection, embolization technique, 43. van Beijnum J, Lovelock CE, Cordonnier C, Rothwell PM, Klijn
and results. AJNR Am J Neuroradiol. 2012;33(7):1299–304. CJM, Salman RA. Outcome after spontaneous and arteriovenous
doi:10.3174/ajnr.A2947. malformation-related intracerebral haemorrhage: population-
22. Karlsson B, Lindqvist M, Blomgren H, Wan-Yeo G, Soderman M, based studies. Brain. 2009;132:537–43.
Lax I, Yamamoto M, Bailes J. Long-term results after fractionated 44. Hartmann A, Stapf C, Hofmeister C, Mohr JP, Sciacca RR, Stein
radiation therapy for large brain arteriovenous malformations. BM, Faulstich A, Mast H. Determinants of neurological outcome
Neurosurgery. 2005;57:42–9. after surgery for brain arteriovenous malformations. Stroke.
23. Spetzler RF, Martin NA. A proposed grading system for arteriove- 2000;31:2361–4.
nous malformations. J Neurosurg. 1986;65:476–83. 45. Lawton MT, Du R, Tran MN, Nelson M, Achrol A, McCulloch
24. de Oliveria E, Tedeschi H, Raso J. Comprehensive management of CE, Johnston SC, Quinnine NJ, Young WL. Effect of presenting
arteriovenous malformations. Neurol Res. 1998;20:673–83. hemorrhage on outcome after microsurgical resection of brain
25. Hamilton MG, Spetzler RF. The prospective application of a arteriovenous malformations. Neurosurgery. 2005;56:485–93.
grading system for arteriovenous malformations. Neurosurgery. 46. Mohr JP, Moskowitz AJ, Parides M, Stapf C, Young WL. Hull
1994;34:2–7. down on the horizon. A randomized trial of unruptured brain arte-
26. Heros RC, Korosue K, Diebold PM. Surgical excision of cerebral riovenous malformations (ARUBA) trial. Stroke. 2012;43:
arteriovenous malformations. Neurosurgery. 1990;26:570–8. 1744–5.
27. Pik JHT, Morgan MK. Microsurgery for small arteriovenous mal- 47. Cockroft KM, Jayaraman MV, Amin-Hanjani S, Derdeyn CP,
formations of the brain: results in 110 consecutive patients. McDougall CG, Wilson JA. A perfect storm: how a randomized
Neurosurgery. 2000;47:571–7. trial of unruptured brain arteriovenous malformations’ (ARUBA’s)
28. Pikus HJ, Beach ML, Harbaugh RE. Microsurgical treatment of trial design challenges notions of external validity. Stroke.
arteriovenous malformations: analysis and comparison with ste- 2012;43:1979–81.
reotactic radiosurgery. J Neurosurg. 1998;88:641–6. 48. Pollock BE, Flickinger JC. A proposed radiosurgery-based grad-
29. Schaller C, Schramm J. Microsurgical results for small arteriove- ing system for arteriovenous malformations. J Neurosurg.
nous malformations accessible for radiosurgical or embolization 2002;96:79–85.
treatment. Neurosurgery. 1997;40:664–74. 49. Pollock BE, Flickinger JC. Modification of the radiosurgery-
30. Steiner L, Leksell L, Grietz T, Forster DMC. Stereotaxic radiosur- based arteriovenous malformation grading system. Neurosurgery.
gery for cerebral arteriovenous malformations: report of a case. 2008;63:239–43.
Acta Chir Scand. 1972;138:459–64. 50. Englot DJ, Young WL, Han SJ, McCulloch CE, Chang EF, Lawton
31. Leksell L. Stereotactic radiosurgery. J Neurol Neurosurg MT. Seizure predictors and control after microsurgical resection
Psychiatry. 1983;46:797–803. of supratentorial arteriovenous malformations in 440 patients.
32. Kjellberg RN, Hanamura T, Davis KR, Lyons SL, Adams Neurosurgery. 2012;71:572–80.
RD. Bragg-peak proton-beam therapy for arteriovenous malfor- 51. Piepgras DG, Sundt Jr TM, Ragoowansi AT, Stevens L. Seizure
mations. N Engl J Med. 1983;309:269–74. outcome in patients with surgically treated cerebral arteriovenous
33. Steinberg GK, Fabrikant JI, Marks MP, Levy RP, Frankel KA, malformations. J Neurosurg. 1993;78:5–11.
Phillips MH, Shuer LM, Silverberg GD. Stereotactic heavy- 52. Schäuble B, Cascino GD, Pollock BE, Gorman DA, Weigand S,
charged-particle Bragg-peak radiation for intracranial arteriove- Cohen-Godal A, McClelland RL. Seizure outcomes after stereo-
nous malformations. N Engl J Med. 1990;323:96–101. tactic radiosurgery for cerebral arteriovenous malformations.
34. Betti OO, Munari C, Rosler R. Stereotactic radiosurgery with the Neurology. 2004;63:683–7.
linear accelerator: treatment of arteriovenous malformations. 53. Yang S, Kim DG, Chung H, Paek SH. Radiosurgery for unrup-
Neurosurgery. 1989;24:311–21. tured cerebral arteriovenous malformations. Long-term seizure
35. Columbo F, Pozza F, Chierego G, Casentini L, De Luca G, outcome. Neurology. 2012;78:1292–8.
Francescon P. Linear accelerator radiosurgery of cerebral arterio- 54. Blatt DR, Friedman WA, Bova FJ. Modifications based on com-
venous malformations: an update. Neurosurgery. 1994;34:14–21. puted tomographic imaging in planning the radiosurgical treat-
36. Friedman WA, Bova FJ, Mendenhall WM. Linear accelerator ment of arteriovenous malformations. Neurosurgery. 1993;33:
radiosurgery for arteriovenous malformations: the relationship of 588–95.
size to outcome. J Neurosurg. 1995;82:180–9. 55. Flickinger JC, Pollock BE, Kondziolka D, Lunsford LD. A dose-
37. Heros RC, Yu YK. Is surgical therapy needed for unruptured response analysis of arteriovenous malformation obliteration by
arteriovenous malformations? Neurology. 1987;37:279–86. radiosurgery. Int J Radiat Oncol Biol Phys. 1996;36:873–9.
602 B.E. Pollock
56. Karlsson B, Lindquist C, Steiner L. Prediction of obliteration after 73. Gallina P, Merienne L, Meder JF, Schlienger M, Lefkopoulos D,
gamma knife surgery for cerebral arteriovenous malformations. Merland J. Failure in radiosurgery treatment of cerebral arteriove-
Neurosurgery. 1997;40:425–31. nous malformations. Neurosurgery. 1998;42:996–1004.
57. Flickinger JC. An integrated logistic formula for prediction of 74. Buis DR, Lagerwaard FJ, Barkhof F, Dirven CMF, Lycklama GJ,
complications from radiosurgery. Int J Radiat Oncol Biol Phys. Meijer OWM, van den Berg R, Langendijk HA, Slotman BJ,
1989;17:879–85. Vandertopet WP. Stereotactic radiosurgery for brain AVMs: role
58. Flickinger JC, Kondziolka D, Lunsford LD, Liscak R, Phuong of interobserver variation in target definition on digital subtrac-
LK, Pollock BE. Development of a model to predict permanent tion angiography. Int J Radiat Oncol Biol Phys. 2005;62:
symptomatic post-radiosurgery injury for arteriovenous malfor- 246–52.
mation patients. Int J Radiat Oncol Biol Phys. 2000;46:1143–8. 75. Yu C, Petrovich Z, Apuzzo ML, Zelman V, Giannotta SL. Study of
59. Lax I, Karlsson B. Prediction of complications in gamma knife magnetic resonance imaging-based arteriovenous malformation
radiosurgery of arteriovenous malformations. Acta Oncol. delineation without conventional angiography. Neurosurgery.
1996;35:49–56. 2004;54:1104–7.
60. Pollock BE, Kondziolka D, Lunsford LD, Patel A, Bissonette DJ, 76. Zipfel GJ, Bradshaw P, Bova FJ, Friedman WA. Do the morpho-
Flickinger JC. Magnetic resonance imaging: an accurate method logical characteristics of arteriovenous malformations affect the
to evaluate arteriovenous malformations after stereotactic radio- results of radiosurgery? J Neurosurg. 2004;101:393.
surgery. J Neurosurg. 1996;85:1044–9. 77. Taeshineetanakul P, Krings T, Geibprasert S, Menezes R, Agid R,
61. Buis DR, Bot JCJ, Barkhof F, Knol DL, Lagerwaard FJ, Slotman Terbrugge K, Schwartz ML. Angioarchitecture determines oblit-
BJ, Vandertop WP, van den Berg R. The predictive value of 3D eration rate after radiosurgery in brain arteriovenous malforma-
time-of-flight angiography in assessment of brain arteriovenous tions. Neurosurgery. 2012;71(6):1071–8. doi:10.1227/NEU.
malformation obliteration after radiosurgery. AJNR Am J ob013e31826f79ec.
Neuroradiol. 2012;33:232–8. 78. Pollock BE, Lunsford LD, Kondziolka D, Maitz A, Flickinger
62. Lindquist C, Steiner L. Stereotactic radiosurgical treatment of JC. Patient outcomes after stereotactic radiosurgery for “operable”
arteriovenous malformations. In: Lunsford LD, editor. Modern arteriovenous malformations. Neurosurgery. 1994;35:1–8.
stereotactic neurosurgery. Boston: Martinus Nijhoff Publishing; 79. Friedman WA, Blatt DL, Bova FJ, Buatti JM, Mendenhall WM,
1988. p. 491–505. Kubilis PS. The risk of hemorrhage after radiosurgery for arterio-
63. Schneider BF, Eberhard DA, Steiner LE. Histopathology of venous malformations. J Neurosurg. 1996;84:912–9.
arteriovenous malformations after gamma knife radiosurgery. 80. Nataf F, Ghossoub M, Schlienger M, Moussa R, Meder J, Roux
J Neurosurg. 1997;87:352–7. F. Bleeding after radiosurgery for cerebral arteriovenous malfor-
64. Major O, Szeifert GT, Fazekas I, Vitanovics D, Csonka E, Kocsis mations. Neurosurgery. 2004;55:298–305.
B, Bori Z, Kemeny AA, Nagy Z. Effects of single high-dose 81. Pollock BE, Flickinger JC, Lunsford LD, Bissonette DJ,
gamma irradiation on cultured cells in human cerebral arteriove- Kondziolka D. Hemorrhage risk after stereotactic radiosurgery of
nous malformation. J Neurosurg. 2002;97(5 Suppl):459–63. cerebral arteriovenous malformations. Neurosurgery. 1996;38:
65. Szeifert GT, Major O, Kemeny AA. Ultrastructural changes in 652–61.
arteriovenous malformations after gamma knife surgery: an elec- 82. Karlsson B, Lax I, Söderman M. Risk of hemorrhage during the
tron microscopic study. J Neurosurg. 2005;102(Suppl):289–92. 2-year latency period following gamma knife radiosurgery for
66. Kano H, Lunsford LD, Flickinger JC, Yang H, Flannery TJ, Awan arteriovenous malformations. Int J Radiat Oncol Biol Phys.
NR, Niranjan A, Novotny J, Kondziolka D. Stereotactic radiosur- 2001;49:1045–51.
gery for arteriovenous malformations, part 1: management of 83. Maruyama K, Kawahara N, Shin M, Tago M, Kishimoto J, Kurita
Spetzler-Martin grade I and II arteriovenous malformations. H, Kawamoto S, Morita A, Kirino T. The risk of hemorrhage after
J Neurosurg. 2012;116:11–20. radiosurgery for cerebral arteriovenous malformations. N Engl J
67. Kano H, Kondziolka D, Flickinger JC, Yang H, Flannery TJ, Med. 2005;352:146–53.
Niranjan A, Novotny J, Lunsford LD. Stereotactic radiosurgery 84. Lindqvist M, Karlsson B, Guo W, Kihlstrom L, Lippitz B,
for arteriovenous malformations, part 4: management of basal Yamamoto M. Angiographic long-term follow-up data for arterio-
ganglia and thalamus arteriovenous malformations. J Neurosurg. venous malformations previously proven to be obliterated after
2012;116:33–43. gamma knife radiosurgery. Neurosurgery. 2000;46:803–10.
68. Kano H, Kondziolka D, Flickinger JC, Yang H, Flannery TJ, 85. Shin M, Kawahara N, Maruyama K, Tago M, Ueki K, Kirino
Ninanjan A, Novotny J, Lunsford LD. Stereotactic radiosurgery T. Risk of hemorrhage from an arteriovenous malformation con-
for arteriovenous malformations, part 5: management of brainstem firmed to have been obliterated on angiography after stereotactic
arteriovenous malformations. J Neurosurg. 2012;116:44–53. radiosurgery. J Neurosurg. 2005;102:842–6.
69. Shin M, Maruyama K, Kurita H, Kawamoto S, Tago M, Terahara 86. Levegrun S, Hof H, Essig M, Schlegel W, Debus J. Radiation-
A, Morita A, Ueki K, Takakura K, Kirino T. Analysis of nidus induced changes of brain tissue after radiosurgery in patients with
obliteration rates after gamma knife surgery for arteriovenous arteriovenous malformations: correlation with dose distribution
malformations based on long-term follow-up data: the University parameters. Int J Radiat Oncol Biol Phys. 2004;59:796–808.
of Tokyo experience. J Neurosurg. 2004;101:18–24. 87. Pollock BE. Occlusive hyperemia. A radiosurgical phenomenon?
70. Schwartz M, Sixel K, Young C, Kemeny A, Forster D, Walton L, Neurosurgery. 2000;47:1178–84.
Franssen E. Prediction of obliteration of arteriovenous malforma- 88. Chapman PH, Ogilvy CS, Loeffler JS. The relationship between
tions: the obliteration prediction index. Can J Neurol Sci. 1997;24: occlusive hyperemia and complications associated with the radio-
106–9. surgical treatment of arteriovenous malformations: report of two
71. Ellis TL, Friedman WA, Bova FJ, Kubilis PS, Buatti JM. Analysis cases. Neurosurgery. 2004;55:228–34.
of treatment failure after radiosurgery for arteriovenous malfor- 89. Pollock BE, Gorman DA, Brown PD. Radiosurgery for arteriove-
mations. J Neurosurg. 1998;89:104–10. nous malformations of the basal ganglia, thalamus, and brainstem.
72. Kano H, Kondziolka D, Flickinger JC, Yang H, Flannery TJ, Awan J Neurosurg. 2004;100:210–4.
NR, Niranjan A, Novotny J, Lunsford LD. Stereotactic radiosur- 90. Yamamoto M, Ban S, Ide M, Jimbo M. A diffuse white matter isch-
gery for arteriovenous malformations, part 3: outcome predictors emic lesion appearing 7 years after stereotactic radiosurgery for
and risks after repeat radiosurgery. J Neurosurg. 2012;116:21–32. cerebral arteriovenous malformations: case report. Neurosurgery.
1997;41:1405–9.
91. Kihlström L, Guo W, Karlsson B, Lindquist C, Lindqvist 108. Lindvall P, Bergström P, Löfroth P, Hariz M, Henriksson R,
M. Magnetic resonance imaging of obliterated arteriovenous mal- Jonasson P, Bergenheim AT. Hypofractionated conformal stereo-
formations up to 23 years after radiosurgery. J Neurosurg. tactic radiotherapy for arteriovenous malformations. Neurosurgery.
1997;86:589–93. 2003;53:1036–42.
92. Pollock BE, Brown Jr RD. Management of cysts arising after 109. Veznedaroglu E, Andrews D, Benitez R, Downes MB, Werner-
radiosurgery of intracranial arteriovenous malformations. Wasik M, Rosenstock J, Curran WJ, Rosenwasser RH. Fractionated
Neurosurgery. 2001;49:259–65. stereotactic radiotherapy for the treatment of large arteriovenous
93. Yamamoto M, Jimbo M, Hara M, Saito I, Mori K. Gamma knife malformations with or without previous partial embolization.
radiosurgery for arteriovenous malformations: long-term follow- Neurosurgery. 2004;55:519–30.
up results focusing on complications occurring more than 5 years 110. Huang PP, Rush SC, Donahue B, Narayana A, Becske T, Nelson
after irradiation. Neurosurgery. 1996;38:906–14. PK, Han K, Jafar JJ. Long-term outcomes after staged-volume
94. Yamamoto M, Ide M, Jimbo M, Ono Y. Middle cerebral artery radiosurgery for large arteriovenous malformations. Neurosurgery.
stenosis caused by relatively low-dose irradiation with stereotactic 2012;71:632–44.
radiosurgery for cerebral arteriovenous malformations: case 111. Kano H, Kondziolka D, Flickinger JC, Park K, Parry PV, Yang H,
report. Neurosurgery. 1997;41:474–7. Sirin S, Niranjan A, Novotny J, Lunsford LD. Stereotactic radio-
95. Kaido T, Hoshida T, Uranishi R, Akita N, Kotani A, Nishi N, surgery for arteriovenous malformations, part 6: multistaged volu-
Sakaki T. Radiosurgery-induced brain tumor: case report. metric management of large arteriovenous malformations.
J Neurosurg. 2001;95:710–3. J Neurosurg. 2012;116:54–65.
96. Loeffler JS, Niemierko A, Chapman PH. Second tumors after 112. Pollock BE, Kline RW, Stafford SL, Foote RL, Schomberg PJ. The
radiosurgery: tip of the iceberg or a bump in the road? rationale and technique of staged-volume arteriovenous malfor-
Neurosurgery. 2003;52:1436–42. mation radiosurgery. Int J Radiat Oncol Biol Phys. 2000;48:
97. Karlsson B, Kihlstrom L, Lindquist C, Steiner L. Gamma knife 817–24.
surgery for previously irradiated arteriovenous malformations. 113. Lawton MT. Spetzler-Martin grade III arteriovenous malforma-
Neurosurgery. 1998;42:1–5. tions: surgical results and modification of the grading scale.
98. Schlienger M, Nataf F, Lefkopoulos D, Mammar H, Missir O, Neurosurgery. 2003;52:740–8.
Meder J, Huart J, Platoni P, Deniaud-Alexandre E, Merienne L. 114. Morgan MK, Drummond KJ, Grinnell V, Sorby W. Surgery for
Repeat linear accelerator radiosurgery for cerebral arteriovenous cerebral arteriovenous malformations: risks related to lenticulo-
malformations. Int J Radiat Oncol Biol Phys. 2003;56:529–36. striate arterial supply. J Neurosurg. 1997;86:801–5.
99. Stahl JM, Chi Y, Friedman WA. Repeat radiosurgery for intracra- 115. Andrade-Souza YM, Zadeh G, Ramani M, Scora D, Tsao MN,
nial arteriovenous malformations. Neurosurgery. 2012;70:150–4. Schwartz ML. Testing the radiosurgery-based arteriovenous mal-
100. Miyawaki L, Dowd C, Wara W, Goldsmith B, Albright N, Gutin P, formation score and the modified Spetzler-Martin grading system
Halbach V, Hieshima G, Higashida R, Lulu B, Pitts L, Schell M, to predict radiosurgical outcome. J Neurosurg. 2005;103:642–8.
Smith V, Weaver K, Wilson C, Larson D. Five year results of 116. Andrade-Souza Y, Zadeh G, Scora D, Tsao MN, Schwartz
LINAC radiosurgery for arteriovenous malformations: outcome ML. Radiosurgery for basal ganglia, internal capsule, and thala-
for large AVMs. Int J Radiat Oncol Biol Phys. 1999;44:1089–96. mus arteriovenous malformations: clinical outcomes.
101. Blackburn SL, Ashley Jr WM, Rich KM, Simpson JR, Drzymala Neurosurgery. 2005;56:56–64.
RE, Ray WZ, Moran CJ, Cross III DT, Chicoine MR, Dacey Jr 117. Cohen-Gadol AA, Pollock BE. Radiosurgery for pediatric arterio-
RG, Derdeyn CP, Zipfel GJ. Combined endovascular emboliza- venous malformations. J Neurosurg. 2006;104(6 Suppl):388–91.
tion and stereotactic radiosurgery in the treatment of large 118. Colombo F, Cavedon C, Casentini L, Francescon P, Causin F,
arteriovenous malformations. J Neurosurg. 2011;114:1758–67. Pinna V. Early results of CyberKnife radiosurgery for arteriove-
102. Gobin YP, Laurent A, Merienne L, Schlienger M, Aymard A, nous malformations. J Neurosurg. 2009;111:807–19.
Houdart E, Casasco A, Lefkopoulos D, George B, Merland 119. Lee JS, Girvigian MR, Miller MJ, Rahimian J, Chen JCT,
JJ. Treatment of brain arteriovenous malformations by emboliza- Greathouse HE, Tome M. Validation of a radiosurgery-based grad-
tion and radiosurgery. J Neurosurg. 1996;85:19–28. ing system for arteriovenous malformations. Radiosurgery. 2006;
103. Lawton MT, Hamilton MG, Spetzler RF. Multimodality treatment 6:221–8.
of deep arteriovenous malformations: thalamus, basal ganglia, and 120. Maruyama K, Kondziolka D, Niranjan A, Flickinger JC, Lunsford
brain stem. Neurosurgery. 1995;37:29–36. LD. Stereotactic radiosurgery for brainstem arteriovenous malfor-
104. Pandey P, Marks MP, Harraher CD, Westbroek EM, Chang SD, mations: factors affecting outcome. J Neurosurg. 2004;100:
Levy RP, Dodd RL, Steinberg GK. Multimodality management 407–13.
of Spetzler-Martin Grade III arteriovenous malformations. 121. Maruyama K, Shin M, Tago M, Kurita H, Kawamoto S, Morita A,
J Neurosurg. 2012;116:1279–88. Kirino T. Gamma knife surgery for arteriovenous malformations
105. Redekop GJ, Elisevich KV, Gaspar LE, Wiese KP, Drake involving the corpus callosum. J Neurosurg. 2005;102(Suppl):
CG. Conventional radiation therapy of intracranial arteriovenous 49–52.
malformations: long-term results. J Neurosurg. 1993;78:413–22. 122. Pollock BE, Brown Jr RD. Use of the modified rankin scale to
106. Chang T, Shirato H, Aoyama H, Ushikoshi S, Kato N, Kuroda S, assess outcome after arteriovenous malformation radiosurgery.
Ishikawa T, Houkin K, Iwasaki Y, Miyasaka K. Stereotactic irra- Neurology. 2006;67:1630–4.
diation for intracranial arteriovenous malformations using stereo- 123. Raffa SJ, Chi YY, Bova FJ, Friedman WA. Validation of the
tactic radiosurgery or hypofractionated stereotactic radiotherapy. radiosurgery-based arteriovenous malformation score in a large
Int J Radiat Oncol Biol Phys. 2004;60:861–70. linear accelerator radiosurgery experience. J Neurosurg. 2009;
107. Hattangadi JA, Chapman PH, Bussiere MR, Niemierko A, Ogilvy 111:832–9.
CS, Rowell A, Daartz J, Loeffler JS, Shih HA. Planned two- 124. Wegner RE, Oysul K, Pollock BE, Sirin S, Kondziolka D, Niranjan
fraction proton beam stereotactic radiosurgery for high-risk inop- A, Lunsford LD, Flickinger JC. A modified radiosurgery-based
erable cerebral arteriovenous malformations. Int J Radiat Oncol arteriovenous malformation grading scale and correlation with
Biol Phys. 2012;83:533–41. outcomes. Int J Radiat Oncol Biol Phys. 2011;79:1147–50.
Arteriovenous Malformations:
Eric M. Deshaies and Surasak Komonchan
Introduction Natural History of Arteriovenous

Malformations
Arteriovenous malformations (AVMs) are congenital vascular
lesions comprising of direct connections between arteries and The prevalence of AVMs in the general population is less
veins without an intervening capillary network. The central than 1 % [8–11]. AVMs can manifest with seizures, head-
collection of intertwined vessels formed by these arterial-to- ache, cognitive deficit, pulsatile tinnitus, or other focal neu-
venous shunts, called the nidus, is often embedded within the rological deficits. The most common presentation is
parenchyma of the central nervous system. The etiology of intracranial hemorrhage associated with a 50 % morbidity
AVMs is uncertain, but current thought is that they represent and mortality [12, 13]. AVMs are the most common cause of
nests of abnormal regression in the fetal vasculature [1, 2]. intracranial hemorrhage in children and young adults [13,
AVMs vary in size, location, and angioarchitecture, each 14]. The risk of hemorrhage in unruptured AVMs is esti-
central to preoperative planning. They are frequently sup- mated to be about 2–4 % annually [9, 11, 15–22] but is
plied by multiple arterial pedicles causing abnormally high cumulative such that patients diagnosed at a younger age are
flow and pressure within the nidus and draining veins at a higher risk of bleeding during the course of their life-
(Fig. 48.1). The abnormal hemodynamic forces within the time. After the initial hemorrhage, the risk of another bleed
AVM and surrounding vasculature can result in hemorrhage. is 18 % in the first 6 months and returns to baseline by 3–5
Traditionally, surgical resection has been the only treatment years [15, 20, 22].
for these lesions. However, with the development of mini- The abnormally high flow and pressure within the AVM
mally invasive techniques including endovascular emboliza- contribute to the bleeding and can also result in aneurysm
tion and stereotactic radiosurgery (SRS), AVM treatment development. AVM-associated aneurysms further increase
plans have become more complex and frequently require a the annual bleeding risk to 41 % [12, 15, 23]. These aneu-
multimodality approach [3–7]. Occasionally, depending rysms are likely the result of abnormal shear stress on the
upon the size, location, and natural history of the AVM, and arterial walls creating further weakening with aneurysm for-
the patients risk-factors and medical comorbidities, the saf- mation. Venous aneurysms, or varices, form in the venous
est treatment may be no treatment at all [3–5]. system and can bleed, particularly when the vein distal to the
varix is stenotic, further increasing the pressure within the
venous aneurysm.
Arterial aneurysms are classified by location relative to
the AVM: (1) intranidal, (2) extranidal, and (3) flow related
(Fig. 48.2). The first two types are directly associated with
the vasculature of the AVM, whereas the flow-related aneu-
rysms are located elsewhere on the circle of Willis (COW)
E.M. Deshaies, M.D. (*) [24]. Intranidal aneurysms were reported in 13.6 % of AVMs,
Department of Neurosurgery, SUNY Upstate Medical University, extranidal aneurysms in 10.9 % of AVMs, with the latter
750 East Adams Street, Syracuse, NY 13210, USA
being flow-related 72 % of the time and in the COW 27 % of
e-mail: deshaieE@upstate.edu
the time [24]. The etiology of the AVM-associated aneu-
S. Komonchan, M.D.
rysms seems related to the abnormally high-flow within the
Department of Neurology, Prasat Neurological Institute,
312 Rajavithl Road, Rajthevee 10400, Thailand AVM vasculature, but it is less clear how the distant ones
606 E.M. Deshaies and S. Komonchan
Fig. 48.1 (a) Digital subtraction angiography demonstrating the three rather than blue, depicting early venous drainage from the
main components of an AVM: arterial feeding pedicles (small arrow), AVM. Additionally, cortical veins not associated with the AVM are blue
nidus (circle), draining vein (thick arrow) into the superior sagittal (small arrows) suggesting that the AVM does not drain through these
sinus. (b) Digital subtraction angiography using iFlow analysis veins. (c) Intraoperative visualization of well-demarcated flow (arrow)
(Siemens, software version 20) of an AVM; arterial (red) and venous between arterial (red) and venous (blue) blood in an arterialized cortical
(blue) phases. Notice that the superior sagittal sinus (SSS) is yellow vein associated with an AVM
Clinical Presentation and Initial Assessment
Patients with unruptured AVMs most often present with

chronic headaches or seizures, while ruptured ones present
with severe acute headache and neurological deficits [12, 21,
25]. Risk factors for AVM bleeding are often controversial
with conflicting reports in the literature. However, some of
the more accepted risk factors include prior hemorrhage,
associated aneurysm, venous varix, and venous outflow ste-
nosis, while nidus size remains controversial [15, 16, 23, 26].
In the case of patients who present with ruptured AVMs,
initial assessment can be artificially divided into two catego-
ries: hemorrhage with and without symptomatic mass effect.
Clearly, patients with AVM bleeding significant enough to
cause severe neurological deficit from mass effect should be
considered for operative intervention to evacuate hematoma
and resect the AVM. However, the extent of surgical resec-
tion in these emergent situations is often decided intraopera-
tively depending on the size and location of the lesion. For
example, the entire AVM may not be removed at the time of
Fig. 48.2 3D digital subtraction angiogram demonstrating AVM nidus
clot evacuation if it is located in eloquent regions because it
(circle) and an extranidal AVM-associated multilobulated aneurysm
(arrow) could exacerbate neurological deficits. In this situation, it
would be better to evacuate clot leaving the AVM behind to
limit further neurological deficits and prevent profuse bleed-
ing intraoperatively.
form. Though speculative, distant aneurysms may result In situations where there is no symptomatic mass effect,
from structural wall defects that arise either from congenital timing of treatment becomes more complex. Some neurosur-
or acquired etiologies that increase the susceptibility of these geons prefer to delay treatment until the clot retracts, cerebral
vessels to aneurysm formation. edema resolves, and elevated intracranial pressure subsides,
48 Arteriovenous Malformations: Viewpoint—Surgery 607
using only aggressive medical management as needed [27]. Table 48.1 Spetzler-Martin grading of AVMs [28]
Acute surgical intervention in this setting can make treat- Grading feature Points assigned
ment more precarious and potentially increase the risk of Size of nidus
dangerously elevating intracranial pressure. For example, if Small (<3 cm) 1
preoperative embolization is being considered in the acute Medium (3–6 cm) 2
phase, a second bleed, and an ischemic event or redirection Large (>6 cm) 3
of the flow in the AVM from vascular occlusion, could result Eloquence of adjacent brain
in additional edema and mass effect, forcing surgical inter- Non-eloquent 0
vention. Similarly, preoperative angiography in the setting of Eloquent 1
brain shift and edema may not allow full and accurate visu- Pattern of venous drainage
Superficial 0
alization of the AVM angioarchitecture, making complete
Deep 1
resection less likely and more difficult. Delaying interven-
tion until the hematoma, edema, and elevated intracranial
pressure resolve followed by elective noninvasive functional
Table 48.2 Spetzler-Martin grade and predicted surgical risk [28]
brain mapping and angiography will enable the neurosur-
Grade No deficit (%) Minor deficit (%) Major deficit (%)
geon to operate more safely with better outcomes than in the
1 100 0 0
acute scenario.
2 95 5 0
3 84 12 4
4 73 20 7
Imaging Modalities for Presurgical Planning 5 69 19 12
In order to appropriately choose candidates for surgical

resection, imaging modalities assessing the angioarchitec- Table 48.3 Supplementary grading system [30]
ture of the AVM, surrounding cortical eloquence, and anat- Grading feature Points assigned
omy of the white matter tracks can be helpful during Age
preoperative planning and discussions with the patient about <20 years 1
surgical risks. Not all patients with AVMs are surgical candi- 20–40 years 2
dates particularly with the advent of endovascular emboliza- >40 years 3
tion and SRS [6]. Here, we will discuss the utility of the Unruptured presentation
Spetzler-Martin grade and imaging modalities during preop- No 0
erative treatment planning. Yes 1
Diffuse
No 0
Yes 1
AVM Surgical Grading Systems
The AVM size (<3 cm, 3–6 cm, >6 cm), location (eloquent or
non-eloquent), and venous drainage pattern (deep or superfi-
cial) were found to be important AVM characteristics when outcomes associated with the SMG will be different for indi-
determining surgical risk and predicting postoperative neu- vidual surgeons depending upon their training, technical
rological outcomes [28]. The Spetzler-Martin grading sys- abilities, case volume, and experience. Additionally, interob-
tem (SMG) was developed using size, location, and venous server variability for the SMG was 28 %, with venous drain-
drainage patterns to assign a numeric score (1–5) with the age having the least interobserver variability and nidus size
higher number predicting higher surgical risk with worse having the greatest [29].
neurological outcomes (Table 48.1); an SMG of 6 has been A modification of the SMG was subsequently described
used to imply an “unresectable” AVM. This system was vali- where age, rupture status, and nidus diffuseness were consid-
dated and predicts that an SMG of “1” has a 0 % risk of defi- ered important characteristics for assessing surgical outcome.
cit from surgical resection, while an SMG of “5” has a 31 % This modification was published as the “supplementary
risk. The authors further divided surgical risk into “major” grading system” (SGS) developed by reviewing 300 patients
and “minor” neurological deficits (Table 48.2). Though the who underwent microsurgical AVM resection and determin-
SMG has been very useful to many surgeons planning for ing their modified Rankin score, SMG, and SGS (Table 48.3)
AVM resection, it does have pitfalls. For example, surgical [30, 31]. The predictive accuracy of the SGS for postsurgical
neurological deficits was higher (0.73) than the SMG (0.66). However, only catheter-based digital subtraction angiography
However, the SGS is intended to be an adjunct to the SMS (DSA) can demonstrate the anatomical and hemodynamic fea-
and not a substitute. Therefore, the modification includes tures of the AVM. Differentiating between arterial feeding
adding the SMG and SGS scores to obtain a combined score pedicles and draining veins can be challenging on the static
ranging from 0 to 10. Using this scoring system, a grade of imaging such as CTA and MRA (Fig. 48.3). Additionally,
1–3 predicts low surgical risk, 4–6 moderate risk, and 7–10 clearly identifying the extent of the AVM nidus on MRA and
high risk, for surgical AVM resection. CTA can also be challenging at times, particularly with diffuse
A three-tier classification system was recently described lesions. Static images, however, unlike DSA, depict the brain
in order to guide AVM management based on the SMG [32]. tissue relative to the AVM and are therefore very important for
AVMs were classified into “A” through “C” based on SMG determining the location of the AVM within the cerebral hemi-
with recommended management options according to the spheres. These static imaging modalities are particularly useful
classification (Table 48.4). The recommendations based for determining AVM proximity to eloquent cortical and
upon this classification system were Class “A” (SMG 1, 2), subcortical areas. Newer static imaging modalities such as
surgical resection; Class “B” (SMG 3), multimodality treat- functional MRI (fMRI) and diffusion tensor imaging (DTI) can
ment; and Class “C” (SMG 4, 5), no treatment. be particularly useful for preoperative planning.
Imaging Studies Functional MRI
Various imaging modalities can be used to depict the anatomical The emergence of FMRI in the 1990s and its continued
features of an AVM, including computed tomography angio- development enable clinicians to identify eloquent cortical
graphy (CTA), magnetic resonance angiography (MRA), and areas using real-time noninvasive mapping techniques. The
magnetic resonance venography (MRV) (Fig. 48.3). patient is presented with a series of tasks during an MRI and
the cortex can be mapped based on changes in blood flow
and oxygen delivery, making this a functional and physiolog-
Table 48.4 Three-tier classification of cerebral AVMs [32] ical test for detection of eloquent brain tissue. The proximity
Class Spetzler-Martin grade (SMG) Treatment of eloquent areas relative to the AVM can then be viewed
A 1 and 2 Surgical resection when planning the surgical approach and assessing surgical
B 3 Multimodality risk with the patient (Fig. 48.4). The combined use of nonin-
C 4 and 5 No treatment vasive static imaging (MRA, fMRI, CTA) and minimally
Fig. 48.3 Coronal section of a CTA (a) and MRA (b) depicting the to distinguish between arterial and venous structures and determining
three main architectural components of an AVM: arterial feeding pedi- the precise boundaries of the AVM nidus
cles, nidus, and venous drainage. These static images make it difficult
invasive dynamic imaging (DSA) allows the neurosurgeon to Diffusion Tensor Imaging
apply the SMG in a more precise manner when considering
potential treatment options [33]. For example, an AVM DTI is a relatively new form of MRI technology that uses
clearly located in the eloquent cortex or deep white matter water diffusion in the brain tissue to map the white matter
structures would likely result in an unacceptably high surgi- tracks (tractography) throughout the hemispheres (Fig. 48.6).
cal risk, making embolization or SRS a safer option. fMRI is The result is a color-coded map of the white matter tracts that
also helpful with AVM treatment planning because a small can be viewed relative to the location of the AVM. White
percentage of individuals exhibit cortical rearrangement, matter tracts originating from eloquent structures passing
such that functionality is not where one would typically through the AVM nidus could be an important deterrent to
expect. These functional anomalies could have important surgical resection. For example, surgical resection of an
implications for surgical planning when they are identified AVM nidus adjacent to the white matter tracts connecting
ahead of time and avoid unexpected poor outcomes from Broca and Wernicke areas may result in a conductive aphasia
neurological deficits (Fig. 48.5). from disconnection of these fibers. Another example would
be an AVM in Meyer’s loop resulting in visual field deficits
after resection. In this way, DTI tractography can be an
important tool for presurgical risk assessment and may guide
the decision towards a minimally invasive therapy that is less
likely to injure critical brain structures.
Presurgical Embolization
The first reported AVM embolization was performed by Drs.

Luessenhop and Spence in 1960. They reported near-total
angiographic occlusion of an AVM using methyl methacry-
late embospheres [34]. Since that time many different mate-
rials have been developed for embolization, including
calibrated particles, silk thread, absolute ethyl alcohol
(EtOH), and polyvinyl alcohol (PVA). Liquid embolic agents
have been developed with the advantage over particles
because they form a vessel cast resulting in better occlusion
and are less likely to migrate into the venous system when
there is a high-flow fistulae.
Fig. 48.4 fMRI depicting vascular flow voids from the AVM in the left One of the first and more commonly used embolic agents
frontal lobe anterior to the right-hand motor cortex (yellow blob) was the liquid adhesive N-butyl-2-cyanoacrylate (NBCA,
Fig. 48.5 fMRI demonstrating cortical rearrangement in a left-handed individual who was asked to tap his right fingers. Notice bilateral cortical
involvement (red blobs) with the left cortical activity lying deep, medial to the AVM flow voids
Fig. 48.6 DTI tractography demonstrating white matter tracks sur- left parieto-occipital AVM in a patient who presented with receptive
rounding the AVM (a) with corresponding MRI (b) allowing localiza- aphasia, visual field cut, and dyslexia after hemorrhage. The arrow in
tion of the AVM relative to critical brain structures. Depicted here is a (a) points to a disruption in the white matter tracts seen as a black void
Trufil, Codman), first introduced in the 1980s for intravascu-

lar embolization. This was followed by a nonadhesive agent
called ethylene vinyl alcohol (Onyx, eV3/Covidien
Neurovascular) in the 1990s, which occludes flow by pre-
cipitating in the vessel, allowing for a more controlled injec-
tion, longer working time, and lower risk of microcatheter
retention. Each of these embolic agents has its pros and cons,
which are discussed in greater detail elsewhere in this book.
Onyx has become the more popular liquid embolic agent
because its nonadhesive properties make it easier to work
with for reasons mentioned previously. Embolization alone
does not always cure AVMs and has a more important role as
an adjunctive treatment prior to surgical resection or
SRS. Complete angiographic obliteration of AVMs after
embolization has been reported in several studies in up to
57 % of cases [35–37]. However, AVMs have been reported Fig. 48.7 Onyx glue (black) identified in an embolized arterial feeding
to hemorrhage even with angiographic evidence of complete pedicle intraoperatively
embolization, supporting the notion that nests of the AVM
may remain despite “angiographic cure” [38–40]. This post-
embolization hemorrhage may be from small nidal nests that the AVM nidus throughout the resection and increase the
can’t be visualized angiographically or from AVM recur- difficulty of resection. Therefore, prior embolization can
rence secondary to recanalization of the occluded arterial reduce the complexity of surgery resulting in improved clinical
feeding pedicles. The low cure rate with current emboliza- outcomes. Data suggest that occlusion of >75 % of the AVM
tion materials and techniques makes endovascular therapy an nidus facilitates surgical resection, while elimination of <50 %
important adjunct to surgical resection or SRS, but not a helps little during surgical resection [45]. An additional benefit
practical sole treatment. of presurgical embolization is the easy identification of the
Presurgical embolization reduces the number of arterial feeding pedicles intraoperatively guiding the neurosurgeon to
feeding pedicles to the AVM ultimately making surgery safer the nidus, particularly those parts that are covered with brain
with less blood loss and shorter anesthesia time [7, 35, 41–44]. tissue and difficult to visualize initially (Fig. 48.7).
It is particularly useful for those larger AVMs with multiple Interestingly, bipolar cauterization of the Onyx casts
arterial feeding pedicles and those with deep feeding arteries causes sparks, with seemingly no risk to the patient.
that would not be accessible until the end of resection. These Additionally, unlike n-BCA, solidified Onyx is moderately
deep-seated arteries, when not embolized, continue to feed compressible allowing placement of AVM or aneurysm clips
during surgery. Glue casts also act as formidable arterial seizures. These issues ideally are addressed ahead of time to
plugs after transecting the arterial feeding pedicles during minimize complications and optimize surgical success and
nidus resection. patient outcome.
When using multimodality therapy, the cumulative risk of
AVM embolization (single or staged treatment) plus the risk
of surgical resection (determined by the SMG or SGS) must Anesthesia Considerations
be taken into account [7]. Complications from AVM emboli-
zation include transient (11.5 %) and permanent non- AVM resection is performed via craniotomy under general
disabling (2.6 %) neurological deficits and permanent anesthesia and has the potential for large volumes of blood
disabling deficits or death (1.6 %) [7, 38, 43, 46, 47]. The loss. Therefore, preoperative preparation should include
cumulative risk of embolization plus surgical resection must radial artery line for stringent blood pressure control, periph-
be discussed with the patient as part of their individualized eral or central intravenous access for crystalloid and colloid
surgical treatment plan. resuscitation, and a urinary catheter for fluid output monitor-
ing [49, 50]. The patient should be cross matched for at least
two units of blood and have them available for transfusion. A
Presurgical Stereotactic Radiosurgery specific range of systolic blood pressure or mean arterial
pressure should be stated clearly to the anesthesiologist
SRS is not often used as a presurgical adjunct to AVM resec- along with goals for hematocrit, end-tidal pCO2, and the type
tion. It generally takes 15–30 months to be effective, tempo- of anesthetic needed for neuromonitoring.
rarily increases the risk of bleeding over baseline, and affords By this point, the decision to use intraoperative neuro-
little or no protection from hemorrhage during the first cou- monitoring should have been made and the specific need for
ple of years. SRS can, however, reduce the size of the AVM SSEP, MEP, VEP, BAER, and cortical mapping discussed
and modify the structure of the arterial walls, making surgery with the monitoring team [51]. This is imperative because
easier to perform [48]. SRS is particularly useful as an the type of anesthetic (gas anesthesia versus total intrave-
adjunct to surgical resection postoperatively if residual or nous anesthesia or TIVA) can affect the neuromonitoring
inoperable AVM must be left behind. SRS can be helpful signals and if not attended to properly, could render them
when a patient presents with an AVM hemorrhage requiring useless [52].
emergent surgical evacuation from an otherwise inoperable
lesion. In this situation, clot evacuation and brain decom-
pression can be performed leaving the inoperable AVM for Tenets of Surgical Resection
SRS at a later date once the patient recovers. A more in-
depth discussion of SRS for AVMs can be found elsewhere Before draping the patient, femoral artery access should be
in this textbook. prepared and draped for possible intraoperative angiography
without contaminating the surgical field. Imaging studies
pertinent to the resection should be readily available for
Surgical Considerations viewing during the surgery along with neuronavigation and
multimodality image fusion.
Surgical success is not only dependent upon technical prow- The patient should be positioned to expose the area of
ess but upon perioperative medical management skills as interest and allow for a wide craniotomy that encompasses
well. A multidisciplinary approach increases the safety of the entire nidus and exposes the arterial feeding pedicles and
surgical resection and likelihood of good outcome. When the draining veins. Too little bone removal could result in a pre-
situation allows, preoperative assessment should include (1) carious situation where the feeding and draining vessels are
the patient’s primary care physician to address any comor- under the skull and unable to be accessed sufficiently. If neu-
bidities that would increase anesthesia and surgical risks ronavigation is being used, it can aid in planning the crani-
such as hypertension, pulmonary and cardiac disease; (2) otomy incision and bone removal to ensure adequate
anesthesia assessment focusing on intraoperative blood pres- exposure of all major components of the AVM. Additional
sure control, seizure prophylaxis, anesthetic type, and blood room may be needed if neuromonitoring grids will be placed
transfusion; (3) neuromonitoring team for cortical mapping for cortical mapping or simultaneous resection of an epilep-
in situations where the AVM is in proximity to the eloquent tic focus is planned.
cortex or to help guide resection of an epileptic focus; and General principles of AVM resection are to first occlude
(4) discussions with an intensivist about postoperative blood the arterial feeders to stop flow into the nidus followed by
pressure management, potential cerebral edema, stroke, and resection of nidus and occlusion of the venous outflow
Fig. 48.8 Surgical resection of an AVM in the occipital cortex fed by the occipital lobe is partially cauterized, and Onyx glue is seen in the
the right posterior cerebral artery. (a) The arterial feeding pedicles on artery between the bipolar tips. (c) Partially resected AVM with clips
the cortical surface contain blackish-gray Onyx glue before entering the seen in view alongside cauterized nidus. (d) Resection cavity after
nidus seen between the bipolar tips. (b) AVM nidus being resected from complete nidus resection
vessels (Fig. 48.8). Resecting in this order minimizes the risk tent arterial occlusion although the surgeon may consider
of AVM nidus rupture and severe bleeding. Temporary AVM adding a clip or further cauterize the vessel, depending on
or aneurysm clips can be used to test vessels when there is its size, as an additional precaution against bleeding from
uncertainty as to whether the vessel is an arterialized vein or recanalization.
arterial feeders that needs to be ligated. At times it can be If bleeding becomes difficult to control during resection
difficult to distinguish between the two types of abnormal of the AVM, the surgical approach should be reassessed for
vessels, and permanently occluding an arterialized vein early the premature occlusion of a venous outflow conduit or
in the resection could result in severe bleeding. AVM or entrance into the nidus itself. Bleeding from the nidus is dif-
small aneurysm clips and cauterization are used to perma- ficult to control because of the pressure from high blood flow
nently occlude the vessels. En passage cerebral vessels need and the pathological vessel walls, making them less respon-
to be preserved to avoid ischemic stroke, and these too can sive to cauterization. In this situation, the surgeon should
be tested with temporary aneurysm clips while watching for begin working peripherally, away from the actual nidus, until
changes in neuromonitoring signals. a complete resection can be performed. Irrigating bipolar
If preoperative embolization was performed, the glue cautery tips are recommended to minimize “sticking” of the
cast can be followed to the AVM nidus. Occlusive glue vessels to the tips resulting in tearing of the nidal vessels and
casts can be transected and left behind to ensure persis- worsened bleeding.
Fig. 48.9 Digital subtraction angiography of the AVM fed from the posterior cerebral and superior cerebellar arteries (a) before and (b) after
surgical resection
Intraoperative Angiography is likely the result of the acute hemodynamic changes that
occur within the local vasculature after AVM resection.
After a gross total resection of the AVM is suspected, intra- Systolic blood pressures should be kept within low-normal
operative angiography is recommended [53]. Small nests of ranges (90–120 mmHg) postoperatively, monitored by an
AVM nidus can remain after apparent “gross total resection” arterial line, and antihypertensive intravenous drips and pain
that may be difficult to visualize in pockets of white matter control used when necessary. Antiepileptic agents should
or the sulci. Intravenous fluorescein angiography can be per- also be continued after supratentorial AVM resection.
formed under the microscope if this feature is available to the Postoperative angiography should be obtained at this
surgeon, but small rests of AVM nidus enveloped by the point if not performed in the operating room to evaluate for
brain parenchyma likely will not be seen with this technique. residual AVM. If there is residual nidus, consider re-
Therefore, DSA is recommended instead and remains the exploration potentially with image guidance, endovascular
gold standard vascular imaging study (Fig. 48.9). In situa- embolization, or SRS.
tions where intraoperative angiography is not available, post-
operative angiography can be performed at a later time. This
is not ideal, however, because if residual AVM is not identi- Conclusion
fied until after surgery, re-exploration, endovascular emboli-
zation, or SRS will be required to treat the remnant. AVM treatment frequently requires multimodality therapy.
Thorough preoperative investigation with static imaging,
angiography, fMRI, and DTI will help to assess surgical risk.
Postsurgical Care Once the neurosurgeon has decided on AVM resection, pre-
operative embolization should be considered and periopera-
Postoperative evaluation should look for signs of cerebral tive treatment discussed with members of a multidisciplinary
edema, seizures, ischemic stroke, and hemorrhage from team who will be involved with the surgery. Neuromonitoring
“normal pressure breakthrough syndrome (NPBS)” [54]. capabilities for intraoperative monitoring and localization of
Typically these patients are observed in the intensive care an epileptic focus when necessary should be available during
unit or a unit where detailed neurological examinations by surgery. Postoperative care management in an ICU includes
the nursing staff and tight blood pressure control can be per- close neurological examinations by a trained nursing staff
formed safely. NPBS can be problematic when it occurs and and monitoring for seizure activity and NPBS.
20. Wilkins RH. Natural history of intracranial vascular malformations:

References a review. Neurosurgery. 1985;16(3):421–30.
21. McLaughlin MR, Kondziolka D, Flickinger JC, Lunsford S, Lunsford
LD. The prospective natural history of cerebral venous malforma-
1. Davidson AS, Morgan MK. The embryologic basis for the anatomy tions. Neurosurgery. 1998;43(2):195–200; discussion 200–201.
of the cerebral vasculature related to arteriovenous malformations. 22. Kondziolka D, Lunsford LD, Kestle JR. The natural history of cere-
J Clin Neurosci. 2011;18(4):464–9. bral cavernous malformations. J Neurosurg. 1995;83(5):820–4.
2. Di Rocco C, Tamburrini G, Rollo M. Cerebral arteriovenous mal- 23. Langer DJ, Lasner TM, Hurst RW, Flamm ES, Zager EL, King Jr
formations in children. Acta Neurochir (Wien). 2000;142(2): JT. Hypertension, small size, and deep venous drainage are associated
145–56; discussion 156–8. with risk of hemorrhagic presentation of cerebral arteriovenous mal-
3. Pandey P, Marks MP, Harraher CD, Westbroek EM, Chang SD, Do formations. Neurosurgery. 1998;42(3):481–6; discussion 487–9.
HM, et al. Multimodality management of Spetzler-Martin Grade III 24. Lv X, Wu Z, Li Y, Jiang C, Yang X, Zhang J. Cerebral arteriovenous
arteriovenous malformations. J Neurosurg. 2012;116(6):1279–88. malformations associated with flow-related and circle of Willis
4. Stapf C. The rationale behind “A Randomized Trial of Unruptured aneurysms. World Neurosurg. 2011;76(5):455–8.
Brain AVMs” (ARUBA). Acta Neurochir Suppl. 2010;107:83–5. 25. Josephson CB, Bhattacharya JJ, Counsell CE, Papanastassiou V,
5. Mohr JP, Moskowitz AJ, Stapf C, Hartmann A, Lord K, Marshall Ritchie V, Roberts R, et al. Seizure risk with AVM treatment or con-
SM, et al. The ARUBA trial: current status, future hopes. Stroke. servative management: prospective, population-based study.
2010;41(8):e537–40. Neurology. 2012;79(6):500–7.
6. Yashar P, Amar AP, Giannotta SL, Yu C, Pagnini PG, Liu CY, et al. 26. Stapf C, Mast H, Sciacca RR, Choi JH, Khaw AV, Connolly ES,
Cerebral arteriovenous malformations: issues of the interplay et al. Predictors of hemorrhage in patients with untreated brain arte-
between stereotactic radiosurgery and endovascular surgical ther- riovenous malformation. Neurology. 2006;66(9):1350–5.
apy. World Neurosurg. 2011;75(5–6):638–47. 27. Zacharia BE, Vaughan KA, Jacoby A, Hickman ZL, Bodmer D,
7. Muñoz F, Clavel P, Molet J, Castaño C, de Teresa S, Solivera J, Connolly Jr ES. Management of ruptured brain arteriovenous
et al. Current management of arteriovenous malformations. malformations. Curr Atheroscler Rep. 2012;14(4):335–42.
Retrospective study of 31 cases and literature review. Neurocirugia 28. Spetzler RF, Martin NA. A proposed grading system for arteriove-
(Astur). 2007;18(5):394–404. nous malformations. J Neurosurg. 1986;65(4):476–83.
8. Laakso A, Hernesniemi J. Arteriovenous malformations: epidemiol- 29. Du R, Dowd CF, Johnston SC, Young WL, Lawton MT. Interobserver
ogy and clinical presentation. Neurosurg Clin N Am. 2012;23(1):1–6. variability in grading of brain arteriovenous malformations using
9. Stapf C, Mast H, Sciacca RR, Berenstein A, Nelson PK, Gobin YP, the Spetzler-Martin system. Neurosurgery. 2005;57(4):668–75; dis-
et al. The New York Islands AVM Study: design, study progress, cussion 668–75.
and initial results. Stroke. 2003;34(5):e29–33. 30. Lawton MT, Kim H, McCulloch CE, Mikhak B, Young WL. A
10. Stapf C, Labovitz DL, Sciacca RR, Mast H, Mohr JP, Sacco supplementary grading scale for selecting patients with brain arte-
RL. Incidence of adult brain arteriovenous malformation hemor- riovenous malformations for surgery. Neurosurgery. 2010;66(4):
rhage in a prospective population-based stroke survey. Cerebrovasc 702–13; discussion 713.
Dis. 2002;13(1):43–6. 31. Rodríguez-Hernández A, Kim H, Pourmohamad T, Young WL,
11. Stapf C, Mohr JP, Pile-Spellman J, Solomon RA, Sacco RL, Lawton MT. Cerebellar arteriovenous malformations: anatomical
Connolly Jr ES. Epidemiology and natural history of arteriovenous subtypes, surgical results, and increased predictive accuracy of the
malformations. Neurosurg Focus. 2001;11(5):e1. supplementary grading system. Neurosurgery. 2012;71(6):1111–24.
12. Luo J, Lv X, Jiang C, Wu Z. Brain AVM characteristics and age. 32. Spetzler RF, Ponce FA. A 3-tier classification of cerebral arteriovenous
Eur J Radiol. 2012;81(4):780–3. malformations. Clinical article. J Neurosurg. 2011;114(3):842–9.
13. Kondziolka D, Humphreys RP, Hoffman HJ, Hendrick EB, Drake 33. Cannestra AF, Pouratian N, Forage J, Bookheimer SY, Martin NA,
JM. Arteriovenous malformations of the brain in children: a forty Toga AW. Functional magnetic resonance imaging and optical
year experience. Can J Neurol Sci. 1992;19(1):40–5. imaging for dominant-hemisphere perisylvian arteriovenous mal-
14. Fullerton HJ, Achrol AS, Johnston SC, McCulloch CE, Higashida formations. Neurosurgery. 2004;55(4):804–12; discussion 812–4.
RT, Lawton MT, et al. Long-term hemorrhage risk in children ver- 34. Luessenhop AJ, Spence WT. Artificial embolization of cerebral
sus adults with brain arteriovenous malformations. Stroke. 2005; arteries. Report of use in a case of arteriovenous malformation.
36(10):2099–104. J Am Med Assoc. 1960;172:1153–5.
15. Pollock BE, Flickinger JC, Lunsford LD, Bissonette DJ, Kondziolka 35. Liu HM, Huang YC, Wang YH. Embolization of cerebral arteriove-
D. Factors that predict the bleeding risk of cerebral arteriovenous nous malformations with n-butyl-2-cyanoacrylate. J Formos Med
malformations. Stroke. 1996;27(1):1–6. Assoc. 2000;99(12):906–13.
16. Hernesniemi JA, Dashti R, Juvela S, Väärt K, Niemelä M, Laakso 36. Panagiotopoulos V, Gizewski E, Asgari S, Regel J, Forsting M,
A. Natural history of brain arteriovenous malformations: a long- Wanke I. Embolization of intracranial arteriovenous malformations
term follow-up study of risk of hemorrhage in 238 patients. with ethylene-vinyl alcohol copolymer (Onyx). AJNR Am J
Neurosurgery. 2008;63(5):823–9; discussion 829–31. Neuroradiol. 2009;30(1):99–106.
17. Kim H, McCulloch CE, Johnston SC, Lawton MT, Sidney S, Young 37. van Rooij WJ, Jacobs S, Sluzewski M, Beute GN, van der Pol
WL. Comparison of 2 approaches for determining the natural his- B. Endovascular treatment of ruptured brain AVMs in the acute phase
tory risk of brain arteriovenous malformation rupture. Am J of hemorrhage. AJNR Am J Neuroradiol. 2012;33(6):1162–6.
Epidemiol. 2010;171(12):1317–22. 38. Lv X, Wu Z, Jiang C, Li Y, Yang X, Zhang Y, et al. Complication
18. Halim AX, Johnston SC, Singh V, McCulloch CE, Bennett JP, risk of endovascular embolization for cerebral arteriovenous mal-
Achrol AS, et al. Longitudinal risk of intracranial hemorrhage in formation. Eur J Radiol. 2011;80(3):776–9.
patients with arteriovenous malformation of the brain within a 39. Lv X, Wu Z, Li Y, Yang X, Jiang C. Hemorrhage risk after partial
defined population. Stroke. 2004;35(7):1697–702. endovascular NBCA and ONYX embolization for brain arteriove-
19. Laakso A, Dashti R, Seppänen J, Juvela S, Väärt K, Niemelä M, nous malformation. Neurol Res. 2012;34(6):552–6.
et al. Long-term excess mortality in 623 patients with brain 40. Lv X, Wu Z, Jiang C, Li Y, Yang X, Zhang Y, et al. Endovascular
arteriovenous malformations. Neurosurgery. 2008;63(2):244–53; treatment accounts for a change in brain arteriovenous malforma-
discussion 253–5. tion natural history risk. Interv Neuroradiol. 2010;16(2):127–32.
41. Xu F, Ni W, Liao Y, Gu Y, Xu B, Leng B, et al. Onyx embolization malformations with onyx: patient selection, embolization technique,
for the treatment of brain arteriovenous malformations. Acta and results. AJNR Am J Neuroradiol. 2012;33(7):1299–304.
Neurochir (Wien). 2011;153(4):869–78. 48. Steinberg GK, Chang SD, Levy RP, Marks MP, Frankel K,
42. Natarajan SK, Ghodke B, Britz GW, Born DE, Sekhar Marcellus M. Surgical resection of large incompletely treated intra-
LN. Multimodality treatment of brain arteriovenous malformations cranial arteriovenous malformations following stereotactic radio-
with microsurgery after embolization with onyx: single-center surgery. J Neurosurg. 1996;84(6):920–8.
experience and technical nuances. Neurosurgery. 2008;62(6): 49. Hashimoto T, Young WL. Anesthesia-related considerations for cere-
1213–25; discussion 1225–6. bral arteriovenous malformations. Neurosurg Focus. 2001;11(5):e5.
43. Weber W, Kis B, Siekmann R, Jans P, Laumer R, Kühne D. Preoperative 50. Logvinova AV, Litt L, Young WL, Lee CZ. Anesthetic concerns in
embolization of intracranial arteriovenous malformations with onyx. patients with known cerebrovascular insufficiency. Anesthesiol
Neurosurgery. 2007;61(2):244–52; discussion 252–4. Clin. 2010;28(1):1–12.
44. Yuki I, Kim RH, Duckwiler G, Jahan R, Tateshima S, Gonzalez N, 51. Gabarrós A, Young WL, McDermott MW, Lawton MT. Language
et al. Treatment of brain arteriovenous malformations with and motor mapping during resection of brain arteriovenous malfor-
high-flow arteriovenous fistulas: risk and complications associated mations: indications, feasibility, and utility. Neurosurgery. 2011;
with endovascular embolization in multimodality treatment. 68(3):744–52.
Clinical article. J Neurosurg. 2010;113(4):715–22. 52. Miller C, Mirski M. Anesthesia considerations and intraoperative
45. Viñuela F, Dion JE, Duckwiler G, Martin NA, Lylyk P, Fox A, et al. monitoring during surgery for arteriovenous malformations and dural
Combined endovascular embolization and surgery in the manage- arteriovenous fistulas. Neurosurg Clin N Am. 2012;23(1):153–64.
ment of cerebral arteriovenous malformations: experience with 101 53. Chalouhi N, Theofanis T, Jabbour P, Dumont AS, Gonzalez LF,
cases. J Neurosurg. 1991;75(6):856–64. Starke RM, et al. Safety and efficacy of intra-operative angiography
46. Jayaraman MV, Marcellus ML, Hamilton S, Do HM, Campbell D, in craniotomies for cerebral aneurysms and arteriovenous malfor-
Chang SD, et al. Neurologic complications of arteriovenous mal- mations: a review of 1093 consecutive cases. Neurosurgery. 2012;
formation embolization using liquid embolic agents. AJNR Am J 71(6):1162–9.
Neuroradiol. 2008;29(2):242–6. 54. Luessenhop AJ, Ferraz FM, Rosa L. Estimate of the incidence and
47. van Rooij WJ, Jacobs S, Sluzewski M, van der Pol B, Beute GN, importance of circulatory breakthrough in the surgery of cerebral
Sprengers ME. Curative embolization of brain arteriovenous arteriovenous malformations. Neurol Res. 1982;4(3–4):177–90.
Cerebral Arteriovenous
Malformations Viewpoint: 49
Endovascular Therapy Perspective
M. Yashar S. Kalani, Richard W. Williamson,

Felipe C. Albuquerque, and Cameron G. McDougal
have attempted to shed light on the annual risk of hemor-

Introduction rhage from cerebral AVMs and suggest that the risk of hem-
orrhage from AVMs is between 2 and 4 % [6, 7]. In the event
Arteriovenous malformations are some of the most treacher- of a hemorrhage, the risk subsequently increases to 7–17 %
ous lesions in medicine. Although initially postulated to be for the year immediately after the hemorrhage [8]. Although
congenital lesions, more recent data suggest that these mal- results vary greatly between series, each episode of hemor-
formations undergo dynamic remodeling, and can expand or rhage is associated with a 10–15 % and 15–30 % risk of mor-
involute [1]; rare cases of de novo formation have been bidity [9, 10]. Most practitioners agree that hemorrhage and
reported in the literature [2, 3]. The recent publication of the alterations in physiology of the AVM, as manifested by new
ARUBA trial [4] has led to much confusion on the role of neurological deficits or seizure, should lead to detailed
intervention versus medial management of these lesions. work-up and possibly intervention to eliminate the risk of
Although most practitioners agree that the results of the hemorrhage.
ARUBA trial are marred by design and execution flaws, the Although guidelines for treatment of AVMs are based on
study serves as a reminder that careful patient selection and expert opinion, the recently published randomized control
the availability of a multidisciplinary team with expertise in trial ARUBA attempted to address the role of intervention
neurological surgery, interventional techniques, and radia- versus medical therapy for unruptured AVMs [4]. Despite
tion therapy are cornerstones of successful treatment. best intentions, ARUBA had major shortcomings associated
with study design and execution. The ARUBA investigators
reviewed 1,740 patients aged 18 years or greater without a
Natural History history of hemorrhage for inclusion in the study. Of the
cohort reviewed, 726 patients fit inclusion criteria but 177
The decision to treat any pathology rests on a solid under- patients were treated outside the trial and 323 refused enroll-
standing of the natural history of the disease, the morbidity ment; 226 were actually included in the study. The authors
and mortality associated with intervention, and the expertise do not address the reasons for a large number of patients
of the practitioner(s) involved in the care of the patient. being treated outside the trial. A second critique of the trial
Therefore any discussion on treatment of AVMs depends on design is the lack of elaboration on what the study defined as
an understanding of the presentation and natural history of an “intervention group.” This group consisted of patients
these lesions in the general population. treated with surgery, endovascular embolization, and radio-
Hemorrhage is the most common presenting symptom of therapy, but few details are provided on the techniques of
patients with AVMs, followed by seizures and focal neuro- embolization and radiotherapy, use of embolysates, number
logical deficits, likely caused by microhemorrhage or steal of endovascular interventions, or the fractionation regimen
phenomena from adjacent brain [5]. Several large studies or dose of radiation. Furthermore, the study followed patients
for a mean of 33 months, a time point that is certainly inad-
equate for assessing lesions that leave patients with a life-
M.Y.S. Kalani, M.D., Ph.D. • R.W. Williamson, M.D. time risk for hemorrhage. These shortcomings have resulted
F.C. Albuquerque, M.D. (*) • C.G. McDougal, M.D. in many practitioners questioning the utility of the results of
Division of Neurological Surgery, Barrow Neurological Institute,
ARUBA [11, 12].
St. Joseph’s Hospital and Medical Center,
300 West Thomas Road, Phoenix, AZ 85013, USA In the absence of level I evidence, one has to turn to
e-mail: Yashar.kalani@bnaneuro.net; neuropub@dignityhealth.org expert guidelines for management. Spetzler and Martin [13]
618 M.Y. Kalani et al.
Table 49.1 The Spetzler–Martin [13] and the updated Spetzler–Ponce (ev3, Irvine, CA) has greatly expanded the indications and
grading scale [15] for cerebral arteriovenous malformations safety of this approach [18]. In rare circumstances, emboliza-
Description Points tion may be used as a curative modality for the complete
Size occlusion of AVMs. This strategy is associated with an
<3 cm 1 increased risk of complications, due to the aggressive nature
3–6 cm 2 of embolization needed to completely occlude the AVM
>6 cm 3 nidus, and is rarely the strategy of choice in most centers [18,
Eloquence
19]. A third indication for endovascular treatment is its appli-
Yes 1
cation as an adjunct before radiosurgery. Yet other indications
No 0
are the use of embolization for targeted treatment of high-risk
Deep venous drainage
features of AVMs (such as intranidal aneurysms) or as a pal-
Yes 1
No 0
liative tool to minimize patient symptomatology. In general,
the indication to use embolization in the treatment strategy
Reprinted from Spetzler RF, Martin NA. A proposed grading system for
arteriovenous malformations. J Neurosurg. 1986;65:473–83, with should take into consideration the added risk associated with
permission from the American Association of Neurological Surgeons this intervention. We adhere to the guidelines mentioned
Points are attributed to size, deep venous drainage, and eloquence of above and rarely embolize grade I or II AVMs. Embolization
brain involved. The sum of the points dictates the Spetzler–Martin is commonly performed for grade III AVMs and is utilized as
grade of the AVM. Experience suggests that grade I and II AVMs and
grade IV and V AVMs behave similarly and have similar morbidity and a part of multimodality regimen for more complex lesions.
mortality profiles. This led Spetzler and Ponce to suggest a modified
three-tier grading scheme where grades I and II (now grade A) and
grades IV and V (now grade C) lesions are combined, with grade III Preoperative Embolization
lesions re-classified as grade B
The goal of preoperative embolization is reduction in the

introduced a simple grading scheme based on angiographic number of arterial pedicles feeding the nidus of the AVM,
features, size, and location of AVMs that has provided a especially those vascular feeders that reside deep to the nidus
reproducible benchmark for assessing morbidity and mor- and are difficult to access early in the surgical approach
tality associated with surgical intervention for AVMs (Fig. 49.1). At our center we perform embolization proce-
(Table 49.1) [14]. Hamilton and Spetzler [14] reported mor- dures on the day prior to planned microsurgical resection.
bidity and mortality rates of 1 % and 3 % for treatment of Although the timing of embolization and surgery are contro-
low grade (I/II and III, respectively) AVMs. The combined versial, we are reluctant to leave the patient at risk for
morbidity and mortality rates for treating high-grade (IV hemorrhage after the hemodynamic changes induced by the
and V) AVMs approach 50 % in the immediate posttreat- embolization. In the post-embolization period and prior to
ment period and decline to 17–22 % at follow-up. Based on resection, patients with complex AVMs are kept sedated with
data on morbidity and mortality after treatment, current strict blood pressure control. Others have postulated that
guidelines suggest surgery for grade I and II AVMs, emboli- surgery should be performed in a delayed fashion (2–3 weeks
zation followed by surgery for most grade III AVMs, and after embolization) to allow for normalization of hemody-
observation for most grade IV and V AVMs [15]. Although namic parameters after embolization [20, 21].
this is a great oversimplification and each patient requires
independent evaluation based on angiographic features of
the AVM, AVM location, presence of high-risk features, the Curative Strategies
capabilities of the treating team, and patient and family
wishes, this paradigm for treatment serves as guideline for Angiographic cures have been achieved in a subset of AVMs
intervention. with embolization as a stand-alone therapy. Using
cyanoacrylate-based glues (e.g., N-butyl cyanoacrylate
[NBCA]), a complete occlusion rate of 20 % has been
Endovascular Treatment reported [19]; however, the more recent introduction of ethyl
vinyl alcohol copolymer (EVOH) or Onyx has increased
Endovascular embolization has become a pillar in the multi- total obliteration rates to as high as 51 % [18] among all
disciplinary treatment of AVMs [16, 17]. Endovascular treat- AVMs and up to 96 % for select AVMs with simple angio-
ment can be used as a preoperative adjunct for the curative graphic features [22]. AVMs most amenable to complete
resection of AVMs. This indication constitutes the largest use obliteration are those lesions with a small nidus, readily
of endovascular techniques for AVMs, and the advent of the visible draining vein, and few arterial pedicles. These char-
embolysate ethyl vinyl alcohol copolymer (EVOH) or Onyx acteristics are common among grade I and II AVMs, which
49 Cerebral Arteriovenous Malformations Viewpoint: Endovascular Therapy Perspective 619
Fig. 49.1 Embolization as an adjunct to operative resection. This Onyx was used to embolize a branch of the PCA feeding the nidus. AP
52-year-old male presented to our clinical service after an intraventricu- (g) and lateral (h) projections post-Onyx embolization reveal dimin-
lar and parenchymal hemorrhage. (a) Axial computed tomography (CT) ished flow to the AVM with preservation of the draining vein. The
demonstrates the bleed pattern. Axial (b), sagittal (c), and coronal (d) patient underwent microsurgical resection of the AVM via a transcorti-
CT angiography revealed an intraventricular AVM fed by branches of cal approach. AP (i) and lateral (j) ICA injections demonstrate minimal
the posterior cerebral artery (PCA) with deep venous drainage. residual nidus left apposed to the brainstem. The patient was scheduled
Anteroposterior (AP) (e) and lateral (f) right internal carotid artery to receive radiosurgery to the residual nidus. Used with permission from
(ICA) injections demonstrate that the primary arterial feeder of the Barrow Neurological Institute
nidus is the PCA and provides evidence of the early draining vein.
can readily be treated with surgery, with low resultant Pre-radiosurgery

morbidity and mortality. As such, we rarely attempt a com-
plete angiographic cure with embolization alone at our cen- Embolization can be selectively used in patients with deep-
ter. Another issue of concern with curative embolization is seated or complex lesions being considered for radiation ther-
the question of durability of an angiographic cure with apy (Fig. 49.2). Embolization can shrink the AVM nidus
embolysates. Experience with surgical resection suggests thereby decreasing the volume that must be targeted for ther-
that despite angiographic cure, residual blood flow is noted apy with radiation. Despite obvious benefits of downsizing
at the time of surgery through embolized vessels. and reducing the complexity of AVMs with embolization, the
Fig. 49.2 Embolization as an adjunct prior to radiotherapy. This tion in the size of the nidus. The lesion was felt to be amenable to
28-year-old male with a grade V AVM presented to our service with a microsurgical resection. (c) Anteroposterior angiography with vertebral
large diffuse lesion involving the internal capsule with deep venous artery injection demonstrates the size of residual nidus and pure super-
drainage. (a) Axial T2-weighted magnetic resonance (MR) scan reveals ficial drainage of the lesion. The patient underwent craniotomy with
the diffuse and extensive nature of the AVM nidus. The patient under- resection of the lesion. (d) Postoperative T1-weighted axial MR scan
went pre-radiosurgery embolization with Onyx. (b) T2-weighted axial reveals complete resection of the lesion. Used with permission from
MR images 3 years after radiosurgery demonstrate a significant reduc- Barrow Neurological Institute
Fig. 49.3 Targeted embolization of AVMs. In select cases angiographic Given the location of this lesion in the thalamus, it was felt that the
features of AVMs may predispose them to higher rates of hemorrhage. patient would benefit from treatment of high-risk features and close
High-risk features predisposing to hemorrhage include intranidal aneu- follow-up. (b) AP vertebral artery angiography demonstrates residual
rysms and venous outflow obstruction. (a) Anteroposterior (AP) angi- nidus after targeted embolization of the high-risk features with Onyx.
ography of the left vertebral artery demonstrates a thalamic AVM with Used with permission from Barrow Neurological Institute
multiple intranidal aneurysms fed from the left posterior cerebral artery.
embolysate may actually make targeting of lesions difficult Targeted Embolization

by causing distortion of the nidus, thereby reducing the rate
of AVM obliteration during the period of latency [23–25]. Targeted embolization can be used to treat angiographic fea-
The process of embolization also creates discrete foci of tures of an AVM that may predispose it to a higher risk of
AVM with intervening embolized nidus, which can make rupture. In general this strategy is used in cases when defini-
planning for radiosurgery more challenging [26]. Furthermore, tive treatment is not possible or is too risky. Features that
the possibility of recanalization of embolyzed segments leads increase the likelihood of rupture, so called high-risk fea-
to uncertainty of the fidelity of the treatment. Nonetheless, for tures, include intranidal aneurysms, outflow venous stenosis,
select complex AVMs, embolization and radiosurgery may be and the presence of a single deep draining vein (Fig. 49.3)
used as a last resort treatment option [27]. [28–30]. Although partial treatment of an AVM affords no
49 Cerebral Arteriovenous Malformations Viewpoint: Endovascular Therapy Perspective 621
Fig. 49.4 Palliative embolization of AVMs. High-grade AVMs not grade V AVM. A decision was made to embolize external carotid artery
amenable to multimodality therapy may present with symptoms sug- feeders (mainly from the middle meningeal artery and occipital artery)
gestive of steal. In select cases, embolization may result in improve- feeding the AVM. An unsubtracted image (c) demonstrates the Onyx
ment in symptomatology and quality of life. This patient presented with cast and lateral carotid angiography (d) demonstrates reduced flow of
worsening intractable headaches. It was felt that her symptoms were the nidus. The patient had significant improvement in headaches. Used
likely due to vascular steal. Lateral angiograms of the right external with permission from Barrow Neurological Institute
(a) and internal (b) carotid artery demonstrate a large right occipital
protection over the natural history of the lesion, treatment of 3. Stevens J, Leach JL, Abruzzo T, Jones BV. De novo cerebral arte-
riovenous malformation: case report and literature review. AJNR
high-risk features, such as intranidal aneurysms, decreases
Am J Neuroradiol. 2009;30(1):111–2.
the likelihood of hemorrhage [31]. 4. Mohr JP, Parides MK, Stapf C, et al. Medical management with or
without interventional therapy for unruptured brain arteriovenous
malformations (ARUBA): a multicentre, non-blinded, randomised
trial. Lancet. 2014;383(9917):614–21.
Palliative Embolization 5. Hofmeister C, Stapf C, Hartmann A, et al. Demographic, morpho-
logical, and clinical characteristics of 1289 patients with brain arte-
In select cases, AVMs may become symptomatic due to vas- riovenous malformation. Stroke. 2000;31(6):1307–10.
cular steal. In these instances, palliative embolization may 6. Hernesniemi JA, Dashti R, Juvela S, Vaart K, Niemela M, Laakso
A. Natural history of brain arteriovenous malformations: a long-
decrease steal, thereby palliating the symptoms (Fig. 49.4)
term follow-up study of risk of hemorrhage in 238 patients.
[32]. Although data in support of this paradigm are limited to Neurosurgery. 2008;63(5):823–9; discussion 829–31.
small series, in select cases palliative embolization does result 7. Ondra SL, Troupp H, George ED, Schwab K. The natural history of
in improvement in quality of life of the patient [31, 33, 34]. symptomatic arteriovenous malformations of the brain: a 24-year
follow-up assessment. J Neurosurg. 1990;73(3):387–91.
8. Itoyama Y, Uemura S, Ushio Y, et al. Natural course of unoperated
intracranial arteriovenous malformations: study of 50 cases.
Conclusions J Neurosurg. 1989;71(6):805–9.
9. Laakso A, Dashti R, Juvela S, Isarakul P, Niemela M, Hernesniemi
J. Risk of hemorrhage in patients with untreated Spetzler-Martin
More so than any other lesion, AVMs require an interdisciplin-
grade IV and V arteriovenous malformations: a long-term follow-
ary team with expertise in microsurgery, endovascular emboli- up study in 63 patients. Neurosurgery. 2011;68(2):372–7; discus-
zation, and radiation therapy. Despite the disappointing results sion 378.
of the ARUBA trial, mainly due to flaws in study design and 10. Laakso A, Hernesniemi J. Arteriovenous malformations: epidemi-
ology and clinical presentation. Neurosurg Clin N Am. 2012;
execution, a preponderance of evidence suggests that carefully
23(1):1–6.
selected patients gain benefit from aggressive and timely resec- 11. Lawton MT, Abla AA. Management of brain arteriovenous malfor-
tion of AVMs. These lesions should be treated in high-volume mations. Lancet. 2014;383(9929):1634–5.
centers where the necessary expertise is available. 12. Russin J, Spetzler R. Commentary: The ARUBA trial. Neurosurgery.
2014;75(1):E96–7.
13. Spetzler RF, Martin NA. A proposed grading system for arteriove-
nous malformations. J Neurosurg. 1986;65(4):476–83.
References 14. Hamilton MG, Spetzler RF. The prospective application of a grad-
ing system for arteriovenous malformations. Neurosurgery.
1. Lasjaunias P, Berenstein A, ter Brugge K. Clinical vascular anat- 1994;34(1):2–6; discussion 6–7.
omy and variations, Surgical neuroangiography. Berlin: Springer; 15. Spetzler RF, Ponce FA. A 3-tier classification of cerebral
2001. arteriovenous malformations. Clinical article. J Neurosurg. 2011;
2. Fujimura M, Kimura N, Ezura M, Niizuma K, Uenohara H, 114(3):842–9.
Tominaga T. Development of a de novo arteriovenous malformation 16. Chang SD, Marcellus ML, Marks MP, Levy RP, Do HM, Steinberg
after bilateral revascularization surgery in a child with moyamoya GK. Multimodality treatment of giant intracranial arteriovenous
disease. J Neurosurg Pediatr. 2014;13(6):647–9. malformations. Neurosurgery. 2007;61(1 Suppl):432–42; discus-
sion 442–434.
17. Pandey P, Marks MP, Harraher CD, et al. Multimodality manage- nous malformations. Neurosurgery. 2012;71(6):1139–47; discus-
ment of Spetzler-Martin Grade III arteriovenous malformations. sion 1147–38.
J Neurosurg. 2012;116(6):1279–88. 26. Shtraus N, Schifter D, Corn BW, et al. Radiosurgical treatment
18. Saatci I, Geyik S, Yavuz K, Cekirge HS. Endovascular treatment of planning of AVM following embolization with Onyx: possible
brain arteriovenous malformations with prolonged intranidal Onyx dosage error in treatment planning can be averted. J Neurooncol.
injection technique: long-term results in 350 consecutive patients 2010;98(2):271–6.
with completed endovascular treatment course. J Neurosurg. 27. Blackburn SL, Ashley Jr WW, Rich KM, et al. Combined endovascu-
2011;115(1):78–88. lar embolization and stereotactic radiosurgery in the treatment of large
19. Yu SC, Chan MS, Lam JM, Tam PH, Poon WS. Complete oblitera- arteriovenous malformations. J Neurosurg. 2011;114(6):1758–67.
tion of intracranial arteriovenous malformation with endovascular 28. Brown Jr RD, Wiebers DO, Forbes GS. Unruptured intracranial
cyanoacrylate embolization: initial success and rate of permanent aneurysms and arteriovenous malformations: frequency of intracra-
cure. AJNR Am J Neuroradiol. 2004;25(7):1139–43. nial hemorrhage and relationship of lesions. J Neurosurg. 1990;
20. Natarajan SK, Ghodke B, Britz GW, Born DE, Sekhar LN. 73(6):859–63.
Multimodality treatment of brain arteriovenous malformations with 29. da Costa L, Wallace MC, Ter Brugge KG, O’Kelly C, Willinsky
microsurgery after embolization with onyx: single-center experi- RA, Tymianski M. The natural history and predictive features of
ence and technical nuances. Neurosurgery. 2008;62(6):1213–25; hemorrhage from brain arteriovenous malformations. Stroke.
discussion 1225–16. 2009;40(1):100–5.
21. Starke RM, Meyers PM, Connolly ES. Adjuvant endovascular man- 30. Kader A, Young WL, Pile-Spellman J, et al. The influence of
agement of brain arteriovenous malformations. In: Youmans JR, hemodynamic and anatomic factors on hemorrhage from cerebral
Winn HR, editors. Youmans neurological surgery. Philadelphia: arteriovenous malformations. Neurosurgery. 1994;34(5):801–7;
Elsevier; 2011. p. 4058. discussion 807–8.
22. van Rooij WJ, Jacobs S, Sluzewski M, van der Pol B, Beute GN, 31. Krings T, Hans FJ, Geibprasert S, Terbrugge K. Partial “targeted”
Sprengers ME. Curative embolization of brain arteriovenous mal- embolisation of brain arteriovenous malformations. Eur Radiol.
formations with onyx: patient selection, embolization technique, 2010;20(11):2723–31.
and results. AJNR Am J Neuroradiol. 2012;33(7):1299–304. 32. Han PP, Ponce FA, Spetzler RF. Intention-to-treat analysis of
23. Andrade-Souza YM, Ramani M, Beachey DJ, et al. Liquid Spetzler-Martin grades IV and V arteriovenous malformations: nat-
embolisation material reduces the delivered radiation dose: a ural history and treatment paradigm. J Neurosurg. 2003;98(1):3–7.
physical experiment. Acta Neurochir. 2008;150(2):161–4; 33. Batjer HH, Devous Sr MD, Seibert GB, et al. Intracranial
discussion 164. arteriovenous malformation: relationships between clinical and
24. Kano H, Kondziolka D, Flickinger JC, et al. Stereotactic radiosur- radiographic factors and ipsilateral steal severity. Neurosurgery.
gery for arteriovenous malformations after embolization: a case- 1988;23(3):322–8.
control study. J Neurosurg. 2012;117(2):265–75. 34. Kusske JA, Kelly WA. Embolization and reduction of the “steal”
25. Schwyzer L, Yen CP, Evans A, Zavoian S, Steiner L. Long-term syndrome in cerebral arteriovenous malformations. J Neurosurg.
results of gamma knife surgery for partially embolized arteriove- 1974;40(3):313–21.
Radiosurgery for Cavernous
Malformations and Other Vascular 50
Diseases
Ajay Niranjan, Greg Bowden, John C. Flickinger,

and L. Dade Lunsford
Cavernous malformations are considered to be vascular

Intracranial Cavernous Malformations hamartomas composed of closely approximated endothelial-
lined sinusoidal collections without significant amounts of
Intracranial cavernous malformations are angiographically interspersed neural tissue. The lack of intervening neural tis-
occult malformations of the vascular bed. These abnormal sue distinguishes these lesions from capillary telangiecta-
vascular connections although usually congenital may sias. De novo pathogenesis may occur spontaneously or in
enlarge over time from bleeding events or even develop de association with initiating events, such as biopsy, a preexist-
novo [1, 2]. The lesions can even occur on a familial basis ing venous malformation, or perhaps following radiation
[3, 4]. Patients may be asymptomatic, although they often therapy. Pathologically, 9–88 % of cavernous malformations
present with headaches, seizures, or parenchymal hemor- show evidence of recent and remote hemorrhage depending
rhages. Substantial advances in the cellular and molecular on source and bleed criteria [5]. Evidence of this is from
biology of the vasculogenesis, angiogenesis, and cardiovas- hemosiderin-laden macrophages, cholesterol crystals, and
cular physiology of these anomalies along with findings hemosiderin-stained parenchymal tissues. Cavernous mal-
from detailed clinicopathologic and clinicoradiologic retro- formations are discrete multilobulated lesions and grossly
spective and prospective longitudinal studies have led to a resemble small mulberries. Clots and blood products in vari-
better understanding of these vascular malformations as a ous stages of evolution can be seen within these lesions.
whole. However, controversy still remains regarding the Calcification and gliosis are also often seen. The architecture
treatment options. of the component vessels consists of a single layer of endo-
thelium and varying quantities of subendothelial fibrous
stroma. Smooth muscle and elastic fibers are absent.
Pathophysiology Re-endothelialization of the hemorrhagic cavities, growth of
new blood vessels, and proliferation of granulation tissue
Cavernous malformations can be found in any part of the may account for the apparent growth of some cavernous
brain because they can occur at any location along the vascu- malformations.
lar bed. Most (80–90 %) of the lesions are supratentorial, and
the frontal and temporal lobes are the most common sites.
The deep cerebral white matter, corticomedullary junction, Epidemiology
and basal ganglia are common supratentorial sites, whereas
the pons and cerebellar hemispheres are common posterior Cavernous malformations represent approximately 1 % of
fossa sites. intracranial vascular lesions and 15 % of cerebrovascular
malformations although these numbers may be underestima-
tions [6]. In early studies of major autopsy reports, the calcu-
A. Niranjan, M.D., M.B.A. (*) • G. Bowden, M.D., M.Sc. lated prevalence was 0.02–0.53 %. With the advent of MRI,
Department of Neurological Surgery, University of Pittsburgh cavernous malformations are currently the most commonly
Medical Center, 200 Lothrop Street, Pittsburgh, PA 15213, USA
e-mail: niranjana@upmc.edu identified brain vascular malformations. The detection of
previously unidentified asymptomatic lesions by MRI has
J.C. Flickinger, M.D. • L.D. Lunsford, M.D.
Department of Radiation Oncology, University of Pittsburgh, recently raised the estimated overall prevalence to 0.45–0.9 %
5230 Centre Avenue, Pittsburgh, PA 15213, USA [7–9]. Multiple lesions are seen in approximately 15–33 % of
spontaneous cases, although one series reported an incidence slowly, the lesions may also produce seizures and a variety
as high as 50 %. Multiple lesions are more common in the of neurological findings similar to those expected of intra-
familial form, occurring in as many as 84 % of patients [4, 10]. cranial tumors. Once patients become symptomatic,
The familial form of the disorder is inherited as an autosomal 40–50 % present with seizures, 20 % present with focal
dominant trait with variable expression. Cavernous malfor- neurological deficits, and 10–25 % present with hemor-
mations are the most common CNS vascular malformation rhage. Symptoms may progress rapidly, remain stable for
subtype in patients with mixed vascular lesions. Associated years, or wax and wane. Headache is estimated to be a
developmental venous anomalies (DVAs) are present in symptom in as many as 25 % of patients. Acute headaches
approximately 10–30 % of patients with cavernous malfor- may result from parenchymal irritation secondary to gross
mations [11]. or repeated extralesional hemorrhage. Chronic headaches
could be the result of mass effect in slow-growing larger
lesions as a result of multiple intralesional hemorrhages.
Molecular Basis The clinical symptoms usually result from the location of
the lesion and, at times, their slow expansion. Infratentorial
Although most cavernous malformations are believed to be location and previous gross hemorrhage are associated with
sporadic, many familial cases have been observed over the an increased risk of subsequent and progressive neurological
last 2 decades. These cases exhibit an autosomal dominant disability. Large hemorrhages can cause both obstructive and
pattern of inheritance [12] and seem to affect the Hispanic nonobstructive hydrocephalus. Any hemorrhage found on CT
population in particular. Recent research has demonstrated at scans in a relatively young patient should be investigated
least three separate genes (CCM1, CCM2, and CCM3) further, and cavernous malformation must be considered a
related to the familial form of the disease. Although research possible etiology. In the workup of a patient with a seizure
is still pending, the three CCM protein products appear disorder, cavernous malformation must be considered as a
involved in new blood vessel stabilization. Linkage analyses potential cause, especially if the patient is aged 20–40 years.
using autosomal dominant families manifesting CCMs have The bleeding rate is estimated to be 0.5–2.5 % per lesion-year
identified three different causative loci on chromosomes and 0.25–16.5 % per patient-year [11]. This risk is increased
7q21.2 (CCM1), 7p13 (CCM2), and 3q25.2-q27 (CCM3). in patients with established prior hemorrhage. A study at the
Mutations in the gene Krit1 are responsible for CCM1, muta- University of Pittsburgh concluded that in patients who had
tions in the gene MGC4607 are responsible for CCM2, and at least one prior hemorrhage, the subsequent annual risk of
mutations in the gene PDCD10 were recently reported to be hemorrhage was 33.9 %. After the first bleed, annual hemor-
responsible for CCM3. Of familial cavernous malforma- rhage rates in years 1–5 were 52 %, 35 %, 39 %, 24 %, and
tions, 40 % can be linked to a protein created by the mutated 32 %, respectively [17]. In addition, women are at a slightly
CCM1 gene. This is the gene responsible for most of the higher risk for hemorrhage, especially. However, pregnancy
cases of familial multiple cavernous malformations in does not appear to change the rate of hemorrhage events com-
Mexican-American families and in a number of other fami- pared to natural history [18, 19]. The hemorrhage is rarely
lies. CCM1 is responsible for creating KRIT1 protein or life-threatening. Hemorrhages in critical locations can have
Krev interaction-trapped one protein. The exact function of more severe effects, and thus, they are more likely to produce
KRIT1 protein is not known. If both copies of the CCM1 symptoms. Progressive neurological deficits are more often
gene mutate, the KRIT1 protein cannot function and associated with cavernous malformations in the infratentorial
cavernous malformations form [13]. CCM2 gene controls space and with lesions that demonstrate slow enlargement
the production of a protein named malcavernin. Of familial because of rebleeding episodes.
cavernous malformations, 20 % can be linked to a CCM2
mutation [14–16].
Radiologic Features
Clinical Presentation Although cavernous malformations can be diagnosed using

CT scans, MRI is the modality of choice for the long-term
Cavernous malformations can occur at any age, but they follow-up of patients as well as the assessment of family
are most likely to become clinically apparent in patients members in whom similar lesions are suspected. Cavernous
aged 20–40 years. Patients with cavernous malformations malformations can be divided in three components: the
may remain asymptomatic, but they most often present peripheral pseudocapsule composed of gliotic hemosiderin-
with headache or neurological symptoms after a hemorrhage laden tissue, the irregular intersecting connective tissue
or repeated hemorrhage. Because of the extruded blood septa separating the sinusoidal spaces, and the central vas-
products and the fact that some malformations can enlarge cular area composed of slow-flowing sinusoidal spaces [20].
50 Radiosurgery for Cavernous Malformations and Other Vascular Diseases 625
MRI findings of parenchymal cavernous malformations The rate of these complications must be carefully balanced
demonstrate typical, popcorn-like, well-circumscribed, against the natural history of the lesion itself (Table 50.1)
well-delineated lesions. The core is formed by multiple foci [2, 4, 7, 10, 11, 22–25]. It is important to ascertain the pres-
of mixed signal intensities, which represents hemorrhage in ence of DVAs, as these are composed of functional venous
various stages of evolution. Acute hematoma containing channels that should not be resected or radiated because of
deoxyhemoglobin is isointense on T1-weighted images and the risk for venous infarction. Increasingly, DVAs are identi-
markedly hypointense on T2-weighted images. Subacute fied in association with cavernous malformations [6].
hematoma, which contains extracellular methemoglobin, In older patients or patients with low-risk lesions which
displays hyperintensity on both T1- and T2-weighted images are difficult to reach surgically, observation with periodic
because of the paramagnetic effect of the methemoglobin. imaging may be appropriate. When patients present with
The interspersed fibrous-containing elements demonstrate recurrent hemorrhage, progressive neurological deteriora-
mild hypointensity on both T1- and T2-weighted images tion, or intractable epilepsy, then treatment in the form of
because they contain a combination of calcification and surgery or radiosurgery should be considered. The manage-
hemosiderin. The heterogeneous core typically is sur- ment options for a patient with a cavernous malformation
rounded completely by a low-signal-intensity hemosiderin must be based on age, symptoms, hemorrhage risk, and loca-
rim on T1-weighted images. The hypointensity of this rim tion, especially as it relates to surgical accessibility. Resection
becomes more prominent on T2-weighted images because clearly provides immediate advantages for accessible lesions
of the magnetic susceptibility effects. Smaller cavernous and complete resection is achieved in 91 % (62 % bleed
malformations may appear as focal hypointense nodules rate in partially resected lesions) [26]. However, surgery on
with both T1- and T2-weighted sequences. Typically, cav- average has a 45 % postoperative and 15 % long-term mor-
ernous malformations are not associated with mass effect or bidity [26]. These results can be much higher when resection
edema and do not demonstrate a feeding artery or draining of deep-seated lesions is conducted (Table 50.2) [27–33].
vein, except when associated with other vascular malfor- For younger patients with more accessible lesions, the
mations. Cavernous malformations can be associated with removal of lesions may help control epilepsy, improve neu-
DVAs, which typically demonstrate a prominent medullary rological deficits, and prevent any subsequent hemorrhage.
vein with the classic caput medusae pattern of drainage. Radiosurgery is a minimally invasive option for patients
MRI has largely replaced conventional angiography in the presenting with recurrent hemorrhages from deep-seated
diagnosis of cavernous malformations. However, when the cavernous malformations.
lesions occur in combination with other types of vascular
malformations, as they do in as many as 30 % of patients,
MRI characteristics become more complicated and less spe- The Role of Radiosurgery
cific. In these patients, angiography can help further define
the lesions. Because of the extremely slow flow of blood The successful management of cerebral AVMs with stereo-
through these lesions, cerebral angiograms are often normal. tactic radiosurgery prompted the exploration of its role in the
Most cavernous malformations (37–48 %) are avascular management of cavernous malformations. The benefits of
masses on conventional angiograms. The avascular appear- radiosurgery for cavernous malformations are difficult to
ance is the result of compression or destruction of vascular assess because of its unclear natural history and the lack of
channels by hemorrhage, thrombosis, and generalized slow an imaging technique that can document a “cure.” The role
flow because of the small size of the connecting sinusoidal of radiosurgery in this disease is still considered controver-
vessels with the peripheral normal parenchymal vessels. sial by many physicians. However, the lack of options for
When lesions are smaller and not associated with hemato- surgically inaccessible cavernous malformations has made
mas, 20–27 % of angiograms demonstrate normal findings. radiosurgery a possible alternative to conservative manage-
Capillary blush which is not a specific finding is demon- ment for lesions with a high risk of symptomatic bleeding
strated at 12–20 % [21]. (Table 50.3) [17, 34–40].
Management Options Dose Planning Technique for Cavernous

Malformation Radiosurgery
When a patient is diagnosed with a cavernous malformation,
the reason that led to the detection of the lesion must be fac- All patients must have stereotactic magnetic resonance imag-
tored into the final decision as to what, if any, treatment ing. We prefer contrast-enhanced MRI using gradient
should be offered. Severe complications can occur after treat- recalled acquisitions (1-mm slice thickness) supplemented
ment of lesions in critical locations in the brain and brainstem. with T2-weighted sequences. The malformation is defined as
Table 50.1 Natural history of cavernous malformations

First author Patients FU length Lesion location Clinical presentation Annual bleeding risk Comment
Robinson 66 patients 26 months Supratentorial: 55 Seizures: 34 patients 0.7 % per lesion – HRG common in
[7] (1991) patients female and in
Deep: 4 patients Focal deficit: 30 patients infratentorial lesions
Cerebellar: 9 patients HRG: 6 patients
Brainstem: 8 patients Incidental: 9 patients
Zabramski 31 patients 2.2 years Supratentorial: 92 % Symptomatic: 61 % HRG + symptom: 6 % – Developed new
[4] (1994) Multi: 84 % Infratentorial: 8 % Incidental: 39 % HRG overall: 13 % cavernoma: 29 %
Aiba [22] 110 4.71 years Supratentorial: 65 % HRG: 62 patients Re-HRG: 22.9 % per – Younger and female
(1995) patients lesion patients had higher
Deep: 11 % Seizures: 25 patients HRG risk
Brainstem: 18 % Incidental: 23 patients
Kondziolka 122 34 months BG/thalamus: 17 % HRG = 1: 41 % Total: 2.6 % – No influence on
[23] (1995) patients location, sex, and
Multi: 20 % Brainstem: 35 % HRG > 1: 9 % No HRG: 0.6 % number on HRG risk
Seizures: 23 % Previous HRG: 4.5 %
Porter [2] 173 46 months Superficial: 109 HRG: 25.4 % 4.2 % annual rate – Significantly greater
(1997) patients patients
Multi: 17.9 % Deep: 64 patients Seizure: 35.8 % 10.6 % deep location Bleed rate for deep
Neurological deficit: 20.2 % ~37 % recovered from
Incidental: 12.1 % new deficits
Labauge 40 patients 3.2 years Supratentorial: 176 HRG: 19 patients 11 % per patient – 1/3 bleeds,
[11] (2000) symptomatic
Multi: 93 % Cerebellar: 30 Seizures: 12 patients 2.5 % per lesion – 27.5 % of patients
develop new lesions
Brainstem: 26 Focal signs: 4 patients
Kupersmith 37 patients 4.9 years – 12 midbrain HRG: 73 % 2.5 % per patient – Young patient with
[24] (2011) – 18 pons Mass effect only: 22 % 5.1 % rebleeding risk lesion of >10 mm had
– 7 medulla Asymptomatic: 5 % higher risk of HRG
Al-Holou 92 patients 3.5 years Supratentorial: 75 Acute symptoms: 17 HRG: 1.6 %/year – Brainstem location
[10] (2012) patients patients
Multi: 30 % Infratentorial: 1 Chronic symptoms: 5 Re-HRG: 8 %/year and age increased
patient patients
Incidental: 34 0.2 % if incidental bleed rate
Salmon 139 5 years Lobar: 67 % Incidental: 47 % HRG: 2.4 %/year – Risk declines over
[25] (2012) patients
Multi: 17 % Brainstem: 14 % Seizure: 25 % Re-HRG: 29.5 %/year 5 years and is higher
Cerebellum: 13 % HRG: 12 % HRG year 1: 19.8 % for women
Deep structures: 6 % Neurological deficit: HRG year 5: 5 %
15 %
HRG hemorrhage, ARE adverse radiation effect, BG basal ganglia
the region characterized by mixed signal intensity within an to prescribe margin dose. Associated DVAs are excluded
outer hemosiderin ring, typified by low signal intensity. from dose planning (Fig. 50.2). Selection of the radiosurgery
Hematoma eccentric from the malformation is excluded dose is dependent upon malformation volume and brain
from dose planning. The prescription isodose covers the location. Typically, the dose is less than that used for AVMs
mixed signal intensity malformation but not the low-signal- of similar volumes. The margin dose to the cavernous mal-
intensity region surrounding it. Targeting within the MRI- formation margin varies from 12 to 20 Gy, which is depen-
defined T2 signal hemosiderin rim is very important because dent upon the brain volume and location. Patients usually
radiation delivered to the iron-impregnated gliotic brain sur- receive a single dose (40 mg) of methylprednisolone at the
rounding the CM may increase the risk of adverse radiation conclusion of the radiosurgery procedure and, if pertinent,
effects if this area is included in prescription isodose [41]. maintain their current dose of antiepileptic medication.
Single or multiple isocenter dose plans are used to construct Follow-up imaging studies and clinical evaluations are
a conformal irradiation volume that matches the malforma- requested at 6-month intervals for the first 2 years following
tion margin (Fig. 50.1). The 50 % or greater isodose is used radiosurgery and then annually (Fig. 50.3).
Table 50.2 Summary of reports on resection of deep-seated cavernous malformations

Immediate postoperative
First author Patients Location Presentation outcome Long-term outcome
Steinberg [27] 56 patients Brainstem: 42 HRG: Average of 2.1 16 % improved 43 % improved
(2000) Thalamus: 5 55 % stable 52 % stable
Basal ganglia: 10 29 % worse 5 % worse
No death 4 patients had
re-HRG
Mathiesen [28] 68 patients Basal ganglia: 11 Varied with location 25/29 complete Re-HRG: 4 partial
(2003) resection resections
Thalamus: 12 HRG 1 or 2: 68 patients 69 % neuro-decline 20 patients improved
Mesencephalon: 5 Only 2 (5 %) permanent (10 patients intact)
Pons: 31 4 stable, 5 worse
Medulla: 5 1 died of rebleed
Wang [29] (2003) 137 Midbrain: 29 HRG = 1: 45 patients 131/137 total resection Re-HRG: 3 partial
patients resections
Pons: 85 HRG > 1: 58 patients 72.3 % stable or improved Death: 1
Medulla: 19 HRG 3 or more: 34 27.7 % new deficits 7.8 % living
patients dependently
Cerebellar peduncle: 8 60 % annual rebleeding
Ferroli [30] 52 patients Medulla: 7 HRG = 1: 32 patients 4 repeat surgeries 19 % permanent
(2005) Pontomedullary: 3 HRG > 1: 18 patients 29 patients stable deficits
Pons: 31 No HRG: 2 patients 23 patients worse
Pontomesencephalic: 3 HRG: 3.8 % annual risk 1 death
Midbrain: 6 Re-HRG: 34.7 % annual
Hauck [31] (2009) 44 patients Pons: 25 HRG = 1: 20 patients Improved: 30 % Residual: 1
Midbrain: 11 HRG > 1: 23 patients Stable: 59 % Re-HRG: 2 patients
Medulla: 8 Worse: 11 % Revisions: 8 patients
Abla [32] (2011) 260 Pons: 112 HRG: 252 patients New deficits: 53 % Death: 3
patients Medulla: 29 New deficits: 7 patients Residual lesion: 29 New deficit: 36 %
Midbrain: 49 Re-HRG: 7.7 Improved: 45 %
Pontomedullary: 40 Died: 2 Re-HRG: 2.0 %
Pontomesencephalic: 31
Pandley [33] 176 Pons: 94 HRG = 1: 70 patients Second operation: 17 Death: 3
(2012) patients patients
20 pediatric Midbrain: 28 HRG > 1: 102 patients New deficit: 31.3 % Improved: 61.8 %
Medulla: 14 Seizure: 4 patients Improved: 19.9 Stable: 25.9 %
BG/thalamus: 43 Re-HRG: 5 patients
HRG hemorrhage, ARE adverse radiation effect, BG basal ganglia
Results of Cavernous Malformation cavernous malformations. Kondziolka et al. reported the

Radiosurgery results of radiosurgery in 47 patients who harbored a hemor-
rhagic malformation in a critical intraparenchymal location
The results of multiple studies indicated that radiosurgery is an [43]. Of these, 44 patients had experienced at least two hemor-
option for some patients with deep-seated cavernous malfor- rhages before radiosurgery. At a mean follow-up of 3.6 years
mations. Kida et al. reported 20 cases of symptomatic angio- after radiosurgery, these authors reported significant reduc-
graphically occult vascular malformations treated using tions in the proportion of patients with hemorrhage and the
radiosurgery [42]. Among 20 lesions, 14 were located supra- mean number of hemorrhages per patient. The annual hemor-
tentorially, 4 in the brainstem, and 2 in the cerebellar hemi- rhage rate in the first 2 years after radiosurgery was 8.8 %
spheres. The lesions were treated using 15–20 Gy at the lesion which further decreased to 1.1 % in the 2- to 6-year interval
margins. These investigators reported a significant reduction after radiosurgery. After radiosurgery, 12 patients (26 %) sus-
of rebleeding risk, as well as improvement in seizure control. tained neurological worsening, of which eight improved on
Adverse effects were generally mild and well controlled by medications, two underwent surgical resection and died, and
medication. Thus, the preliminary results indicated a certain two had new permanent deficits. This study provided further
usefulness of radiosurgery in the treatment of symptomatic support for the role of radiosurgery in selected patients.
Table 50.3 Summary of some recent Gamma Knife radiosurgical series of cavernous malformations
First author Patients/FU Location Presentation Dose/volume Outcome Morbidity
Karlsson 22 patients Cortical: 8 HRG: 16 patients Margin: 18 Gy Post-GK HRG rate: ARE: 6 patients
[34] (1998) (9–35) 8 %/year
Deep cerebral: 7 Epilepsy: 6 patients – Years 1–2: 10 % – 5 symptoms
GK 6.9 years Brainstem: 6 Prior resection: 1 – Years 3–4: 12 % Onset 16 months
patient
Cerebellum: 1 – Years 5–6: 5 %
– Years 7–8: 7 %
Hasegawa 82 patients Supratentorial: HRG: 82 patients Margin: 16.2 Gy HRG: 12.3 % first 2 ARE: 11 patients
[17] (2002) 16 (1–7 bleeds) (12–20) years
Deep: 13 HRG rate: 33.9 %/ HRG: 0.76 % years None after 1992
year 2–12
GK 4.89 years Brainstem: 52 Prior surgery: 20 Vol.: 1.85 cm3
patients (0.12–6.98)
Liu [35] 125 patients Brainstem: 49 HRG: 112 patients Margin: 12.1 Gy HRG rate: 6.5 %/year ARE: 13.1 %
(2005) (9–20)
BG/thalamus: 14 HRG >1: 45 HRG: 10.3 % first 2 Symptoms: 2.5 %
patients years
GK 5.4 years Cortical: 39 HRG rate: 29.2 %/ Vol.: 3.12 cm3 HRG: 3.3 % after
year (.032–25.9) year 2
Cerebellum: 10 Seizures: 28
patients
Kim [36] 42 patients Cortical: 21 HRG: 26.6 % Margin: 14.55 Gy Seizure cessation: Edema: 5 patients
(2005) (10–25) 75 %
Brainstem: 6 Seizures: 28.6 % – Mean of – Symptoms: 5
31.3 months patients
GK 29.6 months Basal ganglia: 5 Neurological HRG: 1 – Recovered: 5
deficit: 45.2 % patients
Liscak [37] 112 patients Brainstem: 33 HRG in 59 patients Margin: 16 Gy HRG rate: 1.6 %/year Edema: 30
(2005) (9–36) patients
Cortical: 54 Seizures in 40 Vol.: 0.9 cm3 – 2 deaths from – Symptoms: 17
patients (0.06–12.5) HRG patients
GK 48 months BG/thalamus: 17 Neurological Neurological deficits: – Recovered: 16
deficit: 51 patients 33 % improved patients
Cerebellum: 8 7 patients prior
partial resection
Nagy [38] 113 patients Brainstem: 79 HRG high risk >1: Margin: 12–15 Gy High risk <2 years: ARE: 6 symptoms
(2010) 41 patients (10–20) 15 %/year
BG/thalamus: 39 HRG low risk <1: Vol.: High risk >2 years:
54 patients 0.33–0.83 cm3 2.4 %/year
GK 48 months Neurological deficit Low risk <2 years:
high risk: 72 % 5.1 %/year
Neurological deficit Low risk >2 years:
low risk: 43 % 1.3 %/year
HRG pre GK:
30.5 %/year
Lunsford 103 patients Brainstem: 66 HRG >1: 103 Margin: 16 Gy HRG <2 years: ARE: 19 patients
[39] (2010) patients (12–20) 10.8 %/year
BG/thalamus: 27 HRG pre GK: Vol.: 1.31 cm3 HRG > 2 years: 1 %/ Symptoms: 12
32.5 %/year year patients
GK 67.8 months Deep lobar: 10 Seizures: 8 patients 75 % seizure
cessation
Lee [40] 49 patients Pons: 34 HRG >1: 49 Margin: 11 Gy HRG <2 years: ARE: 3 patients
(2012) patients (9–15) 3.33 %/year
Midbrain: 10 HRG pre GK: Vol.: 3.2 cm3 HRG > 2 years: Symptoms: 2
31.3 %/year 1.74 %/year patients
GK 40.6 months Medulla: 3 Neurological Cyst: 1 patient
deficits: 49 patients
HRG hemorrhage, Vol. volume, ARE adverse radiation effect, BG basal ganglia
Fig. 50.1 Gamma Knife radiosurgery dose plan for cavernous malfor- struction covers the malformation but stays within the low-signal
mation of the brainstem. Fifty percent 14-Gy isodose line projected on region. An 8-Gy line is also present for perspective
axial contrast-enhanced MR images with sagittal and coronal recon-
Fig. 50.2 Gamma Knife radiosurgery dose plan for cavernous malfor- MR images with sagittal and coronal reconstruction covers the malfor-
mation of the left thalamus with a history of multiple symptomatic mation. Multiple developmental venous anomalies were noted
bleeds. Fifty percent isodose line projected on axial contrast-enhanced
Fig. 50.3 A 28-year-old female presented with history of two hemor- 16 Gy was delivered to 50 % isodose line (middle). A 4-year follow-up
rhagic episodes from cavernous malformation in her anterior right MRI shows regression in cavernous malformation with stable appear-
pons. Axial MR gadolinium-enhanced MR (left) has shown pontine ance of the brainstem
cavernous malformation at the time of radiosurgery. A margin dose of
Pollock et al. evaluated the efficacy and safety of Kim et al. evaluated the efficacy of radiosurgery in
radiosurgery for patients with cavernous malformations [44]. symptomatic cavernous malformations for which the surgical
Seventeen patients underwent radiosurgery for high-surgical- risk was thought to be unacceptable. These investigators
risk cavernous malformations that were located in eloquent treated the first 11 cases with LINAC radiosurgery using
regions (the thalamus/basal ganglia, brainstem, and corpus CT-based planning and the next 11 cases with GK radiosur-
callosum). All patients had at least two documented hemor- gery using MR-based planning. The volume of the lesion
rhages before undergoing radiosurgery. The median margin ranged from 0.09 to 4.8 cm3 (mean, 1.42 cm3) and the mean
radiation dose was 18 Gy and the median maximum dose marginal dose was 16.1 Gy (range, 8–24 Gy). The median
was 32 Gy. The annual hemorrhage rate during the 51 months follow-up period after radiosurgery was 38.3 months. In the
preceding radiosurgery was 40.1 %, compared with 8.8 % in group with prior hemorrhage, the bleeding rate of cavernous
the first 2 years following radiosurgery and 2.9 % thereafter. malformation after radiosurgery (1.55 %/year) was lower
These investigators reported a high (41 %) permanent than that during the pre-radiosurgical period (35.5 %/year).
radiation-related morbidity rate with 18 Gy as margin dose. Six patients showed neurological deterioration following
Hasegawa et al. studied outcome of Gamma Knife radio- radiosurgery; however, the neurological deficits persisted in
surgery for 82 symptomatic patients who had imaging- only two of the patients treated with LINAC. These authors
confirmed hemorrhages for which resection was believed to suggested a possible correlation of LINAC radiosurgery with
be associated with high risk [17]. Most patients had multiple radiation-induced neurological deficits. The size of the
hemorrhages from the brainstem or diencephalic cavernous lesion decreased in 11 patients, increased in one, and did not
malformations. During an observation of 354 patient-years change in ten cases. A perilesional hyperintense signal on
prior to radiosurgery, 202 hemorrhages were observed, for an T2-weighted MR images was seen in nine patients and
annual hemorrhage rate of 33.9 %, excluding the first hemor- correlated with the treatment modality used, being more fre-
rhage. After radiosurgery, during a total of 401 patient-years, quent after LINAC radiosurgery [45]. Kim et al. evaluated
19 hemorrhages were identified, of which 17 occurred in the 42 cavernous malformation patients treated with Gamma
first two posttreatment years, and only two after 2 years had Knife surgery with 14.55 Gy as mean margin dose [36]. The
elapsed. The annual hemorrhage rate was 12.3 % per year for mean follow-up period after radiosurgery was 29.6 months
the first 2 years after radiosurgery. The annual hemorrhage (range, 5–93 months). The tumor decreased in size in 29
rate dropped to 0.76 % per year after 2 years. Eleven patients cases, was unchanged in 12, and increased in size in one. In
developed new neurological symptoms without hemorrhage the seizure group, seizures were controlled without anticon-
after radiosurgery (13.4 %). The symptoms were minor in six vulsant medication in nine cases (81.8 %) after 31.3 months.
of these patients and temporary in five. These investigators Rebleeding occurred in only one case (2.3 %). On
concluded that radiosurgery conferred a reduction in the risk T2-weighted imaging, changes were seen in 11 cases
of hemorrhage for high-risk cavernous malformations. This (26.2 %), and neurological deterioration was correlated with
reduction was most pronounced after 2 years [17]. imaging changes in three (7.1 %). These authors concluded
that patients who receive a marginal dose below 15 Gy had any conclusion regarding the mechanism of efficacy [48].
better outcome than those treated with higher doses [36]. However, a recent case report suggests that the histopatho-
Liscak et al. followed 107 patients treated with Gamma logical findings after irradiation on these lesions may be
Knife radiosurgery using an average of 16-Gy margin dose similar to those described in arteriovenous malformations
[37]. After a median of 48 months, lesion regression was after Gamma Knife surgery [49].
observed in many patients. These authors reported a tran- A significant number of centers have published data indi-
sient morbidity (rebleeding and edema) rate of 20.5 % and cating a significant decrease in the symptomatic hemorrhage
permanent morbidity rate of 4.5 % [37]. rate following stereotactic radiosurgery of cerebral cavern-
Nagy et al. published on 113 patients with primarily ous malformations [38, 40, 50–52]. This demonstrates that
brainstem and thalamic lesions that were followed over a this technique may be an effective management strategy for
mean of 48 months [38]. This group demonstrated a hemor- patients with hemorrhagic malformations in high-risk
rhage rate of 30.5, which is compatible with other recent brain locations when administered by skilled personnel.
studies [40]. They separated malformations into high and The patients that were selected for radiosurgery harbored
low risk based on whether they had greater than 1 bleed pre- malformations with a rate after two hemorrhages of
viously. The higher-risk group had a hemorrhage rate of 30–32.5 % annual risk for another symptomatic event in con-
15 % for 2 years that decreased to 2.4 % after that time. The trast to a 0.6 % annual risk in those who had not had a prior
low-risk group had 5.1 and 1.3 % risks for the first 2 years symptomatic hemorrhage [38–40]. There is a subset of
and beyond, respectively. Adverse symptomatic outcomes patients who seem to bleed more frequently than others [22,
were observed in six patients. This demonstrates the hetero- 23, 53]. The reduced hemorrhage rate achieved after a 2-year
geneity of cavernous malformation activity and impact of latency from radiosurgery (1 %) approaches the low risk
previous bleeds. Lundsford et al. presented a review of 103 (0.6 %) identified in many natural history studies for
patients with deep lesions, including 65 % within the brain- nonhemorrhagic cavernous malformations. The probability
stem [39]. Two or more bleeds were required and pre Gamma of developing post-radiosurgery imaging changes compati-
Knife bleed rate was 32.5 %. The 67.8-month follow-up ble with ARE depends on the CM marginal dose and treat-
demonstrated a 10.8 % hemorrhage rate in the first 2 years ment volume. The volume of tissue receiving 12 Gy or more
which decreased to 1 % per year after that point. (the 12-Gy volume) provides predictive information about
One of the most frequent presentations of cavernomas is the probability of developing such imaging changes [41]. In
seizure. The ability to control these episodes or eliminate addition, care must be taken to achieve a highly conformal
them is a major impetus for treatment. Baumann et al. pro- treatment area and use caution around the hemosiderin depo-
vided information from the surgical perspective by analyzing sition zone at the edge of the cavernous malformation.
168 patients who underwent resection for supratentorial soli- All reports on cavernous malformation radiosurgery are
tary cavernous malformations with at least three prior sei- based on clinical outcome because cavernous malformation
zure episodes [46]. At 3 years 65 % of patients were seizure obliteration cannot be documented using any of the available
free and on no medication (Engel 1); these results were imaging modality. Without this documentation, patients can-
roughly in context with results seen by other sources [47]. not be declared as “cured” after radiosurgery. Although
Lunsford et al. and Kim et al. have both published from the patients may be encouraged by a reduction in hemorrhage
Gamma Knife perspective with results of 75 % of patients at risk, they cannot be told that this risk is eliminated. Even
Engel 1 at 3.6 years and 31.3 months, respectively [36, 39]. though radiosurgery significantly lowers the risk of hemor-
These results are based on a combined 20 patients. rhage, there can be complications. The complex natural his-
tory of cavernous malformations (solitary, multiple, and de
novo) makes clarity even harder to obtain. These risks and
Radiobiological Considerations benefits must be carefully balanced against the natural history
of untreated lesions if the use of radiosurgery is considered.
Post-radiosurgery AVM obliteration is likely due to endothe-
lial cell proliferation leading to vessel wall hyalinization,
fibrosis, and myofibroblast-induced vessel wall contracture. Radiosurgery for Other Vascular
We believe that the response of cavernous malformation to Abnormalities
irradiation is similar. It is reasonable to expect a latency
interval of several years for hemorrhage protection as it Dural Arteriovenous Fistulas
occurs before AVM obliteration. At present, there are no
good histologic studies of irradiated cavernous malforma- Dural arteriovenous fistulas (AVFs) represent approximately
tions in patients who were thought to be successfully man- 10–15 % of all intracranial vascular abnormalities [54]. They
aged. Histologic studies of irradiated patients who later are direct connections between the meningeal arterial system
underwent cavernous malformation resection could not draw and dural venous sinuses or cortical veins. Most are thought
Fig. 50.4 Cerebral angiogram revealed a dural vascular malformation to MR images supplemented with cerebral angiograms. Gamma plan shows
the sigmoid sinus. This malformation was treated with radiosurgery fol- dose plan projected on axial MR poster with sagittal and coronal recon-
lowed by embolization. Radiosurgery dose planning was performed on struction. A margin dose of 20 Gy was delivered to 50 % isodose line
to be acquired lesions, secondary to dural venous channel obstructed sinus, while type IIb fistulas drain retrogradely
occlusion leading to recruitment of collateral blood supply. into leptomeningeal veins through a venous sinus. Type
Trauma, infection, and hypercoagulable states are risk fac- IIa + b combines the features of both previous categories.
tors for the development of dural AVFs. The majority involve Type III fistulas involve direct connections between menin-
the transverse, sigmoid, and cavernous sinuses. These are geal arteries and cortical veins, without intervening dural
generally low-flow lesions, which present with bruits and sinus. Type IV is the same as type III, with added variceal
pulsatile tinnitus. Cavernous sinus lesions can also present dilation of the cortical veins. Type V represents a fistula
with chemosis and ophthalmoplegia and, eventually, lead to draining into perimedullary veins. The risk of bleeding
loss of vision. Other notable locations include the superior increases directly with the fistula grade for types I–IV. Type
sagittal sinus and tentorium, and these tend to follow a more V fistulas usually present with myelopathy.
aggressive course. Regardless of their locations, dural AVFs Treatment of dural AVFs varies according to the location
can cause subarachnoid or intraparenchymal hemorrhage. and grade of the lesion. Endovascular embolization through a
The bleeding risk is related primarily to the presence of lep- transarterial or transvenous route is the mainstay of treatment
tomeningeal venous drainage, dural sinus occlusion, and for simple, low-flow lesions involving the transverse-sigmoid
aneurismal venous dilation [55]. or cavernous regions, with high cure rates [57]. High-risk
Dural AVFs can be classified based on the pattern of lesions are usually treated with multimodality management
venous drainage and patency of dural sinus involved [56]. comprising embolization and microsurgery. Stereotactic
Type I fistulas drain in an anterograde fashion in a patent radiosurgery has emerged as a minimally invasive adjunct or
dural sinus. Type IIa fistulas drain retrogradely in an alternative management modality for AVFs (Fig. 50.4).
Yang et al. published results out of Pittsburgh on 40 Pan et al. published the results of Gamma Knife radio-
patients treated for 44 dural AVFs [58]. The average volume surgery as primary treatment of patients (n = 20) with
was 2.0 cm3 and the median dose was 21 Gy. Three groups transverse-sigmoid AVFs [62]. Tinnitus and headache were
were examined: Gamma Knife therapy alone, Gamma Knife the most common symptoms. Three patients had previously
with embolization performed together, and delayed Gamma bled. Two patients had failed prior surgery and embolization.
Knife after previous embolization. The follow-up was 45 Marginal doses ranged from 16.5 to 19 Gy. Complete oblit-
months and 28 patients with 32 dural AVFs demonstrated eration associated with resolution of symptoms was seen in
obliteration. Gamma Knife alone had an obliteration rate of 58 %, with an additional 16 % having near-total occlusion
67 %. When coupled with embolization, 83 % was achieved. after median follow-up of 19 months. Temporary hair loss
The delayed radiosurgery group saw results of 71 % likely over the mastoid region was the only adverse effect of irra-
due to difficulty identifying peripheral areas treated with diation, seen in ten patients.
embolization material. The authors came to the conclusion Pollock et al. managed 20 patients with cavernous sinus
that small low-flow dural AVFs could be managed with fistulas using radiosurgery alone (seven patients) or fol-
Gamma Knife radiosurgery alone if necessary [58]. lowed by embolization (13 patients) [63]. Clinical symp-
Hanakita et al. performed a retrospective analysis of 22 toms improved in 95 % of patients. Follow-up angiogram
patients treated with Gamma Knife radiosurgery [59]. Fifty- was obtained for 15 patients and demonstrated total oblit-
five percent of patients had obliteration of their dural AVF eration in 13 and near-total in one patient. No subsequent
within 5 years. The obliteration rate without cortical venous hemorrhage occurred. Morbidity was seen in two patients
drainage was 86 %, but only 47 % with cortical drainage. No following embolization procedures. Finally, Lewis et al.
hemorrhages or complications were identified. They indi- reported the results of transarterial embolization followed
cated that Gamma Knife is a viable option in the multidisci- by LINAC radiosurgery (seven patients) or surgery (two
plinary but should be considered carefully when cortical patients) in a cohort of patients with high-risk tentorial dural
venous drainage is present. AVFs [64]. Five patients had experienced prior hemorrhage.
Koebbe et al. described 18 dural AVF patients managed In the radiosurgical group, dose delivered ranged from 8 to
with Gamma Knife radiosurgery [60]. Three patients pre- 20 Gy (mean, 15.6 Gy). Four of seven patients showed com-
sented with intracerebral hemorrhage. Embolization was plete thrombosis of the lesion 24 months after the procedure,
performed in ten patients (prior to radiosurgery in nine and and the others had significant occlusion. One patient suffered
after in two patients). Margin dose ranged from 15 to 30 Gy from transient radiation injury to the brainstem 14 months
(mean, 20 Gy). The mean nidus volume was 2.16 cm3. after radiosurgery.
Complete resolution of symptoms was seen in nine patients, The outcome data suggests that stereotactic radiosurgery
with the nine others experiencing significant symptom provides significant benefit in the management of dural AVFs.
reduction. Follow-up imaging demonstrated fistula oblitera- The obliteration rate is high with minimal radiation-induced
tion in 12 patients (eight on angiography and four on MRA) morbidity. It is a valuable addition in the multimodality man-
and decrease in size in three patients. Three patients did not agement of high-risk dural AVFs and also as primary treat-
have follow-up imaging. No further hemorrhage was seen ment of low-flow dural AVFs of the transverse, sigmoid, and
after the Gamma Knife procedure. One patient experienced cavernous regions, with or without adjunct embolization.
transient symptomatic radiation-induced edema, and two Radiosurgery followed by embolization appears to be a sound
patients with cavernous sinus fistulas had a permanent com- strategy for patients with bothersome symptoms such as tin-
plication from embolization (facial numbness and sixth nitus. Radiosurgery should be performed first because the
nerve palsy). Clinical outcomes were better for patients whole nidus can be better defined and treated prior to emboli-
treated with radiosurgery alone compared to the combined zation. It is followed by embolization, which can provide
treatment. immediate relief from acute symptoms.
Friedman et al. reported the outcome of 23 patients with
transverse-sigmoid fistulas treated with combined radiosur-
gery followed by transarterial embolization in most cases Vein of Galen Malformations
[61]. Symptoms completely resolved in 87 % of patients,
with 9 % also experiencing significant relief. Out of 17 Vein of Galen malformations (VGMs, also known as vein of
patients who had follow-up angiography, seven (41 %) were Galen aneurysms) are rare vascular abnormalities that are
cured, four (24 %) had more than 90 % obliteration, and six usually found in the pediatric population. They constitute
(35 %) had more than 50 % obliteration of the fistula. No approximately 1 % of all intracranial vascular malformations
patient had radiation-induced complication and one patient [65]. They are direct fistulous connections between the cere-
suffered a transient embolization-related ischemic episode. bral arterial vasculature and the galenic venous system, often
Two patients had ischemic events following diagnostic angi- associated with straight sinus stenosis or occlusion and per-
ography (transient in one case and permanent in the other). sistence of a falcine sinus. Clinical presentation is dependent
on the age at which it manifests. Neonates usually show deterioration with or without hemorrhage. J Neurosurg.
signs of overt high-output cardiac failure, while older patients 1997;87(2):190–7.
3. Kivelev J, Niemelä M, Kivisaari R, Dashti R, Laakso A,
can present with hydrocephalus or intracranial hemorrhage. Hernesniemi J. Long-term outcome of patients with multiple cere-
Yasargil classified VGM into four different types based on bral cavernous malformations. Neurosurgery. 2009;65(3):450–5;
the arterial supply of the fistula [66]. Type I malformations discussion 455.
are direct anastomosis between branches of the distal poste- 4. Zabramski JM, Wascher TM, Spetzler RF, Johnson B, Golfinos J,
Drayer BP, Brown B, Rigamonti D, Brown G. The natural history
rior cerebral artery or the pericallosal artery and the vein of of familial cavernous malformations: results of an ongoing study.
Galen. Type II malformations constitute connections between J Neurosurg. 1994;80(3):422–32.
thalamoperforating arteries and the vein of Galen. Type III 5. Dalyai RT, Ghobrial G, Awad I, Tjoumakaris S, Gonzalez LF,
combines the features of types I and II. Type IV malforma- Dumont AS, Chalouhi N, Randazzo C, Rosenwasser R, Jabbour
P. Management of incidental cavernous malformations: a review.
tions are classic parenchymal AVMs with venous drainage in Neurosurg Focus. 2011;31(6):E5.
the galenic system and are not true VGM. 6. McLaughlin MR, Kondziolka D, Flickinger JC, Lunsford S,
Treatment of VGM usually consists of staged endovascu- Lunsford LD. The prospective natural history of cerebral venous
lar embolization, with the intent of progressively reducing malformations. Neurosurgery. 1998;43(2):195–200; discussion
200–1.
the arteriovenous shunt and avoiding complications of cere- 7. Robinson JR, Awad IA, Little JR. Natural history of the cavernous
bral venous congestion. Neonates with cardiac decompensa- angioma. J Neurosurg. 1991;75(5):709–14.
tion need to be medically stabilized and have urgent 8. Washington CW, McCoy KE, Zipfel GJ. Update on the natural his-
treatment, but elective embolization can be done for stable tory of cavernous malformations and factors predicting aggressive
clinical presentation. Neurosurg Focus. 2010;29(3):E7.
patients. Microsurgery is technically demanding and is best 9. Kondziolka D, Monaco EA, Lunsford LD. Cavernous malforma-
reserved for type I and simple type III malformations which tions and hemorrhage risk. Prog Neurol Surg. 2013;27:141–6.
are not amenable to endovascular procedures [67]. 10. Al-Holou WN, O’Lynnger TM, Pandey AS, Gemmete JJ, Thompson
There have been few reports of radiosurgical management BG, Muraszko KM, Garton HJL, Maher CO. Natural history and
imaging prevalence of cavernous malformations in children and
of VGM. Watban et al. reported three infantile cases which young adults. J Neurosurg Pediatr. 2012;9(2):198–205.
were successfully obliterated by transarterial embolization 11. Labauge P, Brunereau L, Lévy C, Laberge S, Houtteville JP. The
18–24 months after having failed radiosurgery [68]. Margin natural history of familial cerebral cavernomas: a retrospective
doses used varied from 20 to 25 Gy. Payne et al. studied the MRI study of 40 patients. Neuroradiology. 2000;42(5):327–32.
12. Cavalcanti DD, Kalani MYS, Martirosyan NL, Eales J, Spetzler
outcomes of Gamma Knife radiosurgery in a series of nine RF, Preul MC. Cerebral cavernous malformations: from genes to
patients with VGM [69]. The population consisted of eight proteins to disease. J Neurosurg. 2012;116(1):122–32.
children (aged from 4 to 14 years) and one adult. Four 13. Bacigaluppi S, Retta SF, Pileggi S, Fontanella M, Goitre L, Tassi L,
patients presented with hemorrhage, two with increased head La Camera A, Citterio A, Patrosso MC, Tredici G, Penco S. Genetic
and cellular basis of cerebral cavernous malformations: implica-
circumference, two with headaches, and one with hydro- tions for clinical management. Clin Genet. 2013;83(1):7–14.
cephalus. Three patients had a Yasargil type I lesion; one, a 14. Liquori CL, Berg MJ, Squitieri F, Ottenbacher M, Sorlie M,
type II lesion; two, type III lesions; and three, type IV lesions. Leedom TP, Cannella M, Maglione V, Ptacek L, Johnson EW,
Four had failed prior endovascular treatment. Three patients Marchuk DA. Low frequency of PDCD10 mutations in a panel of
CCM3 probands: potential for a fourth CCM locus. Hum Mutat.
required two Gamma Knife procedures. The mean marginal 2006;27(1):118.
dose used was 20.5 Gy (range, 15–25 Gy). Out of eight 15. Liquori CL, Berg MJ, Siegel AM, Huang E, Zawistowski JS,
patients who underwent follow-up cerebral angiogram, four Stoffer T, Verlaan D, Balogun F, Hughes L, Leedom TP, Plummer
had complete obliteration and four showed significant oblit- NW, Cannella M, Maglione V, Squitieri F, Johnson EW, Rouleau
GA, Ptacek L, Marchuk DA. Mutations in a gene encoding a novel
eration (50, 80, 80, and 90 %) with a follow-up ranging from protein containing a phosphotyrosine-binding domain cause type 2
15 months to 8 years. One patient had transient symptomatic cerebral cavernous malformations. Am J Hum Genet. 2003;73(6):
radiation-induced edema after two Gamma Knife proce- 1459–64.
dures, and MRI showed evidence of edema in another patient 16. Verlaan DJ, Laurent SB, Rochefort DL, Liquori CL, Marchuk DA,
Siegel AM, Rouleau GA. CCM2 mutations account for 13% of
who remained asymptomatic. These promising results dem- cases in a large collection of kindreds with hereditary cavernous
onstrate that stereotactic radiosurgery can be of benefit in the malformations. Ann Neurol. 2004;55(5):757–8.
management of stable VGM patients who can tolerate the 17. Hasegawa T, McInerney J, Kondziolka D, Lee JYK, Flickinger JC,
time interval needed for obliteration. Lunsford LD. Long-term results after stereotactic radiosurgery for
patients with cavernous malformations. Neurosurgery. 2002;50(6):
1190–7; discussion 1197–8.
18. Witiw CD, Abou-Hamden A, Kulkarni AV, Silvaggio JA, Schneider
References C, Wallace MC. Cerebral cavernous malformations and pregnancy:
hemorrhage risk and influence on obstetrical management.
1. Gross BA, Lin N, Du R, Day AL. The natural history of intracranial Neurosurgery. 2012;71(3):626–30; discussion 631.
cavernous malformations. Neurosurg Focus. 2011;30(6):E24. 19. Kalani MYS, Zabramski JM. Risk for symptomatic hemorrhage of
2. Porter PJ, Willinsky RA, Harper W, Wallace MC. Cerebral cavern- cerebral cavernous malformations during pregnancy. J Neurosurg.
ous malformations: natural history and prognosis after clinical 2013;118(1):50–5.
20. Ide C, De Coene B, Baudrez V. MR features of cavernous angioma. 40. Lee C-C, Pan DH-C, Chung W-Y, Liu K-D, Yang H-C, Wu H-M,
JBR-BTR. 2000;83(6):320. Guo W-Y, Shih Y-H. Brainstem cavernous malformations: the role
21. Hallam DK, Russell EJ. Imaging of angiographically occult cere- of Gamma Knife surgery. J. Neurosurg. 2012;117 Suppl:164–169.
bral vascular malformations. Neuroimaging Clin N Am. 1998;8(2): 41. Niranjan A, Lunsford LD. Stereotactic radiosurgery guidelines for
323–47. the management of patients with intracranial cavernous malforma-
22. Aiba T, Tanaka R, Koike T, Kameyama S, Takeda N, Komata T. tions. Prog Neurol Surg. 2013;27:166–75.
Natural history of intracranial cavernous malformations. J 42. Kida Y, Kobayashi T, Tanaka T. Treatment of symptomatic AOVMs
Neurosurg. 1995;83(1):56–9. with radiosurgery. Acta Neurochir Suppl. 1995;63:68–72.
23. Kondziolka D, Lunsford LD, Kestle JR. The natural history of cere- 43. Kondziolka D, Lunsford LD, Flickinger JC, Kestle JR. Reduction
bral cavernous malformations. J Neurosurg. 1995;83(5):820–4. of hemorrhage risk after stereotactic radiosurgery for cavernous
24. Kupersmith MJ, Kalish H, Epstein F, Yu G, Berenstein A, Woo H, malformations. J Neurosurg. 1995;83(5):825–31.
Jafar J, Mandel G, De Lara F. Natural history of brainstem cavern- 44. Pollock BE, Garces YI, Stafford SL, Foote RL, Schomberg PJ, Link
ous malformations. Neurosurgery. 2001;48(1):47–53; discussion MJ. Stereotactic radiosurgery for cavernous malformations. J
53–4. Neurosurg. 2000;93(6):987–91.
25. Al-Shahi Salman R, Hall JM, Horne MA, Moultrie F, Josephson 45. Kim DG, Choe WJ, Paek SH, Chung HT, Kim IH, Han
CB, Bhattacharya JJ, Counsell CE, Murray GD, Papanastassiou V, DH. Radiosurgery of intracranial cavernous malformations. Acta
Ritchie V, Roberts RC, Sellar RJ, Warlow CP. Scottish Audit of Neurochir (Wien). 2002;144(9):869–78; discussion 878.
Intracranial Vascular Malformations (SAIVMs) collaborators. 46. Baumann CR, Acciarri N, Bertalanffy H, Devinsky O, Elger CE,
Untreated clinical course of cerebral cavernous malformations: a Russo Lo G, Cossu M, Sure U, Singh A, Stefan H, Hammen T,
prospective, population-based cohort study. Lancet Neurol. Georgiadis D, Baumgartner RW, Andermann F, Siegel AM. Seizure
2012;11(3):217–24. outcome after resection of supratentorial cavernous malformations:
26. Gross BA, Batjer HH, Awad IA, Bendok BR, Du R. Brainstem cav- a study of 168 patients. Epilepsia. 2007;48(3):559–63.
ernous malformations: 1390 surgical cases from the literature. 47. Stavrou I, Baumgartner C, Frischer JM, Trattnig S, Knosp E. Long-
World Neurosurg. 2013;80(1–2):89–93. term seizure control after resection of supratentorial cavernomas: a
27. Steinberg GK, Chang SD, Gewirtz RJ, Lopez JR. Microsurgical retrospective single-center study in 53 patients. Neurosurgery.
resection of brainstem, thalamic, and basal ganglia angiographi- 2008;63(5):888–96; discussion 897.
cally occult vascular malformations. Neurosurgery. 2000;46(2):260– 48. Gewirtz RJ, Steinberg GK, Crowley R, Levy RP. Pathological
70; discussion 270–1. changes in surgically resected angiographically occult vascular
28. Mathiesen T, Edner G, Kihlström L. Deep and brainstem caverno- malformations after radiation. Neurosurgery. 1998;42(4):738–42;
mas: a consecutive 8-year series. J Neurosurg. 2003;99(1):31–7. discussion 742–3.
29. Wang C-C, Liu A, Zhang J-T, Sun B, Zhao Y-L. Surgical manage- 49. Nyáry I, Major O, Hanzély Z, Szeifert GT. Histopathological find-
ment of brain-stem cavernous malformations: report of 137 cases. ings in a surgically resected thalamic cavernous hemangioma 1 year
Surg Neurol. 2003;59(6):444–54; discussion 454. after 40-Gy irradiation. J. Neurosurg. 2005;102 Suppl:56–58.
30. Ferroli P, Sinisi M, Franzini A, Giombini S, Solero CL, Broggi 50. Monaco EA, Khan AA, Niranjan A, Kano H, Grandhi R, Kondziolka
G. Brainstem cavernomas: long-term results of microsurgical D, Flickinger JC, Lunsford LD. Stereotactic radiosurgery for the
resection in 52 patients. Neurosurgery. 2005;56(6):1203–12; dis- treatment of symptomatic brainstem cavernous malformations.
cussion 1212–4. Neurosurg Focus. 2010;29(3):E11.
31. Hauck EF, Barnett SL, White JA, Samson D. Symptomatic brainstem 51. Kim C-Y, Paek SH, Jeong SS, Chung H-T, Han JH, Park C-K, Jung
cavernomas. Neurosurgery. 2009;64(1):61–70; discussion 70–1. H-W, Kim DG. Gamma knife radiosurgery for central neurocytoma:
32. Abla AA, Lekovic GP, Turner JD, de Oliveira JG, Porter R, Spetzler primary and secondary treatment. Cancer. 2007;110(10):2276–84.
RF. Advances in the treatment and outcome of brainstem cavernous 52. Park S-H, Hwang S-K. Gamma Knife radiosurgery for symptom-
malformation surgery: a single-center case series of 300 surgically atic brainstem intra-axial cavernous malformations. World
treated patients. Neurosurgery. 2011;68(2):403–14; discussion Neurosurg. 2013;80(6):e261–6.
414–5. 53. Tung H, Giannotta SL, Chandrasoma PT, Zee CS. Recurrent intra-
33. Pandey P, Westbroek EM, Gooderham PA, Steinberg GK. Cavernous parenchymal hemorrhages from angiographically occult vascular
malformation of brainstem, thalamus and basal ganglia- a series of malformations. J Neurosurg. 1990;73(2):174–80.
176 patients. Neurosurgery. 2013;72(4):573–89. 54. Houser OW, Baker HL, Rhoton AL, Okazaki H. Intracranial dural
34. Karlsson B, Kihlström L, Lindquist C, Ericson K, Steiner L. arteriovenous malformations. Radiology. 1972;105(1):55–64.
Radiosurgery for cavernous malformations. J Neurosurg. 1998; 55. Awad IA, Little JR, Akarawi WP, Ahl J. Intracranial dural arteriove-
88(2):293–7. nous malformations: factors predisposing to an aggressive neuro-
35. Liu K-D, Chung W-Y, Wu H-M, Shiau C-Y, Wang L-W, Guo W-Y, logical course. J Neurosurg. 1990;72(6):839–50.
Pan DH-C. Gamma knife surgery for cavernous hemangiomas: an 56. Cognard C, Gobin YP, Pierot L, Bailly AL, Houdart E, Casasco A,
analysis of 125 patients. J. Neurosurg. 2005;102 Suppl:81–86. Chiras J, Merland JJ. Cerebral dural arteriovenous fistulas: clinical
36. Kim MS, Pyo SY, Jeong YG, Lee SI, Jung YT, Sim JH. Gamma and angiographic correlation with a revised classification of venous
knife surgery for intracranial cavernous hemangioma. J. Neurosurg. drainage. Radiology. 1995;194(3):671–80.
2005;102 Suppl:102–106. 57. Viñuela F. Van V Halbach. Raven Pr: Interventional neuroradiol-
37. Liscak R, Vladyka V, Simonová G, Vymazal J, Novotny J. Gamma ogy; 1992.
knife surgery of brain cavernous hemangiomas. J. Neurosurg. 58. Yang H, Kano H, Kondziolka D, Niranjan A, Flickinger JC,
2005;102 Suppl:207–213. Horowitz MB, Lunsford LD. Stereotactic radiosurgery with or
38. Nagy G, Razak A, Rowe JG, Hodgson TJ, Coley SC, Radatz MWR, without embolization for intracranial dural arteriovenous fistulas.
Patel UJ, Kemeny AA. Stereotactic radiosurgery for deep-seated Prog Neurol Surg. 2013;27:195–204.
cavernous malformations: a move toward more active, early inter- 59. Hanakita S, Koga T, Shin M, Shojima M, Igaki H, Saito N. Role of
vention. Clinical article. J Neurosurg. 2010;113(4):691–9. Gamma Knife surgery in the treatment of intracranial dural arterio-
39. Lunsford LD, Khan AA, Niranjan A, Kano H, Flickinger JC, venous fistulas. J. Neurosurg. 2012;117 Suppl:158–163.
Kondziolka D. Stereotactic radiosurgery for symptomatic solitary 60. Koebbe CJ, Singhal D, Sheehan J, Flickinger JC, Horowitz M,
cerebral cavernous malformations considered high risk for resec- Kondziolka D, Lunsford LD. Radiosurgery for dural arteriovenous
tion. J Neurosurg. 2010;113(1):23–9. fistulas. Surg Neurol. 2005;64(5):392–8; discussion 398–9.
61. Friedman JA, Pollock BE, Nichols DA, Gorman DA, Foote RL, with stereotactic radiation or surgery. J Neurosurg. 1994;81(6):
Stafford SL. Results of combined stereotactic radiosurgery and 851–9.
transarterial embolization for dural arteriovenous fistulas of the 65. Drake CG. Cerebral arteriovenous malformations: considerations
transverse and sigmoid sinuses. J Neurosurg. 2001;94(6):886–91. for and experience with surgical treatment in 166 cases. Clin
62. Pan H-C, Sun M-H, Chen CC-C, Chen C-J, Lee C-H, Sheehan Neurosurg. 1979;26:145–208.
J. Neuroimaging and quality-of-life outcomes in patients with brain 66. Yaşargil MG. Microneurosurgery. New York: Thieme; 1996.
metastasis and peritumoral edema who undergo Gamma Knife sur- 67. Mericle RA, Mickle JP, Quisling RG. Vein of Galen malformation.
gery. J. Neurosurg. 2008;109 Suppl:90–98. Pediatr Neurosurg. 1999;31(5):279–80.
63. Pollock BE, Nichols DA, Garrity JA, Gorman DA, Stafford 68. Watban JA, Rodesch G, Alvarez H, Lasjaunias P. Transarterial
SL. Stereotactic radiosurgery and particulate embolization for embolization of vein of Galen aneurysmal malformation after
cavernous sinus dural arteriovenous fistulae. Neurosurgery. 1999; unsuccessful stereotactic radiosurgery. Report of three cases.
45(3):459–66; discussion 466–7. Childs Nerv Syst. 1995;11(7):406–8.
64. Lewis AI, Tomsick TA, Tew JM. Management of tentorial dural 69. Payne BR, Prasad D, Steiner M, Bunge H, Steiner L. Gamma surgery
arteriovenous malformations: transarterial embolization combined for vein of Galen malformations. J Neurosurg. 2000;93(2):229–36.
Cerebral Cavernous Malformations:
Robert L. Dodd and Gary K. Steinberg
developmental venous anomalies (DVAs), which are often

Introduction found in association with CCMs [8, 9]. The incidence of these
lesions is approximately 0.5–0.6 per 100,000 people per year
Cerebral cavernous malformations (CCMs), also known as [1, 2], in whom over 40 % are clinically symptomatic and
cavernous angiomas, cavernomas, or cryptic vascular mal- most of whom present with seizures. The age distribution of
formations, are rare low-flow vascular lesions [1, 2]; how- patients with clinically symptomatic cavernous malforma-
ever, because their hemorrhage can result in significant tions is bell shaped, with the highest incidence between the
morbidity and mortality [3, 4], a great deal of effort has been third and fifth decades [10]. New presentation of cavernous
devoted to their detection and treatment. Although previ- malformations in infants or adults over the age of 80 years,
ously thought to be congenital, it is now well recognized that though documented, is rare. One-quarter of cases, however,
most cavernous malformations are acquired [5], arising de are estimated to occur during childhood [11–15]. Whereas a
novo or occasionally following radiotherapy. The natural few published reports have found slight male or female pre-
history of cavernous malformations is only now being eluci- ponderance, most studies with large patient populations con-
dated with the widespread use of magnetic resonance imag- cur that there is no difference in the sex incidence of CCMs.
ing (MRI). Preliminary reports suggest the annual risk of Cavernous malformations are defined as an assembly of
clinically significant hemorrhage to be less than 1 % [1–4], thin-walled vascular sinusoids without intervening brain
although many factors may increase this risk. In general, parenchyma, lined by a single layer of endothelium, devoid of
results from surgical removal of accessible symptomatic cav- mature vessel elements such as smooth muscle and elastin. In
ernomas are excellent, with improved control of medically contrast to other vascular lesions, cavernous malformations
intractable seizures, restoration of neurological function, and rarely bleed into the subarachnoid space or into the ventricu-
decreased risk of future hemorrhage. Surgically inaccessible lar system; more commonly, these lesions bleed into the sinu-
lesions remain a difficult clinical challenge. soidal space of the malformation itself, resulting in acute
enlargement or intraparenchymal hemorrhage. Although
these vascular malformations are low flow and undergo con-
Epidemiology tinual microhemorrhaging, clinically significant intralesional
hemorrhage or gross hemorrhage beyond the lesion is rare,
The prevalence of CCMs remains undetermined; however, with an estimated risk between 0.25 and 6 % per year [3, 11,
most reports estimate their occurrence in approximately 0.3– 16–20]. Hypertension, cigarette smoking, oral contraceptive,
0.9 % of the general population [1, 2]. These estimates are alcohol consumption, and cocaine use are all known risk fac-
based on autopsy results and prospective cohort studies using tors for intracerebral hemorrhage in general, but no study to
MRI, which suggest cavernous malformations comprise date has shown increased risk in patients who harbor cavern-
8–15 % of all vascular lesions [6, 7], therefore accounting for ous malformations. The most widely cited risk factor for
the second most common type of vascular malformation after clinically significant hemorrhage, apart from family history,
is prior hemorrhage [3, 17, 19, 20]. Interestingly, while the
high frequency, rapidity, and gravity of hemorrhagic recur-
R.L. Dodd, M.D., Ph.D. • G.K. Steinberg, M.D., Ph.D. (*) rences after the first intracranial hemorrhage have been
Department of Neurosurgery, Stanford University School
of Medicine, 300 Pasteur Drive, Room 281, Stanford,
stressed by some, other series have shown little influence.
CA 94305, USA Several authors have found a female preponderance in bleed-
e-mail: gsteinberg@stanford.edu ing risk, suggesting that endocrine factors may influence
638 R.L. Dodd and G.K. Steinberg
hemorrhage tendencies, particularly since some of the bleeding malformation induction in 40–53 % of familial cases [36–39],
episodes occurred during pregnancy [3, 9, 19]. These obser- with mutations in genes at CCM2 and CCM3 accounting for
vations are supported by the detection of estrogen receptors in 15–20 % [38, 39] and 10–40 % [38, 39], respectively [10].
a few cavernous malformations obtained from females [21], Interestingly, the majority of Hispanic-American patients of
as well as by speculation that hormonal fluctuations during Mexican descent have the same mutation in CCM1, indicating
pregnancy increase endothelial cell proliferation in these a common ancestor [40]. Reports of cavernous malforma-
lesions and their propensity to hemorrhage [19]. tions in familial series indicate the majority of these indi-
viduals had multiple lesions (up to 93 %), were clinically
symptomatic, and presented at a younger age.
Anatomical Distribution Proteins expressed by the CCM genes have been identi-
fied histologically in neurons, astrocytes, and vascular endo-
Cavernous malformations can be found throughout the cen- thelium [37, 41–43]. Their primary and secondary structures
tral nervous system including every region of the brain, reveal ultrastructural regions that suggest binding, and
brainstem, spinal cord, cranial nerves, and ventricles. The research has established that the three CCM proteins interact
cerebral hemispheres are the most frequently affected areas with each other to form ternary complexes [44–47], which in
(in one study 21 % frontal, 16 % parietal, 15 % temporal, and turn interact with other proteins [48, 49]. KRIT1 seems to be
3.5 % occipital), with most being subcortical in location predominantly localized in the cytoplasm although the pro-
[22]. Recently, more cases of lesions in the basal ganglia, tein has both a nuclear localization signal and a nuclear
corpus callosum, thalamus, and third ventricle have been export signal, suggesting it can be shuttled to and from the
reported because of the interest generated by their therapeu- nucleus [50]. Malcavernin also localizes to the cytoplasm,
tic complexity. Between 10 and 23 % of cavernous malfor- binds KRIT1, and may inhibit its nuclear translocation
mations are located infratentorially, where the pons is the [32, 36, 51, 52]. PDCD10 binds to malcaverin and also has
preferred site [21–25]. In 40–60 % of cases, multiple lesions structural domains that interact with the (PI3K-) PIP3-
exist, and multiplicity should always raise the suspicion of PDPK1-Akt signaling pathway [34, 53].
the familial form of the condition [26]. Multiple intracranial The precise pathophysiologic mechanisms that explain
cavernous malformations are sometimes associated with how an alteration in the CCM complex results in manifesta-
similar vascular lesions in other organs, suggesting the pres- tions of the clinical disease remain unclear. However, animal
ence of a wider and more systemic condition. models and in vitro biochemical experiments, while imper-
fect, suggest that CCM gene products in concert with other
cellular machinery form ternary complexes that regulate key
Molecular and Genetic Basis of CCMs steps in blood vessel formation and maintenance [54–58]. It is
postulated that the inability to compose these protein com-
CCMs can occur in either a sporadic or familial form of the plexes leads to the dysfunction in cytoskeletal processes
disease, and while the occurrence of cavernous malforma- responsible for cell–cell adhesion, cell adhesion to the extra-
tions in more than one family member was initially believed cellular matrix, and tight junction assembly important for
uncommon, more recent evidence suggests as many as half cell polarity [59]. This ultimately results in dysregulation of
of CCM cases maybe familial in nature [27–29]. Hereditary vascular development and endothelial permeability [60, 61].
forms of CCM exhibit Mendelian autosomal dominant
inheritance with variable penetrance caused by loss-of-
function mutations which to date have been associated with Natural History
three genes: KRIT1 (also known as CCM1) on chromosome
7q11.2-21 [30, 31]; CCM2 (encoding the malcavernin pro- Decisions on treatment recommendations are based to a great
tein), on chromosome 7p13 [30, 32]; and PDCD10 (also extent on the estimated risk of further morbidity; therefore,
known as CCM3) on chromosome 3q26.1 [10, 30, 33, 34]. knowledge regarding the natural history of cavernous malfor-
In addition, discrepancies in the observed relative incidence mations is an important consideration in clinical practice.
of mutations in these three genes between the original link- Despite recognition of this point, our present knowledge is
age analysis data and the mutation screening data suggest a incomplete because of several factors: the low frequency of
potential fourth CCM locus (CCM4) [35]. occurrence limits study population size, the relatively recent
The importance of CCM genes is highlighted by the introduction of MRI technology required for accurate diagno-
observation that at least one of the CCM genes is mutated in sis limits the length of clinical follow-up, the paucity of pro-
most familial cavernous malformations. Over 100 different spective studies limits analysis of the natural history of
single-gene (KRIT1) mutations have been found in CCM1- cavernous malformations, and confusion in the literature exists
linked families, which are believed to cause cavernous on what constitutes a clinically significant hemorrhage.
51 Cerebral Cavernous Malformations: Viewpoint—Surgery 639
Nevertheless, several recent studies have provided some spaces, resulting in growth of the vascular nidus; (2) recurrent
estimates for risk of bleeding and growth [1, 4, 62, 63]. The “external” microhemorrhages or red blood cell diapedesis
risk of clinically relevant hemorrhage after detection of a with thrombosis, followed by organization, fibrosis, and cal-
cavernous malformation depends on the presenting symp- cification; (3) acute gross hemorrhage, resulting in acute size
toms as well as the location of the lesion. Preliminary reports increase; and (4) cyst formation “internal” hemorrhage that
indicate that the frequency of clinical hemorrhage among ruptures septae between adjacent sinusoids. Further growth
patients with supratentorial nonfamilial cavernous malfor- and perifocal edema may be related to secondary microhem-
mations who present either with incidental diagnosis or with orrhages and fluid exudation from the neocapillary mem-
seizures is approximately 0.25–2 % per year [2, 3, 17, 19]. brane in the cyst capsule.
The percentage appears to be significantly higher (6.5 % per As our understanding of bleeding and growth rates is far
patient per year; 1.3 % per lesion per year) in patients with from exact, and indeed many patients who experience bleed-
the familial form of the disease [20]. Likewise, worsening of ing from cavernous malformations may show good clinical
neurologic signs and symptoms is more frequent in the recovery, some authors have encouraged conservative treat-
familial group, occurring in more than 15 % per year [20]. ment, particularly for brainstem cavernous malformations.
Among patients with nonfamilial cavernomas who present Kupersmith et al. report that the majority of 23 nonsurgically
with symptomatic hemorrhage, the annual recurrent hemor- treated patients with symptomatic brainstem cavernous mal-
rhage rate may be higher (about 4–5 %) in the next year than formations followed over a mean of almost 5 years either
in patients with no prior hemorrhage (0.6 %) [2]. Barker improved (78 %) or remained clinically stable (17.4 %) [66].
et al. provide evidence for temporal clustering of hemor- Other larger series of similar patients however detail a much
rhages, suggesting rebleed rates as high as 2.1 % per month more ominous outlook, with a 3.5 % perioperative mortality
for the first 2.5 years after the initial hemorrhage before rate, an 8 % overall mortality rate during follow-up after
decreasing to 1 % per month [62]. The risk of hemorrhage surgery, and a 20 % mortality rate for nonsurgically treated
may also depend somewhat on location, with some reports of patients after 35.7 months of follow-up [21]. Such variability
patients with deep lesions having an initial annual clinical demonstrates the sometimes unpredictable clinical courses
hemorrhage risk of 4.1 % compared with 0.4 % among those of these lesions and highlights our deficient knowledge of
with superficial lesions [2]. However, a recent meta-analysis how they behave.
of 11 natural history studies suggests that this difference Furthermore, survival from and successful treatment for a
largely reflects differences in the definition of hemorrhage cavernous malformation that has hemorrhaged may not
(radiographic versus clinical) [64], with deep/eloquent loca- always provide a cure for the disease. Clinical reports have
tions behaving more clinically aggressive without radio- documented the de novo formation of cavernomas for both
graphic changes on MRI. Younger patients may also have a familial and sporadic forms of the disease [5, 20, 65]. In
higher risk than older patients of a second hemorrhage. patients with genetic predispositions, the new cavernoma
Cavernous malformations are often surprisingly dynamic, formation has been estimated to have an incidence of 0.2–0.4
with changes in size, number, and MRI signal characteris- lesions per patient-year [20, 67]. The incidence of de novo
tics. The size of these lesions in the literature has been lesions in patients without known genetic predisposition is
reported to vary between less than 1 mm up to more than unclear, but multiple cavernomas have been described to
10 cm in diameter [23]. The mean cavernoma size in several occur in one report, in 14 of 55 sporadic cases (25.4 %) [65].
large series was approximately 15–19 mm, with the majority
of symptomatic lesions being greater than 1 cm in diameter
[16, 19, 23]. However, clinical symptoms are more related to Clinical Presentation
location than to size, as very small tectal cavernomas can
cause ophthalmoplegia, while giant frontal lobe lesions may The clinical symptomatology of cavernous malformations is
be asymptomatic. Clatterbuck et al. prospectively followed highly variable, ranging from an asymptomatic incidental
114 lesions in 68 patients over 352.9 patient-years and dem- finding to discovery in autopsy after fatal hemorrhage.
onstrated that most lesions either increased or decreased in Typically patients come to medical attention because of sei-
size with less than 22 % being stable in volume [65]. Three zures, focal neurological deficits, or headaches, though there
de novo lesions were observed during this period and may have been an increasing number of patients with malforma-
represent the growth of very small lesions without symptom- tions found incidentally following imaging for other pur-
atic hemorrhage. In addition to changes in volume, they also poses such as trauma. All of these symptoms are believed to
demonstrated a stereotypical progression of changes in be secondary to either lesion hemorrhage or mass effect on
appearance. The mechanisms purported to cause growth and neural structures. In one series of 565 cases of symptomatic
MRI signal changes include: (1) progressive ectasia of vas- cavernous malformations, 217 (38.4 %) presented with sei-
cular channels and budding of capillaries from cavernous zures, 191 (33.8 %) with tumor-like symptoms, 66 (11.6 %)
presented with headache, and 101 (17.8 %) with overt hem- with acute symptoms. The prevalence of this type of hemor-
orrhage [68]. Numerous other reports have confirmed epilep- rhage in large series ranges between 6 and 30 % of cases
tic seizures constitute the most frequent clinically presenting [9, 11, 16, 22, 65]. The clinical presentation of hemorrhage
symptom. Between 35 and 70 % of symptomatic cavernomas is acute, but usually not as dramatic as those that occur from
are associated with recurrent seizures, which are drug resis- intracranial aneurysms or AVMs. Sudden onset headache
tant in 40 % of cases [69]. All seizure types including simple followed by a focal neurological deficit is typical, depending
seizures, complex partial seizures, and generalized seizures on the location and size of the hematoma.
[10] have been known to present in patients with supratento-
rial cavernous malformations. The pathogenesis of seizures
is debated, but most believe they are related to the presence Diagnostic Imaging
of iron products after red cell breakdown secondary to mul-
tiple microhemorrhages. Most patients with seizures present The most sensitive and therefore the most important diagnos-
with lesions in their frontal or temporal lobes. The estimated tic tool is MRI (Figs. 51.1, 51.2, and 51.3). The high sensitiv-
risk of developing seizure is 1–2 % per person-year exposure ity of gradient-echo sequences is particularly helpful in
and mean age at time of first seizure is 42 ± 3.78 years [9, 11]. detecting both acute and chronic hemorrhages that may not
The second most frequent presentation is that of focal be seen with conventional spin-echo techniques. Indeed the
neurological deficits, which correspond to lesion location, widespread use of MRI has led to the detection of asymptom-
and are usually produced by intralesional or perilesional atic cavernous malformations, therefore increasing previous
hemorrhage. They account for 35–50 % of reported cases estimates regarding the prevalence of these lesions. The MRI
[14]. Vertigo, diplopia, ataxia, hemiparesis, associated sen- appearance of cavernous malformations has been detailed by
sory disturbances, and changes in level of consciousness are many authors including Zabramski et al., who distinguish
typical of infratentorial cavernous malformations, which, four types of MR characteristics [20]. Type I lesions are
occasionally, may even produce signs of increased intracra- hyperintense on T1- and T2-weighted images and indicate
nial pressure due to obstructive hydrocephalus. Aphasia and subacute hemorrhage. Type II lesions show a mixed-signal
apraxias produced by supratentorial lesions, visual loss by intensity on T1- and T2-weighted imaging with a surround-
lesions within the optic nerve or chiasm, and symptoms of ing hemosiderin ring. Type III lesions are hypo- to isointense
trigeminal neuralgia from extra-axial cavernous malforma- on T1- and T2-weighted images and most probably indicate
tions of the middle fossa have also been reported. chronic hemorrhage. Type IV lesions are poorly visualized
Neurological deficits may be transient, progressive, or fixed. except on gradient-echo sequences. Histopathological corre-
In rare instances cavernous malformations may simulate lations to these MRI patterns have been confirmed. In addi-
multiple sclerosis due to fluctuating progressive neurological tion to establishing the diagnosis, MRI also defines the exact
deficits, such as repeated exacerbation of complaints and size, location, and extent of cavernous malformations as well
alternating periods of remission. Cavernous malformations as provides information about multiplicity, hydrocephalus,
have also been found in the spinal cord where they present and the condition of the adjacent brain parenchyma.
with an acute or subacute myelopathy.
Headache accompanies cavernous malformations in many
patients and may have prompted diagnostic evaluation Treatment
uncovering the lesion. While this is a controversial present-
ing complaint due to its nonspecific and nonlocalizing nature, Given the excellent results of microsurgical resection of cav-
it represents a major symptom in 25–30 % of reported cases ernous malformations, it is generally easy to endorse the phi-
[14]. The prevalence of asymptomatic lesions in the general losophy of operative management for most of these lesions.
population is unknown, but recent evidence suggests it is However, increasing evidence suggesting a relatively benign
much higher than previously suspected. Studies among con- natural history of some lesions has indicated the need for
secutive patients undergoing MRI have detected asymptom- reevaluation of management decisions. In particular, when
atic cavernous malformations in 14–19 % of patients [14]. dealing with a patient harboring several lesions, a family
Other reviews of the literature have found as many as 21 % affected by a hereditary form of this condition, or deep-
of patients to be without symptoms at first discovery [3, 22]. seated lesions in the brainstem, it becomes apparent that an
The incidence of cavernous malformations that present aggressive approach is not always advisable. The only clear
with clinical hemorrhage is uncertain. This is because of indications for surgery are for medically intractable epilepsy,
variability in hemorrhage patterns of cavernous malforma- documented recurrent hemorrhage, and progressive neuro-
tions as well as variability in the literature of what defines a logical deficit. Patients with an established diagnosis of
clinically significant event. Some have described clinically CCM who present without clinical hemorrhage, seizures, or
significant hemorrhages as those that are sizable and present other specific symptoms are usually candidates for clinical
Fig. 51.1 Selected (a) axial FLAIR, (b) coronal T1 postcontrast, and hypersomnolence, loss of appetite, and memory difficulties. A cavern-
(c) sagittal T1 MRI scans of a 31-year-old female medical student who ous malformation is centered in the midline involving the massa inter-
presented with four clinical hemorrhages resulting in headaches, gait media, hypothalamus, midbrain, third ventricle, and extending to the
instability, and symptoms of hypothalamic dysfunction including suprasellar cistern
observation and repeat imaging. Conservative management quent and noneloquent areas. More than 75 % of all cavern-
is also appropriate for many patients with purely incidental ous malformations are located supratentorially [22], most of
lesions or malformations located deep within functional which are located in cortical and subcortical regions that are
areas that do not present to a pial or ventricular surface. readily accessible with modern surgical techniques. Even
prior to MRI, surgical series have documented the successful
removal of cavernous malformations with low risk. Vaquero
Microsurgical Treatment of Supratentorial et al. report successful treatment of 19 supratentorial cavern-
Lesions ous malformations located in the cerebral hemispheres from
1977 to 1987. Of the 19 patients, 17 had excellent results (no
Surgical resection of most cavernous malformations is seizures without medication) and two were classified as good
straightforward because these lesions have no major arterial outcomes (occasional seizures with medication) [70]. There
supply and do not bleed much at surgery. The most difficult were no poor outcomes or deaths in this cohort. Similarly,
task in resection is often intraoperative localization and excellent results have been documented in the post-MRI era.
choosing an appropriate surgical corridor. This has been McCormick reviewed his experience with removal of supra-
facilitated with the use of stereotaxic MRI computer- tentorial cavernous malformations. In 63 of 65 patients, com-
guidance systems for neuronavigation, which allows plan- plete removal was obtained, two patients experienced
ning of surgical trajectories to avoid functional cortex and worsening of an existing deficit, a third patient developed a
major blood vessels, as well as target localization to within new deficit, and there was only one postoperative death [71].
1–2 mm. The risk of open microsurgery is generally related When performed for medically intractable epilepsy, several
to lesion size, location, and proximity to critical structures. In published results confirm a greater than 90 % likelihood of a
addition, advanced age, medical comorbidities, and poor seizure-free outcome with surgical extirpation of cavernous
neurological condition are all negative prognostic factors for malformations and the surrounding epileptogenic tissue [69].
a good surgical result. Several surgical series have reported The surgical approach is defined by the lesion location,
excellent results in removal of superficial lesions of both elo- and in general the craniotomy is centered over the lesion.
Fig. 51.2 Selected immediately postoperative (a) axial T1, (b) coronal T1 malformation. (e) A 3.5-cm cavernous malformation was resected piece-
postcontrast, and (c) sagittal T1 MRI scans demonstrate gross total resec- meal from the third ventricle, diencephalon, and midbrain. There was
tion of the cavernous malformation seen in Fig. 51.1. (d) Microsurgical acute, subacute, and chronic hemorrhage within the malformation, which
removal was performed via an interhemispheric transcallosal approach to was composed of thin-walled, fragile vascular channels. No attempt was
the third ventricle. The asterisk (*) is on the corpus callosum, through made to remove the surrounding hemosiderin-stained brain. Postoperatively,
which a small incision was made that exposed an enlarged foramen of the patient quickly recovered to her neurologic baseline and has continued
Monro. The laser pointer is focused on the superior aspect of the cavernous to improve, though some memory difficulties still remain
The use of neuronavigation allows precise targeting that cated for seizures, the hemosiderin- or hematoidin-laden
minimizes the size of the craniotomy and risk to exposed gliotic tissue should be resected due to its role in seizure gen-
brain (Fig. 51.3). Lesions that involve the anterior corpus eration. Intraoperative electrocorticography may be helpful
callosum near the genu and cingulate gyrus can be approached for confirming the extent of abnormal electrical activity, par-
through a small frontal craniotomy. Lesions that are more ticularly for lesions near the motor or sensory cortices. For
basal in the subfrontal or septal regions often require a lower dominant hemisphere lesions, awake intraoperative speech
frontal craniotomy through a bicoronal incision or an inter- mapping may be required to preserve language function.
hemispheric approach. Posterior parasagittal approaches are
generally used for lesions in the body of the corpus callosum,
cingulate gyrus, and midline posterior frontal, parietal, or Microsurgical Treatment of Infratentorial
occipital areas (Fig. 51.4), as well as for some thalamic CMs. Lesions
Medial anterior temporal lesions involving the amygdala, the
hippocampus, or the uncus are approached through the Lesions located deep within the brain are difficult to remove
Sylvian fissure via a standard pterional craniotomy. and present special challenges (Figs. 51.1, 51.2, 51.3, and
Superficial cortical lesions are often easy to identify because 51.4), though recent series describe improved outcomes with
of overlying cortical discoloration, whereas small subcorti- image-guided techniques. Approximately 10–30 % of all
cal lesions often require stereotaxis. Both trans-sulcal and intracranial cavernous malformations are located in the pos-
trans-gyral approaches can be employed depending on lesion terior fossa [21–25], and some reports suggest these lesions
geometry and the location of cortical vasculature. Once iden- may be more frequently symptomatic [21]. Surgical removal
tified, a well-defined gliotic plane typically allows for easy is indicated for symptomatic lesions located in the cerebellum
separation of the lesion from surrounding tissue so that com- or superficially in the brainstem when eloquent parenchyma
plete lesion resection can be achieved. When surgery is indi- can be spared. Several surgical approaches have been
Fig. 51.3 (a) Sagittal T1, (b) coronal T1 postcontrast, and (c) axial T2 Microsurgical resection of the cavernous malformation was therefore
and (d) T1 MRI scans illustrate a 1-cm, right midpontine cavernous performed. (e) A right subtemporal approach was used as illustrated,
malformation with subacute hemorrhage surrounded by a rim of hemo- aided by the intraoperative surgical navigation. Brainstem functional
siderin and surrounding brainstem edema (arrow in c). Although this mapping was employed prior to making the few-millimeter incision in
lesion did not present to a pial or ventricular surface, three clinically the lateral surface of the pons (right upper panel, e). Postoperatively,
significant hemorrhages were observed over a 7-month period, result- the patient’s hemiparesis transiently worsened, but he recovered within
ing in diplopia, facial weakness, sensory loss, and mild hemiparesis. 1 week and currently has no focal deficits
Fig. 51.4 Selected 3-year postoperative (a) sagittal, (b) coronal, and (c) axial T1 MRI scans demonstrate gross total resection of the cavernous
malformation seen in Fig. 51.3
detailed in the literature including suboccipital, infratento- the selection of an approach that optimizes exposure to the
rial supracerebellar, occipital transtentorial, far lateral, lesion with minimum disruption to surrounding structures.
transtemporal, subtemporal transtentorial, and combined In general, lesions of the cerebellar vermis, medial cerebellar
petrosal. Furthermore, much emphasis has been placed on hemispheres, floor of the fourth ventricle, and dorsal medulla
are easily approached through a standard suboccipital radiosurgery may decrease the risk of clinical hemorrhage,
approach. Lesions of the tectum and pineal region are often compared with the statistically derived natural history [13, 73,
best approached via infratentorial supracerebellar access. 74]. Our combined Stanford University–University of
The occipital transtentorial approach provides exposure of California, Berkeley, program previously treated 57 cavern-
the superior cerebellar peduncles and vermis, the anterior ous malformations in the brainstem, thalamus, and basal
medullary velum, posterior third ventricle, the splenium, and ganglia using Bragg peak helium ion radiosurgery (47
the quadrigeminal plate. The far lateral approach provides lesions) or LINAC radiosurgery (10 lesions), with mixed
excellent exposure of the anterior and lateral medulla and results [74]. Eighteen patients (32 %) experienced symp-
cervicomedullary junction. Transtemporal approaches allow tomatic bleeding (20 hemorrhaging episodes) after radio-
access to lesions in and around the internal auditory meatus surgery. Sixteen hemorrhaging episodes occurred within 36
such as pontomedullary junction cavernous malformations months after radiosurgery (9.4 % annual bleeding rate; 16
and can be employed in patients without serviceable hearing. hemorrhaging episodes/171 patient-years), and 4 hemor-
Cavernous malformations involving the anterior midbrain– rhaging episodes occurred more than 36 months after treat-
interpeduncular fossa region can be approached through a ment (1.6 % annual bleeding rate; 4 hemorrhaging
pterional trans-sylvian, subtemporal, or orbitozygomatic episodes/257 patient-years) (P < 0.001). The radiosurgery
craniotomy, and occasionally lesions located in the upper group at Harvard University reported a decrease in the
two-thirds of the brainstem require a combined petrosal annual hemorrhage rate from 17.3 % before radiosurgery to
approach. CSF drainage for brain relaxation, electrophysio- 4.5 % after a latency period of 2 years after radiosurgery
logical monitoring, and functional mapping are all important [73]. Similar symptomatic hemorrhage rates after Gamma
for minimizing complications and reducing the risk of per- Knife radiosurgery for cavernous malformations were
manent neurological deficit. reduced from 8.8 % during the first 2 years after radiosur-
Unlike supratentorial cavernous malformations, deep gery to 1.1 % thereafter [13]. A more recent study of 103
lesions can be extremely adherent to the normal parenchyma patients with solitary cavernous malformations who were
and are sometimes removed in a piecemeal fashion following treated with Gamma Knife radiosurgery at the University of
circumferential separation and devascularization from the Pittsburgh with follow-up from 2 to 20 years demonstrated
surrounding gliosis. All perforating arterial vessels must be a 10.8 % annual hemorrhage rate for the first 2 years after
dissected meticulously and preserved. Furthermore, the treatment and then a 1.06 % annual bleed rate after 2 years
hemosiderin staining surrounding the malformation should [82]. However, statistical modeling analysis showed that the
be spared from surgical removal. An association between hemorrhage rate from untreated cavernous malformations
cavernous malformations and venous malformation has been appeared to demonstrate temporal clustering of hemor-
described and may be as high as 16 % [72], and while the rhages such that the rehemorrhage rate from untreated cav-
goal of surgery is complete excision of the cavernoma, inter- ernous malformations was found to be high initially (2 %
ruption of a venous malformation is to be avoided because of monthly rehemorrhage rate during the first 2.5 years follow-
the risk of venous infarction. The risk of open microsurgery ing hemorrhage; 14 % cumulative incidence of a second
for superficial cerebellar lesions is low, similar to that for hemorrhage during the first year) with decreases 2–3 years
most supratentorial cavernous malformations. The risks after a previous hemorrhage (to less than 1 % per month)
associated with surgical exploration of deep-brain and brain- [62]. These results therefore question whether the data from
stem cavernous malformations is considerably higher, as dis- the radiosurgical series represent protective effects of treat-
cussed below. ment or simply the natural bleeding history of these lesions.
Additionally, in ten patients who underwent microsurgical
resection of their cavernous malformations 1–10 years fol-
Radiosurgical Treatment of Cavernous lowing radiosurgical treatment, none of the cavernous mal-
Malformations formations were found to be thromobosed on pathological
examination [83]. Furthermore, in our series of patients
A number of groups have used stereotactic-focused radio- with cavernous malformations who underwent radiosur-
surgery with heavy-charged particles (protons or helium gery, 7 % experienced symptomatic radiation edema and
ions) or photons (Gamma Knife or LINAC radiosurgery) in 2 % experienced radiation necrosis [74] (Fig. 51.5). Other
attempts to improve the natural history of deep, difficult-to- groups have documented a 13.5 % incidence of new neuro-
resect cavernous malformations [13, 73–81]. Although there logic deficits [82] and a 16 % incidence of permanent neu-
has been no prospective randomized study, recent reports rological deficits with a 3 % mortality rate secondary to
from several groups suggest that, for certain patients, radiation-induced complications [73].
Fig. 51.5 (a) Axial proton-density MRI of a 35-year-old woman who 2 days after surgery without new deficits. (c) Hematoxylin and eosin-
initially presented with generalized seizures shows a 1-cm vascular stained specimen from a surgically removed brainstem cavernous mal-
malformation located in the left thalamus adjacent to the posterior limb formation 5 years after stereotactic heavy-charged-particle radiosurgery.
of the internal capsule. She was first treated with stereotactic heavy- Patent vascular channels (arrow) are observed. (d) Surgical specimen
charged-particle Bragg peak radiosurgery. Nine months after radiosur- from tissue adjacent to a cavernous malformation removed from another
gery, she experienced worsening right hemiparesis; (b) axial T2 patient with delayed radiation-induced injury 3 years after stereotactic
imaging at that time demonstrated extensive radiation-induced changes radiosurgery of a right parietal cavernous malformation. Histopathologic
extending into the basal ganglia and thalamus. She subsequently under- changes consisted of thick-walled hyalinized vessels with fibrinoid
went stereotactically guided microsurgical resection via a small left necrosis (arrow), postradiation gliosis, and reactive astrocytosis and
posterior parietal, interhemispheric transcallosal approach through the necrosis (arrowhead), characteristic of radiation effects
pulvinar. The patient recovered extremely well and was discharged
persistent deficit. Various series have demonstrated recurrent

Brainstem Cavernous Malformations hemorrhage rates of 5–60 % [21, 66, 84–87]. Furthermore,
when multiple bleeding episodes do occur, the time interval
Brainstem cavernomas, though very rare, deserve special between hemorrhages is increased and the mortality rate for
mention because their clinical presentation, natural history, each hemorrhage may be as high as 20 %.
and treatment decisions often differ from those of cavernous Surgical treatment of brainstem cavernous malformations
malformations in other locations. Such lesions account for is considerably more risky than operative intervention in
9–35 % of all cavernomas and for 18–22 % of all intracranial other less eloquent areas of the nervous system. Difficult
cavernomas [21, 23, 25]. Fifty to 70 % of these lesions occur access, narrow surgical corridors, and sensitivity to injury of
in the pons, 20–35 % in the midbrain, and 15–25 % in the critical structures near these lesions present unique obsta-
medulla. Symptoms vary from cranial nerve deficits, sensory cles. Despite the high risk of permanent morbidity and tech-
dysfunction, paresis/plegia, ataxia, dysmetria, speech diffi- nical challenges of microsurgical excision, several centers
culty to headaches and decreased levels of consciousness. have demonstrated expertise in removing these lesions with
The clinical presentation can often be confusing and many good outcomes. Pandey et al. reported a series of 176 patients
patients who have experienced episodic clinical deterioration with 179 deep-brain cavernous malformations located in the
secondary to hemorrhaging have been diagnosed with hav- brainstem (n = 136), thalamus (n = 16), and basal ganglia
ing multiple sclerosis. (n = 26), which were microsurgically resected [86].
There is evidence that the natural history of brainstem Neuronavigation, mild hypothermia (32–33 °C), electro-
cavernous malformations differs from that of other locations, physiological monitoring, cranial nerve mapping, and use of
suggesting that this subgroup has a higher propensity for the CO2 laser assisted in the safe removal of these lesions.
clinical bleeding. Porter et al. reported nearly all of their 100 Over a 20-year period with a mean follow-up of 3.5 years,
brainstem cavernomas presented with evidence of clinically the overall results were excellent (mRS = 0–1) in 54 %,
significant hemorrhage, 56 % with multiple bleeds, and 22 % good (mRS = 2) in 29 %, and poor (mRS = 3–6) in 18 %; 5 %
with more than two clinical bleeds [21]. While some patients of patients died. Ninety-six percent of these lesions were
with hemorrhage from brainstem cavernous malformations completely resected after a single surgery (86 %) or 2 surger-
make good recoveries without intervention, it is also clear ies (9.6 %), 88 % of the patients were either neurologically
that multiple bleeding episodes increase the likelihood of a improved or unchanged after 6 months, and only 12 % of the
patients were worse than their preoperative status at the last Acknowledgement This work was supported in part by support from
follow-up. Of 135 patients undergoing microsurgical resec- Bernard and Ronni Lacroute, the William Randolph Hearst Foundation,
and Russell and Beth Siegelman (to GKS).
tion of 136 brainstem cavernous malformations in this series,
although slightly less than 1/3 of the patients were immedi-
ately worse after surgery, at long-term follow-up (>6
months), 87 % of the patients were neurologically unchanged References
or improved, while 13 % had worsened. Porter et al. reported
their experience with 100 patients with brainstem cavernous 1. Al-Shahi R, Bhattacharya JJ, Currie DG, Papanastassiou V, Ritchie
V, Roberts RC, et al. Prospective, population-based detection of
malformations, of which 86 underwent microsurgical resec-
intracranial vascular malformations in adults: the Scottish
tion. Over a 12-year period, 87 % were the same or better, Intracranial Vascular Malformation Study (SIVMS). Stroke. 2003;
10 % were worse, and 4 % died; the long-term neurological 34(5):1163–9.
morbidity rate was 12 %. Bertalanffy et al. reviewed their 2. Brown Jr RD, Flemming KD, Meyer FB, Cloft HJ, Pollock BE, Link
ML. Natural history, evaluation, and management of intracranial
series of 24 brainstem cavernous malformations over a
vascular malformations. Mayo Clin Proc. 2005;80(2):269–81.
5-year period [23]. Seventy percent had Karnofsky scores of 3. Aiba T, Tanaka R, Koike T, Kameyama S, Takeda N, Komata
90 or better at follow-up, 21 % had scores of 60–70, and 4 % T. Natural history of intracranial cavernous malformations.
had scores less than 60. Their rate of long-term morbidity J Neurosurg. 1995;83(1):56–9.
was 5.5 % and there were no deaths. Therefore, while the 4. Brown Jr RD, Wiebers DO, Torner JC, O’Fallon WM. Frequency of
intracranial hemorrhage as a presenting symptom and subtype
risks of surgical intervention for brainstem and deep-brain analysis: a population-based study of intracranial vascular malfor-
cavernous malformations are not low, the natural history of mations in Olmsted Country, Minnesota. J Neurosurg. 1996;85(1):
such lesions is likely worse than that of cavernous malforma- 29–32.
tions at other CNS sites, and we recommend that surgical 5. Detwiler PW, Porter RW, Zabramski JM, Spetzler RF. De novo for-
mation of a central nervous system cavernous malformation: impli-
resection be considered for symptomatic lesions that present cations for predicting risk of hemorrhage. Case report and review of
to a pial or ependymal surface and in certain circumstances the literature. J Neurosurg. 1997;87(4):629–32.
even for those malformations that do not present to a pial/ 6. Giombini S, Morello G. Cavernous angiomas of the brain. Account
ependymal surface, if they have bled repetitively and can be of fourteen personal cases and review of the literature. Acta
Neurochir (Wien). 1978;40(1–2):61–82.
exposed through a safe corridor [86]. 7. Sarwar M, McCormick WF. Intracerebral venous angioma. Case
report and review. Arch Neurol. 1978;35(5):323–5.
8. Maraire JN, Awad IA. Intracranial cavernous malformations: lesion
Conclusions behavior and management strategies. Neurosurgery. 1995;37(4):
591–605.
9. Moriarity JL, Clatterbuck RE, Rigamonti D. The natural history of
Surgical resection of symptomatic cavernous malformations cavernous malformations. Neurosurg Clin N Am. 1999;10(3): 411–7.
provides an excellent treatment option and, when completely 10. Raychaudhuri R, Batjer HH, Awad IA. Intracranial cavernous angi-
excised, is curative for most lesions. When cavernous mal- oma: a practical review of clinical and biological aspects. Surg
formations are removed in patients with medically refractory Neurol. 2005;63(4):319–28; discussion 28.
11. Del Curling O, Kelly Jr DL, Elster AD, Craven TE. An analysis of
epilepsy, improved seizure control is typical and many cen- the natural history of cavernous angiomas. J Neurosurg.
ters report high rates of “seizure-free” patients. In general, 1991;75(5):702–8.
the surgical morbidity associated with the resection of super- 12. Herter T, Brandt M, Szuwart U. Cavernous hemangiomas in chil-
ficial lesions in the cerebral and cerebellar cortices is mini- dren. Childs Nerv Syst. 1988;4(3):123–7.
13. Kondziolka D, Lunsford LD, Flickinger JC, Kestle JR. Reduction
mal, and mortality is considered unusual. Resection of of hemorrhage risk after stereotactic radiosurgery for cavernous
lesions in and around eloquent cortex requires electrophysi- malformations. J Neurosurg. 1995;83(5):825–31.
ological monitoring and often functional mapping, but can 14. Robinson JR. Clinical spectrum and natural history. In: Awad IA,
be performed safely. Deep-seated cavernous malformations Barrow DL, editors. Cavernous malformations. Park Ridge: AANS;
1993. p. 25–36.
in the thalamus, basal ganglia, and brainstem are associated 15. Scott RM, Barnes P, Kupsky W, Adelman LS. Cavernous angiomas
with higher surgical risks; however, for symptomatic lesions of the central nervous system in children. J Neurosurg. 1992;
that demonstrate recurrent hemorrhage and present to a pial 76(1):38–46.
or ventricular surface, resection by an experienced neurosur- 16. Kim DS, Park YG, Choi JU, Chung SS, Lee KC. An analysis of the
natural history of cavernous malformations. Surg Neurol.
geon is well reported and should be considered. The role of 1997;48(1):9–17; discussion 8.
radiosurgery for unresectable malformations remains uncer- 17. Kondziolka D, Lunsford LD, Kestle JR. The natural history of cere-
tain and at our institution was associated with significant bral cavernous malformations. J Neurosurg. 1995;83(5):820–4.
risks of radiation-induced injury without any significant 18. Pollock BE, Garces YI, Stafford SL, Foote RL, Schomberg PJ, Link
MJ. Stereotactic radiosurgery for cavernous malformations.
change in the natural history of hemorrhaging. Until better J Neurosurg. 2000;93(6):987–91.
evidence of its efficacy is established, we do not recommend 19. Robinson JR, Awad IA, Little JR. Natural history of the cavernous
radiosurgery for treatment of cavernous malformations. angioma. J Neurosurg. 1991;75(5):709–14.
20. Zabramski JM, Wascher TM, Spetzler RF, Johnson B, Golfinos J, 38. Riant F, Bergametti F, Ayrignac X, Boulday G, Tournier-Lasserve
Drayer BP, et al. The natural history of familial cavernous malfor- E. Recent insights into cerebral cavernous malformations: the
mations: results of an ongoing study. J Neurosurg. 1994;80(3): molecular genetics of CCM. FEBS J. 2010;277(5):1070–5.
422–32. 39. Wang QK. Update on the molecular genetics of vascular anomalies.
21. Porter RW, Detwiler PW, Spetzler RF, Lawton MT, Baskin JJ, Lymphat Res Biol. 2005;3(4):226–33.
Derksen PT, et al. Cavernous malformations of the brainstem: expe- 40. Gunel M, Awad IA, Finberg K, Anson JA, Steinberg GK, Batjer
rience with 100 patients. J Neurosurg. 1999;90(1):50–8. HH, et al. A founder mutation as a cause of cerebral cavernous mal-
22. Hsu F, Rigamonti D, Huhn S. Epidemiology of cavernous malfor- formation in Hispanic Americans. N Engl J Med. 1996;334(15):
mations. In: Awad IA, Barrow DL, editors. Cavernous malforma- 946–51.
tions. Park Ridge: AANS; 1993. p. 13–23. 41. Denier C, Gasc JM, Chapon F, Domenga V, Lescoat C, Joutel A,
23. Bertalanffy H, Benes L, Miyazawa T, Alberti O, Siegel AM, Sure et al. Krit1/cerebral cavernous malformation 1 mRNA is preferen-
U. Cerebral cavernomas in the adult. Review of the literature and tially expressed in neurons and epithelial cells in embryo and adult.
analysis of 72 surgically treated patients. Neurosurg Rev. Mech Dev. 2002;117(1–2):363–7.
2002;25(1–2):1–53; discussion 4–5. 42. Guzeloglu-Kayisli O, Kayisli UA, Amankulor NM, Voorhees JR,
24. Sandalcioglu IE, Wiedemayer H, Secer S, Asgari S, Stolke Gokce O, DiLuna ML, et al. Krev1 interaction trapped-1/cerebral
D. Surgical removal of brain stem cavernous malformations: surgi- cavernous malformation-1 protein expression during early angio-
cal indications, technical considerations, and results. J Neurol genesis. J Neurosurg. 2004;100(5 Suppl Pediatrics):481–7.
Neurosurg Psychiatry. 2002;72(3):351–5. 43. Petit N, Blecon A, Denier C, Tournier-Lasserve E. Patterns of
25. Steinberg GK, Chang SD, Gewirtz RJ, Lopez JR. Microsurgical expression of the three cerebral cavernous malformation (CCM)
resection of brainstem, thalamic, and basal ganglia angiographically genes during embryonic and postnatal brain development. Gene
occult vascular malformations. Neurosurgery. 2000;46(2):260–70; Expr Patterns. 2006;6(5):495–503.
discussion 70–1. 44. Hilder TL, Malone MH, Bencharit S, Colicelli J, Haystead TA,
26. Campbell PG, Jabbour P, Yadla S, Awad IA. Emerging clinical Johnson GL, et al. Proteomic identification of the cerebral cavernous
imaging techniques for cerebral cavernous malformations: a sys- malformation signaling complex. J Proteome Res. 2007;6(11):
tematic review. Neurosurg Focus. 2010;29(3):E6. 4343–55.
27. Denier C, Goutagny S, Labauge P, Krivosic V, Arnoult M, Cousin 45. Voss K, Stahl S, Schleider E, Ullrich S, Nickel J, Mueller TD, et al.
A, et al. Mutations within the MGC4607 gene cause cerebral cav- CCM3 interacts with CCM2 indicating common pathogenesis for
ernous malformations. Am J Hum Genet. 2004;74(2):326–37. cerebral cavernous malformations. Neurogenetics. 2007;8(4):249–56.
28. Labauge P, Laberge S, Brunereau L, Levy C, Tournier-Lasserve 46. Zawistowski JS, Serebriiskii IG, Lee MF, Golemis EA, Marchuk
E. Hereditary cerebral cavernous angiomas: clinical and genetic DA. KRIT1 association with the integrin-binding protein ICAP-1: a
features in 57 French families. Societe Francaise de Neurochirurgie. new direction in the elucidation of cerebral cavernous malforma-
Lancet. 1998;352(9144):1892–7. tions (CCM1) pathogenesis. Hum Mol Genet. 2002;11(4):389–96.
29. Rigamonti D, Hadley MN, Drayer BP, Johnson PC, Hoenig- 47. Zhang J, Rigamonti D, Dietz HC, Clatterbuck RE. Interaction
Rigamonti K, Knight JT, et al. Cerebral cavernous malformations. between krit1 and malcavernin: implications for the pathogenesis
Incidence and familial occurrence. N Engl J Med. 1988;319(6): of cerebral cavernous malformations. Neurosurgery.
343–7. 2007;60(2):353–9; discussion 9.
30. Kleaveland B, Zheng X, Liu JJ, Blum Y, Tung JJ, Zou Z, et al. 48. Ma X, Zhao H, Shan J, Long F, Chen Y, Chen Y, et al. PDCD10
Regulation of cardiovascular development and integrity by the interacts with Ste20-related kinase MST4 to promote cell growth
heart of glass-cerebral cavernous malformation protein pathway. and transformation via modulation of the ERK pathway. Mol Biol
Nat Med. 2009;15(2):169–76. Cell. 2007;18(6):1965–78.
31. Laberge-le Couteulx S, Jung HH, Labauge P, Houtteville JP, 49. Uhlik MT, Abell AN, Johnson NL, Sun W, Cuevas BD, Lobel-Rice
Lescoat C, Cecillon M, et al. Truncating mutations in CCM1, KE, et al. Rac-MEKK3-MKK3 scaffolding for p38 MAPK activation
encoding KRIT1, cause hereditary cavernous angiomas. Nat Genet. during hyperosmotic shock. Nat Cell Biol. 2003;5(12):1104–10.
1999;23(2):189–93. 50. Zhang J, Basu S, Rigamonti D, Dietz HC, Clatterbuck RE. Krit1
32. Liquori CL, Berg MJ, Siegel AM, Huang E, Zawistowski JS, modulates beta 1-integrin-mediated endothelial cell proliferation.
Stoffer T, et al. Mutations in a gene encoding a novel protein con- Neurosurgery. 2008;63(3):571–8; discussion 8.
taining a phosphotyrosine-binding domain cause type 2 cerebral 51. Faurobert E, Albiges-Rizo C. Recent insights into cerebral cavern-
cavernous malformations. Am J Hum Genet. 2003;73(6):1459–64. ous malformations: a complex jigsaw puzzle under construction.
33. Bergametti F, Denier C, Labauge P, Arnoult M, Boetto S, Clanet M, FEBS J. 2010;277(5):1084–96.
et al. Mutations within the programmed cell death 10 gene cause 52. Labauge P, Denier C, Bergametti F, Tournier-Lasserve E. Genetics
cerebral cavernous malformations. Am J Hum Genet. 2005;76(1): of cavernous angiomas. Lancet Neurol. 2007;6(3):237–44.
42–51. 53. Schleider E, Stahl S, Wustehube J, Walter U, Fischer A, Felbor
34. Dibble CF, Horst JA, Malone MH, Park K, Temple B, Cheeseman U. Evidence for anti-angiogenic and pro-survival functions of the
H, et al. Defining the functional domain of programmed cell death cerebral cavernous malformation protein 3. Neurogenetics.
10 through its interactions with phosphatidylinositol-3,4,5- 2011;12(1):83–6.
trisphosphate. PLoS One. 2010;5(7):e11740. 54. Cavalcanti DD, Kalani MY, Martirosyan NL, Eales J, Spetzler RF,
35. Liquori CL, Berg MJ, Squitieri F, Ottenbacher M, Sorlie M, Preul MC. Cerebral cavernous malformations: from genes to pro-
Leedom TP, et al. Low frequency of PDCD10 mutations in a panel teins to disease. J Neurosurg. 2012;116(1):122–32.
of CCM3 probands: potential for a fourth CCM locus. Hum Mutat. 55. Hogan BM, Bussmann J, Wolburg H, Schulte-Merker S. ccm1 cell
2006;27(1):118. autonomously regulates endothelial cellular morphogenesis and
36. Brouillard P, Vikkula M. Genetic causes of vascular malformations. vascular tubulogenesis in zebrafish. Hum Mol Genet. 2008;17(16):
Hum Mol Genet. 2007;16(Spec No. 2):R140-9. 2424–32.
37. Revencu N, Vikkula M. Cerebral cavernous malformation: new 56. Plummer NW, Gallione CJ, Srinivasan S, Zawistowski JS, Louis
molecular and clinical insights. J Med Genet. 2006;43(9):716–21. DN, Marchuk DA. Loss of p53 sensitizes mice with a mutation in
Ccm1 (KRIT1) to development of cerebral vascular malformations. 73. Amin-Hanjani S, Ogilvy CS, Candia GJ, Lyons S, Chapman
Am J Pathol. 2004;165(5):1509–18. PH. Stereotactic radiosurgery for cavernous malformations:
57. Plummer NW, Squire TL, Srinivasan S, Huang E, Zawistowski JS, Kjellberg’s experience with proton beam therapy in 98 cases at the
Matsunami H, et al. Neuronal expression of the Ccm2 gene in a new Harvard Cyclotron. Neurosurgery. 1998;42(6):1229–36; discussion
mouse model of cerebral cavernous malformations. Mamm 36–8.
Genome. 2006;17(2):119–28. 74. Chang SD, Levy RP, Adler Jr JR, Martin DP, Krakovitz PR,
58. Whitehead KJ, Plummer NW, Adams JA, Marchuk DA, Li Steinberg GK. Stereotactic radiosurgery of angiographically occult
DY. Ccm1 is required for arterial morphogenesis: implications for vascular malformations: 14-year experience. Neurosurgery.
the etiology of human cavernous malformations. Development. 1998;43(2):213–20; discussion 20–1.
2004;131(6):1437–48. 75. Fabrikant JI, Levy RP, Steinberg GK, Phillips MH, Frankel KA,
59. Haasdijk RA, Cheng C, Maat-Kievit AJ, Duckers HJ. Cerebral cavern- Silverberg GD. Stereotactic charged-particle radiosurgery: clinical
ous malformations: from molecular pathogenesis to genetic counsel- results of treatment of 1200 patients with intracranial arteriovenous
ling and clinical management. Eur J Hum Genet. 2012;20(2):134–40. malformations and pituitary disorders. Clin Neurosurg. 1992;38:
60. Whitehead KJ, Chan AC, Navankasattusas S, Koh W, London NR, 472–92.
Ling J, et al. The cerebral cavernous malformation signaling path- 76. Kida Y, Kobayashi T, Tanaka T. Treatment of symptomatic AOVMs
way promotes vascular integrity via Rho GTPases. Nat Med. with radiosurgery. Acta Neurochir Suppl. 1995;63:68–72.
2009;15(2):177–84. 77. Kondziolka D, Lunsford LD, Coffey RJ, Bissonette DJ, Flickinger
61. Yadla S, Jabbour PM, Shenkar R, Shi C, Campbell PG, Awad JC. Stereotactic radiosurgery of angiographically occult vascular
IA. Cerebral cavernous malformations as a disease of vascular per- malformations: indications and preliminary experience.
meability: from bench to bedside with caution. Neurosurg Focus. Neurosurgery. 1990;27(6):892–900.
2010;29(3):E4. 78. LeDoux MS, Aronin PA, Odrezin GT. Surgically treated cavernous
62. Barker 2nd FG, Amin-Hanjani S, Butler WE, Lyons S, Ojemann angiomas of the brain stem: report of two cases and review of the
RG, Chapman PH, et al. Temporal clustering of hemorrhages from literature. Surg Neurol. 1991;35(5):395–9.
untreated cavernous malformations of the central nervous system. 79. Levy RP, Fabrikant JI, Frankel KA, Phillips MH, Lyman
Neurosurgery. 2001;49(1):15–24; discussion 5. JT. Charged-particle radiosurgery of the brain. Neurosurg Clin N
63. Barrow DL, Krisht A. Cavernous malformations and hemorrhage. Am. 1990;1(4):955–90.
In: Awad IA, Barrow DL, editors. Cavernous malformations. Park 80. Seo Y, Fukuoka S, Takanashi M, Nakagawara J, Suematsu K,
Ridge: AANS; 1993. p. 65–80. Nakamura J, et al. Gamma Knife surgery for angiographically
64. Gross BA, Lin N, Du R, Day AL. The natural history of intracranial occult vascular malformations. Stereotact Funct Neurosurg.
cavernous malformations. Neurosurg Focus. 2011;30(6):E24. 1995;64 Suppl 1:98–109.
65. Clatterbuck RE, Moriarity JL, Elmaci I, Lee RR, Breiter SN, 81. Steinberg GK, Levy RP, Fabrikant JI, Frankel KA, Phillips MH,
Rigamonti D. Dynamic nature of cavernous malformations: a pro- Marks MP. Stereotactic helium ion Bragg peak radiosurgery for
spective magnetic resonance imaging study with volumetric analy- angiographically occult intracranial vascular malformations.
sis. J Neurosurg. 2000;93(6):981–6. Stereotact Funct Neurosurg. 1991;57(1–2):64–71.
66. Kupersmith MJ, Kalish H, Epstein F, Yu G, Berenstein A, Woo H, 82. Lunsford LD, Khan AA, Niranjan A, Kano H, Flickinger JC,
et al. Natural history of brainstem cavernous malformations. Kondziolka D. Stereotactic radiosurgery for symptomatic solitary
Neurosurgery. 2001;48(1):47–53; discussion 4. cerebral cavernous malformations considered high risk for resec-
67. Labauge P, Brunereau L, Levy C, Laberge S, Houtteville JP. The tion. J Neurosurg. 2010;113(1):23–9.
natural history of familial cerebral cavernomas: a retrospective 83. Gewirtz RJ, Steinberg GK, Crowley R, Levy RP. Pathological
MRI study of 40 patients. Neuroradiology. 2000;42(5):327–32. changes in surgically resected angiographically occult vascular
68. Rigamonti D, Hsu F, Huhn S. Angiograhically occult vascular mal- malformations after radiation. Neurosurgery. 1998;42(4):738–42;
formations. In: Carter LP, Spetzler RF, editors. Neurovascular sur- discussion 42–3.
gery. New York: McGraw-Hill; 1994. p. 521–40. 84. Ferroli P, Sinisi M, Franzini A, Giombini S, Solero CL, Broggi
69. Awad IA, Robinson JR. Cavernous malformations and epilepsy. In: G. Brainstem cavernomas: long-term results of microsurgical
Awad IA, Barrow DL, editors. Cavernous malformations. Park resection in 52 patients. Neurosurgery. 2005;56(6):1203–12; dis-
Ridge: AANS; 1993. p. 49–63. cussion 12–4.
70. Vaquero J, Salazar J, Martinez R, Martinez P, Bravo G. Cavernomas 85. Fritschi JA, Reulen HJ, Spetzler RF, Zabramski JM. Cavernous
of the central nervous system: clinical syndromes, CT scan diagno- malformations of the brain stem. A review of 139 cases. Acta
sis, and prognosis after surgical treatment in 25 cases. Acta Neurochir (Wien). 1994;130(1–4):35–46.
Neurochir (Wien). 1987;85(1–2):29–33. 86. Pandey P, Westbroek EM, Gooderham PA, Steinberg GK.
71. McCormick PC. Management of intracranial cavernous and venous Cavernous malformation of brainstem, thalamus, and basal gan-
malformations. In: Barrow DL, editor. Intracranial vascular malfor- glia: a series of 176 patients. Neurosurgery. 2013;72(4):573–89;
mations: neurosurgical topics. Park Ridge: AANS; 1990. p. 197–218. discussion 88–9.
72. Wascher TM, Spetzler RF. Cavernous Malformations of the brain 87. Wang CC, Liu A, Zhang JT, Sun B, Zhao YL. Surgical management
stem. In: Carter LP, Spetzler RF, editors. Neurovascular surgery. of brain-stem cavernous malformations: report of 137 cases. Surg
New York: McGraw-Hill; 1994. p. 541–55. Neurol. 2003;59(6):444–54; discussion 54.
Trigeminal Neuralgia
52
Lawrence S. Chin, Seung S. Hahn, Shilpen Patel,
Thomas Mattingly, and Young Kwok
The diagnosis is based on history. Typically, the neurologic

Introduction exam is normal, unless the syndrome is secondary to another
process such as a tumor in the posterior fossa or multiple scle-
Trigeminal neuralgia (TN) is a facial pain syndrome charac- rosis. Imaging studies obtained in the workup include mag-
terized by a severe electrical, shooting pain lasting seconds netic resonance imaging (MRI) to rule out structural causes,
to minutes that is confined to one or more distributions of the but usually this is normal. An atypical form of TN exists and
trigeminal nerve. More recently, the International Headache is characterized by pain of longer duration that may be asso-
Society has defined TN pain as the following: (a) paroxys- ciated with a burning sensation and facial numbness [4].
mal attacks, lasting from a second to 2 min, affecting one or Occasionally, patients with typical TN may develop atypical
more divisions of the trigeminal nerve; (b) pain that is features over time. It is not unusual for TN to have periods of
intense, sharp, superficial or stabbing, precipitated from trig- remissions and exacerbations; seasonal variation is also com-
ger areas or factors; (c) attacks stereotyped in the individual mon with winter being a particularly difficult time for most
patient; (d) no clinically evident neurologic deficit; and (e) sufferers. Patients with TN usually note a progression of their
pain not attributed to another disorder [1]. The pain, typi- symptoms with time: the pain becomes more intense and the
cally unilateral, may be bilateral in approximately 5 % of remissions more rare.
cases and may occur either spontaneously or be triggered by The pathophysiology of this process is somewhat contro-
innocuous sensations or activities such as a gust of wind, versial. As a result of surgical observations in the posterior
chewing, drinking, brushing teeth, shaving, or washing the fossa, Dandy postulated that vascular compression, usually
face. Severely affected patients may lose weight from not from a branch of the superior cerebellar artery at the trigemi-
eating. The median age of presentation is in the sixties, and nal root, was the primary cause of TN [5]. This idea was
it affects women more often than men [2]. The incidence of advanced by Gardner, Jannetta, and others who developed
trigeminal neuralgia is 4–13 per 100,000 people and approx- microvascular decompression (MVD) via retromastoid cra-
imately 15,000 new cases occur in USA each year [3]. niectomy as a treatment [6]. It is postulated that arterial pul-
sation of the trigeminal nerve root entry zone results in
demyelination and subsequent ephaptic transmission [7].
L.S. Chin, M.D., F.A.C.S., F.A.A.N.S. (*)
Department of Neurosurgery, SUNY Upstate Medical University, However, not all patients with typical TN have vascular com-
750 East Adams Street, Syracuse, NY 13210, USA pression found at MVD, and vascular compression of the
e-mail: chinL@upstate.edu trigeminal roots has been observed in patients without TN. In
S.S. Hahn addition, patients with multiple sclerosis (MS) and brain-
Department of Radiation Oncology, SUNY Upstate Medical stem infarcts may develop TN, indicating that TN may be
University, 750 East Adams Street, Syracuse, NY 13210, USA
generated by central mechanisms as well [8].
S. Patel
Department of Radiation Oncology, University of Washington
Medical Center, Seattle, WA, USA
Treatment Options
T. Mattingly
Department of Neurosurgery, University of Maryland,
Baltimore, MD, USA Patients with TN have a number of treatment options. The first
line is medical therapy. Carbamazepine (Tegretol) is the drug
Y. Kwok
Department of Radiation Oncology, University of Maryland of choice, but other anticonvulsants such as gabapentin
Medical Center, Baltimore, MD, USA (Neurontin), phenytoin (Dilantin), oxcarbazepine (Trileptal),
Table 52.1 Results of linac-based or CyberKnife SRS in idiopathic TN

No. of Median follow-up BNI I–III Time to Median Recurrence Complication
patients (months) relief (%) pain relief dose (Gy) rate (%) rate (%)
Chen et al. [11] (Kaiser Los Angeles) 32 8 78 1.5 months NR NR 6
Richards et al. [12] (Wisconsin) 28 12 75 1 month 80 46 14
Lim et al. [13] CyberKnife (Stanford) 41 11 93 7 days 78 16 51
Smith et al. [14] (UCLA) 60 23 72 2.7 months 90 26 25
NR not reported
lamotrigine (Lamictal), and topiramate (Topamax) can also be With recent improvements in technology and imaging,
effective. Baclofen (Kemstro) has also been used though it has linear accelerator (linac)-based and CyberKnife SRS for
a different mechanism of action; it acts by inhibiting mono- TN have become more common. Table 52.1 compares
synaptic and polysynaptic spinal reflexes. In severe cases, two some of the major series reported in the literature. The
or more drugs may be required. Narcotic pain medications are range of patients achieving good relief (Barrow
typically not recommended because most trigeminal attacks Neurological Institute [BNI] I–III) is 72–93 %. The median
do not last for a significant duration to justify routine use of time to pain relief varies from 1 week to 2.7 months. An
such medications. Medication may only help in approxi- analysis of recurrence revealed rates that ranged from 6 to
mately 75 % of patients, and side effects such as drowsiness, 51 % [13–16].
cognitive dysfunction, bone marrow suppression, and hepatic Ma from the University of Maryland performed a
dysfunction can be significant. In addition, patients will fre- comparative analysis of linac and GK-SRS for TN therapy.
quently become refractory to medication over time [9]. Specifically, the dose fall-off characteristics and set-up error
If medical therapy fails or is not tolerated, there exist two tolerance of the two modalities were examined. This study
primary surgical therapies: MVD and ablative percutaneous found equivalent dose fall-off properties between Gamma
procedures. Performing an MVD requires general anesthesia Knife and linac-based SRS provided only if a sufficient num-
and carries a risk of stroke, infection, hemorrhage, cerebro- ber of arcs (>7) and small intra-arc error (<0.5 mm) were
spinal fluid (CSF) leak, facial numbness, weakness, and satisfied for linac-based deliveries. Furthermore, because of
hearing loss. For patients over the age of 70 years and for the large number of arcs that are required, the treatment time
those with surgical contraindications, MVD may not be an for linac-based SRS compared with Gamma Knife is signifi-
ideal treatment. The percutaneous approaches (radiofre- cantly increased [17]. At this time, robust safety and long-
quency ablation, glycerol injection, and balloon compres- term efficacy data exist only for Gamma Knife treatment of
sion) to the trigeminal ganglion are performed under local TN; however, the clinical experience in this disease with
anesthesia but are associated with a significant risk of tri- other forms of radiosurgery is growing and will lead to
geminal dysfunction. Stereotactic radiosurgery (SRS) pro- greater patient options.
vides an alternative treatment that is minimally invasive yet
effective and causes few side effects.
SRS was first described in 1951 by Lars Leksell as an Mechanism of GK-SRS Effect
adaptation of his arc-centered stereotactic frame with a radia-
tion source replacing the lesioning electrode. In his initial The relationship between post-procedure numbness and effi-
attempts at SRS, Leksell coupled a dental X-ray tube to his cacy suggests that GK-SRS works by blocking axonal trans-
stereotactic arc to irradiate the trigeminal ganglion [10]. After mission. As predicted by models of radiation injury, both the
trying several different radiation sources, he subsequently time to effective pain relief and numbness are delayed,
reported the successful treatment of TN in two patients with although pain relief frequently occurs many months before
Gamma Knife stereotactic radiosurgery (GK-SRS) using any side effects are experienced. Kondziolka et al. used a pri-
cobalt-60 [11]. With the development of high-resolution mate model to explore the effects of 80 or 100 Gy to trigemi-
MRI, the proximal portion of trigeminal nerve is usually well nal nerves and observed a combination of axonal degeneration
identified between the pons and the Meckel’s cave, and con- and edema [18]. Necrosis was seen in nerves that received
sidered to be a well-defined and safer target for SRS. the higher dose, and both myelinated and unmyelinated
Thus, the radiosurgery target has moved toward root entry fibers were equally affected. Why the functional improve-
zone of trigeminal nerve which allowed much higher radio- ment is seen in patients before these histologic changes are
surgery dose (70–90 Gy) without risking significant side seen is unknown, but an effect of GK-SRS on ephaptic trans-
effects [12]. mission provides a possible mechanism.
52 Trigeminal Neuralgia 651
Targeting, Dose, and Dose Rate
Early SRS treatments for TN targeted the trigeminal ganglion

because the trigeminal cistern in the Meckel’s cave could be
inferred from landmarks on skull films, a practice later
refined by stereotactic cisternography. In fact, glycerol rhi-
zotomy was discovered incidentally during the use of glyc-
erol as a carrier for tantalum dust in outlining the trigeminal
cistern [19]. In 1991, Lindquist et al. reported 46 cases of
GK-SRS directed at the trigeminal ganglion with 22 patients
localized using cisternography [20]. They did not report pre-
scription doses. The results were disappointing as only 4 of
22 patients had prolonged pain relief. Rand et al. reported a
series of 12 patients treated by GK-SRS doses ranging from
57 to 74 Gy with only mild success and concluded that the
ganglion was not the appropriate target [21]. Based on these
findings and with the development of high-resolution MRI,
which made it possible to reliably target the trigeminal nerve
anywhere along its course from the pons to the Meckel’s Fig. 52.1 Gamma Knife treatment plan for a patient with right trigemi-
cave, the optimal target moved toward the trigeminal nerve nal neuralgia. Isodose coverage of the nerve is seen in the right upper
root entry zone. panel. The 20 % line touches the brain stem, and the 10 % line cuts into
In 1996, Kondziolka et al. published the results of a mul- it. Beam shaping is accomplished by plugging 32 beams as shown in the
left lower panel. The result, seen in the right lower panel, is a flattened
ticenter trial of GK-SRS that standardized the target [22]. 10 and 20 % isodose line such that the brain stem receives 10 % only
A single 4-mm isocenter was applied to the trigeminal nerve along the edge
2–4 mm anterior to the junction of the nerve at the pons as
seen on axial and coronal MRI scans. This location has since
been confirmed effective by several large series [23]. The In Leksell’s original 1971 paper, he described successful
20 % isocenter line typically contacts the brain stem using treatment at maximum doses of 16.5 and 22 Gy [11]. Patients
this protocol, which is well tolerated at the current maximum studied in the 1996 multicenter trial received doses ranging
doses of 70–80 Gy to the isocenter. Evaluation of brain-stem from 60 to 90 Gy [22]. Patients who received at least 70 Gy
dose as well as evidence of compression was evaluated at the fared significantly better, but no difference was seen by
University of Maryland. Doses to the brain stem ranged from increasing the dose to 90 Gy. At these doses, the edge of the
10 to 20 %, and evidence of compression was seen in 12 % brain stem receives no more than 16 Gy, which is well
of patients. Both brain-stem dose and evidence of compres- tolerated.
sion did not have any effect on treatment response, medica- At the University of Maryland, MRI imaging is performed
tion use, or development of facial numbness [24]. Beam with gadolinium-enhanced short-TR sequences (T1-weighted)
shaping is accomplished by plugging some of the 201 sources with axial volume acquisitions using 512 × 216 matrices at
during a treatment and can be used to accomplish a flattening 1.5-mm slice intervals. Occasionally, it is difficult to visual-
of the lower isodose curves such that the 10 % line is parallel ize the affected prepontine trigeminal nerve on the axial MRI
to the brain stem (Fig. 52.1). Theoretically, this should view. This was evaluated and it was found that MRI quality
decrease the risk of radiation necrosis to the brain stem. did not affect treatment response, medication use, and devel-
Some centers have also tried using two isocenters to irradiate opment of facial numbness [24]. Based on this study, it is
a greater length of nerve. A randomized, prospective study recommended that if the trigeminal nerve is difficult to image,
reported by Flickinger et al. found no increase in efficacy but additional axial long-TR sequences (T2-weighted) or use of
did show a statistically significant increase in trigeminal dys- the coronal and sagittal views is useful (Case 52.1). A single
function [25]. Efficacy was reported to be 67.7 ± 5.1 % in 4-mm isocenter shot is placed just anterior to the entry of the
both arms, and facial numbness and mild and severe pares- trigeminal nerve into the pons with the 20 % isocenter barely
thesias developed in 19 %, 12 %, and 2 % in two isocenter touching the pons surface. Coronal MRI scans are examined
patients versus 7 %, 9 %, and 0% in one isocenter patient, to ensure that the 50 % line surrounds the nerve, and 90 and
respectively. In the multivariate analysis, the length of nerve 99 % lines are evaluated to make sure the center of the nerve
treated was the only significant factor predicting for facial receives the maximum dose. A customized plugging pattern
numbness. is then applied that elongates the lower isodose curves and
Case 52.1 Case 52.2

This 42-year-old man with a 10-year history of well- This 60-year-old man had a 5-year history of typical
controlled multiple sclerosis began experiencing typi- trigeminal neuralgia that initially responded to glyc-
cal right V3 trigeminal neuralgia 4 years prior to erol rhizotomy. His pain recurred and he required
treatment. He had been placed on increasing doses of treatment with Tegretol, which was becoming ineffec-
Tegretol and Neurontin with decreasing effectiveness. tive. One year prior to GK-SRS, he had a cardiac defi-
The increased doses of medication also worsened his brillator placed for arrhythmias. A cisternogram was
MS symptoms. He had typical lancinating pain with no performed through a C1–2 puncture. Axial CT scans
numbness that was triggered by touching the face and were obtained that showed the right trigeminal nerve
talking. At the time of GK-SRS, he had T1- and (Fig. 52.3). The nerve was targeted according to proto-
T2-weighted MRI scans done. Because the T1 scans col with a 4-mm shot of radiation with a total dose of
did not show the nerve, we used the T2 scan to target 75 Gy delivered to the isocenter. Three weeks after
the nerve (Fig. 52.2a, b). A 4-mm shot was used with a treatment, his pain disappeared. He remains pain-free
plugging pattern to shape the 10 and 20 % isodose 1 year after treatment.
lines. A dose of 75 Gy was delivered to the 100 % line.
One month after GK-SRS, he noticed the sudden
improvement in his pain. At 2 months after treatment,
he was weaned off both medications. He remains pain-
free for 2 years.
Fig. 52.2 MRI images from the Gamma Knife treatment of a patient
with right trigeminal neuralgia. (a) T1-weighted MRI scan that shows
the left nerve (arrow) but not the right. (b) T2-weighted MRI scan at the
same level illustrates the right nerve well (arrow)
Fig. 52.3 CT cisternogram of a patient with right trigeminal neuralgia
and a cardiac defibrillator. CT contrast in the CSF cisterns around the
trigeminal nerve highlights it well (arrow)
minimizes the volume of brain stem covered by the 10 % line.
Finally, a dose of 75 Gy is delivered to the 100 % isocenter
(37.5 Gy to the 50 %). dose. As the dose rate is lowered and the exposure time
Patients with a pacemaker or other MRI contraindications extended, the biologic effect of a given dose is generally
present a dilemma. Provided the patient is not on anticoagu- reduced. The University of Maryland examined this effect in
lants, computed tomography (CT) contrast cisternography is a series of patients treated the year prior to a source change
used for localizing the trigeminal nerve (Case 52.2). In rare (median dose rate, 162 cGy/min: range, 151 to 179 cGy/min)
circumstances, an MRI scan in patients with a pacemaker is compared with patients treated in the year immediately after
required. A cardiologist is present in these cases in the MRI a source change (median dose rate, 343 cGy/min: range, 321
suite with a defibrillator, and the pacemaker is programmed to 366 cGy/min) and found the difference in pain relief was
before the scan to avoid extraventricular pacing and is then 61 % compared with 83 %, respectively (p = 0.05). These
reset immediately after the MRI. data suggest that if dose rate falls below 179 cGy/min, con-
For gamma rays, dose rate is one of the principal factors sideration should be made for dose escalation in the treat-
that determine the biologic consequences of a given absorbed ment of TN [26].
A comparison of selected major series reported in the

Treatment Response literature is shown in Table 52.4 [22, 23, 27–32]. In general,
there is good correlation between the different series.
Assessing positive treatment responses in TN is challenging The range of patients achieving good relief (BNI I to III) is
because the symptoms are subjective. The BNI pain intensity 69–94 % with 22–76 % achieving complete pain relief on no
score combines severity of pain with medication use medications (BNI I). The median time to pain relief varies
(Table 52.2) [27]. Patients able to stop all medications are clas- from 2 to 4 weeks. An actuarial analysis of recurrence has
sified as BNI I if they have complete pain relief or BNI II if shown rates of 23 %, 33 %, and 39 % at 1, 2, and 3 years,
they have occasional tolerable pain. Patients remaining on respectively [28]. A follow-up of the original Maryland
medication are BNI III if they have good pain relief, but are series (median follow-up 5.6 years) shows that recurrent pain
BNI IV if they are not adequately controlled. Patients receiving become more common over time. At 5 and 7 years after treat-
no relief are classified BNI V. Most patients are BNI V before ment, only 34 and 22 % of patients were free of recurrence
their treatment. For comparison with other series, we consider [33]. Quality of life assessments have documented that most
BNI I to III to be a good outcome. The McGill pain score is patients believe their GK-SRS treatments were successful
useful in the preoperative patient assessment (Table 52.3). even when the relief was temporary. The lack of side effects
Patients are asked to rate their TN pain and their worst non-TN appears significant in this regard, and a high proportion of
headache using a range from mild (I) to excruciating (V). patients who fail initial GK-SRS elect for an additional
treatment.
Table 52.2 Barrow Neurological Institute pain intensity score A variety of prognostic factors have been found to corre-
I No trigeminal pain, no medications
late with treatment outcome. A universal finding in all forms
II Occasional pain that is well tolerated, no medications of surgical therapy for TN is that the patient’s initial treat-
III Occasional pain that requires medication to be tolerated ment is the most effective; the results with GK-SRS are no
IV Pain that is not adequately controlled with medication different. Most studies find that the absence of a prior surgi-
V Severe pain without relief cal treatment correlates with a better outcome [23, 28, 29,
32]. Other preoperative prognostic factors for treatment suc-
cess include a shorter duration of symptoms and a patient
Table 52.3 McGill pain assessment score description (McGill pain scale) of their TN pain as severe
I Mild and their non-TN headaches as mild [28]. The implications
II Discomforting of this finding are unclear. Patients with more classic TN
III Distressing pain may have fewer secondary gain issues or conflicting
IV Horrible social issues that may impair their treatment. A prognostic
V Excruciating finding that has had conflicting results is the presence of
Table 52.4 Results of initial GK-SRS in idiopathic TN

Time to
No. of Median BNI I BNI I–III pain relief Recurrence Complication
patients follow-up relief (%) relief (%) (weeks) Median dose rate (%) rate (%)
Kondziolka et al. [22] 50 18 months 58 94 4 70 Gy to the 100 % 6 10
(multicenter)
Maesawa et al. [23] 220 22 months 40 69 8 80 Gy to the 100 % 13.6 7.7
(University of Pittsburgh)
Rogers et al. [27] (Barrow 54 12 months 35 89 2 35 Gy prescribed 21 9
Neurological Institute) to the 50 %
Petit et al. [28] (University 96 30 months 42 75 3 75 Gy to the 100 % 29 7.3
of Maryland)
Pollock et al. [32] (Mayo 117 26 months 58 75 3 80 Gy to the 100 % 23 37
Clinic)
Young et al. [29] (Northwest 110 19.8 months 76 88 2 70 Gy to the 100 % 34 2.7
Gamma Knife Center)
Sheehan et al. [30] 151 19 months 47 70 3.5 80 Gy to the 100 % 27 9
(University of Virginia)
Brisman et al. [31] 293 4.6 years 22 76 NR 76.8 Gy to the 24 12
(Columbia) 100 %
NR not reported
atypical TN symptoms. Whereas some studies indicate atyp-

ical symptoms carry an unfavorable prognosis, others have Treatment Options for Recurrent
not confirmed this [22, 23, 27–29]. Trigeminal Neuralgia
A minimum treatment dose of 70 Gy appears necessary
for a good outcome, but doses of 90 Gy and above do not Although the initial response to GK-SRS is often quite good,
appear to provide additional pain relief. The issue of dose the response may not be durable. As seen in patients treated
escalation, however, remains an open question for a number by MVD and the percutaneous procedures, recurrence rates
of reasons. The possibility of a type II error in the published increase with time. Using an actuarial analysis, this recur-
clinical series may mask a small benefit from a higher pre- rence rate for GK-SRS may be as high as 39 % by 3 years
scription dose. Also, the development of posttreatment facial after treatment. Petit found that absence of a prior surgical
numbness has been shown to correlate with good results, and treatment, duration of TN less than 50 months before
because facial numbness is dependent on the dose, an GK-SRS, and a BNI I response to GK-SRS correlated with
increased dose should result in an improved or more durable smaller risk of recurrence [28].
outcome [27, 32]. The development of more dysesthesias Most patients choose to undergo a repeat GK-SRS at first
with higher doses, however, may temper the desire to use recurrence. Data from four institutions on repeat GK-SRS
higher doses. for recurrent TN are presented in Table 52.5 [35–38].
Although there are variations in technique, primarily related
to the dose given at the second treatment, the results of repeat
Complications GK-SRS are generally quite good. The University of
Maryland reported 18 patients from a total series of 112
The distinguishing characteristic of GK-SRS for TN is the patients who underwent repeat GK-SRS for TN [35]. BNI I
very low morbidity and absence of mortality. Similar to the result was seen in 45 % and BNI I–III was seen in 33 %.
percutaneous ablative procedures, the primary side effect is Repeat GK-SRS was more effective for patients that experi-
trigeminal dysfunction. There is no risk to any cranial nerves enced a longer period of pain relief after their initial
other than cranial nerve V. The reported risk of posttreatment GK-SRS. Patients that had no response to initial GK-SRS
numbness or paresthesias varies from 2.7 to 37 % (Table 52.4) also failed to respond to their second treatment. Only 11 %
and is not bothersome for the majority. The primary risk fac- experienced increased numbness in this experience. The
tor for developing this complication is receiving a dose of target for the second treatment is usually not different from
90 Gy. Only one case of deafferentation pain/pain due to loss the initial treatment, although Hasegawa advocated placing
of sensory input into the central nervous system, the most the second lesion just anterior to the first one to reduce the
feared complication because it is so difficult to treat, has brain-stem dose [36]. The optimum dose to be used at second
been reported [23]. Corneal numbness, another dangerous treatment has not been established. Some centers use a lower
complication because of the potential risk of permanent dose at repeat treatment, whereas others have advocated
damage to the eye, has been reported in only three patients, increasing the second dose to 90 Gy.
all treated with 90 Gy [34]. No mastication motor deficits The University of Maryland recently presented data com-
were seen in any study. As expected, patients who undergo a paring treatment outcomes for patients treated with repeat
repeat GK-SRS treatment to the same trigeminal nerve have radiosurgery for refractory or recurrent TN with a low dose
a higher incidence of facial numbness, but there has not been (70–75 Gy) performed at the University of Maryland and a
an increase of dysesthesias or symptomatic radiation necro- high dose (90 Gy) performed at the University of Kentucky.
sis reported. A complication that is not reported but has been Similar rates of pain control were seen in both groups. The
observed in our patients is the temporary exacerbation of tri- higher dose was associated with an increased probability of
geminal pain in the initial month after GK-SRS. In no discontinuing all medications. It is worth noting though that
patients has this become permanent. 25 % of these patients had increased or new-onset numbness
Table 52.5 Results of repeat GK-SRS for idiopathic TN

Follow-up BNI class BNI class First dose Second Trigeminal
N (months) I (%) I–III (%) (Gy) dose (Gy) dysfunction (%)
Herman [35] (University of Maryland) 18/112 28 45 78 75 70 11
Hasegawa et al. [36] (University of Pittsburgh) 27/387 20.4 19 85 75 64 11
Shetter [37] (Barrow Neurological Institute) 19/240 13.5 53 85 78 46.6 42
Pollock [38] (Mayo Clinic) 10/100 15 80 80 70 90 80
or dysesthesias [39]. Not unexpectedly, using the higher GK-SRS for secondary TN has a much lower rate of
doses results in 80 % of patients experiencing trigeminal postoperative trigeminal dysfunction (4 % versus 17–45 %),
dysfunction, but the percentage of patients having an excel- and essentially no risk of other cranial nerve deficit or motor
lent result (BNI I) is also increased. Because there is no pre- weakness. In addition, patients avoid the customary compli-
cedence for performing a third GK-SRS on the same nerve cations of craniotomy and general anesthesia. The mecha-
until recently, the risk of radiation necrosis of the brain stem, nism of pain relief is not known because these benign tumors
which is presumed to be high, is unknown. At SUNY, the do not change significantly in size after GK-SRS, and the
authors have recently treated a case of right-sided trigeminal doses given are significantly smaller than typical doses for
neuralgia three times over 10 years, the last one delivered a idiopathic TN. Compensating for the lower doses may be the
year ago. This patient has a relatively long segment of tri- fact that the prescription isodose lines in the tumor cases
geminal nerve between the pons and Meckel’s cave. The are generally much closer to the brain stem indicating that
single 4-mm target isocenter for the third GK-SRS was made the brain stem at the root entry zone may receive a dose simi-
7.5 mm away from the pons. The treatment prescription was lar to patients treated by the standard 4-mm shot protocol.
70 Gy to 100 % and the maximum dose to the pons was cal- These results raise the interesting hypothesis that the effec-
culated to be 9.98 Gy, which was considered to be safe. The tiveness of GK-SRS in TN is related more to the brain-stem
patient experienced good pain relief after the third GK-SRS dose than to the nerve dose [42]. Further studies will be
without any side effects including no facial numbness at one needed to clarify this issue.
year follow-up.
The alternative to a repeat GK-SRS is either MVD or a
percutaneous procedure. Pollock et al. found that an excel- Multiple Sclerosis, Post-herpetic Neuralgia,
lent outcome was seen in 75 % of patients undergoing MVD and Atypical Facial Pain
and in 71 % undergoing glycerol rhizotomy [32]. Interestingly,
five of eight patients were observed to have a region of pre- TN is a well-recognized symptom of MS, and its manifesta-
sumed radiation injury on the segment of the superior cere- tions are indistinguishable from classic, idiopathic TN. It
bellar artery that contacted the trigeminal nerve. In four of occurs in 1–2 % of patients with MS but rarely as the pre-
these five cases, the patient either received a high dose senting symptom. Patients with MS present at an earlier age
(90 Gy) or had a repeat GK-SRS treatment. Our anecdotal and frequently have bilateral pain [42]. Pharmacological
experience with MVD after failed GK-SRS is that arachnoidal treatment of MS-TN patients is complicated by their greater
adhesions do form on the trigeminal nerve and can complicate sensitivity to gait imbalance and other side effects of medica-
mobilization of arterial loops. Good outcomes, however, are tions. Both MVD and percutaneous procedures have a lower
still possible with MVD in this setting [40]. rate of effectiveness for patients with MS, but GK-SRS has
been shown to be effective in up to 80 % of MS-TN patients
[43, 44]. The treatment protocol for these patients is
Secondary Trigeminal Neuralgia unchanged with respect to dose and target. Because MS
patients comprise a small portion of the major series pub-
Patients with secondary TN experience symptoms caused by lished to date, their sensitivity to GK-SRS compared with
the compressive effects of a tumor. The two most common typical TN patients is unknown [28, 29]. An increase in tri-
tumors are meningioma and trigeminal schwannoma, but geminal complications has not been seen [43].
other tumor types such as metastases, chordoma, and epider- The use of GK-SRS for atypical facial pain syndromes such
moids are also possible. These tumors are usually located in as post-herpetic neuralgia is controversial. Pharmacological
the cavernous sinus, Meckel’s cave, petroclival region, or therapy is frequently ineffective, and MVD and percutaneous
tentorial notch. In addition to TN, patients may present with treatments are contraindicated because they may worsen the
facial numbness, weakness, diplopia, visual loss, and hemi- pain. Only nucleus caudalis dorsal root entry zone (DREZ)
paresis. Treatment with GK-SRS in these patients is both lesions have been documented to show some benefit. Urgosik
effective at controlling tumor growth and TN symptoms. reported on 16 patients who received GK-SRS and found a
Regis et al. reported 46 patients with tumors amenable to successful result (excellent, very good, and good) in only 44 %
GK-SRS who were treated with a multiple-shot dose plan to of patients. Pain relief occurred after a median interval of 1
a mean dose of 14 Gy (range, 8–25 Gy) [41]. Forty-five month, and no radiation-related side effects have been observed
patients were followed for a median time of 55 months. in these patients [45]. The University of Maryland experience
Initially, 80 % had complete pain relief (BNI I) and 16 % with atypical facial pain syndromes is similar. In this study,
were improved. Recurrent pain was experienced by 6 approximately half of the patients presented with atypical TN
(13.3 %) patients, and slight hypesthesias were seen in 2 and half initially presented with classic TN that then pro-
(4.4 %). In comparison with published microsurgical series, gressed to atypical TN. Seventy-two percent of patients
reported pain relief with a median time to relief of 5.8 weeks 15. Lim M, Villavicencio AT, Burneikiene S, Chang SD, Romanelli P,
(range, 0–24 weeks). There was no significant difference McNeely L, McIntyre M, Thramann JJ, Adler JR. CyberKnife
radiosurgery for idiopathic trigeminal neuralgia. Neurosurg Focus.
between those that presented with atypical pain and those that 2005;18:E9.
progressed to atypical TN. Approximately 60 % of patients 16. Smith ZA, De Salles AA, Frighetto L, Goss B, Lee SP, Seich M,
were able to discontinue or decrease the use of pain medica- Wallace RE, Cabatan-Awang C, Solberg T. Dedicated linear
tion. Bothersome numbness was reported in five (27 %) accelerator radiosurgery for the treatment of trigeminal neuralgia.
J Neurosurg. 2003;99(3):511–6.
patients. Of the patients with sustained pain relief, quality of 17. Ma L, Kwok Y, Chin LS, Yu C, Regine WF. Comparative analyses
life improved an average of 82 %, and of the patients in whom of linac and Gamma Knife radiosurgery for trigeminal neuralgia
pain returned, quality of life improved an average of 63 % treatments. Phys Med Biol. 2005;50:5217–27.
[46]. The standard treatment protocols are used, and an 18. Kondziolka D, Lacomis D, Niranjan A, Mori Y, Maesawa S,
Fellows W, Lunsford LD. Histological effects of trigeminal nerve
increase in complications or worsened pain has not been noted. radiosurgery in a primate model: implications for trigeminal neural-
gia radiosurgery. Neurosurgery. 2000;46:971–6.
19. Hakanson S. Transovale trigeminal cisternography. Surg Neurol.
Conclusion 1978;10:137–44.
20. Lindquist C, Kihlstrom L, Hellstrand E. Functional neurosurgery—
a future for the Gamma-Knife. Stereotact Funct Neurosurg. 1991;
GK-SRS for TN is effective and should be considered a stan- 57:72–81.
dard alternative to conventional surgical treatments for idio- 21. Rand RW, Jacques DB, Melbye RW, Copcutt BG, Levenick
pathic TN. Initial results are comparable with other MN. Fisher MR Leksell Gamma Knife treatment of tic douloureux.
Stereotact Funct Neurosurg. 1993;61 Suppl 1:93–102.
treatments, but a significant recurrence rate is observed. It is, 22. Kondziolka D, Lunsford LD, Flickinger JC, Young RF, Vermeulen
however, an extraordinarily safe procedure particularly for S, Duma CM, Jacques DB, Rand EW, Regis J, Peragut JC, Manera
elderly patients and those wishing to avoid surgery or have L, Epstein MH, Lindquist C. Stereotactic radiosurgery for trigemi-
surgical contraindications. nal neuralgia: a multi-institutional study using the gamma unit.
J Neurosurg. 1996;84:940–5.
23. Maesawa S, Salame C, Flickinger JC, Pirris S, Kondziolka D,
Lunsford LD. Clinical outcomes after stereotactic radiosurgery for
idiopathic trigeminal neuralgia. J Neurosurg. 2001;94:14–20.
References 24. Cheuk AV, Chin LS, Petit JH, Herman JM, Fang HB, Regine
WF. Gamma Knife surgery for trigeminal neuralgia: outcome,
1. International Headache Society. International classification of imaging, and brainstem correlates. Int J Radiat Oncol Biol Phys.
headache disorders. 2nd ed. Oxford: Blackwell; 2003. 2004;60(2):537–41.
2. Wilkins R. Trigeminal neuralgia: introduction. In: Wilkins R, 25. Flickinger JC, Pollock BE, Kondziolka D, Phuong LK, Foote RL,
Rengachary SS, editors. Neurosurgery. New York: McGraw-Hill; Stafford SL, Lunsford LD. Does increased nerve length within the
1996. p. 3921–9. treatment volume improve trigeminal neuralgia radiosurgery? A
3. Pollock BE, Ecker RD. A prospective cost-effective study of tri- prospective double-blind, randomized study. Int J Radiat Oncol
geminal neuralgia surgery. Clin J Pain. 2005;21:317–22. Biol Phys. 2001;51:449–54.
4. Hughes B. Atypical trigeminal neuralgia. Br Dent J. 1950;89(11): 26. Patel S. Evaluating the influence of dose-rate on outcome with
243–9. Gamma-Knife stereotactic radiosurgery in the treatment of trigemi-
5. Dandy WE. Concerning the cause of trigeminal neuralgia. Am J nal neuralgia. Int J Radiat Oncol Biol Phys. 2005;63(S1):S429.
Surg. 1934;24:447–55. 27. Rogers CL, Shetter AG, Fiedler JA, Smith KA, Han PP, Speiser
6. Lunsford LD, Apfelbaum RI. Choice of surgical therapeutic modal- BL. Gamma knife radiosurgery for trigeminal neuralgia: the initial
ities for treatment of trigeminal neuralgia: microvascular decom- experience of The Barrow Neurological Institute. Int J Radiat Oncol
pression, percutaneous retrogasserian thermal, or glycerol Biol Phys. 2000;47:1013–9.
rhizotomy. Clin Neurosurg. 1985;32:319–33. 28. Petit JH, Herman JM, Nagda S, DiBiase SJ, Chin LS. Radiosurgical
7. Gardner WJ. Trigeminal neuralgia. Clin Neurosurg. 1968;15:1–56. treatment of trigeminal neuralgia: evaluating quality of life and
8. Loeser JD, Calvin WH, Howe JF. Pathophysiology of trigeminal treatment outcomes. Int J Radiat Oncol Biol Phys. 2003;56:
neuralgia. Clin Neurosurg. 1977;24:527–37. 1147–53.
9. Canavero S, Bonicalzi V. Drug therapy of trigeminal neuralgia. 29. Young RF, Vermeulen S, Posewitz A. Gamma Knife radiosurgery
Expert Rev Neurother. 2006;6(3):429–40. for the treatment of trigeminal neuralgia. Stereotact Funct
10. Leksell L. The stereotaxic method and radiosurgery of the brain. Neurosurg. 1998;70:192–9.
Acta Chir Scand. 1951;102:316–9. 30. Sheehan J, Pan H-C, Stroila M, Steiner L. Gamma Knife surgery
11. Leksell L. Stereotaxic radiosurgery in trigeminal neuralgia. Acta for trigeminal neuralgia: outcomes and prognostic factors.
Chir Scand. 1971;137:311–4. J Neurosurg. 2005;102:434–41.
12. Kondziolka D, Perez B, Flickinger JC, Habeck M, Lunsford 31. Brisman R. Gamma Knife surgery with a dose of 75 to 76.8 Gray
LD. Gamma Knife radiosurgery for trigeminal neuralgia. Arch for trigeminal neuralgia. J Neurosurg. 2004;100:848–54.
Neurol. 1998;55:1524–9. 32. Pollock BE, Phuong LK, Gorman DA, Foote RL, Stafford
13. Chen JCT, Girvigian M, Greathouse H, Miller M. Rahimian J SL. Stereotactic radiosurgery for idiopathic trigeminal neuralgia.
Treatment of trigeminal neuralgia with linear accelerator radiosur- J Neurosurg. 2002;97:347–53.
gery: initial results. J Neurosurg. 2004;101 Suppl 3:346–50. 33. Dhople AA, Adams JR, Maggio WW, Naqvi SA, Regine WF, Kwok
14. Richards GM, Bradley KA, Tomé WA, Bentzen SM, Resnick DK, YA. Long-term outcomes of Gamma Knife radiosurgery for classic
Mehta MP. Linear accelerator radiosurgery for trigeminal neural- trigeminal neuralgia: Implications of treatment and critical review
gia. Neurosurgery. 2005;57:1193–200. of the literature. J Neurosurg. 2009;111:351–8.
34. Pollock BE, Phuong LK, Foote RL, Staffod SL, Gorman DA. High- intra-operative findings and treatment outcomes. J Neurosurg.
dose trigeminal neuralgia radiosurgery associated with increased 2005;102(Suppl):259–61.
risk of trigeminal nerve dysfunction. Neurosurgery. 2001;49:58–62. 41. Regis J, Metellus P, Dufour H, Roche PH, Muracciole X, Pellet W,
35. Herman JM, Petit JH, Amin P, Kwok Y, Dutta PR, Chin LS. Repeat Grisoli F, Peragut JC. Long-term outcome after gamma knife sur-
gamma knife radiosurgery for refractory or recurrent trigeminal gery for secondary trigeminal neuralgia. J Neurosurg. 2001;95:
neuralgia: treatment outcomes and quality of life assessment. Int J 199–205.
Radiat Oncol Biol Phys. 2004;59(1):112–6. 42. Brisman R. Trigeminal neuralgia and multiple sclerosis. Arch
36. Hasegawa T, Kondziolka D, Spiro R, Flickinger JC, Lunsford Neurol. 1987;44:379–81.
LD. Repeat radiosurgery for refractory trigeminal neuralgia. 43. Rogers CL, Shetter AG, Ponce FA, Fiedler JA, Smith KA, Speiser
Neurosurgery. 2002;50:494–500. BL. Gamma knife radiosurgery for trigeminal neuralgia associated
37. Shetter AG, Rogers CL, Ponce F, Fiedler JA, Smith K, Speiser with multiple sclerosis. J Neurosurg. 2002;97:529–32.
BL. Gamma Knife: radiosurgery for recurrent trigeminal neuralgia. 44. Zorro O, Lobato-Polo J, Kano H, Flickinger JC, Lunsford LD,
Neurosurgery. 2002;97(5 Suppl):536–8. Kondziolka D. Gamma Knife radiosurgery for multiple sclerosis-
38. Pollock BE, Foote RL, Stafford SL, Link MJ, Gorman DA, Schomberg related trigeminal neuralgia. Neurology. 2009;73(14):1149–54.
PJ. Results of repeated gamma knife radiosurgery for medically 45. Urgosik D, Vymazal J, Vladyka V, Liscak R. Treatment of posther-
unresponsive trigeminal neuralgia. J Neurosurg. 2000;93(3):162–4. petic trigeminal neuralgia with the gamma knife. J Neurosurg.
39. Dutta P. Comparison of repeat GK-SRS for refractory or recurrent 2000;93:165–8.
trigeminal neuralgia: does dose matter. Int J Radiat Oncol Biol 46. Dhople A, Kwok Y, Chin L, Shepard D, Slawson R, Amin P, Regine
Phys. 2005;60(S1):S547. W. Efficacy and quality of life outcomes in patients with atypical
40. Shetter AG, Zabramski JM, Speiser BL. Microvascular decom- trigeminal neuralgia treated with gamma knife stereotactic radio-
pression after gamma knife surgery for trigeminal neuralgia: surgery. Int J Radiat Oncol Biol Phys. 2007;69:397–403.
Trigeminal
Neuralgia: Viewpoint—Surgery 53
Craig Rabb and Ahmed Cheema
Technique
Introduction
The procedure is carried out under sedation as patients need
Trigeminal neuralgia is a severely debilitating disease with to be awake for examination during rhizotomy. The patient is
annual incidence of 4.3 cases per 100,000 and accounting for placed in the supine position and Hartel’s technique can be
15,000 new cases diagnosed each year [1]. Its lifetime preva- used to visualize the path to foramen ovale and eventually to
lence is estimated to be around 0.3 % [2]. It is a clinical diag- Meckel’s cave. The Hartel’s technique describes three points
nosis usually characterized by unilateral sharp stabbing pain [15]. Point 1 is located inferiorly at mid pupil, Point 2 is
exacerbated by touch with pain-free intervals in between epi- located 3 cm anteriorly to external auditory meatus along
sodes [3]. It is well established that the majority of cases are zygoma, and Point 3 is located 2–3 cm lateral to oral com-
caused by some kind of compression at the root entry zone missure. The foramen ovale is located at the junction of
[4]. The pathophysiology of trigeminal neuralgia has been vertical axis (Point 1) and horizontal axis (Point 2). Point 3
extensively studied yet poorly understood. Prior to undertak- signifies the entry point of needle. Cannulation of the fora-
ing treatment, trigeminal neuralgia must be differentiated men ovale is carried under local anesthesia and confirmed
from patients with atypical facial pain [5]. Although vascular with fluoroscopy, wince reflex, contraction of masseter mus-
compression is most likely the culprit for neuralgia, other cle, and flow of cerebrospinal fluid (CSF). Additionally, cis-
compressive and demyelinating causes account for about ternography can be used to confirm the location of needle in
6–16 % of the cases and should be entertained during evalu- appropriate location. Lately, frameless stereotaxy has also
ation including tumor, aneurysm, venous compression, and been used to increase accuracy and decrease time to cannula-
multiple sclerosis [6–10]. Surgical options for trigeminal tion of foramen ovale with promising results [16–18]. Once
neuralgia can be broadly divided into either destructive or the needle is appropriately placed in Meckel’s cave, cisternal
nondestructive. This chapter reviews these techniques fol- volume is estimated and anhydrous glycerol is injected. The
lowed by a comparison with stereotactic radiosurgery (SRS). patient is kept in sitting flexed position for an hour, then
watched overnight in the hospital before being discharged.
Percutaneous Glycerol Rhizotomy

Results
Since its inception by Håkanson in 1981 and subsequent testing
and progression by Beck and Lunsford, percutaneous retrogas- Lunsford described a series of 112 patients undergoing
serian glycerol rhizotomy still is used for treatment of trigemi- percutaneous glycerol rhizotomy (PGR), with 67 % having
nal neuralgia [11–13]. It involves identification of Meckel’s complete pain relief, 23 % improved requiring minimal med-
cave and then successful injection of 0.2–0.4 mL glycerol [14]. ical therapy, and 10 % having an unfavorable outcome [19].
It is a chemically destructive method of treatment. Fujimaki et al. followed 122 patients who underwent PGR
for 38–54 months and noticed a median pain-free interval of
32 months and a recurrence rate of 72 % [20]. They also
C. Rabb, M.D. (*) • A. Cheema noticed 63 % of patients having post-procedure hypesthesias,
Department of Neurosurgery, University of Oklahoma 29 % having dysesthesias, and two patients with anesthesia
College of Medicine, 1000 N. Lincoln, Suite 400,
Oklahoma City, OK 73104, USA dolorosa. Blomstedt et al. evaluated the complications rela-
e-mail: craig-rabb@ouhsc.edu ted to the procedure. They observed persistent debilitating
660 C. Rabb and A. Cheema
complications in 28.1 % of patients including labial herpes Technique

(1.8 %), anesthesia dolorosa (0.8 %), dysesthesias (22.7 %),
hearing loss (1.9 %), and meningitis (1.5 %) [21]. When Balloon compression is carried out under general anesthesia as
compared to Gamma Knife surgery, both have good initial there is a risk of bradycardia that should be closely monitored
outcomes; however, there were more recurrences in the PGR and watched during the extent of procedure. An external pace-
group compared to Gamma Knife and a later treatment maker is placed. The foramen ovale is cannulated as described
response in the Gamma Knife group [22]. before. Once the foramen ovale accessed, a Fogarty balloon is
inflated with intraluminal pressures of 1,200–1,500 mmHg
and resultant tissue compression of 650–950 mmHg. The bal-
Percutaneous Radiofrequency Ablation loon shows characteristic pear shape. The balloon is inflated
for 1 min and then withdrawn.
Radiofrequency ablation is a thermal destructive method for
treatment of trigeminal neuralgia. Its technique and results
are described below. Results
Belber et al. reported improvement in 76 % of patients after

Technique balloon compression [24]. In elderly cohort, Mizuno et al.
reported promising results with 10/16 patients benefiting
This procedure can be carried out under local anesthesia from the procedure and others having at least 50 % improve-
with mild sedation. Access to the foramen ovale and confir- ment in their pain [25].
mation are carried out as discussed in glycerol rhizotomy
section. Once the foramen ovale is cannulated, either a
straight or curved thermal-coupled or cordotomy-type elec- Microvascular Decompression
trode can be used for selection of trigeminal nerve divi-
sions. Curved electrodes give the surgeon an opportunity to Microvascular decompression (MVD) has long stood the test
selectively ablate rootlets. The nerve roots can be located of time for treatment of trigeminal neuralgia in appropriately
using the clivus as anatomical landmark. The maxillary selected patients. It is an invasive but nondestructive method
nerve is located at the clivus, the ophthalmic division is of treatment for trigeminal neuralgia focusing mainly on
located distally, and the mandibular is located proximal to decompression of trigeminal nerve. It has been extensively
clivus. studied and is considered the gold standard for comparison
Once the desired nerve branch is selected, a preliminary with other treatments while newer methods are still being
lesion is produced using 70 °C for 70 s. The patient’s facial investigated.
sensation is assessed and sometimes facial flushing can be
noticed that is present in the selected rootlet.
Technique
Results The procedure requires the patient to be under general anes-

thesia and appropriately positioned to have CSF diverted to
Ischia et al. in his series of 124 patients noted greater than allow for maximal cerebellar retraction with minimal mechan-
90 % immediate relief with only 67 % of patients having ical retraction. The patient is positioned in the lateral decubitus
no pain at 3.7-year follow-up [23]. They also noted that position with the head secured using three-point fixation with
2–3 % of patients have hypesthesia, dysesthesia, decreased head holder, flexed, and rotated to allow the maximal access to
visual acuity, and masseter dysfunction following the the cerebellopontine angle. The ipsilateral arm is carefully
procedure. padded, and placed on an “airplane,” and is taped under mild
tension so as to move the shoulder out of the line of vision, an
access to the cerebellopontine angle (CPA). Reverse
Percutaneous Balloon Compression Trendelenburg position can be used for optimal venous flow.
The entry point can be determined by drawing a line run-
Balloon micro-compression is a mechanical destructive ning from the inion to the zygoma, which in the retromastoid
method for treatment of trigeminal neuralgia. area identifies the transverse sinus. The digastric grove is then
53 Trigeminal Neuralgia: Viewpoint—Surgery 661
the anatomical landmark for the sigmoid-transverse junction It has been shown that isolated ophthalmic division
[26]. In studies, it has been shown to be posteroinferior to the symptoms are most likely in cigarette-smoking older patient
asterion [27, 28]. Image guidance can be used to further population and are via caudal compression [32]. Maxillary
assess the junction, although we have not found this neces- division symptoms are usually found in woman and usually
sary. A curvilinear “lazy S” incision is made and carried down secondary to a venous compression [33]. Mandibular nerve
to the bone. A high-speed drill is used for making a burr hole, symptoms are usually caused by rostral compression from
to initiate exposure. Kerrison rongeurs are then used to the superior cerebellar artery [32].
remove bone in an anterior-superior direction until the stereo- The wound is copiously irrigated and any venous oozing
typic bluish appearance of the sinus is identified, and the stopped meticulously. Bone edges are waxed and the dura is
transverse sigmoid junction is partially exposed. The dura is closed watertight with 4-0 or 5-0 suture, and augmented with
open in a T-shaped manner inferior and posterior to the fibrin sealant. Cranioplasty is performed using calcium phos-
sinuses [26]. The procedure may be carried out while continu- phate cement, and scalp is closed in usual fashion. Patients
ous brain stem auditory evoked potentials are being moni- are watched in the intensive care setting overnight and usu-
tored [29]. CSF is allowed to liberally drain during exposure. ally discharged in 2–3 days.
The microscope is then brought in. Typically, the tentorial
dura above the cerebellum is used as an orientation point, and
bony and dural exposure should be continued until the under- Results
surface of the tentorium is readily visualizable. Then, one can
trace this anteriorly until the junction with the petrous bone Several studies have evaluated patients undergoing MVD
forms an unmistakable angle, which is the direct route to the years following surgery. Originally, Apfelbaum in a series of
trigeminal nerve. The arachnoid of the CPA cistern is opened 273 patients followed for 55 months showed immediate
using microdissection techniques, and CSF is further drained. relief in 94 % of patients and satisfactory pain relief in 70 %
The cerebellum then relaxes considerably. We usually employ of patients on follow-up [34]. Kolluri and Heros described
extremely gentle retraction on the upper surface of the cere- similar results with 78 % pain relief in patients undergoing
bellum until sufficient CSF has drained that it is no longer MVD with a mean follow-up of 5 years [35]. Burchiel et al.
needed. In the lateral position, once CSF is drained, gravity had a higher recurrence rate of 47 % in his study with major
becomes an excellent retractor. recurrence rate of 3.5 % annually and minor recurrences of
The eighth nerve can be encountered en route to the 1.5 % annually [36]. In these studies, it was noted that patient
trigeminal nerve, and can be confused for the fifth nerve by with definitive arterial compression had lower rate of recur-
less experienced surgeons. Maintaining awareness of this is rences when compared to venous and/or bony compression.
crucial. The trigeminal nerve, once encountered, is usually In a larger study, Barker et al. mentioned that at 6.2 years, the
unmistakable by its characteristic location at the apex of the treatment success in patients undergoing MVD was 70 %
tentorium/petrous junction and by its brilliant white appear- [37]. The recurrence rate after 10 years of follow-up was less
ance. The surgeon then must use careful microdissection than 1 % and most of the recurrences were in the initial
to mobilize the offending vessels from the nerve. The most 2-year follow-up. Similar results were also reported by Olson
common artery compressing the nerve is the superior cere- et al. in their 25-year follow-up of MVD [38]. Special atten-
bellar artery. One must be careful not to assume there is just tion should be paid to patients with multiple sclerosis and
one vessel causing symptoms. Occasionally, compression trigeminal neuralgia. This subpopulation of patient has been
can originate from petrosal veins or the anterior-inferior cer- thought to be refractory to treatment with MVD. Resnick
ebellar artery. Veins causing compression usually require et al. in their analysis showed that MVD alone is insufficient
cauterization and division [30]. Our preference for treatment for treatment, however should be considered on a case-by-case
of the arteries is to mobilize them extensively, and place a basis [39–42].
sling of Teflon around the vessel, and glue it to the tentorium McLaughlin et al. in their analysis of complication rate of
well away from the nerve using fibrin sealant. The Teflon felt MVD in a large series have observed that in experienced
requires teasing it with Adson forceps until a thin strip hands the complications associated with MVD are low.
results. Only when the artery cannot be mobilized suffi- Since 1990, after modifications of MVD technique under
ciently do we place a pad of Teflon felt between the artery Dr. Jannetta, the rate of cerebellar injuries is 0.45 %, CSF
and nerve. The whole nerve should be examined for any leak is 1.85 %, and hearing loss is 0.8 %, based upon analysis
offending loops from its origin in brain stem all the way to of 1995 patients who underwent MVD [43]. In another anal-
the Meckel’s cave [31]. ysis by Kondo et al., they witnessed increasing satisfaction
and pain-free interval in patients undergoing surgery after following Gamma Knife radiosurgery (mean 24 days, range
1990, where satisfaction rate grew from 92.9 to 96.7 % with 1–180 days) and a 33 % recurrence rate at 1-year follow-up
recurrence decreased from 10.2 to 6.5 % [44]. Risk stratifi- [54]. These results stress the need for further studies however
cation for MVD has been done to predict outcomes. also show that in patients suitable for MVD, surgery should
Theodosopoulos et al. stratified patients undergoing MVD as be offered, with Gamma Knife reserved for patients not ame-
high risk based upon age less than 53 years, female sex, men nable to surgery, or those who decline open surgery.
with left-sided trigeminal neuralgia and pain lasting greater
than 11 years and found that in high risk group 58 % of the
patients are pain free at 4-year follow-up compared to Microvascular Decompression in Elderly
80–89 % of patients in lower risk profile [45].
Jodicke et al. in their limited analysis of eight patients with
median age of 70 years showed that 6/8 patients were pain
Discussion free after MVD and propose consideration of MVD even in
patients older than 65 years of age if they are medically sta-
Microvascular Decompression Versus ble to undergo surgery [55]. Ogungbo et al. in a 42-patient
Glycerol Rhizotomy retrospective analysis found no differences in outcomes
in patients older than 65 years of age when compared to
Fergusson described a series of 45 patients who were younger cohort [56]. Similar results were also reported by
followed after PGR compared to MVD in 24 patients, which Ferroli et al. and Pollock et al. in their analysis [57–59].
showed 58 % pain free at 30 months compared to 71 %,
respectively [46]. Suwa et al. retrospectively analyzed the
pain-free interval and found 7 % recurrence in case of MVD Economical and Patient’s Perspective
at 95 months compared to 49 % in PGR group at 84 months of Treatment Options
[47]. Similar results were also seen in a retrospective study
for a 20-year follow-up that showed 50 % recurrence rate Pollock and Ecker, in their prospective cost-based analysis,
with PGR at 2-year follow-up compared to MVD with 64 % showed “The cost per quality adjusted pain-free year was
pain-free and lesser sensory deficits after 20 years [48, 49]. $6,342, 8,174, and 8,269 for glycerol rhizotomy, microvas-
cular decompression, and SRS, respectively” [60]. This chapter
emphasizes that in older patients percutaneous methods may be
Microvascular Decompression Versus more cost-effective and in younger population MVD remains a
Radiofrequency Ablation better surgical option when considered from an economic
standpoint of view. Zakrzewska et al. based upon questionnaire
Apfelbaum originally described comparison of percutaneous showed that 89 % of the patients undergoing MVD were satis-
radiofrequency ablation (PRFA) with MVD and described a fied with the surgical outcome compared to 72 % with percuta-
series of 103 patients followed for 36 months, with results neous rhizotomy [61]. The MVD group was also less dissatisfied
showing 88 % relief in PRFA group versus 96 % in MVD at 4 % compared to 20 % in percutaneous rhizotomy group.
with recurrence rates of 13 % and 5 %, respectively [50].
Conclusion
Microvascular Decompression Versus
Gamma Knife There are multiple invasive and noninvasive treatments
available for trigeminal neuralgia. While percutaneous pro-
Kondziolka et al. in their analysis for SRS have shown that cedures offer a less invasive modality for treatment, their
96 % of the patients initially treated with Gamma Knife have long-term recurrence rates are higher when compared to the
good to excellent result at 1-year follow-up [51]. In patients MVD. Even in patients refractory to initial MVD, recur-
previously treated with MVD, SRS proved to be beneficial to rences have been successfully treated with re-exploration
85.6 % of patients, further elucidating its important role in [62]. These reoperations have shown arterial (22 %), venous
patient population with recurrent and refractory trigeminal (13 %), and Teflon compression (13 %) in these patients.
neuralgia and as an adjuvant therapy [52]. Pollock et al. Percutaneous methods show promise in selected patient
showed in their analysis that when compared with SRS, populations who cannot undergo surgery due to comorbidi-
MVD offers more prolonged pain-free interval at 3-year ties. SRS offers another alternative to MVD as it can be
follow-up (72 % versus 59 %) [53]. Sheehan et al., in another performed in patients with significant medical comorbidi-
3-year follow-up study, confirmed the delayed onset of relief ties. It has a delayed onset yet usually gives good pain relief.
53 Trigeminal Neuralgia: Viewpoint—Surgery 663
treatment of trigeminal neuralgia. Neurosurgery. 2006;59(4 Suppl

References 2):ONS394–401; discussion ONS2; PubMed PMID: 17041509.
19. Lunsford LD, Bennett MH. Percutaneous retrogasserian glycerol
1. Katusic S, Beard CM, Bergstralh E, Kurland LT. Incidence and rhizotomy for tic douloureux: Part 1. Technique and results in 112
clinical features of trigeminal neuralgia, Rochester, Minnesota, patients. Neurosurgery. 1984;14(4):424–30.
1945-1984. Ann Neurol. 1990;27(1):89–95; PubMed PMID: 20. Fujimaki T, Fukushima T, Miyazaki S. Percutaneous retrogasserian
2301931. glycerol injection in the management of trigeminal neuralgia: long-
2. Mueller D, Obermann M, Yoon MS, Poitz F, Hansen N, Slomke term follow-up results. J Neurosurg. 1990;73(2):212–6; PubMed
MA, et al. Prevalence of trigeminal neuralgia and persistent PMID: 2195139.
idiopathic facial pain: a population-based study. Cephalalgia : an 21. Blomstedt PC, Bergenheim AT. Technical difficulties and periop-
international journal of headache. 2011;31(15):1542–8; PubMed erative complications of retrogasserian glycerol rhizotomy for
PMID: 21960648. trigeminal neuralgia. Stereotact Funct Neurosurg. 2002;79(3–4):
3. Brisman R. Trigeminal neuralgia: diagnosis and treatment. World 168–81; PubMed PMID: 12890975.
Neurosurg. 2011;76(6):533–4; PubMed PMID: 22251500. 22. Henson CF, Goldman HW, Rosenwasser RH, Downes MB, Bednarz
4. Jannetta PJ. Observations on the etiology of trigeminal neuralgia, G, Pequignot EC, et al. Glycerol rhizotomy versus gamma knife
hemifacial spasm, acoustic nerve dysfunction and glossopharyn- radiosurgery for the treatment of trigeminal neuralgia: an analysis
geal neuralgia. Definitive microsurgical treatment and results in of patients treated at one institution. Int J Radiat Oncol Biol Phys.
117 patients. Neurochirurgia. 1977;20(5):145–54. 2005;63(1):82–90; PubMed PMID: 16111575.
5. Eller JL, Raslan AM, Burchiel KJ. Trigeminal neuralgia: definition 23. Ischia S, Luzzani A, Polati E, Ischia A. Percutaneous controlled
and classification. Neurosurgical Focus. 2005;18(5):E3. thermocoagulation in the treatment of trigeminal neuralgia. Clin J
6. Balasundram S, Cotrufo S, Liew C. Case series: non vascular Pain. 1990;6(2):96–104; PubMed PMID: 2135010.
considerations in trigeminal neuralgia. Clin Oral Investig. 2012; 24. Belber CJ, Rak RA. Balloon compression rhizolysis in the surgical
16(1):63–8; PubMed PMID: 21210165. management of trigeminal neuralgia. Neurosurgery. 1987;20(6):
7. Baldwin NG, Sahni KS, Jensen ME, Pieper DR, Anderson RL, Young 908–13; PubMed PMID: 3614571.
HF. Association of vascular compression in trigeminal neuralgia 25. Mizuno M, Saito K, Takayasu M, Yoshida J. Percutaneous micro-
versus other facial-pain syndromes by magnetic-resonance-imaging. compression of the trigeminal ganglion for elderly patients with
Surgical Neurology. 1991;36(6):447–52; PubMed PMID: WOS: trigeminal neuralgia and patients with atypical trigeminal neural-
A1991GP01000006; English. gia. Neurol Med Chir. 2000;40(7):347–50; PubMed PMID:
8. Yen J. [Vascular relationships of the trigeminal root with trigeminal 10927900; discussion 50–1.
neuralgia]. Zhonghua er bi yan hou ke za zhi. 1992;27(3):164–5, 26. Jannetta PJ, McLaughlin MR, Casey KF. Technique of microvascu-
91; PubMed PMID: 1419194. lar decompression. Technical note Neurosurgical focus. 2005;18(5):
9. Zeng Q, Zhou Q, Liu Z, Li C, Ni S, Xue F. Preoperative detection E5; PubMed PMID: 15913281.
of the neurovascular relationship in trigeminal neuralgia using 27. Ucerler H, Govsa F. Asterion as a surgical landmark for lateral
three-dimensional fast imaging employing steady-state acquisition cranial base approaches. Journal of cranio-maxillo-facial surgery :
(FIESTA) and magnetic resonance angiography (MRA). Journal of official publication of the European Association for Cranio-Maxillo-
clinical neuroscience : official journal of the Neurosurgical Society Facial Surgery. 2006;34(7):415–20; PubMed PMID: 16963269.
of Australasia. 2013;20(1):107–11; PubMed PMID: 23098388. 28. Martinez F, Laxague A, Vida L, Prinzo H, Sgarbi N, Soria VR, et al.
10. Zhou Q, Liu ZL, Qu CC, Ni SL, Xue F, Zeng QS. Preoperative Topographic anatomy of the asterion. Neurocirugia. 2005;16(5):
demonstration of neurovascular relationship in trigeminal neuralgia 441–6; PubMed PMID: 16276452, Anatomia topografica del
by using 3D FIESTA sequence. Magn Reson Imaging. 2012;30(5): asterion.
666–71; PubMed PMID: 22405984. 29. Wilkins RH, Radtke RA, Erwin CW. Value of intraoperative brain-
11. Hakanson S. Trigeminal neuralgia treated by the injection of glyc- stem auditory evoked potential monitoring in reducing the auditory
erol into the trigeminal cistern. Neurosurgery. 1981;9(6):638–46; morbidity associated with microvascular decompression of cranial
PubMed PMID: 7322329. nerves. Skull base surgery. 1991;1(2):106–9; PubMed PMID:
12. Lunsford LD. Percutaneous glycerol treatment of trigeminal neural- 17170830; PubMed Central PMCID: 1656316.
gia. Neurosurgery. 1982;10(2):299–300; PubMed PMID: 7070629. 30. Lee SH, Levy EI, Scarrow AM, Kassam A, Jannetta PJ. Recurrent
13. Beck DW, Olson JJ, Urig EJ. Percutaneous retrogasserian glycerol trigeminal neuralgia attributable to veins after microvascular
rhizotomy for treatment of trigeminal neuralgia. J Neurosurg. decompression. Neurosurgery. 2000;46(2):356–61; PubMed PMID:
1986;65(1):28–31; PubMed PMID: 3712026. 10690724; discussion 61–2.
14. Lunsford LD. Identification of Meckel cave during percutaneous 31. Broggi G, Broggi M, Ferroli P, Franzini A. Surgical technique for
glycerol rhizotomy for tic douloureux. AJNR American journal of trigeminal microvascular decompression. Acta Neurochir. 2012;
neuroradiology. 1982;3(6):680–2; PubMed PMID: 6816048. 154(6):1089–95; PubMed PMID: 22531963.
15. Taha J. Trigeminal Neuralgia: Percutaneous Procedures. Seminars 32. Goya T, Wakisaka S, Kinoshita K. Microvascular decompression
in Neurosurgery. 2004;15(2/3):115–34. for trigeminal neuralgia with special reference to delayed recur-
16. Lin MH, Lee MH, Wang TC, Cheng YK, Su CH, Chang CM, et al. rence. Neurol Med Chir. 1990;30(7):462–7; PubMed PMID:
Foramen ovale cannulation guided by intra-operative computed 1701856.
tomography with integrated neuronavigation for the treatment of 33. Sekula RF, Frederickson AM, Jannetta PJ, Bhatia S, Quigley MR,
trigeminal neuralgia. Acta Neurochir. 2011;153(8):1593–9; PubMed Abdel Aziz KM. Microvascular decompression in patients with
PMID: 21503836. isolated maxillary division trigeminal neuralgia, with particular
17. Mandat T, Brozyna B, Krzymanski G, Podgorski JK. An image- attention to venous pathology. Neurosurg Focus. 2009;27(5):E10;
guided, noninvasive method of cannulation of the foramen ovale for PubMed PMID: 19877788.
awake, percutaneous radiofrequency rhizotomy. J Neurosurg. 2009; 34. Apfelbaum RI. Surgery for tic douloureux. Clin Neurosurg.
111(6):1223–5; PubMed PMID: 19374497. 1983;31:351–68; PubMed PMID: 6333952.
18. Bale RJ, Laimer I, Martin A, Schlager A, Mayr C, Rieger M, et al. 35. Kolluri S, Heros RC. Microvascular decompression for trigeminal
Frameless stereotactic cannulation of the foramen ovale for ablative neuralgia. A five-year follow-up study Surgical neurology. 1984;
22(3):235–40; PubMed PMID: 6463831.
36. Burchiel KJ, Clarke H, Haglund M, Loeser JD. Long-term efficacy 49. Pollock BE, Phuong LK, Gorman DA, Foote RL, Stafford
of microvascular decompression in trigeminal neuralgia. J Neuro- SL. Stereotactic radiosurgery for idiopathic trigeminal neuralgia.
surg. 1988;69(1):35–8; PubMed PMID: 2454303. J Neurosurg. 2002;97(2):347–53; PubMed PMID: 12186463.
37. Barker 2nd FG, Jannetta PJ, Bissonette DJ, Larkins MV, Jho HD. The 50. Apfelbaum RI. A comparison of percutaneous radiofrequency
long-term outcome of microvascular decompression for trigeminal trigeminal neurolysis and microvascular decompression of the tri-
neuralgia. N Engl J Med. 1996;334(17):1077–83; PubMed PMID: geminal nerve for the treatment of tic douloureux. Neurosurgery.
8598865. 1977;1(1):16–21; PubMed PMID: 615946.
38. Olson S, Atkinson L, Weidmann M. Microvascular decompression 51. Kondziolka D, Lunsford LD, Flickinger JC. Gamma knife radiosur-
for trigeminal neuralgia: recurrences and complications. Journal of gery as the first surgery for trigeminal neuralgia. Stereotact Funct
clinical neuroscience : official journal of the Neurosurgical Society Neurosurg. 1998;70 Suppl 1:187–91; PubMed PMID: 9782250.
of Australasia. 2005;12(7):787–9; PubMed PMID: 16165362. 52. Kao MC. Gamma knife surgery for trigeminal neuralgia. J Neurosurg.
39. Resnick DK, Jannetta PJ, Lunsford LD, Bissonette DJ. 2002;96(1):160–1; PubMed PMID: 11794600.
Microvascular decompression for trigeminal neuralgia in patients 53. Pollock BE. Comparison of posterior fossa exploration and stereo-
with multiple sclerosis. Surg Neurol. 1996;46(4):358–61; PubMed tactic radiosurgery in patients with previously nonsurgically treated
PMID: 8876717; discussion 61–2. idiopathic trigeminal neuralgia. Neurosurg Focus. 2005;18(5):E6;
40. Broggi G, Ferroli P, Franzini A, Pluderi M, La Mantia L, Milanese PubMed PMID: 15913282.
C. Role of microvascular decompression in trigeminal neuralgia 54. Sheehan J, Pan HC, Stroila M, Steiner L. Gamma knife surgery for
and multiple sclerosis. Lancet. 1999;354(9193):1878–9; PubMed trigeminal neuralgia: outcomes and prognostic factors. J Neurosurg.
PMID: 10584731. 2005;102(3):434–41; PubMed PMID: 15796376.
41. Broggi G, Ferroli P, Franzini A, Servello D, Dones I. Microvascular 55. Jodicke A, Winking M, Deinsberger W, Boker DK. Microvascular
decompression for trigeminal neuralgia: comments on a series of decompression as treatment of trigeminal neuralgia in the elderly
250 cases, including 10 patients with multiple sclerosis. J Neurol patient. Minimally invasive neurosurgery : MIN. 1999;42(2):92–6;
Neurosurg Psychiatry. 2000;68(1):59–64; PubMed PMID: PubMed PMID: 10422706.
10601403; PubMed Central PMCID: 1760596. 56. Ogungbo BI, Kelly P, Kane PJ, Nath FP. Microvascular decompres-
42. Eldridge PR, Sinha AK, Javadpour M, Littlechild P, Varma TR. sion for trigeminal neuralgia: report of outcome in patients over 65
Microvascular decompression for trigeminal neuralgia in patients years of age. Br J Neurosurg. 2000;14(1):23–7; PubMed PMID:
with multiple sclerosis. Stereotact Funct Neurosurg. 2003; 81(1–4): 10884880.
57–64; PubMed PMID: 14742965. 57. Ferroli P, Acerbi F, Tomei M, Tringali G, Franzini A, Broggi
43. McLaughlin MR, Jannetta PJ, Clyde BL, Subach BR, Comey CH, G. Advanced age as a contraindication to microvascular decom-
Resnick DK. Microvascular decompression of cranial nerves: lessons pression for drug-resistant trigeminal neuralgia: evidence of preju-
learned after 4400 operations. J Neurosurg. 1999;90(1):1–8; PubMed dice? Neurological sciences : official journal of the Italian
PMID: 10413149. Neurological Society and of the Italian Society of Clinical
44. Kondo A. Microvascular decompression surgery for trigeminal Neurophysiology. 2010;31(1):23–8; PubMed PMID: 19806315.
neuralgia. Stereotact Funct Neurosurg. 2001;77(1–4):187–9; 58. Pollock BE, Stien KJ. Posterior fossa exploration for trigeminal
PubMed PMID: 12378075. neuralgia patients older than 70 years of age. Neurosurgery.
45. Theodosopoulos PV, Marco E, Applebury C, Lamborn KR, Wilson 2011;69(6):1255–9; PubMed PMID: 21734615; discussion 9–60.
CB. Predictive model for pain recurrence after posterior fossa sur- 59. Sekula Jr RF, Frederickson AM, Jannetta PJ, Quigley MR, Aziz
gery for trigeminal neuralgia. Arch Neurol. 2002;59(8):1297–302; KM, Arnone GD. Microvascular decompression for elderly patients
PubMed PMID: 12164727. with trigeminal neuralgia: a prospective study and systematic review
46. Ferguson GG, Brett DC, Peerless SJ, Barr HW, Girvin JP. Trigeminal with meta-analysis. J Neurosurg. 2011;114(1):172–9; PubMed
neuralgia: a comparison of the results of percutaneous rhizotomy PMID: 20653393.
and microvascular decompression. The Canadian journal of neuro- 60. Pollock BE, Ecker RD. A prospective cost-effectiveness study of
logical sciences Le journal canadien des sciences neurologiques. trigeminal neuralgia surgery. The Clinical journal of pain.
1981;8(3):207–14; PubMed PMID: 6169418. 2005;21(4):317–22; PubMed PMID: 15951649.
47. Suwa H, Hanakita J, Mizuno M, Namura S, Ohtsuka T, Asahi M, 61. Zakrzewska JM, Lopez BC, Kim SE, Coakham HB. Patient
et al. [Long-term follow up of trigeminal neuralgic patients treated reports of satisfaction after microvascular decompression and
by retrogasserian rhizotomy or by microvascular decompression]. partial sensory rhizotomy for trigeminal neuralgia. Neurosurgery.
No shinkei geka. Neurological surgery. 1993;21(1):29–36. 2005;56(6):1304–11; PubMed PMID: 15918947; discussion
48. Tronnier VM, Rasche D, Hamer J, Kienle AL, Kunze S. Treatment 11–2.
of idiopathic trigeminal neuralgia: comparison of long-term out- 62. Cho DY, Chang CG, Wang YC, Wang FH, Shen CC, Yang
come after radiofrequency rhizotomy and microvascular decompres- DY. Repeat operations in failed microvascular decompression for
sion. Neurosurgery. 2001;48(6):1261–7; PubMed PMID: 11383728; trigeminal neuralgia. Neurosurgery. 1994;35(4):665–9; PubMed
discussion 7–8. PMID: 7808609; discussion 9–70.
Trigeminal Neuralgia:
Viewpoint—Medical Management 54
Neil C. Porter
Introduction Pathophysiology
Trigeminal neuralgia is a well-recognized neuropathic pain The pathophysiology and pathogenesis of trigeminal neural-
syndrome characterized by brief paroxysms of facial pain in gia are poorly understood. A number of theories abound,
the distribution of the trigeminal nerve. The syndrome can however [2, 3]. Given that the condition occurs in multiple
occur secondary to a number of medical conditions includ- sclerosis, some investigators have argued that demyelination
ing multiple sclerosis, basilar artery aneurysm, and brain at the entry zone of the nerve root is a common feature to
tumor [1]. The disorder can also be idiopathic, the latter both primary and secondary trigeminal neuralgia [4]. Others
instance being known as Tic Delouroux. Trigeminal neural- have focused on structural concerns with the “vascular com-
gia can strike persons of any age but typically affects older pression theory” [5]. In this latter circumstance, the pain is
individuals. Treatment options include a number of neuro- believed to be the consequence of compression on the tri-
pathic pain medications as well as neurosurgical and radio- geminal nerve by some vascular anomaly. In any event, the
surgical interventions. The mainstay of treatment for most major focus in treating the patient is the elimination of the
individuals, however, remains medication. subjective pain.
Clinical Features Neuropathic Pain
Trigeminal neuralgia is characterized by recurrent facial pain Trigeminal neuralgia is a condition characterized by neuro-
in the distribution of the trigeminal nerve. The trigeminal pathic pain. Other notable conditions that also involve neuro-
nerve is divided into three branches, V1, V2, and V3, provid- pathic pain include painful diabetic neuropathy, post-herpetic
ing sensory innervation to the forehead, cheek, and chin, neuralgia, and complex regional pain syndrome. In contrast
respectively. The pain of trigeminal neuralgia most fre- to nociceptive pain, whereby the nervous system detects
quently occurs in the distribution of V2 or V3. The pain is some unpleasant stimulus alerting the body to potential dam-
usually lancinating in nature, occurring as fleeting shocks. age, in neuropathic pain the nervous system generates the
This pain can be extremely intense, producing a great deal of impulses that are appreciated as pain by the patient.
distress in sufferers. The pain is typically triggered by chew- Neuropathic pain can be constant or intermittent in nature. It
ing, ingesting cold foods or liquids, cold winds, or simple may be characterized as burning, aching, stabbing, or shoot-
tactile stimulation on the face. ing. The pain may occur spontaneously or, as with trigeminal
neuralgia, be elicited by tactile stimulation.
Neuropathic Pain Agents: Overview
Medications used to treat neuropathic pain can be conve-

N.C. Porter, M.D. (*)
niently organized into a number of classes including antiepi-
Department of Neurology, University of Maryland School
of Medicine, 110 S. Paca Street, Baltimore, MD 21201, USA leptic drugs, GABA-related medications, antidepressants, and
e-mail: Nport001@umaryland.edu local anesthetics or topical agents. Although not traditionally
666 N.C. Porter
viewed as neuropathic pain agents, narcotics are also effective Carbamazepine is associated with a dose-related leukopenia
in treating neuropathic pain. Particular benefits and draw- but in rare circumstances has been linked to irreversible
backs are associated with each class of agent as well as spe- aplastic anemia (risk at 1:40,000). Additional side effects
cific agents within each class. Most of the agents are well include dizziness and hyponatremia. Drug interactions are
tolerated, but cognitive side effects may limit the utility of also problematic given that Carbamazepine affects the levels
many of the antiepileptic drugs and antidepressants in certain of other medications that are metabolized by the P450
cases. For trigeminal neuralgia, the antiepileptic medications system.
dominate the treatment strategy followed by the GABA ago- In summary, Carbamazepine is the mainstay of medical
nists. Antidepressants have little role while narcotics provide treatment for trigeminal neuralgia, with proven efficacy in
only adjunctive therapy. An excellent review of the literature randomized clinical trials. Efficacy initially ranges from 70
regarding pharmacologic treatment options is contained to 90%, but response rates may diminish over time to 50%.
within a Practice Parameter from the American Academy of The medication must be started cautiously and titrated
Neurology in 2008 on issues surrounding the diagnosis and relatively slowly.
treatment of trigeminal neuralgia [6].
Oxcarbazepine
Antiepileptic Drugs
Oxcarbazepine is a newer antiepileptic drug structurally
Carbamazepine related to Carbamazepine. Like its structural analog,
Oxcarbazepine is a sodium channel blocker. Because of
Carbamazepine is clearly the drug of choice for trigeminal structural modifications, however, this newer drug has a
neuralgia. Known mostly for treatment of partial and sec- number of theoretical and practical advantages over
ondarily generalized seizures, Carbamazepine is also a well- Carbamazepine. Oxcarbazepine has much less effect on the
known neuropathic pain agent. Structurally similar to the P450 system and therefore is associated with far fewer drug
tricyclic antidepressants, Carbamazepine’s effectiveness for interactions. Furthermore, Oxcarbazepine cannot be metabo-
seizures and neuropathic pain is believed to occur through its lized to an epoxide derivative that is responsible for some of
action on sodium channels. Carbamazepine acts to block the side effects seen with Carbamazepine. Obviously, though,
sodium channel function in a rate-related fashion. This given its lesser track record, Oxcarbazepine is less well
allows the medication to affect dysfunctioning neurons established than Carbamazepine as a treatment for trigemi-
greater than those firing normally. nal neuralgia.
Carbamazepine was first reported to be effective in treat- Although Oxcarbazepine has not been proven effective
ing trigeminal neuralgia in 1962 [7]. This agent was proven for trigeminal neuralgia in a randomized, placebo-controlled
efficacious in a number of randomized placebo-controlled trial, it has been shown to have similar efficacy as
clinical trials [8–10]. In each trial initial response rates were Carbamazepine in two randomized trials. Response rates in
as high as 70–90 %. Long-term efficacy, however, dimin- these trials approached 90% [11, 12]. Furthermore, the drug
ished to approximately 50 %. appeared to be better tolerated than Carbamazepine.
Carbamazepine has a relatively short half-life, necessitat- Oxcarbazepine is typically started at 150–300 mg BID,
ing a dosing schedule of 3–4 times per day. Because it auto- with the lower dosing associated with less side effects.
induces its own metabolism, a patient cannot be loaded with Typical target dosages range from 300 to 600 mg BID with
Carbamazepine. Instead, Carbamazepine must be started at a some refractory patients requiring up to 1,200 mg BID.
relatively low dose and then slowly titrated upwards. The Oxcarbazepine appears to possess several advantages
patient should be started at 100 mg BID or TID. The dose over Carbamazepine, including better tolerability and fewer
can be increased every 3–5 days by 200 mg total up to drug interactions. One disadvantage, however, is cost.
200–400 mg TID. Some refractory patients may require Being a newer agent, Oxcarbazepine is significantly more
doses as high as 2,400 mg/day (800 mg TID). expensive than the older Carbamazepine. Additionally,
Carbamazepine’s major advantage is its efficacy in most Oxcarbazepine has been associated with a number of side
patients. The medication is relatively well tolerated with effects similar to other sodium channel blockers, including
only minor side effects in the majority of patients. dizziness, headache, gait instability, nausea, and vomiting.
Additionally, Carbamazepine is relatively inexpensive, being In summary, Oxcarbazepine holds great promise to poten-
available in generic form. tially replace Carbamazepine as the first-line treatment for
Unfortunately, Carbamazepine is not without its shortcom- trigeminal neuralgia. Although both medications have simi-
ings. Given its relatively short half-life, the medication requires lar efficacy, Oxcarbazepine appears to be safer and better
relatively frequent dosing that may affect compliance. tolerated.
54 Trigeminal Neuralgia: Viewpoint—Medical Management 667
Gabapentin [15, 16]. In one randomized, double-blind, placebo-

controlled, crossover study, Lamotrigine again performed
Gabapentin has assumed a major role as an antiepileptic drug well, being superior to placebo [17].
used for neuropathic pain. A number of trials have proven its Lamotrigine must be started at a low dose and increased
efficacy in treating painful diabetic neuropathy and post- slowly to minimize the risk of a serious rash. The medication
herpetic neuralgia. Despite this, no clinical trials have proven is typically started at 25 mg qd and increased by 25 mg every
Gabapentin’s role in trigeminal neuralgia. Nonetheless, 3–7 days up to 200 mg qd. The major side effect with
Gabapentin has been advocated for trigeminal neuralgia as Lamotrigine is the aforementioned rash; otherwise, the med-
well as other disorders characterized by neuropathic pain. ication is safe and well tolerated.
Gabapentin was developed as an antiepileptic drug with In summary Lamotrigine is another promising agent
structural similarity to GABA. The medication, however, whose use in trigeminal neuralgia should increase over time.
appears to have no action at the GABA receptor. Instead, The medication is safe when used correctly, apparently
Gabapentin is believed to exert its effect through the α[alpha]- effective and well tolerated.
2-δ[delta] calcium channel.
Gabapentin has been proven efficacious in randomized
clinical trials for neuropathic pain syndromes such as painful Topiramate
diabetic neuropathy and post-herpetic neuralgia but not spe-
cifically for trigeminal neuralgia. The drug has been reported, Topiramate is a novel antiepileptic agent that is believed to
however, to provide adequate pain relief for trigeminal neu- work by blocking sodium channels, blocking certain gluta-
ralgia in small, uncontrolled series [13, 14]. mate receptors, and enhancing activity of GABA via a non-
Like many of the other neuropathic pain agents, Gabapentin benzodiazepine site of the GABA receptor. Although it has
is best tolerated when started at a very low dose. Patients may never been studied in randomized trials, a number of small
be started at 300 mg daily with their dose being increased by series have demonstrated its efficacy in trigeminal neuralgia
300 mg every 3 days up to 600 mg TID. Studies have failed [18, 19]. Topiramate is typically started at 25 mg b.i.d. and
to show superiority of higher doses of Gabapentin, but some increased by 50 mg at weekly intervals up to 200 mg BID.
individuals may require doses as high as 900–1,200 mg TID. On the positive side, Topiramate is safe and reasonably
The most attractive aspect of Gabapentin is its apparent well tolerated, but it does have some notable drawbacks.
safety. Gabapentin is extremely well tolerated with no appre- As a new drug, it is relatively expensive. Additionally,
ciable serious medical side effects, and negligible drug inter- Topiramate appears to cause mental clouding, including
actions. Furthermore, overdosage has not been seen, given word-finding difficulty. Lastly, it is associated with kidney
the lack of dose-limiting side effects. stone formation.
Given the large dosages required by patients for adequate To summarize, Topiramate may be a reasonable third-line
pain relief, Gabapentin can be quite costly even as a generic. agent that is safe and reasonably well tolerated. Future studies
Furthermore, problematic side effects include dizziness, are needed to specifically justify its use in trigeminal neuralgia.
lower extremity edema, weight gain, and headaches. Lastly,
a withdrawal syndrome has been reported with rapid cessa-
tion of the medication. Pregabalin
In short, Gabapentin is a very popular drug for neuro-
pathic pain because of its favorable side effect profile. Pregabalin is a compound designed to be similar to
Evidence to date supports its use in trigeminal neuralgia fol- Gabapentin. It has been proven efficacious in treating painful
lowing trials of the better-established medications. diabetic neuropathy but is not specifically approved for tri-
geminal neuralgia. The mechanism of action is believed to
be similar to that of Gabapentin, namely via action at the
Lamotrigine α[alpha]-2-δ[delta] calcium channel. Because of its more
recent development, experience with Pregabalin is more lim-
Lamotrigine is a newer antiepileptic drug that acts through a ited than with many of the other agents. Although no specific
novel mechanism. It is believed to exert its effect through the studies using Pregabalin for trigeminal neuralgia have been
stabilization of the inactive form of the sodium channel. performed, the medication’s affect on neuropathic pain in
Although Lamotrigine has a relatively short track record general makes it a reasonable consideration once the more
with regard to neuropathic pain, this agent has already been established agents have been exhausted.
shown to be efficacious in treating patients with trigeminal The target dose of Pregabalin is 300 mg daily divided
neuralgia. In a number of uncontrolled series, Lamotrigine BID. The medication is typically started at 75 mg BID and
conveyed adequate pain relief to a majority of patients increased up to 150 mg BID after approximately 1 week.
668 N.C. Porter
Pregabalin appears to be safe and well tolerated, similar Given the advent of newer agents, however, the role of
to Gabapentin. Its greater potency, however, allows for lower Baclofen for trigeminal neuralgia has become less certain.
dosages and therefore lower pill counts than its predecessor. Baclofen was first reported as a treatment for trigeminal
Furthermore, cost is similar to that of Gabapentin. Pregabalin neuralgia in the 1980s [23]. Similar to a number of other
is associated with similar side effects as Gabapentin. agents used to treat trigeminal neuralgia, Baclofen has never
Additionally, though, thrombocytopenia and arthralgias have been proven efficacious in a randomized clinical trial. The
been reported. drug was shown to be effective in a double-blind crossover
In summary, Pregabalin may be an important medication trial, though [24].
in the future for trigeminal neuralgia. Further studies may Baclofen is typically dosed in a TID fashion, starting at
help support its use for this indication. 5–10 mg TID and increasing moderately slowly up to 20 mg
TID. Baclofen is safe and well tolerated, with dizziness,
drowsiness, and gait instability being the predominate side
Phenytoin effects.
In summary, Baclofen is a safe and reasonably effective
Although reported as the first antiepileptic drugs effective in medication for trigeminal neuralgia. Because of its safety
trigeminal neuralgia, Phenytoin’s role has greatly diminished profile and long-term track record, Baclofen should be con-
in recent years. Like Carbamazepine, Phenytoin is a sodium sidered early for the treatment of trigeminal neuralgia.
channel blocker that works in a rate-related fashion. Phenytoin
was first reported effective in treating trigeminal neuralgia in
1942 [20]. Although a number of reports have been published Clonazepam
since that time [21, 22], no clinical trials demonstrating effi-
cacy of Phenytoin have surfaced. Nonetheless, Phenytoin Clonazepam is a long-acting benzodiazepine that has been
may be a reasonable choice for selected patients with trigem- reported efficacious at high doses in some patients with tri-
inal neuralgia. geminal neuralgia [25]. Like other benzodiazepines,
Phenytoin has a relatively long half-life, allowing for Clonazepam acts by binding to GABA receptors and poten-
daily or twice daily dosing. Although BID dosing is optimal, tiating the effects of that neurotransmitter. In past years
daily dosing may improve compliance. Dosing of the medi- Clonazepam was used as a third-line agent. With the avail-
cation in a TID fashion is not necessary even though it is ability of more effective neuropathic pain agents in recent
commonplace. Phenytoin can be loaded, but this may not be years, however, its use has become negligible.
necessary in patients with trigeminal neuralgia. Typical dos- In contrast to dosing for epilepsy and mood disorders,
ing schedules consist of 100–200 mg BID with daily doses relatively high doses of Clonazepam are required to treat tri-
around 6 mg/kg per day. geminal neuralgia. The major advantage of Clonazepam is
Phenytoin’s primary strong suit is its low cost. Its major dis- its modest cost and long-term safety. Although Clonazepam
advantages are its modest efficacy and its accompanying cos- is well tolerated at lower doses, the high doses required to
metic side effects. Phenytoin is believed to cause “coarsening” treat trigeminal neuralgia are associated with more profound
of facial features, gum hypertrophy and hirsutism in women. side effects such as sedation and cognitive impairment.
Furthermore, like other older antiepileptic agents, long-term Furthermore, dependence must be a consideration, although
use of Phenytoin has been associated with osteoporosis. this concern is usually less prominent given the long half-life
In summary, Phenytoin may be a reasonable option in of the drug.
selected patients in whom multiple other medications have In summary, Clonazepam like Phenytoin may have a role
failed or who cannot afford the newer agents and who do not in extremely selected patients with trigeminal neuralgia averse
wish to undergo invasive procedures. Cosmetic side effects to invasive procedures and refractory to the numerous neuro-
and modest efficacy markedly limit its role in trigeminal pathic agents that are generally considered more effective.
neuralgia.
Local Anesthetics
GABA-Related Agents
Capsaicin
Baclofen
Although it has not been widely used in trigeminal neuralgia,
Baclofen, an agent used predominately to diminish spasticity Capsaicin has been shown to be efficacious in some uncon-
in patients with disturbances of the central nervous system, trolled series. Capsaicin is a compound derived from chili
has historically been used as a second-line agent behind peppers. The commercially available formulation consists of
Carbamazepine in treating patients with trigeminal neuralgia. a cream containing the active substance as well as Lidocaine
54 Trigeminal Neuralgia: Viewpoint—Medical Management 669
to buffer the burning caused by the medication. Capsaicin many oral and parenteral products. Fentanyl patches are
acts by depleting substance P, a neurotransmitter involved in applied to the skin, and changed every 3 days. The patches
pain pathways, from nerve terminals. come in varying strengths. For those individuals naive to
At least two uncontrolled series have been published doc- narcotics, the starting dosage should be 25 μg/h, but the dose
umenting the apparent efficacy of Capsaicin in trigeminal can be increased as tolerated up to 100 μg/h. As with other
neuralgia. Adequate pain control was obtained in approxi- narcotics, the major initial side effects include sedation and
mately 30% [26] of patients in one study and 60% in another nausea. The most concerning long-term side effect remains
[27]. Pain relief was sustained despite only transient use of dependence.
the medication.
Capsaicin should be applied to the affected area 2–4 times
daily. Duration of use should be tailored to the patient. Summary
Caution must be used to avoid exposure to the eye.
Capsaicin is extremely safe with no systemic side effects. In conclusion, trigeminal neuralgia is a serious medical con-
Additionally, cost is highly reasonable. The major drawback dition that can be debilitating when individuals are severely
to Capsaicin, however, is tolerability. Prior to causing anal- affected. A number of medical treatment options exist for the
gesia the agent may cause severe, intolerable burning in the disorder, including various antiepileptic drugs, GABA-
areas to which the medication is applied. mimetic agents, topical anesthetics, and even narcotics.
In summary, Capsaicin appears to be a reasonable option Selection of any specific agent should be dictated by the
in refractory cases of trigeminal neuralgia. Patients should be medication’s relative efficacy and side-effect profile as well
cautioned against exposure of the agent to sensitive areas. as the patient’s medical history and personal preferences. In
Additionally, patients should be forewarned of the potential refractory cases nonpharmacological treatments such as
for extreme burning pain in areas exposed to the medication. glycerol injections, neurosurgical manipulation of the tri-
geminal nerve, and local irradiation serve as attractive alter-
natives to medication. For the foreseeable future, though,
Narcotic Agents medical management will remain the first-line treatment for
trigeminal neuralgia.
Oral Agents
Oral narcotics of varying potency may ameliorate excruciat- References

ing episodes of pain due to trigeminal neuralgia. Although
narcotics are not typically considered neuropathic pain 1. Rozen TD. Trigeminal neuralgia and glossopharyngeal neuralgia.
Neurol Clin. 2004;22:185–206.
agents, they are nonetheless effective in treating this type of
2. Joffroy A, Levivier M, Massager N. Trigeminal neuralgia: patho-
pain. Narcotics act at opioid receptors, providing relatively physiology and treatment. Acta Neurol Belg. 2001;101:20–5.
nonspecific pain reduction. Short-acting agents may be use- 3. Rappaport ZH, Devor M. Trigeminal neuralgia: the role of self-
ful for “breakthrough pain.” None, however, has been stud- sustaining discharge in the trigeminal ganglion. Pain. 1994;56:
127–38.
ied specifically with regard to trigeminal neuralgia. Dosing
4. Love S, Coakham HB. Trigeminal neuralgia: pathology and patho-
of oral narcotics must be individualized to the patient. The genesis. Brain. 2001;124:2347–60.
key to management is finding the minimally effective dose. 5. Kaye AH. Trigeminal neuralgia: vascular compression theory. Clin
Used judiciously, narcotics can be very useful. Obvious con- Neurosurg. 2000;46:499–506.
6. Grosneth G, Cruccu G, Alksne J, et al. Practice parameter: the diag-
cerns, though are tolerance and dependence.
nostic evaluation and treatment of trigeminal neuralgia (an evidence-
based review): report of the quality standards subcommittee of the
American Academy of Neurology and the European Federation of
Fentanyl Patch Neurological Societies. Neurology. 2008;71:1183–90.
7. Blom S. Trigeminal neuralgia: its treatment with a new anticonvul-
sants drug. Lancet. 1962;1:839–40.
Long-acting narcotics may provide more sustained pain 8. Campbell FG, Graham JG, Zikha KJ. Clinical trial of Carbamazepine
relief than the short-acting agents. Unfortunately, given the (Tegretol) in trigeminal neuralgia. J Neurosurg Neurol Psychiatry.
degree of bad press surrounding some oral agents, many 1966;29:265–7.
9. Nicol C. A four-year double-blind randomized study of Tegretol in
physicians are reluctant to prescribe such medications. With
facial pain. Headache. 1969;9:54–7.
this in mind, controlled release formulations such as the 10. Rockliff BW, Davis EH. Controlled sequential trials of carbamaze-
Fentanyl patch represent attractive solutions to the problem. pine in trigeminal neuralgia. Arch Neurol. 1966;15:129–36.
Fentanyl is a short-acting narcotic. The patch is comprised 11. Beydoun A. Safety and efficacy of Oxcarbazepine: results of ran-
domized, double-blind trials. Pharmacotherapy. 2000;20:152S–8.
of a matrix that allows slower release of Fentanyl into the
12. Carrazana E, Mikoshiba I. Rationale and evidence for the use of
skin over many days. Thus, these patches are less likely to Oxcarbazepine in neuropathic pain. J Pain Symptom Manage.
produce any euphoria that may be the reinforcing aspect of 2003;25:S31–5.
670 N.C. Porter
13. Sist T, Filadora V, Miner M, Lema M. Gabapentin for idiopathic 20. Bergouignan M. Cures heureuses de nevralgies facials essentielles
trigeminal neuralgia: reported to cases. Neurology. 1997; par diphenyl-hidantoinate de soude. Rev Laryngol Otol Rhinol.
48:1467. 1942;63:34–41.
14. Khan OA. Gabapentin relieves trigeminal neuralgia in multiple 21. Raskin NH. Facial pain. In: Headache. New York: Churchill
sclerosis patients. Neurology. 1998;51:611–4. Livingstone; 1988. p. 333-74.
15. Lunardi G, Leandri M, Albano C, Cultrera S, Fracassi M, Rubino V, 22. Zakrzewska JM. Trigeminal neuralgia. In: Medical management.
Favale E. Clinical effectiveness of Lamotrigine and plasma levels in London: W.B. Saunders; 1955. p. 80–107.
essential and symptomatic trigeminal neuralgia. Neurology. 1997; 23. Fromm GH, Terrence CF, Chatta AS, Glass JD. Baclofen in trigem-
48:1714–7. inal neuralgia: its effect on the spinal trigeminal nucleus: a pilot
16. Canavero S, Bonicalzi V. Lamotrigine control of trigeminal neural- study. Arch Neurol. 1980;37(12):768–71.
gia: an expanded study. J Neurol. 1997;244:527–32. 24. Fromm GH, Terrence CF, Chatta AS. Baclofen in the treatment of
17. Zakrzewska JM, Chaudhry Z, Nurmikko TJ, Patton DW, Mullens trigeminal neuralgia: double-blind study and long-term follow-up.
EL. Lamotrigine (Lamictal) in refractory trigeminal neuralgia: Arch Neurol. 1984;15(3):240–4.
results from a double-blind placebo-controlled crossover trial. Pain. 25. Court JE, Kase CS. Treatment of Tic Douloureux with a new anti-
1997;73:223–30. convulsant (Clonazepam). J Neurol Neurosurg Psychiatry. 1976;
18. Zvartau-Hind M, Din MU, Gilani A, Lisak RP, Khan OA. Topiramate 39:297–9.
relieves refractory trigeminal neuralgia in MS patients. Neurology. 26. Epstein JB, Marcoe JH. Topical application of Capsaicin for treat-
2000;55:1587–8. ment of oral neuropathic pain and trigeminal neuralgia. Oral Surg
19. Solaro C, Messmer MM, Brichetto G, Gasperini C, Mancardi Oral Med Oral Pathol. 1994;77:135–40.
G. Topiramate relieves idiopathic and symptomatic trigeminal neu- 27. Fusco BM, Alessandri M. Analgesics effect of Capsaicin in idio-
ralgia. J Pain Symptom Manage. 2001;21:367–8. pathic trigeminal neuralgia. Anesth Analg. 1992;74:375–7.
for Movement Disorders 55
Yoshinori Higuchi
Patients with a high-risk factor for surgery (e.g., those

Introduction receiving anticoagulant therapy, the aged, or patients with sys-
temic complications) call for an alternative treatment option.
Stereotactic neurosurgery has evolved rapidly with the devel- Leksell treated five cases of tremor using GK surgery between
opment of the computed imaging systems, the advances in 1968 and 1970; however, beneficial results were limited,
the understanding of disease pathophysiology, the discovery mainly due to technical difficulties such as the lack of comput-
of new therapeutic targets, the increasingly widespread use erized imaging techniques [4]. The introduction of magnetic
of neurophysiological technology such as microelectrode resonance imaging allowed for more refined targeting. Rand
recording, and the use of deep brain stimulation (DBS), all of et al. reported the use of radiosurgery treatment for movement
which have contributed to the accessibility of functional ste- disorders, based on this modern imaging technology [5].
reotaxy. Radiosurgery involves the precise destruction of a Thereafter, numerous reports have been published. Careful
target containing neural structures or tumors, without dam- review of several such reports revealed that the results of
age to the adjacent tissues [1]. Leksell first described the radiosurgery were not always satisfactory and included some
concept of using radiosurgery for the destruction of normal cases with extremely poor outcomes. In fact, this was raised as
brain tissue in 1951 [2]. Leksell treated two patients with a major criticism of the technique [6]. Worldwide, 5 % of GK
intractable cancer pain using the gamma knife (GK) and surgery procedures are carried out for the treatment of func-
published the results in 1968 [3]. The target could not then tional disorders, predominantly trigeminal neuralgia. The
be visualized directly; instead, indirect target coordination ablation of deep brain structures using GK surgery is less
was provided by pneumoencephalography. The target tissue common, because the target for many functional disorders
was irradiated at 200–250 Gy using gamma units. In these cannot be visualized even with modern neuroimaging tech-
patients, autopsy findings revealed a well-circumscribed niques, and there is no opportunity to identify the final target
lesion in the posterior thalamus, at the pre-calculated site. using neurophysiological evaluation. A great deal of experi-
The application of stereotactic radiosurgery for the treat- ence with stereotactic surgery would be required to make the
ment of functional disorders by the pioneers of stereotactic decision of the target coordination. A recent prospective study
radiosurgery has received much attention. Extensive intraop- proved the efficacy of treatment using GK thalamotomy for
erative neurophysiological studies have provided extensive intractable tremor [7]. GK surgery is still one of the surgical
knowledge of the pathophysiology of movement disorders, options for treating tremor-dominant PD and ET.
particularly of essential tremor (ET) and Parkinson’s disease This chapter will review the current literature and discuss
(PD). Currently, the standard techniques for the surgical treat- the use of stereotactic radiosurgery for treatment of move-
ment for ET and advanced PD are DBS and ablation surgery. ment disorders.
Gamma Knife Thalamotomy
Thalamotomy using GK surgery was initially applied for the

Y. Higuchi, M.D., Ph.D. (*) treatment of intractable pain. In the majority of recent
Department of Neurological Surgery, Chiba University Graduate
School of Medicine, 1-8-1 Inohana, Chuo-ku, Chiba City, Chiba
reports, intractable tremor was the main target disease for
260-8670, Japan GK thalamotomy. Table 55.1 summarizes the recent reports
e-mail: yhiguchi@faculty.chiba-u.jp of gamma knife thalamotomy [7–16].
Table 55.1 Summary of gamma knife thalamotomy
Unilateral Prescription
Author Use or bilateral dose (maximum Complication
(year) n Disease Targeting of plug treatment dose, Gy) Follow-up Effects rate (%) Complication
Pan 8 Parkinson’s disease Based on Schaltenbrand and Wahren atlas Yes Unilateral 160–180 4.5 months Tremor 12.5 Transient
(1996) [8] disappears in hemiparesis and
3 cases, reduction edema
in 3 cases
Friedman 15 Parkinson’s disease 3 7–8 mm anterior to PC, 11–13 mm lateral No Unilateral 120–140 3 months 2.1 decrease 46.7 Edema, etc.
(1999) [9] Essential tremor 12 to ACPC line, 2 mm above ACPC line in TRS
Duma 38 Parkinson’s disease 6 mm posterior to the midpoint of ACPC, No Unilateral 34 120–160 Median Reducing and 2.6 Dysarthria
(1999) [10] 15 mm lateral to ACPC line, 4 mm superior Bilateral 4 30 months eliminating tremor
to ACPC line in 90 %
50 % isodose line medially to IC
Duma 42 Parkinson’s disease 6 mm posterior to the midpoint of ACPC, No Unilateral 38 120–160 Median Good or 1.3 Paresthesia, motor
(2007) [11] Essential tremor 15 mm lateral to ACPC line, 4 mm superior Bilateral 4 30 months excellent 54 % weakness, dysphasia
to ACPC line
30 % isodose line medially to IC
Niranjan 11 Essential tremor 8 25 % of the ACPC distance anterior to PC, No Unilateral 130–150 Median Tremor 9.1 Hemiparesis,
(2000) [12] Multiple sclerosis 3 laterality 1/2 third ventricle width + 11 mm (model U) 6 months reduction in all dysarthria
from ACPC, 2 mm superior to the ACPC line
Mathieu 6 Multiple sclerosis 25 % of the ACPC distance + 1 or 2 mm No Unilateral 130–150 Median Tremor 16.7 Hemiparesis
(2007) [13] anterior to PC, laterality 1/2 third ventricle 27.5 months reduction in all
width + 11 mm from ACPC, 2 or 2.5 mm
superior to the ACPC line
Kondziolka 31 Essential tremor 25 % of the ACPC distance + 1 mm anterior Yes Unilateral 130–140 Median Improvement 7.7 Hemiparesis,
(2008) [14] to PC, laterality 1/2 third ventricle 36 months in 88.0 % speech disturbance,
width + 11 mm from ACPC, 2.5 mm dysphagia
superior to the ACPC line. The 20 % isodose
line medial to the internal capsule
Young 161 Essential tremor Based on Schaltenbrand and Wahren atlas No Unilateral 130 141–152 Mean Writing, 77 %; 8.4 Sensory loss, motor
(2010) [15] Staged 44 ± 33 drawing, 81 % impairment, speech
bilateral 42 months disturbance, etc.
Lim 18 Parkinson’s disease 3 25 % of ACPC + 1 mm to PC, 11 mm lateral Yes Unilateral 130–140 Mean Improvement 16.7 Thalamic
(2010) [16] Essential tremor 15 to third ventricle wall, 2–3 mm above ACPC 19.2 ± 7.3 in tremor rating hemorrhage,
months score ADL score sensory disturbance
Ohye 72 Parkinson’s disease 59 Longitudinal thalamic length 45 %, 2 mm Yes Unilateral 130 24 months Improvement 1.4 Motor weakness
(2012) [7] Essential tremor 13 inside external border of internal capsule, in 81.1 %
3–4 dorsally to ACPC
TRS tremor rating scale, IC internal capsule
55 Stereotactic Radiosurgery for Movement Disorders 673
Indication of GK Thalamotomy Extensive experience using stereotactic thalamotomy with

microelectrode recordings might support GK thalamotomy,
Indications of GK thalamotomy for the treatment of intrac- which has the disadvantage of involving no neurophysiologi-
table tremor are determined in accordance with more con- cal examination.
ventional stereotactic surgery. While DBS or conventional
thalamotomy is available, the use of GK thalamotomy has
been reserved for specific patients, such as those who are Tolerance of the Internal Capsule
poor candidates for conventional surgery (e.g., due to
advanced age or ongoing anticoagulant therapy). Accurate lateral coordinates are crucial to prevent radiation
Intractable tremors could be classified as components of injury to the internal capsule. Duma et al. adjusted the 50 %
ET, PD, or multiple sclerosis. The most common disease isodose line medially to the internal capsule [24]. Kondziolka
treated by GK thalamotomy is ET. ET is a common adult et al. retained 20 % of the isodose line of the 4-mm collima-
movement disorder, with prevalence estimates derived from tor medially to the internal capsule [14]. The tolerance of the
population studies ranging from 0.4 to 5 % [17, 18]. The internal capsule thus remains unclear. However, Maruyama
incidence and prevalence of ET both increase with advanc- et al. used diffusion tensor tractography of the pyramidal
ing age [18], and conventional thalamotomy or DBS is not tract when dose planning of GK surgery for treatment of an
often recommended for aged patients with ET. arteriovenous malformation and calculated the tolerance of
PD is another indication for GK thalamotomy; however, the pyramidal tract [25, 26]. The maximum dose to the pyra-
subthalamic DBS improves symptoms of rigidity, bradykine- midal tract and 20- or 25-Gy volume in the pyramidal tract
sia and motor fluctuation, as well as tremor [19]. In a system- significantly correlated with the motor complication.
atic review of PD progression, the non-tremor dominant Interestingly, the internal capsule was found to be more frag-
subtype at disease onset appeared to favor rapid progression ile than the coronal radiation. A plugging technique to reduce
to Hoehn and Yahr stage III [20]. Tremor seems to predict a the dose of radiation to the internal capsule was applied in
slow progression of PD. If intractable tremor causes signifi- several recent studies (Fig. 55.1) [7, 14, 16]. It is now gener-
cant functional impairment, GK thalamotomy is an option ally accepted that the maximum dose of radiation to the
for tremor reduction [7]. internal capsule should be less than 20–25 Gy.
Target Coordination Prescribed Dose
Although the efficacy of GK thalamotomy for intractable GK treatment has been commonly performed using a 4-mm
tremor has been evaluated, the prescription dose and target collimator. The prescribed dose varied among the different
selection varied greatly among the studies reviewed (see studies (see Table 55.1). Steiner et al. reported that a dose of
Table 55.1). Previous studies have selected the ventralis 150 Gy was necessary for the reproducible creation of a
intermedius nucleus (VIM) as the target for treating intrac- human brain lesion, using an autopsy series of GK thalamot-
table tremor. In the majority of studies, target determination omy that used to treat intractable cancer pain [27]. In the
was based on the anterior commissure-posterior commissure reports published in the early 1990s, a relatively high dose
(ACPC) line. Anterior-posterior coordination was 6–8 mm was selected for lesioning; however, in more recent reports,
anterior to PC, or 25 % of the ACPC line from PC, while a maximum dose of 130–140 Gy was recommended for thal-
right-left coordination was 11–13 mm lateral to the ACPC amotomy using GK [7, 14, 16]. Okun et al. reported five
line, and rostral-caudal coordination was 2–4 mm above the cases with complications of GK thalamotomy, including two
ACPC line. Anatomical investigation of formalin-fixed cases of bilateral lesioning [6]. Radiation effects were found
human brains indicated the anterior tip of ventroposterior to have expanded into the internal capsule and mesencepha-
nucleus [21]. Ohye et al. adjusted the tentative target based lon, which caused further motor complications and pseudo-
on the ACPC line, so that the center of the radiation beam bulbar symptoms. However, the prescribed dose should be
would cover 45 % of the longitudinal axis of the thalamus for noted; the target was irradiated at a maximum dose of
actual irradiation [22]. No neurophysiological information 200 Gy. Such a high dose is not currently used for the patients
aids the refinement of the target and is in fact a disadvantage with intractable tremor.
in GK thalamotomy. Sato et al. performed depth recordings
around the region of the supposedly optimum target for
tremor, as determined in usual thalamotomy, to provide indi- Radiation Efficacy
rect support for GK thalamotomy, and demonstrated a rhyth-
mic discharge time locked to tremor, and the presence of Previous reports demonstrated the effectiveness of GK thala-
kinesthetic neurons within the expected target area [23]. motomy in treatment of tremor; however, the evaluation of
674 Y. Higuchi
Fig. 55.1 Representative GK thalamotomy treatment plan. Treatment plan (MRI) of right hemisphere GK thalamotomy for intractable left-hand
tremor and isodose lines (15 and 50 %) is indicated. The dose to the internal capsule is reduced by using the plugging technique
tremor reduction was not uniform (Table 55.1). The Unified Table 55.2 Onset of clinical effects after gamma knife thalamotomy
Parkinson’s Disease Rating Scale (UPDRS) [28], the Fahn- Author (year) Dose (Gy) Onset of clinical effects
Tolosa-Marin Tremor Rating Scale [29], and tremor-cessation Lindquist (1992) [31] 180 1 month
rates were all used to grade tremor and other symptoms, before Friehs (1995) [32] 160 3–4 weeks
and after treatment. Overall, 70–90 % of the tremor patients Pan (1996) [8] 160–180 3 days (1 case)
obtained an anti-tremor effect from GK surgery. Differences Young (1998) [33] 120–160 2–3 months
in the outcomes of GK surgery in different primary diseases Duma (1998) [24] 120 (160) 1 week to 8 months (2 months)
remain unclear due to insufficient numbers of patients. Friedman (1999) [9] 120–140 4 weeks
In order to clarify the safety and ascertain optimally effec- Ohye (2012) [7] 130 3 months
tive conditions for performing unilateral GK thalamotomy
for treatment of intractable tremors, a prospective multi-
institutional study was published by the Japan Leksell approximately 1 year (Table 55.2) [7–9, 24, 30–32]. A single
Gamma Knife Society (JLGK0301). The study included session of high-dose irradiation requires time to produce the
blinded videotaped assessments and demonstrated the effi- desired lesion. Leksell described autopsy cases in his first
cacy of GK surgery for the treatment of both tremor and report of GK thalamotomy [3]. Lesions irradiated at 200–
rigidity in PD; however, no effects on bradykinesia or gait 250 Gy were confirmed pathologically 2 and 5 months after
disturbance were found [7]. treatment, and the lesions contained necrosis with a marginal
A slow improvement over several months tended to gliotic zone. Radiosurgical lesions are dependent on the dose
characterize the clinical course following GK thalamotomy, received, volume of tissue irradiated, and time [33, 34].
in contrast to the immediate effect of radiofrequency The low dose of irradiation (130–140) used more recently
coagulation. The period over which improvements in clinical would require some time to produce the lesion following
symptoms were observed varied widely, from several days to GK thalamotomy. However, some of the patients treated by
GK thalamotomy experienced treatment effects even before Most of the reported complications were transient; however,
coagulation necrosis occurred. Kondziolka demonstrated some of patients suffered from severe sequelae [6]. A high
that the lower doses (80 Gy) used to treat trigeminal neural- prescribed dose of radiation seemed to be associated with
gia cause incomplete axonal degeneration in a baboon model more complications.
[35]. Early changes, prior to coagulative necrosis in the thal- Siderowf et al. reported involuntary movements resulting
amus, might occur after GK surgery and cause an early from GK thalamotomy [41]. Magnetic resonance (MR)
improvement in clinical symptoms. images revealed extended edema involving the thalamus,
basal ganglia, and midbrain. Dystonia and choreoathetosis
emerged, despite a good response to treatment with a marked
Comparison of GK Thalamotomy reduction in the tremor.
and Conventional Thalamic Surgery Okun et al. reported pseudobulbar laughter following GK
thalamotomy [42]; in this case, the tremor persisted follow-
The anti-tremor effects of GK thalamotomy should be coming GK thalamotomy. MR images demonstrated extended
pared with those of other surgical treatment options. edema, involving the internal capsule and posterior thala-
However, a head-to-head comparison with other surgical mus. The prescribed dose was unknown in both of these
procedures was limited, because various methods of tremor cases. Extended perifocal edema to the neuronal structures
evaluation and variable follow-up terms were applied in each surrounding the target site is the expected cause of these
study. Based on tremor-cessation rates, VIM DBS resulted in complications. Ohye et al. reported thalamic reaction on MR
a tremor-absent state in 67–96 % of both PD and ET patients images following GK thalamotomy [7]. Extended T2 elon-
[36, 37]. VIM thalamotomy also abolished tremor in gation surrounding the GK lesion occurred in 10.9 % of the
65–89 % of PD and ET patients. The tremor-cessation rate patients, and the prediction of a thalamic reaction was diffi-
achieved with GK thalamotomy was, however, relatively cult in spite of a uniform treatment protocol.
low; 79 % of the patients had good tremor control, including
slight or infrequent tremors, which is similar to the results of
VIM thalamotomy [37, 38]. Gamma Knife Pallidotomy
Posteroventral pallidotomy (PVP) has been used to treat

Microelectrode Recording After GK patients with advanced PD, since Laitinen et al. [43] reintro-
Thalamotomy duced PVP and confirmed the results that Leksell and
Svennilson et al. [44] and others described earlier [45, 46].
Some failed cases of GK thalamotomy have been reported. Reviews of controlled studies demonstrated an improvement
Some of these cases received subsequent conventional thala- in quality of life following surgery [47, 48]. Procedures with-
motomy, or DBS. In these cases, intraoperative monitoring out the need for battery replacement, risk of infection, or
revealed the radiation effects through electrical activity in hardware problems seemed to be beneficial for some patients
the thalamus. Normal neuronal and tremor rhythm were with PD. To date, DBS of the subthalamic nucleus and pal-
demonstrated in and adjacent to the lesion of the previously lidum has become the dominant surgical procedure used for
irradiated target [39]. Target deviation to the anterior and the treatment of PD.
medial direction, or rhythmic tremor discharge located close GK pallidotomy has been reported in a limited number
to the internal capsule, was less effective. This seems to of publications (Table 55.3) [5, 9, 10, 32, 49]. GK pallidot-
reflect technical limit of GK thalamotomy at present. We omy resulted in clinical improvement of rigidity, bradykine-
have relatively limited knowledge of the effects of radiation sia, and dyskinesia, in several of the patients. These
on neuronal activity in the thalamus. Terao et al. reported treatment results indicate less effectiveness compared with
their microelectrode findings after GK thalamotomy and PVP [50, 51]. However, some of the patients treated using
demonstrated a reorganization of kinesthetic cells in the irra- GK pallidotomy had an inadequate diagnosis of PD (e.g.,
diated VIM [40]. Changes in the somatotopography of kines- short disease duration (<5 years), nonresponsive to levodopa
thetic cells might make it difficult to perform a secondary therapy). At present, these patients rarely make good surgi-
surgery on these failed cases. cal candidates, because of the possibility of them having
parkinsonism of non-Parkinson’s disease. Young et al.
carried out a comparison of treatment results between
Complications GK pallidotomy and radiofrequency pallidotomy for PD
[52]. The authors reported an improvement of rigidity, bra-
Reported complications from GK thalamotomy varied dykinesia, and dyskinesia with both treatment modalities.
between 1.4 and 46.7 % (Table 55.1). These complications However, neither group showed a significant improvement
mainly consisted of hemiparesis and speech disturbances. in UPDRS score.
676
Table 55.3 Summary of gamma knife pallidotomy

Author Unilateral/ Prescription dose Follow-up Complication
(year) n Target coordination bilateral (maximum dose, Gy) (months) Effects rate (%) Complication
Rand (1992) [5] 8 2 mm anterior or posterior to the MC point, Unilateral 140–165 NA Rigidity NA NA
23–25 mm lateral to ACPC, 1–2 mm above reduction
ACPC line in 50 %
Optic tract outside 10 % isodose line
Friedman (1996) [49] 4 2–3 mm anterior to the MC point, 18–21 mm Unilateral 180 12 No effects 25 Severe edema
lateral to ACPC line, 3–6 mm below ACPC line
Young (1998) [32] 28 In the medial globus pallidus just superior Unilateral 24 120–160 12 Improvement 7.10 Visual field defect, hypophonia
to the optic tract and just inferomedial to the Bilateral 4 in bradykinesia
internal capsule in coronal images at 2–3 mm and rigidity in
anterior to MC point 64.3 %
Duma (1999) [10] 18 50 % isodose line of a single or double Unilateral 160 4–40 Improvement 50 Visual field defect, speech and
isocenter on the Gpi in 33 % swallowing disturbance
Friedman (1999) [9] 2 2–3 mm anterior to MC point, 18–23 mm Unilateral 120–140 8 Improvement – Mild edema, lacunar infarction
lateral to ACPC line, 4–6 mm below ACPC line of dyskinesia Hypophonia, dysphagia
ACPC anterior commissure-posterior commissure, MC mid-commissural
Y. Higuchi
Target coordination is again critical, because the medial for subthalamotomy using GK technology. Visualization of
globus pallidum is located in the angle between the roof of the subthalamic nucleus using MR imaging has improved
the optic tract and the lateral border of the internal capsule. significantly in the last decade, and recent MR imaging tech-
In most reports regarding GK pallidotomy, the isocenter was nology demonstrates the possibility of direct targeting, as
located at Laitinen’s PVP point [43], or the medial globus well as targeting using internal fiducial points (e.g., the red
pallidum, referring to the optic tract and the internal capsule. nucleus) [64]. Future advances will no doubt allow for the
The optic tract is highly vulnerable to irradiation, and accurate identification of different functional parts of the
therefore, careful dose planning is required for the successful subthalamic nucleus and the correlation between the location
treatment of benign disease. Dose and volume in the irradi- of subthalamic lesions and therapeutic effects obtained from
ated optic pathway are strongly associated with optic neu- radiofrequency subthalamotomy.
ropathy. In previous reports, more than 10 Gy of single-dose
irradiation caused a high incidence of optic neuropathy [53],
and previous radiation therapy increased the incidence of GK Surgery for Other Movement Disorders
optic neuropathy, even if dose to the optic tract was 8 Gy or
less [54]. Therefore, the incidence of visual field impairment Although there have been no studies designed to specially
was high following GK pallidotomy. Use of GK pallidotomy address this issue, GK surgery applied for the treatment of
has not become widespread due to the anatomical disadvan- movement disorders has been described in several case
tage of the pallidum for radiosurgery. reports. Kwon et al. treated hemiballism in a patient with a
history of tuberculous meningitis using GK pallidotomy
[65]. Pan et al. reported clinical improvement following GK
Gamma Knife Subthalamotomy thalamotomy in two cases of cervical dystonia [66], and
Ohye et al. treated a case with dystonia using GK thalamot-
The subthalamic nucleus is a standard treatment target for omy [67]. All patients were reported to experience signifi-
DBS in advanced PD [55]. The relationship between the cant improvements following treatment.
subthalamic nucleus and parkinsonian conditions has Treatment of secondary movement disorders treated using
been demonstrated using an MPTP (1-methyl, 4-phenyl, GK surgery has also been reported. Kurita et al. reported
1,2,3,6-tetrahydropyridine) monkey model, and hyperactiv- relief of hemiballism resulting from basal ganglia arteriove-
ity of the subthalamic nucleus is a fundamental pathophysi- nous malformation [68]. Yen et al. treated cervical dystonia
ological sign in the parkinsonian state [56–58]. Lesioning of caused by basal ganglia venous angioma and obtained sig-
the subthalamic nucleus reverses parkinsonian symptoms in nificant clinical improvement [69], and Karampelas et al.
the MPTP monkey model [59–61]. The scientific evidence treated a subthalamic nucleus metastasis, which was causing
promotes stereotaxy of the subthalamic nucleus for treat- contralateral hemichorea-hemiballism, and confirmed a
ment of PD. The efficacy of radiofrequency subthalamotomy reduction in involuntary movement [70]. Movement disor-
has been evaluated in several studies. Alvarez et al. report the ders due to an intracranial lesion, especially a deep-seated
therapeutic efficacy of unilateral subthalamotomy in PD lesion, would be a good indication for GK surgery.
[62]. Unilateral subthalamotomy provided significant and
sustained motor benefits contralateral to the lesion. However,
15 % of the patients developed postoperative hemichorea- Conclusions
ballism, which subsequently required pallidotomy.
Subthalamotomy using GK surgery has been described in GK thalamotomy for treatment of intractable tremor is a safe
only one case report [63]. The preliminary target was identi- and effective alternative to radiofrequency or DBS proce-
fied at 2 mm posterior to the mid-commissural line, 13 mm dures for high-risk patients. Extensive experience of conven-
lateral to the midline, and 5 mm inferior to the mid- tional stereotactic surgery is expected to stabilize treatment
commissural line. Small modifications of the target coordi- results. Collaboration with neurologists or a movement dis-
nates were made based on the patient’s brain anatomy and on order team is mandatory to assess treatment options and to
the Schaltenbrand and Wahren atlas. The target was irradi- evaluate the treatment results of GK surgery for movement
ated at 120 Gy, using a 4-mm collimator. The patient experi- disorders. Indication of GK pallidotomy in PD has been lim-
enced improvement of parkinsonism and was able to reduce ited due to relatively high complication rate arising from the
their levodopa dosage following treatment. This limited clin- anatomical disadvantageous position of the globus pallidum.
ical experience has not clarified the potential of GK subthal- For other movement disorders, the worldwide experience in
amotomy in PD yet. treatment with GK surgery is limited.
The final decision of electrode placement depends on the
findings of microelectrode recordings in most of DBS facilities. Conflict of Interest Notification No actual or potential conflicts of
Further refinement of targeting techniques will be required interest exist.
678 Y. Higuchi
21. Brierley JB, Beck E. The significance in human stereotactic brain

References surgery of individual variation in the diencephalon and globus palli-
dus. J Neurol Neurosurg Psychiatry. 1959;22:287–98.
22. Ohye C. From selective thalamotomy with microrecording to
1. Larsson B, Leksell L, Rexed B, Sourander P, Mair W, Andersson gamma thalamotomy for movement disorders. Stereotact Funct
B. The high-energy proton beam as a neurosurgical tool. Nature. Neurosurg. 2006;84(4):155–61.
1958;182(4644):1222–3. 23. Sato S, Ohye C, Shibazaki T, Zama A, Cai X. Neurophysiological
2. Leksell L. The stereotaxic method and radiosurgery of the brain. evaluation of the optimum target in gamma thalamotomy: indirect
Acta Chir Scand. 1951;102(4):316–9. evidence. Stereotact Funct Neurosurg. 2005;83(2–3):108–12; dis-
3. Leksell L. Cerebral radiosurgery. I. Gammathalamotomy in two
cussion 113–4.
cases of intractable pain. Acta Chir Scand. 1968;134(8):585–95.
24. Duma CM, Jacques DB, Kopyov OV, Mark RJ, Copcutt B, Farokhi
4. Lindquist C, Kihlstrom L, Hellstrand E. Functional neurosurgery—
HK. Gamma knife radiosurgery for thalamotomy in parkinsonian
a future for the gamma knife? Stereotact Funct Neurosurg.
tremor: a five-year experience. J Neurosurg. 1998;88(6):1044–9.
1991;57(1–2):72–81.
25. Maruyama K, Kamada K, Ota T, Koga T, Itoh D, Ino K, et al.
5. Rand RW, Jacques DB, Melbye RW, Copcutt BG, Fisher MR,
Tolerance of pyramidal tract to gamma knife radiosurgery based on
Levenick MN. Gamma Knife thalamotomy and pallidotomy in
diffusion-tensor tractography. Int J Radiat Oncol Biol Phys.
patients with movement disorders: preliminary results. Stereotact
2008;70(5):1330–5.
Funct Neurosurg. 1993;61 Suppl 1:65–92.
26. Maruyama K, Kamada K, Shin M, Itoh D, Aoki S, Masutani Y,
6. Okun MS, Stover NP, Subramanian T, Gearing M, Wainer BH,
et al. Integration of three-dimensional corticospinal tractography
Holder CA, et al. Complications of gamma knife surgery for
into treatment planning for gamma knife surgery. J Neurosurg.
Parkinson disease. Arch Neurol. 2001;58(12):1995–2002.
2005;102(4):673–7.
7. Ohye C, Higuchi Y, Shibazaki T, Hashimoto T, Koyama T, Hirai T,
27. Steiner L, Forster D, Leksell L, Meyerson BA, Boethius
et al. Gamma knife thalamotomy for Parkinson’s disease and essen-
J. Gammathalamotomy in intractable pain. Acta Neurochir (Wien).
tial tremor: a prospective multicenter study. Neurosurgery.
1980;52(3–4):173–84.
2012;70(3):526–35; discussion 535–6.
8. Pan L, Dai JZ, Wang BJ, Xu WM, Zhou LF, Chen XR. Stereotactic 28. Fahn S, Elton RL, Committee MotUD. Unified Parkinson’s Disease
Gamma thalamotomy for the treatment of parkinsonism. Stereotact Rating Scale. In: Fahn S, Marsden CD, Calne DB, Goldstein M
Funct Neurosurg. 1996;66 Suppl 1:329–32. (eds). Recent developments in Parkinson’s disease. Florham Park:
9. Friedman DP, Goldman HW, Flanders AE, Gollomp SM, Curran Jr Macmillan Health Care Information; 1987. pp. 153-164.
WJ. Stereotactic radiosurgical pallidotomy and thalamotomy with 29. Fahn S, Tolosa E, Martin C. Clinical rating scale for tremor. In:
the gamma knife: MR imaging findings with clinical correlation– Jankovic J, Tolosa E, editors. Parkinson’s disease and movement dis-
preliminary experience. Radiology. 1999;212(1):143–50. orders. 2nd ed. Baltimore: Williams & Wilkins; 1993. p. 271–80.
10. Duma CM, Jacques D, Kopyov OV. The treatment of movement 30. Lindquist C, Steiner L, Hindmarsh T. Gamma knife thalamotomy
disorders using Gamma Knife stereotactic radiosurgery. Neurosurg for tremor. Report of two cases. In: Steiner L (ed). Radiosurgery.
Clin N Am. 1999;10(2):379–89. New York: Raven Press; 1992. pp. 237-243.
11. Duma CM. Movement disorder radiosurgery–planning, physics 31. Friehs GM, Ojakangas CL, Pachatz P, Schrottner O, Ott E, Pendl
and complication avoidance. Prog Neurol Surg. 2007;20:249–66. G. Thalamotomy and caudatotomy with the Gamma Knife as a
12. Niranjan A, Kondziolka D, Baser S, Heyman R, Lunsford treatment for parkinsonism with a comment on lesion sizes.
LD. Functional outcomes after gamma knife thalamotomy for Stereotact Funct Neurosurg. 1995;64 Suppl 1:209–21.
essential tremor and MS-related tremor. Neurology. 2000;55(3): 32. Young RF, Shumway-Cook A, Vermeulen SS, Grimm P, Blasko J,
443–6. Posewitz A, et al. Gamma knife radiosurgery as a lesioning tech-
13. Mathieu D, Kondziolka D, Niranjan A, Flickinger J, Lunsford nique in movement disorder surgery. J Neurosurg. 1998;89(2):
LD. Gamma knife thalamotomy for multiple sclerosis tremor. Surg 183–93.
Neurol. 2007;68(4):394–9. 33. Kondziolka D, Lunsford LD, Claassen D, Maitz AH, Flickinger
14. Kondziolka D, Ong JG, Lee JY, Moore RY, Flickinger JC, Lunsford JC. Radiobiology of radiosurgery: Part I. The normal rat brain
LD. Gamma Knife thalamotomy for essential tremor. J Neurosurg. model. Neurosurgery. 1992;31(2):271–9.
2008;108(1):111–7. 34. Kondziolka D, Niranjan A, Lunsford LD, Flickinger JC.
15. Young RF, Li F, Vermeulen S, Meier R. Gamma Knife thalamotomy Radiobiology of radiosurgery. Prog Neurol Surg. 2007;20:16–27.
for treatment of essential tremor: long-term results. J Neurosurg. 35. Kondziolka D, Lacomis D, Niranjan A, Mori Y, Maesawa S,
2010;112(6):1311–7. Fellows W, et al. Histological effects of trigeminal nerve radiosur-
16. Lim SY, Hodaie M, Fallis M, Poon YY, Mazzella F, Moro E. Gamma gery in a primate model: implications for trigeminal neuralgia
knife thalamotomy for disabling tremor: a blinded evaluation. Arch radiosurgery. Neurosurgery. 2000;46(4):971–6; discussion 976–7.
Neurol. 2010;67(5):584–8. 36. Yamamoto T, Katayama Y, Kano T, Kobayashi K, Oshima H, Fukaya
17. Louis ED, Ottman R, Hauser WA. How common is the most com- C. Deep brain stimulation for the treatment of parkinsonian, essential,
mon adult movement disorder? Estimates of the prevalence of and poststroke tremor: a suitable stimulation method and changes in
essential tremor throughout the world. Mov Disord. 1998;13(1): effective stimulation intensity. J Neurosurg. 2004;101(2):201–9.
5–10. 37. Schuurman PR, Bosch DA, Bossuyt PM, Bonsel GJ, van Someren
18. Benito-Leon J, Bermejo-Pareja F, Morales JM, Vega S, Molina EJ, de Bie RM, et al. A comparison of continuous thalamic stimula-
JA. Prevalence of essential tremor in three elderly populations of tion and thalamotomy for suppression of severe tremor. N Engl J
central Spain. Mov Disord. 2003;18(4):389–94. Med. 2000;342(7):461–8.
19. Kleiner-Fisman G, Herzog J, Fisman DN, Tamma F, Lyons KE, 38. Linhares MN, Tasker RR. Microelectrode-guided thalamotomy for
Pahwa R, et al. Subthalamic nucleus deep brain stimulation: sum- Parkinson’s disease. Neurosurgery. 2000;46(2):390–5; discussion
mary and meta-analysis of outcomes. Mov Disord. 2006;21 Suppl 395–8.
14:S290–304. 39. Ohye C, Shibazaki T, Ishihara J, Zhang J. Evaluation of gamma
20. Post B, Merkus MP, de Haan RJ, Speelman JD. Prognostic factors thalamotomy for parkinsonian and other tremors: survival of neu-
for the progression of Parkinson’s disease: a systematic review. rons adjacent to the thalamic lesion after gamma thalamotomy.
Mov Disord. 2007;22(13):1839–51. quiz 1988. J Neurosurg. 2000;93 Suppl 3:120–7.
40. Terao T, Yokochi F, Taniguchi M, Kawasaki T, Okiyama R, Hamada 56. Mitchell IJ, Clarke CE, Boyce S, Robertson RG, Peggs D,
I, et al. Microelectrode findings and topographic reorganisation of Sambrook MA, et al. Neural mechanisms underlying parkinsonian
kinaesthetic cells after gamma knife thalamotomy. Acta Neurochir symptoms based upon regional uptake of 2-deoxyglucose in mon-
(Wien). 2008;150(8):823–7; discussion 827. keys exposed to 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine.
41. Siderowf A, Gollump SM, Stern MB, Baltuch GH, Riina HA. Neuroscience. 1989;32(1):213–26.
Emergence of complex, involuntary movements after gamma knife 57. Miller WC, DeLong MR. Altered tonic activity of neurons in the
radiosurgery for essential tremor. Mov Disord. 2001;16(5): globus pallidus and subthalamic nucleus in the primate model of
965–7. parkinsonism. In: Carpenter MB, Jayaraman A (eds). The basal
42. Okun MS, Heilman KM, Vitek JL. Treatment of pseudobulbar ganglia II. Structure and function. New York: Plenum; 1987.
laughter after gamma knife thalamotomy. Mov Disord. 2002; pp. 415-427.
17(3):622–4. 58. Levy R, Herrero MT, Ruberg M, Villares J, Faucheux B, Guridi J,
43. Laitinen LV, Bergenheim AT, Hariz MI. Leksell’s posteroventral et al. Effects of nigrostriatal denervation and L-dopa therapy on the
pallidotomy in the treatment of Parkinson’s disease. J Neurosurg. GABAergic neurons in the striatum in MPTP-treated monkeys and
1992;76(1):53–61. Parkinson’s disease: an in situ hybridization study of GAD67
44. Svennilson E, Torvik A, Lowe R, Leksell L. Treatment of mRNA. Eur J Neurosci. 1995;7(6):1199–209.
parkinsonism by stereotatic thermolesions in the pallidal region. 59. Bergman H, Wichmann T, DeLong MR. Reversal of experimental
A clinical evaluation of 81 cases. Acta Psychiatr Scand. parkinsonism by lesions of the subthalamic nucleus. Science.
1960;35:358–77. 1990;249(4975):1436–8.
45. Gillingham FJ, Kalyanaraman S, Donaldson AA. Bilateral stereo- 60. Guridi J, Herrero MT, Luquin MR, Guillen J, Ruberg M, Laguna J,
taxic lesions in the management of parkinsonism and the dyskine- et al. Subthalamotomy in parkinsonian monkeys. Behavioural and
sias. Br Med J. 1964;2(5410):656–9. biochemical analysis. Brain. 1996;119(Pt 5):1717–27.
46. Narabayashi H. Surgical treatment in the levodopa era. In: Stern G, 61. Aziz TZ, Peggs D, Sambrook MA, Crossman AR. Lesion of the
editor. Parkinson’s disease. London: Chapman & Hall; 1990. subthalamic nucleus for the alleviation of 1-methyl-4-phenyl-
p. 597–646. 1,2,3,6-tetrahydropyridine (MPTP)-induced parkinsonism in the
47. Goetz CG, Poewe W, Rascol O, Sampaio C. Evidence-based medi- primate. Mov Disord. 1991;6(4):288–92.
cal review update: pharmacological and surgical treatments of 62. Alvarez L, Macias R, Pavon N, Lopez G, Rodriguez-Oroz MC,
Parkinson’s disease: 2001 to 2004. Mov Disord. 2005;20(5): Rodriguez R, et al. Therapeutic efficacy of unilateral subthalamot-
523–39. omy in Parkinson’s disease: results in 89 patients followed for up to
48. Martinez-Martin P, Deuschl G. Effect of medical and surgical inter- 36 months. J Neurol Neurosurg Psychiatry. 2009;80(9):979–85.
ventions on health-related quality of life in Parkinson’s disease. 63. Keep MF, Mastrofrancesco L, Erdman D, Murphy B, Ashby
Mov Disord. 2007;22(6):757–65. LS. Gamma knife subthalamotomy for Parkinson disease: the sub-
49. Friedman JH, Epstein M, Sanes JN, Lieberman P, Cullen K, thalamic nucleus as a new radiosurgical target. Case report.
Lindquist C, et al. Gamma knife pallidotomy in advanced J Neurosurg. 2002;97(5 Suppl):592–9.
Parkinson’s disease. Ann Neurol. 1996;39(4):535–8. 64. Brunenberg EJ, Platel B, Hofman PA, Ter Haar Romeny BM,
50. Vitek JL, Bakay RA, Freeman A, Evatt M, Green J, McDonald W, Visser-Vandewalle V. Magnetic resonance imaging techniques for
et al. Randomized trial of pallidotomy versus medical therapy for visualization of the subthalamic nucleus. J Neurosurg. 2011;115(5):
Parkinson’s disease. Ann Neurol. 2003;53(5):558–69. 971–84.
51. de Bie RM, Schuurman PR, Bosch DA, de Haan RJ, Schmand B, 65. Kwon Y, Whang CJ. Stereotactic Gamma Knife radiosurgery for
Speelman JD. Outcome of unilateral pallidotomy in advanced the treatment of dystonia. Stereotact Funct Neurosurg. 1995;64
Parkinson’s disease: cohort study of 32 patients. J Neurol Neurosurg Suppl 1:222–7.
Psychiatry. 2001;71(3):375–82. 66. Pan L, Zhang N, Dai JZ, Wang EM. Gamma knife radiosurgery for
52. Young RF, Vermeulen S, Posewitz A, Shumway-Cook torsion spasm. Report of two cases. J Neurosurg 2000;93 Suppl
A. Pallidotomy with the gamma knife: a positive experience. 3:189-90.
Stereotact Funct Neurosurg. 1998;70 Suppl 1:218–28. 67. Ohye C, Shibazaki T, Zhang J, Andou Y. Thalamic lesions produced
53. Leber KA, Bergloff J, Pendl G. Dose-response tolerance of the by gamma thalamotomy for movement disorders. J Neurosurg.
visual pathways and cranial nerves of the cavernous sinus to stereo- 2002;97(5 Suppl):600–6.
tactic radiosurgery. J Neurosurg. 1998;88(1):43–50. 68. Kurita H, Kawamoto S, Sasaki T, Shin M, Ueki K, Momose T, et al.
54. Hasegawa T, Kobayashi T, Kida Y. Tolerance of the optic apparatus Relief of hemiballism from a basal ganglia arteriovenous malfor-
in single-fraction irradiation using stereotactic radiosurgery: evalu- mation after radiosurgery. Neurology. 1999;52(1):188–90.
ation in 100 patients with craniopharyngioma. Neurosurgery. 69. Yen CM, Sheehan J, Pan HC. Successful treatment of cervical dys-
2010;66(4):688–94; discussion 694–5. tonia induced by basal ganglion venous angioma with gamma knife
55. Benabid AL, Mitrofanis J, Chabardes S, Seigneuret E, Torres N, thalamotomy. J Clin Neurosci. 2012;19(3):470–1.
Piallat B, et al. Subthalamic nucleus stimulation for Parkinson’s 70. Karampelas I, Podgorsak MB, Plunkett RJ, Fenstermaker RA.
disease. In: Lozano AM, Gildenberg PL, Tasker RR (eds). Textbook Subthalamic nucleus metastasis causing hemichorea-hemiballism
of stereotactic and functional neurosurgery: Springer; 2009. treated by gamma knife stereotactic radiosurgery. Acta Neurochir
pp. 1603-1630. (Wien). 2008;150(4):395–6; discussion 397.
Perspective on Radiosurgery Versus
Conventional Surgery for Movement 56
Disorders
Zion Zibly, Andrew B. Shaw, John Y.K. Lee,

and Ali R. Rezai
and 1982, 762 patients were treated for multiple brain and
Introduction neurologic disorders; of them, five were treated for parkin-
sonism [6]. The logic behind the utilization of radiosurgical
Among the different surgical techniques of treating movement techniques is that it is even less invasive than chronic brain
disorders, deep brain stimulation (DBS) has become the stimulation (DBS), thus eliminating the need for an open sur-
procedure of choice around the world, due to its safety com- gical procedure and hardware-related complications such as
pared to other lesioning procedures. Since the introduction infection and device-related complications such as lead
and technical refinements of DBS for the treatment of move- breakage or failure. In addition, radiosurgery is less time-
ment disorders in the 1980s, more than 80,0000 procedures consuming than microelectrode-guided radiofrequency
have been done worldwide. For patients with medically lesioning and can be performed in a subset of patients who
refractory movement disorders, DBS procedures have largely are at higher surgical risk for open procedures secondary to
replaced neuroablative techniques. This shift away from medical comorbidities including cardiorespiratory, coagu-
ablative procedures is multifactorial. Among the more lopathy, or any other surgical contraindication. It is also
important factors are its proven efficacy, the reversibility, and suitable for patients who cannot tolerate an awake procedure.
the perceived minimally invasive nature of DBS compared Nevertheless, radiosurgery has its limitations, disadvantages,
with destructive procedures. During the last several years, and complications, which will be further discussed in this
there has been an increase in the use of radiosurgical tech- chapter. This chapter reviews the evidence for DBS versus
niques such as Gamma Knife (GK), linear accelerator radiosurgical lesioning in the treatment of patients with med-
(Linac), and more recently high-intensity focused ultra- ically refractory movement disorders and defines the role of
sound (HIFU) to perform thalamotomies and pallidotomies. radiosurgery in the treatment of these patients. We will also
The preference between radiosurgical techniques and DBS compare the safety and efficacy of DBS compared to radio-
depends mostly on the surgeons experience and patients surgery. Finally, we will review the emerging role of HIFU in
comorbidities [1–5]. treating movement disorders.
The first use of cobalt-based GK was done in 1968 by the
Swedish neurosurgeon Lars Leksell, and between 1968
Thalamic DBS Versus Radiosurgical
Z. Zibly, M.D.
Thalamotomy
Department of Neurological Surgery, The Ohio State University
Medical Center, Columbus, OH, USA Thalamic DBS
A.B. Shaw, M.D. (*)
Department of Neurological Surgery and Neuromodulation, High-frequency stimulation of the ventralis intermedius
Wexner Medical Center at The Ohio State University, (VIM) nucleus of the thalamus was first introduced and
Columbus, OH, USA
described by Benabid in 1987. In his paper, Benabid
e-mail: Andrew.Shaw@osumc.edu
described the relief of extrapyramidal tremor in Parkinson’s
J.Y.K. Lee, M.D.
disease patients who had bilateral symptoms. In this revo-
Department of Neurosurgery, University of Pennsylvania,
Philadelphia, PA, USA lutionary report, DBS was introduced in one side of the
brain, while the other side was treated with radiofrequency
A.R. Rezai, M.D.
Department of Neurological Surgery and Neuroscience, ablation [7]. This subset of patients was chosen because of
The Ohio State University, Columbus, OH, USA the recognition of cognitive complications in patients who
682 Z. Zibly et al.
had undergone bilateral thalamotomies [8, 9]. In this initial An important complication of DBS pertinent to all DBS
report, all patients achieved some relief of tremor, and this cases whether targeting VIM, globus pallidus interna (GPi),
was the foundation of VIM DBS for the treatment of tremor. or STN is the risk of intraoperative/procedural hemorrhage.
Over the next few years, practitioners in multiple centers Benabid reported a microhemorrhage rate of six incidences
in both North America and Europe confirmed and defined out of a total of 177 operations, of which three were asymp-
the benefits of DBS of the VIM nucleus in the treatment of tomatic [11]. The intracranial hemorrhage rate in the North
parkinsonian tremor. In Parkinson’s disease (PD), Benabid American and European trials was 2 % and 5 %, respectively
et al. followed approximately 100 patients for up to 8 years [13, 20]. Starr recently reported a hemorrhage rate of 3.3 %,
and have documented an 88 % improvement in tremor [10, but only a subset of these were symptomatic, thus resulting
11]. Only a minority of patients developed tolerance to DBS in a symptomatic hemorrhage rate of 0.6 %. In addition,
(e.g., a diminished effect over time requiring greater voltages there has been some retrospective study evidence that the
and higher battery use). Thalamic DBS appeared to improve hemorrhage rate may increase with the aggressive use of
parkinsonian tremor but not bradykinesia and rigidity nor the microelectrode recording (MER). Hariz et al., in a meta-
levodopa-induced dyskinesias which are the other cardinal analysis of reported series, noted that non-MER techniques
features of Parkinson’s disease [5, 12]. In another study, were at least five times less likely to have hemorrhagic com-
Koller et al. reported the results of a double-blind, North plications [23]. For example, Palur et al. demonstrated a
American, multicenter study (four centers) for 24 PD patients slightly higher rate of hemorrhage in cases performed with
[13]. They randomized unilateral thalamic DBS against pla- MER versus macrostimulation alone (0.4 % versus 0.2 %)
cebo, and double-blinded assessments demonstrated a sig- [24]. The DBS Study Group reported a relationship between
nificant reduction in resting, postural, and action tremor. the number of microelectrode tracks and the risk of hemor-
However, DBS of VIM again did not improve the rigidity rhage: patients without hemorrhage had a mean of 2.9 ± 1.8
and bradykinesia of PD patients, and hence VIM DBS in par- passes compared with 4.1 ± 2.0 among those who had hemor-
kinsonian patients is currently reserved for the rare unilateral rhage (p = 0.05) [25]. There are no prospective, randomized
tremor-predominant patient. The majority of patients with studies of MER versus non-MER strategies, and there is a
medically refractory PD will undergo STN DBS at this time, large variability in MER techniques used at different centers
which will be discussed later. with regard to number of passes and equipment employed. In
DBS of the VIM nucleus for the treatment of essential all studies noted, however, the overall small risk of bleeding
tremor is highly effective with more than 90 % of patients in both MER and non-MER groups indicates the overall
demonstrating a satisfactory result [11, 14–18]. It is the sur- relative safety of both techniques [24–26]. The risk of hem-
gical treatment of choice for patients with essential tremor orrhage is not negligible, but not all hemorrhages result in
who experience diminished quality of life despite maximal obvious clinical deficit [27], and the risk of hemorrhage must
medical therapy. With VIM DBS as many as 50 % of patients be considered in relation to the benefit obtained in patients
have complete abolition of upper extremity and head tremor with medically refractory movement disorders.
with significant improvements in their quality of life [19]. In Another specific complication of DBS is related to the
the North American and European multicenter trials, DBS of need to implant permanent hardware into the body. Bhatia
the VIM thalamus was effective, and beneficial results were et al. reported in their systematic review of 3,550 DBS surger-
maintained after several years [13, 20]. ies from 35 studies an infection rate of 4.7 % [28]. Boviatsis
Adverse effects from VIM DBS are infrequent and usu- et al. reported a total hardware-related complication rate of
ally mild. Benabid reported mild contralateral paresthesias, 4.3 % in their study of 106 DBS procedures. These included
limb dystonia, and cerebellar dysmetria, which were all con- electrode breakage, lead migrations, and infection [29, 30].
trolled by adjustments of stimulation parameters [21]. Kondziolka et al. documented a similar hardware-related
Dysarthria and gait disequilibrium were uncommon and complication rate in Pittsburgh: 27 % of 66 patients devel-
nearly always limited to patients receiving bilateral stimula- oped hardware complications [31]. In contrast with these
tion or who had undergone prior thalamotomy. Significant two reports, other studies have documented lower rates of
cognitive deficits have not been observed after thalamic hardware-related complications. Fenoy et al. reported in a
stimulation, although mild deficits in verbal fluency have large series of 1,333 DBS implanted electrodes and 1,218
been documented [19]. The great benefit of DBS compared implantable pulse generators (IPG) postoperative rate of
with permanent lesioning procedures is the ability to control infection that required reoperation as low as 1.24 % [32]. In
multiple parameters, including voltage, pulse width, fre- Koller’s series of 53 patients with unilateral thalamic stimu-
quency, and contacts allowing for a myriad of ways to modu- lators, the incidence of infection was 3.8 % with a 1.9 %
late the affects of DBS. Hence, the therapeutic index (e.g., malfunction rate [13]. The Grenoble group’s large series of
the difference between beneficial effect and adverse effect) 197 patients contained three patients who experienced infec-
can be navigated through careful programming [22]. tions and five with scalp erosions leading to exposed hardware,
56 Perspective on Radiosurgery Versus Conventional Surgery for Movement Disorders 683
Fig. 56.1 Axial T1-weighted contrast-enhanced and T2-weighted MRI scans 4 months following GK thalamotomy
for a total rate of 2.5 % [12]. The European Multicenter Study applications such as movement disorders. Currently, Gamma
yielded an erosion/infection rate of 2.7 %, and the North Knife thalamotomy (GKT) is the primary SRS method used
American trial reported 2.9 % [20]. None of these large stud- to treat movement disorders, although the application of lin-
ies reported data on lead migration or fracture, and in both, ear accelerator by creating precise thalamic lesion has been
the mean follow-up time was 12 months or less thus limiting described and was in fact one of the first functional radiosur-
the evaluation of long-term complications. Among the 143 gical procedures [4, 27, 33, 34] (Fig. 56.1).
patients in the multicenter prospective series of patients In order to discuss the merits of radiosurgical thalamot-
undergoing GPi or STN DBS, there were 5 leads that omy, it is beneficial to consider the role of conventional
migrated, 4 infected leads, 2 broken leads, 1 scalp erosion, radiofrequency thalamotomy in the treatment of patients
and 1 incidence of equipment malfunction [25]. With with tremor. Tasker retrospectively studied patients who had
improved electrode and hardware design as well as improve- radiofrequency thalamotomy and thalamic stimulation [1].
ments in surgical technique, hardware-related complication He noted tremor recurrence in 15 % of thalamotomy patients,
rates appear to decrease. Nevertheless, hardware failure con- as opposed to 5 % of DBS patients. Importantly, 23 % of
tributes to morbidity in patients who undergo the placement the thalamotomy patients required repeat procedures for
of permanent deep brain stimulators. These complications are tremor control. In addition, thalamotomy resulted in a decline
obviously averted in patients who undergo a radiosurgical in writing performance in 23 % of patients, as opposed to no
lesioning procedure, and the relative risks must be weighed patients in the DBS group. All of the DBS patients achieved
for each patient. better than 50 % improvement in dexterity, writing, and
In summary, thalamic DBS for the treatment of tremor drinking ability, where only half of the thalamotomy cohort
is effective in Parkinson’s disease and essential tremor as obtained that level of benefit. With respect to complications,
demonstrated in multiple trials that include double- Tasker demonstrated that chronic stimulation had a lower
blinded, prospective studies. The risk of intraoperative/ incidence of ataxia and dysarthria, which are common
procedural hemorrhage is approximately 3 % with less complications associated with bilateral thalamotomy [1].
than one third of these being symptomatic. The risk of Moreover, when these complications occurred in the DBS
hardware-related complications ranges from 3 to 27 %, but group, simply adjusting the stimulation parameters served to
improvements in technique and hardware development abolish these problems. In another study that was a prospec-
will limit these problems. It is the opinion of most practi- tive, blinded study, Schuurman et al. randomized 68 patients
tioners in the field that the reversibility, symptomatic con- (45 with PD, 13 with ET, and 10 with multiple sclerosis)
trol, and ability to minimize side effects with adjustment to radiofrequency thalamotomy versus VIM DBS [35]. Both
of the DBS system warrant taking the risk associated with groups of patients had excellent suppression of tremor, but
the implantation of hardware. the VIM DBS group had greater improvements in activities
of daily living (ADL), and more importantly, the VIM DBS
group had fewer complications. Patients who underwent
Radiosurgical Thalamotomy thalamotomy were more likely to develop somnolence, cog-
nitive deterioration, and dysarthria, as well as gait or balance
During the last decade, stereotactic radiosurgery (SRS) has disturbance. Thus, Tasker and Schuurman et al. have both
been widely used to treat a diversity of brain and spine dis- confirmed the higher complication rates associated with thal-
eases. Advances in neuroimaging, planning systems, and amotomy compared with VIM DBS, and it is in this context
dosimetry have resulted in the expansion of SRS for functional that the use of radiosurgical thalamotomy must be considered.
684 Z. Zibly et al.
Table 56.1 Summary of Gamma Knife thalamotomy for movement disorders

Author Year published Treated disorder Target Number of procedures Follow-up (months) Complication rate (%)
Young 1998 PD ViM 27 22.3 (mean) 0
Friedman 1999 PD tremor ViM 15 13 46.7
ET
Niranjan 1999 12 24 (median) 8.3
Duma 1998 PD ViM 38 28 (median) 0
Niranjan 2000 ET ViM 11 6 (median) 9.1
MS tremor
Young 2000 PD tremor 158 12–96 1.3
ET
Mathieu 2007 MS tremor ViM 6 25.5 (mean) 16.6
Kondziolka 2008 ET ViM 31 36 (median) 6.4
Lim 2010 PD tremor 18 19.2 (mean) 16.7
ET
Young 2010 Essential tremor ViM 161 44 (mean) 6.9
Elaimy 2010 ET ViM 1 144 0
Ohye 2012 PD tremor ViM 72 24 N/A
ET
Yen 2012 Dystonia ViM 1 72 0
PD Parkinson’s disease, ET essential tremor, MS multiple sclerosis, ViM ventralis intermedius nucleus
As opposed to both radiofrequency thalamotomy and who had a previously mapped lesion [38]. At present, Ohye
DBS, GKT offers a noninvasive lesioning method that does has reported radiosurgical thalamotomy in 70 patients with
not require a skin incision, burr hole, or passage of elec- PD and has reported tremor suppression of at least two thirds
trodes. In addition, it does not harbor any foreign hardware in more than 80 % of patients with no loss of effect even after
placement, and the risk of infection/erosion is eliminated. 10 years in some patients [37, 39]. Duma et al. has reported
Hence, the surgical risk of complications is potentially their series of 42 radiosurgical thalamotomies in 38 patients
lower than that of radiofrequency thalamotomy or DBS. over a 7-year period. In their series of patients, 24 % had
This technique may be ideal for patients who are not suitable complete relief of tremor, 55 % had “successful” results in
to undergo a surgical procedure due to medical comorbidities improvement of their tremor, and 21 % of patients had only
or to those who want to avoid open surgical procedures “mild” or no relief of tremor. Median follow-up in this series
with hardware implantations. GKT is effective and reported was 28 months [40, 41]. Niranjan et al. reported favorable
to treat parkinsonian tremor, essential tremor, multiple results in 12 patients who were not candidates for open surgi-
sclerosis-related tremor, and posttraumatic-related tremor cal procedures (nine patients with essential tremor and three
with different results [36]. A major disadvantage of GKT is patients with MS-related tremor) [42]. Eight of their 12
the inability to confirm the target physiologically by using patients with essential tremor had complete tremor arrest,
MER and macrostimulation (electrophysiological monitor- while the remaining four patients had improvement in tremor.
ing). One has to rely purely on imaging-defined anatomic The largest series of patients undergoing GKT has been pub-
targeting, which lacks the accuracy of brain mapping. In lished by Young et al. who retrospectively reviewed 158
addition, the method of lesion generation is delayed, and as patients who underwent radiosurgical thalamotomy [43]. Of
has been suggested in some recent publications, the lesion the 74 patients with parkinsonian tremor followed at least 4
can be variable in size [37]. Hence, GKT has both advan- years, 79.7 % were still tremor-free, while 70.6 % of the 16
tages and disadvantages that may give it a role in the treat- patients with essential tremor were tremor-free. In this series,
ment of tremor. only one patient sustained a transient complication, and two
Compared with the large multicenter studies of deep brain patients sustained permanent complications including hemi-
stimulation for tremor, there are only a few published case paresis and mild facial paresthesias. No cognitive changes
series of patients treated by Gamma Knife radiosurgical thal- were reported. Kondziolka et al. [44] reported a series of 31
amotomy for the treatment of parkinsonian tremor and essen- essential tremor patients treated with SRS followed for a
tial tremor [6] (Table 56.1). Ohye was one of the first to mean of 36 months. Of their patient cohort, 18 (66.6 %)
advocate radiosurgical lesioning of the thalamus in patients reported improvement in action tremor and writing, 6
who had a contralateral radiofrequency lesion or in patients (22.2 %) showed improvement in action tremor alone, and 3
(11.1 %) showed no improvement. In this study, there were the thalamotomy procedures were performed at exceptionally
two complications (7.4 %) that were both permanent, includ- high doses using 200 Gy (one patient was treated at 150 Gy).
ing hemiparesis, dysphagia, and speech impairment. Recently, Unfortunately, a systematic study of dose–response compli-
Lim et al. reported their experience with GKT for treating cation rate has not been performed to date.
disabling tremor [36]. Although this study is derived from a Because of the delay in effect after GKT, an accurate doc-
single center, it is the only prospective blinded independent umentation of complications requires vigilant follow-up.
observer study that was done for GKT. In this series, assess- The complications that can occur mainly include radiation
ment of tremor, ADL, complications, and long-term outcomes necrosis, brain edema, and bleeding. There isn’t an estab-
were analyzed in 14 patients with mean follow-up period of lished time frame for the occurrence of complication, and
7.3 months. They concluded that an improvement is achieved they can happen even years following GKT treatment.
only in tremor and ADL, but this improvement seems to be Rothstein et al. [47] reported an intrathalamic hemorrhage in
modest, non-sustained, and less impressive than achieved the precise site of prior GKT 7.5 years after treatment. In
with DBS. In this study, three patients developed delayed Young’s series, 3 of 74 patients developed a complication
complications including hemiparesis and speech deficit. [43]. Other reports are single-center reports of complications
In contrast to the major studies that were done to evaluate without a denominator to define the total number of proce-
DBS, GKT reports lack the stringent standards to prove effi- dures or patients who underwent GKT. Siderowf et al.
cacy and safety (multicenter, double-blinded, randomization, reported one case of a complex, involuntary movement after
long-term follow-up). This makes it difficult to compare Gamma Knife radiosurgical thalamotomy [2]. Unfortunately,
these different modalities; however, the results of these stud- specific dose parameters were not specified in that case
ies suggest that radiosurgical thalamotomy is less superior to report. Okun et al. reported a series of eight patients who
DBS or even radiofrequency thalamotomy. developed multiple complications, including hemiplegia,
In opposition to the immediate response and relief of homonymous visual field deficits, hand weakness, dysar-
tremor following DBS or thalamic radiofrequency ablation, thria, hypophonia, aphasia, arm and face numbness, pseudo-
Gamma Knife radiosurgical thalamotomy effect is delayed, bulbar laughter, and death [46]. Mathieu et al. reported
and different studies reported different time frames from progressive contralateral hemiplegia following GKT with T2
anywhere between immediate affects to 12 months [37, 40, changes around targeted lesion [48]. These complications all
42]. Niranjan et al. [44] reported immediate relief in tremor developed after treatment at different Gamma Knife centers,
following GKT. Moro et al. [36] reported in their study of 18 and it does not appear that the same rate of complications
patients an improvement in tremor with a delay of 6–12 have developed at all radiosurgical centers. Nevertheless, the
months following GKT. Ohye et al. [36] reported in their true incidence of complications after radiosurgical thalamot-
study improvement in tremor starting 1 year following omy is not well known.
GKT. This gap between treatment and functional progress One particular explanation of the complications associ-
must be taken under consideration when making a treatment ated with radiosurgical thalamotomy may be related to the
decision, since most patients with movement disorders suffer variability in lesion size. Animal studies in rodents and
daily and some cannot tolerate the time between treatment nonhuman primates have demonstrated the ability to create
and clinical effect. This delay is related to neurobiological necrotic lesions in animal brain parenchyma [45, 49, 50].
sequences that take place in the brain parenchyma following Young et al. reported a clear correlation between lesion vol-
GKT and the time it takes the ionizing radiation to function- ume and complication. In their study, the mean lesion vol-
ally damage or to destroy the tissue targeted. The time to ume in which no complications were noticed was 188 m,
lesion effect can be modified and accelerated with increased while the mean volume for complications was 871 mm
doses [45]; however, the optimal technique for Gamma Knife [51]. Dose, volume, and time are the three key factors that
radiosurgical lesioning has not been established. Due to the determine the nature of the functional ablation [45]. Higher
different response of tissue to ionizing radiation, lesion size doses can create necrosis faster as can the use of a larger-
can be smaller than expected, while in other cases can be volume collimator, although almost all movement disorder
larger and expand to eloquent adjacent structures resulting in GK lesions employ a 4-mm-diameter collimator. This
serious adverse effects [46]. Peak central doses can range dose–volume–time response curve can roughly be corre-
from 120 to 200 Gy. Duma et al. reported a difference in lated in humans; however, despite this apparent consistency
clinical outcome with different doses, suggesting an improve- in dose delivery and tissue response, Friehs et al. report
ment with higher doses [41]. Despite this correlation, most variable size on follow-up imaging after GK radiosurgical
surgeons at this time now perform radiosurgical lesioning for ablation for functional procedures. Friehs et al. examined
functional disorders at lower doses (e.g., 140 Gy) [42]. This follow-up imaging after 140 Gamma Knife radiosurgical
decision is partially based on the series of eight patients all lesions produced in the treatment of patients with PD, pain,
treated at the same institution who developed complications and ET [52]. In all cases, the 4-mm collimator was used for
published by Okun et al. [46]. In that series, all except one of the Gamma Knife, but the postoperative scans demonstrated
686 Z. Zibly et al.
lesion sizes that ranged from undetectable to 4,000 mm3. The most consistent effect of pallidal ablation appears to be
Friehs correlated the higher doses with greater lesion volume the relief of contralateral L-dopa-induced dyskinesias,
and suggested that doses of 160 Gy and above were associ- whereas the amount of improvement in contralateral bra-
ated with inordinately large volumes. Franzini et al. reported dykinesia, rigidity, and tremor differs among the many
in their study that older patients have a higher chance of studies [55]. In contrast with the number of studies docu-
having an unexpected response, and lesion size cannot be menting the effects of ablation of the posteroventral GPi,
predicted [53]. Thus, although lesion size appears to be cor- there are remarkably few studies that extend the initial obser-
related with radiation dose, there appears to be a patient to vation of Siegfried and Lippitz that chronic high-frequency
patient variability in response to Gamma Knife radiosurgical stimulation of the internal pallidum may effectively treat
lesioning. In addition, Ohye reports two different types of PD [56]. GPi DBS is also becoming increasingly common
lesions after radiosurgical lesions using the single 4-mm col- procedure as STN DBS for the treatment of Parkinson’s
limator isocenter with central doses of 130 Gy: “circum- disease. GPI DBS has also been used for the treatment of
scribed round high signal area of approximately 7- to 8-mm psychiatric disorders such as Tourette’s syndrome. Cannon
diameter surrounding a smaller low signal area. The other et al. reported the implantation of bilateral GPI DBS in 11
is characterized by an irregular-shaped high signal zone patients with medically intractable Tourette’s syndrome. In
extending into the medial thalamic area and/or internal cap- their study, there was 48 % improvement in motor tics and
sule and often accompanied by streaking along the thalamo- 56.5 % reduction in phonic tics [57]. Other studies have
capsular border” [37, 41]. Ohye speculates that the delivery showed that a unilateral GPI DBS may also be effective for
time may influence lesion size. Since the cobalt sources of the treatment of Tourette’s syndrome [58].
the Gamma Knife need to be replaced over time, delivery of DBS of the STN is the most common application for
the same 143-Gy dose may require double the length of time Parkinson’s disease. DBS of the internal segment of the glo-
with old sources compared with new sources [37]. “After bus pallidus is also effective [54, 55]. The most consistent
reloading, the restricted lesion was more frequent and the effect produced by pallidal DBS is a marked reduction of
lesion volume was smaller.” contralateral L-dopa-induced dyskinesias [56, 59–68].
In conclusion, both DBS of the VIM or GKT have been Volkmann et al. demonstrated a 54 % improvement in the
employed in the treatment of tremor. Several well-conducted “off”-period UPDRS motor score at 1-year follow-up, with
trials have demonstrated the efficacy and complication rate significant improvement in bradykinesia, tremor, posture,
of DBS of the VIM thalamic nucleus in the treatment of both and gait. “On”-period motor symptoms did not improve sig-
parkinsonian tremor and essential tremor. In contrast, large, nificantly after surgery except for dyskinesias, which were
well-conducted trials of Gamma Knife radiosurgical thala- reduced by 83 % at 1-year follow-up [68]. It appears that GPi
motomy have not been performed. Thus, it is difficult to DBS may directly reduce levodopa-induced dyskinesias,
compare directly the efficacy and risks of GKT of the VIM whereas STN DBS reduces levodopa-induced dyskinesias
nucleus for the treatment of tremor. Nevertheless, lesioning indirectly via reduction of overall medication intake. Loher
of the thalamus is an alternative treatment for tremor, and et al. reported 1-year results of 16 patients showing a 38 %
Gamma Knife ablation may be effective when properly per- improvement in medication-off UPDRS motor scores and a
formed. Hence, the first-line surgical therapy in the treatment 33 % improvement in the ADL score in patients receiving
of patients with medically refractory tremor is VIM DBS, unilateral stimulation [69]. Bilateral stimulation led to a
and GKT remains an option for patients who cannot undergo slight improvement in these results. Some improvement in
an open surgical procedure secondary to medical comorbidi- the medication-on state was also noted. A larger series of 36
ties. Gamma Knife radiosurgery of the VIM thalamus, how- patients [25] demonstrated that bilateral pallidal stimulation
ever, should only be performed by neurosurgeons who have results in a median motor improvement of 37 % and an
experience in both deep brain stimulation and radiofrequency increase from 28 to 64 % of the day without disabling invol-
thalamic procedures to optimize results. untary movements [25]. Many reports demonstrate the ben-
eficial effects of GPi stimulation on dyskinesias, on–off
fluctuations, and tremor [3, 61, 66, 68, 70]. Also, pallidal
Pallidal DBS Versus Radiosurgical stimulation has been validated in those patients who have
Pallidotomy previously undergone contralateral pallidotomy. In a cohort
of four patients who underwent GPi stimulation contralateral
Pallidal DBS to a prior pallidotomy, motor scores improved by almost
50 %, while bradykinesia was decreased by 37 % and tremor
Since its reintroduction by Laitinen and colleagues in 1992, by 93 % without serious adverse cognitive or motor effects
many successful pallidotomies have been performed through- [71]. Despite the benefits of GPi DBS, most centers currently
out the world in the treatment of Parkinson’s disease [54]. favor STN DBS in the treatment of Parkinson’s disease.
Dystonia is one movement disorder whose primary surgi- Radiosurgical Pallidotomy

cal target is the GPi as opposed to the STN or VIM; however,
the treatment of dystonia with DBS remains in its infancy. The pallidotomy is a time-honored surgical procedure that
Unlike the rapid improvement seen in PD, GPi DBS in dys- was performed in the 1950s by many surgeons, including
tonia results only in gradual improvement over several Guiot, Spiegel and Wycis, Talairach, Riechert, and Leksell.
months. Preliminary results suggest that more mobile, fluid The reintroduction of Leksell’s posteroventral pallidotomy
abnormal posturing may respond quicker, whereas fixed by Laitinen in 1992 (after its disappearance, which coincided
tonic abnormal posturing may require a longer period to with the benefits of L-dopa therapy) initiated a resurgence of
improve [71–74]. Given these encouraging results, there interest in the surgical treatment of movement disorders. One
have been more patients with both primary and secondary of the most consistent benefits of radiofrequency pallidot-
dystonia who have undergone bilateral GPi DBS. Patients omy is improvement in contralateral drug-induced dyskine-
with DYT1 and primary dystonia have proven to respond sias. Nearly every study has documented improvement in
more favorably to GPi DBS than some of the secondary dys- L-dopa dyskinesias from 61 to 82 % [55, 79–87]. Tremor
tonias. This may relate to variability in the etiology of the appears to respond by 33 to 90 % at 6 months, although
secondary dystonias. Currently, DBS for dystonia is not tremor alone is not significant indication for pallidotomy.
FDA approved and has a humanitarian exemption that limits The results for the symptoms of parkinsonian rigidity and
the ability for high-quality studies to be performed. bradykinesia are not as consistent with studies demonstrat-
The complication rate of GPi DBS is similar to that of ing some benefit. In addition, there have only been minimal
VIM DBS. The risk of hemorrhage, infection, and hardware- benefits in gait. Given the results described above, conven-
related complications are similar. Although there are no tional pallidotomy is only recommended for patients who are
direct studies that compare the incidence of cognitive com- severely disabled by asymmetrical L-dopa dyskinesias and
plications after bilateral pallidal DBS, the overall reported who are not candidates for implanted hardware systems.
complication rate after bilateral DBS does not appear to be Bilateral pallidotomies are not considered a good option for
as high as the rate after bilateral pallidal destructive lesions. most patients given the high rates of cognitive complications
Global scores of cognitive function exhibit little change after after bilateral destructive lesions in the globus pallidus [61].
unilateral or bilateral pallidal stimulation, although subtle Gamma Knife radiosurgical pallidotomy differs from con-
worsening of frontal lobe scores including verbal fluency ventional radiofrequency pallidotomy primarily in its less
have been described [61, 65, 74, 75]. Risk factors for cogni- invasive nature. It can theoretically expect to provide excel-
tive deterioration include preoperative L-dopa dosages and lent relief of levodopa-induced dyskinesias and tremor with
advanced age [74]. fewer complications than conventional pallidotomies.
Despite the lack of direct comparison between the effi- The evidence supporting the use of Gamma Knife palli-
cacy of STN DBS versus GPi DBS, most neurologists/neu- dotomy is even more sparse than the evidence supporting the
rosurgeons have selective preference for STN versus GPi use of GKT. Only case reports and case series are available
DBS. Recently, the European NSTAPS study assessed for study. Indeed, the last published case series of patients
whether GPi DBS gives grater functional improvement treated with Gamma Knife pallidotomy dates back to 1998
than STN DBS in advanced PD. This was a multicenter ran- [87], which most likely reflects the established success of
domized study with a total number of 128 patients [76]. STN DBS in the treatment of Parkinson’s disease and the
The results of this study showed improvement in the off- high complication rates of GK pallidotomy. Friedman et al.
medication state in the STN group compared to the GPi reported four patients in 1995 who were treated with 180 Gy
group in the UPDRS score (20.3 versus 11.4), but failed to to the right posteroventral pallidum for advanced PD [88].
show any other benefit of STN DBS over GPi DBS. There One patient had improvement in dyskinesias but also became
are some who believe that STN DBS may have additional transiently psychotic and demented. The three other patients
risk for cognitive compromise in patients who are exhibit- neither improved nor suffered a complication. In a later pub-
ing significant cognitive deficits. On the other hand, recent lication, Friedman et al. reported two patients who underwent
randomized, controlled multicenter studies showed that radiosurgical pallidotomy with lower doses of 120 to 140 Gy
DBS of the STN does not reduce overall cognition or affec- [89]. They chose to perform the procedure only in patients
tivity [77, 78]. Nevertheless, due to the potential risk for who had undergone a prior radiofrequency pallidotomy with
cognitive deficits following STN DBS, baseline neuropsy- physiologic mapping. They created a mirror-image lesion,
chological studies and evaluation are important and should assuming symmetry between the two sides. Both of these
be done prior to surgery. Overall both STN and GPi have patients, however, developed complications: hemiparesis in
similar efficacies for PD; however, STN is the more com- one patient and hypophonia and swallowing difficulties in
mon target for PD DBS, while the primary target of choice the other patient. In addition, Friedman et al. reported a stroke
for dystonia is the GPi. in one patient as a complication from GK pallidotomy [90].
688 Z. Zibly et al.
Fig. 56.2 T1 contrast-enhanced MRI scan with a superimposed reformatted Schaltenbrand and Wahren atlas with bilateral electrodes trajectory
targeting the STN
This was attributed to vascular hyalinization and thrombosis.

The largest series of GK pallidotomies comes from Young
et al. who reported a series of 29 patients who underwent
Gamma Knife pallidotomy [87]. He reports that two thirds of
his patients demonstrated improvements in bradykinesia and
rigidity. Only one of these 29 patients developed a complica-
tion—hemianopsia 9 months after the procedure. UPDRS
scores were not consistently evaluated by blinded observers
in this study. With respect to complications, Okun et al.
reported three complications after Gamma Knife pallidotomy Fig. 56.3 Postoperative anteroposterior and lateral skull X-rays dem-
performed in one center [46]. Three patients who had Gamma onstrating the implanted STN DBS electrodes
Knife pallidotomy had been treated with doses of 200, 150,
and 100 Gy. All three experienced visual field loss and/or
hemiparesis and dysarthria.
Compared with GKT, there appears to be a shortage of
published case series of patients treated with GK pallidot-
omy for the treatment of Parkinson’s disease. This most
likely represents a lack of enthusiasm on the part of the treat-
ing surgeons and referring neurologists. There are few stud-
ies that truly document the benefit of GK pallidotomy in a
blinded and unbiased fashion, and there are several reports of
complications after GK pallidotomy. Hence, at this stage,
GK pallidotomy is not performed as a viable method of
lesioning in the treatment of PD. GK pallidotomy have been
reported in case studies as a successful treatment option for Fig. 56.4 Postoperative axial and sagittal T1-weighted MRI scans
dystonia and hemidystonia [91]. showing the implanted STN DBS electrodes
Subthalamic DBS Versus Radiosurgical One-year follow-up of 24 patients implanted with bilateral
Subthalamotomy STN DBS demonstrated that UPDRS ADL and motor
scores improved by 60 % in the off-medication state [93].
Subthalamic DBS The UPDRS subscores for akinesia (56 %), tremor (80 %),
rigidity (68 %), and gait (55 %) also improved, in contrast
At the present time, DBS of the subthalamic nucleus is the with the results seen with VIM stimulation [11]. However,
most widely practiced surgical treatment for patients with the effect in the on-medication state was not as pro-
idiopathic Parkinson’s disease (Figs. 56.2, 56.3, and 56.4). nounced, with only a 10 % improvement in the motor score
Benabid et al. were the first group to demonstrate the efficacy noted. Levodopa drug use was cut in half. These results have
of STN DBS in the treatment of PD in the mid-1990s [92]. held up at 5-year follow-up [94]. Kumar et al. performed a
prospective, randomized, double-blinded study comparing Radiosurgical Subthalamotomy

STN DBS with the stimulator on and off [94]. They found a
65 % reduction in off-period motor UPDRS scores, a 40 % Subthalamotomy in the form of ischemic strokes or hemor-
reduction in on-period motor UPDRS scores, and an 85 % rhages have generally been known to cause hemiballism.
reduction in L-dopa-induced dyskinesias in the seven patients However, despite this conventional teaching, there are case
who were evaluated. Limousin et al. found a 60 % improve- reports of parkinsonian patients with spontaneous STN
ment in off-period UPDRS scores but only a 10 % improve- hemorrhage who have experienced motor improvement
ment in on-period UPDRS scores. without the occurrence of hemiballism [109, 110]. In addi-
These two prospective double blind studies comparing tion, there are centers outside of the USA that have reported
STN DBS with the stimulator on and off found a 60–65 % on the benefits of therapeutic STN ablative surgery using
reduction in off-period motor UPDRS scores and a 10–40 % traditional radiofrequency lesioning techniques. The Cuban
reduction in on-period motor UPDRS scores with an asso- group reported their results in 18 patients after bilateral dor-
ciated 85 % reduction in L-dopa-induced dyskinesias [93]. sal subthalamotomy [111]. After a 16-month average fol-
In summary, STN DBS is quite effective in the treatment of low-up period, there was a 58 % improvement in UPDRS
Parkinson’s disease. Levodopa-responsive motor features motor scores in the “off” period. Three patients developed
typically respond well to STN DBS with the possible severe generalized chorea, dysarthria, and balance instabil-
exception of tremor. The axial features of PD such as dys- ity that lasted 6 months from which they spontaneously
arthria, postural instability, and freezing of gait may not recovered. The Bristol group reported their results on 50
reliably respond to STN DBS. Disease-related progression subthalamotomies in 39 patients [112]. After unilateral STN
of these features over time may compromise overall post- lesions, the patients demonstrated a 46 % improvement in
surgical benefit despite continuing response of limb motor the motor UPDRS score at 24 months. In addition, they
function. were able to decrease their levodopa dose by 50 %. With
Preoperative levodopa responsiveness (by varying defini- respect to complications, one patient developed an intrace-
tions) is often reported as a predictive factor for a positive rebral hemorrhage, and one patient developed postoperative
response to surgery [12, 95–102]. A positive response to a hemiballism on the side contralateral to the STN lesion that
levodopa challenge and correlation with good outcome may lasted for 3 weeks and then spontaneously disappeared.
be more evident with STN DBS than with GPi DBS [103, Hence, subthalamotomy via radiofrequency lesioning is a
104]. The mean percent improvement in motor UPDRS dur- procedure in its infancy with a much diminutive number of
ing the levodopa test ranges from 40 to 70 % in published patients treated compared to DBS in the reported literature.
studies. As the levodopa response is the main predictor of The results are promising, but safety and efficacy remains to
outcome after STN DBS [96, 102, 103, 105], the benefit will be proved.
be greatest in those patients who have a high off-drug score There is only a single case report in the literature with
and a low on-drug score. In other words, the “best on” may one patient treated with Gamma Knife subthalamotomy
be a better predictor for functional outcome than the numeri- for Parkinson’s disease [113]. This one patient had a
cal magnitude of the response. prior radiofrequency pallidotomy and then had a Gamma
The hardware complications of STN DBS are similar to Knife procedure with a single 120-Gy, 4-mm shot to the
those listed in the section on VIM DBS as are the issues subthalamic nucleus. Forty-two months after the proce-
related to hemorrhage rates (e.g., approximately 3 % hemor- dure, her postoperative Parkinson’s disease disability rat-
rhage rate with less than one third being symptomatic). ing score had decreased from 28 to 11, and the patient
Complications related to STN stimulation, however, are dif- had no side effects. Gamma Knife subthalamotomy
ferent from VIM or GPi stimulation. Involuntary movements results in a small area of ischemia which can be seen and
can occur during the initial phase of STN stimulation but tend followed with magnetic resonance imaging studies
to be mild and generally respond to changes in the stimulation (Fig. 56.1). Although this is perhaps a promising case
parameters [106, 107]. In a retrospective study, Volkmann report, Gamma Knife subthalamotomy cannot be recom-
noted that STN stimulation appeared to be associated with a mended as a safe option in patients with Parkinson’s disease
higher incidence of adverse events compared with GPi stimu- at this time.
lation at 1-year follow-up periods [108]. The STN DBS group
had higher rates of depression, anhedonia, hypophonia, dys-
arthria, and apraxia of eyelid opening. Nevertheless, the com- Future of Movement Disorder Surgery
plication rates of STN DBS generally are mild and can be
improved with changes in stimulator settings, and STN is the During the last decade, new technologies have emerged with
most common DBS surgery performed for PD. applications to movement disorder surgery.
690 Z. Zibly et al.
Intraoperative Magnetic Resonance Imaging erated functional lesions in the mammalian brain using what
in Treatment of Movement Disorders they termed “neurosonic surgery” in the late 1940s [116].
These developments led to the prototypical HIFU devices
During the last years, an increasing number of medical cen- described in 1949 by Lars Leksell, which gained notoriety in
ters have implicated the use of intraoperative MRI-guided the 1950s for the ablation of soft tissue lesions [117].
DBS. The surgical procedure can be done with and without a Furthermore, by the late 1950s, Fry and colleagues were suc-
frame and uses an image-guided system for targeting and cessfully using ultrasound to treat movement disorders. The
implanting the electrode. Currently, there is only one real- histological changes in the brain tissue consist of thermal
time intraoperative MRI-guided system that has been injury and necrosis. However, ultrasound therapy as a neuro-
approved by the FDA to be used. The main drawback of surgical approach was largely ignored at the time, with most
using an intraoperative MRI-guided DBS system is the procedures needing intensive craniotomies [118].
inability to perform MER and macrostimulation in order to The combination of MRI and HIFU as a treatment modal-
functionally confirm the location, refine initial targeting, ity was first introduced and received FDA approval for the
account for anatomical variations, and assess side effects treatment of uterine fibroids. More recently, MRI-guided
during stimulation. Also it is only applicable for targeting HIFU (MRgHIFU) was approved for the treatment of pros-
areas of the brain which can be identified with an MR imag- tate cancer and breast fibroadenomas. MRgHIFU allows for
ing, such as the STN, GPi, GPe, and thalamus. The benefits target acquisition and guidance through “real-time” ther-
of MR guidance during DBS surgery include targeting per- mometry giving this modality far greater accuracy than ear-
formed after the dura is opened and brain shift has occurred, lier HIFU technology [119]. Additional clinical trials are
high-resolution MR images for confirmation of electrode ongoing in the treatment of a variety of movement disorders
position, decreasing the number of MER electrodes passing including essential tremor, dystonia, and Parkinson’s disease
through the brain, and reducing surgical time [114]. It also and essential tremor. Its utility in neurosurgery is in its
allows the surgeons to assess the occurrence of intracranial infancy, and ongoing research may open the door to new pos-
bleeding and eliminates the need to withhold antiparkinso- sibilities. The HIFU treatment for essential tremor has been
nian medications and the discomfort and additional stress to approved recently in Europe, and a multicenter trial is
the patient being awake. The surgical method chosen is expected to begin in the USA. A pilot study utilizing HIFU
dependent on both the surgeon and patient preference while in the treatment of Parkinson’s disease is currently been eval-
considering the risks and benefits. uated in the USA.
Gamma Knife Radiosurgery Conclusion
GK for the treatment of movement disorders is a growing There are several options for neurosurgeons to treat advanced
field, and since there is a lack of neurosurgeons familiar and and intractable movement disorders such as essential tremor,
experienced with DBS, it has gained popularity. The proce- Parkinson’s disease, and dystonia. DBS of the VIM thala-
dure appears to be safe, and patients can easily tolerate it as mus, STN, and GPi is the most common approach in devel-
it is short and does not necessitate hospitalization. In addi- oped countries and is proven to be effective and safe in the
tion, it is noninvasive and does not require open surgery or treatment of patients with Parkinson’s disease, essential
hardware implantation. The patient and practitioner must tremor, and dystonia. Radiofrequency lesioning has also
weigh the potential complications with the benefits when been performed for decades with established efficacy in
choosing any surgical procedure. experienced hands.
Gamma Knife radiosurgery for the treatment of movement
disorders has a specific limited role and needs to be further
High-Intensity Focused Ultrasound investigated (Table 56.2). At this time, GKT has been shown
to have some benefits in specific scenarios such as patients
High-intensity focused ultrasound (HIFU) is a new, evolving with coagulopathy, older age, and surgical contraindications
technology. This modality provides a promising new noninva- to DBS or RF surgery. GK radiosurgery’s role in GPi and
sive and alternative treatment for a number of conditions such STN for PD and dystonia needs further investigation particu-
as movement disorders, tumors, and other conditions [115]. larly by the same neurosurgeons who have expertise in DBS.
Edmund Harvey Newton in 1929 first demonstrated the In addition to movement disorders, a number of other condi-
central nervous system’s (CNS) sonographic response with tions are being investigated with DBS, RF lesioning, and
using ultrasound to stimulate nerves and muscles. gamma knife including obsessive–compulsive disorder, major
Subsequently, William Fry and colleagues successfully gen- depression, traumatic brain injury, and Alzheimer’s disease.
Table 56.2 Landmark differences between radiosurgery and deep brain stimulation for movement disorders
Intraoperative MRI-guided
Radiosurgery Radiofrequency lesioning Deep brain stimulation DBS
Experience 45 years 55 years >25 years 7 years
Adjustable? No Lesion may be “sculpted” Yes Yes
in or
Time to clinical 1–12 months Minutes to hours Minutes to hours Minutes to hours
effect?
Reversibility? No No Yes Yes
Repeatable? No Yes N/A N/A
Targeting method Imaging only Imaging plus Imaging plus Imaging only
microelectrode recording microelectrode recording
and macrostimulation and macrostimulation
Possible Necrosis Lesion spread to adjacent Hardware infection/ Hardware infection/erosion
complications structures erosion
Lesion spread to Higher incidence of Electrode breakage Electrode breakage
adjacent complications when
structures performed bilaterally
12. Benabid AL, Benazzouz A, Hoffmann D, Limousin P, Krack P,

References Pollak P. Long-term electrical inhibition of deep brain targets in
movement disorders. Mov Disord. 1998;13 Suppl 3:119–25; Epub
1. Tasker RR. Deep brain stimulation is preferable to thalamotomy 1998/11/25.
for tremor suppression. Surg Neurol. 1998;49(2):145–53; discus- 13. Koller W, Pahwa R, Busenbark K, Hubble J, Wilkinson S, Lang A,
sion 53–4; Epub 1998/02/11. et al. High-frequency unilateral thalamic stimulation in the treat-
2. Siderowf A, Gollump SM, Stern MB, Baltuch GH, Riina ment of essential and parkinsonian tremor. Ann Neurol. 1997;42(3):
HA. Emergence of complex, involuntary movements after gamma 292–9; Epub 1997/10/23.
knife radiosurgery for essential tremor. Mov Disord. 2001;16(5): 14. Speelman JD, Bosch DA. [Continuous electric thalamus stimula-
965–7; Epub 2001/12/18. tion for the treatment of tremor resistant to pharmacotherapy].
3. Peppe A, Pierantozzi M, Altibrandi MG, Giacomini P, Stefani A, Ned Tijdschr Geneeskd. 1995;139(18):926–30; Epub 1995/05/06.
Bassi A, et al. Bilateral GPi DBS is useful to reduce abnormal (Continue elektrische thalamusstimulatie voor de behandeling van
involuntary movements in advanced Parkinson’s disease patients, farmacotherapie-resistente tremor).
but its action is related to modality and site of stimulation. Eur J 15. Alesch F, Pinter MM, Helscher RJ, Fertl L, Benabid AL, Koos
Neurol. 2001;8(6):579–86; Epub 2002/01/11. WT. Stimulation of the ventral intermediate thalamic nucleus in
4. Frighetto L, Bizzi J, Annes RD, Silva Rdos S, Oppitz P. Stereotactic tremor dominated Parkinson’s disease and essential tremor. Acta
radiosurgery for movement disorders. Surgical Neurol Int. 2012;3 Neurochir. 1995;136(1–2):75–81; Epub 1995/01/01.
Suppl 1:S10–6; Epub 2012/07/25. 16. Ondo W, Jankovic J, Schwartz K, Almaguer M, Simpson
5. Pollak P, Benabid AL, Limousin P, Benazzouz A. Chronic intrace- RK. Unilateral thalamic deep brain stimulation for refractory
rebral stimulation in Parkinson’s disease. Adv Neurol. 1997;74: essential tremor and Parkinson’s disease tremor. Neurology.
213–20; Epub 1997/01/01. 1998;51(4):1063–9; Epub 1998/10/22.
6. Elaimy AL, Arthurs BJ, Lamoreaux WT, Demakas JJ, Mackay 17. Kumar K, Kelly M, Toth C. Deep brain stimulation of the ventral
AR, Fairbanks RK, et al. Gamma knife radiosurgery for move- intermediate nucleus of the thalamus for control of tremors in
ment disorders: a concise review of the literature. World J Surg Parkinson’s disease and essential tremor. Stereotact Funct
Oncol. 2010;8:61; Epub 2010/07/29. Neurosurg. 1999;72(1):47–61; Epub 2000/01/21.
7. Benabid AL, Pollak P, Louveau A, Henry S, de Rougemont 18. Hubble JP, Busenbark KL, Wilkinson S, Pahwa R, Paulson GW,
J. Combined (thalamotomy and stimulation) stereotactic surgery Lyons K, et al. Effects of thalamic deep brain stimulation based on
of the VIM thalamic nucleus for bilateral Parkinson disease. Appl tremor type and diagnosis. Mov Disord. 1997;12(3):337–41; Epub
Neurophysiol. 1987;50(1–6):344–6; Epub 1987/01/01. 1997/05/01.
8. Matsumoto K, Asano T, Baba T, Miyamoto T, Ohmoto T. Long- 19. Troster AI, Fields JA, Pahwa R, Wilkinson SB, Strait-Troster KA,
term follow-up results of bilateral thalamotomy for parkinsonism. Lyons K, et al. Neuropsychological and quality of life outcome
Appl Neurophysiol. 1976;39(3–4):257–60; Epub 1976/01/01. after thalamic stimulation for essential tremor. Neurology.
9. Hirai T, Miyazaki M, Nakajima H, Shibazaki T, Ohye C. The cor- 1999;53(8):1774–80; Epub 1999/11/24.
relation between tremor characteristics and the predicted volume 20. Limousin P, Speelman JD, Gielen F, Janssens M. Multicentre
of effective lesions in stereotactic nucleus ventralis intermedius European study of thalamic stimulation in parkinsonian and
thalamotomy. Brain. 1983;106(Pt 4):1001–18; Epub 1983/12/01. essential tremor. J Neurol Neurosurg Psychiatry. 1999;66(3):289–
10. Benabid AL, Pollak P, Seigneuret E, Hoffmann D, Gay E, Perret 96; Epub 1999/03/20.
J. Chronic VIM thalamic stimulation in Parkinson’s disease, 21. Benabid AL, Pollak P, Gervason C, Hoffmann D, Gao DM,
essential tremor and extra-pyramidal dyskinesias. Acta Neurochir Hommel M, et al. Long-term suppression of tremor by chronic
Suppl. 1993;58:39–44; Epub 1993/01/01. stimulation of the ventral intermediate thalamic nucleus. Lancet.
11. Benabid AL, Pollak P, Gao D, Hoffmann D, Limousin P, Gay E, 1991;337(8738):403–6; Epub 1991/02/16.
et al. Chronic electrical stimulation of the ventralis intermedius 22. Volkmann J, Herzog J, Kopper F, Deuschl G. Introduction to the
nucleus of the thalamus as a treatment of movement disorders. programming of deep brain stimulators. Mov Disord. 2002;17
J Neurosurg. 1996;84(2):203–14; Epub 1996/02/01. Suppl 3:S181–7; Epub 2002/04/12.
692 Z. Zibly et al.
23. Hariz MI, Fodstad H. Do microelectrode techniques increase 41. Duma CM, Jacques DB, Kopyov OV, Mark RJ, Copcutt B, Farokhi
accuracy or decrease risks in pallidotomy and deep brain stimula- HK. Gamma knife radiosurgery for thalamotomy in parkinsonian
tion? A critical review of the literature. Stereotact Funct tremor: a five-year experience. J Neurosurg. 1998;88(6):1044–9;
Neurosurg. 1999;72(2–4):157–69; Epub 2000/06/15. Epub 1998/06/03.
24. Palur RS, Berk C, Schulzer M, Honey CR. A metaanalysis com- 42. Niranjan A, Kondziolka D, Baser S, Heyman R, Lunsford
paring the results of pallidotomy performed using microelectrode LD. Functional outcomes after gamma knife thalamotomy for
recording or microelectrode stimulation. J Neurosurg. 2002;96(6): essential tremor and MS-related tremor. Neurology. 2000;55(3):
1058–62; Epub 2002/06/18. 443–6; Epub 2000/08/10.
25. The Deep-Brain Stimulation for Parkinson’s Disease Study 43. Young RF, Jacques S, Mark R, Kopyov O, Copcutt B, Posewitz A,
Group. Deep-brain stimulation of the subthalamic nucleus or the et al. Gamma knife thalamotomy for treatment of tremor: long-
pars interna of the globus pallidus in Parkinson’s disease. N Engl term results. J Neurosurg. 2000;93 Suppl 3:128–35; Epub
J Med. 2001;345(13):956–63; Epub 2001/09/29. 2001/01/06.
26. Hariz MI. Safety and risk of microelectrode recording in surgery 44. Kondziolka D, Ong JG, Lee JY, Moore RY, Flickinger JC,
for movement disorders. Stereotact Funct Neurosurg. 2002; Lunsford LD. Gamma Knife thalamotomy for essential tremor. J
78(3–4):146–57; Epub 2003/03/26. Neurosurg. 2008;108(1):111–7; Epub 2008/01/05.
27. Binder DK, Rau GM, Starr PA. Risk factors for hemorrhage dur- 45. Kondziolka D, Lunsford LD, Claassen D, Maitz AH, Flickinger
ing microelectrode-guided deep brain stimulator implantation for JC. Radiobiology of radiosurgery: Part I. The normal rat brain
movement disorders. Neurosurgery. 2005;56(4):722–32; discus- model. Neurosurgery. 1992;31(2):271–9; Epub 1992/08/01.
sion 32; Epub 2005/03/29. 46. Okun MS, Stover NP, Subramanian T, Gearing M, Wainer BH,
28. Piacentino M, Pilleri M, Bartolomei L. Hardware-related infec- Holder CA, et al. Complications of gamma knife surgery for
tions after deep brain stimulation surgery: review of incidence, Parkinson disease. Arch Neurol. 2001;58(12):1995–2002; Epub
severity and management in 212 single-center procedures in the 2001/12/26.
first year after implantation. Acta Neurochir. 2011;153(12):2337– 47. Rothstein TL. A late complication of gamma knife radiosurgery.
41; Epub 2011/08/30. Rev Neurol Dis. 2010;7(4):150–1; discussion 7–9; Epub 2011/01/06.
29. Oh MY, Abosch A, Kim SH, Lang AE, Lozano AM. Long-term hard- 48. Mathieu D, Kondziolka D, Niranjan A, Flickinger J, Lunsford
ware-related complications of deep brain stimulation. Neurosurgery. LD. Gamma knife thalamotomy for multiple sclerosis tremor.
2002;50(6):1268–74; discussion 74–6; Epub 2002/05/23. Surg Neurol. 2007;68(4):394–9; Epub 2007/10/02.
30. Boviatsis EJ, Stavrinou LC, Themistocleous M, Kouyialis AT, 49. Andersson B, Larsson B, Leksell L, Mair W, Rexed B, Sourander
Sakas DE. Surgical and hardware complications of deep brain P, et al. Histopathology of late local radiolesions in the goat brain.
stimulation. A seven-year experience and review of the literature. Acta Radiol Ther Phys Biol. 1970;9(5):385–94; Epub 1970/10/01.
Acta Neurochir. 2010;152(12):2053–62; Epub 2010/07/27. 50. Larsson B, Leksell L, Rexed B, Sourander P, Mair W, Andersson
31. Kondziolka D, Whiting D, Germanwala A, Oh M. Hardware- B. The high-energy proton beam as a neurosurgical tool. Nature.
related complications after placement of thalamic deep brain stim- 1958;182(4644):1222–3; Epub 1958/11/01.
ulator systems. Stereotact Funct Neurosurg. 2002;79(3–4):228–33; 51. Young RF, Li F, Vermeulen S, Meier R. Gamma Knife thalamot-
Epub 2003/08/02. omy for treatment of essential tremor: long-term results. J
32. Fenoy AJ, Simpson Jr RK. Management of device-related wound Neurosurg. 2010;112(6):1311–7; Epub 2009/11/10.
complications in deep brain stimulation surgery. J Neurosurg. 52. Friehs GM, Noren G, Ohye C, Duma CM, Marks R, Plombon J,
2012;116(6):1324–32; Epub 2012/03/13. et al. Lesion size following Gamma Knife treatment for functional
33. Frighetto L, De Salles A, Wallace R, Ford J, Selch M, Cabatan- disorders. Stereotact Funct Neurosurg. 1996;66 Suppl 1:320–8;
Awang C, et al. Linear accelerator thalamotomy. Surg Neurol. Epub 1996/01/01.
2004;62(2):106–13; discussion 13–4; Epub 2004/07/21. 53. Franzini A, Marchetti M, Brait L, Milanesi I, Messina G, Forapani
34. Smith ZA, De Salles AA, Frighetto L, Goss B, Lee SP, Selch M, E, et al. Deep brain stimulation and frameless stereotactic radio-
et al. Dedicated linear accelerator radiosurgery for the treatment surgery in the treatment of bilateral parkinsonian tremor: target
of trigeminal neuralgia. J Neurosurg. 2003;99(3):511–6; Epub selection and case report of two patients. Acta Neurochir.
2003/09/10. 2011;153(5):1069–75; Epub 2011/02/22.
35. Schuurman PR, Bosch DA, Bossuyt PM, Bonsel GJ, van Someren 54. Laitinen LV, Bergenheim AT, Hariz MI. Leksell’s posteroventral
EJ, de Bie RM, et al. A comparison of continuous thalamic stimu- pallidotomy in the treatment of Parkinson’s disease. J Neurosurg.
lation and thalamotomy for suppression of severe tremor. N Engl 1992;76(1):53–61; Epub 1992/01/01.
J Med. 2000;342(7):461–8; Epub 2000/02/17. 55. Lang AE, Lozano AM, Montgomery E, Duff J, Tasker R,
36. Lim SY, Hodaie M, Fallis M, Poon YY, Mazzella F, Moro Hutchinson W. Posteroventral medial pallidotomy in advanced
E. Gamma knife thalamotomy for disabling tremor: a blinded Parkinson’s disease. N Engl J Med. 1997;337(15):1036–42; Epub
evaluation. Arch Neurol. 2010;67(5):584–8; Epub 2010/05/12. 1997/10/09.
37. Ohye C, Shibazaki T, Sato S. Gamma knife thalamotomy for 56. Siegfried J, Lippitz B. Bilateral chronic electrostimulation of
movement disorders: evaluation of the thalamic lesion and clinical ventroposterolateral pallidum: a new therapeutic approach for
results. J Neurosurg. 2005;102(Suppl):234–40; Epub 2005/01/25. alleviating all parkinsonian symptoms. Neurosurgery. 1994;35(6):
38. Ohye C, Shibazaki T, Hirato M, Inoue H, Andou Y. Gamma thala- 1126–9; discussion 9–30; Epub 1994/12/01.
motomy for parkinsonian and other kinds of tremor. Stereotact 57. Cannon E, Silburn P, Coyne T, O’Maley K, Crawford JD, Sachdev
Funct Neurosurg. 1996;66 Suppl 1:333–42; Epub 1996/01/01. PS. Deep brain stimulation of anteromedial globus pallidus interna
39. Ohye C, Shibazaki T, Ishihara J, Zhang J. Evaluation of gamma for severe Tourette’s syndrome. Am J Psychiatry. 2012;169(8):860–
thalamotomy for parkinsonian and other tremors: survival of neu- 6; Epub 2012/07/10.
rons adjacent to the thalamic lesion after gamma thalamotomy. J 58. Dong S, Zhuang P, Zhang XH, Li JY, Li YJ. Unilateral deep brain
Neurosurg. 2000;93 Suppl 3:120–7; Epub 2001/01/06. stimulation of the right globus pallidus internus in patients with
40. Duma CM, Jacques D, Kopyov OV. The treatment of movement Tourette’s syndrome: two cases with outcomes after 1 year and a
disorders using Gamma Knife stereotactic radiosurgery. Neurosurg brief review of the literature. J Int Med Res. 2012;40(5):2021–8;
Clin N Am. 1999;10(2):379–89; Epub 1999/03/31. Epub 2012/12/05.
59. Burchiel KJ, Anderson VC, Favre J, Hammerstad JP. Comparison disease: neurobehavioral functioning before and 3 months after
of pallidal and subthalamic nucleus deep brain stimulation for electrode implantation. Neurology. 1997;49(4):1078–83; Epub
advanced Parkinson’s disease: results of a randomized, blinded 1997/10/27.
pilot study. Neurosurgery. 1999;45(6):1375–82; discussion 82–4; 76. Odekerken VJ, van Laar T, Staal MJ, Mosch A, Hoffmann CF,
Epub 1999/12/22. Nijssen PC, et al. Subthalamic nucleus versus globus pallidus
60. Durif F, Lemaire JJ, Debilly B, Dordain G. Long-term follow-up bilateral deep brain stimulation for advanced Parkinson’s disease
of globus pallidus chronic stimulation in advanced Parkinson’s (NSTAPS study): a randomised controlled trial. Lancet Neurol.
disease. Mov Disord. 2002;17(4):803–7; Epub 2002/09/05. 2013;12(1):37–44; Epub 2012/11/22.
61. Ghika J, Villemure JG, Fankhauser H, Favre J, Assal G, Ghika- 77. Witt K, Daniels C, Reiff J, Krack P, Volkmann J, Pinsker MO,
Schmid F. Efficiency and safety of bilateral contemporaneous pal- et al. Neuropsychological and psychiatric changes after deep brain
lidal stimulation (deep brain stimulation) in levodopa-responsive stimulation for Parkinson’s disease: a randomised, multicentre
patients with Parkinson’s disease with severe motor fluctuations: a study. Lancet Neurol. 2008;7(7):605–14; Epub 2008/06/10.
2-year follow-up review. J Neurosurg. 1998;89(5):713–8; Epub 78. Saez-Zea C, Escamilla-Sevilla F, Katati MJ, Minguez-Castellanos
1998/11/17. A. Cognitive effects of subthalamic nucleus stimulation in
62. Gross C, Rougier A, Guehl D, Boraud T, Julien J, Bioulac B. High- Parkinson’s disease: a controlled study. Eur Neurol. 2012;68(6):
frequency stimulation of the globus pallidus internalis in 361–6; Epub 2012/10/26.
Parkinson’s disease: a study of seven cases. J Neurosurg. 1997; 79. Baron MS, Vitek JL, Bakay RA, Green J, Kaneoke Y, Hashimoto
87(4):491–8; Epub 1997/10/10. T, et al. Treatment of advanced Parkinson’s disease by posterior
63. Krack P, Pollak P, Limousin P, Hoffmann D, Benazzouz A, Le Bas GPi pallidotomy: 1-year results of a pilot study. Ann Neurol.
JF, et al. Opposite motor effects of pallidal stimulation in 1996;40(3):355–66; Epub 1996/09/01.
Parkinson’s disease. Ann Neurol. 1998;43(2):180–92; Epub 80. Baron MS, Vitek JL, Bakay RA, Green J, McDonald WM, Cole
1998/03/04. SA, et al. Treatment of advanced Parkinson’s disease by unilateral
64. Kumar R, Lozano AM, Montgomery E, Lang AE. Pallidotomy posterior GPi pallidotomy: 4-year results of a pilot study. Mov
and deep brain stimulation of the pallidum and subthalamic Disord. 2000;15(2):230–7; Epub 2000/04/07.
nucleus in advanced Parkinson’s disease. Mov Disord. 1998;13 81. Iacono RP, Shima F, Lonser RR, Kuniyoshi S, Maeda G, Yamada
Suppl 1:73–82; Epub 1998/06/05. S. The results, indications, and physiology of posteroventral palli-
65. Merello M, Nouzeilles MI, Kuzis G, Cammarota A, Sabe L, Betti dotomy for patients with Parkinson’s disease. Neurosurgery.
O, et al. Unilateral radiofrequency lesion versus electrostimula- 1995;36(6):1118–25; discussion 25–7; Epub 1995/06/01.
tion of posteroventral pallidum: a prospective randomized com- 82. Johansson F, Malm J, Nordh E, Hariz M. Usefulness of pallidot-
parison. Mov Disord. 1999;14(1):50–6; Epub 1999/01/26. omy in advanced Parkinson’s disease. J Neurol Neurosurg
66. Pahwa R, Wilkinson S, Smith D, Lyons K, Miyawaki E, Koller Psychiatry. 1997;62(2):125–32; Epub 1997/02/01.
WC. High-frequency stimulation of the globus pallidus for the 83. Kishore A, Turnbull IM, Snow BJ, de la Fuente-Fernandez R,
treatment of Parkinson’s disease. Neurology. 1997;49(1):249–53; Schulzer M, Mak E, et al. Efficacy, stability and predictors of out-
Epub 1997/07/01. come of pallidotomy for Parkinson’s disease. Six-month follow-
67. Tronnier VM, Fogel W, Kronenbuerger M, Steinvorth S. Pallidal up with additional 1-year observations. Brain. 1997;120(Pt 5):
stimulation: an alternative to pallidotomy? J Neurosurg. 1997;87(5): 729–37; Epub 1997/05/01.
700–5; Epub 1997/11/05. 84. Kondziolka D, Bonaroti E, Baser S, Brandt F, Kim YS, Lunsford
68. Volkmann J, Sturm V, Weiss P, Kappler J, Voges J, Koulousakis A, LD. Outcomes after stereotactically guided pallidotomy for
et al. Bilateral high-frequency stimulation of the internal globus advanced Parkinson’s disease. J Neurosurg. 1999;90(2):197–202;
pallidus in advanced Parkinson’s disease. Ann Neurol. Epub 1999/02/09.
1998;44(6):953–61; Epub 1998/12/16. 85. Masterman D, DeSalles A, Baloh RW, Frysinger R, Foti D,
69. Loher TJ, Burgunder JM, Pohle T, Weber S, Sommerhalder R, Behnke E, et al. Motor, cognitive, and behavioral performance fol-
Krauss JK. Long-term pallidal deep brain stimulation in patients lowing unilateral ventroposterior pallidotomy for Parkinson dis-
with advanced Parkinson disease: 1-year follow-up study. J ease. Arch Neurol. 1998;55(9):1201–8; Epub 1998/09/18.
Neurosurg. 2002;96(5):844–53; Epub 2002/05/15. 86. Vitek JL, Bakay RA, DeLong MR. Microelectrode-guided palli-
70. Adler CH, Kumar R. Pharmacological and surgical options for the dotomy for medically intractable Parkinson’s disease. Adv Neurol.
treatment of cervical dystonia. Neurology. 2000;55(12 Suppl 1997;74:183–98; Epub 1997/01/01.
5):S9–14; Epub 2001/02/24. 87. Young RF, Vermeulen S, Posewitz A, Shumway-Cook
71. Galvez-Jimenez N, Lozano A, Tasker R, Duff J, Hutchison W, A. Pallidotomy with the gamma knife: a positive experience.
Lang AE. Pallidal stimulation in Parkinson’s disease patients with Stereotact Funct Neurosurg. 1998;70 Suppl 1:218–28; Epub
a prior unilateral pallidotomy. Can J Neurol Sci. 1998;25(4):300– 1998/10/23.
5; Epub 1998/11/25. 88. Friedman JH, Epstein M, Sanes JN, Lieberman P, Cullen K,
72. Coubes P, Cif L, El Fertit H, Hemm S, Vayssiere N, Serrat S, et al. Lindquist C, et al. Gamma knife pallidotomy in advanced
Electrical stimulation of the globus pallidus internus in patients Parkinson’s disease. Ann Neurol. 1996;39(4):535–8; Epub
with primary generalized dystonia: long-term results. J Neurosurg. 1996/04/01.
2004;101(2):189–94; Epub 2004/08/18. 89. Friedman DP, Goldman HW, Flanders AE, Gollomp SM, Curran
73. Coubes P, Roubertie A, Vayssiere N, Hemm S, Echenne Jr WJ. Stereotactic radiosurgical pallidotomy and thalamotomy
B. Treatment of DYT1-generalised dystonia by stimulation of the with the gamma knife: MR imaging findings with clinical correla-
internal globus pallidus. Lancet. 2000;355(9222):2220–1; Epub tion–preliminary experience. Radiology. 1999;212(1):143–50;
2000/07/06. Epub 1999/07/16.
74. Vingerhoets G, van der Linden C, Lannoo E, Vandewalle V, 90. Friedman JH, Fernandez HH, Sikirica M, Stopa E, Friehs
Caemaert J, Wolters M, et al. Cognitive outcome after unilateral G. Stroke induced by gamma knife pallidotomy: autopsy result.
pallidal stimulation in Parkinson’s disease. J Neurol Neurosurg Neurology. 2002;58(11):1695–7; Epub 2002/06/12.
Psychiatry. 1999;66(3):297–304; Epub 1999/03/20. 91. Kwon Y, Whang CJ. Stereotactic Gamma Knife radiosurgery for
75. Troster AI, Fields JA, Wilkinson SB, Pahwa R, Miyawaki E, the treatment of dystonia. Stereotact Funct Neurosurg. 1995;64
Lyons KE, et al. Unilateral pallidal stimulation for Parkinson’s Suppl 1:222–7; Epub 1995/01/01.
694 Z. Zibly et al.
92. Benabid AL, Pollak P, Gross C, Hoffmann D, Benazzouz A, Gao brain stimulation of the subthalamic nucleus. J Neurol.
DM, et al. Acute and long-term effects of subthalamic nucleus 1999;246(10):907–13; Epub 1999/11/07.
stimulation in Parkinson’s disease. Stereotact Funct Neurosurg. 106. Kumar R, Lozano AM, Sime E, Halket E, Lang AE. Comparative
1994;62(1–4):76–84; Epub 1994/01/01. effects of unilateral and bilateral subthalamic nucleus deep brain
93. Limousin P, Krack P, Pollak P, Benazzouz A, Ardouin C, Hoffmann stimulation. Neurology. 1999;53(3):561–6; Epub 1999/08/17.
D, et al. Electrical stimulation of the subthalamic nucleus in 107. Limousin P, Pollak P, Hoffmann D, Benazzouz A, Perret JE,
advanced Parkinson’s disease. N Engl J Med. 1998;339(16):1105– Benabid AL. Abnormal involuntary movements induced by sub-
11; Epub 1998/10/15. thalamic nucleus stimulation in parkinsonian patients. Mov
94. Kumar R, Lozano AM, Kim YJ, Hutchison WD, Sime E, Halket E, Disord. 1996;11(3):231–5; Epub 1996/05/01.
et al. Double-blind evaluation of subthalamic nucleus deep brain 108. Volkmann J, Allert N, Voges J, Weiss PH, Freund HJ, Sturm V.
stimulation in advanced Parkinson’s disease. Neurology. Safety and efficacy of pallidal or subthalamic nucleus stimula-
1998;51(3):850–5; Epub 1998/09/25. tion in advanced PD. Neurology. 2001;56(4):548–51; Epub
95. Bejjani BP, Gervais D, Arnulf I, Papadopoulos S, Demeret S, 2001/02/27.
Bonnet AM, et al. Axial parkinsonian symptoms can be 109. Sellal F, Hirsch E, Lisovoski F, Mutschler V, Collard M, Marescaux
improved: the role of levodopa and bilateral subthalamic stimula- C. Contralateral disappearance of parkinsonian signs after
tion. J Neurol Neurosurg Psychiatry. 2000;68(5):595–600; Epub subthalamic hematoma. Neurology. 1992;42(1):255–6; Epub
2000/04/15. 1992/01/01.
96. Charles PD, Van Blercom N, Krack P, Lee SL, Xie J, Besson G, 110. Vidakovic A, Dragasevic N, Kostic VS. Hemiballism: report of 25
et al. Predictors of effective bilateral subthalamic nucleus stimula- cases. J Neurol Neurosurg Psychiatry. 1994;57(8):945–9; Epub
tion for PD. Neurology. 2002;59(6):932–4; Epub 2002/09/26. 1994/08/01.
97. Deuschl G, Fogel W, Hahne M, Kupsch A, Muller D, Oechsner M, 111. Alvarez L, Macias R, Lopez G, Alvarez E, Pavon N, Rodriguez-
et al. Deep-brain stimulation for Parkinson’s disease. J Neurol. Oroz MC, et al. Bilateral subthalamotomy in Parkinson’s disease:
2002;249(Suppl 3:III):36–9; Epub 2003/01/11. initial and long-term response. Brain. 2005;128(Pt 3):570–83;
98. Kleiner-Fisman G, Fisman DN, Sime E, Saint-Cyr JA, Lozano Epub 2005/02/04.
AM, Lang AE. Long-term follow up of bilateral deep brain stimu- 112. Gill SS, Heywood P. Bilateral dorsolateral subthalamotomy for
lation of the subthalamic nucleus in patients with advanced advanced Parkinson’s disease. Lancet. 1997;350(9086):1224; Epub
Parkinson disease. J Neurosurg. 2003;99(3):489–95; Epub 1998/07/04.
2003/09/10. 113. Keep MF, Mastrofrancesco L, Erdman D, Murphy B, Ashby
99. Pahwa R, Wilkinson SB, Overman J, Lyons KE. Bilateral subtha- LS. Gamma knife subthalamotomy for Parkinson disease: the sub-
lamic stimulation in patients with Parkinson disease: long-term thalamic nucleus as a new radiosurgical target. Case report. J
follow up. J Neurosurg. 2003;99(1):71–7; Epub 2003/07/12. Neurosurg. 2002;97(5 Suppl):592–9; Epub 2003/01/01.
100. Simuni T, Jaggi JL, Mulholland H, Hurtig HI, Colcher A, Siderowf 114. Martin AJ, Larson PS, Ostrem JL, Keith Sootsman W, Talke P,
AD, et al. Bilateral stimulation of the subthalamic nucleus in Weber OM, et al. Placement of deep brain stimulator electrodes
patients with Parkinson disease: a study of efficacy and safety. J using real-time high-field interventional magnetic resonance
Neurosurg. 2002;96(4):666–72; Epub 2002/05/07. imaging. Magn Reson Med. 2005;54(5):1107–14; Epub
101. Vingerhoets FJ, Villemure JG, Temperli P, Pollo C, Pralong E, 2005/10/06.
Ghika J. Subthalamic DBS replaces levodopa in Parkinson’s dis- 115. Jagannathan J, Sanghvi NT, Crum LA, Yen CP, Medel R,
ease: two-year follow-up. Neurology. 2002;58(3):396–401; Epub Dumont AS, et al. High-intensity focused ultrasound surgery of
2002/02/13. the brain: part 1–A historical perspective with modern applica-
102. Welter ML, Houeto JL, Tezenas du Montcel S, Mesnage V, Bonnet tions. Neurosurgery. 2009;64(2):201–10; discussion 10–1; Epub
AM, Pillon B, et al. Clinical predictive factors of subthalamic 2009/02/05.
stimulation in Parkinson’s disease. Brain. 2002;125(Pt 3):575–83; 116. Bradley Jr WG. MR-guided focused ultrasound: a potentially dis-
Epub 2002/03/02. ruptive technology. J Am Coll Radiol. 2009;6(7):510–3; Epub
103. Krack P, Pollak P, Limousin P, Hoffmann D, Xie J, Benazzouz A, 2009/06/30.
et al. Subthalamic nucleus or internal pallidal stimulation in young 117. Cohen ZR, Zaubermann J, Harnof S, Mardor Y, Nass D, Zadicario
onset Parkinson’s disease. Brain. 1998;121(Pt 3):451–7; Epub E, et al. Magnetic resonance imaging-guided focused ultrasound
1998/04/29. for thermal ablation in the brain: a feasibility study in a swine
104. Krause M, Fogel W, Heck A, Hacke W, Bonsanto M, model. Neurosurgery. 2007;60(4):593–600; discussion 600; Epub
Trenkwalder C, et al. Deep brain stimulation for the treatment 2007/04/07.
of Parkinson’s disease: subthalamic nucleus versus globus palli- 118. Tyler WJ, Tufail Y, Pati S. Pain: Noninvasive functional neurosur-
dus internus. J Neurol Neurosurg Psychiatry. 2001;70(4):464–70; gery using ultrasound. Nat Rev Neurol. 2010;6(1):13–4; Epub
Epub 2001/03/20. 2010/01/09.
105. Pinter MM, Alesch F, Murg M, Helscher RJ, Binder 119. Jolesz FA. MRI-guided focused ultrasound surgery. Annu Rev
H. Apomorphine test: a predictor for motor responsiveness to deep Med. 2009;60:417–30; Epub 2009/07/28.
Movement Disorders:
Viewpoint—Medical Therapy 57
Abraham N. Lieberman and Sara S. Dhanani
neuroleptics (dystonia, parkinsonism), and valproate (tremor).

Movement Disorders: Viewpoint—Medical The term tardive dyskinesia is used to describe drug-induced
Therapy movement disorders. Tardive can signify a delay in the appear-
ance of the movement disorder after the drug has begun, or
Movement disorders are a diverse group of diseases that delay in the appearance of the movement disorder after the
constitute either an excess of movement or a paucity of drug has been discontinued. The focus of this chapter will be
movement. The former are known as hyperkinetic and the on the two most common disorders: PD and ET.
latter as hypokinetic disorders. Parkinson disease (PD), next
to essential tremor (ET), is the most common movement dis-
order, and unlike ET is more debilitating. PD encompasses Parkinson Disease
both hypokinesia (slowness and paucity of movement) and
hyperkinesia (tremor). In addition many PD patients on PD is a disorder characterized by four cardinal features that
levodopa treatment develop dyskinesias, a hyperkinetic include tremor at rest, bradykinesia (a slowness and paucity
movement disorder, a combination of chorea (dance-like of movement), rigidity, and postural instability. Other fea-
involuntary movements) and dystonia (altered tone). tures that at times can be as troubling as the cardinal features
Parkinson disease shares many features with a number of include mental changes ranging from minimal cognitive
Parkinson-plus or Parkinson-like disorders. These disorders impairment to dementia, depression, apathy, difficulty walk-
usually do not respond to treatment with levodopa, usually ing, freezing of gait, and autonomic insufficiency. Additional
run more rapid courses, and rarely have hyperkinetic features. features may include a postural kinetic tremor, speech and
These disorders have pathology different from PD; they swallowing difficulties, sialorrhea, and abnormal flexed pos-
include multiple system atrophy (MSA), progressive supra- tures of the hands, the legs, and the trunk. Although tremor is
nuclear palsy (PSP), and corticobasilar degeneration (CBD). strongly associated with PD, 30 % of PD patients have no
Essential tremor (ET) is the most common movement tremor.
disorder. Although troublesome, it is usually not debilitating. The disorder was first described in 1817 by James
Among the more prominent hyperkinetic movement disorders Parkinson who wrote An Essay on the Shaking Palsy [1]. PD
are Huntington disease and Tourette’s syndrome. Hyperkinetic usually begins asymmetrically with slowness of movement of
disorders include movements such as athetosis, ballism, cho- an arm or leg and/or a rest tremor (of the hand). PD is a rarity
rea, myoclonus, tics, and tremor. These can occur as isolated among movement disorders in that it usually appears asym-
phenomena or as the results of strokes or in association with metrically. Corticobasilar degeneration (CBD), a disorder
liver, kidney, endocrine, or metabolic disorders. They can also that occurs 1/100th as often as PD, also presents asymmetri-
occur as the result of drugs including amiodarone (tremor), cally. In PD, over several years the disorder progresses to
lithium (tremor), metoclopramide (dystonia, parkinsonism), involve the contralateral side and trunk [2]. If a disorder diag-
nosed as PD presents with symmetrical signs and progresses
A.N. Lieberman, M.D. (*) rapidly, the evolving disorder which may have initially been
Department of Neurology, Barrow Neurological Institute, diagnosed as PD will likely turn out to be one of PD-plus
St. Joseph’s Hospital and Medical Center, Phoenix, AZ, USA
disorders. This is especially so if postural instability with
S.S. Dhanani, M.D. falls appears early in the evolution of the disorder [3].
Department of Neurology, Banner Sunreach Research Institute,
Sun City, AZ, USA Several studies have identified non-motor features of
e-mail: sara.dhanani@dignityhealth.com PD, features that may precede the diagnosis of PD [4].
696 A.N. Lieberman and S.S. Dhanani
These include loss of olfaction, a rapid eye movement (REM) that early treatment may provide benefits that may not be
sleep behavior disorder (RBD), constipation, autonomic achieved if the drugs are introduced later [10].
insufficiency, apathy, depression, and mild cognitive impair- Monoamine oxidase (MAO) type B inhibitors, such as
ment [5]. Loss of olfaction may be one of the first signs Rasagiline or Selegiline, are effective as monotherapy in
of PD, appreciated in retrospect. Loss of olfaction is not patients with early PD and as adjunctive therapy to dopamine
improved or worsened by anti-PD drugs [6]. Although the agonists or levodopa in patients with more advanced PD [11].
cause of PD is unknown, it has been speculated, by some, a Monoamine oxidase B inhibitors, unlike MAO-A inhibitors
prion, utilizes the olfactory pathway to enter the brain and (drugs used in psychiatry), do not result in elevations of tyra-
cause PD. While RBD is not a specific marker for PD, it is mine, an amine found in certain cured meats, red wines, and
present in a higher frequency in people who develop PD. The cheese. Unlike with MAO-A inhibitors, dietary restrictions
locus ceruleus, a noradrenergic nucleus in the brainstem, a are not necessary with MAO-B inhibitors provided the rec-
nucleus affected in PD, may also be affected in RBD. The ommended doses are used. There is a question as to whether
progression of PD from caudal sites such as the dorsal vagal Rasagiline, through mechanisms unrelated to monoamine
nucleus in the medulla to the locus ceruleus to the substantia oxidase B inhibition, may slow the progression of PD [11].
nigra suggests to some, a causative process that progresses in Approximately 50 % of patients improve mildly or moder-
a caudal rostral direction [7]. ately. Improvement continues throughout the illness.
The pathology in PD is characterized by progressive Other drugs used in PD, drugs that can be used as initial
degeneration of the dopaminergic nigrostriatal system and treatment, include amantadine and the dopamine agonists.
also by degeneration of other neuronal systems including Amantadine, originally introduced as a drug to prevent the
involvement of acetylcholine neurons in the nucleus basalis flu, both enhances dopaminergic transmission, resulting in
of Meynert, serotonin neurons in the dorsal raphe nucleus, an anti-PD effect, and blocks NMDA glutamate receptor.
and noradrenalin neurons in the locus ceruleus as well as The anti-glutaminergic affect may account for the ability of
neurons in the cortex, the spinal cord, and the central and amantadine to lessen the dyskinesias that often complicate
peripheral autonomic nervous system [8]. Recent studies long-term levodopa treatment [12]. Dopamine agonists,
utilizing positron emission tomography (PET scans) and drugs that mimic the effects of dopamine in the brain, may be
single photon emission tomography (SPECT scans) suggest used as initial treatment and as adjuncts to levodopa.
that despite the presence of PD for many years, degeneration Dopamine agonists, including Ropinirole, Pramipexole, and
of the dopamine neurons and the loss of dopamine in the Rotigotine (administered as a skin patch), are longer acting
putamen and caudate nucleus are incomplete [9]. than levodopa. In early PD they have an anti-PD less than
The loss of dopamine neurons in the nigra and the loss of levodopa but they do not result in “wearing off” or “on off”
dopamine in the putamen may account for the hypokinetic and they do not result in dyskinesias. Used in combination
component of PD. These symptoms are helped by levodopa with levodopa they may allow for a reduction in the dose of
(which replaces the depleted dopamine). An impairment of levodopa, a lessening of the “wearing off” and “on off effect”
circuits including tracts from the motor cortex to the putamen and a decrease in dyskinesias [13]. Approximately 50–75 %
then to the globus pallidus, then to the subthalamic nucleus, of patients with early PD improve. Improvement maintained
then to the thalamus, and back to the supplementary motor throughout most of the course of the disease. Adverse effects
cortex may account for the hyperkinetic components of PD. associated with the dopamine agonists include hallucinations
The manifestations of PD range from a barely perceptible (which also occur with levodopa), peripheral edema, hyper-
rest tremor to significant dysfunction resulting in freezing of somnolence, and impulse control disorders including patho-
gait with falls, loss of postural stability, falls, and cognitive logic gambling, hypersexuality, and hyperphagia. Although
decline progressing to dementia. Most of the motor manifes- impulse control disorders also occur with levodopa, they are
tations of PD are related to the deficiency of dopamine in the more prominent with the dopamine agonists.
striatum. A major advantage in identifying a prodromal PD Levodopa remains the single most effective drug for the
state would be the opportunity to introduce therapy at an relief of PD symptoms. Levodopa is combined with carbi-
earlier time point than is currently possible with, perhaps, a dopa, a peripherally acting dopa-decarboxylase inhibitor, a
delay in the progression of some symptoms. drug that blocks the nausea associated with levodopa and
Therapy in early PD is guided by the effect of symptoms allows more levodopa to enter the brain. Carbidopa is usually
on function and quality of life, consideration of complica- combined with levodopa in a 1:4 ratio, carbidopa/levodopa
tions associated with the long-term use of drugs, particularly 25 mg/100 mg, or sometimes a 1:10 ratio. Long-term use of
levodopa and the likelihood of the development of response levodopa may be associated with complications including
fluctuations such as “wearing off” (gradual loss of efficacy of “wearing off,” “on off,” and dyskinesias. In patients with
a dose of levodopa), “on off” (abrupt loss of efficacy of a cognitive impairment levodopa, like the dopamine agonists
dose of levodopa), and dyskinesias. Several studies suggest and like amantadine, may be associated with delusions,
57 Movement Disorders: Viewpoint—Medical Therapy 697
hallucinations, and paranoia. As PD advances drug treatment Other symptoms that may rarely complicate ET include
becomes less satisfactory. In part this is related to progres- ataxia and postural instability. Recent studies have revealed
sion of the underlying disease with involvement of non- the presence of abnormalities in the cerebellum (including
dopaminergic systems and in part to the complications of the loss of Purkinje neurons) and the locus ceruleus. The func-
drugs [14, 15]. All patients improve on levodopa which tion of the Purkinje neurons is to inhibit the outflow nuclei of
failed to improve usually. It indicates the disorder is not PD the cerebellum that in turn stimulate the thalamus. Loss of
but a PD-plus disorder, such as progressive supranuclear Purkinje neurons result in less inhibition of the cerebellar
palsy or multiple system atrophy. Improvement is main- outflow nuclei with increased stimulation of the thalamus—
tained but as the disease progresses higher doses are needed. resulting in a hyperkinetic movement disorder: tremor. DBS
The introduction of deep brain stimulation (DBS) has of the thalamus (presumably with inhibition) is treatment for
altered the treatment of PD. DBS of selected nuclei, unilater- refractory ET. In addition to the presence of motor signs,
ally or bilaterally, may improve tremor, lessen dyskinesia, and there may be, in a few patients, cognitive impairment and
lessen “wearing off.” Targets of DBS include the subthalamic depression [18–20].
nucleus, the globus pallidum interna, and the thalamus. The decision to treat patients with ET is based primarily
Properly choosing patients for DBS and selecting the appro- on the functional impact of the tremor. The most commonly
priate targets is an ongoing work in evolution. used, and generally most effective, medications for essential
tremor are propranolol (a beta-blocker) and primidone (an
anticonvulsant), administered either as monotherapy or in
Essential Tremor combination. In addition, there are other drugs that can be
used to treat tremor in patients not responsive to propranolol
Essential tremor (ET) is the most common movement or primidone. These medications include alprazolam (a ben-
disorder. It’s estimated that 5.0 % of the population has zodiazepine), atenolol (a beta-blocker), gabapentin (an anti-
ET. However in only about 5.0 % is it of sufficient concern to convulsant), topiramate (an anticonvulsant), and zonisamide
warrant a consultation with a neurologist. The tremor of ET, (an anticonvulsant) [21]. Some hand tremors that are diag-
unlike the tremor of PD, begins bilaterally. Unlike the tremor nosed as ET are really dystonic tremors and may respond to
of PD, it usually involves the index, middle, ring, and little botulinum toxin injections [22]. In those patients who have
finger before it involves the thumb. The tremor is present on marked tremors that results in disability or prevent them
sustaining or maintaining a posture (sustention or postural from working and who do not respond to medical therapy,
tremor) or on movement (action or kinetic tremor). Rarely the surgical options are available. The benzodiazepines, beta-
tremor of ET may be present at rest. In ET the rest tremor is blockers, and anticonvulsants are beneficial to a moderate
usually a late manifestation, whereas in PD it is an early man- degree in approximately 40 % of patients. DBS is 95–100 %
ifestation. The tremor of ET can involve the head. It uncom- successful in alleviating tremors. The most common proce-
monly involves the legs. A tremor that begins in one hand and dure performed is DBS targeting the ventralis intermedius
spreads to the ipsilateral leg is usually PD. There is often a (VIM) nucleus of the thalamus [23].
family history in ET. A family history is less common in PD.
Initially it may be difficult, on the basis of a single obser-
vation, to distinguish the postural and/or kinetic tremor of References
ET from a postural or kinetic tremor of PD. Usually within
two years the differences become obvious as the patient with 1. Parkinson J. An essay on the shaking palsy. J Neuropsychiatry Clin
PD develops, in addition to the tremor other symptoms of Neurosci. 2002;14(2):223–36.
PD. A specialized SPECT scan called a DAT scan that labels 2. Uitti RJ, Baba Y, Whaley NR, Wszolek ZK, Putzke JD. Parkinson
disease: Handedness predicts asymmetry. Neurology. 2005;64(11):
the dopamine transporter protein in the caudate and putamen
1925–30.
can distinguish ET from PD with high sensitivity and speci- 3. Gomez-Esteban JC, Tijero B, Ciordia R, Berganzo K, Somme J,
ficity. In ET dopamine is present in the caudate and putamen; Lezcano E, et al. Factors influencing the symmetry of Parkinson’s
in PD, dopamine is depleted in the putamen. These patterns disease symptoms. Clin Neurol Neurosurg. 2010;112(4):302–5.
4. Siderowf A, Lang AE. Premotor Parkinson’s disease: concepts and
are readily recognized on the DAT scan.
definitions. Mov Disord. 2012;27(5):608–16.
The tremor of ET is usually slowly progressive but over 5. Postuma RB, Aarsland D, Barone P, Burn DJ, Hawkes CH, Oertel
the years many patients may experience difficulties perform- W, et al. Identifying prodromal Parkinson’s disease: pre-motor dis-
ing basic daily activities [16], thereby impairing their quality orders in Parkinson’s disease. Mov Disord. 2012;27(5):617–26.
6. Doty RL, Stern MB, Pfeiffer C, Gollomp SM, Hurtig HI. Bilateral
of life. In addition, although postural and kinetic tremors are
olfactory dysfunction in early stage treated and untreated idiopathic
the main features of ET, intention tremors, which impair Parkinson’s disease. J Neurol Neurosurg Psychiatry. 1992;55(2):
activities of daily living, may also occur in some patients [17]. 138–42.
698 A.N. Lieberman and S.S. Dhanani
7. Iranzo A, Molinuevo JL, Santamaria J, Serradell M, Marti MJ, 15. Bayulkem K, Lopez G. Clinical approach to nonmotor sensory fluc-
Valldeoriola F, et al. Rapid-eye-movement sleep behaviour disorder tuations in Parkinson’s disease. J Neurol Sci. 2011;310(1–2):82–5.
as an early marker for a neurodegenerative disorder: a descriptive 16. Busenbark KL, Nash J, Nash S, Hubble JP, Koller WC. Is essential
study. Lancet Neurol. 2006;5(7):572–7. tremor benign? Neurology. 1991;41(12):1982–3.
8. Jellinger KA. Neuropathology of sporadic Parkinson’s disease: 17. Teive HA. Essential tremor: phenotypes. Parkinsonism Relat
evaluation and changes of concepts. Mov Disord. 2012;27(1): Disord. 2012;18 Suppl 1:S140–2 [Review].
8–30. 18. Troster AI, Woods SP, Fields JA, Lyons KE, Pahwa R, Higginson
9. Djaldetti R, Lorberboym M, Karmon Y, Treves TA, Ziv I, Melamed CI, et al. Neuropsychological deficits in essential tremor: an expres-
E. Residual striatal dopaminergic nerve terminals in very long- sion of cerebello-thalamo-cortical pathophysiology? Eur J Neurol.
standing Parkinson’s disease: a single photon emission computed 2002;9(2):143–51.
tomography imaging study. Mov Disord. 2011;26(2):327–30. 19. Louis ED, Benito-Leon J, Bermejo-Pareja F. Self-reported depression
10. Parkinson Study Group. A controlled, randomized, delayed-start and anti-depressant medication use in essential tremor: cross-sectional
study of rasagiline in early Parkinson disease. Arch Neurol. 2004; and prospective analyses in a population-based study. Eur J Neurol.
61(4):561–6. 2007;14(10):1138–46.
11. Olanow CW, Rascol O, Hauser R, Feigin PD, Jankovic J, Lang A, 20. Lorenz D, Schwieger D, Moises H, Deuschl G. Quality of life and per-
et al. A double-blind, delayed-start trial of rasagiline in Parkinson’s sonality in essential tremor patients. Mov Disord. 2006;21(8):1114–8.
disease. N Engl J Med. 2009;361(13):1268–78. 21. Lyons KE, Pahwa R. Pharmacotherapy of essential tremor : an over-
12. Lees A. Alternatives to levodopa in the initial treatment of early view of existing and upcoming agents. CNS Drugs. 2008; 22(12):
Parkinson’s disease. Drugs Aging. 2005;22(9):731–40. 1037–45.
13. Jankovic J, Stacy M. Medical management of levodopa-associated 22. Brin MF, Lyons KE, Doucette J, Adler CH, Caviness JN, Comella CL,
motor complications in patients with Parkinson’s disease. CNS et al. A randomized, double masked, controlled trial of botulinum toxin
Drugs. 2007;21(8):677–92. type A in essential hand tremor. Neurology. 2001; 56(11):1523–8.
14. Pahwa R, Lyons KE. Levodopa-related wearing-off in Parkinson’s 23. Pahwa R, Lyons KE, Wilkinson SB, Simpson Jr RK, Ondo WG,
disease: identification and management. Curr Med Res Opin. 2009; Tarsy D, et al. Long-term evaluation of deep brain stimulation of
25(4):841–9. the thalamus. J Neurosurg. 2006;104(4):506–12.
Radiosurgery for Drug-Resistant
Epilepsies: State of the Art, Results, 58
and Perspectives
Jean Régis, Romain Carron, Fabrice Bartolomei,

and Patrick Chauvel
dose of 50 Gy using the GK. Biochemical analyses

Rationale demonstrated a stability of the level of the GABA Acetyl
decarboxylase in spite of the dramatic decrease of the level
Radiosurgery is a neurosurgical technique where energy is of the CAT indicating an injury of the catecholaminergic sys-
delivered in a single session to a small sharply limited target tem in spite of the sparing of the GABAergic system.
with a stereotactic accuracy, with the aim either to create a Similarly direct dosage of the GABA itself has shown its sta-
lesion or to induce a desired biological effect [1, 2]. Since the bility in spite of the major decrease of the amino excitatory
first attempt of the pioneers in Stockholm, the practice of acids (glutamate and aspartate). This experimental demon-
radiosurgery has dramatically changed with, nowadays in the stration led us to consider radiosurgery as a neuromodulation
vast majority of the indications, the use of nondestructive therapy [1–3] and encouraged us to organize several prospec-
low dosage inducing subtle biological effect like apoptoses tive clinical trials [4–8]. More recently the demonstration of
in tumors or endothelial proliferation in AVM. the existence of a differential effect at the cellular level came
from the Charlottesville group [9]. In epileptic rats irradiated
with 40 Gy in the temporal lobe using the GK immunohisto-
The Differential Effect Concept chemical study suggested that at least one subtype of hippo-
campal interneurons is selectively vulnerable to Gamma
This is a classical clinical observation that in 85 % of the Knife (GK) radiosurgery. Neuronal cells appear to have
AVM associated to a resistant epilepsy radiosurgery is fol- undergone a phenotypic shift with respect to calbindin and
lowed by seizure cessation or dramatic improvement occur- GAD-67 expression. There is a growing body of evidence
ring much before the occlusion of the AVM itself. When the nowadays in favor of a neuromodulatory effect [1, 2, 10].
AVM is located in a highly functional area, the seizure cessa- A series of successive clinical trials have been organized
tion is obtained without clinical deficit. This observation has in Marseille in order to evaluate Gamma Knife surgery in
led us in 1992 to propose the concept of clinical differential epilepsy. In 1993 we organized a phase II prospective trial in
effect: radiosurgery can induce a functional effect as making four MTLE patients with a goal of dose ranging and toxicity
the cortex surrounding the AVM no longer epileptic without evaluation [6, 8]. In 1995, the good safety and the impressive
destroying the underlying function of this cortical area thanks efficacy in the patient receiving the 24 Gy dosage led us to
to its capacity to alter specifically some systems while spar- organize a phase III prospective single center study in four
ing others. The first proof of concept came from the demon- mesial temporal lobe epilepsy (MTLE) patients (24 Gy
stration of the existence of such an effect at the biochemical 7–8 cc) in order to evaluate the reproducibility of the efficacy
level in the striatum of rats [3]. A group of rats received a [4]. In 1996, we organized a prospective multicentric
single isocenter of 4 mm in the left striatum with a maximum European study (21 MTLE patients) confirming the repro-
ducibility of the safety and efficacy [7]. In 1998 a dose de-
escalation study (24, 20, 18 Gy) has shown that the efficacy
J. Régis, M.D. (*) • R. Carron, M.D., Ph.D. was decreasing dramatically when the marginal doses were
Department of Functional Neurosurgery, Aix-Marseille
lower than 24 Gy [5]. Finally, our neurologists performed a
e-mail: jregis@ap-hm.fr long-term evaluation (>5years FU) in the first 15 consecutive
patients treated according to our standard protocol [11]. This
F. Bartolomei, M.D., Ph.D. • P. Chauvel, M.D.
Department of Neurophysiology, Aix-Marseille paper confirmed the good safety and efficacy of GKS in this
University—Timone University Hospital, Marseille, France group of patient on the long term with a rate of 60 % of Engel
700 J. Régis et al.
I at a mean FU of 8 years [6–10] comparing well with the Unfortunately, these data were never published in peer-reviewed
safety and efficacy of open surgery on the long term. More papers and precise data are unavailable.
recently a multicentric prospective trial in the USA has con- The Department of Stereotactic and Functional Surgery in
firmed all our findings [12]. Since this time, radiosurgery is Marseille has two major fields of expertise: epilepsy surgery
the current practice for pure MTLE in our group [13]. and radiosurgery. This context has therefore facilitated the
There are convincing arguments for such an investigation investigation and development of a potential role for GK
of the potential role of radiosurgery in epilepsy surgery. We radiosurgery in the treatment of intractable epilepsy. The
know that: first attempt to treat an MTLE occurred in Marseille in March
• Radiosurgery (since its introduction in the 1950s) has 1993 [6]. Since 1993, we have performed 217 cases of epi-
been demonstrated to have advantages in terms of safety lepsy surgery using GK radiosurgery.
and efficacy, for the treatment of numerous small deeply Among 11,066 Gamma Knife surgery procedures accom-
seated intracerebral lesions. plished in our neurosurgical unit in a 20-year period (between
• Radiosurgical treatment of small cortico-subcortical July 1992 and December 2012), only 217 were proposed in
lesions associated with epilepsy has been demonstrated to patients referred for epilepsy surgery (roughly ten patients/
lead to seizure cessation in a high percentage (58–80 % in year). During the same period, we have performed 759 non
AVM) of cases, long before the expected treatment of the radiosurgical neurosurgical intervention operations for epi-
lesion and sometimes even in spite of failing to cure the lepsy surgery. By the way, the philosophy of our team is to
lesion itself. define the niche of patients in whom the safety-efficacy ratio
• Radiotherapeutic treatment of epilepsies with or without makes it advantageous or at least comparing favorably to
space-occupying lesions can lead to a reduction in seizure open neurosurgery [34]. Obviously, this represents a small
frequency and/or severity. subset of patients in our present experience (23 %). However,
• Experimental models of epilepsies treated with radiation a part of the patients we are actually operating by GKS are
therapy have demonstrated a dose-dependent positive directly referred to us for this specific procedure. The real
effect of radiation on the frequency and severity of the percentage of patients coming from our own clinical pro-
seizures, and on the extent of discharge propagation. gram of investigation for epilepsy and operated by GKS is
Lars Leksell initially conceived of Gamma Knife (GK) only 14.6 %.
radiosurgery as a tool for functional neurosurgery [14, 15].
Accordingly, he used GK in movement disorders, trigeminal
neuralgia, and other pain syndromes, but not for epilepsy Hypothalamic Hamartomas
surgery [16]. The first radiosurgical treatments for epilepsy
surgery were performed by Talairach in the 1950s [17]. Hypothalamic hamartomas (HH) may be asymptomatic,
Talairach was another pioneering expert in stereotaxis. associated with precocious puberty or with neurological dis-
Unlike Leksell, he had specific involvement in epilepsy sur- orders (including epilepsy, behavior disturbances, and cog-
gery and led one of the first large comprehensive programs nitive impairment), or both. Usually the seizures begin early
for epilepsy surgery. As early as 1974, he reported on the use in life and are often particularly drug resistant from the out-
of radioactive Yttrium implants in patients with MTLE with- set. The evolution is unfavorable in the majority of the
out space-occupying lesions, and showed a high rate of sei- patients because of behavioral symptoms (particularly
zure control in patients with epilepsies confined to the mesial aggressive behavior) and mental decline, which occur as a
structures of the temporal lobe [17]. In 1980, Elomaa [18], direct effect of the seizures [35], due to an epileptic encepha-
apparently unaware of Talairach’s work, promoted the idea lopathy. Interestingly, in our experience, the reversal of this
of the use of focal irradiation for the treatment of temporal encephalopathy after radiosurgery seems to start even before
lobe epilepsy, based on the preliminary reports of Tracy, Von complete cessation of the seizures and seems to be correlated
Wieser, and Baudouin [19, 20]. Furthermore, clinical experi- to the improvement in background EEG activity. We may
ence of the use of GK- and Linac-based radiosurgery in arte- speculate that these continuous discharges are leading to the
riovenous malformations (AVMs) and cortico-subcortical disorganization of several systems, including the limbic sys-
tumors (mostly metastases and low-grade glial tumors) tem, and that their disappearance accounts for the improve-
revealed an antiepileptic effect of radiosurgery in absence of ment seen in attention, memory, cognitive performance,
necrotizing effect [21–23]. A series of experimental studies impulsive behavior, etc. Here, the goal of radiosurgery is the
in small animals confirmed this effect [24, 25] and has reversal of the epileptic encephalopathy more than seizure
emphasized its relationship to the dose delivered [26–29]. cessation. Consequently, we consider that it is essential to
Barcia Salorio et al., and later Lindquist et al., reported small operate on these young patients as early as possible, what-
and heterogeneous groups of patients treated with the aim of ever the surgical approach considered (resection or
seizure cessation; results were however poor [30–33]. radiosurgery).
58 Radiosurgery for Drug-Resistant Epilepsies: State of the Art, Results, and Perspectives 701
The intrinsic epileptogenicity of HH has been demon-

strated [36, 37] even though the mechanisms of the epilepsy
associated with HH are still debatable. The boundary of the
target zone of treatment is that of the lesion visualized on
magnetic resonance (MR) imaging. This contrasts greatly
with cases of MTLE where there is no such clear delineation
of an epileptogenic zone on the images used for planning
radiosurgical intervention.
We retrospectively analyzed radiosurgery in a series of
ten patients collected from centers around the world [38].
The very good safety-efficacy ratio (all improved, 50 %
cured, and no adverse effects except one case of poikilother-
mia) led us to organize a prospective multicenter trial. Our
trial of 64 prospectively evaluated patients is unique by the
number of patients and the strict methodology of this evalu- Fig. 58.1 Hypothalamic hamartomas topological classification. The
exact location of the lesion in relation to the interpeduncular fossa and
ation. We have published preliminary reports of this study
the walls of the third ventricle correlates with the extent of excision
[39–41] and the final evaluation is to be published. These 64 required, the seizure control, and the complication rate. Although it
patients operated between 1999 and 2007 have been all fol- may be an exceptional observation, Type V (pedunculated) tend not to
lowed more than 3 years (36–107 months). According to our have neurological symptoms (no epilepsy, no cognitive deterioration,
and no behavioral disturbances). They may present with precocious
policy, the patient and the family are offered a second radio-
puberty or be symptom-free. Types I, II, III, and IV may cause seizures
surgery in case of partial benefit when the lesion is anatomi- in many cases, as well as mental retardation, behavioral abnormalities,
cally small and well defined. Due to significant but incomplete and precocious puberty. Type VI HH are frequently found in patients
efficacy, 25 patients (62.5 %) were treated twice. The preop- with especially severe clinical presentations. We consider radiosurgery
as a first-line treatment for small Type I, II, III, or IV HH. From Regis
erative cognitive deficits, behavioral disturbances, and inves-
J, Hayashi M, Eupierre LP, Villeneuve N, Bartolomei F, Brue T, Chauvel
tigated relationship of seizure severity and anatomical type P. Gamma Knife surgery for epilepsy related to hypothalamic hamarto-
to cognitive abilities were characterized [42, 43]. The goal of mas. Acta Neurochir Suppl 91:33–50, 2004; Used with permission
the preoperative work-up was to adequately select the candi-
dates for inclusion and to evaluate the baseline neurological
and endocrinological functions. All radiosurgical procedures formed in six patients (9.3 %) with quite large HH and poor
were carried out using a Gamma Knife model B, C, 4C, or efficacy of radiosurgery. After this microsurgical approach, 1
Perfexion (Elekta Instrument, Stockholm). Consistently we is cured (Engel I), 2 improved (Engel III), and 3 not improved
elaborated multi-isocentric complex dose planning of high (Engel IV). We found no permanent or even transient neuro-
conformity and selectivity. We used low peripheral doses to logical deficit. A transient increase in seizures was observed
take into account the close relationship with optic pathways in seven patients (17.5 %). A transient non-disabling poikilo-
and hypothalamus (median 17 Gy; range 13–26 Gy). The thermia was observed in three cases.
lesions treated are generally small one (median 9.5 mm; Topological classification (Fig. 58.1) of the lesion based
range 5–26). We pay special attention to the dose delivered on a good high resolution MRI is a key feature in the
to the mammillary body and to the fornix and we always try decision-making process [39]. Previous classification is
to tailor the dose plan for each patient, based on the use of a based on anatomical [44–46] or surgical [47] consideration.
single run of shots with the 4 mm collimator. Patients were These classifications are not describing the large diversity of
evaluated with respect to seizures, cognition, behavior, and these lesions and their therapeutic consequences. As under-
endocrine status 6, 12, 18, 24, and 36 months after radiosur- lined by Palmini and coworkers, the exact location of the
gery and then every year. Results are demonstrating that in lesion in relation to the interpeduncular fossa and the walls
65 % of these patient an Engel I+II is achieved. An Engel III of the third ventricle correlates with the extent of excision,
is observed in 20 % of the patients. The frequency of seizures seizure control, and complication rate [48]. On this basis, we
before radiosurgery was 92 seizures/month (Mean, classify the HH according to their topology based on our
427 ± 1,009; min 3.3) and after fell to six seizures/month original classification [39, 40]. In our experience this classi-
(Mean 34.6 ± 78; min = 0; max = 425). In the majority of fication correlates with the clinical semiology and severity
these patients, a dramatic behavioral and cognitive improve- and is especially critical for surgical strategy selection. Type
ment is observed. Psychiatric and cognitive comorbidity was I (small HH located inside the hypothalamus extending more
cured in 28 %, improved in 56 %, and stable in 8 % of the or less in the third ventricle) are certainly the best candidates
patients. Globally a very good result has been obtained in for GKS. In this population the risk of microsurgical removal
60 % of the patients. A microsurgical approach has been per- is likely to be potentially high.
In Type II (when the lesion is small and mainly in the floor) but has been systematically disappointing. In our opin-
third ventricle) radiosurgery is certainly the safer alternative. ion, this strategy may cause loss of a precious time for the
Even though the endoscopic and transcallosal interforniceal child to be treated effectively. Consequently, we do not advo-
approaches have been proposed, the risks of short-term cate for such a strategy. When microsurgical resection has
memory worsening, endocrinological disturbance (hyper- left a small remnant in the third ventricle and a still active
phagia with obesity, low thyroxine, sodium metabolism dis- epilepsy, re-operation can be envisaged by GKS.
turbance), and thalamic or thalamocapsular infarcts have Two major questions remain. First, we know that com-
been reported also by the more enthusiastic and skillful neu- plete treatment or resection of the lesion is not always man-
rosurgeons. However in cases of very severe repeated status datory [50–52] but we do not know how to predict in an
epilepticus we propose as a salvage surgery either a transcal- individual patient the amount (and mapping) of the HH that
losal interforniceal approach or an endoscopic approach must be treated in order to obtain a complete antiepileptic
(depending on the width of third ventricle). In an emergency effect. Secondly, we know that these patients frequently
situation if the lesion is small and the third large, the endo- present with an electro-clinical semiology suggesting
scopic approach is chosen. involvement of the temporal or frontal lobe and which can
In Type III (lesion located essentially in the floor) the mimic a secondary epileptogenesis phenomenon [37, 49].
extremely close relationship between the mammillary body, In our experience, some of these patients can be completely
the fornix, and the lesion is clearly leading us to prefer cured by the isolated treatment of the HH, while in others, a
GKS. We speculate that sessile hypothalamic hamartomas partial result is obtained, with residual seizures despite a sig-
have always more or less an extension in the hypothalamus nificant overall psychiatric and cognitive improvement. In
close to the mammillary body. Thus when a lesion is classi- this second group, it is tempting to propose that such a sec-
fied as a Type II, it means that the lesion appears on the MR ondary epileptogenic area accounts for the partial failure.
like mainly located in the third ventricle but is likely to have Our initial results indicate that GKS is as effective as
a root in the hypothalamus. The same assumption is made for microsurgical resection and much safer [53]. GKS also
Type III. avoids the vascular risk related to radiofrequency lesioning
In Type IV (the lesion sessile in the cistern) a disconnection or stimulation. With transcallosal interforniceal approach,
can be discussed (pterional approach with or without orbitozy- both cognitive (long-term memory impairment) and severe
gomatic osteotomy). However, if the lesion is small, GKS can endocrinological complications (23 % long-term appetite
be recommended due to its safety and to its capability to reach stimulation and major weight gain) have been reported [54,
at the same time also the small associated part of the lesion in 55]. Stabell et al. have reported with endoscopy a serious
the hypothalamus itself, frequently visible on the high resolu- deterioration of memory and reading skills associated with a
tion MR. In Delalande experience only two patients among 14 permanent oculomotor paresis [56]. Using interstitial
are seizure free after a single disconnection through a pteri- implants (Iodine 125 seeds), Schulze-Bonhage et al. [57]
onal approach [49]. Consequently, we do use this approach in reported in a series of 24 patients, 5 patients developing a
case of lesion too large for GKS as a first step of a staged symptomatic edema; 4 patients having a weight gain of more
approach. In most circumstances, the patient is improved but than 5 kg, which was severe in 2; and a persistent decline of
not seizure free after the first surgical step and GKS is orga- episodic memory in 2 patients. None of these complications
nized at 3 months as a second step of the treatment. have been observed after Gamma Knife surgery. The main
Type V (pediculate) are rarely epileptic and can be easily disadvantage of radiosurgery is its delayed action. Longer
cured by radiosurgery or disconnection through a pterional follow-up is mandatory for proper evaluation of the role of
approach. In case of severe epilepsy the second therapeutic GKS. Results are faster and more complete in patients with
modality will allow certainly a faster seizure cessation. smaller lesions inside the III ventricle (Stage II). The early
However a distant extension of the HH in the hypothalamus effect on subclinical EEG discharges appears to play a major
close to the mammillary bodies must be cautiously searched role in the dramatic benefit to sleep quality, behavior, and
on high resolution MR and its discovery will eventually lead cognitive-developmental improvement. Gamma Knife sur-
to prefer GKS allowing to treat both parts of the lesion, espe- gery can safely lead to the reversal of the epileptic encepha-
cially in cases in which if the cisternal component is small. lopathy [38, 40, 41, 58].
Type VI (giant) do not represent good indications for first Due to the very poor clinical prognosis of the majority of
intention radiosurgery, as in nearly all the cases a combina- these patients with HH and the invasiveness of microsurgical
tion of several therapeutic modalities should be utilized. resection, GK can now be considered the first-line interven-
Even if GKS does not seem to be suitable when the lesion is tion for small middle size HH associated with epilepsy, as it
large, radiosurgical disconnection can be considered (target- can lead to dramatic improvements to the future of these
ing only the superior part in the hypothalamus and/or the young patients. The role of secondary epileptogenesis or of
third ventricle leaving untreated all the lesion lower than the widespread cortical dysgenesis in these patients needs to be
better evaluated and understood, in order to optimize patient first few months. Thereafter, there is a rapid and dramatic
selection and define the best treatment period. increase in auras for some days or weeks and then the sei-
zures disappear. Usually the peak in seizure cessation is
observed around the 8th–18th month with a clear variability
Mesial Temporal Lobe Epilepsy in the delay in onset. In one patient, this occurred 26 months
after GK radiosurgery. We usually consider a delay of 2
The first Gamma Knife surgery operations for MTLE were years as a minimum for post-radiosurgery follow-up. In the
performed in Marseille in March 1993. As far as no similar absence of initial radiological changes or clinical benefit, the
experience was available at this time in the literature, we recommendation is to wait for the onset of the MRI changes
were obliged to base our technical choices on hypothesis and their subsequent disappearance. All our patients had the
and experience of radiosurgery for other pathological condi- same pattern of MR changes regardless of marginal dose
tions. Four patients were treated with different technical (18–24 Gy) and treatment volume (5–8.5 cc). However, the
strategies (dose, volume, target definition). The delayed degree of these changes and their delay of onset varied
huge radiological changes observed some months after according to the dose delivered to the margin, the volume
radiosurgery [59] led us to stop such treatment and follow treated, and the individual patient. In order to allow an opti-
these first four patients. Due to the clinical safety of the pro- mal evaluation, we recommend that subsequent microsur-
cedure in these patients and the gradual disappearance of the gery not be considered before the third year after radiosurgery.
acute MR changes after some months, we treated several Similarly, we believe that a patient who undergoes a corti-
new series of patients under strict prospective controlled cectomy before the onset of the MR changes has occurred
trial conditions (with ethical committee approval). The treat- cannot be assumed to have failed radiosurgical treatment. Of
ment for the following 16 patients was based upon that of course, before consideration of any further surgery, the ques-
the first patient who had a successful outcome (as opposed tion of the reason for the failure needs to be addressed. After
to the three others who had partial or no effect). This “classic reviewing files of patients treated for MTLE with radiosur-
planning” was based on the use of two 18 mm shots, cover- gery, it was sometimes possible to identify likely causes of
ing a volume of around 7 cc at the 50 % isodose (24 Gy), and failure, such as:
has turned out to produce a high rate of seizure cessation 1 Poor patient selection (e.g., patients with epilepsy involv-
[60, 61]. For epileptological reasons, as well as for safety ing more than the MTL structures)
reasons, the targeting was very much centered on the para- 2 Patients with the diagnosis of “treatment failure”
hippocampal cortex and spared a significant part of the (<3 years) who had been operated upon too early after
amygdaloid complex and hippocampus. The refinement of radiosurgery [65]
the Gamma Knife surgery technique, and the desire to find a 3 Targeting of the amygdala and hippocampus (which is not
dose which would create less transient acute MR changes, in our opinion the optimal target in terms of safety and
led us to reduce the dose from 24 Gy to 20 and 18 Gy at the efficacy) instead of parahippocampal cortex [66]
margin. However, this brought about a significant decrease 4 Insufficient dosage [65–67]
in the rate of seizure cessation. We have reviewed the long- Our current strategy of treatment is based on our first
term follow-up of our first 15 patients operated by GKS for series of MTLE patients who were strictly selected and
MTLE at the state of the art (24 Gy). The mean follow-up treated systematically with a very simple but very reproduc-
was 8 years and at the last follow-up 73 % were seizure free. ible dose planning strategy [4, 6]. The identification of puta-
These long-term results are comparing favorably to micro- tive improvements in the methodology requires a systematic
surgical one. No permanent neurological deficit was reported analysis of the influence of the technical data from our
out of a visual field deficit in nine patients [11]. After micro- experience and from the literature on the outcome of those
surgery for MTLE on the dominant side, a verbal memory patients.
deficit is typically observed in 30–50 % of the patients [62, 63].
This is of special importance to note that none of our patients
have observed a neuropsychological worsening (using the The “Technical” Questions
evaluation published by Clusman et al.) and specifically no
verbal memory decline [4, 5, 7, 11]. This finding of our four The Dose Issue
prospective trials has been confirmed by the US prospective
trial [64]. The first targets used in functional GK radiosurgery (capsu-
The timetable of events after radiosurgery and the follow- lotomy, thalamotomy of VIM or the centromedianum, palli-
up are quite standardized. Patients are informed that delayed dotomy) were treated using high dose (300–150 Gy)
efficacy of radiosurgery is its main drawback. Typically, the delivered in very small volumes (3–5 mm in diameter) [16].
frequency of the seizures is not modified significantly for the The goal was to destroy a predefined very small anatomical
structure with stereotactic precision. Quite a significant

variability in the delay and amplitude of the MR changes
has been reported with fixed regimen of doses [28, 68].
Barcia Salorio et al. have presented several times a small
and heterogeneous group of patients treated with different
kinds of devices and dosage regimens [69]. Apparently
some of those patients had no expanding lesion and were
treated with very large volumes and very low dosage (around
10 Gy). Based on this experience, several teams have made
the assumption that very low doses, as low as 10–20 Gy at
the margin, should be as effective as the 24 Gy protocol (at
the margin) that we used for our first series of patients with
MTLE [4]. A cautious examination of the last proceeding of
Barcia Salorio et al. shows that the individual information
concerning the dose at the margin, the volume, and the
topography of the epileptogenic zone are not provided.
Moreover, among the 11 patients reported, the real rate of
seizure cessation is apparently only 36 % (4/11), which is
much lower than what we would expect with resection in
MTLE [30]. In a heterogeneous group of 176 patients, Yang
et al. confirmed that only a very low rate of seizure control
is achieved when low doses (from 9 to 13 Gy at the margin)
are used [67].
The experience of the radiosurgical treatment of HH indi-
cates that 18 Gy at the margin appears to be a threshold in Fig. 58.2 Dosimetry in a case of a typical right mesial temporal lobe
terms of probability of seizure cessation [38]. In this group epilepsy (24 Gy at the 50 %)
of patient (36 cases), only one showed MR changes. The
majority of the AVM cases with worsening of the epilepsy
were treated with a range of doses between 15 and 18 Gy. to our clinical experience in humans, a similar maximum
Similarly, poor results have been reported by Cmelak et al. in dose range of 40–50 Gy is currently the range of dose pro-
one case of MTLE treated with Linac-based radiosurgery, viding the optimal safety-efficacy ratio.
with 15 Gy at the 60 % isodose line, who underwent surgical
resection 1 year later. In this case, the authors first observed
a slight improvement followed by an obvious worsening The Target Definition
[65]. A recent de-escalation study has allowed us to demon-
strate poorer results in patients receiving doses of 18 or When the target is a lesion that is precisely defined radio-
20 Gy at the margin as compared to 24 Gy [59, 70]. Due to logically, the question of the selection of the marginal dose
the rate of seizure cessation that is achievable by conven- can be quite easily addressed by correlating safety-efficacy
tional resection, a radiosurgical strategy associated with a individual outcome to the marginal dose (Fig. 58.2). This
much lower rate of seizure cessation appears unacceptable. can be refined based upon stratification according to volume,
Fractionated stereotactically guided radiotherapy has been location, age, etc. However, in patients presenting with
demonstrated to fail systematically in controlling seizures. MTLE, this process is invalid for two reasons. Firstly, there
Among 12 patients treated by Grabenbauer et al., none have is no consensus regarding the requirement for extent of
achieved seizure cessation [71, 72]; only seizure reduction mesial temporal lobe resection. Secondarily, the concept of
was obtained in this series. MTLE syndrome with a stable extent of the epileptogenic
Experimental studies on small animals have demonstrated zone and surgical target is increasingly the topic of debate
the antiepileptic effect of radiosurgery [26, 28, 73], the dose [75, 76].
dependence of this effect [26, 27, 28, 74], and the possibility The volume (in association with marginal dose) is well
of obtaining clear antiepileptic effect without macroscopic known to be a major determinant of the tissue effect, as
necrosis using certain doses [27]. Of course, the rat models shown in integrated risk/dose volume formulae [77]. In the
of epilepsy are far from being good models of human first series of patients that we treated, this marginal isodose
MTLE. However, taking into account the huge difference in volume (or prescription isodose volume) was approximately
volume of the target, it is intriguing to notice that according 7 cc (range 5–8,5).
An attempt to correlate dose/volume and the effect on

seizures and on the MR changes (as evaluated by volume Patient Selection
of the contrast enhancement ring, extent of the high T2 sig-
nal, and the importance of the mass effect) has been pub- Whang (without having first performed specific preoperative
lished recently [70]. In this study, we found, not epileptological work-up) treated patients with epilepsy asso-
surprisingly, that the higher the dose and the volume, the ciated with slowly growing lesions and observed seizure ces-
higher the risk of having more severe MR changes, but also sation in only 38 % (12/31) of the patients [82]. This kind of
the higher the chance of achieving seizure cessation. observation emphasizes the importance of preoperative
However, these data have limited value. Hence, more pre- definition of the extent of the epileptic zone and of its
cise identification of those structures of the mesial tempo- relationship with the lesion [78, 83]. In our institution, the
ral lobe which need to be “covered” by the radiosurgical philosophy is to adapt the investigations for each individual
treatment may allow more selective, but just as efficacious, case. In some patients, the electro-clinical data, the structural
dose planning strategies, in spite of smaller prescription and functional imaging, and the neuropsychological exami-
isodose volumes. nation are sufficiently concordant for surgery of the temporal
There is growing evidence to support the organization of lobe to be proposed without depth electrode recording.
the epileptogenic zone in networks, meaning that several dif- In other cases, the level of evidence for MTLE is judged
ferent and possibly distant structures are discharging simulta- insufficient, and a stereoelectroencephalographic (SEEG)
neously at the onset of the electro-clinical seizure. This kind study is performed. The strategy of SEEG implantation is
of organization explains why the risk of failure is so high based on the primary hypothesis (mesial epileptogenic zone)
when a simple topectomy (without preoperative investiga- and alternative hypotheses (early involvement of the tempo-
tions) is performed in severe drug-resistant epilepsies associ- ral pole, lateral cortex, basal cortex, insular cortex, or other
ated with a benign lesion [78]. This has been also reported in cortical areas). The goal of these studies is to record the
MTLE [75, 76]. Certain nuclei of the amygdaloid complex, patient’s habitual seizures, in order to establish the temporo-
the head body and tail of the hippocampus, the perirhinal, spatial pattern of involvement of the cortical structures dur-
entorhinal (EC) and parahippocampal cortices may be asso- ing these seizures. Clearly in these patients, the high
ciated with the genesis of the seizures. The role of the EC resolution of depth electrode recording allows fine tailoring
cortex in epilepsy is supported by experimental studies in of surgical resection, according to the precise temporospatial
animal [79, 80]. The EC is considered to be the amplifier of course of the seizures. The main limitation of radiosurgery is
the “amygdalohippocampal epileptic system.” The pattern of that of size of the target (prescription isodose volume). The
the associated structures, including that of the structure play- radiosurgical treatment of MTLE is certainly the most selec-
ing the leader role, can vary significantly from one patient to tive surgical therapy for this group of patients. The require-
another [75, 76]. There is a subgroup of patients who have ments for precision and accuracy in the definition of the
clonic discharges and the involvement of the EC, amygdala, epileptogenic zone are consequently higher. Furthermore, if
and head of the hippocampus, with a clear leader role of the depth electrode investigation enables demonstration of a par-
EC. Wieser et al. have analyzed the postoperative MR images ticular subtype of MTLE, this can lead to tailoring of the
of patients operated by Yasargil (amygdalohippocampec- treatment volume and frequently allows this to be reduced.
tomy) and were able to correlate the quality of the resection
of each substructure of the mesial temporal lobe area and the
outcome with respect to seizures [81]. Only the quality of the The Potential Concerns
removal of the anterior parahippocampal cortex was corre-
lated strongly with a higher chance of seizure cessation [81]. The risk of long-term complications must always be cau-
We tried to perform a similar study in patients treated with tiously scrutinized in functional neurosurgery. Radiotherapy
GK radiosurgery [70]. We defined and manually drew the is most frequently used in the brain for short-term life-
limits of subregions on the stereotactic images of all these threatening pathologies. The use of radiotherapy in young
patients. The amygdala, the head, the body, and the tail of the patients with benign disease, such as pituitary adenomas or
hippocampus were first delineated. The white matter, the craniopharyngiomas, has been associated with a significant
parahippocampal cortex, and the cortex of the anterior wall rate of cognitive decline [61, 68] and tumor genesis [84]
of the collateral fissure were then separately drawn and including some carcinogenesis [85]. If the risk of radiation-
divided into four sectors in the rostrocaudal axis, correspond- induced tumor was similar with radiosurgery, we should
ing to the amygdala, the head, the body, and the tail of the have by now already observed numerous cases. However,
hippocampus [70]. such reported cases [86–88] are extremely rare and fre-
quently fail to meet the classical criteria by which tumors are tory due to the rarity of surgical complications and a high
deemed to be “radiation induced” [89]. In fact it is consid- rate of seizure freedom. In our experience, the most impor-
ered that, if this risk exists, it is likely to be around 1/10,000 tant selection parameters are the demonstration of the purely
which is far lower than the mortality risk associated with mesial location of the epileptogenic zone, as well as clear
temporal lobectomy [60, 90–93]. understanding by the patient of the advantages, disadvan-
Epilepsy is a life-threatening condition. The risk of sud- tages, and limitations. Best candidates are young patients,
den unexplained death in epileptic patients (SUDEP) is with middle severity epilepsy (working, socially well
higher than in the general population [94, 95]. This risk is inserted), with a high level of functioning (able to understand
higher in patients treated with more than two antiepileptic well the limits and constraints of radiosurgery), a quite high
drugs and IQ lower than 70 (as independent factors). Because risk of memory deficit with microsurgery (MTLE on the
seizure cessation after surgery reduces the mortality risk to dominant side with few or no atrophy, few deficit of the
that of the general population [95], microsurgical resection verbal memory preoperatively), and potentially huge social
of the epileptogenic zone may confer a benefit in terms of the and professional consequences in case of postoperative
possibility of immediate seizure cessation and therefore memory deficit [2, 13, 96]. One other very good indication in
reduced mortality risk, as compared to the more delayed our experience is that of patients with proven MTLE but pre-
benefits of radiosurgical treatment. Our patients are system- vious failure of microsurgery, supposedly due to insufficient
atically informed about this disadvantage of radiosurgery. posterior extent of the resection.
What Are the Current Indications?
The demonstrated advantages of radiosurgery are the com-

fort of the procedure, the absence of general anesthesia, the
absence of surgical complications and mortality, the very
short hospital stay, and the immediate return to the previous
level of functioning and employment. In MTLE the potential
sparing of memory function is still a matter of debate and
needs to be established using comparative studies. There is
also a requirement for further demonstration of long-term
Fig. 58.3 Radiosurgical anterior corpus-callosotomy in a young
efficacy and safety of radiosurgery. Worldwide, microsurgi- patient with a bilateral mesial frontal lobe epilepsy (MR negative).
cal corticectomies for MTLE are proving to be very satisfac- Dose planning displayed on a sagittal view (160 Gy at max)
Fig. 58.4 Periventricular

heterotopia in a young female
with severe drug-resistant
epilepsy (24 Gy at the 50 %).
SEEG investigation has allowed
location of the epileptogenic
zone as confined to a tiny part
of the periventricular heterotopia
in front of the posterior part of
the left temporal lobe. Seizure
cessation has occurred more than
1 year after radiosurgery. The
patient has been able to stop the
medication
Callosotomy can be performed by radiosurgery quite 7. Regis J, Rey M, Bartolomei F, Vladyka V, Liscak R, Schrottner O,
safely as previously reported by Shrottner et al. (Fig. 58.3). Pendl G. Gamma knife surgery in mesial temporal lobe epilepsy: a
prospective multicenter study. Epilepsia. 2004;45:504–15.
When the epileptological context makes realistic an anterior 8. Regis J, Semah F, Bryan R, Levrier O, Rey M, Samson Y, Peragut
callosotomy and when the invasivity of a microsurgical J. Early and delayed MR and PET changes after selective tem-
approach is a concern, radiosurgery can be a very appealing poromesial radiosurgery in mesial temporal lobe epilepsy. AJNR
alternative (six cases in our material). Am J Neuroradiol. 1999;20:213–6.
9. Tsuchitani S, Drummond J, Kamiryo T, Anzivino M, Chen Z,
Finally, we have started some years ago to treat epilepsies Steiner L, Lee K. Selective vulnerability of interneurons to low dos-
associated with benign lesions in highly functional areas (21 age radiosurgery. Presented at Society for Neuroscience Annual
patients). We have treated small benign lesions like dysplasia Meeting [abstract], New Orleans, 2003.
or periventricular heterotopia (Fig. 58.4). The high func- 10. Quigg M, Rolston J, Barbaro NM. Radiosurgery for epilepsy: clini-
cal experience and potential antiepileptic mechanisms. Epilepsia.
tional risk of any epilepsy surgeries in these patients has led 2012;53:7–15.
us first to perform a cautious analysis of the extent of the 11. Bartolomei F, Hayashi M, Tamura M, Rey M, Fischer C, Chauvel P,
epileptogenic zone relying on SEEG investigation. When the Regis J. Long-term efficacy of gamma knife radiosurgery in mesial
SEEG is confirming the ictal onset in the lesion or its close temporal lobe epilepsy. Neurology. 2008;70:1658–63.
12. Barbaro NM, Quigg M, Broshek DK, Ward MM, Lamborn KR,
vicinity and when the EZ is confined to a small area, radio- Laxer KD, Larson DA, Dillon W, Verhey L, Garcia P, Steiner L,
surgery is systematically discussed. Heck C, Kondziolka D, Beach R, Olivero W, Witt TC, Salanova V,
Goodman R. A multicenter, prospective pilot study of gamma
knife radiosurgery for mesial temporal lobe epilepsy: seizure
response, adverse events, and verbal memory. Ann Neurol.
Conclusions 2009;65:167–75.
13. Regis J, Bartolomei J, Chauvel P. Epilepsy. Prog Neurol Surg.
The field of epilepsy surgery is a new and promising one for 2007;20:267–78.
radiosurgery. However, determination of the extent of the 14. Leksell L. The stereotaxic method and radiosurgery of the brain.
Acta Chir Scand. 1951;102:316–9.
epileptogenic zone requires specific expertise, which is cru- 15. Leksell L. Stereotaxic radiosurgery in trigeminal neuralgia. Acta
cial in order to achieve a reasonable rate of seizure cessation. Chir Scand. 1971;137:311–4.
In addition, the huge impact of fine technical detail on the 16. Lindquist C, Kihlstrom L, Hellstrand E. Functional neurosurgery—
efficacy and eventual toxicity of the procedure means that, at a future for the gamma knife? Stereotact Funct Neurosurg. 1991;
57:72–81.
present, its use for these indications remains under evalua- 17. Talairach J, Bancaud J, Szikla G, Bonis A, Geler S, Vedrenne
tion, and further prospective work is absolutely required. It is C. Approche nouvelle de la neurochirurgie de l’epilepsie.
difficult to know whether we really are at the dawning of a Méthodologie stéréotaxique et résultats thérapeutiques.
broader indication for the use of radiosurgery in forthcoming Neurochirurgie. 1974;20:92–8.
18. Elomaa E. Focal irradiation of the brain: an alternative to temporal
years; our ability to identify the correct technical strategies lobe resection in intractable focal epilepsy ? Med Hypotheses.
should determine whether this is so! 1980;6:501–3.
19. Baudouin M, Stuhl L, Perrard A. Un cas d’épilepsie focale traité par
la radiothérapie. Rev Neurol. 1951;84:60–3.
20. Von Wieser W. Die Roentgentherapie der traumatischen Epilepsie.
References Mschr Psychiat Neurol. 1939;101:422–4.
21. Heikkinen ER, Konnov B, Melnikow L. Relief of epilepsy by radio-
1. Regis J, Bartolomei F, Hayashi M, Chauvel P. Gamma Knife sur- surgery of cerebral arteriovenous malformations. Stereotact Funct
gery, a neuromodulation therapy in epilepsy surgery! Acta Neurosurg. 1989;53:157–66.
Neurochir Suppl. 2002;84:37–47. 22. Rogers L, Morris H, Lupica K. Effect of cranial irradiation on sei-
2. Regis J, Carron R, Park M. Is radiosurgery a neuromodulation ther- zure frequency in adults with low-grade astrocytoma and medically
apy?: a 2009 Fabrikant award lecture. J Neurooncol. 2010;98: intractable epilepsy. Neurology. 1993;43:1599–601.
155–62. 23. Rossi G, Scerrati M, Roselli R. Epileptogenic cerebral low grade
3. Régis J, Kerkerian-Legoff L, Rey M, Viale M, Porcheron D, tumors: effect of interstitial stereotactic irradiation on seizures.
Nieoullon A, Peragut J-C. First biochemical evidence of differen- Appl Neurophysiol. 1985;48:127–32.
tial functional effects following gamma knife surgery. Stereotact 24. Barcia Salorio JL, Roldan P, Hernandez G, Lopez Gomez
Funct Neurosurg. 1996;66:29–38. L. Radiosurgery treatment of epilepsy. Appl Neurophysiol. 1985;
4. Régis J, Bartolomei F, Rey M, Genton P, Dravet C, Semah F, 48:400–3.
Gastaut J, Chauvel P, Peragut J. Gamma knife surgery for mesial 25. Gaffey C, Monotoya V, Lyman J, Howard J. Restriction of the
temporal lobe epilepsy. Epilepsia. 1999;40:1551–6. spread of epileptic discharges in cats by mean of Bragg Peak intra-
5. Regis J, Levivier M, Hayashi M. Radiosurgery for intractable epi- cranial irradiation. Int J Appl Radiat Isotope. 1981;32:779–87.
lepsy. Tech Neurosurg. 2003;9:191–203. 26. Chen ZF, Kamiryo T, Henson SL, Yamamoto H, Bertram EH,
6. Régis J, Peragut JC, Rey M, Samson Y, Levrier O, Porcheron D, Schottler F, Patel F, Steiner L, Prasad D, Kassell NF, Shareghis S,
Regis H, Sedan R. First selective amygdalohippocampic radiosur- Lee KS. Anticonvulsant effects of gamma surgery in a model of
gery for mesial temporal lobe epilepsy. Stereotact Funct Neurosurg. chronic spontaneous limbic epilepsy in rats. J Neurosurg.
1994;64:191–201. 2001;94:270–80.
27. Maesawa S, Kondziolka D, Dixon C, Balzer J, Fellows W, Lunsford 46. Valdueza JM, Cristante L, Dammann O, Bentele K, Vortmeyer A,
L. Subnecrotic stereotactic radiosurgery controlling epilepsy produced Saeger W, Padberg B, Freitag J, Herrmann HD. Hypothalamic ham-
by kainic acid injection in rats. J Neurosurg. 2000;93:1033–40. artomas: with special reference to gelastic epilepsy and surgery.
28. Mori Y, Kondziolka D, Balzer J, Fellows W, Flickinger J, Lunsford Neurosurgery. 1994;34:949–58; discussion 958.
L, Thulborn K. Effects of stereotactic radiosurgery on an animal 47. Delalande O, Fohlen M. Disconnecting surgical treatment of hypo-
model of hippocampal epilepsy. Neurosurgery. 2000;46:157–65; thalamic hamartoma in children and adults with refractory epilepsy
discussion 165–58. and proposal of a new classification. Neurol Med Chir (Tokyo).
29. Ronne-Engström E, Kihlström L, Flink R, Hillered L, Carlson H, 2003;43:61–8.
Linquist C, Lin K. Gamma Knife surgery in epilepsy: an experi- 48. Palmini A, Chandler C, Andermann F, Costa Da Costa J, Paglioli-
mental model in the rat. Presented at European Society for Neto E, Polkey C, Rosenblatt B, Montes J, Martinez JV, Farmer JP,
Stereotactic and Functional Neurosurgery, 1993. Sinclair B, Aronyk K, Paglioli E, Coutinho L, Raupp S, Portuguez
30. Barcia Salorio JL, Garcia JA, Hernandez G, Lopez Gomez M. Resection of the lesion in patients with hypothalamic hamarto-
L. Radiosurgery of epilepsy: long-term results. Presented at European mas and catastrophic epilepsy. Neurology. 2002;58:1338–47.
Society for Stereotactic and Functional Neurosurgery, 1993. 49. Cascino GD, Andermann F, Berkovic SF, Kuzniecky RI, Sharbrough
31. Lindquist C. Gamma knife surgery in focal epilepsy. 1 year follow- FW, Keene DL, Bladin PF, Kelly PJ, Olivier A, Feindel W. Gelastic
up in 4 cases (Leksell Gamma Knife Society, Stockholm, 1992). seizures and hypothalamic hamartomas: evaluation of patients
32. Lindquist C, Hellstrand E, Kilström L, Abraham-Fuchs K, Jernberg undergoing chronic intracranial EEG monitoring and outcome of
B, Wirth A. Stereotactic localisation of epileptic foci by magneto- surgical treatment. Neurology. 1993;43:747–50.
encephalography and MRI followed by gamma surgery. Presented 50. Pascual-Castroviejo I, Moneo JH, Viano J, Garcia-Segura JM,
at International Stereotactic Radiosurgery Symposium, 1991. Herguido MJ, Pascual Pascual SI. [Hypothalamic hamartomas:
33. Lindquist C, Kihlström L, Hellstrand E, Knutsson E. Stereotactic control of seizures after partial removal in one case]. Rev Neurol.
radiosurgery instead of conventional epilepsy surgery. Presented at 2000;31:119–22.
European Society for Stereotactic and Functional Neurosurgery, 1993. 51. Rosenfeld JV, Harvey AS, Wrennall J, Zacharin M, Berkovic
34. Regis J, Bartolomei F, Hayashi M, Roberts D, Chauvel P, Peragut SF. Transcallosal resection of hypothalamic hamartomas, with con-
JC. The role of gamma knife surgery in the treatment of severe trol of seizures, in children with gelastic epilepsy. Neurosurgery.
epilepsies. Epileptic Disord. 2000;2:113–22. 2001;48:108–18.
35. Deonna T, Ziegler AL. Hypothalamic hamartoma, precocious 52. Watanabe T, Enomoto T, Uemura K, Tomono Y. Nose T: [Gelastic
puberty and gelastic seizures: a special model of “epileptic” devel- seizures treated by partial resection of a hypothalamic hamartoma].
opmental disorder. Epileptic Disord. 2000;2:33–7. No Shinkei Geka. 1998;26:923–8.
36. Kuzniecky R, Guthrie B, Mountz J, Bebin M, Faught E, Gilliam F, 53. Wait SD, Abla AA, Killory BD, Nakaji P, Rekate HL. Surgical
Liu HG. Intrinsic epileptogenesis of hypothalamic hamartomas in approaches to hypothalamic hamartomas. Neurosurg Focus. 2011;
gelastic epilepsy. Ann Neurol. 1997;42:60–7. 30:E2.
37. Munari C, Kahane P, Francione S, Hoffmann D, Tassi L, Cusmai R, 54. Anderson JF, Rosenfeld JV. Long-term cognitive outcome after trans-
Vigevano F, Pasquier B, Betti OO. Role of the hypothalamic ham- callosal resection of hypothalamic hamartoma in older adolescents
artoma in the genesis of gelastic fits (a video-stereo-EEG study). and adults with gelastic seizures. Epilepsy Behav. 2010;18:81–7.
Electroencephalogr Clin Neurophysiol. 1995;95:154–60. 55. Freeman JL, Zacharin M, Rosenfeld JV, Harvey AS. The endocri-
38. Regis J, Bartolomei F, de Toffol B, Genton P, Kobayashi T, Mori Y, nology of hypothalamic hamartoma surgery for intractable epi-
Takakura K, Hori T, Inoue H, Schrottner O, Pendl G, Wolf A, Arita lepsy. Epileptic Disord. 2003;5:239–47.
K, Chauvel P. Gamma knife surgery for epilepsy related to hypotha- 56. Stabell KE, Bakke SJ, Egge A. Cognitive and neurological sequelae
lamic hamartomas. Neurosurgery. 2000;47:1343–51; discussion after stereoendoscopic disconnection of a hypothalamic hamar-
1351–42. toma. A case study. Epilepsy Behav. 2012;24:274–8.
39. Regis J, Hayashi M, Eupierre LP, Villeneuve N, Bartolomei F, Brue 57. Schulze-Bonhage A, Trippel M, Wagner K, Bast T, Deimling FV,
T, Chauvel P. Gamma knife surgery for epilepsy related to hypotha- Ebner A, Elger C, Mayer T, Keimer R, Steinhoff BJ, Spreer J, Fauser
lamic hamartomas. Acta Neurochir Suppl. 2004;91:33–50. S, Ostertag C. Outcome and predictors of interstitial radiosurgery in
40. Regis J, Scavarda D, Tamura M, Nagayi M, Villeneuve N, the treatment of gelastic epilepsy. Neurology. 2008;71:277–82.
Bartolomei F, Brue T, Dafonseca D, Chauvel P. Epilepsy related to 58. Mathieu D, Deacon C, Pinard CA, Kenny B, Duval J. Gamma
hypothalamic hamartomas: surgical management with special refer- Knife surgery for hypothalamic hamartomas causing refractory epi-
ence to gamma knife surgery. Childs Nerv Syst. 2006;22:881–95. lepsy: preliminary results from a prospective observational study.
41. Regis J, Scavarda D, Tamura M, Villeneuve N, Bartolomei F, Brue J Neurosurg. 2010;113 Suppl:215–21.
T, Morange I, Dafonseca D, Chauvel P. Gamma knife surgery for 59. Hayashi M, Regis J. Hori T: [Current treatment strategy with
epilepsy related to hypothalamic hamartomas. Semin Pediatr gamma knife surgery for mesial temporal lobe epilepsy]. No
Neurol. 2007;14:73–9. Shinkei Geka. 2003;31:141–55.
42. Frattali CM, Liow K, Craig GH, Korenman LM, Makhlouf F, Sato 60. Ganz JC. Gamma knife radiosurgery and its possible relationship to
S, Biesecker LG, Theodore WH. Cognitive deficits in children with malignancy: a review. J Neurosurg. 2002;97:644–52.
gelastic seizures and hypothalamic hamartoma. Neurology. 61. Glosser G, McManus P, Munzenrider J, Austin-Seymour M,
2001;57:43–6. Fullerton B, Adams J, Urie MM. Neuropsychological function in
43. Weissenberger AA, Dell ML, Liow K, Theodore W, Frattali CM, adults after high dose fractionated radiation therapy of skull base
Hernandez D, Zametkin AJ. Aggression and psychiatric comorbid- tumors. Int J Radiat Oncol Biol Phys. 1997;38:231–9.
ity in children with hypothalamic hamartomas and their unaffected 62. Clusmann H, Schramm J, Kral T, Helmstaedter C, Ostertun B,
siblings. J Am Acad Child Adolesc Psychiatry. 2001;40:696–703. Fimmers R, Haun D, Elger CE. Prognostic factors and outcome
44. Arita K, Ikawa F, Kurisu K, Sumida M, Harada K, Uozumi T, after different types of resection for temporal lobe epilepsy.
Monden S, Yoshida J, Nishi Y. The relationship between magnetic J Neurosurg. 2002;97:1131–41.
resonance imaging findings and clinical manifestations of hypotha- 63. Stroup E, Langfitt J, Berg M, McDermott M, Pilcher W, Como
lamic hamartoma. J Neurosurg. 1999;91:212–20. P. Predicting verbal memory decline following anterior temporal
45. Debeneix C, Bourgeois M, Trivin C, Sainte-Rose C, Brauner lobectomy (ATL). Neurology. 2003;60:1266–73.
R. Hypothalamic hamartoma: comparison of clinical presentation 64. Quigg M, Broshek DK, Barbaro NM, Ward MM, Laxer KD, Yan G,
and magnetic resonance images. Horm Res. 2001;56:12–8. Lamborn K. Neuropsychological outcomes after Gamma Knife
radiosurgery for mesial temporal lobe epilepsy: a prospective mul- and epileptogenesis in vitro. In: Avanzini G, Engel J, Fariello R,
ticenter study. Epilepsia. 2011;52:909–16. Heinemann U, editors. Neurotransmitters in epilepsy. Amsterdam:
65. Cmelak AJ, Abou-Khalil B, Konrad PE, Duggan D, Maciunas Elsevier Science; 1992. p. 173–80.
RJ. Low-dose stereotactic radiosurgery is inadequate for medically 80. Wilson W, Swartzwelder HS, Anderson WW, Lewis DV. Seizure
intractable mesial temporal lobe epilepsy: a case report. Seizure. activity in vitro: a dual focus model. Epilepsy Res. 1988;2:
2001;10:442–6. 289–93.
66. Kawai K, Suzuki I, Kurita H, Shin M, Arai N, Kirino T. Failure 81. Wieser HG, Siegel AM, Yasargil GM. The Zurich amygdalo-
of low-dose radiosurgery to control temporal lobe epilepsy. hippocampectomy series: a short up-date. Acta Neurochir Suppl
J Neurosurg. 2001;95:883–7. (Wien). 1990;50:122–7.
67. Yang KJ, Wang KW, Wu HP, Qi ST. Radiosurgical treatment of 82. Whang CJ, Kwon Y. Long-term follow-up of stereotactic gamma
intractable epilepsy with low radiation dose. Di Yi Jun Yi Da Xue knife radiosurgery in epilepsy. Stereotact Funct Neurosurg. 1996;66
Xue Bao. 2002;22:645–7. suppl 1:349–56.
68. McCord MW, Buatti JM, Fennell EM, Mendenhall WM, Marcus Jr 83. Kitchen N. Experimental and clinical studies on the putative thera-
RB, Rhoton AL, Grant MB, Friedman WA. Radiotherapy for pitu- peutic efficacy of cerebral irradiation (radiotherapy) in epilepsy.
itary adenoma: long-term outcome and sequelae. Int J Radiat Oncol Epilepsy Res. 1995;20:1–10.
Biol Phys. 1997;39:437–44. 84. Strasnick B, Glasscock ME, Haynes D. The natural history of
69. Barcia-Salorio JL, Barcia JA, Hernandez G, Lopez-Gomez untreated acoustic neuromas. Laryngoscope. 1994;104:1115–9.
L. Radiosurgery of epilepsy. Long-term results. Acta Neurochir 85. Simmons NE, Laws E. Gliomas occurrence after sellar irradiation:
Suppl (Wien). 1994;62:111–3. case report and review. Neurosurgery. 1998;42:172–8.
70. Hayashi M, Bartolomei F, Rey M, Farnarier P, Chauvel P, Regis 86. Kaido T, Hoshida T, Uranishi R, Akita N, Kotani A, Nishi N, Sakaki
J. MR changes after Gamma knife radiosurgery for Mesial Temporal T. Radiosurgery-induced brain tumor. Case report. J Neurosurg.
Lobe Epilepsy: an evidence for the efficacy of subnecrotic doses. 2001;95:710–3.
In: Kondziolka D, editor. Radiosurgery. Karger: Basel; 2002. 87. Shamisa A, Bance M, Nag S, Tator C, Wong S, Noren G, Guha
p. 192–202. A. Glioblastoma multiforme occurring in a patient treated with
71. Grabenbauer GG, Reinhold C, Kerling F, Muller RG, Lambrecht U, gamma knife surgery. Case report and review of the literature.
Pauli E, Ganslandt O, Sauer R, Stefan H. Fractionated stereotacti- J Neurosurg. 2001;94:816–21.
cally guided radiotherapy of pharmacoresistant temporal lobe epi- 88. Yu JS, Yong WH, Wilson D, Black KL. Glioblastoma induction
lepsy. Acta Neurochir. 2002;84:65–70. after radiosurgery for meningioma. Lancet. 2000;356:1576–7.
72. Stefan H, Hummel C, Grabenbauer GG, Muller RG, Robeck S, 89. Cahan W, Woodard H, Highinbotham N, Stewart F, Coley
Hofmann W, Buchfelder M. Successful treatment of focal epilepsy B. Sarcoma arising in irradiated bone: report of eleven cases.
by fractionated stereotactic radiotherapy. Eur Neurol. 1998;39: Cancer. 1948;1:3–29.
248–50. 90. Loeffler JS, Niemierko A, Chapman PH. Second tumors after radio-
73. Barcia Salorio JL, Vanaclocha V, Cerda M, Roldan P. Focus irradia- surgery: tip of the iceberg or a bump in the road? Neurosurgery.
tion in epilepsy. Experimental study in the cat. Appl Neurophysiol. 2003;52:1436–40; discussion 1440–32.
1985;48:152. 91. Lunsford LD, Niranjan A, Flickinger JC, Maitz A, Kondziolka
74. Maesawa S, Kondziolka D, Balzer J, Fellows W, Dixon E, Lunsford D. Radiosurgery of vestibular schwannomas: summary of experi-
LD. The behavioral and electroencephalographic effects of stereo- ence in 829 cases. J Neurosurg. 2005;102 Suppl:195–9.
tactic radiosurgery for the treatment of epilepsy evaluated in the rat 92. Muracciole X, Cowen D. Regis J: [Radiosurgery and brain radio-
kainic acid model. Stereotact Funct Neurosurg. 1999;73:115. induced carcinogenesis: update]. Neurochirurgie.
75. Bartolomei F, Wendling F, Bellanger JJ, Regis J, Chauvel P. Neural 2004;50:414–20.
networks involving the medial temporal structures in temporal lobe 93. Rowe J, Grainger A, Walton L, Silcocks P, Radatz M, Kemeny
epilepsy. Clin Neurophysiol. 2001;112:1746–60. A. Risk of malignancy after gamma knife stereotactic radiosurgery.
76. Spencer S, Spencer D. Entorhinal-hippocampal interactions in Neurosurgery. 2007;60:60–5; discussion 65–6.
medial temporal lobe epilepsy. Epilepsia. 1994;35:721–7. 94. Ficker DM, So EL, Shen WK, Annegers JF, O’Brien PC, Cascino
77. Flickinger JC. An integrated logistic formula for prediction of com- GD, Belau PG. Population-based study of the incidence of
plications from radiosurgery. Int J Radiat Oncol Biol Phys. sudden unexplained death in epilepsy. Neurology.
1989;17:879–85. 1998;51:1270–4.
78. Régis J, Bartolomei F, Kida Y, Kobayashi T, Vladyka V, Liscàk R, 95. Sperling MR, Feldman H, Kinman J, Liporace JD, O’Connor
Forster D, Kemeny A, Schröttner O, Pendl G. Radiosurgery of epi- MJ. Seizure control and mortality in epilepsy. Ann Neurol. 1999;
lepsy associated with cavernous malformation: retrospective study 46:45–50.
in 49 patients. Neurosurgery. 2000;47:1091–7. 96. Regis J, Arkha Y, Yomo S, Bartolomei F, Peragut JC. Chauvel P:
79. Jones R, Heinemann U, Lambert J. The entorhinal cortex and gen- [Radiosurgery for drug-resistant epilepsies: state of the art, results
eration of seizure activity: studies of normal synaptic transmission and perspectives]. Neurochirurgie. 2008;54:320–31.
Epilepsy: Viewpoint—Surgery
59
David Carter and Zulma Tovar-Spinoza
imaging (magnetoencephalography [8], magnetic resonance

Introduction spectroscopy [9], fluorodeoxyglucose positron emission
tomography (FDG-PET), and single-photon emission com-
Epilepsy is the occurrence of two or more unprovoked puted tomography (SPECT). Phase II consists of a cerebral
seizures. About 50 million people in the world suffer from epi- angiogram and a bilateral intracarotid amobarbital procedure
lepsy, with age-adjusted prevalence rates of 0.2–4.1 % [1]. (Wada test) to evaluate lateralization of language function
Fundamentally, there are two types of epilepsy, primary as well to evaluate the integrity of memory function on
generalized and localization-related epilepsy. Primary gener- each side. Phase III involves invasive (intracranial) EEG
alized epilepsy is considered to have a genetic etiology, monitoring (Fig. 59.1). Phase IV is the actual resective
whereas most localization-related epilepsy is presumed to be surgery with intraoperative functional mapping and elec-
the result of a cerebral insult, even though the insult cannot trocorticography as needed.
be determined in about half of all epileptic patients, regard- Based on the results of the neuroimaging studies, patients
less of age [2]. Localization-related epilepsy accounts for are divided into those with and without space-occupying
about 60–75 % of all epilepsy. It has mostly a temporal lobe lesions. A space-occupying lesion is generally a tumor, a
location in adults and an extra temporal source in children. vascular anomaly, or a well-defined developmental abnor-
Candidates for epilepsy surgery have medically refractory mality such as a hamartoma or focal cortical dysplasia.
epilepsy, defined as inadequate seizure control in spite of Resective surgery could proceed without phase III inves-
appropriate medical therapy with at least two of the specific tigation if an interictal or ictal EEG or both demonstrated
antiepileptic drugs (AED) for the specific type of seizure, or abnormal activity in the same area as the space-occupying
with controlled seizures but with unacceptable drug-related lesion, and there is no contradictory localizing information
side effects [3–5]. from other tests. In these cases, the resection includes the
lesion and is extended to surrounding cortex to an extent
determined by results of intraoperative electrocorticography,
Presurgical Evaluation pathological evidence of normal cytoarchitectonic margins,
or both. If the resection area is adjacent to or in important
In order to obtain the best result for seizure control and functional cortex, intraoperative mapping/monitoring can be
minimize complications, the selection of the appropriate used. For patients with space-occupying lesions who do not
surgical intervention is essential by identifying the patient’s fit these criteria, phase III is done to provide EEG localiza-
epilepsy syndrome and, where relevant, localization. Most tion as well as extraoperative functional mapping as needed.
centers follow a standard, staged protocol for evaluation of Patients without space-occupying lesions fall into two
patients [6, 7]. Phase I involves noninvasive testing. This categories: those with suspected neocortical epilepsy or
includes history, neurologic evaluation, MRI, video-EEG those with temporal epilepsy. In patients with mesial tempo-
monitoring, neuropsychological evaluation and functional ral lobe epilepsy, if there is concordance with the informa-
tion from phases I–II, resective surgery could proceed
without phase III investigations. If no area of resection is
D. Carter, M.D., Ph.D. (*) • Z. Tovar-Spinoza, M.D.
defined or further functional information is required, then
Department of Neurosurgery, SUNY Upstate
Medical University, Syracuse, NY, USA intracranial recording (phase III) is considered necessary for
e-mail: carterda@upstate.edu localization of epileptogenic area and functional mapping.
712 D. Carter and Z. Tovar-Spinoza
On MRI, the most common lesion is hippocampal sclerosis

(hippocampal shrinkage and increased T2 signal), but other
abnormalities can be seen such as dysplasias (some are
microdysplasias and only appreciated pathologically),
tumors, and vascular malformations. These lesions can be
considered examples of “dual pathology” in the temporal
lobe and may suggest a more generous resection than lesio-
nectomy is needed [22].
When an adequate constellation of features is present,
they permit a diagnosis of MTLE with enough confidence to
consider surgical resection for seizure control. Corroborative
testing for localization can include magnetoencephalogra-
phy, PET, ictal SPECT, magnetic resonance spectroscopy, or
Fig. 59.1 Intracranial monitoring with grids and strips invasive monitoring [23]. Neuropsychological testing and
testing for language and memory (e.g., WADA, fMRI) can
provide further information on function and localization,
help plan resections, predict functional reserve after surgery,
After the evaluation is complete, patients are deter- and establish baselines. The precise “origin” of the seizure(s)
mined to be eligible for resective surgery or not. Ineligible is generally not known; but rather the investigations impli-
patients for resective surgery are considered for surgery cate the mesial structures but also more widespread con-
that disconnect and prevent propagation of epileptic dis- nected regions [21, 24–26].
charges such as multiple subpial transections [10] or
vagus nerve stimulation [11].
Novel techniques include radiosurgery [12] and magnetic Procedures
resonance-guided laser-induced thermal lesioning protocols
of epileptogenic foci [13]. Penfield and Baldwin [19] described the basic resection,
beginning with anterolateral temporal cortex, limited
posteriorly to the vein of Labbe, and below the Sylvian fissure,
Medial Temporal Lobe Epilepsy but they often individualized this by pathologic findings and
electrocorticography with mapping, frequently with the
Background patient conscious. As mentioned above, the mesial structures
were removed to the pia overlying the basal cisterns, includ-
The temporal lobe was postulated as a site for onset of sei- ing uncus, inferior amygdala, and hippocampus [19, 20]
zures in the late 1800s [14]. The behavioral and electroen- (Fig. 59.2).
cephalographic characteristics of these seizures were Over the past 60 years, one of the interests in the surgical
becoming increasingly understood by the middle of the last procedure for MTLE has been to decrease the extent of tissue
century [15–18] enough that subtotal anterior temporal removal aiming to minimize any deficit while maintaining
resections—to include the mesial structures (inferior amyg- the benefit of surgery. Also, in this endeavor, there was often
dala, hippocampus, and parahippocampus), as well as a goal to better identify the structure(s) whose removal is
anterolateral neocortex—began to be performed for intrac- critical for success at relieving seizures. This conceivably
table temporal seizures [19, 20]. Surgeons found the mesial could lead to increasing surgical success rate. Study of modi-
structures often were sclerosed—but the cause was uncertain fications of subtotal anterior temporal resections continues
(and remains incompletely understood [18]). today. However, no modification is known to be most effica-
The syndrome of medial temporal lobe epilepsy (MTLE) cious for seizure control, nor is there consensus on the
is now well described, allowing it to be generally distin- structure(s) to be removed for surgical cure of seizures [27].
guished from seizures beginning outside the temporal lobe, This is reflected in the fact that the basic procedure described
and it remains important to aim to differentiate them from by Penfield and Baldwin [19] and Falconer [28] has a good
seizures arising in the lateral temporal lobe [21]. Febrile sei- success rate (comparable to the range of techniques
zures are a risk factor. Family history of seizures may be described), favorable side effect profile and remains a stan-
reported. Clinically, visceral sensations are the most com- dard contemporary surgical treatment [29]. A major determi-
mon auras; the next most common is fear. Electrophysiological nant in developing operative variations in the temporal lobe
features of temporal seizure activity, particularly medial was the location of temporal neural tissue critical to language
temporal lobe location are observed, both interictal and ictal. function, generally felt to be largely in the superior and more
59 Epilepsy: Viewpoint—Surgery 713
In addition to these frequently used open surgical

approaches, stereotactic approaches have also been evolving
for lesioning of mesial structures. Radiofrequency lesions of
mesial structures have been described to result in success
rates for seizure control similar to that of open surgery, but
experience is limited [45]. These centers have also examined
the question of which structure must be lesioned to obtain
seizure control. Although there was a general relationship
between lesioning of mesial structures and outcome, they did
not find a significant correlation between lesion of particular
mesial structures (e.g., hippocampus, amygdala, entorhinal
cortex) and outcome [46]. Recently, successful magnetic
resonance-guided stereotactic laser-induced thermal lesion-
ing of mesial structures for epilepsy has been described in a
case report [13].
Fig. 59.2 Schematic of cross section of the temporal lobe demonstrat-

ing the medial temporal lobe (a) and lateral temporal lobe (b). Temporal Outcomes
resections for epilepsy may focus on lateral temporal cortex, medial
temporal structures, or resection of both. Specific approaches through Outcomes in terms of seizure control, for these variations on
or “around” the lateral temporal structures, to approach the medial
structures more specifically, include: transsylvian (1), transtemporal
this procedure, have been described by many groups and
(2), and subtemporal (3) have been notably similar. In surveys, Engel et al. [47]
reported seizure freedom before 1985 at 56 % (2,336 pts)
and 1986–1990 at 68 % (3,579 pts) (1 year or more follow-up).
posterior temporal lobe. The operation of Penfield and Wiebe et al. [48] performed a randomized trial of epilepsy
Baldwin [30] and Falconer [20] frequently used lateral tai- surgery (subtotal anterior temporal lobectomy) compared to
loring during awake craniotomy to avoid this language cor- medical treatment and reported that at 1 year 58 % of patients
tex. Ojemann et al. [31] emphasized and further developed in the surgical group were free of disabling seizures com-
this issue. pared to 8 % in the medical group (Class I evidence for sig-
For surgery under general anesthesia, more standardized nificant benefit). Engel et al. [36], in 2003, surveyed the
or measured resections were developed: the lateral neocortex surgical literature since 1990 and included 24 papers with
was resected approximately 5–6 cm from the temporal pole retrospective descriptive data and reported 67 % of 1,952
on the nondominant cortex (slightly less on the dominant patients were free of disabling seizures—emphasizing the
side) with mesial structures removed [32–34]. Although similarity of this result to the evidence from the Class I study
such margins if electrically mapped could include some pos- [48]. Seizure outcomes across the different methods of par-
itive language sites, experience indicates including some of tial temporal resection, as mentioned above, continue to
these sites with standardized resections is still safe for lan- appear similar [27].
guage outcome [29, 32, 35, 36]. These “standard” lateral The critical structure(s) for excision for successful treat-
resection margins have continued to decrease in recent ment of seizures remains uncertain despite many years of
decades. Frequently the superior temporal gyrus resection is attention. Although most evidence implicates the mesial
minimized [32, 37], while lateral resections of the other lat- structures (hippocampus, inferior amygdala, parahippocam-
eral temporal gyri (both dominant and nondominant) have pus), the correlation between their removal and outcome is
been considered at 4.5 or 3.5 cm [29, 37, 38] or even smaller not tight [27] (for review see Schramm et al. [27]). For
at about 2 cm [39]—although our understanding of the instance, resections that retained the hippocampus could pro-
effects of these variations remains incompletely understood. vide seizure control [49], but not as great as when the hip-
Other approaches have been described, including: the pocampus was also removed. Jack [50] reported that seizure
approach of Niemeyer [40] who described a lateral approach outcome did not seem related to the extent of removal of any
through the lateral temporal cortex into the ventricle, thus particular mesial structure, but to some damage to all based
permitting resection of mesial structures; transsylvian on post-op MRIs. Also based on post-op MRIs, Siegel [51]
approaches for mesial excision [25, 41]; and subtemporal reported with transsylvian mesial resections that seizure out-
approaches [42, 43] (Fig. 59.2). come appeared more related to the extent of removal of the
The development of intraoperative neuronavigation has parahippocampal gyrus than the amygdala or hippocampus.
lead to further refinement of the surgical approach [44] Wyler et al. [52], in a randomized study, reported that seizure
(Fig. 59.2). outcome was significantly better with surgery aimed at hip-
pocampal removal to the level of the posterior midbrain resection. Tanriverdi et al. [65] described that verbal memory
(69 % seizure-free at 1 year) compared to mesial resections decreases more with left-sided selective amygdalohippo-
only to the anterior midbrain (38 % at 1 year). Bonilha et al. campectomies compared to more standard left temporal
[53], from post-op MRIs, reported a significant correlation resections. On the right, nonverbal memory decreased espe-
between the extent of hippocampal removal and outcome cially after the standard temporal resections compared to the
and further improvement when combined with entorhinal selective resection.
cortex removal. Schramm et al. [54] in a randomized study, There is a general agreement that an important effect on
reported that 2.5 cm vs. a 3.5 cm hippocampal removal was neuropsychology after resections involves the functionality of
not associated with difference in seizure outcome (74 % vs. the resected tissue; for example, there are higher risks to mem-
73 % free of disabling seizures) (included post-op MRI con- ory associated with resections that include a normal-appearing
firmation of significantly differing resection volumes). hippocampus vs. sclerotic [66]. However, this remains an
It has been suggested the variety of resections utilized in evolving area—partly because of ongoing interest in defining
the anterior temporal lobe for MTLE may be effective by meaningful behavioral results that may vary between surgical
“severing a critical proportion of the connections” [55], nec- methods described above and radiosurgery [12].
essary to support a seizure. Although this circuitry is likely Complications can occur with any of these resective pro-
concentrated medially in the temporal lobe, its more wide- cedures, for example, diplopia from irritation of cranial
spread and possibly variable nature [56] may be why nerve III or IV, a stroke from injury of the arteries or veins in
20–40 % of surgically treated patients are not cured by sur- the basal cisterns or insula, and injury to the basal ganglia
gery and a consistently critical structure for removal has not superior to the amygdala (for further discussion, see [29, 38,
been identified. It should also be noted that long-term out- 67]). Postoperative dysphasia can occur but is usually tran-
comes need to be considered and have been reported to sient [29]. Lastly an upper quadrantanopsia is common from
decrease gradually at longer post-op periods [57], but further involvement of visual fibers variably sweeping around the
longer-term data is needed. Reported success rates of reop- temporal horn [68].
eration for continuing or recurrent seizures are variable but
most commonly below 50 %. Patient selection appears to be
important in this situation [58]. Hypothalamic Hamartoma
Since the effectiveness for seizures is similar for the dif-
ferent resective approaches, focus on other outcomes has Hypothalamus hamartomas (HH) are rare, congenital hetero-
also been of major interest. Wiebe [48] in their randomized topic lesions in the region of the third ventricle and tuber
study demonstrated a significant improvement in quality of cinereum that are intrinsically epileptogenic when closely
life (QOL) in the surgical arm. Engel et al. [36] reviewed and connected to the mammillary bodies [69]. The prevalence is
described that improved QOL measures were primarily 1 in 50,000–100,000 [70].
related to freedom from disabling seizures post-op. Other Patients classically present with isolated fits of ictal
evaluations have included neuropsychological testing and laughter (gelastic epilepsy) or a combination of gelastic and
this remains a major area of interest today. In a recent review other types of seizures during the first years of life that are
Sherman et al. [59] describe decreases in verbal memory and often particularly drug resistant from the onset [71].
naming after left-sided surgery as being the most consistent Although there are several classifications available for
change after temporal resections (different resection types these lesions, two clinical syndromes have been recognized:
included), but with patients usually not self-reporting precocious puberty and intractable epilepsy including gelas-
changes. Loss of visual memory was less common but simi- tic seizures. A significant number of children have both
lar after surgery on either side. problems. The natural history is unfavorable in the majority
Advantages in neuropsychological outcomes after more of patients because of behavioral symptoms and mental
selective resections compared to resections with generous decline; these include oppositional defiant disorder (83.3 %),
lateral removal have been reported [60–62], but this was not attention-deficit/hyperactivity disorder (75 %), high rates of
seen in other studies [63, 64]. Further complicating the situ- conduct disorder (33.3 %), speech retardation/learning
ation, there may be different effects of different procedures impairment (33.3 %), and anxiety and mood disorders
in the different hemispheres. For example, Helmstaedter (16.7 %). Significant rates of aggression have been noted
et al. [39] describe how left resections were worse for verbal with 58 % of the seizure patients meeting criteria for the
memory with selective amygdalohippocampectomy being affective subtype of aggression and 30.5 % having the preda-
more deleterious than an anterior neocortical temporal resec- tory aggression subtype [72].
tion plus mesial resection. Right resections were worse for Typically, medium/large sessile lesions within the third
figural memory, but on this side the anterior temporal resection ventricle are more likely to lead to the seizure disorder [73]
plus mesial resection was worse than the selective mesial than pedunculated lesions, which hang inferiorly from the
tuber cinereum and are more likely to result in central preco- radiofrequency obliteration [83–85], or magnetic resonance
cious puberty. Resection of pedunculated lesions causing stereotactic laser ablation [13], and also nonsurgical thera-
precocious puberty reverses the hormonal abnormalities. pies such as radiosurgery [86].
The first successful and safe removal of an HH was Removal, disconnection, or ablation of hypothalamic
reported by Paillas et al. in [74], but the interest for the sur- hamartomas affects the seizure and behavioral outcomes of
gical cure of this specific group of patients was primarily these often-desperate patients. The ideal approach to the
developed in the 1990s. The association between the HHs treatment of the patient with a hypothalamic hamartoma and
and epilepsy was obscured by the fairly generalized spike intractable epilepsy has yet to be defined [87–89].
and wave EEG patterns usually observed. Consequently,
patients underwent temporal and frontal resections without
seizure relief. The emphasis changed, however, due to the Surgical Techniques
information gleaned from ictal single-photon emission com-
puted tomography (iSPECT) and ictal depth recordings There is literature describing the success rate of resection of
from the lesion itself [75] implicating the epileptogenicity just the lesion (lesionectomy) vs. lesion resection plus sur-
of the hamartoma. High-quality magnetic resonance (MR) rounding tissue (that may demonstrate electrographic abnor-
imaging has strongly facilitated the diagnosis of HH in malities). Although lesionectomy is quite successful for
cases of severe epilepsy in children. Delalande classified seizure control, the excision of adjacent irritated cortex likely
HHs into four types: Type I has a horizontal implantation can provide further benefit—particularly in patients with a
plane inferior to the floor of the third ventricle; Type II longer history of seizures [90–92]. Disconnective procedures
within the third ventricle, with a vertical insertion plane; are based on the concept of interrupting the spreading of the
Type III is a combination of Types I and II; and Type IV epileptic discharge in the brain, by isolating the primary epi-
includes all giant hamartomas [76]. leptogenic zone. This is the basic concept for hemispherec-
Delalande et al. [76, 77] and Choi et al. [78] have empha- tomy, multilobar disconnection, hypothalamic hamartoma
sized that the goal of surgery is to resect and/or disconnect disconnection, corpus callosotomy, and multiple subpial
the hamartoma from the adjacent hypothalamus and to pre- transections.
serve the mammillary bodies (which may be malpositioned
and deformed), mammillothalamic tracts, tuber cinereum,
and hypothalamic nuclei. The spread of seizures occurs in Hemispherectomy and Hemispherotomy
part via the mammillary body. The surgery should ideally be
performed before the onset of secondary generalized epi- Dandy [93] first reported the anatomical hemispherectomy
lepsy [70, 79, 80]. Larger lesions might require more than in 1928 for the treatment of gliomas. In 1938, McKenzie [94]
one disconnection procedure, and some might require a mul- first reported its use in the treatment of epilepsy. Anatomical
tistep surgical approach. hemispherectomy was popular until 1966, but became more
The microsurgical resection in this critical area is attended selectively utilized because of postoperative complications,
by a significant risk of oculomotor palsy, hemiparesis, hypo- which could include superficial cerebral hemosiderosis and a
thalamic deficit, and/or visual field defect, and many epi- significant incidence of hydrocephalus [95, 96]. A major fac-
lepsy surgeons have therefore abandoned this type of tor leading to the development of hemosiderosis was the
approach [81]. Resections have often been subtotal and only large cavity created following removal of the hemisphere.
partially effective in controlling seizures, but still associated Hence, various surgical modifications have evolved as alter-
with significant complications. natives, attempting to minimize the removal of brain but
There are three typical microsurgical approaches: maximize the disconnection of white matter.
• Transcallosal interforniceal approach, entering the third One of the most popular techniques, the functional hemi-
ventricle from superiorly [82] spherectomy, was described by Rasmussen [97] in 1983,
• Pterional approach, exposing the inferior and lateral who modified the technique of the anatomical hemispherec-
aspect of the hypothalamus [76] tomy to remove part of the central and temporal regions
• Anterior midline approach, entering the third ventricle while disconnecting the rest. In the early 1990s, a new era in
through the lamina terminalis [72] functional hemispherectomy was initiated with the intro-
Comparing the three approaches, the transcallosal ante- duction of the hemispherotomy, with the aim of removing as
rior interforniceal offers the best chance of seizure freedom little brain as possible, and two different approaches were
with less morbidity than the pterional route. Concerns described almost simultaneously by Delalande [98] and
regarding the impairment of short-term memory in children, Villemure [99, 100]. Since the first description of those tech-
who are cognitively intact, have promoted the exploration niques, a number of modifications have been made to reduce
of surgical alternatives, e.g., endoscopic disconnection, the number of complications [70]. The hemispherotomy
techniques described include the vertical parasagittal Multiple Subpial Transections

hemispherotomy by Delalande and associates [98], the
periinsular hemispherotomy by Villemure and Mascott First described by Morrel in 1969, the multiple subpial tran-
[100–102], and the transsylvian keyhole hemispherotomy sections (MST) procedure was designed to treat patients with
by Schramm et al. [103]. epileptogenic zones in functionally critical cortical areas
Other Disconnective Techniques where respective surgery could cause unacceptable neuro-
Daniel et al. [104] reported the posterior quadrantectomy logical deficits [114]. The rationale is based on the principle
for patients with refractory temporo-parieto-occipital epi- that vertical columns are essential in the propagation of nor-
lepsy. Three variants were utilized in this series, namely, (1) mal cortical activity from cortex to subcortical structures.
anatomical posterior quadrantectomy, which entails the Therefore, the transection of horizontal cortical-cortical
removal of the temporal, parietal, and occipital lobe; (2) interneuronal connections will interrupt local synchroniza-
functional posterior quadrantectomy, where an extended tion of epileptogenic activity and the spread of the epileptic
temporal lobectomy is performed and then parieto-occipital discharges [115].
lobe is disconnected; and (3) periinsular posterior quadran- Although it is a well-known technique, MST is the least
tectomy; in this surgery no lobe is resected, and the entire frequently performed procedure in epilepsy surgery (on the
temporo-parieto-occipital is disconnected but remains viable order of 0.6 %). It is difficult to assess the effectiveness of
because of the preservation of vessels supplying or draining MST because the majority of the reports used MST in com-
the disconnected lobes [105]. They reported excellent results. bination with resective surgery [116–118]. The use of MST
Chabardes et al. [106] published a series of 47 patients with as a stand-alone procedure has been reported by Schramm
temporal disconnection procedures guided by neuronaviga- [119], Smith [120], and Whisler [121] with a seizure-free
tion. Outcomes and complications are within the range of outcome of 5 %, 37.5 %, and 63 %, respectively.
that expected with resective procedures described in a prior Tellez-Zenteno et al. [122] in a meta-analysis study con-
section. cluded that MST has the lowest rate of long-term seizure-
free outcome (16 %) among all epilepsy procedures.
Corpus Callosotomy
Electrical Stimulation for Epilepsy
First described in 1940 [107], corpus callosotomy has been
reserved for the treatment of generalized epilepsy or partial Advantages of electrical stimulation therapy for epilepsy
epilepsy with secondary generalization where the onset include reversibility, adjustability and that inaccessible and
cannot be localized. Corpus callosotomy is particularly multiple targets may be influenced. Presumably the stimula-
effective for the drop seizures of Lennox-Gastaut syndrome tion disrupts the abnormal electrical “buildup” of the seizure,
[108], a severe childhood epilepsy disorder characterized but both the mechanism and circuitry remain unclear.
by encephalopathy and multiple, often intractable, seizure
types. The drop attack is the most frequently recognizable
seizure type in this patient population and is also the most Vagal Nerve Stimulation
dangerous physically, thus severely limiting quality of life.
The diagnosis is confirmed by electroencephalography, for Vagal nerve stimulation (VNS) had been known to amelio-
which the classic pattern is a slow 2.5 Hz generalized spike rate experimental seizure activity since the 1950s [123],
and wave [109]. Seizure improvement ranges between 50 and a series of clinical studies supported its utility in the
and 81 % with complete callosotomy [110–112]. The com- mid- and late 1990s. In patients with primarily focal sei-
plete callosotomy is a palliative procedure reserved for zures, after VNS implant, patients were randomized
nonverbal patients with severe intractable epilepsy because between stimulation that was either “high” (“typical” and
it often produces a hemisphere disconnection syndrome felt to be active) or “low” (low frequency and hypothesized
that can result in disabling symptoms of intermanual con- to be ineffective). Blinding was not likely complete as some
flict [113]. The 2/3 anterior partial callosotomy does not patients and observers may have intuited the group they
usually have significant cognitive consequences. It is were in. In the initial randomized study (n = 114), there was
because of the potential for the development of this discon- a 24.5 % reduction of seizures in the “high” stimulation
nection syndrome that partial callosotomy is usually under- group at 14 weeks—6 % in control [124]. In the 2nd study
taken initially, and the section can be completed if there is (n = 196) after 3 months, the seizure frequency declined 28 %
an inadequate seizure control. in the “high” stimulation group—vs. 15 % in the “low” [125].
DeGiorgio et al. [126] reported on patients from this latter structure thus aborting the seizure [131]. An example of this
trial who were followed unblinded with “high” stimulation type of system has been in clinical trial, which demonstrated
and reported a 45 % decrease at 1 year. Over 400 patients a significant benefit [134]. In this trial, 191 patients were
from early trials [127] were followed, and seizure reduc- implanted (97 randomized to stimulation and 94 control) and
tions of 35 % at 1 year, 44 % at 2 years, and 44 % at 3 years assessed in a blinded fashion over a 12-week period. There
were reported. Approval for the device was received for was a 37.9 % decrease in seizures in the stimulated patients
adjunctive treatment for intractable focal seizures in the vs. a 17.3 % decrease in the placebo patients. Benefits com-
United States in the late 1990s. pared to baseline were sustained after unblinding. While
Beneficial effects appear to be maintained in the longer promising this treatment remains under assessment. Other
term (for at least a decade) according to Elliott et al. [128], experimental neuromodulation strategies include cell graft-
with seizure reduction rates reported in excess of 50 % at a ing, gene therapy, focal cooling, and drug delivery [131].
mean of 5 years as part of medical management in a com-
prehensive program. Although only approved for use in Acknowledgment We want to thank Sean Huckins for his help with
patients over 12 in the USA for focal seizures, there is evi- illustrations.
dence for general usefulness in generalized seizures [129].
It is approved in Europe for this indication in pediatric
patients [130]. The effect presumably occurs by retrograde References
stimulation thru the vagus to the brainstem and then ros-
1. Treiman DM. Management of refractory complex partial seizures:
trally into thalamus and limbic system. A number of side
current state of the art. Neuropsychiatr Dis Treat. 2010;6:
effects can occur (e.g., voice alteration, coughing) but these 297–308. Epub 2010/07/16.
are generally tolerated with adjustment of stimulation 2. Olafsson E, Ludvigsson P, Gudmundsson G, Hesdorffer D,
intensity and time. Kjartansson O, Hauser WA. Incidence of unprovoked seizures and
epilepsy in Iceland and assessment of the epilepsy syndrome clas-
sification: a prospective study. Lancet Neurol. 2005;4(10):627–34.
3. Aicardi J. Evolution of epilepsy surgery in childhood: the neurolo-
Brain Stimulation gist’s point of view. Epileptic Disord. 1999;1(4):243–7. Epub
2000/08/11.
4. Bourgeois M, Di Rocco F, Roujeau T, Boddaert N, Lelouch-
Electrical stimulation of the brain has also received consider-
Tubiana A, Varlet P, et al. Epilepsy and focal lesions in children.
able interest as a treatment for epilepsy. Studies in animals Surgical management. Neurochirurgie. 2008;54(3):362–5. Epub
have supported the concept that deep brain stimulation 2008/04/29. Epilepsie et lesions focales chez l’enfant. Traitement
(DBS) can be effective and clinical studies have raised hopes chirurgical.
5. Shields WD. Surgical treatment of refractory epilepsy. Curr Treat
for a variety of effective stimulation sites [131, 132].
Options Neurol. 2004;6(5):349–56. Epub 2004/07/29.
Currently most attention is focused on electrical stimula- 6. Akai T, Otsubo H, Pang EW, Rutka JT, Chitoku S, Weiss SK, et al.
tion of the anterior nucleus of the thalamus for treatment of Complex central cortex in pediatric patients with malformations
refractory epilepsy (SANTE trial—Fisher R. et al. [133]). of cortical development. J Child Neurol. 2002;17(5):347–52.
Epub 2002/08/02.
The nucleus has extensive interconnections with the limbic
7. Berg AT, Vickrey BG, Langfitt JT, Sperling MR, Walczak TS,
system. In the multicenter trial, data from 110 patients with Shinnar S, et al. The multicenter study of epilepsy surgery: recruit-
intractable partial seizures (± secondary generalization and ment and selection for surgery. Epilepsia. 2003;44(11):1425–33.
not amenable to resection) were randomized into low (pre- Epub 2003/11/26.
8. Tovar-Spinoza ZS, Ochi A, Rutka JT, Go C, Otsubo H. The role of
sumed ineffective) or high (hypothesized to be effective)
magnetoencephalography in epilepsy surgery. Neurosurg Focus.
stimulation of the anterior nucleus of the thalamus. At the 2008;25(3):E16. Epub 2008/09/02.
end of the blinded 3-month period, seizure reduction was 9. Ochi A, Otsubo H. Magnetoencephalography-guided epilepsy sur-
45 % in the high stimulation group compared to 14 % in the gery for children with intractable focal epilepsy: SickKids experi-
ence. Int J Psychophysiol. 2008;68(2):104–10. Epub 2008/03/04.
low stimulation group. After 3 months all patients received
10. Benifla M, Otsubo H, Ochi A, Snead 3rd OC, Rutka JT. Multiple
high stimulation and followed at least 13 months with 54 % subpial transections in pediatric epilepsy: indications and out-
having greater than 50 % reduction of seizures. Complications comes. Childs Nerv Syst. 2006;22(8):992–8. Epub 2006/06/21.
were within an acceptable range for typical DBS procedures 11. Benifla M, Rutka JT, Logan W, Donner EJ. Vagal nerve stimula-
tion for refractory epilepsy in children: indications and experience
and the underlying condition [133]. It was shown to be most
at The Hospital for Sick Children. Childs Nerv Syst.
effective for seizures of temporal and frontal origin. The 2006;22(8):1018–26. Epub 2006/07/04.
treatment is currently approved in Europe and Canada. 12. Quigg M, Rolston J, Barbaro NM. Radiosurgery for epilepsy:
Another approach for electrical stimulation for epilepsy clinical experience and potential antiepileptic mechanisms.
Epilepsia. 2012;53(1):7–15. Epub 2011/12/24.
consists of “responsive neurologic stimulation.” Electrodes
13. Curry DJ, Gowda A, McNichols RJ, Wilfong AA. MR-guided ste-
that sense seizure onset are implanted along with a stimulat- reotactic laser ablation of epileptogenic foci in children. Epilepsy
ing electrode targeted to the focus or seizure-modulating Behav. 2012;24(4):408–14. Epub 2012/06/13.
14. Jackson J. Case of epilepsy with tasting movements and “dreaming Subcommittee of the American Academy of Neurology, in
state”-very small patch of softening in the left uncinate gyrus. association with the American Epilepsy Society and the American
Brain. 1898;21:580–90. Association of Neurological Surgeons. Neurology. 2003;60(4):
15. Gibbs F, Gibbs E, Lennox W. Cerebral dysrhythmias of epilepsy. 538–47. Epub 2003/02/26.
Arch Neurol Psychiatry. 1938;39:298–314. 37. Spencer DD, Spencer SS, Mattson RH, Williamson PD, Novelly
16. Gastaut H, Terzian H, Naquet R, Luschnat K. Correlations RA. Access to the posterior medial temporal lobe structures in the
between automatism in temporal crises and electroencephalo- surgical treatment of temporal lobe epilepsy. Neurosurgery.
graphic phenomena which accompany them. Rev Neurol (Paris). 1984;15(5):667–71. Epub 1984/11/01.
1952;86(6):678–82. 38. Abson-Kraemer D, Spencer D. Temporal lobectomy under gen-
17. Jasper H, Kershman J. Electroencephalographic classification of eral anesthesia. In: Starr P, Barbaro N, Larson P, editors. Functional
the epilepsies. Arch Neurol Psychiatry. 1941;45:905–43. neurosurgery (neurosurgical operative atlas). New York: Thieme;
18. Feindel W, Penfield W. Localization of discharge in temporal lobe 2008. p. 29–35.
automatism. AMA Arch Neurol Psychiatry. 1954;72(5):603–30. 39. Helmstaedter C, Richter S, Roske S, Oltmanns F, Schramm J,
19. Penfield W, Baldwin M. Temporal lobe seizures and the technic of Lehmann TN. Differential effects of temporal pole resection with
subtotal temporal lobectomy. Ann Surg. 1952;136(4):625–34. amygdalohippocampectomy versus selective amygdalohippo-
Epub 1952/10/01. campectomy on material-specific memory in patients with mesial
20. Falconer M. Discussion on the surgery of the temporal lobe epi- temporal lobe epilepsy. Epilepsia. 2008;49(1):88–97. Epub
lepsy: surgical and pathological aspects. Proc R Soc Med. 1953; 2007/10/19.
46:971–4. 40. Niemeyer P. The transventricular amygdala-hippocampectomy in
21. Mathern G, Wilson C, Beck H. Hippocampal Sclerosis. In: Engel temporal lobe epilepsy. In: Baldwin M, Bailey P, editors. Temporal
JJ, Pedley TA, editors. Epilepsy: a comprehensive textbook. lobe epilepsy. Springfield IL: Charles C Thomas; 1958.
New York: Wolters Kluwer-Lippincott Williams & Wilkins; 2008. p. 461–82.
p. 121–36. 41. Wieser HG, Yasargil MG. Selective amygdalohippocampectomy
22. Li LM, Cendes F, Andermann F, Watson C, Fish DR, Cook MJ, as a surgical treatment of mesiobasal limbic epilepsy. Surg Neurol.
et al. Surgical outcome in patients with epilepsy and dual pathol- 1982;17(6):445–57. Epub 1982/06/01.
ogy. Brain. 1999;122(Pt 5):799–805. Epub 1999/06/04. 42. Hori T, Yamane F, Ochiai T, Kondo S, Shimizu S, Ishii K, et al.
23. Engel Jr J. Surgical treatment for epilepsy: too little, too late? Selective subtemporal amygdalohippocampectomy for refractory
JAMA. 2008;300(21):2548–50. Epub 2008/12/04. temporal lobe epilepsy: operative and neuropsychological out-
24. Kemmotsu N, Girard HM, Bernhardt BC, Bonilha L, Lin JJ, comes. J Neurosurg. 2007;106(1):134–41. Epub 2007/01/24.
Tecoma ES, et al. MRI analysis in temporal lobe epilepsy: cortical 43. Hori T, Tabuchi S, Kurosaki M, Kondo S, Takenobu A, Watanabe
thinning and white matter disruptions are related to side of seizure T. Subtemporal amygdalohippocampectomy for treating medi-
onset. Epilepsia. 2011;52(12):2257–66. Epub 2011/10/07. cally intractable temporal lobe epilepsy. Neurosurgery.
25. Yasargil MG, Wieser HG, Valavanis A, von Ammon K, Roth 1993;33(1):50–6; discussion 6–7; Epub 1993/07/01.
P. Surgery and results of selective amygdala-hippocampectomy in 44. Olivier A. Transcortical selective amygdalohippocampectomy in
one hundred patients with nonlesional limbic epilepsy. Neurosurg temporal lobe epilepsy. Can J Neurol Sci. 2000;27 Suppl 1:S68–
Clin N Am. 1993;4(2):243–61. Epub 1993/04/01. 76; discussion S92–6.
26. Fahoum F, Lopes R, Pittau F, Dubeau F, Gotman J. Widespread 45. Liscak R, Malikova H, Kalina M, Vojtech Z, Prochazka T, Marusic
epileptic networks in focal epilepsies: EEG-fMRI study. Epilepsia. P, et al. Stereotactic radiofrequency amygdalohippocampectomy
2012;53(9):1618–27. Epub 2012/06/14. in the treatment of mesial temporal lobe epilepsy. Acta Neurochir.
27. Schramm J. Temporal lobe epilepsy surgery and the quest for opti- 2010;152(8):1291–8. Epub 2010/04/03.
mal extent of resection: a review. Epilepsia. 2008;49(8):1296– 46. Malikova H, Liscak R, Vojtech Z, Prochazka T, Vymazal J,
307. Epub 2008/04/16. Vladyka V, et al. Stereotactic radiofrequency amygdalohippocam-
28. Hill D, Falconer MA, Pampiglione G, Liddell DW. Discussion on pectomy: does reduction of entorhinal and perirhinal cortices
the surgery of temporal lobe epilepsy. Proc R Soc Med. influence good clinical seizure outcome? Epilepsia.
1953;46(11):965–76. Epub 1953/11/01. 2011;52(5):932–40. Epub 2011/04/02.
29. Bingaman W, Najm I. Standard temporal lobectomy. In: Winn H, 47. Engel JJ, Van Ness P, Rasmussen TB. In: Engel JJ, editor. Surgical
editor. Youmans neurological surgery. Philadelphia: Elsevier; 2012. treatment of the epilepsies. New York: Raven; 1993. p. 609–22.
30. Penfield W, Jasper H. Epilepsy and the functional anatomy of the 48. Wiebe S, Blume WT, Girvin JP, Eliasziw M. A randomized, con-
human brain. Boston: Little, Brown and Company; 1954. trolled trial of surgery for temporal-lobe epilepsy. N Engl J Med.
31. Ojemann G, Ojemann J, Lettich E, Berger M. Cortical language 2001;345(5):311–8. Epub 2001/08/04.
localization in left, dominant hemisphere. An electrical stimula- 49. Olivier A. Relevance of removal of limbic structures in surgery for
tion mapping investigation in 117 patients. J Neurosurg. temporal lobe epilepsy. Can J Neurol Sci. 1991;18(4 Suppl):628–
1989;71(3):316–26. 35. Epub 1991/11/01.
32. Falconer M, Hill D, Meyer A, Mitchell W, Pond DA. Treatment of 50. Jack Jr CR, Sharbrough FW, Marsh WR. Use of MR imaging for
temporal lobe epilepsy by temporal lobectomy: a survey of find- quantitative evaluation of resection for temporal lobe epilepsy.
ings and results. Lancet. 1955;23:827–35. Radiology. 1988;169(2):463–8. Epub 1988/11/01.
33. Falconer M. Anterior temporal lobectomy for epilepsy. In: Logue V, 51. Siegel AM, Wieser HG, Wichmann W, Yasargil GM. Relationships
editor. Neurosurgery. Philadelphia: Lippincott; 1971. p. 142–9. between MR-imaged total amount of tissue removed, resection
34. Olivier A. Commentary: cortical resections. In: Jr EJ, editor. Surgical scores of specific mediobasal limbic subcompartments and clini-
treatment of epilepsies. New York: Raven; 1987. p. 405–16. cal outcome following selective amygdalohippocampectomy.
35. Schaffler L, Luders HO, Morris 3rd HH, Wyllie E. Anatomic dis- Epilepsy Res. 1990;6(1):56–65. Epub 1990/05/01.
tribution of cortical language sites in the basal temporal language 52. Wyler AR, Hermann BP, Somes G. Extent of medial temporal
area in patients with left temporal lobe epilepsy. Epilepsia. resection on outcome from anterior temporal lobectomy: a ran-
1994;35(3):525–8. Epub 1994/05/01. domized prospective study. Neurosurgery. 1995;37(5):982–90;
36. Engel Jr J, Wiebe S, French J, Sperling M, Williamson P, Spencer discussion 90–1; Epub 1995/11/01.
D, et al. Practice parameter: temporal lobe and localized neocorti- 53. Bonilha L, Yasuda CL, Rorden C, Li LM, Tedeschi H, de Oliveira E,
cal resections for epilepsy: report of the Quality Standards et al. Does resection of the medial temporal lobe improve the
outcome of temporal lobe epilepsy surgery? Epilepsia. 2007; hypothalamic hamartomas. Acta Neurochir Suppl. 2004;91:
48(3):571–8. Epub 2007/03/01. 33–50. Epub 2005/02/15.
54. Schramm J, Lehmann TN, Zentner J, Mueller CA, Scorzin J, 70. De Ribaupierre S, Delalande O. Hemispherotomy and other dis-
Fimmers R, et al. Randomized controlled trial of 2.5-cm versus 3.5- connective techniques. Neurosurg Focus. 2008;25(3):E14. Epub
cm mesial temporal resection—Part 2: volumetric resection extent 2008/09/02.
and subgroup analyses. Acta Neurochir. 2011;153(2):221–8. 71. Parvizi J, Le S, Foster BL, Bourgeois B, Riviello JJ, Prenger E,
55. Abosch A, Bernasconi N, Boling W, Jones-Gotman M, Poulin N, Saper C, Kerrigan JF. Gelastic epilepsy and hypothalamic hamar-
Dubeau F, et al. Factors predictive of suboptimal seizure control tomas: neuroanatomical analysis of brain lesions in 100 patients.
following selective amygdalohippocampectomy. J Neurosurg. Brain. 2011;134:2960–8.
2002;97(5):1142–51. Epub 2002/11/27. 72. Maixner W. Hypothalamic hamartomas—clinical, neuropatho-
56. Spencer SS. Neural networks in human epilepsy: evidence of and logical and surgical aspects. Childs Nerv Syst. 2006;22(8):867–
implications for treatment. Epilepsia. 2002;43(3):219–27. Epub 73. Epub 2006/06/10.
2002/03/22. 73. Valdueza JM, Cristante L, Dammann O, Bentele K, Vortmeyer A,
57. McIntosh AM, Kalnins RM, Mitchell LA, Fabinyi GC, Briellmann Saeger W, et al. Hypothalamic hamartomas: with special reference
RS, Berkovic SF. Temporal lobectomy: long-term seizure out- to gelastic epilepsy and surgery. Neurosurgery. 1994;34(6):949–
come, late recurrence and risks for seizure recurrence. Brain. 58; discussion 58; Epub 1994/06/01.
2004;127(Pt 9):2018–30. Epub 2004/06/25. 74. Paillas JE Roger J, Toga M. Hamartoma of the hypothalamus.
58. Jehi LE, Silveira DC, Bingaman W, Najm I. Temporal lobe epi- Clinical, radiological and histological study. Results of excision.
lepsy surgery failures: predictors of seizure recurrence, yield of Rev Neurol (Paris). 1969;120:177–94.
reevaluation, and outcome following reoperation. J Neurosurg. 75. Andermann F, Arzimanoglou A, Berkovic SF. Hypothalamic ham-
2010;113(6):1186–94. Epub 2010/09/14. artoma and epilepsy: the pathway of discovery. Epileptic Disord.
59. Sherman EM, Wiebe S, Fay-McClymont TB, Tellez-Zenteno J, 2003;5(4):173–5. Epub 2004/02/21.
Metcalfe A, Hernandez-Ronquillo L, et al. Neuropsychological 76. Delalande O, Fohlen M. Disconnecting surgical treatment of
outcomes after epilepsy surgery: systematic review and pooled hypothalamic hamartoma in children and adults with refractory
estimates. Epilepsia. 2011;52(5):857–69. Epub 2011/03/24. epilepsy and proposal of a new classification. Neurol Med Chir
60. Hermann B, Davies K, Foley K, Bell B. Visual confrontation nam- (Tokyo). 2003;43(2):61–8. Epub 2003/03/12.
ing outcome after standard left anterior temporal lobectomy with 77. Dorfmuller G, Fohlen M, Bulteau C, Delalande O. Surgical dis-
sparing versus resection of the superior temporal gyrus: a random- connection of hypothalamic hamartomas. Neurochirurgie.
ized prospective clinical trial. Epilepsia. 1999;40(8):1070–6. 2008;54(3):315–9. Epub 2008/05/03. Deconnexion chirurgicale
Epub 1999/08/17. des hamartomes hypothalamiques.
61. Clusmann H, Schramm J, Kral T, Helmstaedter C, Ostertun B, 78. Choi JU, Yang KH, Kim TG, Chang JH, Chang JW, Lee BI, et al.
Fimmers R, et al. Prognostic factors and outcome after different Endoscopic disconnection for hypothalamic hamartoma with
types of resection for temporal lobe epilepsy. J Neurosurg. intractable seizure. Report of four cases. J Neurosurg. 2004;100(5
2002;97(5):1131–41. Epub 2002/11/27. Suppl Pediatrics):506–11. Epub 2004/08/04.
62. Paglioli E, Palmini A, Portuguez M, Paglioli E, Azambuja N, da 79. Ng YT, Rekate HL, Prenger EC, Wang NC, Chung SS, Feiz-Erfan
Costa JC, et al. Seizure and memory outcome following temporal I, et al. Endoscopic resection of hypothalamic hamartomas for
lobe surgery: selective compared with nonselective approaches for refractory symptomatic epilepsy. Neurology. 2008;70(17):1543–
hippocampal sclerosis. J Neurosurg. 2006;104(1):70–8. Epub 8. Epub 2008/04/23.
2006/03/03. 80. Rosenfeld JV. The evolution of treatment for hypothalamic hamar-
63. Wolf RL, Ivnik RJ, Hirschorn KA, Sharbrough FW, Cascino GD, toma: a personal odyssey. Neurosurg Focus. 2011;30(2):E1. Epub
Marsh WR. Neurocognitive efficiency following left temporal 2011/03/05.
lobectomy: standard versus limited resection. J Neurosurg. 81. Stewart L, Steinbok P, Daaboul J. Role of surgical resection in the
1993;79(1):76–83. Epub 1993/07/01. treatment of hypothalamic hamartomas causing precocious
64. Jones-Gotman M, Zatorre RJ, Olivier A, Andermann F, Cendes puberty. Report of six cases. J Neurosurg. 1998;88:340–5.
F, Staunton H, et al. Learning and retention of words and 82. Harvey AS, Freeman JL, Berkovic SF, Rosenfeld JV. Transcallosal
designs following excision from medial or lateral temporal- resection of hypothalamic hamartomas in patients with intractable
lobe structures. Neuropsychologia. 1997;35(7):963–73. Epub epilepsy. Epileptic Disord. 2003;5(4):257–65. Epub 2004/02/21.
1997/07/01. 83. Fujimoto Y, Kato A, Saitoh Y, Ninomiya H, Imai K, Sakakibara
65. Tanriverdi T, Dudley RW, Hasan A, Al Jishi A, Al Hinai Q, Poulin RI, et al. Stereotactic radiofrequency ablation for sessile hypotha-
N, et al. Memory outcome after temporal lobe epilepsy surgery: lamic hamartoma with an image fusion technique. Acta
corticoamygdalohippocampectomy versus selective amygdalo- Neurochirurgica (Wein). 2003;145(8):697–700; discussion 700–
hippocampectomy. J Neurosurg. 2010;113(6):1164–75. Epub 1. Epub 2003/10/02.
2009/12/08. 84. Khan S, Wright I, Javed S, Sharples P, Jardine P, Carter M, et al.
66. Seidenberg M, Hermann B, Wyler AR, Davies K, Dohan Jr FC, High frequency stimulation of the mamillothalamic tract for the
Leveroni C. Neuropsychological outcome following anterior tem- treatment of resistant seizures associated with hypothalamic ham-
poral lobectomy in patients with and without the syndrome of artoma. Epilepsia. 2009;50(6):1608–11. Epub 2009/02/27.
mesial temporal lobe epilepsy. Neuropsychology. 1998;12(2):303– 85. Parrent AG. Stereotactic radiofrequency ablation for the treatment
16. Epub 1998/04/29. of gelastic seizures associated with hypothalamic hamartoma.
67. Spencer S, Huh L. Outcomes of epilepsy surgery in adults and Case report. J Neurosurg. 1999;91(5):881–4. Epub 1999/11/30.
children. Lancet Neurol. 2008;7(6):525–37. Epub 2008/05/20. 86. Drees C, Chapman K, Prenger E, Baxter L, Maganti R, Rekate H,
68. Nilsson D, Malmgren K, Rydenhag B, Frisen L. Visual field et al. Seizure outcome and complications following hypothalamic
defects after temporal lobectomy—comparing methods and ana- hamartoma treatment in adults: endoscopic, open, and Gamma
lysing resection size. Acta Neurol Scand. 2004;110(5):301–7. Knife procedures. J Neurosurg. 2012;117(2):255–61. Epub
Epub 2004/10/13. 2012/06/12.
69. Regis J, Hayashi M, Eupierre LP, Villeneuve N, Bartolomei F, 87. Unger F, Schrottner O, Haselsberger K, Korner E, Ploier R, Pendl
Brue T, et al. Gamma knife surgery for epilepsy related to G. Gamma knife radiosurgery for hypothalamic hamartomas in
patients with medically intractable epilepsy and precocious Deconnexion du lobe temporal dans les epilepsies temporales
puberty. Report of two cases. J Neurosurg. 2000;92(4):726–31. pharmacoresistantes: techniques, complications et resultats.
Epub 2000/04/13. 107. Vanwagenen WP, Hereen RY. Surgical division of commissural
88. Regis J, Bartolomei F, de Toffol B, Genton P, Kobayashi T, Mori Y, pathways in the corpus callosum: relation to spread of an epileptic
et al. Gamma knife surgery for epilepsy related to hypothalamic attack. Arch Neurol Psychiatry. 1940;44:740–59.
hamartomas. Neurosurgery. 2000;47(6):1343–51. discussion 51–2; 108. Gates J, Rosenfeld W. Maxwell Rea. Response of multiple seizure
Epub 2000/12/29. types to corpus callosum section. Epilepsia. 1987;28:28–34.
89. Rekate HL, Feiz-Erfan I, Ng YT, Gonzalez LF, Kerrigan 109. Vanstraten A, Ng Y. Update on the management of lennox-gastaut
JF. Endoscopic surgery for hypothalamic hamartomas causing syndrome. Pediatr Neurol. 2012;47(3):153–61.
medically refractory gelastic epilepsy. Childs Nerv Syst. 2006; 110. Hodaie M, Musharbash A, Otsubo H, Snead 3rd OC, Chitoku S,
22:874–80. Ochi A, et al. Image-guided, frameless stereotactic sectioning of
90. Cohen DS, Zubay GP, Goodman RR. Seizure outcome after lesio- the corpus callosum in children with intractable epilepsy. Pediatr
nectomy for cavernous malformations. J Neurosurg. 1995;83(2): Neurosurg. 2001;34(6):286–94. Epub 2001/07/17.
237–42. Epub 1995/08/01. 111. Rahimi SY, Park YD, Witcher MR, Lee KH, Marrufo M, Lee
91. Rassi-Neto A, Ferraz FP, Campos CR, Braga FM. Patients with MR. Corpus callosotomy for treatment of pediatric epilepsy in
epileptic seizures and cerebral lesions who underwent lesionec- the modern era. Pediatr Neurosurg. 2007;43(3):202–8. Epub
tomy restricted to or associated with the adjacent irritative area. 2007/04/06.
Epilepsia. 1999;40(7):856–64. Epub 1999/07/14. 112. Wong TT, Kwan SY, Chang KP, Hsiu-Mei W, Yang TF, Chen YS,
92. Kim SK, Wang KC, Hwang YS, Kim KJ, Cho BK. Intractable epi- et al. Corpus callosotomy in children. Childs Nerv Syst. 2006;
lepsy associated with brain tumors in children: surgical modality 22(8):999–1011. Epub 2006/07/11.
and outcome. Childs Nerv Syst. 2001;17(8):445–52. Epub 113. Reeves A, Risse G. Neurological effects of callosotomy. In:
2001/08/18. Reeves A, Roberts D, editors. Epilepsy and the Corpus Callosum
93. Dandy W. Removal of right hemisphere for certain tumors with 2. New York: Plenum Press; 1995. p. 241–51.
hemiplegia. Preliminary report. JAMA. 1928;90:823–5. 114. Morrell F, Whisler W, Bleck T. Multiple subpial transection: a new
94. McKenzie K. The present status of a patient who had the right approach to the surgical treatment of focal epilepsy. J Neurosurg.
cerebral hemisphere removed. JAMA. 1938;111:168–83. 1989;70:231–9.
95. Wilson P. Cerebral hemispherectomy for infantile hemiplegia. 115. Tovar-Spinoza ZS, Rutka J. Multiple subpial transections in chil-
A report of 50 cases. Brain. 1970;93:147–80. dren with refractory epilepsy. In: Cataltepe O, Jallo GI, editors.
96. Oppenheimer D, Griffith H. Persistent intracranial bleeding as a Pediatric epilepsy surgery. New York: Thieme; 2010. p. 268–73.
complication of hemispherectomy. J Neurol Neurosurg Psychiatry. 116. Shimizu T, Maehara T, Hino T, Komori T, Shimizu H, Yagishita A,
1966;29:229–40. et al. Effect of multiple subpial transection on motor cortical excit-
97. Rasmussen T. Hemispherectomy for seizures revisited. Can J ability in cortical dysgenesis. Brain. 2001;124(Pt 7):1336–49.
Neurol Sci. 1983;10:71–8. Epub 2001/06/16.
98. Delalande O, Pinard J, Basdevant C, Gauthe M, Plouin P, Dulac O. 117. Blount JP, Langburt W, Otsubo H, Chitoku S, Ochi A, Weiss S,
Hemispherotomy: a new procedure for central disconnection. et al. Multiple subpial transections in the treatment of pediatric
Epilepsia. 1992;33 Suppl 3:99–100. epilepsy. J Neurosurg. 2004;100(2 Suppl Pediatrics):118–24.
99. Villemure JG, Rasmussen T. Functional hemispherectomy in Epub 2004/02/05.
children. Neuropediatrics. 1993;24(1):53–5. Epub 1993/02/01. 118. Spencer SS, Schramm J, Wyler A, O’Connor M, Orbach D, Krauss
100. Villemure JG, Mascott CR. Peri-insular hemispherotomy: surgical G, et al. Multiple subpial transection for intractable partial epi-
principles and anatomy. Neurosurgery. 1995;37(5):975–81. Epub lepsy: an international meta-analysis. Epilepsia. 2002;43(2):
1995/11/01. 141–5. Epub 2002/03/21.
101. Villemure JG, Vernet O, Delalande O. Hemispheric disconnec- 119. Schramm J, Aliashkevich AF, Grunwald T. Multiple subpial transec-
tion: callosotomy and hemispherotomy. Adv Tech Stand tions: outcome and complications in 20 patients who did not undergo
Neurosurg. 2000;26:25–78. Epub 2000/09/21. resection. J Neurosurg. 2002;97(1):39–47. Epub 2002/07/24.
102. Villemure JG, Meagher-Villemure K, Montes JL, Farmer JP, 120. Smith MC, Byrne R. Multiple subpial transection in neocortical
Broggi G. Disconnective hemispherectomy for hemispheric epilepsy: Part I. Adv Neurol. 2000;84:621–34. Epub 2000/11/25.
dysplasia. Epileptic Disord. 2003;5 Suppl 2:S125–30. Epub 121. Whisler W. Multiple subpial transection. Tech Neurosurg.
2003/11/18. 1995;1:40–4.
103. Schramm J, Kral T, Clusmann H. Transsylvian keyhole functional 122. Tellez-Zenteno JF, Dhar R, Wiebe S. Long-term seizure outcomes
hemispherectomy. Neurosurgery. 2001;49(4):891–900; discussion following epilepsy surgery: a systematic review and meta-analysis.
900–1. Epub 2001/09/21. Brain. 2005;128(Pt 5):1188–98. Epub 2005/03/11.
104. Daniel RT, Meagher-Villemure K, Farmer JP, Andermann F, 123. Zanchetti A, Wang SC, Moruzzi G. The effect of vagal afferent
Villemure JG. Posterior quadrantic epilepsy surgery: technical stimulation on the EEG pattern of the cat. Electroencephalogr Clin
variants, surgical anatomy, and case series. Epilepsia. Neurophysiol. 1952;4(3):357–61. Epub 1952/08/01.
2007;48(8):1429–37. Epub 2007/04/20. 124. A randomized controlled trial of chronic vagus nerve stimulation
105. Daniel R, Babu K, Jacob R, Villemure J. Posterior quadrantic for treatment of medically intractable seizures. The Vagus Nerve
resection and disconnection. In: Cataltepe O, Jallo GI, editors. Stimulation Study Group. Neurology. 1995;45(2):224–30. Epub
Pediatric epilepsy surgery: preoperative assessment and surgical 1995/02/01
treatment. New York: Thieme Medical Publishers, Inc; 2010. 125. Handforth A, DeGiorgio CM, Schachter SC, Uthman BM,
p. 196–204. Naritoku DK, Tecoma ES, et al. Vagus nerve stimulation therapy
106. Chabardes S, Minotti L, Hamelin S, Hoffmann D, Seigneuret E, for partial-onset seizures: a randomized active-control trial.
Carron R, et al. Temporal disconnection as an alternative treatment Neurology. 1998;51(1):48–55. Epub 1998/07/23.
for intractable temporal lobe epilepsy: techniques, complications 126. DeGiorgio CM, Schachter SC, Handforth A, Salinsky M,
and results. Neurochirurgie. 2008;54(3):297–302. Epub 2008/04/18. Thompson J, Uthman B, et al. Prospective long-term study of
vagus nerve stimulation for the treatment of refractory seizures. refractory epilepsy. Neurosurgery. 2000;47(6):1353–7; discussion
Epilepsia. 2000;41(9):1195–200. Epub 2000/09/22. 7–8; Epub 2000/12/29.
127. Morris 3rd GL, Mueller WM. Long-term treatment with vagus 131. Al-Otaibi FA, Hamani C, Lozano AM. Neuromodulation in
nerve stimulation in patients with refractory epilepsy. The Vagus epilepsy. Neurosurgery. 2011;69(4):957–79; discussion 79; Epub
Nerve Stimulation Study Group E01-E05. Neurology. 2011/07/01.
1999;53(8):1731–5. 132. Rolston JD, Englot DJ, Wang DD, Shih T, Chang EF. Comparison
128. Elliott RE, Morsi A, Kalhorn SP, Marcus J, Sellin J, Kang M, of seizure control outcomes and the safety of vagus nerve, tha-
et al. Vagus nerve stimulation in 436 consecutive patients with lamic deep brain, and responsive neurostimulation: evidence from
treatment-resistant epilepsy: long-term outcomes and predic- randomized controlled trials. Neurosurg Focus. 2012;32(3):E14.
tors of response. Epilepsy Behav. 2011;20(1):57–63. Epub Epub 2012/03/03.
2010/12/15. 133. Fisher R, Salanova V, Witt T, Worth R, Henry T, Gross R, et al.
129. Holmes MD, Silbergeld DL, Drouhard D, Wilensky AJ, Electrical stimulation of the anterior nucleus of thalamus for treat-
Ojemann LM. Effect of vagus nerve stimulation on adults with ment of refractory epilepsy. Epilepsia. 2010;51(5):899–908. Epub
pharmacoresistant generalized epilepsy syndromes. Seizure. 2010/03/25.
2004;13(5):340–5. Epub 2004/05/26. 134. Morrell MJ. Responsive cortical stimulation for the treatment of
130. Patwardhan RV, Stong B, Bebin EM, Mathisen J, Grabb medically intractable partial epilepsy. Neurology. 2011;77(13):
PA. Efficacy of vagal nerve stimulation in children with medically 1295–304. Epub 2011/09/16.
Epilepsy: Viewpoint—Medical
60
Gregory K. Bergey
Introduction Initial Presentation: Seizure Types

and Treatment Considerations
A number of the intracranial lesions that would appropriately
be treated with stereotactic radiosurgery are associated with Epileptic seizures produced by focal lesions are always par-
epileptic seizures. Indeed meningiomas, metastatic brain tial seizures. Even if a patient presents with a generalized
lesions, cavernous malformations, arteriovenous malforma- convulsive seizure, it will be a secondarily generalized sei-
tions, and primary brain tumors often have seizures as the zure, reflecting propagation of the partial seizure from a
presenting manifestation, leading to the evaluation that focal brain region. Although a new classification of epileptic
detects the lesion for the first time. Lesions located in supra- seizures is being discussed, it is not yet formally accepted; it
tentorial brain cortex are much more likely to produce epi- is still useful to think of seizures in terms of partial (simple
leptic seizures than will lesions in white matter or deep brain and complex) or focal seizures and generalized seizures, and
structures (e.g., pituitary tumors, pineal tumors). Those these concepts will be largely preserved in any new classifi-
located in the posterior fossa (e.g., acoustic tumors) are cation. Generalized seizures can be either primary general-
unlikely to produce seizures. Lesions in the frontal or tempo- ized or secondarily generalized. A further classification of
ral lobes, particularly those involving or adjacent to mesial seizures can be into idiopathic, cryptogenic, or symptomatic;
temporal structures (amygdala, hippocampus), can fre- when discussing seizures related to lesions treated with
quently be epileptogenic. radiosurgery, these seizures are always symptomatic. The
In addressing the medical management of seizures associ- symptomatology of partial seizures observed will reflect
ated with lesions being treated with stereotactic radiosur- the region of the brain involved. For example, lesions in the
gery, it is appropriate to think of medical treatment in terms temporal lobe may produce complex partial seizures, often
of the natural history of this treatment, before radiotherapy, with an aura (e.g., unusual sensation, déjà vu, abdominal
in the period soon after radiotherapy (the first 1–2 years), and feelings) with associated automatisms and alteration of con-
then in the remote period after successful treatment. While sciousness. Lesions involving the primary motor cortex may
these delineations are conceptually operational, it is useful to produce focal motor seizures. There are some brain regions
think in these terms since the brain pathology and treatment (e.g., orbitofrontal) that, although potentially quite epilepto-
considerations evolve over time. genic, are silent regions, and clinical manifestations only
occur if the seizures propagate to adjacent areas capable of
producing clinical signs.
Therefore all epileptic seizures produced by structural
brain lesions will be symptomatic partial seizures with or
without secondary generalization. Seizures due to structural
lesions are more likely to recur than are idiopathic seizures,
and because the risk is probably in the range of 30–50 % or
greater, antiepileptic (AED) therapy is warranted. Structural
lesions do not produce primary generalized seizures (myo-
G.K. Bergey, M.D. (*)
Department of Neurology, Johns Hopkins Hospital,
clonic seizures, absence seizures, or primary generalized
Baltimore, MD, USA tonic–clonic seizures). The patient who presents with brief
e-mail: gbergey@jhm.edu periods of staring and altered awareness in conjunction with a
724 G.K. Bergey
newly diagnosed structural lesion is almost certainly having and that many focal EEG abnormalities may be relatively
complex partial seizures not petit mal or absence seizures nonspecific (i.e., only focal slow activity rather than epilep-
which are primary generalized seizures. Complex partial sei- togenic features). Sometimes continuous video-EEG moni-
zures by definition have associated altered awareness. Patients toring of patients during actual ictal events is necessary to
experiencing complex partial seizures will typically be con- provide a definitive association between a lesion and the par-
fused and have difficulty responding or processing informa- tial seizures. Such monitoring is usually reserved for the pre-
tion, but their eyes will be open and they will not fall (although surgical evaluation prior to resective surgery and is not
they may). Complex partial seizures typically will last only necessary for medical management of seizures.
1–2 min [1] and then will be followed by a postictal recovery There is no evidence to support prophylactic treatment
period of variable length but often several minutes. In contrast with AEDs in patients with primary brain tumors, metasta-
to complex partial seizures, simple partial seizures, although ses, or vascular malformations who have not yet had seizures.
also typically of short duration, do not produce alteration of Most of the best studies on AED prophylaxis in lesional epi-
consciousness. As noted above, the clinical manifestations lepsy have been done on brain tumors, primary and second-
reflect the region of the brain involved by the seizure. ary [7, 8]. Indeed there is a published AAN guideline [9]
In assessing the patient with a seizure disorder, it is impor- regarding brain tumors, based on published evidence-based
tant to determine whether the lesion to be treated with radio- studies that have failed to demonstrate a benefit of AED pro-
surgery is the symptomatic cause of the seizures. As phylaxis. While there is a paucity of controlled trials of sei-
mentioned, such lesions as meningiomas, oligodendroglio- zure prophylaxis in patients with vascular malformations
mas, arteriovenous malformations if they involve cortical or (e.g., cavernomas, arteriovenous malformations) at the pres-
mesial temporal structures can be highly epileptogenic. ent time, it is appropriate also not to administer prophylactic
Seizures are the most common initial symptom in patients AEDs to these populations before an actual seizure occurs.
with low-grade gliomas [2]. About 40 % of patients with pri- Any patient presenting with seizures thought due to a
mary brain tumors will present with seizures and about lesion that will be treated with radiosurgery should be placed
30–70 % of patients with supratentorial tumors will have sei- on antiepileptic drug (AED) therapy. As mentioned, since
zures at some time [3]. In one series, oligodendrogliomas these patients have experienced a symptomatic seizure, the
and grade 2 astrocytomas were the most likely to present risk of recurrence is significant. Obviously the MRI will be
with seizures [3]. About 30 % of patients with arteriovenous abnormal; these patients should receive AED therapy even if
malformations may present with seizures [4]. Brain metasta- the EEG is unremarkable or reveals only nonspecific abnor-
ses, a common indication for radiosurgery, frequently cause malities (e.g., focal slowing) if the history is convincing for
seizures because of their locations involving cortical or subclinical seizures. Even patients presenting only with simple
cortical structures. In one study [5], 12 % of patients with partial seizures (i.e., no alteration of consciousness) that are
metastatic lung carcinoma experienced seizures. Patients not disabling seizures should be started on AED therapy
with metastatic melanoma may have a higher incidence of because of the potential for simple partial seizures to propa-
associated seizures, in part because of the propensity of met- gate and produce disabling complex partial symptomatology
astatic melanoma to have associated hemorrhage [6]. Indeed or to secondarily generalize to convulsive events. However,
an epileptic seizure, either partial or generalized, may be the patients already on AED therapy who are only experiencing
first presentation prompting evaluation; in one study, 20 % of non-disabling simple partial seizures (e.g., several seconds
patients with metastases presented with seizures [3]. of visual symptoms due to an occipital lesion) do not neces-
Conversely, some lesions can be incidental and not pro- sarily need AED adjustment, and this usually does not war-
duce seizures or other symptoms. Most patients with multi- rant adding an additional agent. Indeed the simple partial
ple cavernous malformations or multiple metastases have seizures reflecting the lesion at the seizure onset zone may be
seizures originating from only one site. MRI imaging can more difficult to prevent than are seizures resulting from
determine whether the lesion is in an area where it reason- propagation (e.g., secondarily generalized tonic–clonic sei-
ably could be expected to be epileptogenic, but only the EEG zures) and aggressive AED polytherapy may produce addi-
can determine this definitively. It is also important to corre- tive side effects without abolishing the seizures. The goal of
late the seizure type and symptoms with the location of the AED therapy should be to control and prevent disabling sei-
structural lesion. An EEG may also be useful in establishing zures (i.e., seizures with alteration of consciousness).
a relationship between the lesion and the seizures. If a rou- Since, as mentioned, the seizures due to focal structural
tine interictal EEG demonstrates potentially epileptogenic lesions will be partial or focal with or without secondary gen-
activity from the region of the lesion (e.g., focal spikes or eralization, the selection of an AED includes many possible
sharp transients), this is good supporting evidence. It should choices. Indeed all AEDs, with the exception of ethosuxi-
be recognized that about 30 % of patients with known seizure mide, are effective agents for partial seizures (Table 60.1).
disorders may have unremarkable EEGs between seizures The lack of an FDA monotherapy indication does not mean
60 Epilepsy: Viewpoint—Medical 725
Table 60.1 Antiepileptic drugs for partial seizures (with or without rapidly are governed by a number of factors including
secondary generalization) sedative or cognitive side effects (e.g., phenobarbital,
First generation Second generation Third generation topiramate), the risk of hypersensitivity reactions (e.g.,
Carbamazepine Felbamatea Clobazam lamotrigine), or pharmacokinetic considerations. It takes five
Phenytoinb,c Gabapentin Ezogabinea drug half-lives to reach a steady state; an AED with a long
Phenobarbitalb Lamotrigine Lacosamideb,c half-life like phenobarbital or perampanel (~100 h t 1/2)
Primidone Levetiracetamb,c Perampanel can take weeks to reach a steady state. Conversely AEDs
Valproateb,c Oxcarbazepine Pregabalinb with short half-lives such as levetiracetam or lacosamide
Tiagabine Rufinamide
(7–8 h t 1/2) reach steady state levels about 35 h after starting
Topiramate Vigabatrina
a maintenance dose.
Zonisamide
The past popularity of phenytoin was in large part due to
a
Use limited by significant safety or side effect concerns
b
the ability to load this agent, either orally or parenterally,
Available as a parenteral formulation
c
Can be introduced rapidly without undue side effects or sedation providing early effective therapy. To a large degree, leveti-
racetam has replaced phenytoin for acute therapy.
Levetiracetam has a much more favorable pharmacokinetic
that an AED is not effective as monotherapy, only that specific and side effect profile compared to phenytoin, with no
monotherapy trials were not performed. In contrast to enzyme induction, no drug–drug interactions, few hypersen-
European regulatory agencies which only require non- sitivity reactions, linear kinetics, and no significant sedative
inferiority trials for a monotherapy indication, the FDA still side effects. There is good level A evidence to support the
requires superiority trials, and these are difficult to do, even use of levetiracetam as monotherapy for partial seizures [11].
with the recent modification to allow historical controls Available in both oral and parenteral formulations, oral load-
[10, 11]. It is generally felt by epilepsy experts that all AEDs ing with levetiracetam achieves a cMax in 45 min [13] and is
that are effective as adjunctive therapy are effective as well tolerated [14]. The generic formulation of the immedi-
monotherapy and that the standard of care allows for such use ate release formulation has excellent bioavailability com-
independent of the formal indications. pared with the proprietary compound [15]. Table 60.1 lists
Treatment selection of an AED should include consider- other AEDs that can be introduced relatively rapidly and are
ations regarding efficacy and tolerability as well as side effect available in parenteral formulations. Lacosamide, a new
profile, pharmacokinetics, and the need for rapid introduc- third-generation AED, can also be introduced rapidly in
tion. Although there are comparative trials [12] between the many patients, and a parenteral formulation is available,
first-generation AEDs that have shown phenytoin and carba- although rapid loading is not quite as well tolerated as leve-
mazepine to be superior (combined efficacy and tolerability) tiracetam. Lacosamide has no significant hepatic induction
to the other older agents, no such comparative trials between and few drug interactions. In patients requiring two AEDs,
newer AEDs have been performed that have demonstrated the combination of levetiracetam and lacosamide is particu-
superiority between the newer agents or between the newer larly well tolerated. Gabapentin and pregabalin have identi-
and older agents. There is a general trend away from initial cal mechanisms of action; pregabalin is preferred because of
therapy with sedating AEDs (e.g., phenobarbital, primidone) its linear kinetics which increases CNS bioavailability at
and AEDs that have powerful hepatic enzyme induction (e.g., higher doses. Pregabalin is very safe and has no drug interac-
phenytoin, phenobarbital, carbamazepine) that may result in tions. At high doses, particularly if introduced rapidly, some
drug–drug interactions or promote bone mineral density loss. nonspecific CNS side effects can occur with pregabalin;
Agents with nonlinear kinetics (e.g., phenytoin) can be more often these are initiation side effects and abate with time.
likely to produce toxicity at higher doses and serum levels. Carbamazepine, an excellent AED for partial onset sei-
Most of the newer second- and third-generation AEDs require zures, produces autoinduction, requiring slow introduction
much less monitoring of hematologic and metabolic param- over several weeks to avoid toxicity. Carbamazepine also is a
eters, and titration of these medications can be done based on powerful hepatic enzyme inducer with multiple drug interac-
clinical parameters (i.e., seizure control and lack of dose- tions and the potential for rare but serious blood dyscrasias.
related side effects) without a need for frequent monitoring The congener oxcarbazepine has much less hepatic induc-
of serum levels of these AEDs. tion and no epoxide metabolite, allowing somewhat more
When a patient has experienced a disabling seizure (i.e., rapid introduction than carbamazepine, although tolerability
with alteration of consciousness), it is desirable to institute is enhanced with slow titration. Hyponatremia, much like
antiepileptic therapy promptly. As mentioned above, in these SIADH, can occur particularly in the elderly and can be sig-
patients with symptomatic epilepsy, it is not necessary nificant (i.e., serum sodium levels below 125 mEq) [16]. The
to wait for a second seizure; this is likely to occur without acute use of lamotrigine, an otherwise well-tolerated agent
AED therapy. The limits on the ability to introduce an AED with little sedation and no hepatic induction, is limited by the
726 G.K. Bergey
need for slow introduction over weeks to months to mini- There is no data to support prophylactic treatment with AEDs
mize the risk of hypersensitivity reactions which can on in patients undergoing radiosurgery who do not have a seizure
occasion be severe (e.g., Stevens–Johnson syndrome). With history; there are no good controlled studies of AED prophy-
appropriate slow introduction, the risk of such severe rashes laxis in patients undergoing radiosurgical therapy.
is no greater than phenytoin or carbamazepine [17].
Topiramate and zonisamide need to be introduced over sev-
eral weeks because of the potential for cognitive side effects, Treatment Implications After Successful
particularly with topiramate. Radiosurgery
In women of childbearing age, levetiracetam and
lamotrigine have good profiles regarding associated major The obvious major benefit of stereotactic radiosurgery is the
fetal malformations and development. Other AEDs can also ability to avoid a craniotomy and resective surgery. Successful
be used in this patient population, but phenobarbital and val- treatment will slow or stop tumor growth or obliterate arte-
proate should be avoided unless absolutely necessary. riovenous malformations, eliminating the risk of serious
intracranial hemorrhage. But obviously the original lesion,
although significantly altered anatomically and pathologi-
Acute/Subacute Effects of Radiosurgery cally, still remains. What happens to previously symptomatic
on Epileptic Seizures seizures in these patients? While radiosurgery has been suc-
cessfully used to treat patients with drug-resistant partial epi-
Treatment with radiosurgery is designed to produce focal lepsy due to mesial temporal sclerosis [21, 22], the selection
injury to the targeted lesion. In the weeks and months after of radiosurgery for treatment of intracranial lesions is typi-
stereotactic radiosurgery, while the maximal effects and ben- cally done independent of considerations of seizure control,
efits of treatment are being accrued, there is local tissue reac- specifically to reduce or eliminate risk of hemorrhage from a
tion which can include edema and other local tissue effects vascular malformation or to stop tumor growth. Indeed many
[18, 19], typically with onset after 3–5 months. Patients with patients may have not have had seizures or have had seizures
preexisting seizures due to the underlying lesion may have that were readily controlled and not drug-resistant prior to
transient worsening of seizures requiring adjustment of AED radiosurgery. In all these patients, the decision to treat a
therapy. Series studying complications of radiosurgery typi- lesion with radiosurgery is independent of seizure control
cally include heterogenous patient populations of patients but is based on these other important considerations.
with and without seizure histories and with lesions unlikely The success of resective surgical therapy for drug-resistant
to produce seizures (e.g., acoustic neuromas, schwannomas) lesional epilepsy is best with total resection of the epilepto-
making assessment of these studies difficult. Patients with- genic lesion. Residual hemosiderin after cavernoma resec-
out a history of previous seizures may develop seizures after tion, or residual tumor, even if benign and nonprogressive
radiotherapy of potentially epileptogenic lesions. In one (e.g., DNET, ganglioglioma) can compromise total seizure
report [20], 21 % of patients undergoing radiosurgery had control or at least limit the ability to discontinue AEDs. It
seizures attributed to the radiosurgery treatment, with 15 % would be reasonable, therefore, to infer that since the under-
of patients having seizures in the early months after therapy. lying lesion is not removed with radiosurgery, the lesion
Interestingly four of six (67 %) patients with treated lesions including the associated hemosiderin, although altered, is
near the motor cortex experienced seizures, suggesting that still potentially epileptogenic. But there is reasonable evi-
lesion location might be predictive of post-radiosurgery sei- dence that radiosurgery may also reduce or eliminate the
zures. This is a reasonable hypothesis since even untreated associated epileptogenicity of the lesion much like total
lesions located in the frontal and temporal regions, involving resection. In the trials of treatment of drug-resistant seizures
neocortical or mesial temporal structures, may have a higher in patients with mesial temporal sclerosis (MTS), the per-
incidence of epileptogenicity. centage of patient achieving seizure freedom was compara-
In patients with preexisting seizure disorders, AEDs should ble to results with standard open anterior temporal lobectomy
be maintained during the acute and subacute radiotherapy with 77 % of patients in a recent small randomized trial
period, with adjustments being made as needed based on sei- achieving seizure freedom [21] with the higher regimen. As
zure recurrence. In patients with new onset seizures shortly expected, the maximal beneficial effects of radiosurgery are
after radiotherapy, AEDs should be introduced employing the not achieved until >1 year after treatment. The mechanisms
considerations for selection discussed above. The seizures of this beneficial effect of radiosurgery are not known but
in some patients may be more difficult to control during may include neuromodulatory effects or ischemic necrosis of
this subacute time period after radiotherapy when local tis- epileptogenic tissue [22]. There are no comparative studies
sue reactions are most prominent. As the local reaction sub- of AED discontinuation in patients successfully treated with
sides, seizures typically are much easier to control or manage. anterior temporal lobectomy and stereotactic radiosurgery,
but the goal of any surgical procedure for drug-resistant

epilepsy, resective or radiosurgical, is to provide seizure
control in these patients with previously uncontrolled partial
epilepsy. Therefore radiosurgery should best be thought of as
adjunctive therapy with AEDs for these patients.
One indication for open resection of a favorably located
cavernous malformation is drug-resistant epilepsy, and
radiosurgery offers an alternative for lesions in critical loca-
tions. A recent report [23] of 49 patients with cavernomas in
or near eloquent cortex and drug-resistant epilepsy treated
with radiosurgery found 53 % seizure-free with another
20 % achieving a significant reduction, and only 26 % show-
ing little or no improvement. These patients often remained
on some medical therapy for their seizures. While these
numbers for seizure control are less than would be expected
with open resective surgery [24], radiosurgery of caverno- Fig. 60.1 T-weighted FLAIR MRI image of 30-year-old female who
presented >10 years ago with right occipital hemorrhage from docu-
mas and arteriovenous malformations in patients with drug-
mented 2.5 cm AVM. Successfully treated with radiosurgery; posttreat-
resistant therapy remains an option in patients where open ment angiography documented obliteration of vascular malformation.
resection poses unacceptable risks [25, 26]. No history of seizures until possibly after hemorrhage. Was given a trial
Arteriovenous malformations often present with seizures off AEDs 2 years after radiosurgery and had GTCS 6 months after ces-
sation of AEDs. Has now been seizure-free on AED monotherapy for
or hemorrhage. Radiotherapy is one modality used in treating
>9 years. Image shows residual cavity from hemorrhage and surround-
these lesions to reduce the risk of subsequent hemorrhage. A ing increased signal
report of 65 patients with AVM and single or recurrent sei-
zures found that 51 % were seizure-free after radiosurgery at
3-year follow-up including over 50 % of patients with medi- Are there other patients with a previous history of seizures
cally intractable partial epilepsy [27]. The patients with low who are candidates for AED reduction or withdrawal after
seizure frequency (<4) and small AVM size did best from the radiotherapy? Even in patients undergoing resective surgery
standpoint of seizure control. However, a recent report of 229 for drug-resistant partial seizures, there is about a 10–15 %
patients found no difference in the 5-year risk of seizures risk of seizure recurrence in patients who achieve two years
whether the patients received AVM treatment or were treated of seizure freedom [29]. Because of this risk and the excel-
conservatively [28]. The incidence of patients with seizures lent side effect profile of selected second- and third-
and patients with seizures who achieved seizure control was generation AEDs, many epileptologists, including this
not significantly different in either group. The incidence of author, often leave these patients with controlled seizures,
patients having seizures associated with AVM is higher in seizure on a small amount of a single AED (e.g., 1,000 mg/
patients who have experienced an intracranial bleed [28]. day of levetiracetam or 200 mg/day of lamotrigine), unless
In assessing the patient with epilepsy due to lesions treated the patient clearly understands the risk of seizure recurrence
with radiosurgery, patients can be grouped based on response and wishes to be off medication. Because radiosurgery is
to therapy and degree of seizure control. Patients with ongo- non-resective, it is reasonable to assume that the risk of sei-
ing seizures obviously should remain on AEDs, recognizing, zure recurrence treated with radiosurgery is somewhat higher
as discussed above, that after the acute and subacute local tis- than those treated with surgical resection of the epileptogenic
sue reactions subside, that seizure control may be more easily lesion. Figure 60.1 illustrates an example of a young woman
attained. Patients being treated with radiosurgery for lesions with a small occipital AVM that presented with hemorrhage
that have not progressed, who only had seizures associated and a single seizure. Radiotherapy successfully obliterated
with the therapy, i.e., occurring in the first year after treat- her lesion, eliminating the risk of subsequent hemorrhage,
ment, may be candidates for AED withdrawal. It is reasonable and she had no seizures on 200 mg daily of lamotrigine until
to maintain these patients on AED therapy for 1–2 years until an attempt was made to wean her off seizure medications
all reversible changes of therapy (e.g., edema) have resolved resulting in a GTCS. She has achieved long-term seizure
and then, if a routine interictal EEG does not reveal poten- control (>5 years) on the previous medical regimen. In sum-
tially epileptogenic activity, to wean the patient off AEDs. mary, there may be patients with symptomatic seizures
During this period of AED reduction, the patient should mini- treated with radiosurgery who can successfully be weaned
mize potentially dangerous activities. In patients with seizure off all seizure medications and the percentage may be half or
recurrence after AED withdrawal, most seizures occur in the more than half of patients with small lesions, but with a risk
first few months after discontinuation of therapy. of recurrence that is not trivial and may be in the range of
728 G.K. Bergey
30–50 %, it is reasonable to continue low-dose therapy that is

well tolerated. This recommendation is based on the fact that References
the previously epileptogenic lesion, while clearly altered by
radiosurgery, still remains. In patients with previously hem- 1. Afra P, Jouny CC, Bergey GK. Duration of complex partial sei-
zures: an intracranial EEG study. Epilepsia. 2008;49(4):677–84.
orrhagic vascular malformations, the associated hemosiderin
2. Smits A, Guffau H. Seizures and the natural history of World Health
is also present and thought to contribute to epileptogenicity. Organization grade II gliomas: a review. Neurosurgery.
If a patient with total seizure control on low-dose AED 2011;68(5):1326–33.
monotherapy wishes to be given a trial off AEDs, an EEG 3. Lynam LM, Lyons MK, Drazkowski JF, Sirven JI, Noe KH, Rs Z,
Wilkens JA. Frequency of seizures in patients with newly diag-
should be performed. An interictal EEG showing potentially
nosed brain tumors: a retrospective review. Clin Neurol and
epileptogenic activity should raise concern for a high risk Neurosurg. 2007;109(7):634–8.
(i.e., >50 %) of seizure recurrence. If the EEG is unremark- 4. Galletti F, Costa C, Cupini LM, Eusebi P, Hamam M, Caputo N,
able, the patient understands the risks of recurrence and can Siliquini S, Conti C, Moschini E, Lunardi P, Carletti S, Calabresi
P. Brain arteriovenous malformations and seizures: an Italian study.
refrain from driving or other potentially dangerous activity
J Neurol Neurosurg Psychiatry. 2013;85(3):284–8.
for 6 months, then a cautious trial off AEDs may be under- 5. Lee MH, Kong DS, Seol HJ, Nam DH, Lee JI. Risk of seizure and
taken. Any seizures occurring more than 2 years after radio- its clinical implication in the patients with cerebral metastasis from
surgery indicate the need for lifelong AED therapy. lung cancer. Acta Neurochir (Wien). 2013;155(10):1833–7.
6. Goldlust SA, Hsu M, Lassman AB, Panageas KS, Avila EK. Seizure
prophylaxis and melanoma brain metastases. J Neurooncol. 2012;
108(1):109–14.
Conclusion 7. Forsyth PA, Weaver S, Fulton D, Brasher PM, Sutherland G,
Stewart D, Hagen NA, Barnes P, Cairnross JH, DeAngelis
LM. Prophylactic anticonvulsants in patients with brain tumor. Can
Seizures are the presenting symptom of many patients with
J Neurol Sci. 2003;30(2):106–12.
cerebral lesions destined to be treated with radiosurgery. 8. Sirven JI, Wingerchuk DM, Drazkowski JF, Lyons MK, Zimmerman
Partial seizures with or without secondary generalization RS. Seizure prophylaxis in patients with brain tumor: a meta-
will be the seizure type. All patients with partial seizures, analysis. Mayo Clin Proc. 2004;79(12):1489–94.
9. Glantz MG, Cole BF, Forsyth PA, Recht LD, Wen PY, Chamberlain
even those with only simple partial seizures, deserve AED
MC, Grossman SA, Cairncross JG. Practice parameter: anticonvul-
therapy to reduce the risk of disabling seizures. Comparative sant prophylaxis in patients with newly diagnosed brain tumor.
trials between AEDs demonstrating superiority are limited Report of the Quality Standards Subcommittee of the American
and are not available for the newer agents. Therefore the Academy of Neurology. Neurology. 2000;43(10):1886–93.
10. French JA, Wang S, Warnock B, Temkin N. Historical control
selection of AED should be based on consideration other
monotherapy design in the treatment of epilepsy. Epilepsia. 2013;
than relative efficacy, considerations such as side effect pro- 51(10):1936–43.
file or pharmacokinetics. Some of the newer second- and 11. Glauser T, Ben-Menachem E, Bourgeois B, Cnaan A, Guerreiro C,
third-generation AEDs offer distinct advantages over the Kälviäinen R, Mattson R, French JA, Perucca E, Tomson T. Updated
ILAE evidence review of antiepileptic drug efficacy and effective-
first-generation agents, and the use of phenytoin, phenobar-
ness as initial monotherapy for epileptic seizures and syndromes.
bital, primidone, and valproate has appropriately declined in Epilepsia. 2013;54(3):551–63.
recent years. Levetiracetam is a particularly useful agent 12. Mattson RH, Cramer JA, Collins JF, Smith DB, Delgado-Esqueta
both for acute and long-term use, and newly approved lacos- AV, Browne TR, Williamson PD, Treiman DM, McNamara JO,
McCutchen CB, Homan RW, Crill WE, Lybozynski MF, Rosenthal
amide offers similar benefits.
NP, Mayersdorf. Comparison of carbamazepine, phenobarbital,
The local tissue reaction produced by radiosurgery may phenytoin, and primidone in partial and secondarily generalized
temporarily exacerbate preexisting seizures or on occasion tonic-clonic seizures. N Engl J Med. 1985;313(3):145–51.
produce seizures in patients with no previous seizure history. 13. Ramael S, De Smedt F, Toublanc N, Otoul C, Boulanger P,
Riethuisen JM, Stockis A. Single-dose bioavailability of levetirace-
These effects abate with time. Successful radiosurgical treat-
tam intravenous infusion relative to oral tablets and multiple-dose
ment in addition to its primary treatment goals may also pharmacokinetics and tolerability of levetiracetam intravenous
reduce the epileptogenicity of lesion, but because the patho- infusion compared with placebo in healthy subjects. Clin Ther.
logically altered (e.g., obliterated vascular malformation) 2006;28(5):734–44.
14. Khoubessi MZ, Amina S, Pita I, Bergey GK, Werz MA. Tolerability
lesion remains, it is recommended that most patients
and efficacy of oral loading of levetiracetam. Neurology. 2008;
with controlled seizures remain on low-dose AED therapy 70(22 Pt2):2166–70.
unless seizures occurred only in the period shortly after 15. Krauss GL, Caffo B, Chang YT, Hendrix CW, Chuang K. Assessing
radiotherapy treatment. Patients with surgically accessible bioequivalence of generic antiepileptic drugs. Ann Neurol. 2011;
70(2):221–8.
lesions and drug-resistant disabling partial seizures may be
16. Dong X, Leppik IE, White J, Rarick J. Hyponatremia from
candidates for resective seizure surgery even after successful oxcarbazepine and carbamazepine. Neurology. 2005;65(12):
radiosurgery treatment. 1976–8.
17. Mochenhaupt M, Messenheimer J, Tennis P, Schlingmann J. Risk 23. Leveque M, Carron R, Bartolomei F, Regis J. Radiosurgical treat-
of Stevens-Johnson syndrome and toxic epidermal necrolysis in ment for epilepsy associated with cavernomas. Prog Neurol Surg.
new users of antiepileptics. Neurology. 2005;64(7):1134–8. 2013;27:157–65.
18. Girvigian MR, Chen JC, Rahimian J, Miller MJ, Tome 24. Von der Brelie C, Malter M, Niehusmann P, Elger C, Von Lehe M,
M. Comparison of early complications for patients with convexity Schramm J. Surgical management and long-term seizure outcome
and parasagittal meningiomas treated with either stereotactic radio- after epilepsy surgery for different types of epilepsy associated with
surgery or fractionated stereotactic radiotherapy. Neurosurgery. cerebral cavernous malformations. Epilepsia. 2013;54(9):1699–706.
2008;62(Suppl):A19–27. 25. Kim W, Stramotas S, Choy W, Dye J, Nagasawa D, Yang I. Prognostic
19. Unger KR, Lominska CE, Chanyasulkit J, Randolph-Jackson P, factors for post-operative seizure outcomes after cavernous malfor-
White RL, Aulisi E, Jacobson J, Jean W, Gagnon GJ. Risk factors mation treatment. J Clin Neurosci. 2011;18(7):877–90.
for posttreatment edema in patients with stereotactic radiosurgery 26. Alonso-Vanegas MA, Cisneros-Franco JM, Otsuki T. Surgical man-
for meningiomas. Neurosurgery. 2012;70(3):639–45. agement of cavernous malformations presenting with drug-resistant
20. Gelblum DY, Lee H, Bilsky M, Pinola C, Longford S, Wallner epilepsy. Front Neurol. 2012;2:86.
K. Radiographic findings and morbidity in patients treated with ste- 27. Schauble B, Cascino GD, Pollock BE, Gorman DA, Weigand S,
reotactic radiosurgery. Int J Radiat Oncol Biol Phys. 1998;42(2): Cohen-Gadol AA, McClelland RL. Seizure outcomes after stereo-
391–5. tactic radiosurgery for cerebral arteriovenous malformations.
21. Barbaro NM, Quigg M, Broshek DK, Ward MM, Lamborn KR, Neurology. 2004;63(4):683–7.
Laxer KD, Larson DA, Dillon W, Verhey L, Garcia P, Steiner L, 28. Josephson DB, Leach JP, Duncan R, Roberts RC, Counsell CE,
Heck C, Kondziolka D, Beach R, Olivero W, Witt TC, Salanova V, Al-Shahi Salman R, Scottish Audit of Intracranial Vascular
Goodman. A multicenter, prospective pilot study of gamma knife Malformations (SAIVMs) steering committee and collaborators.
radiosurgery for mesial temporal lobe epilepsy: seizure response, Seizure risk from cavernous or arteriovenous malformations: pro-
adverse events, and verbal memory. Ann Neurol. 2009;65(2): spective population-base study. Neurology. 2011;76(18):1548–54.
167–75. 29. Spencer SS, Berg AT, Vickrey BG, Sperling MR, Bazil CW, Shinnar
22. Quigg M, Rolston J, Barbara NM. Radiosurgery for epilepsy: clini- S, Langfitt JT, Walczak TS, Pacia SV. Multicenter study of epilepsy
cal experience and potential antiepileptic mechanisms. Epilepsia. surgery predicting long-term seizure outcome after resective epi-
2012;53(1):7–15. lepsy surgery: the multicenter study. Neurology. 2005;65(6):912–8.
for Psychiatric and Pain Disorders 61
Chun-Po Yen and Jason P. Sheehan
Radiosurgery and Somatic Pain

Introduction and Historical Perspective
Leksell had a special interest in pain disorders likely in part
Despite great advances in the medical management of pain due to the fact that a person close to him died a painful death.
and psychiatric disease, there remains a group of patients He focused upon acute pain associated with cancer and tri-
who are refractory to the best medical therapies. For exam- geminal neuralgia. Leksell did not really explore treatment
ple, despite therapeutic progress in recent years, conven- for chronic pain. When Leksell started to treat acute pain, the
tional treatment of anxiety disorders fails or has only a understanding of the pathophysiology of pain was limited.
temporary effect in 20 % of patients. Pain and psychiatric Therefore, the results for surgical treatment of pain were
disorders are often severely disabling and are associated with fairly discouraging.
rates of suicide comparable to those of depression. When He was perhaps overly optimistic about the radiosurgical
medical therapy fails, options include open surgical proce- results for pain syndromes. As soon as he developed a new
dures, consisting of lesioning or implantation of deep brain technique, Leksell would try it in the management of both
stimulators (DBS), and radiosurgery. Radiosurgical options intractable pain and tic douloureux. In fact, Leksell designed
are well-suited for functional neurosurgery, in which the the radiation pattern of the first Gamma Knife to be lens-
neurosurgeon strives to create a focused lesion without injur- shaped so that its effect would be similar to that of a knife
ing or affecting adjacent structures. blade cutting pain fibers. Leksell wanted to prove that radio-
Historically, the Gamma Knife was developed in the late surgery could selectively ablate pain transmitted by C fibers.
1960s as an alternative to open stereotactic lesioning for In Brain Fragments, Leksell described the treatment of a
functional disorders. With the development of the Gamma 33-year-old accountant with facial dolorosa secondary to
Knife and a limited understanding of the neurophysiology of basal cell cancer of the tongue using the Gamma Knife. “The
pain and psychiatric disorders, Dr. Leksell was obsessively treatment lasts 1 h, and the dose given is 15,000 rad.
interested in the treatment of patients with intractable pain Meanwhile we eat sandwiches and drink Pilsner in the con-
and psychiatric disorders [1–4]. In his book Brain Fragments, trol room. When the radiation treatment is finished, the pain
Dr. Leskell wrote, “One can accept death, but one cannot has already diminished” [2].
accept the deep, devastating pain. Sharp, intractable pain is
like hell ‘without escape, without hope and without
Heliotrope when Venom burns.’ Standing at the bedside Radiosurgery and Trigeminal Neuralgia
without ever having experienced pain, it is impossible to
imagine the patient’s agony, and it is impossible to under- With the Gamma Knife unit installed in Stockholm, he
stand that a short time without pain can be extreme happi- asked Dr. Hakanson to find the two patients he previously
ness” [2]. Since then, the Gamma Knife and other treated with stereotactic technique using orthovoltage x
radiosurgical devices have been utilized to treat patients with rays for trigeminal neuralgia. The two patients remained
medically refractory psychiatric and pain disorders. pain free 17 years after the treatment. Following the suc-
cessful long-term outcome of these two patients, Dr.
Leksell and Dr. Hakanson treated 48 patients between
C.-P. Yen, M.D. • J.P. Sheehan, M.D., Ph.D. (*)
1970 and 1978, using plain stereotactic skull X-rays and
Department of Neurological Surgery, University of Virginia,
1215 Lee Street, Charlottesville, VA 22908, USA transoval cisternography using tantalum dust for stereotactic
e-mail: Cy4f@virginia.edu; jps2f@virginia.edu targeting [5].
732 C.-P. Yen and J.P. Sheehan
Radiosurgery and Psychiatric Disorders normal parietal lobe of rats was studied. A dose of 50 Gy
caused astrocytic swelling and fibrin deposition in capillary
Leksell also first described psychoradiosurgery, in which he walls without changes in neuronal morphology or break-
targeted the frontolimbic connections in both anterior inter- down of the blood–brain barrier at 12 months. At 75 Gy,
nal capsules (capsulotomy) for selected cases of intractable more vigorous morphological changes were seen in astro-
anxiety and severe central pain [6]. Leksell first employed cytes within 4 months. In addition, necrosis, breakdown of
stereotactic radiosurgery to treat some psychiatric disorders. the blood–brain barrier, and hemispheric swelling were
noted. At 120 Gy, astrocytic swelling occurred within 1 week
of irradiation and necrosis was seen at 4 weeks, but it was not
Disease Pathophysiology and Radiosurgical associated with hemispheric swelling [8].
Targeting for Pain and Psychiatric Disorders Unlike treating a tumor or arteriovenous malformation,
there is not a pathological target for functional radiosurgery.
The radiobiology of treating these disorders differs from Effective radiosurgical management of pain and psychiatric
other disease entities in that the targeted area is a brain struc- disease depends upon selecting appropriate targets. This, in
ture that appears radiographically normal; targets for such part, relies on an understanding of the anatomical circuitry
procedures include the thalamus, pituitary gland, or internal and pathophysiological mechanisms underlying the func-
capsule. The radiosurgical treatment of pain and psychiatric tional disorder. The organization of the limbic system is the
disease relies heavily on the concept that adjacent normal basis of understanding for pain and psychoneurosurgery
structures receive a markedly lower dose (due to rapid radia- (Fig. 61.1) [9]. Neuroanatomical sites for various pain and
tion dose fall-off). A higher dose rate (larger dose in an psychosurgeries are depicted in Fig. 61.2 [10].
equivalent amount of time or same total dose applied over a
shorter period of time) increases the lethality of the dose to
the target due to greater interference with intrinsic cellular Pain
repair mechanisms during irradiation. The significance of
this effect is seen most clearly at a threshold dose rate of Radiosurgery has been used to treat different forms of pain
1 Gy/min [7]. including somatic pain (i.e. cancer-related or non-cancer-
In order to understand the radiobiology of a single high related), central pain, trigeminal neuralgia, and sphenopala-
dose of radiation on normal brain, the effect of GKS on the tine neuralgia. All of these types of pain are a manifestation
4
6
8 CINGULUM BUNDLE
9
23
24
CC
ATR FORNIX
SCR ST
AT
25
UM SM HAB
10 PT
SE MTT
13 HPT
12
DB MFB
14
HYP IP
UB OT MB
PIT
INS AMYG
OB HIPPOCAMPUS
CG
LMA
PRESUBICULUM G
HIPPOCAMPAL GYRUS
SOMATIC AND
VISCERAL
AFFERENTS
Fig. 61.1 Limbic system: OB olfactory bulb, LOT lateral olfactory habenulointerpeduncular tract, IP interpeduncular nucleus, LMA limbic
striae, INS insula, UB uncinate bundle, DB diagonal band of Broca, midbrain area, G nucleaus of Gudden, CG central gray, and CC corpus
AMYG amygdala, SCR subcallosal radiations, HYP hypothalamus, AT callosum. From Mindus P: Capsulotomy in Anxiety Disorders- A
anterior thalamus, MB mammillary body, MTT mammillothalamic Multidisciplinary Study, Dissertation at the Karolinska Institute and
tract, ATR anterior thalamic radiations, ST stria terminalis, HAB Hospital, in. Stockholm, Sweden, 1991
habenula, MFB medial forebrain bundle, SM stria medullaris, HPT
61 Stereotactic Radiosurgery for Psychiatric and Pain Disorders 733
Fig. 61.2 Neuroanatomical Coronal plane of anterior tip

sites for various of lateral ventricle
psychosurgeries are
illustrated. From Mindus Ant. commissure
P: Capsulotomy in Anxiety
Post. commissure
Disorders—A
Multidisciplinary Study, Cingulate gyrus
Dissertation at the
Karolinska Institute and
Hospital, in. Stockholm,
Sweden, 1991
Hippocampal gyrus
Corpus callosum
Orbital cortex
Uncus
CORONAL PLANE OF:

BIMEDIAL LEUCOTOMY (FALCONER-SCHURR, JACKSON)
INFERIOR-MEDIAL LEUCOTOMY (BAKER et al)
Cingulotomy: (BALLANTINE et al)
(MEYER et al)
SITE OF LEUCOTOMY: (BAILEY et al)
SITE OF 90Y LESION OF KNIGHT (IN SUBCOSTICAL WHITE MATTER)
CONNECTIONS ORBITAL CORTEX TO UNCUS (VAN HOESEN et al)
RECIPROCAL CONNECTIONS FRONTAL GRANULAR CORTEX GYRUS

CINGULI (NAUTA)
ANTERIOR CAPSULOTOMY (LEKSELL et al)
of different pathophysiological processes. There is not a In pioneering work, Granit, Leksell, and Skoglund (1944)
single anatomical circuit or target that is most appropriate. demonstrated that local pressure on nerve fibers could result
Treatment can target central anatomic structures that mediate in painful afferent discharges from the injured neural seg-
the pain sensation (e.g., thalamus) or can target the nerve that ment [16]. More recently, Jannetta and others have suggested
carries the pain sensation (e.g., trigeminal nerve in trigemi- that vascular compression of the trigeminal nerve may be a
nal neuralgia). The complexity of neuronal network mediat- causal agent in trigeminal neuralgia [17–19]. This fact led to
ing pain makes its management by open or closed stereotactic the hypothesis of a causal relationship between vessel com-
techniques largely unsatisfactory. Chronic pain targeting pression and trigeminal neuralgia and the devising of micro-
with radiosurgery has been typically performed at the level vascular decompression surgery. Despite such hypotheses,
of the medial thalamus. Young targeted the intralaminar, the fact that balloon compression of the nerve can lead to
mediodorsal centromedian, and parafascicular nuclei in symptomatic improvement in some patients and infections and
chronic pain patients [11–13]. PET or functional MRI- demyelinating conditions may also affect the trigeminal nerve
guided cingulotomies have also been proposed for chronic and illicit pain underscores the true lack of understanding as
pain treatment [14]. to the pathophysiology of trigeminal neuralgia [20, 21].
Trigeminal neuralgia (TGN) is a paroxysmal lancinat- With radiosurgery, the usual target for treating trigemi-
ing pain confined to a distribution encompassing one or more nal neuralgia is the trigeminal nerve root entry zone or the
of the branches of the trigeminal nerve on one side of the cisternal segment of the trigeminal nerve at the level of
face. As early as 1941, Olivecrona understood and described the pons. Much has been written about whether or not the
that mechanical pressure along the root or at the level of trigeminal nerve should be targeted more proximally or
the ganglion could be the cause of trigeminal neuralgia [15]. distally [22–24].
Rare cases of glossopharyngeal and sphenopalatine been more simple yet elegant. Drawing from the work of
neuralgia treatment with radiosurgery have been described Henry Wycis in Philadelphia and the anatomical studies of
[25, 26]. Neither the radiosurgical targeting nor the outcome Professor Meyer in London, Leksell used thermal injury to
for these types of neuralgia is well defined. perform a stereotactic capsulotomy on a successful indus-
Obsessive-Compulsive Disorder
and Depression
The most common radiosurgical target for obsessive-

compulsive disorder (OCD) is the anterior limb of the inter-
nal capsule. A capsulotomy can be performed on the right
hemisphere, left hemisphere, or both. The anatomical site for
a capsultomy is depicted in Fig. 61.3 [27] and the appearance
on MRI is shown in Fig. 61.4. Lippitz et al. noted a correla-
tion between clinical success and radiological lesioning of
the middle of the anterior limb of the internal capsule [28].
This was defined as parallel to a plane encompassing the
anterior commissure-posterior commissure line level at the
level of the foramen of Monro and 4 mm rostral by a plane
defined by the internal cerebral vein [29]. Cingulotomy can
also be performed for OCD and depression [30]. In addition,
limbic leucotomy involving bilateral lesions of the cingulum
with subcaudate tractotomy have been advocated for intrac-
table depression [31–33].
Anxiety
Leksell was impressed with the result, yet disappointed

with the prefrontal lobotomy technique made famous by
Fig. 61.3 An axial slice of the brain depicting the target areas for bilat-
Egas Moniz and Almeida Lima. He believed that the fron- eral capsulotomies (asterisk). From Mindus P: Capsulotomy in Anxiety
tal lobes should have been surgically treated with more Disorders—A Multidisciplinary Study, Dissertation at the Karolinska
finesse, and that psychosurgery in general should have Institute and Hospital, in. Stockholm, Sweden, 1991
Fig. 61.4 Gamma Knife dose plan with isocenters depicted in yellow and the 6-month post-Gamma Knife coronal MRIs illustrating the imaging
appearance of bilateral capsulotomies for intractable obsessive-compulsive disorder
trialist who was completely incapacitated by anxiety [2]. Electroconvulsive therapy has been utilized for depression.
The result from their early case was good and prompted It typically has a transient benefit and must be readministered
more psychosurgery using electrodes and later the Gamma with some regularity.
Knife. For intractable anxiety disorders, bilateral capsu-
lotomies have been performed with a similar target to that
for OCD [28, 34, 35]. The Argument in Favor of Radiosurgery
In most cases, radiosurgical treatment of pain and psychiatric

Treatment Alternatives disorders should be reserved for patients who demonstrate
fairly severe symptoms and who fail medical management.
Chronic pain is initially treated with pharmacological treat- However, as noted, this is not an inconsequential number of
ment. Pain specialists utilize opioid agonists, nonsteroidal patients. In patients with psychiatric disease who require
anti-inflammatories, local anesthetics, antidepressants, cer- surgical intervention, functional radiosurgery offers several
tain anticonvulsants, and muscle relaxants to optimize pain important clinical as well as scientific advantages over open
control. More recently, implantable drug delivery systems techniques. The most important is patient tolerance. It is our
(e.g., morphine pumps), transcutaneous electrical nerve experience that this psychologically vulnerable group of
stimulation, spinal cord stimulators, and vagal nerve stimula- patients is much more willing to undergo a closed stereotac-
tors have been implanted to assist with the treatment of tic procedure, which in contrast to open surgery, leaves
chronic pain [36–38]. Pain rehabilitation programs can also minimal, if any, external marks. Theoretically, the gradual
be of use. development of the radiosurgical lesion may also allow the
For trigeminal neuralgia, medical management is the first patient to better psychologically adjust. The psychological
line of treatment for patients. Medications such as carbam- rehabilitation phase is an important part of any psychosurgi-
azepine, oxcarbazepine, phenytoin, gabapentin, and pregab- cal procedure.
alin are used to treat trigeminal neuralgia. However, many Stereotactic radiosurgery also does not carry the same type
patients with this condition eventually fail medical therapy and degree of immediate or severe risks that open surgery
because of refractory pain or intolerable medication side- (e.g., microvascular decompression, deep brain stimulation,
effects. More invasive treatment options include microvascu- etc) does. For instance, in a long-term series of microvascular
lar decompression (MVD) and percutaneous procedures decompression patients, there were the following risks from
such as glycerol rhizolysis, radiofrequency rhizotomy, per- the series by Dr. Jannetta: 0.2 % death; 0.1 % brain stem
cutaneous balloon microcompression, and peripheral nerve infarct; and 1 % hearing loss 1. None of these complications
blocks [1, 18, 20, 21, 39–41]. were observed in our series nor were they associated with
Obsessive-compulsive disorder, depression, and anxiety radiosurgical treatment of trigeminal neuralgia in other major
are initially managed with pharmacological agents and psy- centers. Complications from hardware and the risk of hemor-
chiatric intervention/counseling. Heterocyclic compounds, rhage for deep brain and vagal nerve stimulators combine to
monoamine oxidase inhibitors, selective serotonin reuptake yield a risk of 10–27 % [46–48]. Ultimately, the patient must
inhibitors, bipolar agents, and anxiolytics/hypnotics have all choose the type of intervention he or she is willing to undergo.
been applied to these conditions with limited efficacy. Many Such a choice inevitably involves a weighing of the relative
of the patients with intractable OCD, depression, or anxiety risks and benefits of each intervention.
have a high rate of comorbid Axis I diagnoses including per- Of all the indications, stereotactic radiosurgery has been
sonality disorders and other functional impairments further most embraced for the treatment of medically refractory tri-
complicating the treatment and reducing the chances of a geminal neuralgia. The widespread application of radiosur-
favorable therapeutic outcome [42]. Unfortunately, pharma- gery for trigeminal neuralgia is in part a testament to its
cological therapies only benefit between 50 and 70 % of all acceptable degree of pain relief and few side effects.
OCD patients [43]. Fava and Davidson estimate that 29–46 % Radiosurgery for chronic pain, OCD, intractable anxiety, and
of depressed patients fail to respond to pharmacological depression has been applied with far less frequency. This fact
agents [44]. stems in part from the notion that such patients are usually
In addition to pharmacological treatment, deep brain managed by other types of physicians (e.g., psychiatrists, pain
stimulation, vagal nerve stimulation, and transcranial mag- specialists, etc.) who have little familiarity with radiosurgery.
netic stimulation have been utilized for OCD and depression Moreover, radiation oncologists and neurosurgeons are gener-
[45]. The ability to reverse or modulate these neurological ally uncomfortable with treating these types of disorders with-
interventions adds to the attractiveness of these interventions out the close involvement of appropriate specialists to evaluate
over a more static lesioning achieved with radiosurgery. and assist in the pre- and postoperative management.
Surgical Treatment Alternatives for Pain target area after approximately 3 months. This signal is most
and Psychosurgeries likely produced by local edema. The edema extends a maximal
volume at around 9 months and then slowly subsides. The
Open lesioning—Although the results of stereotactic or per- edema is directly related to the dose and to the volume radi-
cutaneous lesioning (e.g., thalamotomy, radiofrequency rhi- ated. It may be sufficient to use only one isocenter and the
zotomy, glycerol injection) have been well studied, and the 4-mm collimator. With these treatment parameters and a
beneficial effects are immediate, there remains the potential maximal dose of 180 Gy, a lesion measuring approximately
for complications such as intracranial hemorrhage, stroke, 4 to 5 mm three can be expected within several weeks with
and infection. In addition, these procedures carry with them only minimal transient edema. Centers have utilized maxi-
an anesthetic risk too. mal target doses of 120–200 Gy for radiosurgical treatment
Deep brain stimulation and vagal nerve stimulation—The of OCD, intractable anxiety disorders, depression, and
advantages of stimulation techniques include the reversible chronic pain [11, 13, 34, 35, 42, 51, 52].
nature of the process and the ability to modulate the neural Kihlstrom et al. propose that a minimum dose of 110 Gy using
stimulation over time. However, patients are again exposed the 4 mm collimator is required to create a permanent lesion [35].
to the risks of open surgery including infection, hemorrhage, Friehs et al. in 1996 reported the following lesion size after deliv-
and anesthesia. In addition, rates of hardware complications/ ering a dose of 160 Gy using a 4 mm collimator [51]:
failure are not trivial, and the batteries need periodic replace-
ment necessitating another albeit minor surgical procedure Time (months) Diameter of necrosis (mm)
[46–48]. 1 3
Microvascular decompression for trigeminal neuralgia— 3 6
This procedure remains the gold standard for surgical 6 8
approaches to treat trigeminal neuralgia. Barker et al. reported 12 4
excellent pain relief in 70 % of patients and partial pain relief
in another 4 % at 10 years postoperatively from a microvascu-
lar decompression [17]. However, there were the following For trigeminal neuralgia, dose selection has been studied
risks from the series by Dr. Jannetta: 0.2 % death; 0.1 % brain much more extensively. Most centers utilize maximal doses
stem infarct; and 1 % hearing loss [49]. Burchiel et al. in 1988 of 70–85 Gy [53–58]. Doses of 90 Gy or higher have been
noted a 3.5 % major recurrence rate and 1.5 % minor recur- associated with increased risk of post-radiosurgical compli-
rence rate following invasive surgical approaches [50]. All cations [55].
surgical approaches including radiosurgery appear to have a Case reports and small case series glossopharyngeal and
“wearing off” effect in terms of pain relief over the years. sphenopalatine neuralgia treatment with radiosurgery have
been published [25, 26]. We reported on the pain relief of a
patient with glossopharyngeal neuralgia delivering 80 Gy
Radiosurgical Treatment Dosimetry to the glossopharyngeal meatus [59]. Pollock and Boes
(2011) reported on a series of five patients with glossopha-
The dose must be planned so that the steepest isodose gradi- ryngeal neuralgia who underwent Gamma Knife radiosur-
ent of the dose distribution (usually between the 50 and gery with a dose of 80 Gy to the distal portion of the
70 % isodose lines) coincides with the periphery of tissue glossopharyngeal and vagus nerves at the level of the jugu-
being treated. This may require several overlapping fields of lar foramen [60]. Of these five patients, three experience
radiation, each using a different collimator size and a sepa- benefit from radiosurgery, whereas two had no benefit and
rate stereotactic focal point. Changing the relative time of underwent microsurgery too with no apparent pain relief
radiation at each target may also change the isodose distri- [60]. Pollock and Kondziolka performed Gamma Knife
bution. Finally, the radiation field may be altered by blocking surgery twice on a patient with sphenopalatine neuralgia
some of the radiation sources, also known as plugging or using a maximal dose of 90 Gy [25]. The patient underwent
shielding. repeat Gamma Knife surgery for partial pain recurrence
For radiosurgical capsulotomies and thalamotomies, focal and remained pain free 2 years after repeat treatment. The
lesions can be produced using a 4-mm collimator and 1–3 North American Gamma Knife Consortium recently
isocenters on each side for overlapping fields, creating a reviewed 17 patients who underwent Gamma Knife surgery
cylindrical lesion, with a maximum dose within the target for intractable sphenopalatine neuralgia. Favorable pain
volume of 200 Gy. The development of the lesions using relief (Barrow Neurological Institute Grades I–IIIb) was
such a plan has been followed by MRI and CT scans every 3 achieved and maintained in 10 (59 %) of 17 patients at a
months. On T2-weighted images, a high signal appears in the median follow-up of 34 months.
others (33 %) had greater than 50 % pain relief [13].

Outcomes Based upon the Type Additional investigation must be conducted before the role
of Radiosurgical Device of the Gamma Knife for pain treatment can be fully defined.
No trials have been performed to look at the results of

psychosurgeries using different radiosurgical devices. In fact, Psychiatric Disease: Anxiety and Obsessive-
very few psychosurgeries have been done with radiosurgical Compulsive Disorder
devices other than the Gamma Knife. However, other radio-
surgical devices (e.g., the Cyberknife or Novalis) may be Mindus and colleagues at the Karolinska Institute reported
well suited for treating psychiatric conditions too. the effects of bilateral anterior radiosurgical capsulotomies
For trigeminal neuralgia, the results of linear accelerator on the anxiety symptoms and personality characteristics.
(LINAC)-based and Gamma Knife radiosurgery appear Using independent observers, the Comprehensive Psycho-
comparable [57, 61, 62]. Outcomes appear generally favor- pathological Rating Scale (CPRS), and the Karolinska Scales
able so long as the team is experienced and the device is of Personality (KSP, designed to asses frontal lobe dysfunc-
optimized for radiosurgical treatment. tion and anxiety proneness), they compared patients who had
been treated with conventional thermocoagulation and fol-
lowed for 1 year with patients who had been treated with
Radiosurgical Treatment Outcomes Gamma Surgery and followed for 7 years. The two groups
were otherwise similar, each including patients who had a
Pain Other than Trigeminal Neuralgia mean duration of psychiatric illness of 15 years and who had
failed multiple other treatment interventions. The results of
Early results using the Gamma Knife to produce thalamoto- Gamma Knife capsulotomy were found to be comparable to
mies for pain control were published by Steiner et al. in those of capsulotomy performed by the thermocoagulation
1980. All of the 52 patients treated suffered from terminal technique. In both groups, freedom from symptoms or con-
cancer and were treated prior to the advent of CT or siderable improvement was noted in 70–80 % of patients,
MRI. Pneumoencephalography was used to target the tha- and none were worse after the operation. Negative effects on
lamic centrum median-parafasciculus (CM-Pf complex). the personality were not noted. Representative KSP scores
Good pain relief was obtained in 8 patients and moderate are illustrated in Figs. 61.5 [65] and 61.6 [66]. A number of
pain relief in 18. The patients had in general only temporary patients who were preoperatively unable to work or function
relief of pain. Of those with good pain relief, five died with- normally in society due to preoccupation with personal
out recurrence of pain between 1 and 13 months after the cleanliness and the inability to use public transportation
procedure, and three had recurrence of pain at 3, 6, and 9 resulting in domestic confinement, aggravated psychological
months, respectively. Doses between 100 and 250 Gy were problems, deterioration of family relationships, and devasta-
tested. Observation of an actual lesion was only possible in tion of personal economy were able to return to their previ-
21 of 36 patients that had a postmortem examination. Not ous occupation and to a normal social function. In five of
surprisingly, the presence of a lesion was associated with seven patients who underwent Gamma Knife capsulotomies,
relief. Lesions were only reliably created with doses greater a lesion was demonstrated by MRI, and those were the
than 160 Gy. The collimators used were 3 × 5 and 3 × 7 mm. patients who benefited from the procedure. The lowest effec-
The most effective lesions were more medially located near tive target dose was 160 Gy, whereas 100, 120, and 152 Gy
the wall of the third ventricle, and the greatest relief was for failed to produce lesions [29, 35].
face or arm pain [63]. Preliminary results from an ongoing study of bilateral ante-
These results were not particularly encouraging. However, rior capsulotoy using the Gamma Knife, Greenberg et al. in
with improvements in neuroimaging and alternate target 2003 reported 4 out of 15 patients had at least a 35 % decrease
selection, it is possible that more effective lesions can be pro- in the Yale-Brown Obsessive-Compulsive scale and a mini-
duced. Recent reports seem to support this expectation. mum 15 point improvement in the Global assessment scale on
Hayashi et al. in 2003 reported significant pain reduction in 5-year follow-up. In another group of 16 patients who under-
patients with severe cancer pain and post-stroke thalamic went two pairs of bilateral lesions during one session, 10 out of
pain after Gamma Knife lesioning of the hypophysis [64]. 16 patients met the aforementioned criteria at the 3-year fol-
Using the 4 mm collimator and doses of 140–180 Gy, Young low-up time point. Once achieved, this group noted generally
et al. have published effective pain relief in patients with stable benefits [42]. Kondziolka et al. (2011) recently reported
chronic, intractable pain following medial thalamotomy with on the radiosurgical treatment of three patients with medically
the Gamma Knife [11–13]. In a series of 15 patients fol- intractable OCD. They observed no morbidity and significant
lowed for more than 3 months after a radiosurgical-induced improvement in the Yale-Brown Obsessive-Compulsive Scale
medial thalamotomy, four (27 %) were pain free and five in all three patients [67]. However, recently, the group at Brown
Somatic anxiety
Musular temsion
Psychic anxiety
Psychasthenia
Inhibition of aggression
Impulsivity HJ
Monotony avoidance Post
Socialization
Pre
Social desirability
Detachment
Indirect aggression
Verbal aggression
Irritability
Suspicion
Guilt
–1 SD 50 1 SD
Fig. 61.5 Karolinska Scales of Personality (KSP) scores reported by Capsulotomy in Anxiety Disorders—A Multidisciplinary Study,
Mindus for a patient who underwent a capsulotomy for non-obsessive- Dissertation at the Karolinska Institute and Hospital, in. Stockholm,
compulsive disorder anxiety. SD standard deviation. From Mindus P: Sweden, 1991
Fig. 61.6 Karolinska Scales of Personality (KSP) scores reported by Capsulotomy in Anxiety Disorders—A Multidisciplinary Study,
Mindus for a patient who underwent a capsulotomy for obsessive- Dissertation at the Karolinska Institute and Hospital, in. Stockholm,
compulsive disorder anxiety. SD standard deviation. From Mindus P: Sweden, 1991
University has reportedly put its use of the Gamma Knife on thermocoagulation and cryprobes to a perform limbic
hiatus for treatment of OCD patients. This hiatus arose after leucotomy may give direction to future work using radio-
three patients were noted to have cyst formation after treatment surgical devices to treat depression [31, 68].
with the Gamma Knife Perfexion.
Complications of Functional Radiosurgery

Psychiatric Disease: Depression
Although stereotactic radiosurgery is not associated with
The long-term outcomes of radiosurgery for depression some of the immediate risks of open surgical procedures, it
have not been well studied. Radiosurgery of patients with nonetheless has risks. Except for trigeminal neuralgia, dose
intractable depression requires a multidisciplinary approach selection is not well studied. The wide range of doses may
and warrants further investigation [63]. The work using lead to an unpredictability of lesion size and subsequent out-
come. Kihlstrom et al. noted that parenchymal reaction and and chronic pain. Advances in neuro-imaging, neuro-
time course to development of a lesion following radiosur- anatomy, and psychiatry may lead to a resurgence of interest
gery for refractory anxiety disorders may be difficult to pre- in this field. Any future application of radiosurgery to treat
dict [34]. Other potential complications associated with these disorders should be done in the context of a multidisci-
radiosurgery include radiation-induced parenchymal changes plinary team.
(some of which are reversible and others are irreversible),
delayed cyst formation, radiation-induced neoplasia, and
vascular injury [69, 70]. More worrisome side effects of References
radiosurgical or thermocoagulation tractotomies, capsuloto-
1. Helfant MH, Leksell L, Strang RR. Experiences with intractable
mies, cingulotomies, and limbic leucotomies include head-
pain treated by stereotaxic mesencephalotomy. Acta Chir Scand.
ache, confusion, somnolence, transient disinhibition, 1965;129:573–80.
personality changes, seizures, incontinence, exacerbation of 2. Leksell L. Brain fragments. In: Steiner L, Forster D, Backlund E-O,
mania, apathy, and amotivation [29]. The exact incidence of editors. Radiosurgery: baseline and trends. New York: Raven; 1992.
p. 263–92.
these complications is difficult to discern based upon the
3. Leksell L. Sterotaxic radiosurgery in trigeminal neuralgia. Acta
existing reports. Chir Scand. 1971;137:311–4.
4. Leksell L. Trigeminal neuralgia. Some neurophysiologic aspects
and a new method of therapy. Lakartidningen. 1971;68:5145–8.
5. Hakanson S. Transoval trigeminal cisternography. Surg Neurol.
Future Directions 1978;10:137–44.
6. Bingley T, Leksell L, Meyerson BA, et al. Long term results of
The treatment of psychiatric and pain disorders with stereo- stereotactic capsulotomy in chronic obsessive-compulsive neurosis.
tactic radiosurgery will require coordination by a multidisci- In: Sweet WH, Obrador AS, editors. Neurosurgical treatment in
psychiatry, pain and epilepsy. Baltimore: University Park Press;
plinary team. While a neurosurgeon and radiation oncologist
1977. p. 287–99.
can utilize modern stereotactic radiosurgery devices to create 7. Flickinger JC, Lunsford LD, Kondziolka D. Dose-volume consid-
precise lesions in the brain, the long-term efficacy and safety erations in radiosurgery. Stereotact Funct Neurosurg. 1991;57:
of this approach has not been well defined. Advanced neuro- 99–105.
8. Kamiryo T, Kassell NF, Thai QA, Lopes MB, Lee KS, Steiner
imaging modalities such as PET, SPECT, or diffusion tensor
L. Histological changes in the normal rat brain after gamma irradia-
imaging may help to demonstrate overactive areas of the tion. Acta Neurochir (Wien). 1996;138:451–9.
brain in patients afflicted with intractable psychiatric and 9. Mindus P. Capsulotomy in anxiety disorders—a multidisciplinary
pain disorders. Such imaging modalities may provide a bet- study. Dissertation at the Karolinska Institute and Hospital,
Stockholm, Sweden, 1991. p. 8
ter defined target for selected patients and enable clinicians
10. Mindus P. Capsulotomy in anxiety disorders—a multidisciplinary
to track a radiologic response to radiosurgery. study. Dissertation at the Karolinska Institute and Hospital,
11. Young RF, Jacques DS, Rand RW, Copcutt BC, Vermeulen SS,
Posewitz AE. Technique of stereotactic medial thalamotomy with
Conclusions the Leksell Gamma Knife for treatment of chronic pain. Neurol
Res. 1995;17:59–65.
Leksell’s personal battles with pain and his desire to perform 12. Young RF, Jacques DS, Rand RW, Copcutt BR. Medial thalamot-
more elegant psychosurgeries than prefrontal lobotomies led omy with the Leksell Gamma Knife for treatment of chronic pain.
Acta Neurochir Suppl. 1994;62:105–10.
to the use of radiosurgical devices for the treatment of psy-
13. Young RF, Vermeulen SS, Grimm P, Posewitz AE, Jacques DB, Rand
chiatric disorders and pain. With the exception of trigeminal RW, et al. Gamma Knife thalamotomy for the treatment of persistent
neuralgia, the application of radiosurgery to treat these prob- pain. Stereotact Funct Neurosurg. 1995;64 Suppl 1:172–81.
lems has not been widely embraced and is largely considered 14. Romanelli P, Esposito V, Adler J. Ablative procedures for chronic
pain. Neurosurg Clin N Am. 2004;15:335–42.
investigational. Obsessive-compulsive disorder has been the
15. Olivecrona H. Die trigeminusneuralgie und ihre behandlung.
only psychiatric disorder for which radiosurgery has been Nervenarzt. 1941;14:49–57.
reasonably well received. 16. Granit R, Leksell L, Skoglund CR. Fibre interaction in injured or
It is unfortunate that Leksell, who developed the Gamma compressed region of nerve. Brain. 1944;67:125–40.
17. Barker 2nd FG, Jannetta PJ, Bissonette DJ, Larkins MV, Jho
Knife for functional disorders, never saw it embraced for
HD. The long-term outcome of microvascular decompression for
such indications. Competing surgical approaches with deep trigeminal neuralgia. N Engl J Med. 1996;334:1077–83.
brain stimulation and pharmacological alternatives make the 18. Jannetta PJ. Arterial compression of the trigeminal nerve at the pons
widespread embracement of radiosurgery for pain and psy- in patients with trigeminal neuralgia. J Neurosurg. 1967;26(Suppl):
159–62.
chosurgery unlikely. Initial experiences dating over the past
19. Lovely TJ, Jannetta PJ. Microvascular decompression for trigemi-
30 years have demonstrated limited efficacy for radiosurgery nal neuralgia. Surgical technique and long-term results. Neurosurg
to treat some of the most intractable of psychiatric disorders Clin N Am. 1997;8:11–29.
20. Brown JA, Chittum CJ, Sabol D, Gouda JJ. Percutaneous balloon 41. Lunsford LD. Glycerol rhizotomy for trigeminal neuralgia.
compression of the trigeminal nerve for treatment of trigeminal neu- J Neurosurg. 1987;66:151–2.
ralgia. Neurosurg Focus. 1996;1(2):e4; discussion 1 p following e4. 42. Greenberg BD, Price LH, Rauch SL, Friehs G, Noren G, Malone D,
21. Brown JA, Gouda JJ. Percutaneous balloon compression of the et al. Neurosurgery for intractable obsessive-compulsive disorder
trigeminal nerve. Neurosurg Clin N Am. 1997;8:53–62. and depression: critical issues. Neurosurg Clin N Am. 2003;14:
22. Kondziolka D, Lunsford LD, Flickinger JC, Young RF, Vermeulen 199–212.
S, Duma CM, et al. Stereotactic radiosurgery for trigeminal neural- 43. Rasmussen SA, Eisen JL. The epidemiology and clinical features
gia: a multiinstitutional study using the gamma unit. J Neurosurg. of obsessive compulsive disorder. Psychiatr Clin North Am. 1992;
1996;84:940–5. 15:743–58.
23. Massager N, Lorenzoni J, Devriendt D, Desmedt F, Brotchi J, 44. Fava M, Davidson KG. Definition and epidemiology of treatment-
Levivier M. Gamma knife surgery for idiopathic trigeminal neural- resistant depression. Psychiatr Clin North Am. 1996;19:179–200.
gia performed using a far-anterior cisternal target and a high dose of 45. Pallanti S, Hollander E, Goodman WK. A qualitative analysis of
radiation. J Neurosurg. 2004;100:597–605. nonresponse: management of treatment-refractory obsessive-
24. Zhang P, Brisman R, Choi J, Li X. Where to locate the isocenter? compulsive disorder. J Clin Psychiatry. 2004;65 Suppl 14:6–10.
The treatment strategy for repeat trigeminal neuralgia radiosurgery. 46. Blomstedt P, Hariz MI. Hardware-related complications of deep
Int J Radiat Oncol Biol Phys. 2005;62:38–43. brain stimulation: a ten year experience. Acta Neurochir (Wien).
25. Pollock BE, Kondziolka D. Stereotactic radiosurgical treatment of 2005;147:1061–4.
sphenopalatine neuralgia. Case report. J Neurosurg. 1997;87: 47. Kondziolka D, Whiting D, Germanwala A, Oh M. Hardware-
450–3. related complications after placement of thalamic deep brain stimu-
26. Stieber VW, Bourland JD, Ellis TL. Glossopharyngeal neuralgia lator systems. Stereotact Funct Neurosurg. 2002;79:228–33.
treated with gamma knife surgery: treatment outcome and failure 48. Smyth MD, Tubbs RS, Bebin EM, Grabb PA, Blount
analysis. Case report. J Neurosurg. 2005;102(Suppl):155–7. JP. Complications of chronic vagus nerve stimulation for epilepsy
27. Mindus P. Capsulotomy in anxiety disorders—a multidisciplinary in children. J Neurosurg. 2003;99:500–3.
study. Dissertation at the Karolinska Institute and Hospital, 49. Jannetta PJ. Outcome after microvascular decompression for typi-
Stockholm, Sweden, 1991. p. 15 cal trigeminal neuralgia, hemifacial spasm, tinnitus, disabling posi-
28. Leksell L, Herner T, Leksell D, Persson B, Lindquist C. Visualisation tional vertigo, and glossopharyngeal neuralgia (honored guest
of stereotactic radiolesions by nuclear magnetic resonance. lecture). Clin Neurosurg. 1997;44:331–83.
J Neurol Neurosurg Psychiatry. 1985;48:19–20. 50. Burchiel KJ, Clarke H, Haglund M, Loeser JD. Long-term efficacy
29. Lippitz BE, Mindus P, Meyerson BA, Kihlstrom L, Lindquist C. of microvascular decompression in trigeminal neuralgia.
Lesion topography and outcome after thermocapsulotomy or J Neurosurg. 1988;69:35–8.
gamma knife capsulotomy for obsessive-compulsive disorder: 51. Friehs GM, Noren G, Ohye C, Duma CM, Marks R, Plombon J,
relevance of the right hemisphere. Neurosurgery. 1999;44:452–8; et al. Lesion size following Gamma Knife treatment for functional
discussion 458–60. disorders. Stereotact Funct Neurosurg. 1996;66 Suppl 1:320–8.
30. Cosgrove GR, Rauch SL. Stereotactic cingulotomy. Neurosurg Clin 52. Frighetto L, De Salles A, Wallace R, Ford J, Selch M, Cabatan-
N Am. 2003;14:225–35. Awang C, et al. Linear accelerator thalamotomy. Surg Neurol.
31. Kelly D, Mitchell-Heggs N. Stereotactic limbic leucotomy—a 2004;62:106–13; discussion 113–4.
follow-up study of thirty patients. Postgrad Med J. 1973;49:865–82. 53. Brisman R. Gamma knife surgery with a dose of 75 to 76.8 Gray for
32. Kelly D, Richardson A, Mitchell-Heggs N, Greenup J, Chen C, trigeminal neuralgia. J Neurosurg. 2004;100:848–54.
Hafner RJ. Stereotactic limbic leucotomy: a preliminary report on 54. Petit JH, Herman JM, Nagda S, DiBiase SJ, Chin LS. Radiosurgical
forty patients. Br J Psychiatry. 1973;123:141–8. treatment of trigeminal neuralgia: evaluating quality of life and
33. Richardson A. Stereotactic limbic leucotomy: surgical technique. treatment outcomes. Int J Radiat Oncol Biol Phys. 2003;56:
Postgrad Med J. 1973;49:860–4. 1147–53.
34. Kihlstrom L, Guo WY, Lindquist C, Mindus P. Radiobiology of 55. Pollock BE, Phuong LK, Foote RL, Stafford SL, Gorman DA. High-
radiosurgery for refractory anxiety disorders. Neurosurgery. dose trigeminal neuralgia radiosurgery associated with increased
1995;36:294–302. risk of trigeminal nerve dysfunction. Neurosurgery. 2001;49:
35. Kihlstrom L, Hindmarsh T, Lax I, Lippitz B, Mindus P, Lindquist C. 58–62; discussion 62–4.
Radiosurgical lesions in the normal human brain 17 years after 56. Richards GM, Bradley KA, Tome WA, Bentzen SM, Resnick DK,
gamma knife capsulotomy. Neurosurgery. 1997;41:396–401; discus- Mehta MP. Linear accelerator radiosurgery for trigeminal neural-
sion 401–2. gia. Neurosurgery. 2005;57:1193–200; discussion 1193–200.
36. Khadilkar A, Milne S, Brosseau L, Robinson V, Saginur M, Shea B, 57. Sheehan J, Pan HC, Stroila M, Steiner L. Gamma knife surgery for
et al. Transcutaneous electrical nerve stimulation (TENS) for trigeminal neuralgia: outcomes and prognostic factors. J Neurosurg.
chronic low-back pain. Cochrane Database Syst Rev. 2005: 2005;102:434–41.
CD003008 58. Tawk RG, Duffy-Fronckowiak M, Scott BE, Alberico RA, Diaz
37. Multon S, Schoenen J. Pain control by vagus nerve stimulation: from AZ, Podgorsak MB, et al. Stereotactic gamma knife surgery for
animal to man … and back. Acta Neurol Belg. 2005;105:62–7. trigeminal neuralgia: detailed analysis of treatment response.
38. Turk DC. Clinical effectiveness and cost-effectiveness of treat- J Neurosurg. 2005;102:442–9.
ments for patients with chronic pain. Clin J Pain. 2002;18:355–65. 59. Williams BJ, Schlesinger D, Sheehan J. Glossopharyngeal neural-
39. Hamid AI, Qureshi AA, Bhatti IH. Percutaneous radiofrequency gia treated with gamma knife radiosurgery. World Neurosurg.
retrogasserian rhizotomy for trigeminal neuralgia. J Pak Med 2010;73:413–7.
Assoc. 1993;43:132–3. 60. Pollock BE, Boes CJ. Stereotactic radiosurgery for glossopharyn-
40. Henson CF, Goldman HW, Rosenwasser RH, Downes MB, Bednarz geal neuralgia: preliminary report of 5 cases. J Neurosurg. 2011;
G, Pequignot EC, et al. Glycerol rhizotomy versus gamma knife 115:936–9.
radiosurgery for the treatment of trigeminal neuralgia: an analysis 61. Ma L, Kwok Y, Chin LS, Yu C, Regine WF. Comparative analyses
of patients treated at one institution. Int J Radiat Oncol Biol Phys. of linac and Gamma Knife radiosurgery for trigeminal neuralgia
2005;63:82–90. treatments. Phys Med Biol. 2005;50:5217–27.
62. Smith ZA, De Salles AA, Frighetto L, Goss B, Lee SP, Selch M, 67. Kondziolka D, Flickinger JC, Hudak R. Results following gamma
et al. Dedicated linear accelerator radiosurgery for the treatment of knife radiosurgical anterior capsulotomies for obsessive compul-
trigeminal neuralgia. J Neurosurg. 2003;99:511–6. sive disorder. Neurosurgery. 2011;68:28–32.
63. Steiner L, Forster D, Leksell L, Meyerson BA, Boethius J. 68. Price BH, Baral I, Cosgrove GR, Rauch SL, Nierenberg AA, Jenike
Gammathalamotomy in intractable pain. Acta Neurochir (Wien). MA, et al. Improvement in severe self-mutilation following limbic
1980;52:173–84. leucotomy: a series of 5 consecutive cases. J Clin Psychiatry. 2001;
64. Hayashi M, Taira T, Chernov M, Izawa M, Liscak R, Yu CP, et al. 62:925–32.
Role of pituitary radiosurgery for the management of intractable 69. Ganz JC. Gamma knife radiosurgery and its possible relationship to
pain and potential future applications. Stereotact Funct Neurosurg. malignancy: a review. J Neurosurg. 2002;97:644–52.
2003;81:75–83. 70. Lim YJ, Leem W, Park JT, Kim TS, Rhee BA, Kim GK. Cerebral
65. Mindus P. Capsulotomy in anxiety disorders—a multidisciplinary infarction with ICA occlusion after Gamma Knife radiosurgery for
study. Dissertation at the Karolinska Institute and Hospital, pituitary adenoma: a case report. Stereotact Funct Neurosurg.
Stockholm, Sweden, 1991. p. 43 1999;72 Suppl 1:132–9.
66. Mindus P. Capsulotomy in anxiety disorders—a multidisciplinary
study. Dissertation at the Karolinska Institute and Hospital,
Ocular and Orbital Lesions
62
Yan Michael Li, Gabriela Šimonová, Roman Lisčák,
Josef Novotný Jr., Amit Singla, and Lawrence S. Chin
doses in a single session to well-defined intracranial targets.

Introduction The first experiences with ophthalmologic indications
involved the treatment of patients suffering from uveal mela-
Radiation, including radiosurgery, plays an indispensable nomas [3–5]. However, it would appear that the potential of
role in the treatment of benign and malignant orbital and radiosurgery to provide effective treatment for ophthalmo-
ocular diseases, since the clinical benefits of improving sur- logic indications is far greater and, besides uveal melano-
vival and preserving vision in many cases outweigh the risks. mas, the current spectrum of treated indications also includes
Technological advances in medical imaging, treatment plan- vascular lesions, eye metastases, advanced glaucoma, and
ning, and radiation dose delivery have led to dramatic age-related macular degeneration. This chapter reviews the
improvements in radiosurgery and radiotherapy [1, 2]. Now radiobiology, clinical indications and applications, technical
it is also possible to recognize intraocular tumors and other aspects, and side effects and complications of stereotactic
ophthalmologic diseases at an earlier stage because of the radiosurgery in the treatment of benign and malignant orbital
availability and use of both direct and indirect ophthalmos- and ocular lesions.
copy, fluorescein angiography, ultrasonography, and mag-
netic resonance imaging (MRI).
Orbital anatomy and ocular radiosensitivity provide Radiobiology and Technique
unique challenges for radiosurgery. The basic advantage of
radiosurgery is the possibility of applying relatively high Stereotactic radiosurgery (SRS) or stereotactic radiotherapy
(SRT) was originally designed for the treatment of rigid
centrally or almost centrally located intracranial targets [2],
but now also extends to cover the extracranial target [1, 6].
Y.M. Li, M.D., Ph.D. Stereotactic radiosurgery or radiotherapy of ophthalmic
Department of Neurosurgery, UT MD Anderson Cancer Center,
lesions is far different from the standard intracranial treat-
1400 Holcombe Boulevard, Unit 442, Houston, TX 77030, USA
e-mail: YanLi@gmail.com ment due to very eccentric target location and motion of the
eye. These aspects specific to the treatment of ophthalmic
G. Šimonová, M.D., Ph.D.
Department of Stereotactic Radioneurosurgery, Hospital Na lesions need to be taken into account. First of all, a proper
Homolce, Prague, Czech Republic eye fixation or eye motion monitoring system is required
R. Lisčák because the eye can move during the treatment procedure.
Department of Stereotactic Radioneurosurgery, Hospital Na Very eccentric target volume location can also cause some
Homolce, Prague, Czech Republic inaccuracies in the treatment planning calculations including
Department of Neurosurgery, Charles University, U Vojenské relative dose distribution as well as absolute dose calcula-
nemocnice 1200, Praha 6, Střešovice, tions (calculation of treatment time or monitor units). Finally,
Prague 169 02, Czech Republic
there may be also some technical inconveniences to treat
J. Novotný Jr. patients due to limitations in the range of coordinate system
Department of Radiation Oncology, University of Texas
in this very eccentric target location. Despite all these diffi-
Southwestern Medical Center, Dallas, TX, USA
culties, SRS or SRT is an excellent alternative for ophthal-
A. Singla, M.D. • L.S. Chin, M.D., F.A.C.S., F.A.A.N.S. (*)
mic lesion treatment mainly due to its very high accuracy
Department of Neurosurgery, SUNY Upstate Medical University,
750 East Adams Street, Syracuse, NY 13210, USA and delivery of high conformal dose distribution to the small
e-mail: singlaa@upstate.edu; chinL@upstate.edu volume of the treatment target.
744 Y.M. Li et al.
Eye Fixation
For all kinds of SRS or SRT, it is imperative to employ a

patient fixation system that is highly reproducible and meets
the accuracy requirements for high-precision treatments.
Additionally, for treatment of ocular lesions, it is not suffi-
cient to fixate the patient’s head but the eye itself has to be
fixated in the same position during imaging for treatment
planning and during delivery of all treatment fractions.
Fig. 62.1 (a) Example of invasive eye fixation by applying two sutures
Several different eye-immobilization systems and techniques in the rectus muscles and attaching these sutures to the Leksell stereo-
have been described in the literature [4, 7–9]. In principle, tactic frame. (b) Example of noninvasive eye fixation used for image-
eye-immobilization techniques can be divided into two guided/gated SRT (From Petersch B, Bogner J, Dieckmann K, et al.
Automatic real-time surveillance of eye position and gating for stereo-
groups: (1) passive immobilization techniques where the eye
tactic radiotherapy of uveal melanoma. Med Phys 2004; 31:3521–3527.
is immobilized by different mechanical means from outside Used with permission)
and (2) active immobilization techniques where the patient
controls the eye position (e.g., by fixation on a light source).
Tokuuye et al. described a mask fixation technique with a the system is able automatically initiate interruption of
plastic mold gently pressed down over the orbit to restrict patient irradiation.
ocular movements [5]. Zehetmayer et al. reported on a suc- When invasive stereotactic frame fixation to a patient’s
tion immobilization technique for SRS of intracranial malig- head or invasive eye fixation is used, then the treatment itself
nancies at the Leksell Gamma Knife [9]. Langmann et al. is limited to one single fraction or a very small number of
used an invasive fixation method of the globe by means of fractions due to patient comfort. A prerequisite for fraction-
retrobulbar anesthetic blocking for eye treatments with the ated SRT is the reproducible and reliable immobilization of
Leksell Gamma Knife [10]. Simonova et al. reported a series the irradiated structures throughout the whole treatment.
of patients with uveal melanomas treated with the Leksell Whereas this is achieved in the region of the skull by invasive
Gamma Knife when an eye fixation was achieved by apply- or noninvasive fixation systems, treatments of intraocular
ing two sutures in the rectus muscles and attaching these lesions are more complicated due to the additional degrees of
sutures to the Leksell stereotactic frame [4]. The same tech- freedom caused by eye movements. Consequently, special
nique of eye fixation was also reported by Vladyka et al. eye motion monitoring devices as described above have to be
when patients with glaucoma were treated with the Leksell employed during radiation delivery.
Gamma Knife [7]. An example of invasive fixation by means of eye fixation
Active eye-immobilization techniques, where the patient by applying two sutures in the rectus muscles and attaching
controls the eye position (e.g., by fixation on a light source), these sutures to the Leksell stereotactic frame and an exam-
have been described in the literature for proton/helium ion ple of noninvasive eye fixation used for image-guided/gated
beam therapy and recently also for linac-based SRT of SRT are presented in Fig. 62.1.
malignant tumors [11]. For proton/helium ion beam therapy,
a surgical intervention is performed to position tantalum
clips at the border of the visible tumor under diaphanoscopi- Treatment Planning
cal control. These clips are used during treatment planning
to define the target volume and to verify the position of the A prerequisite for correct treatment planning is good imag-
tumor during treatment delivery by using X-rays [12]. ing of the target and surrounding structures, especially
Recently, Petersch et al. described a new, noninvasive eye organs at risk. A typical imaging modality for ophthalmic
fixation system for the application of linac-based image- lesions is MRI, which can be supplemented by computed
guided/gated SRT of uveal melanoma [8]. The system is tomography (CT). Generally, MRI brings better contrast
attached to a commercially available head-and-neck ther- compared with CT when imaging intracranial and intraocu-
moplast mask system and is based on the patient’s fixation lar structures. However, whereas CT is supposed to be free
on a light point. Computer-controlled eye monitoring proof geometric distortions, MRI may produce some geomet-
vides quantitative information about the quality of the ric image distortions (sometimes a few millimeters) that are
patient repositioning, both about the repositioning of the not acceptable for SRS or SRT treatment planning.
patient’s head within the mask and about the eye’s rotational Generally, MRI geometric distortion is mainly a function of
state with respect to the reference position as determined the MRI scanner employed, the MRI sequence used, image
during imaging for treatment planning prior the actual treat- slice orientation, position in imaged object, and material
ment. When eye position exceeds the preset geometric limits, and geometric parameters of the stereotactic frame [13, 14].
62 Ocular and Orbital Lesions 745
Treatment of eye lesions with the Leksell Gamma Knife

requires typically an extremely large neck rake (when supine
position is used) that can be hardly tolerated or even impos-
sible to perform for some patients. Consequently, for a
patient’s better comfort, the treatment is typically done in the
prone patient position (Fig. 62.2).
The most up-to-date system called the LGK PERFEXION
was introduced in 2006 [15]. The entire radiation unit was
redesigned with a beam geometry different from that of the
previous Gamma Knife Models U, B, C, and 4C. A total of
Fig. 62.2 (a) Special y coordinate extender developed at Na Homolce
192 Co-60 sources are arranged in a cylindrical configura-
Hospital used for the treatment of eye lesion when y coordinate exceeds tion in five concentric rings. This differs substantially from
the limits of the commercially available Leksell stereotactic frame. the previous hemispherical arrangements and results in a dif-
(b) Typical treatment of eye lesion on the Leksell Gamma Knife with ferent source-to-focus distance for each ring that varies from
patient in the prone position
374 to 433 mm. The primary and secondary collimators are
replaced by a single large 120-mm-thick tungsten collimator
The issue of image geometric distortion is even more crucial array ring, which eliminates the need for changing different
in the case of eye imaging when a target and other structures collimator helmets and allows greater flexibility in treatment
of interest are in a very frontal location. It is essential to planning by mixing collimator sizes with each dose of radia-
check the extent of geometric distortions for each MRI scan- tion. There are greater time savings in treatment length and
ner and sequence employed for eye imaging. improved safety features for patients and practitioners.
Improved helmet clearance allows for fewer modifications to
Leksell Gamma Knife treat eccentrically located orbital lesions.
Treatment planning for intraocular targets is very similar as
for intracranial targets in the case of the Leksell Gamma Linear Accelerator
Knife. Four different collimator helmets (4, 8, 14, 18 mm) The LINAC is now the most frequently used device for deliv-
are available to create isocenters of essentially spherical dose ery of conventional radiotherapy and radiosurgery. Linear
distributions. To cover optimally the target volume, a combi- accelerators use electromagnetic waves to accelerate a
nation of all four collimator helmets can be used. Dose dis- stream of electrons to nearly the speed of light and then hit a
tribution can be tailored to the target volume by means of high-density metal target. A small portion of the energy is
proper isocenter size selection, proper position in the stereo- converted to X-rays through the process of braking radiation
tactic space, and proper weighting factor for each isocenter. and this produces a poly-energetic beam of photons (X-rays)
To protect surrounding healthy tissue and mainly structures with energies in the MV range. The treatment plans can be
at risk, additional shaping of dose distribution can be done individualized for each patient by adjusting the variables of a
by blocking any of the 201 Cobalt-60 beams. Shielding of multiple arc plan. These variables include the number of
properly selected number of beams can effectively reduce arcs, the gantry start and end angles of each arc (amplitude),
dose to critical structures such as optic nerve or eye lens. The the couch angle (the separation between arcs), different beam
target volume is usually covered by 50 % isodose line. energies (arc weighting), and different sized collimators
Treatment planning process is also a compromise between (5–50 mm).
sufficient conformity required for the treated target and treat- Non-coplanar arcs represent the standard treatment tech-
ment time (with increased number of isocenters, total treat- nique. Non-coplanar arcs treatments use circular collimators
ment time is increased). with fixed diameters. The resulting isodose distribution is
Because of the very frontal location of eye volume, some essentially spherical and very similar to the one created in
mechanical difficulties can occur when treating a patient with the case of the Leksell Gamma Knife. There is usually a
the Leksell Gamma Knife. Depending on a patient’s head wide range (typically 5–40 mm) of circular collimators
shape and volume, it might be difficult to adjust properly the depending on the manufacturer of the system. Individual
frame position to reach all Leksell stereotactic coordinates. dose distribution can be further optimized by combination of
Specifically, y (posterior–anterior) Leksell stereotactic coor- proper number and size of circular collimators. However,
dinate can sometimes exceed limits engraved on the Leksell total number of isocenters is much smaller compared with
stereotactic frame. Because of this, a special y coordinate the Leksell Gamma Knife (usually does not exceed 5).
extender was developed at Na Homolce Hospital to be able to Another option is non-coplanar conformal static fields.
treat also y coordinates exceeding the limits of the commer- Non-coplanar conformal static fields can be applied using
cially available Leksell stereotactic frame (Fig. 62.2). customized blocks or miniature-multileaf collimator with leaf
746 Y.M. Li et al.
width in the millimeter range. The use of miniature-multileaf ing device accurately if beam passed perpendicularly through
collimator has several advantages compared with custom- it. Consequently, Petersch et al. suggested avoidance of fron-
ized blocks: (1) generating miniature-multileaf collimator tal beams when doing treatment planning as a solution to
shapes takes less time than molding blocks, (2) there is no reduce significantly dosimetric errors.
need to reenter the treatment room during multiple field Detailed dosimetric measurements including the treat-
treatments, and (3) any modification in the shape can be ment planning system calculations of target volume relative
performed easily. Single-isocenter technique with selected dose distributions and absolute dose calculations in different
number of conformal fields (typically 5–15) is used. depths were compared with experimental measurements.
Individual fields are shaped using beam’s-eye view tech- Various treatment plans for different ophthalmic indications
nique. Weighting factors of single individual beams, gantry (uveal melanoma, glaucoma, retinoblastoma) calculated for
angles, and table angles are used for final dose distribution the Leksell Gamma Knife were evaluated. There was
optimization. observed good agreement between treatment planning sys-
Dynamic non-coplanar arcs are another treatment tech- tem relative dose distribution calculations for target volume
nique. A selected number of conformal isocentric arcs are and performed measurements (Fig. 62.3). For the calculation
used. Beam’s-eye view technique is used for individual beam of absolute doses, there was observed very good agreement
shaping. at depths 15–20 mm. Typical deviations between treatment
Finally, intensity-modulated SRS or SRT is an option for planning system calculations and actual measurements were
the treatment planning. A selected number of fields opti- within 3 %. However, measurements performed at depth less
mized by proper weighting factor, gantry, and table angles than 10 mm showed large deviations of about 15–20 % com-
are used as in the non-coplanar conformal static fields tech- pared with the treatment planning system calculations. The
nique. Inverse treatment planning is used to generate optimal dose delivered to patient was always smaller than that calcu-
intensity through individual beams. lated by the treatment planning system. To improve inaccu-
Georg et al. compared different dosimetric characteristics racies in absolute dose calculations in surface structures, it
of treatment plans when using the above-mentioned tech- was suggested to use proper tissue equivalent built-up mate-
niques for the treatment planning of uveal melanomas [16]. rial that is attached on the treated eye (Fig. 62.3) and
They concluded that static conformal and dynamic as treat- improves treatment planning system calculations.
ment techniques that have dosimetric advantages over con-
ventional non-coplanar arcs using circular collimators while
being simultaneously highly efficient in treatment planning Clinical Application
and delivery. Dynamic arc SRS or SRT combines the homo-
geneous dose distribution and the high degree of static con- Stereotactic radiosurgery or radiotherapy currently plays an
formal beams with the steep dose gradients and smeared-out important role as primary treatment, adjuvant therapy, or pal-
low-dose volumes of arc beam therapy. Intensity-modulated liation of ocular and orbital tumors. The clinical situation,
SRS or SRT does not show clear impact on further reduction radiation source, and desired result determine the prescribed
of doses to organs at risk and depends probably on the target radiation dose. The following reviews the published experi-
volume location within the eye. ence with stereotactic radiosurgery or radiotherapy for ocu-
lar and orbital lesions (Table 62.1) [3, 4, 17–41].
Dosimetry
Uveal Melanomas
Because of very eccentric target volume location, some inac-
curacies in the treatment planning calculations including Pathophysiology
relative dose distributions as well as absolute dose calcula- Uveal melanoma represents the most common primary malig-
tions (calculation of treatment time or monitor units) can be nant tumor of the eye in adults with a peak incidence between
expected. This issue is even more important for surface ages 55 and 70 years. It affects 6–7 people per million each
structures of eye (e.g., eye lens, eye lid, cornea) than for tar- year, and approximately 50 % of patients with diagnosed
get volumes itself as these structures are located in depths melanomas of the chorioid or ciliary body will die from this
smaller than buildup of employed photon energies. tumor within 15 years [17]. The incidence is rare in patients
Petersch et al. reported excellent agreement between younger than 20 years. The incidence of cutaneous melano-
measurements and treatment planning system calculations mas has been increasing in frequency over the past several
for uveal melanomas at depth 15–20 mm [8]. However, at decades and the dependence on latitude has been observed,
small depths (<10 mm), the treatment planning system can- reflecting levels of exposure to ultraviolet light. This trend
not model the influence of absorption in eye motion monitor- has not been evident for patients with uveal melanomas.
Table 62.1 Indications and radiotherapy doses for ocular and orbital
disease
Orbital radiation doses (Gy)
Pathology Average or recommended dose (Range)
Tumor
Uveal Melanoma 35 (35–50)
[3, 4, 30–33]
Retinoblastoma [35] 43 (40–50)
Orbital metastasis 40 (30–40)
(breast, lung) [38, 52]
Lymphoma [39] 32 (15–46)
Meningioma [53–55] 45 (45–54)
Optic nerve glioma [56] 50 (42–54)
Basal cell carcinoma [57] 50 (34–50)
Squamous carcinoma 50 (50–64)
[58, 59]
Sebaceous carcinoma [60] 55 (45–63)
Rhabdomyosarcoma 45 (32–72)
[61, 62]
Vascular
Cavernous hemangioma 14.5 (12.5–40)
[40, 41]
Choroidal haemangiomas 10 (10)
[42]
Uveal angiomas (Sturge 12 (12–30)
Weber) [63]
Capillary hemangioma— 4.5 (1.5–7.5)
orbit [43]
Others
Age-related macular 12 (12–20)
degeneration [47, 48]
Glaucoma 15 (7–15)
1 Gy = 100 Rads
The uveal tract is a vascular structure with no lymphatic

drainage. Lymphatic node involvement (auricular, subman-
dibular, and cervical nodes) can be diagnosed on rare occa-
sions when subconjunctival extension of primary tumor
Fig. 62.3 (a–c) Example of relative dose distribution treatment plan- has been observed. The major routes of spread are local
ning system calculations verification with polymer gel dosimeter.
growth and by bloodstream, with a predilection for hema-
(a) Three-dimensional treatment plan in LGP. (b) Polymerization that
occurred in the polymer gel. (c) Relative dose profiles (calculated by togenous dissemination to the liver, lung, and brain. Liver
LGP—solid line, measured with polymer gel—dotted line). (d) Tissue- metastases are diagnosed in two-thirds of patients with
equivalent built-up material that is attached to the treated eye to improve disseminated diseases. Systemic examination is recom-
inaccuracies in absolute dose calculations in surface structures.
mended for all patients and includes CT of the chest and
(e) Corresponding MR localization scan visualized in LGP together
with head contour abdomen, and MRI of the brain to detect metastases. Wide
variations in the grading of malignancy can be observed,
from relatively benign types with several years’ survival
Uveal melanomas are neoplasm arising from the uveal tract, without dissemination, to other tumors with fast multiorgan
which includes the iris, ciliary body, and chorioid. Two types dissemination, which lead to the patient’s death within a
of tumor growth were recognized, nodular and infiltrative. few months.
Histologically, the tumor consists of spindle cells, epithelioid
cells, and intermediate cells similar to those of cutaneous Tumor Location
melanoma, but the prognosis is better than that associated The iris melanomas (incidence about 10 %) are single
with cutaneous melanomas. or multiple elevated lesions with changes of iris color.
748 Y.M. Li et al.
These types of melanomas are much more benign, smaller in Several staging systems have been used, one of the most
size, readily visible, and early diagnosed. It can be difficult commonly reported was based on the 2010 American Joint
to assess the potential of the malignant cells, even in those Staging System (Tables 62.2 and 62.3) [19]
with documented growth, because iris nevi may also grow.
For this reason, local resection is usually effective whenever
it seems to be necessary. Treatment Methods
Choroidal melanomas typically grow in nodular form,
and in cases of continuing growth many tumors break Treatment modalities strongly depend on the site of the primary
through the Bruch membrane. The Bruch membrane is a tumor and its volume and metastatic extent. Historically, the
basement membrane–connective tissue complex lying first known treatment modality was enucleation [20].
between the retina pigment epithelium and the choriocapil-
laries. When the melanomas break through the Bruch mem- Surgical Treatment
brane, they extend into the subsensory retina space, which Surgical methods include resection of the eye wall (for very
gives them a mushroom or collar-button shape. These char- small visible melanomas of the iris), enucleation, and orbital
acteristics, as well as the detachment of the retina, are typical exenteration [20]. The role of enucleation in the management
for uveal melanomas. This type of tumor growth is some- of posterior to equator localized melanomas is controversial,
times accompanied by vitreous hemorrhage with symptoms but enucleation remains the standard option for most large
of sudden visual loss. choroidal melanomas that cause severe secondary glaucoma
Melanomas involving the ciliary body are rare, carry a or are invading the optic nerve.
poor prognosis, and are often diagnosed late. Melanomas
can also have diffuse, infiltrative growth to the ciliary body Radiotherapy
and a pattern called ring melanoma. These tumors are diag- Preoperative or postoperative irradiation is a promising
nosed with difficulty and late, because they are accompanied approach under evaluation that could also decrease the
by relatively little visible tumor mass effect and they occur in chances of malignant dissemination, however recent study
the diagnostically “silent” eye area. Any extra scleral exten- observed no benefit [21]. Postoperative external beam radio-
sion of the tumor usually confers a poor prognosis. The dif- therapy has been used in patients with large tumors with
fuse growing melanomas are usually more aggressive, with extrascleral or extraocular extension. The recommended
more malignant cell types. total applied dose is 60 Gy with daily fractionation.
Tumor growth is also influenced on the potential doubling
time (T pot) which varies widely between 60 and 350 days Brachytherapy
(the median of T pot being about 70 days) [18]. A longer T The most frequently used therapeutic system for delivering
pot for a significant proportion of tumors can explain the radiation is brachytherapy (plaque radiotherapy). A number
relatively long survival of these patients and also the rela- of radionuclides, including cobalt-60, gold-198, iodine-125,
tively slow tumor response after irradiation. ruthenium-106, iridium-192, and palladium-103, have been
used for plaque therapy with varying results according to the
retrospective studies [22, 23]. A metal shield containing
Primary Tumor Diagnosis small radioactive seeds is sutured to the outside of the sclera
overlying the tumor with radiation doses principally deliv-
Melanoma of the ciliary body and chorioid have typical ered to the base of the tumor. The radioactive plaque is left in
clinical features, and the correct diagnosis for a majority of place for approximately 5–7 days. Cobalt-60 was one of the
patients can be made by taking a history and performing a first isotopes to be used in brachytherapy, but it is a high-
complete ocular examination. The ophthalmologic exami- energy gamma emitter and thus there exists a risk of damag-
nation includes external ocular examination, indirect oph- ing healthy ocular structures several millimeters from the
thalmoscopy, fluorescein angiography, and ultrasonography source. Other isotopes have been used: iodine-125 is also a
A, B. Magnetic resonance imaging plays a fundamental role gamma emitter but has lower energy than cobalt-60 and,
in the treatment planning of radiosurgery or stereotactic because of this, it is associated with a lower rate of complica-
radiotherapy. tions. Brachytherapy leads to long-term local tumor control
Ultrasonography plays a basic role, not only at the time of in 85–90 % of patients and can be used to treat small mela-
diagnosis but during follow-up and is used to obtain precise nomas [22, 23]. The best results can be achieved in tumors
measurement of the tumor base and height. The height of less than 10 mm in height and with a base diameter not
tumor in particular represents one of the most important fac- exceeding 16 mm. Patients with extrascleral tumor exten-
tors influencing the treatment decision, and it is used during sion, ring melanoma, and tumor involvement of more than
follow-up to evaluate the tumor response. half of the eye are not suitable for plaque therapy.
Table 62.2 Definitions of TNM staging of uveal melanoma (American Joint Staging System, 2010) [19]
Primary tumor (T)
TX Primary tumor cannot be assessed
T0 No evidence of primary tumor
Iris
T1 Tumor limited to the iris
T1a Tumor limited to the iris not more than 3 clock hours in size
T1b Tumor limited to the iris more than 3 clock hours in size
T1c Tumor limited to the iris with secondary glaucoma
T2 Tumor confluent with or extending into the ciliary body, choroid, or both
T2a T2 with secondary glaucoma
T3 Tumor confluent with or extending into the ciliary body, choroid, or both, with scleral extension
T3a T3 and secondary glaucoma
T4 Tumor with extrascleral extension
T4a Tumor with extrascleral extension ≤5 mm in diameter
T4b Tumor with extrascleral extension >5 mm in diameter
Ciliary body and choroid
T1 Tumor size category 1
T1a Tumor size category 1 without ciliary body involvement and extraocular extension
T1b Tumor size category 1 with ciliary body involvement
T1c Tumor size category 1 without ciliary body involvement but with extraocular extension ≤5 mm in
diameter
T1d Tumor size category 1 with ciliary body involvement and extraocular extension ≤5 mm in diameter
diameter
diameter
diameter
T4e Any tumor size category with extraocular extension >5 mm in diameter
Regional lymph nodes (N)
NX Regional lymph nodes cannot be assessed
N0 No regional lymph node metastasis
N1 Regional lymph node metastasis
Distant metastasis (M)
M0 No distant metastasis
M1 Distant metastasis
M1a Largest diameter of the largest metastasis ≤3 cm
M1b Largest diameter of the largest metastasis 3.1–8.0 cm
M1c Largest diameter of the largest metastasis ≥8 cm
750 Y.M. Li et al.
Table 62.3 Anatomic stage/prognostic groups of uveal melanoma applied minimal dose ranged from 25 to 45 Gy in two fractions
(American Joint Staging System, 2010) [19] and 15 Gy in three equal fractions. Local tumor control,
Stage T N M defined as freedom of local progression, was reported in 98 %
I T1a N0 M0 of patients [29]. Further clinical studies using fractionated
IIA T1b–d N0 M0 stereotactic irradiation are justified to optimize the effective
T2a N0 M0 dose and fractionation pattern.
IIB T2b N0 M0
T3a N0 M0 Radiosurgery
IIIA T2c–d N0 M0
The first experiments with radiosurgery for the treatment of
T3b–c N0 M0
uveal melanomas were conducted using the rabbit eye mela-
T4a N0 M0
noma model, with reported tumor lethal dose range from 60
IIIB T3d N0 M0
to 90 Gy applied in a single session. A preliminary study of
T4b–c N0 M0
IIIC T4d–e N0 M0
11 treated patients with uveal melanomas was published in
IV Any T N1 M0 1992 and suggested that local tumor control could be
Any T Any N M1a–c achieved with a minimal dose between 60 and 90 Gy in a
single dose using the Leksell Gamma Knife. Some observa-
tions were made on patients after radiosurgery who had their
Proton Beam Therapy eyes secondarily enucleated. All eyes with secondary enucle-
Patients with tumors too large for brachytherapy or tumors ations presented an initial height exceeding 9 mm and also
considered too close to the optic nerve have been recom- had a large number of ciliary body tumors [30]. Several
mended for proton beam therapy (PBT). Early studies with authors reported the results of radiosurgery in uveal melano-
PBT recognized that severe damage was caused to the eyelid mas and some of them observed relatively high radiation-
and the anterior parts of the eye. Patients with tumors local- related toxicity [4, 30, 31].
ized near the eye surface had a higher incidence of severe Dinca et al. reported the UK experience in treating ocular
toxicity because the sparing effect of Bragg peak was absent. melanoma over the last 20 years [3]. One hundred seventy
Many of these patients had chronically painful and cosmeti- patients who were treated with Gamma Knife radiosurgery
cally unsatisfactory eyes. Local tumor control in 90–98 % of were compared with 620 patients treated with eye enucle-
patients was achieved by this method in a large series [24]. ation. Different peripheral doses (at the 50 % isodose) were
The technique was subsequently improved and the complica- employed. There was no significant difference in survival
tion rate has dropped accordingly. between the 35-Gy, 45-Gy, and 50–70-Gy groups when com-
pared between themselves (p = 0.168) and with the enucle-
Charged Particle Therapy ation group (p = 0.454). The 5-year survival rates were: 64 %
Helium ion irradiation is uniquely suited to the precise local- for 35 Gy, 62.71 % for 45 Gy, 63.6 % for 50–70 Gy, and
ization of irradiation because of the sharp falloff of the dose 65.2 % for enucleated patients. Clinical variables influencing
at the distal end of the Bragg peak, the sharp lateral edges of survival for radiosurgery patients were tumor volume
the beam, and the ability to tailor the depth penetration and (p = 0.014) and location (median 66.4 months for juxtapapil-
spread by a modulated Bragg peak. Uveal melanomas are lary vs. 37.36 months for peripheral tumors, (p = 0.001).
indicated for this treatment. When the charged particles are Regarding complications, using 35 Gy led to more than a
used to treat the uveal melanoma, the normal tissue of the 50 % decrease in the incidence of cataract, glaucoma and
anterior part of the eye can receive a radiation dose compa- retinal detachment, compared with the 45-Gy dose.
rable with the tumor dose. This fact causes radiation toxicity Retinopathy, optic neuropathy, and vitreous hemorrhage
in the anterior parts of the eye leading to side effects such as were not significantly influenced. Blindness decreased dra-
eyelash loss, lacrimal damage, conjunctival changes, dry eye matically from 83.7 % for 45 Gy to 31.4 % for 35 Gy
syndrome, keratitis, and neovascular glaucoma [25, 26]. (p = 0.006), as well as post-radiosurgery enucleation: 23.9 %
Local tumor control is achieved in 96 % of patients and the for 45 Gy vs. 6.45 % for 35 Gy (p = 0.018). Visual acuity,
enucleation rate is 19 % (3 % for local failure and 16 % due recorded up to 5 years post-radiosurgery, was significantly
to complications) [26]. better preserved for 35 Gy than for 45 Gy (p = 0.0003). It
appears that using 35 Gy results in a dramatic decrease in
Stereotactic Radiotherapy complications, visual loss, and salvage enucleation, while
The advantage of stereotactic irradiation using a linear accel- not compromising patient survival.
erator is the possibility of fractionated treatments compared Sarici et al. reported on 50 eyes from 50 consecutive
with the Leksell Gamma Knife. Fractionation may decrease patients with uveal melanoma that were treated with GKRS
late-radiation-related toxicity, and this fact was the subject of and had a median follow-up time of 40 months. They used a
intensive study [8, 9, 11, 16, 27–29]. The reported mean dose of 30 Gy, which is an eye-sparing outpatient option for
Table 62.4 Toxicity scoring system (RTOG/EORTC LENT SOMA)

Grade 1 Grade 2 Grade 3 Grade 4
Cornea Increased tearing on exam Noninfectious keratitis Infectious keratitis, Panophthalmitis, corneal scar,
corneal ulcer ulceration leading to perforation
of globe/loss of globe
Iris Rubeosis only Rubeosis, increased Neovascular Neovascular glaucoma without
intraocular pressure glaucoma, ability to ability to count fingers at 1 m;
count fingers at 1 m complete blindness
Optic nerve Afferent pupillary defect ≤¼ pallor with asymptomatic >¼ pallor or central Profound optic atrophy,
with normal-appearing nerve visual field defect scotoma complete blindness
Lens Asymmetric lenticular Moderate lenticular changes Moderate lenticular Severe lenticular changes
opacities, no visual loss with mild-moderate visual changes with several
loss visual loss
Retina Microaneurysms, nonfoveal Cotton wool spots Massive macular Opaque vitreous hemorrhage,
exudates, minor vessel exudation, focal complete retinal detachment,
attenuation, extrafoveal retinal detachment blindness
pigment changes
Sclera Loss of episcleral vessels ≤50 % scleral thinning >50 % scleral thinning Scleral or periosteal graft
required due to perforation
patients with medium- or large-sized posterior uveal melanomas

who are not eligible for brachytherapy or particle therapy. Imaging and Treatment Planning
The tumor control rate was 90 %, and the eye retention rate
was 82 %. Metastases developed in 18 % of patients with MRI plays a fundamental role in the imaging of ocular
14 % dying of metastases during follow-up. Cataracts (34 %) tumors for treatment planning. The native T1- and
and radiation maculopathy (30 %) were the most frequent T2-weighted sequences, postcontrast T1-weighted sequences
complications, and 14 % of patients developed neovascular and three-dimensional with slice thickness 1–3 mm have
glaucoma. Visual acuity (VA) also decreased significantly been used. The melanotic uveal melanomas are well defined
after treatment (p < 0.0001) [32]. on native T1 (Fig. 62.4a) and T2 (Fig. 62.4b) sequences and
Šimonová, et al. studied a group of 126 patients with have typical postcontrast enhancement on T1-weighted MRI
uveal melanoma treated using the Leksell Gamma Knife (Fig. 62.4c). This postcontrast enhancement of melanomas
over a period of 8 years with all survivors undergoing mini- differentiates the boundary between the tumor volume and
mal follow-up 24 months after the treatment. The late effect the accompanying retinal detachment (Fig. 62.4c). The rare
of normal tissue (LENT) subjective objective management amelanotic melanomas are difficult to differentiate on native
analysis (SOMA) scoring system is acceptable for measure- T1 (Fig. 62.5a) but are well defined on T2-weighted mag-
ment of radiation side effects (Table 62.4). netic resonance sequence (Fig. 62.5b). The postcontrast
Fig. 62.4 Melanotic type of uveal melanoma. (a) Native T1-weighted MRI. (b) T2-weighted MRI. (c) Postcontrast T1-weighted MRI
752 Y.M. Li et al.
Fig. 62.5 Amelanotic type of uveal melanoma. (a) Native T1-weighted MRI. (b) T2-weighted MRI. (c) Postcontrast three-dimensional MRI
Fig. 62.6 (a) Axial, (b) coronal plane, and (c) sagittal plane. Uveal volume is 183.3 mm3, minimum dose is 44 Gy on a 60 % isodose, maxi-
melanoma. Characteristics: The tumor base is 9 mm and the tumor mum dose is 73.3 Gy. The maximum doses to critical structures: lens
height is 3 mm. Gross tumor volume is 84 mm3, the planned treatment 4 Gy, cornea 2 Gy, eyelid 2 Gy
enhancement is not so intensive as in the melanotic forms ultrasonography, indocyanine and fluorescein angiography,
(Fig. 62.5c). The treatment planning volume (PTV) includes optical coherence tomography, ultrabiomicroscopy (for
the growth tumor volume and 1–2-mm safety margins in all tumors located anterior to the equator), visual function, and
three diameters. The minimal dose for radiosurgery ranged biomicroscopy.
from 35 to 45 Gy with a median dose of 41 Gy in our series.
The majority of melanomas are irradiated using 4- and 8-mm
collimator helmets (Fig. 62.6) and the plugging pattern aim Local Tumor Response
to spare critical structures such as cornea (Fig. 62.7) or optic
nerve (Fig. 62.8). Tumor regression is defined as a decrease in tumor height
registered by A and B ultrasonography scans and by control
MRI. In our personal experience, the tumor regression can
Follow-Up be achieved in 70 % of patients. The maximum local effect
has been recorded during the interval of 20–30 months after
Patients should be examined by an ophthalmologist at regu- radiosurgery. Most tumor response has been achieved within
lar intervals (24 h, 1 week, 3 months, and then every 6 24 months after SRS (Figs. 62.9, 62.10, 62.11, and 62.12).
months after SRS; MRI can be repeated every year for the Patients with tumor growth progression or without any
first 2 years after radiosurgery and then every 2–3 years response 30 months after treatment were candidates for
after the treatment). The following ophthalmologic methods retreatment by radiosurgery, and patients with a tumor height
were employed for diagnostic investigation and patient fol- of more than 10 mm or extra scleral growth were indicated
low-up at regular intervals 6 months after radiosurgery: for enucleation (5 %).
Fig. 62.7 Demonstration of the effect of plugged sources for protec- plan with plugging. The dose to cornea in this case did not exceed
tion of cornea for patient with uveal melanoma treated by 8-mm and 14 Gy (given on displayed 20 % isodose). (c) Plug pattern used for
14-mm collimators. (a) Treatment plan without plugging. (b) Treatment 14-mm collimator. Plugged sources are indicated in black
Fig. 62.8 Demonstration of the effect of plugged sources for protec- plan with plugging. (c) Plug pattern used for plugging both 4-mm col-
tion of optic nerve for patient with uveal melanoma treated by two limators. Plugged sources are indicated in black
4-mm collimators. (a) Treatment plan without plugging. (b) Treatment
Fig. 62.9 Post-equatorial uveal melanoma. (a) Pretreatment funduscopy and treatment planning MRI. (b) Native T1-weighted MRI.
(c) T2-weighted MRI
754 Y.M. Li et al.
Fig. 62.10 (a) The tumor response 12 months after radiosurgery and the development of postradiation retinopathy (funduscopy control). (b, c)
Control MRI 12 months after radiosurgery
Prognostic Factors
The main aim of radiosurgery is not only to preserve the bulb

and visual function but also to achieve long-term local tumor
control and to protect against hematologic dissemination of
malignant melanoma. The prognosis of patients with uveal
melanoma is determined by tumor size, cell type, tumor
location (poor results for ciliary body tumors), and the extent
of the disease [4, 12, 20, 21, 23, 25]. Five-year survival rates
Fig. 62.11 (a) Pretreatment, postcontrast T1-weighted MRI. of patients with small tumors (heights less than 8 mm) and
(b) T2-weighted MRI with no detected metastatic disease outside the eye are above
80 %, and for lager tumors (over 10 mm in height) with
metastases outside the eye, the 5-year survival period ranged
from 10 to 30 %. The most important prognostic factor is the
presence or absence of the organ dissemination at the time of
uveal melanoma diagnosis. The potential dissemination has
been discussed several times. Some authors maintain that
dissemination from uveal melanomas starts when the tumor
is larger than 7 mm in diameter, and growth from a 7- to
10-mm diameter increases the risk of metastatic disease to
approximately 16 % [18]. There still exists a serious gap in
our knowledge of growth of melanomas in situ and the docu-
mentation of the time intervals between the diagnosis of pri-
mary tumor and the appearance of the other organ
Fig. 62.12 Control MRI 18 months after radiosurgery. (a) Post- dissemination. Tumor growth strongly depends on tumor
contrast T1-weighted MRI. (b) T2-weighted MRI doubling time. The numerous accounts of tumor doubling
times in uveal melanomas have been reported, and in the
majority of them the estimated doubling time was longer
than 69 days. On theoretical grounds, it has been suggested
Survival that metastatic death (after the primary tumor has released
malignant cells into the systemic circulation) will occur at a
Patients younger than 50 years have the best prognosis, time interval derived from multiplication of the tumor dou-
with a pre-equatorial location of the tumor, when tumor bling time by 35–40 [18]. Taking these two assumptions
height does not exceed 5 mm, gross tumor volume is not together, it may be accepted for practical prognostic consid-
larger than 500 mm3, and there is no other organ eration in individual patients that death from dissemination
dissemination. is unlikely to occur before 6 years after dissemination of the
first viable embolus of malignant tumor cells. The minimum answered the primary study questions and contributed to
latent interval of 6 years implies that metastatic death within clinical and epidemiologic knowledge of choroidal mela-
7–8 years after local treatment on the uveal melanoma is noma. As a consequence of the COMS, a standard approach
nearly always due to pretreatment dissemination [29]. The to I-125 brachytherapy for treatment of choroidal melanoma
chance of metastasis-caused death within 5 years totally became widely available to affected patients [22].
depends on the presence or absence of hematologically dis- Proton beam therapy was analyzed in Switzerland, and
seminated tumor cells or diagnosed metastases at the time of the treatment results over 15-year periods were reported.
local treatment. Some data from the analyzed clinical studies Effective local tumor control was achieved in 90 % and later
strongly supports the assumption that the rate of death by in 96 % of treated patients. Ten years survival of patients
metastatic disease is influenced by local tumor control: with local controlled tumors was 72.6 and 47.5 % in patients
improvement in local tumor control rate results in a longer with local recurrence of melanoma. The helium ions therapy
survival. One of the fundamental experiments and analysis trial documented local tumor control of 85–90 % [25, 26].
of results was reported from study based on 15 years of treat- The radiosurgery series reported the local tumor control of
ment of uveal melanomas with proton beam therapy [26]. about 70–75 % [4, 30, 31].
The minimal latent interval of 5 years implies that metastatic
death 5 years after local treatment is nearly always due to
pretreatment dissemination. This is the one reason that statis- Treatment Methods Comparison
tically significant prognostic factors for survival in tumors
less than 8 mm in diameter should never be established 5 Brachytherapy represents two surgical interventions: the first
years after local treatment, and the recommended interval of is plaque application and the second replacement. Proton
follow-up and analysis seems to be 8–10 years after the treat- beam therapy is accompanied by pretreatment application of
ment. The minimal interval for the evaluation of tumor target markers to the exposed sclera using tantalum clips.
response and treatment-related toxicity is about 3–5 years. Stereotactic radiotherapy uses the fractionation regimes.
Radiosurgery is combined with the stereotactic frame and
eye fixation. The treatment results in terms of tumor response
Other Aspects of Eye Preservation Methods and local control are similar to those of brachytherapy, ste-
reotactic irradiation, helium ions therapy, proton beam ther-
The psychosocial aspect of conservative treatment can also apy, or enucleation and range between 85 and 97 % [20, 25, 26].
be intensively studied. The psychological problems of some Long-term local tumor control of more than 90 % and an
conservatively treated patients can be caused by frequent incidence of rubeosis of iridis in 34 % of patients have been
follow-up tests, treatment of any reoccurrence, treatment of achieved using proton beam therapy after a median follow-
complications, and the possibility of persistent malignant up of 36 months [25]. A major predictive factor for the devel-
cells in the eye involved. On the other hand, the impact of opment of rubeosis iris and later neovascular glaucoma was
enucleation on vision-dependent daily activities must also be a large tumor size unsuitable for brachytherapy. Stereotactic
discussed. irradiation using a linear accelerator with applied doses of
45–70 Gy delivered in one or three fractions is an effective
method with the following results: tumor height reduction
Trials and Treatment Comparison was achieved in 97 %, secondary enucleation were per-
formed in 13 %, and the incidence of secondary glaucoma
The clinical trials and studies about management of uveal was 20 % also including iris or chamber angle neovascular-
melanomas are still going on. The multicenter Collaborative ization [3, 29]. In a prospective study of helium ion therapy,
Ocular Melanoma Study (COMS) was initiated at 32 clinical the incidence of secondary glaucoma was 29 % after a mean
centers in the USA and in Canada in 1986 and ended in 1998. observation period of 53 months [25].
Participating patients were followed for 5–15 years, depend- Radiosurgery data was reported by Rennie, and in this
ing upon date of enrollment, before all clinical follow-up study, after delivering a single fraction of 35–70 Gy, the
ended in 2003. No difference in survival outcomes and little majority of patients developed significant complications
difference in quality-of-life outcomes were observed after a median follow-up of 24 months [3, 32, 33].
between patients in the iodine-125 brachytherapy arm and Retinopathy was found in 78 % and secondary neovascu-
those in the enucleation arm. Five-year survival was substan- lar glaucoma in 43 % of patients. Our radiosurgery data
tially better than anticipated based on a review of the litera- were mentioned above. Comparison of different treatment
ture when the trial was designed. The multidisciplinary methods is given in Table 62.5 [3, 4, 17–20, 35–38, 48,
COMS Group successfully conducted a randomized trial that 52, 54, 55].
756 Y.M. Li et al.
Table 62.5 Comparison of different treatment methods

Median follow-up Minimal Tumor 5-year local Radiation-related Neovascular Enucleation
Treatment method (months) dose (Gy) response (%) tumor control (%) retinopathy (%) glaucoma (%) rate (%)
Brachytherapy [23] 39–96 70–250 69–96 82–92 22–63 3–11 10–21
Proton beam therapy 40 40–70 F 91–100 87–100 NA 8 9
[35–38]
Helium-ions therapy [64] 96 50–80 F 87–96 80 NA 35 20
Stereotactic radiotherapy 36 45–60 F 94 NA 26 9–20 NA
[27, 29]
Radiosurgery [3, 4, 30–33] 36 35–90 70–90 NA 23–70 18–23 18
Fractionated treatment is indicated in the table by F; NA indicates data not available
Treatment Decision Radiosurgery
Rationale criteria for selecting a treatment modality should The role of stereotactic irradiation and radiosurgery in the
be dependent upon the interrelations between the clinical treatment of primary tumors still remains unclear. Pica et al.
characteristics of tumor and the morbidity associated with reported preliminary experience on 11 children (15 eyes)
the available treatment modalities. Small tumors (less than with macular and/or papillary retinoblastoma treated with
10 mm in height) and localized pre-equatorially can be suc- SRT. The mean age was 19 months (range, 2–111). 7, 6, and
cessfully treated using brachytherapy. Tumors localized 2 eyes were classified as International Classification of
equatorially or posterior to the equator and not higher than Intraocular Retinoblastoma Group B, C, and E, respectively.
10 mm can be irradiated using the Leksell Gamma Knife. The delivered dose of SRT was 50.4 Gy in 28 fractions using
Small melanomas (height 5–10 mm) can be treated effec- a dedicated micromultileaf collimator linear accelerator.
tively using Leksell Gamma Knife and in cases with PTV not The median follow-up was 20 months (range, 13–39). Local
exceeding 1,000 mm3 with less than 7 % risk of neovascular control was achieved in 13 eyes (87 %). The actuarial 1- and
glaucoma. Tumors more than 10 mm in height, especially 2-year local control rates were both 82 %. SRT was well tol-
melanomas of the ciliary body, progression of growth after erated. Late adverse events were reported in four patients. Of
irradiation, tumors combined with neovascular glaucoma, or the four patients, two had developed focal microangiopathy
vitreous hemorrhage are candidates for enucleation (Case 20 months after SRT; one had developed a transient recur-
Studies 62.1 and 62.2). rence of retinal detachment; and one had developed bilateral
cataracts. No optic neuropathy was observed [34]. Linear
accelerator-based SRT for papillary and/or macular retino-
Retinoblastoma blastoma in children resulted in excellent tumor control rates
with acceptable toxicity. Additional research regarding SRT
Pathophysiology and its intrinsic organ-at-risk sparing capability is justified in
the framework of prospective trials [35].
Retinoblastoma represents the most common primary malig- Šimonová summarized four patients were admitted for
nant tumor of the eye in children. Although this tumor stereotactic irradiation using the Leksell Gamma Knife and
accounts for only 1 % of all malignant tumors in children, it they had vitreous seeding of malignant cells. All patients
has been studied intensively because of its genetic features. were pretreated according to SIOP (Society International of
Pediatric Oncology) protocol and relapsed after standard
treatment. The only other possibility was enucleation. The
Diagnosis and Treatment whole vitreous body was irradiated using the Leksell Gamma
Knife in a single session, collimator diameter 14 mm, the
The majority of cases are diagnosed within the first 2 years median PTV 2,900 mm3, and the median minimum dose was
of life. The treatment strategy depends on clinical staging 15 Gy. To avoid growth retardation of the irradiated eye and
(disease confined to the retina, disease confined to the globe, bony structures, the minimal dose did not exceed 15 Gy in all
extraocular extension, and presentation of distant metasta- patients, and for this as well as the risk of complications
ses) and includes the following methods: tumor resection, from the previous treatment, the applied doses were rela-
enucleation, chemotherapy, and external beam radiotherapy, tively low and three of the four patients were enucleated
which continues to play an important role in the treatment of with a median of 2 months after radiosurgery for local
extrascleral and intraneural extension of retinoblastoma. progression.
Orbital and Ocular Metastases
Pathophysiology
Metastatic lesions in every subsite of the eye and the orbit

have been described several times but predominately affect
the posterior uveal tract [36, 37]. The autopsy studies esti-
mate the incidence of these metastases to be 4–12 % in
adult patients with cancer of all histologic types. Mostly Fig. 62.13 (a, b) Pretreatment MRI of a spinocellular cell carcinoma
derived from breast cancer in women and from lung cancer case (patient treated by Leksell Gamma Knife, minimum applied dose
in men, other primary sites include the prostate, colon, kid- was 25 Gy on a 50 % isodose curve) (a) in axial and (b) in coronal plane
ney, thyroid, and skin. Most ocular metastases occur
towards the end of life, are asymptomatic, undiagnosed,
and are left untreated. The overall incidence of ocular
metastases is difficult to quantify, as many patients remain
asymptomatic. However, with improvements in systemic
treatment, patients may be living longer and may, therefore,
manifest intraocular and intraorbital lesions more fre-
quently than in the past.
Diagnosis
A careful history and ophthalmologic examination is gener- Fig. 62.14 Control MRI 2 years after radiosurgery documents tumor
ally sufficient to diagnose ocular metastases, and the com- regression
plete examination should include slit-lamp examination,
funduscopy, ultrasonography, and MRI postcontrast
T1-weighted sequence. When the patient has no prior history Stereotactic Radiotherapy
of malignant tumors, a systemic examination should be per- This treatment modality applies higher radiation doses using
formed to identify the primary tumor and metastatic extent a fractionation scheme but, at present, there is still a lack of
outside the eye. Choroidal metastases generally appear as a information about its use in clinical practice.
creamy yellow subretinal mass, often accompanied by reti-
nal detachment. Ultrasonographic findings can usually Radiosurgery
clearly differentiate choroidal metastases from choroidal Radiosurgery may extend treatment options for uveal
melanomas. metastases because of its specific advantages: (1) the pos-
sibility of applying a high radiation dose and (2) evidence
of its sparing effect on the surrounding critical structures.
Treatment The recommended minimal doses are similar to those for
brain metastases, ranging between 20 and 30 Gy, and depend
Brachytherapy on the tumor volume, location, and histology (Figs. 62.13
Plaque therapy is a standard treatment method using the and 62.14).
same radionuclides as for uveal melanomas. Candidates for Additional study of the efficacy and complications of this
this type of irradiation are patients with a solitary lesion and technique compared with external beam radiotherapy is indi-
a mean diameter of the pathologic lesion not exceeding cated as the role of this method still remains open.
8–10 mm.
Radiotherapy Prognosis
External beam fractionated radiotherapy represents the local
effective treatment with a low rate of late complications. The diagnosis of ocular metastases generally carries a poor
The recommended total dose is 30–40 Gy with 2–3 Gy per prognosis with a median survival period of 9 months, with
fraction per day [38]. the literature reporting a range of 4–13 months.
758 Y.M. Li et al.
For adult orbital and intraocular vascular tumors, including

Ocular and Orbital Malignant Lymphomas retinal capillary hemangioma, choroidal hemangioma, radio-
and Leukemia Infiltrates therapy is particularly effective and typically given to older
patients for localized disease. Due to the limited penetration
This special group is represented by orbital malignant lym- and low dose required to treat many of these tumors, elec-
phomas and leukemia pathologic infiltrates. These types of tron-based EBRT can be employed. Kim reported gamma
malignancies are typically radiosensitive and chemosensi- knife radiosurgery for seven symptomatic patients with cho-
tive. The tumor lethal doses ranged from 30 to 40 Gy, with roidal hemangioma (circumscribed form in three patients
2 Gy per fraction per day, and external beam radiotherapy is and diffuse form in four). All patients presented with exuda-
highly effective with a low incidence of radiation-related tive retinal detachment (RD) involving the macula that
toxicity caused by the fact that the effective tumor dose is resulted in severe visual deterioration. The prescription dose
lower than the tolerance doses for critical structures [39]. At to the target margin was 10 Gy in all cases. The mean tumor
present, there is no real indication for using radiosurgery in volume was 536 mm3 (range, 151–1,057) and mean follow-
these types of radiosensitive malignancies, and the question up of 34.4 months (range, 9–76). The resolution of exudative
of its use in persistent or recurring pathologic lesions after RD was achieved within 6 months, and the visual acuity of
previous radiotherapy and chemotherapy is still open. the affected eye had improved at the latest follow-up exami-
nation (p = 0.018) in all patients. No recurrence of exudative
RD occurred. Thinning of the hemangioma was observed in
Vascular Lesions most patients. Symptomatic radiation toxicity had not devel-
oped in any of the patients [42].
Cavernous hemangioma of the orbit typically occurs within Infantile hemangiomas can occur in the eyelids and orbit
the muscle cone and can cause proptosis, optic nerve compres- and cover the visual axis, deviate the eye, and induce ambly-
sion, and corneal exposure. Asymptomatic orbital cavernous opia. When posterior to the eye, these tumors can also pro-
hemangiomas can be followed by serial ophthalmic examina- duce proptosis, corneal exposure, and optic nerve
tion and radiographic imaging. When symptomatic, most compression. Though capillary hemangiomas are very radio-
orbital cavernous hemangiomas can be removed by surgical sensitive, mean dose 4.5 Gy (5–7.5 Gy), those not causing
excision. However, when cavernous hemangioma extends into optic nerve compression, amblyopia, or strabismus should
the orbital apex, or cavernous sinus, superior orbital fissure, is be watched for spontaneous regression over 3–4 years or
within bone, or is nonresectable, these tumors pose diagnostic treated with intralesional or oral steroids, interferon, vincris-
as well as therapeutic challenges for removal. tine, and propranolol. Due to concerns about secondary car-
Khan and Lunsford reported eight symptomatic patients cinogenesis and long-term effects of ocular irradiation,
with cavernous sinus or orbital hemangiomas underwent radiotherapy for infantile hemangiomas should be consid-
SRS. The median target volume was 6.8 mL (range, 2.5– ered a last resort. It could be used as a potentially life-saving
18 mL). The median prescription dose delivered to the mar- method when airway obstruction is involved or when other
gin was 14.5 Gy (range, 12.5–19 Gy). The dose to the optic treatments are not possible or have failed [43].
nerve in all patients was less than 9 Gy (range, 4.5–9 Gy).
The median follow-up period after SRS was 80 months (40–
127 months). Six patients had symptomatic improvement; Age-Related Macular Degeneration
two patients reported persistent diplopia. Follow-up imaging
revealed tumor regression in seven patients and no change in Pathophysiology
tumor volume in one patient. All the patients improved after
SRS [40]. Rootman et al. reported on the use of fractionated Age-related macular degeneration (ARMD) is characterized
stereotactic radiotherapy (SFRT) for the treatment of surgi- by degeneration in the retinal layers accompanied by forma-
cally complicated cavernous malformations. A total dose of tion of subfoveal neovascular membrane that compromises
4,000 cGy divided into 20,200 cGy fractions was applied for vision. The growth factors such as vascular endothelial
three cases, while two other cases were treated with total growth factor plays a fundamental role in age-related macu-
doses of 4,563 and 4,959 cGy divided into 28 × 162 cGy and lar degeneration, and these growth factors are produced by
29 × 171 cGy fractions, respectively. Rapid resolution of inflammatory cells and hypoxic or ischemic retinal tissues,
visual field defect was noted by 3 months, and overall tumor including the neurosensory layer and the retinal pigmented
shrinkage was on average 60 % (range, 32–79). Follow-up epithelium. This pathologic process induces scarring and
was on average 23.4 months (range, 5–50). No complica- destroys the macular retina, causing severe visual changes
tions of treatment were noted [41]. and visual loss in people over 50 years of age [44].
Diagnosis with classic subfoveal neovascular membrane due to ARMD

[47]. The regression was observed in one of ten patients.
A complete ophthalmologic examination, including fluorescein Henderson et al. reported seven patients with ARMD were
angiography, is needed to diagnose age-related macular treated with gamma knife radiosurgery. The median duration
degeneration. The membrane mass can be diagnosed using of follow-up was 2.2 years. Treatment was delivered in a
postcontrast T1-weighted MRI. single shot of 12 Gy. The majority of patients maintained
stable visual acuity after treatment [48].
In our study, we treated 11 patients with neovascular
Treatment Methods membrane (NM) in ARMD with a dose of 15 Gy at a 50 %
isodose in a single fraction. Both ultrasonography and fluo-
Radiotherapy rescein angiography demonstrated a regression of the neo-
The aim of radiotherapy (as for radiosurgery) seems to be (1) vascular complex or its stabilization in 90 % of patients. An
to control growth-factor formation and (2) to alter the degen- enlargement of the NM was found in one patient.
erative process. The persistence of hypoxic stimuli that The patients have to undergo the same procedure as in
upregulate growth factors may explain why the membrane treatment of uveal melanomas. The planning treatment vol-
regrows and disease recurs. ARMD is a degenerative process ume includes the postcontrast enhanced lesion (neovascular-
with proliferative component, and irradiation can play an ization membrane) on a three-dimensional T1-weighted
anti-proliferative role. The experience from radiotherapy sequence. The effective minimal dose based on personal expe-
showed that neovascular membranes are composed of endo- rience is 15 Gy, generally applied on a 50 % isodose curve,
thelial cells with proliferation more rapidly than the endothe- and the use of 8- or 4-mm collimators is recommended to
lial cells of the retina and may be more sensitive than the decrease the doses to the critical structures, especially to the
retinal vasculature [44, 45]. Radiotherapy uses the fraction- disk of the optic nerve. The adequate dose to the optic nerve in
ation scheme and recommended doses ranged from 14 to patients with useful visual function has not to exceed 10 Gy.
20 Gy in 8–10 daily fractions. The experience of radiotherapy Eye structures other than the retina received doses lower than
shows that this modality can affect active subretinal neovas- 10 Gy. Figure 62.15 documents the neovascular membrane on
cularization but is unlikely to prevent new neovascularization postcontrast T1-weighted MRI and 15 Gy and 10 Gy isodose
lesions being produced by this chronic disease. curves. Regression of neovascular membrane after radiosur-
gery (typically late effect) is observed after median time rang-
Proton Beam Therapy ing from 9 to 24 months after radiosurgery (Fig. 62.16),
The proton beam therapy was reported as local effective pretreatment angiography, and ultrasonography and Fig. 62.17
treatment, and the recommended minimal dose is 14 Gy in a documents regression 12 months after radiosurgery). The
single session [45]. application of a stereotactically guided single-dose irradiation
is another promising treatment modality.
Photocoagulation
The management of choroidal subuveal neovascularization
is still controversial and difficult. Laser photocoagulation, Follow-Up
which is indicated for patients with uveal membrane that
adjoins but does not involve the macula, can control the Careful follow-up is necessary to integrate into the treat-
evolving membrane. The strict criteria of the Macular ment, and the intervals at which this takes place depends on
Photocoagulation Study for patient selection shows that only estimates as to when local effect—responses and recurrences
10 % of patients are eligible for this type of treatment [46]. of diseases or radiation-related toxicity—might occur. A rea-
sonably careful follow-up schedule might be every 3 months
Other Methods for the first year after treatment and then every 6 or 12
The pilot studies of surgical excision of membrane and the months for up to 5 years. Radiosurgery could play a signifi-
use of anti-angiogenic agents to prevent new vessel forma- cant role in very advanced forms of neovascular membranes,
tions have not been successful. which are not treatable using other procedures. On this basis,
we have constructed a new prospective, nonrandomized
Radiosurgery study, which is now under way.
Initial results from patients treated by Leksell Gamma Knife The definitive analysis of results and complications in
for ARMD showed that radiosurgery is able to affect tissues age-related macular degeneration after radiosurgery still has
beneath the retina (neovascular membrane in ARMD) with- not been completed, and further investigation is warranted.
out damaging the overlying retinal structures. Haas in his Additional study of the efficacy and complications of this
pilot study had investigated the effect of single-fraction technique in comparison with external beam radiotherapy is
(10 Gy at the 90 % isodose) Gamma Knife surgery in patients indicated, and the role of this method still remains open.
760 Y.M. Li et al.
Fig. 62.15 (a–c) The treatment planning of age-related macular degeneration with 15 and 10 Gy isodose curves
Fig. 62.16 (a, b) Pretreatment angiography and (c) ultrasonography of age-related macular degeneration
Fig. 62.17 (a, b) Control angiography of ARMD and (c) ultrasonography 12 months after radiosurgery
with glaucoma become uni- or bilaterally blind [7, 49].

Glaucoma We currently understand the pathophysiology of glaucoma
to be a progressive loss of ganglion cells resulting in visual
Pathophysiology field damage related to the intraocular pressure. Although
many clinicians now feel that there are several factors
Glaucoma is a chronic, slowly progressive, usually bilateral involved in the pathogenesis of glaucoma, the only rigor-
neuropathy of the optic nerve. Untreated glaucoma eventu- ously proven treatment method is the lowering of intraocular
ally leads to complete loss of vision. About 10 % of patients pressure.
Fig. 62.18 The whole ciliary body irradiated with minimal dose 15 Gy on a 50 % isodose curve shown on (a) axial, (b) coronal, and (c) sagittal
images
Treatment Methods was observed in 77 % with secondary glaucoma, and in all

these patients no medication (analgesics) was required. The
Conventional Treatment median period of the analgesic effect was 6 weeks (range,
The conventional antiglaucomatous treatment—local and 2–32 weeks). In secondary glaucoma, the median of intra-
systemic pharmacotherapy, laser and cryotherapy, and inci- ocular pressure (IOP) was 51.3 mmHg (range, 23–68 mmHg)
sional surgery—sometimes fails, and progressive optic nerve but the median of IOP fell to 27 mmHg (range, 8–48 mmHg)
head neuropathy is manifested. Then we face a difficult situ- after radiosurgery. In primary open angle glaucoma, the less
ation of advanced and sometimes painful glaucoma with raised IOP also fell from a median of 25.3 mmHg (range,
major defects in visual functions and, in the final stages of 17–35 mmHg) to 16.1 mmHg (range, 12–28 mmHg) after
the disease, enucleation of the painful blind eye must be con- radiosurgery [7].
sidered. Therefore, a new effective method of treatment has Patients with advanced glaucoma (after standard treat-
to be sought. ment) have been included into a prospective, nonrandomized
trial, and future investigation and analysis of results and
Radiosurgery complications is warranted. The risk of radiation in benign
During radiosurgical treatment of patients with uveal mela- diseases must be balanced against the risk of alternative
nomas localized near the ciliary body, in the past years at our treatments, which are not negligible.
center we observed that the radiation exerted a positive influ-
ence on painful secondary glaucoma with decrease of intra-
ocular pressure [7]. The influence on pain is well-known Side Effects and Complications
from radiotherapy of degenerative and proliferative benign
processes. The mechanism of this effect has still not been The tolerance of normal tissues is in general the limiting fac-
fully explained. The ciliary body, as the source of intraocular tor for the dose that can be given to target volume. Radiation
aqueous production, had been simultaneously irradiated therapy is associated with a broad spectrum of normal-tissue
using four 8-mm collimators with median of minimal dose reactions, and no reporting of outcome of radiotherapy or
15 Gy on a 50 % isodose curve (Fig. 62.18). The dose to radiosurgery is satisfactory without a thorough description of
cornea should not exceed 7–8 Gy, and in some cases the use the treatment-related toxicity.
of plugging can help to decrease the dose to this structure. Radiation toxicity is relatively well-known from external
The ciliary body as a target volume is located near the beam fractionated radiotherapy, and the most common com-
anterior to equator, and the doses for cornea, lens, and ante- plication is primarily due to radiation vasculopathy.
rior chamber are higher than to the optic nerve and retina. Neovascular glaucoma has been described as the most seri-
A typical acute, transient reaction is lacrimation with inci- ous problem after external beam radiotherapy and radiosur-
dence of 30 %. The transient erythema on eyelid was gery [50]. The development of severe radiation-induced
observed after a single dose of 7 Gy and higher. Late-toxicity toxicity for the various eye critical structures is complicated
changes grade 1 or 2 of cornea was observed after Dmax by other factors (such as hypertension, diabetes), the local
10 Gy. The worsening of lens opacity (grades 3 or 4) was effects of pathologic process, the segment of the eye involved,
recorded after the maximum dose to this structure higher the size and location of the tumor, the type of radiation, and
than 10 Gy. The disappearance of pain after radiosurgery the technique used.
762 Y.M. Li et al.
Complications related to radiation therapy can be acute or and 11 % patients were enucleated for this reason during the
chronic. Chronic changes are delayed in onset and may not first 2 years after radiosurgery. The median time to occur-
improve. Acute reactions are rapid in onset and typically rence of secondary neovascular glaucoma was 18 months
reversible. For example, acute radiation blepharoconjuncti- and so a higher incidence of enucleation during a longer
vitis is commonly seen after EBRT. There are also two other follow-up could be expected. In our series, we did not
types of complications defined: any radiation-related mor- observe any significant influence of the minimum dose and
bidity that occurs within the first 90 days after the end of tumor location, but a significantly lower incidence of sec-
treatment is usually regarded as early and all other complica- ondary glaucoma was noticed when the volume planning
tions as late. The clinical course of the two types of reactions target volume (PTV) of the peripheral (prescribed) isodose
is different, early reactions tending to be transient whereas was less than 1,000 mm3 with an incidence of 6.9 %.
late reactions are often irreversible. Biologically, it has been The worsening of useful vision caused by radiation-
established that higher grades of reactions are on average related toxicity during 3 years after radiosurgery was
seen at later times than lower grades and that increased treat- observed in 35 % of patients and was due to tumor progres-
ment toxicity may considerably shorten the latent period sion in 3 %, optic nerve neuropathy in 8 %, retinopathy in
[54, 55]. Many schemas have been devised for recording the 9 %, and neovascular glaucoma in 14 %.
late effects of radiotherapy treatments including the Radiation There exists a difference in the area of severe complica-
Therapy Oncology Group (RTOG) and LENT SOMA scales tions and it is based on the radiation damage of different eye
widely used in clinical practice. The tables for scoring ocular tissues. The most acceptable are complications concerning
toxicity are documented in Table 62.4 [51]. the lens because ophthalmologists can treat them more eas-
The most common late toxicity for all types of irradiation ily. The main problems in the eye bulb are complications of
is retinopathy, cataracts, secondary glaucoma, and optic neu- the iris, cornea, and optic nerve. The optic nerve tolerance is
ropathy [4, 20, 27, 29, 50, 51]. The presence of iris rubeosis well defined from other radiosurgery procedures especially
can be chosen as a marker of severe ocular damage, and from the irradiation of pituitary adenomas, meningiomas,
many authors have reported that secondary neovascular glau- and craniopharyngiomas. The tolerance dose of the optic
coma is the most common cause of the toxicity-related enu- nerve ranged from 8 to 10 Gy and the same doses are well
cleation [50]. Every new treatment method needs detailed tolerated by other critical ocular structures with an incidence
and qualified radiation morbidity scoring. The initial steps of late complications not higher than 5 % [4].
identify the tolerance doses for critical structures and the Late radiation responses occur after latent periods of 3
effective doses for the target volume. In an analysis of late months and many years [4, 50]. An extended follow-up
toxicity (126 patients), we recorded the following results: period is therefore required in order to obtain a reliable esti-
significantly lower toxicity in the optic nerve was observed mate of the incidence of late complications. Care must be
when the maximum dose was less than 10 Gy (incidence of taken when evaluating late effects in a group of patients with
grades 3 or 4 only in 2.4 %), in the cornea when maximum short life expectancy or a short follow-up, who may not be at
dose did not exceed 10 Gy (incidence of toxicity grades 3 or risk long enough to develop any late complications of inter-
4 in 3 %), in the lens when the maximum dose did not exceed est. The majority of side effects are recorded 9–36 months
7 Gy (incidence of toxicity grades 3 or 4 in 7.7 %), and in the after radiosurgery and radiotherapy and, for example, the
iris when the maximum dose did not exceed 15 Gy (inci- latent period between irradiation and development of cata-
dence of grades 3 or 4 late toxicity in 4.6 %). ract is about 4 years, depending on the dose to this structure,
Detection of risk factors can improve treatment results and the shortest observed latency is 6 months.
and decrease the adverse effects, and an appropriate length
of follow-up is needed for patients after radiosurgery (a min-
imum of 2 years, preferably 5 years) because the hazard of Conclusion
late toxicity is increased with the follow-up time. An accept-
ably low incidence of late radiation-related complications for The treatment of ocular tumors and other eye or orbital
cornea, iris, and optic nerve is observed when the maximum lesions can extend conservative therapeutic options for these
dose to these structures does not exceed 10 Gy. The most types of diseases with vision or eye preservation. Further
frequent and severe radiation-related side effect was secondary clinical studies using radiosurgery in ocular disease are nec-
radiation-related retinopathy, secondary neovascular glau- essary to optimize effective doses and to analyze factors
coma, and optic neuropathy. Severe grades 3 or 4 retinopathy influencing the length of survival with a careful evaluation of
was noticed in 15 % of treated patients, radiation-related late the side effects. The main aim of further clinical studies is to
toxicity grades 3 or 4 in lens was observed in 26 %, optic improve local treatment effects while decreasing severe
neuropathy in 9 %, and secondary glaucoma in 18 % of cases, radiation-related complications.
7. Vladyka V, Liscak R, Simonova G, Pilbauer J, Hejdukova I,

Case Study 62.1 Novacek L. Progress in glaucoma treatment research: a nonran-
domized prospective study of 102 patients with advanced refractory
Uveal melanoma in intimate contact with the disk of glaucoma treated by Leksell gamma knife irradiation. J Neurosurg.
optic nerve. Characteristics: The gross tumor volume 2005;102(Suppl):214–9.
(GTV) was 951 mm3, the treatment planning volume 8. Petersch B, Bogner J, Dieckmann K, Potter R, Georg D. Automatic
(50 % isodose curve) was 2,700 mm3, the tumor height real-time surveillance of eye position and gating for stereotactic
radiotherapy of uveal melanoma. Med Phys. 2004;31(12):3521–7.
was 10 mm, and the minimal dose applied on 50 % iso- 9. Zehetmayer M, Menapace R, Kitz K, Ertl A. Suction attachment for
dose curve was 40 Gy, using 1 isocenter (14-mm col- stereotactic radiosurgery of intraocular malignancies. Ophthal-
limator) (Fig. 62.9). The maximum doses to the critical mologica. 1994;208(3):119–21.
structures: optic nerve 40 Gy, cornea 6 Gy, lens 10 Gy, 10. Langmann A, Langmann G, Unlucerci C, Haller E. [Motility disor-
ders in brachytherapy of choroid melanomas with Ru106 applica-
eyelid 6 Gy. The ultrasonography and control MRI 12 tors]. Ophthalmologe. 1995;92(1):76–8.
months after radiosurgery documented tumor regres- 11. Bellmann C, Fuss M, Holz FG, Debus J, Rohrschneider K, Volcker
sion and the development of postirradiation retinopa- HE, et al. Stereotactic radiation therapy for malignant choroidal
thy grade 2. No acute toxicity was observed. tumors: preliminary, short-term results. Ophthalmology. 2000;107(2):
358–65.
Progression-free survival is 24 months after radiosur- 12. Courdi A, Caujolle JP, Grange JD, Diallo-Rosier L, Sahel J, Bacin
gery (Fig. 62.10). F, et al. Results of proton therapy of uveal melanomas treated in
nice. Int J Radiat Oncol Biol Phys. 1999;45(1):5–11.
13. Novotny Jr J, Vymazal J, Novotny J, Tlachacova D, Schmitt M,
Chuda P, et al. Does new magnetic resonance imaging technology
provide better geometrical accuracy during stereotactic imaging? J
Neurosurg. 2005;102(Suppl):8–13.
Case Study 62.2 14. Walton L, Hampshire A, Forster DM, Kemeny AA. Accuracy of
Patient with useful vision and a diagnosis of uveal stereotactic localisation using magnetic resonance imaging: a com-
melanoma. The gross tumor volume was 480 mm3, the parison between two- and three-dimensional studies. Stereotact
tumor height was 6 mm, the planning treatment vol- Funct Neurosurg. 1996;66 Suppl 1:49–56.
15. Lindquist C, Paddick I. The leksell gamma knife perfexion and
ume (57 % isodose curve) was 890 mm3, and the mini- comparisons with its predecessors. Neurosurgery. 2007;61
mal applied dose 50–57 % isodose curve was 40 Gy. (3 Suppl):130–40. discussion 40–1.
The maximal dose to critical structures: optic nerve 16. Georg D, Dieckmann K, Bogner J, Zehetmayer M, Potter R. Impact
of a micromultileaf collimator on stereotactic radiotherapy of uveal
10 Gy, lens 7 Gy, cornea 4 Gy, eyelid 3 Gy. Pretreatment
melanoma. Int J Radiat Oncol Biol Phys. 2003;55(4):881–91.
MRI scan is shown in Fig. 62.11. 17. Singh AD, Topham A. Incidence of uveal melanoma in the United
Control MRI was performed 18 months after radio- States: 1973–1997. Ophthalmology. 2003;110(5):956–61.
surgery and documented the tumor regressions 18. Manschot WA, van Strik R. Uveal melanoma: therapeutic conse-
quences of doubling times and irradiation results; a review. Int
(Fig. 62.12). No acute or late toxicity was recorded and
Ophthalmol. 1992;16(2):91–9.
no change of vision. Progression-free survival is 30 19. Edge SB, Byrd DR, Compton CC. AJCC cancer staging manual.
months. New York, NY: Springer; 2010.
20. Histopathologic characteristics of uveal melanomas in eyes enucle-
ated from the Collaborative Ocular Melanoma Study. COMS report
no. 6. Am J Ophthalmol. 1998;125(6):745–66.
21. Finger PT. Radiation therapy for choroidal melanoma. Surv
Ophthalmol. 1997;42(3):215–32.
References 22. Hawkins BS. Collaborative ocular melanoma study randomized
trial of I-125 brachytherapy. Clin Trials. 2011;8(5):661–73.
1. Jayarao M, Chin LS. Robotics and its applications in stereotactic 23. Nag S, Quivey JM, Earle JD, Followill D, Fontanesi J, Finger PT.
radiosurgery. Neurosurg Focus. 2007;23(6):E6. The American Brachytherapy Society recommendations for
2. Leksell L. Stereotactic radiosurgery. J Neurol Neurosurg Psychiatry. brachytherapy of uveal melanomas. Int J Radiat Oncol Biol Phys.
1983;46(9):797–803. 2003;56(2):544–55.
3. Dinca EB, Yianni J, Rowe J, Radatz MW, Preotiuc-Pietro D, Rundle 24. Gragoudas ES, Lane AM, Regan S, Li W, Judge HE, Munzenrider
P, et al. Survival and complications following Gamma Knife radio- JE, et al. A randomized controlled trial of varying radiation doses in
surgery or enucleation for ocular melanoma: a 20-year experience. the treatment of choroidal melanoma. Arch Ophthalmol. 2000;
Acta Neurochir (Wien). 2012;154(4):605–10. 118(6):773–8.
4. Simonova G, Novotny Jr J, Liscak R, Pilbauer J. Leksell gamma 25. Char DH, Kroll SM, Castro J. Ten-year follow-up of helium ion
knife treatment of uveal melanoma. J Neurosurg. 2002;97(5 Suppl): therapy for uveal melanoma. Am J Ophthalmol. 1998;125(1):
635–9. 81–9.
5. Tokuuye K, Akine Y, Sumi M, Kagami Y, Ikeda H, Kaneko 26. Castro JR, Char DH, Petti PL, Daftari IK, Quivey JM, Singh RP,
A. Fractionated stereotactic radiotherapy for choroidal melanoma. et al. 15 years experience with helium ion radiotherapy for uveal
Radiother Oncol. 1997;43(1):87–91. melanoma. Int J Radiat Oncol Biol Phys. 1997;39(5):989–96.
6. Wowra B, Zausinger S, Drexler C, Kufeld M, Muacevic A, Staehler 27. Dieckmann K, Georg D, Zehetmayer M, Bogner J, Georgopoulos
M, et al. CyberKnife radiosurgery for malignant spinal tumors: M, Potter R. LINAC based stereotactic radiotherapy of uveal mela-
characterization of well-suited patients. Spine (Phila Pa 1976). noma: 4 years clinical experience. Radiother Oncol. 2003;67(2):
2008;33(26):2929–34. 199–206.
764 Y.M. Li et al.
28. Bogner J, Petersch B, Georg D, Dieckmann K, Zehetmayer M, 46. Laser photocoagulation of subfoveal recurrent neovascular lesions
Potter R. A noninvasive eye fixation and computer-aided eye moni- in age-related macular degeneration. Results of a randomized clini-
toring system for linear accelerator-based stereotactic radiotherapy cal trial. Macular Photocoagulation Study Group. Arch Ophthalmol.
of uveal melanoma. Int J Radiat Oncol Biol Phys. 2003;56(4): 1991;109(9):1232–41.
1128–36. 47. Haas A, Papaefthymiou G, Langmann G, Schrottner O, Feigl B,
29. Zehetmayer M, Kitz K, Menapace R, Ertl A, Heinzl H, Ruhswurm Leber KA, et al. Gamma knife treatment of subfoveal, classic neo-
I, et al. Local tumor control and morbidity after one to three frac- vascularization in age-related macular degeneration: a pilot study. J
tions of stereotactic external beam irradiation for uveal melanoma. Neurosurg. 2000;93 Suppl 3:172–6.
Radiother Oncol. 2000;55(2):135–44. 48. Henderson MA, Valluri S, Lo SS, Witt TC, Worth RM, Danis RP,
30. Zehetmayer M, Menapace R, Kitz K, Ertl A, Strenn K, Ruhswurm et al. Gamma knife radiosurgery in the treatment of choroidal neo-
I. Stereotactic irradiation of uveal melanoma with the Leksell vascularization (wet-type macular degeneration). Stereotact Funct
gamma unit. Front Radiat Ther Oncol. 1997;30:47–55. Neurosurg. 2007;85(1):11–7.
31. Marchini G, Gerosa M, Piovan E, Pasoli A, Babighian S, Rigotti M, 49. Chen PP. Risk and risk factors for blindness from glaucoma. Curr
et al. Gamma Knife stereotactic radiosurgery for uveal melanoma: Opin Ophthalmol. 2004;15(2):107–11.
clinical results after 2 years. Stereotact Funct Neurosurg. 1996;66 50. Kim MK, Char DH, Castro JL, Saunders WM, Chen GT, Stone
Suppl 1:208–13. RD. Neovascular glaucoma after helium ion irradiation for uveal
32. Sarici AM, Pazarli H. Gamma-knife-based stereotactic radiosur- melanoma. Ophthalmology. 1986;93(2):189–93.
gery for medium- and large-sized posterior uveal melanoma. 51. LENT SOMA tables. Radiother Oncol. 1995;35(1):17–60.
Graefes Arch Clin Exp Ophthalmol. 2013;251(1):285–94. 52. Ratanatharathorn V, Powers WE, Grimm J, Steverson N, Han I,
33. Rennie I, Forster D, Kemeny A, Walton L, Kunkler I. The use of Ahmad K, et al. Eye metastasis from carcinoma of the breast: diag-
single fraction Leksell stereotactic radiosurgery in the treatment of nosis, radiation treatment and results. Cancer Treat Rev. 1991;18(4):
uveal melanoma. Acta Ophthalmol Scand. 1996;74(6):558–62. 261–76.
34. Pica A, Russo Krauss I, Castellano I, Rossi M, La Cara F, Graziano G, 53. Turbin RE, Thompson CR, Kennerdell JS, Cockerham KP,
et al. Exploring the unfolding mechanism of gamma-glutamyltrans- Kupersmith MJ. A long-term visual outcome comparison in patients
peptidases: the case of the thermophilic enzyme from Geobacillus with optic nerve sheath meningioma managed with observation,
thermodenitrificans. Biochim Biophys Acta. 2012;1824(4):571–7. surgery, radiotherapy, or surgery and radiotherapy. Ophthalmology.
35. Pica A, Moeckli R, Balmer A, Beck-Popovic M, Chollet-Rivier M, 2002;109(5):890–9. discussion 9–900.
Do HP, et al. Preliminary experience in treatment of papillary and 54. Liu JK, Forman S, Hershewe GL, Moorthy CR, Benzil DL. Optic
macular retinoblastoma: evaluation of local control and local com- nerve sheath meningiomas: visual improvement after stereotactic
plications after treatment with linear accelerator-based stereotactic radiotherapy. Neurosurgery. 2002;50(5):950–5. discussion 5–7.
radiotherapy with micromultileaf collimator as second-line or sal- 55. Andrews DW, Faroozan R, Yang BP, Hudes RS, Werner-Wasik M,
vage treatment after chemotherapy. Int J Radiat Oncol Biol Phys. Kim SM, et al. Fractionated stereotactic radiotherapy for the treat-
2011;81(5):1380–6. ment of optic nerve sheath meningiomas: preliminary observations
36. Bloch RS, Gartner S. The incidence of ocular metastatic carcinoma. of 33 optic nerves in 30 patients with historical comparison to
Arch Ophthalmol. 1971;85(6):673–5. observation with or without prior surgery. Neurosurgery.
37. Goldberg RA, Rootman J, Cline RA. Tumors metastatic to the 2002;51(4):890–902. discussion 3–4.
orbit: a changing picture. Surv Ophthalmol. 1990;35(1):1–24. 56. Kortmann RD, Timmermann B, Taylor RE, Scarzello G, Plasswilm
38. Rudoler SB, Corn BW, Shields CL, De Potter P, Hyslop T, Shields L, Paulsen F, et al. Current and future strategies in radiotherapy of
JA, et al. External beam irradiation for choroid metastases: identifi- childhood low-grade glioma of the brain. Part I: treatment modali-
cation of factors predisposing to long-term sequelae. Int J Radiat ties of radiation therapy. Strahlenther Onkol. 2003;179(8):509–20.
Oncol Biol Phys. 1997;38(2):251–6. 57. Rodriguez JM, Deutsch GP. The treatment of periocular basal cell
39. Bhatia S, Paulino AC, Buatti JM, Mayr NA, Wen BC. Curative carcinomas by radiotherapy. Br J Ophthalmol. 1992;76(4):195–7.
radiotherapy for primary orbital lymphoma. Int J Radiat Oncol Biol 58. Bowyer JD, Sullivan TJ, Whitehead KJ, Kelly LE, Allison RW. The
Phys. 2002;54(3):818–23. management of perineural spread of squamous cell carcinoma to
40. Khan AA, Niranjan A, Kano H, Kondziolka D, Flickinger JC, the ocular adnexae. Ophthal Plast Reconstr Surg. 2003;19(4):
Lunsford LD. Stereotactic radiosurgery for cavernous sinus or orbital 275–81.
hemangiomas. Neurosurgery. 2009;65(5):914–8. discussion 8. 59. Ogawa K, Toita T, Kakinohana Y, Adachi G, Kojya S, Itokazu T,
41. Rootman DB, Rootman J, Gregory S, Feldman KA, Ma et al. Postoperative radiotherapy for squamous cell carcinoma of
R. Stereotactic fractionated radiotherapy for cavernous venous mal- the maxillary sinus: analysis of local control and late complica-
formations (hemangioma) of the orbit. Ophthal Plast Reconstr tions. Oncol Rep. 2001;8(2):315–9.
Surg. 2012;28(2):96–102. 60. Pardo FS, Borodic G. Long-term follow-up of patients undergoing
42. Kim YT, Kang SW, Lee JI. Gamma knife radiosurgery for choroidal definitive radiation therapy for sebaceous carcinoma of the ocular
hemangioma. Int J Radiat Oncol Biol Phys. 2011;81(5):1399–404. adnexae. Int J Radiat Oncol Biol Phys. 1996;34(5):1189–90.
43. Haik BG, Jakobiec FA, Ellsworth RM, Jones IS. Capillary heman- 61. Combs SE, Behnisch W, Kulozik AE, Huber PE, Debus J, Schulz-
gioma of the lids and orbit: an analysis of the clinical features and Ertner D. Intensity modulated radiotherapy (IMRT) and fraction-
therapeutic results in 101 cases. Ophthalmology. 1979;86(5): ated stereotactic radiotherapy (FSRT) for children with
760–92. head-and-neck-rhabdomyosarcoma. BMC Cancer. 2007;7:177.
44. Berson AM, Finger PT, Sherr DL, Emery R, Alfieri AA, Bosworth 62. Haik BG, Jereb B, Smith ME, Ellsworth RM, McCormick
JL. Radiotherapy for age-related macular degeneration: prelimi- B. Radiation and chemotherapy of parameningeal rhabdomyosar-
nary results of a potentially new treatment. Int J Radiat Oncol Biol coma involving the orbit. Ophthalmology. 1986;93(8):1001–9.
Phys. 1996;36(4):861–5. 63. Plowman PN, Hungerford JL. Radiotherapy for ocular angiomas.
45. Slater JM, Archambeau JO, Miller DW, Notarus MI, Preston W, Br J Ophthalmol. 1997;81(4):258–9.
Slater JD. The proton treatment center at Loma Linda University 64. Gragoudas ES, Egan KM, Seddon JM, Walsh SM, Munzenrider
Medical Center: rationale for and description of its development. JE. Intraocular recurrence of uveal melanoma after proton beam
Int J Radiat Oncol Biol Phys. 1992;22(2):383–9. irradiation. Ophthalmology. 1992;99(5):760–6.
Ocular and Orbital: Viewpoint—
Surgery 63
Bryant P. Carruth, Robert H. Hill, and Thomas A. Bersani
Introduction: Surgical Considerations The Surgical Approach

for Orbital Lesions
When considering the surgical approach to an orbital process,
Tumors of the orbit are fortunately rare; however, surgical several factors have to be considered including location
intervention including diagnostic biopsy or excision is often (anterior or deep), surgical space occupied, position relative
necessary. A decision to proceed surgically is based on mul- to the optic nerve, and goal of biopsy (incisional or exci-
tiple factors including clinical and radiographic appearance, sional). The anterior–posterior position of an orbital mass is
anatomic location, and potential risk to important ocular and often the most important factor when considering surgical
orbital structures. Advances in radiation therapy have led to approach. Most anterior lesions can be approached by an
improved outcomes in certain radiosensitive lesions; how- anterior orbitotomy via “swinging” lower eyelid or superior
ever, in a majority of these patients, a tissue diagnosis must lid crease incision. Deeper lesions posterior to the equator of
be established first. the globe require more advanced deep surgical approaches
An orbital biopsy can be either incisional or excisional. often with removal and replacement of bone. These deeper
Infiltrative masses frequently require incisional biopsy to approached are often dictated by the lesion’s position relative
establish the exact nature of the disorder after which addi- to the optic nerve. After considering each of these factors, the
tional surgical or medical treatment can be instituted. safest approach is chosen and then consideration is also given
Alternatively, well-defined or cystic masses tend to be benign to incision placement in an attempt to hide scars [1–3].
and are amenable to complete excisional biopsy. Certainly,
there are exceptions to these rules, the most common being
lymphoid masses which are often well circumscribed. Surgical Spaces of the Orbit
Despite this, lymphoid lesions should be addressed with inci-
sional biopsy as both malignant and benign forms are treated The surgical spaces of the orbit include the intraconal space,
with either radiation if focal or chemotherapy if systemic [1]. the extraconal space, the extraocular muscles, the subperios-
teal space, Tenon’s space, and the extraorbital space.
The intraconal space contains the optic nerve and orbital
B.P. Carruth, M.D. (*)
fat. Examples of tumors commonly found in this location
Department of Ophthalmology, SUNY Upstate Medical University, include optic nerve gliomas and optic nerve meningiomas.
3400 Vickery Rd Suite A, Syracuse, NY 13212, USA The extraconal space contains the lacrimal gland, the superior
Albany Medical Center, Albany, NY, USA oblique muscle, and nerves and vessels that course through
e-mail: carruth426@gmail.com the extraconal orbital fat. The lacrimal gland is a common
R.H. Hill, M.D. source of pathological processes and is accessible with ante-
Department of Ophthalmology, SUNY Upstate Medical University, rior orbitotomy through an upper lid crease incision.
3400 Vickery Rd Suite A, Syracuse, NY 13212, USA The subperiosteal space is a potential space between the
e-mail: rhhill27@gmail.com
periorbita and the orbital bones. A hematoma or abscess may
T.A. Bersani, M.D. collect here, but it is not a common place to find a mass.
Department of Ophthalmology and Otolaryngology,
SUNY Upstate Medical University, 3400 Vickery Rd Suite A,
Also rarely involved in orbital pathologic processes is
Syracuse, NY 13212, USA Tenon’s space, which lies between the eye and the fibrous
e-mail: Tombersani1@gmail.com Tenon’s capsule. Finally, the extraorbital space includes all
766 B.P. Carruth et al.
the structures surrounding the orbit including bone, brain, masses often extending into multiple orbital spaces. A defin-
sinuses, skin, and conjunctiva [1, 4]. itive diagnosis is made by incisional biopsy after which a
systemic workup is required to define the extent of the dis-
ease. If the orbit is the only site of metastasis, radiation ther-
Radiation Responsive Processes apy may be appropriate. However, with systemic involvement,
treatment with chemotherapy is most often employed [7].
Thyroid Eye Disease
Thyroid eye disease, the ophthalmic manifestation of Grave’s Uveal Melanoma

disease, is the most common cause of unilateral and bilateral
exophthalmos in adults [5]. This disorder is characterized by The pigmented, vascular layer of the eye, termed the uveal
progressive inflammation particularly of orbital fat and extra- tract, consists of the iris, ciliary body, and choroid. When
ocular muscles and results in lid swelling, proptosis, lid uveal melanocytes undergo malignant transformation and
retraction, and restrictive strabismus. The treatment of thy- proliferation, a melanoma results. Based on the location, this
roid eye disease is dictated by whether the patient is in the may be an iris, ciliary body, or choroidal melanoma.
active or chronic stage of the disease. It is during the active Melanoma is the most common primary intraocular tumor
phase of the disease that the ophthalmologist must monitor with an incidence of 6–7 cases per million, primarily affect-
the patient’s visual function as severe inflammation can cause ing individuals in their 50s and 60s. Risk factors include
optic nerve compression. Many patients require medical light complexion, smoking, oculodermal melanocytosis, and
management during the acute phase, which often consists of genetic predisposition [8, 9].
oral prednisone (80 mg/day). However, in patients where the Mortality data underscore the significance of this disease.
symptoms are progressing rapidly or when restrictive strabis- Five-year melanoma-related mortality after treatment is
mus predominates, external beam radiation therapy is consid- 10 % for small tumors, 30 % for medium, and 50 % for large
ered with a typical dose of 2,000 rad [6]. In the most drastic tumors. In addition to tumor size, risk factors for mortality
cases of active inflammation in thyroid eye disease, optic include: anterior location, extraocular extension, old age,
neuropathy can be severe and surgical intervention in the tumor regrowth and juxtapapillary location. Histologic and
form of orbital bone and fat decompression may be required. molecular factors include epithelioid cells, high mitotic
index, monosomy 3, trisomy 8, and some newly identified
gene expression profiles [10–13].
Lymphoid Lesions In the distant past, enucleation was the only accepted treat-
ment for uveal melanoma. As eye-sparing treatment modali-
Orbital lymphoproliferative disease usually presents as a ties were developed, enucleation has become less common.
slowly progressive, painless mass. These occur more com- Despite these advances, enucleation as primary treatment
monly in the elderly, and about 25 % of patients carry a pre- remains a viable option in some cases [8]. The defining crite-
vious diagnosis of lymphoma. The most common periocular ria are not without controversy, but the best prospective ran-
lymphoma, representing over 50 % of orbital lesions, is the domized trial to date, the Cumulative Ocular Melanoma Study
MALT type. These lesions most commonly present in the (COMS), outlined those situations in which enucleation is the
superior quadrant of the orbit in the extraconal space and will best option. This study concluded that primary enucleation is
often involve the lacrimal gland. appropriate for large melanomas and that no mortality benefit
Lymphoid lesions continue to be poorly understood; how- exists for presurgical external beam radiotherapy [14].
ever, most periocular lymphomas are non-Hodgkin type. Further, secondary enucleation after failure of other therapies
After a tissue diagnosis is made, all patients with lymphoid leading to a blind and painful eye is often required.
hyperplasia or lymphoma should undergo a systemic evalua- There are two primary surgical options for removing an
tion. In the absence of systemic disease, most of these lesions eye: evisceration and enucleation. Evisceration is removal of
can be treated with external beam radiation with high suc- the contents of the eye leaving the sclera and extraocular
cess rates [1, 5]. muscles intact, whereas enucleation removes the entirety of
the eye leaving only the orbital contents. For many cases of
blind and painful eyes requiring removal, the selected method
Metastatic Tumors to the Orbit is left to surgeon and patient preference; however, the pres-
ence of an intraocular tumor such as choroidal melanoma is
Fortunately, metastatic tumors of the orbit are uncommon. an absolute contraindication to evisceration, and enucleation
The most common primary malignancies metastasizing to is the surgical treatment of choice.
the orbit are lung cancer in men and breast carcinoma in The specific steps of enucleation may vary slightly from
women. On CT scan, metastases will appear as infiltrative one surgeon to another, but the principles remain the same.
63 Ocular and Orbital: Viewpoint—Surgery 767
The conjunctiva and tenon’s capsule are separated from the adjacent structures [20, 21]. Transpupillary thermotherapy
underlying sclera, and the extraocular muscles are removed has been reported to reduce tumor volume in small melano-
from the eye. The globe is brought forward so that the optic mas but may have an increased rate of recurrence relative to
nerve can be identified and cut. The remaining intraconal brachytherapy. Of course, external radiation by traditional or
space is then filled with an orbital implant of silicone, stereotactic means, sometimes combined with surgery, has a
hydroxyapatite, porous polyethylene, or another material. The role in the treatment of melanoma and will be discussed in
recti muscles are secured to the implant or donor scleral wrap- detail by our colleagues.
ping and the tenon’s connective tissue and conjunctiva are Lastly, any physician treating uveal melanoma should be
closed over the implant. A conformer is placed in the conjunc- aware of the current recommendations regarding tumor sur-
tival cul-de-sacs to hold their form while healing occurs, and veillance. A significant body of literature exists regarding the
approximately 6 weeks later, an ocular prosthesis is molded, incidence of metastatic disease with reports of 25 % at 5
crafted, and painted. Inconspicuous custom-made prostheses years, 34 % at 10 years, and as high as 50 % at 25 years
are capable of providing an excellent cosmetic result [15]. [14, 22]. Thus, ongoing surveillance is a necessity. The cur-
Advances in treatment of melanoma in the past decades rent recommendation is for screening every 6–12 months by
have been abundant, and thus, enucleation is less commonly physical examination and liver imaging by ultrasound, CT,
performed [16]. Other treatment modalities include surgical MRI, or PET scan with consideration given to liver function
tumor excision, thermotherapy, photodynamic therapy, and tests and chest X-ray [8, 16, 23].
radiotherapy. The most common treatment for uveal mela-
noma is brachytherapy in which a radioactive plaque, most
commonly Iodine 125, is sutured to the sclera overlying the Retinoblastoma
base of the uveal tumor. This is accomplished with ultrasonic
guidance and posterior globe exposure similar to that Retinoblastoma is a malignant tumor representing uncon-
described for enucleation. The plaque is left in place to trolled growth of neurosensory retinal elements and is the
deliver radiation for 3–7 days and then removed. Local tumor most common malignant ocular tumor in children. Ninety
control rates with brachytherapy have been reported as high percent of cases are diagnosed before 3 years of age and
as 95 % [17]. Dose-dependent radiation complications, how- 30–40 % of patients have bilateral tumors [8, 24–27].
ever, particularly retinopathy and neuropathy, become visu- Retinoblastoma rarely presents with visual symptoms
ally significant in up to 50 % of patients especially those with given the young age of those affected. Rather, the most com-
tumors close to the optic nerve or macula [17]. mon presentation is leukocoria followed by signs of intraoc-
Radiotherapy has been used in a specific localized man- ular inflammation, strabismus, hypopyon, anisocoria, and
ner to treat uveal melanoma by means of charged particle, or proptosis. Extraocular extension of retinoblastoma most
proton-beam, radiation. Only a few centers have the capacity commonly occurs directly via the optic nerve, but malignant
to perform such a procedure, but success comparable to cells can also exit vascular channels or directly extend into
brachytherapy has been reported. A combined surgical the orbit. In rare cases, direct intracranial extension or sys-
approach attaching clips to the sclera outlining the tumor is temic metastasis may occur [8, 28, 29].
necessary with high linear-energy homogenous dose radia- The primary goal of retinoblastoma treatment is, of
tion then applied. Local tumor control has been reported at course, preservation of life with preservation of the eye and
98 %. Unfortunately, a higher dose of radiation is also vision secondary. Retinoblastoma confined to the eye has a
applied to anterior ocular structures. The resulting complica- survival rate of greater than 95 %, but the number drops to
tions of neovascular glaucoma and visual loss have been 50 % with extraocular spread. Globe-sparing treatments for
somewhat prohibitive [8, 18]. retinoblastoma include tumor resection, photocoagulation,
A second arm of the aforementioned COMS trial exam- cryotherapy, brachytherapy, radiation, systemic chemother-
ined mortality in patients with medium-sized tumors under- apy, and most recently, intra-arterial chemotherapy [30, 31].
going brachytherapy vs. enucleation and concluded that While eye-saving treatments are becoming increasingly
survival was equivalent at 82 % in 5 years. Retention of useful particularly in cases of bilateral disease, tumor exci-
visual function in the radiated eyes was somewhat poor with sion by means of enucleation is the definitive treatment.
63 % having visual acuity of 20/200 or worse at 3 years Specifically, enucleation is indicated when a tumor involves
though clearly this is better functioning than those eyes lost greater than 50 % of the eye, has invaded the anterior seg-
to enucleation [17, 19]. ment of the eye or has extended into the optic nerve or orbit.
Globe-sparing tumor excision with or without adjuvant As described above, enucleation is removal of the entire eye
radiation has been successfully reported in cases of small to and contiguous optic nerve with preservation of the extraoc-
medium melanomas. This technique, called partial lamellar ular muscles and other orbital soft tissues. Removal of a long
sclerouvectomy, is surgically complex and suffers from an segment of optic nerve, typically greater than 10 mm, is pref-
inability to monitor margins as well as possible damage to erable to be sure of a negative margin [30–33].
Small, well-circumscribed tumors may sometimes be of supporting cells. Significant visual loss occurs when
treated conservatively with laser photocoagulation, thermal atrophy becomes extensive or when the damage leads to
therapy, or cryotherapy. These techniques are frequently ischemia and neovascularization under the retina. This
augmented by chemotherapy and/or radiation [8]. causes subretinal bleeding, edema, and scarring with sig-
The early onset and rapid growth of retinoblastoma make nificant consequences for vision [9, 41].
this tumor responsive to radiation with globe salvage rates Medical therapy for atrophic ARMD consists of vitamin
reported up to 86 %. The use of EBRT is limited by concerns supplementation to reduce the likelihood of progression to
regarding secondary malignancies, as well as radiation- neovascular ARMD. Antioxidant vitamins have been shown
induced side effects including facial hypoplasia, cataract for- to reduce the risk of progression to neovascular ARMD by
mation, optic neuropathy and retinopathy. The role of EBRT 25 % in patients with intermediate or advanced atrophic
may be linked to chemoreduction and adjuvant use of photo- ARMD [42, 43]. Otherwise, treatment of atrophic macular
coagulation or cryotherapy [34, 35]. degeneration is focused on patient education and regular
The risk of secondary malignancies may be avoided by screening for neovascular disease.
the use of localized radiation with plaque brachytherapy. The Some studies have reported surgical techniques for non-
surgeon attaches a plate containing radioactive material, typ- neovascular ARMD. A technique has been described,
ically iodine 125 or ruthenium 106, onto the sclera overlying whereby choroidal tissue, retinal pigment epithelium, and
the tumor. The plaque is sutured under ultrasound guidance Bruchs membrane are transplanted from the healthy periph-
to ensure accurate placement and then removed 3–7 days ery of the eye to the diseased macula. Despite the theoretical
later. Brachytherapy results in persistent tumor regression in promise, this has been performed only rarely and results are
up to 87 % of cases [36, 37]. thus far underwhelming [44].
When the tumor is small or bilateral, primary or chemore- Treatments for neovascular ARMD may not be consid-
ductive chemotherapy is often used. Tumor shrinkage by ered strictly surgical, but their promise bears noting in this
chemotherapeutic means may allow for subsequent treat- space. Until recently, treatment of ARMD with laser photo-
ment with focal laser, cryotherapy, or radiation. In addition coagulation was the standard of care. Use of an argon laser to
to systemic administration, periocular chemotherapy with thermally coagulate a choroidal neovascular membrane was
subconjunctival carboplatin has proven to be an effective found to be beneficial in stabilizing vision though the num-
adjuvant therapy. With these methods of chemoreduction, ber of patients retaining good visual acuity was still low [45].
control of less advanced cases nears 90 % [38, 39]. Photodynamic therapy (PDT) is the systemic administra-
Administration of low-dose melphalan chemotherapy to the tion of a photosensitizing drug followed by application of
ophthalmic artery under radiologic guidance has recently light to affected tissues inducing reactive oxygen species and
gained favor. The utility of this technique lies in the theoretical leading to destruction of new vessels. PDT resulted in 21 %
ability to cure disease and spare the eye and vision while reduc- of eyes progressing to severe visual loss compared with 48 %
ing systemic side effects and toxicity. Some studies have shown of placebo-controlled eyes [46, 47].
encouraging success, but the jury is still out. Failure and long- While important historically, these modes of treatment
term recurrence with these regimens is not uncommon and sur- have largely been replaced by antiangiogenesis in the form
gical tumor control by enucleation remains important [39, 40]. of anti-vascular endothelial growth factor (VEGF) com-
Vision-sparing and globe-preserving therapies show pounds. VEGF is a growth factor specific to vascular endo-
increasing promise in the treatment of retinoblastoma and thelial cells that can induce angiogenesis, vascular
further advances continue to be made. Absolute tumor con- permeability, and lymphangiogenesis and plays a key role in
trol however remains achievable only through enucleation. the development of choroidal neovascularization. Antibodies
Complicated cases, as well as advanced tumors, are often against VEGF are delivered by intravitreal injection and
best treated with removal of the eye in order to achieve the have revolutionized treatment of neovascular ARMD, not
main goal of any cancer treatment, preservation of life. just slowing visual loss but actually improving visual acuity.
Major trials have demonstrated the effectiveness of ranibi-
zumab, a recombinant antibody fragment directed against
Age-Related Macular Degeneration VEGF. These studies showed an astonishing rate of vision
improvement of 40 % and stabilization of 95 % in the first
Age-related macular degeneration (ARMD) is the most year with 80 % of patients having stable or improved vision
common cause of significant visual loss in the elderly popu- at 24 months. Bevacizumab, a full-length monoclonal anti-
lation, estimated to affect 19.7 % of people over age 75. body against VEGF used off-label for ARMD, has recently
ARMD causes degeneration of the macula, the area of neu- demonstrated equally impressive efficacy [48–51]. Based on
rosensory retina responsible for central visual acuity. This these data, anti-VEGF therapy has become the standard of
occurs by the deposition of waste products and deterioration care for neovascular ARMD.
Surgical therapy for ARMD had been an area of active a chronically blind eye can become physical and cosmeti-
research until the advent of antiangiogenic pharmacologics. cally displeasing. With advances in anophthalmic surgery
Macular translocation was discussed above and had been and ocularistry, these procedures have been made relatively
used in neovascular disease with similar limited efficacy simple, and they give an improved cosmetic result with con-
[44]. Techniques for submacular surgical removal of hemor- sequent improvement in self-esteem and happiness.
rhage and CNV membranes have also been described though In summary, glaucoma is a chronic and progressive optic
a systematic review of the literature reported no benefit [52]. neuropathy that can in some cases be very difficult to treat.
Delivery of anti-VEGF medications carries risks of endo- There are different modalities available for the treatment of
phthalmitis and retinal detachment as well as suffering from refractory glaucoma including gamma knife radiation. We, as
the need for repeated injections. Because of this, methods are ophthalmologists, support cyclophotocoagulation as it histori-
being developed for subretinal drug delivery. These surgical cally has good success rates and avoids the potential risks of
techniques are in their infancy but may herald a new future radiation treatment. For eyes that are blind and painful, evis-
for ARMD treatment. There has also been some interest in ceration and enucleation procedures offer a surgical alterna-
stem cell transplantation in eyes damaged by ARMD though tive that often leads to improved cosmesis and quality of life.
this work remains in the early stages [53].
Low vision aids located outside the eye for patients with
ARMD have been of benefit for many. Recently, however, The Surgeon’s Application of Stereotactic
miniature telescopes and mirrored lenses have been devel- Radiosurgery in the Orbit
oped for intraocular implantation. These fill a small niche at
this time but may become more widely beneficial in the The central tenet of orbital surgery is amelioration of disease
future [54]. with preservation of health and function. This task is made
challenging by the complex anatomy of the orbit which
includes many vital neurologic, vascular, and muscular
Glaucoma structures. Some entities like cavernous hemangioma lend
themselves to careful but complete excision. Others like
Glaucoma is a term used to describe a group of diseases with lymphoproliferative disease require biopsy for tissue diagno-
the common characteristic of optic neuropathy and associated sis but can be well treated with external beam radiation.
visual field loss for which elevated intraocular pressure is a Medicine being the humbling practice it is, however, not
primary risk factor. The World Health Organization (WHO) every pathological process, is so straightforward. Many dis-
estimated from data collected in the 1980s and 1990s that the eases and tumors involve or invest themselves with integral
incidence of primary open angle glaucoma was 2.4 million structures making surgery difficult. Further, many of these
people per year and that the prevalence of blindness for all entities respond only to techniques and dosages of radiation
types of glaucoma was more than 8 million people [55]. that can cause significant morbidity. In these instances, a
The mainstay treatments for glaucoma include local and more nuanced approach involving subtotal surgical excision
systemic pharmacotherapy and incisional surgery. In resis- combined with specific, localized radiation delivered by a
tant types of glaucoma, ciliary body ablation procedures team of orbital surgeons, neurosurgeons, and radiation
including cyclocryotherapy and thermal lasers such as oncologists is emerging as an encouraging alternative to
continuous-wave Nd:YAG, argon, and diode can be used to traditional therapies.
inhibit the secretory activity of the ciliary body epithelium. The location of an orbital tumor determines the level of
There is one report on the use of gamma knife radiation for caution a surgeon must have for two reasons. The first is ease
the treatment of refractory glaucoma showing its ability to of access. Masses located in the anterior orbit, particularly the
decrease pain and intraocular pressure [56]. To the best of lacrimal fossa and lateral and inferior extraconal spaces can
our knowledge, radiation therapy is not routinely utilized for be approached without excessive manipulation of surround-
the treatment of glaucoma, and for reduction of discomfort ing structures. Those in the medial and superior orbits and,
in essentially blind eyes, cyclophotocoagulation has equal if even more importantly, in the posterior orbit and orbital apex
not better results as well as being more cost effective pose a greater challenge. Access to these locations requires
[57, 58]. Cyclophotocoagulation also avoids the potential significant manipulation of orbital structures such as the
secondary risks of radiation. globe, the major vessels, and the optic nerve. Further, some
Finally, in end-stage glaucoma, the eye may become blind lesions involve not just the orbital apex but adjacent struc-
and painful. The authors who support the use of gamma tures like the optic canal, superior orbital fissure, and cavern-
knife radiation for the treatment of glaucoma advocate one ous sinus. Clearly, aggressive surgical exploration in these
of its uses as a means to avoid evisceration or enucleation. locations involves significant risk and is undertaken only
While this certainly is the goal of any ophthalmologist, often when radical control is necessary to preserve life or other
vital functions accepting that ocular motility and/or vision stereotactic radiotherapy may be of benefit. Such a case
may be sacrificed. As an example, Rootman et al. reported was recently presented by Kim et al. in which bilateral meta-
successful treatment of cavernous hemangiomas located at static breast cancer was successfully treated with palliative
the orbital apex with stereotactic fractionated radiotherapy SRT [65].
[59]. This is a typically benign lesion lending itself to com- As discussed above, some tumors simply cannot be com-
plete excision, but when the apex and superior orbital fissures pletely excised from the orbit. Lymphoproliferative disease
are involved, surgical risk increases significantly. is just such an entity. Thankfully, these tumors are highly
The second reason tumor location may give a surgeon radiosensitive using traditional fractionated radiotherapy.
pause relates to the tissue of origin. Some lesions may be These too, however, have been treated with radiosurgery
sacrificed without consequence; however, processes origi- with benefits including lower incidence of dry eyes and cata-
nating from the optic nerve or the meninges do not easily racts owing to the focal delivery of collimated radiation as
lend themselves to physical manipulation. Because the focal well as improved patient comfort with 5-day frameless sur-
nature of radiosurgery may have effects on blood supply to, gery. In addition, lymphoid lesions are frequently ill-defined
and function of, the optic nerve similar to mechanical sur- or irregularly shaped, and therefore the computer-generated
gery, fractionated radiotherapy likely remains the treatment treatment mapping and contouring possible with radiosur-
of choice for optic nerve sheath meningiomas as elegantly gery is particularly beneficial [66].
demonstrated by Bloch et al. Their review found excellent Complications of orbital radiation include dry eyes, cata-
tumor control and good vision preservation with the use of ract formation, retinopathy, optic neuropathy, and diplopia. It
conformal fractionated radiotherapy [60]. is important to keep in mind that these untoward effects of
Further, some tumors located outside the orbital apex may radiation on the optic apparatus, specifically neuropathy and
invest themselves with other integral structures or their retinopathy, are often late sequelae, and therefore the long-
unique pathologic composition may not be amenable to sur- term effects of these treatment modalities remain to be fully
gical excision. In these cases, like a paraganglioma presented elucidated.
by Kim and Lee, radiosurgery may augment intentional For tumors or processes located at the orbital apex and
incomplete excision [61]. those emanating from or involving critical structures, radia-
Because the orbit is contiguous with the CNS via the optic tion and radiosurgery may be the best therapeutic options. In
canal and orbital fissures, many orbital apex tumors share some cases such as sphenoid wing meningiomas, a com-
pathology with intracranial tumors. The application of ste- bined approach can be taken where the surgeon excises as
reotactic radiosurgery on these primary CNS processes is much tumor as can be safely removed and the remainder is
adequately covered elsewhere in this text and the same logic specifically targeted with stereotactic radiation.
applies when the orbit is invaded by these tumors. Kim et al. Intricate anatomy and complex pathology demand a
recently reported on the success of Gamma knife radiosur- unique therapeutic approach. For those patients and diseases
gery for orbital tumors. They reported tumor control in 12 of whose management is not clear-cut or simple, an individual-
15 patients with preservation of vision in 13 of 15 [62]. ized treatment plan involving multiple specialists with surgi-
CyberKnife radiosurgery has also been evaluated in orbital cal, medical, and radiotherapeutic expertise is invaluable.
tumors with Hirschbein et al. reporting encouraging results The progress and development of the field of stereotactic
in 16 patients with good tumor control, vision preservation, radiosurgery has only just begun to potentiate the benefits of
and pain relief [63]. traditional surgery, and we hope that the future is bright for
Just as orbital tumors may invade or affect the central ner- further collaboration.
vous system, CNS tumors can invade the orbit. Even lesions
not located in the orbit can affect the ocular apparatus includ-
ing the optic canal and chiasm. The strongest evidence sup-
References
porting use of stereotactic radiosurgery for lesions abutting
the visual pathways comes from Adler et al. These authors 1. Nerad J. Techniques in ophthalmic plastic surgery. Stanford, CA:
reported remarkable success using CyberKnife multisession Elsevier; 2010. p. 429–32.
radiosurgery to treat perioptic tumors with tumor stability or 2. Nerad J. The diagnostic approach to the patient with proptosis. In The
requisites—oculoplastic surgery. St Louis: Mosby, 2001.p. 348–86.
regression as well as visual field preservation in 94 % of
3. Bonavolonta G. Surgical approaches to the orbit. In: Bosniak S, editor.
patients with mean follow-up of 49 months [64]. Principles and practice of ophthalmic plastic and reconstructive sur-
Another limitation of surgical intervention for orbital gery, vol. II, ch.100. Philadelphia: WB Saunders; 1994. p. 1050–5.
tumors is multifocal or bilateral disease. Specifically, orbital 4. Dutton J. Clinical anatomy of the orbit. In: Yanodd M, Duker J, edi-
tors. Ophthalmology. London: Mosby; 2004. p. 744–51.
metastases may occur in multiple locations and/or bilaterally
5. The American Academy of Ophthalmology. Basic and clinical sci-
and therefore risks of surgery may outweigh the palliative ence course: orbit, eyelids, and lacrimal system, sect. 7. San
benefit of tumor removal. In these cases, palliation with Francisco: The American Academy of Ophthalmology; 2006.
6. Gorman CA, Garrity JA, Fatourechi V, Bahn RS, Petersen IA, et al. 28. Abramson DH, Frank CM, Susman M, Whalen MR, Dunkel IJ,
A prospective randomized double blind, placebo-controlled study Boyd 3rd NW. Presenting signs of retinoblastoma. J Pediatr.
of orbital radiotherapy for Graves’ ophthalmopathy. Ophthalmology. 1998;132:505–8.
2001;108:1523–34. 29. Shields CL, Mashaykekhi A, Demirci H, Meadows AT, Shields
7. Freedman MI, Folk JC. Metastatic tumors to the eye and orbit: JA. Practical approach to the management of retinoblastoma. Arch
patient survival on clinical characteristics. Arch Ophthalmol. Ophthalmol. 2004;122:729–35.
1987;105:1214–19. 30. Abramson DH, Schefler AC. Update on retinoblastoma. Retina.
8. American Academy of Ophthalmology. Ophthalmic pathology and 2004;24:828–48.
intraocular tumors; American Academy of Ophthalmology basic 31. Shields CL, Meadows AT, Leahey AM, Shields JA. Continuing
and clinical science course section 4. Singapore: AAO; 2012. challenges in the management of retinoblastoma with chemother-
9. American Academy of Ophthalmology. Retina and vitreous; apy. Retina. 2004;24:849–62.
American Academy of Ophthalmology basic and clinical science 32. Scheidler V, Dubovy SR, Murray TG. Snare technique for enucle-
course section 12. Singapore: AAO; 2012. ation of eyes with advanced retinoblastoma. Arch Ophthalmol.
10. Shields CL, Kaliki S, Furuta M, Mashayekhi A, Shields JA. Clinical 2007;125:680–3.
spectrum and prognosis of uveal melanoma based on age at presen- 33. Abramson DH, Schefler AC, Almeida D, Folberg R. Optic nerve
tation in 8,033 cases. Retina. 2012;32:1363–72. tissue shrinkage during pathologic processing after enucleation for
11. Harbour JW. Molecular prognostic testing and individualized patient retinoblastoma. Arch Ophthalmol. 2003;121:73–5.
care in uveal melanoma. Am J Ophthalmol. 2009;148:823–9. 34. Hungerford JL, Toma NM, Plowman PN, Doughty D, Kingston
12. Sisley K, Rennie IG, Parsons MA, et al. Abnormalities of chromo- JE. Whole-eye versus lens-sparing megavoltage therapy for retino-
some 3 and 8 in posterior uveal melanoma correlate with prognosis. blastoma. Front Radiat Ther Oncol. 1997;30:81–7.
Genes Chromosomes Cancer. 1997;19:22–8. 35. Murray T. Cancer incidence after retinoblastoma: radiation dose
13. Thomas S, Putter C, Weber S, Bornfeld N, Lohmann DR, Zeschnigk and sarcoma risk. Surv Ophthalmol. 1998;43:288–9.
M. Prognostic significance of chromosome 3 alterations determined 36. Hernandez JC, Brady LW, Shields CL, Shields JA, DePotter
by microsatellite analysis in uveal melanoma: a long-term follow- P. Conservative treatment of retinoblastoma. The use of plaque
up study. Br J Cancer. 2012;106:1171–6. brachytherapy. Am J Clin Oncol. 1993;16:397–401.
14. Hawkins BS, Collaborative Ocular Melanoma Study Group. The 37. Freire JE, DePotter P, Brady LW, Longton WA. Brachytherapy in
Collaborative Ocular Melanoma Study (COMS) randomized trial primary ocular tumors. Semin Surg Oncol. 1997;13:167–76.
of pre-enucleation radiation of large choroidal melanoma IV. Ten- 38. Abramson DH, Frank CM, Dunkel IJ. A phase I/II study of subcon-
year mortality findings and prognostic factors. COMS report no. junctival carboplatin for intraocular retinoblastoma. Ophthalmology.
24. Am J Ophthalmol. 2004;138:936–51. 1999;106:1947–50.
15. Nerad JA. Enucleation, evisceration and exenteration. In: Nerad JA, 39. Shields CL, Shields JA. Intra-arterial chemotherapy for retinoblas-
editor. Techniques in ophthalmic plastic surgery. China: Saunders toma: the beginning of a long journey. Clin Experiment Ophthalmol.
Elsevier; 2010. p. 463–87. 2010;38:638–43.
16. Damato B. Progress in the management of patients with uveal mela- 40. Abramson DH, Dunkel IJ, Brodie SE, Kim JW, Gobin YP. A phase
noma. The 2012 Ashton lecture. Eye (Lond). 2012;26(9):1157–72. I/II study of direct intraarterial (ophthalmic artery) chemotherapy
17. Diner-West M, Earle JD, Fine SL, Hawkins BS, Moy CS, et al. The with melphalan for intraocular retinoblastoma. Ophthalmology.
COMS randomized trial of iodine 125 brachytherapy for choroidal 2008;115:1398–404.
melanoma III: initial mortality findings. COMS report no. 18. Arch 41. Bressler SB, Bressler NM, Gragoudas ES. Age-related macular
Ophthalmol. 2001;119:969–82. degeneration: drusen and geographic atrophy. In: Albert DM,
18. Gragoudas ES, Marie LA. Uveal melanoma: proton beam irradia- Jackobiec FA, editors. Principles and practice of ophthalmology.
tion. Ophthalmol Clin North Am. 2005;18:111–8. 2nd ed. Philadelphia: Sauders; 2000. p. 1982–91.
19. Collaborative Ocular Melanoma Study Group. COMS randomized 42. Age-Related Eye Disease Study Research Group. A randomized,
trial of I-125 brachytherapy for medium choroidal melanoma, I: placebo-controlled, clinical trial of high-dose supplementation with
Visual acuity after 3 years. COMS report no. 16. Ophthalmology. vitamins C and E, beta carotene, and zinc for age related macular
2001;108:348–66. degeneration and vision loss. AREDS report no. 8. Arch Ophthalmol.
20. Damato B, Jones AG. Uveal melanoma: resection techniques. 2001;199:1417–36.
Ophthalmol Clin North Am. 2005;18:119–28. 43. The AREDS2 Research Group. The Age-related Eye Disease Study
21. Ramasubramanian A, Shields CL, Kytasty C, Mahmood Z, Shah 2 (AREDS2): Study design and baseline characteristics (AREDS2
SU, Shields JA. Resection of intraocular tumors (partial lamellar report no. 1). Ophthalmology. 2012;119(11):2282–9.
sclerouvectomy) in the pediatric age group. Ophthalmology. 44. Jousenn AM, Joeres S, Fawzy N, Heussen FM, Llacer H, van Meurs
2012;119(12):2507–13. JC, Kirchof B. Autologous translocation of the choroid and retinal
22. Kujala E, Makitie T, Kivela T. Very long-term prognosis of patients pigment epithelium in patients with geographic atrophy.
with malignant uveal melanoma. Invest Ophthalmol Vis Sci. Ophthalmology. 2007;114:551–60.
2003;44:4651–9. 45. Macular Photocoagulation Study Group. Laser photocoagulation
23. Kaiserman I, Amer R, Pe’er J. Liver function tests in metastatic for juxtafoveal choroidal neovascularization: five year results from
uveal melanoma. Am J Ophthalmol. 2004;137:236–43. randomized clinical trials. Arch Ophthalmol. 1994;112:500–9.
24. Sanders BM, Draper GJ, Kingston JE. Retinoblastoma in Great 46. Treatment of Age-Related Macular Degeneration with
Britain 1969-1980: incidence, treatment, and survival. Br J Photodynamic Therapy (TAP) Study Group. Photodynamic therapy
Ophthalmol. 1988;72:576–83. of subfoveal choroidal neovascularization in age-related macular
25. Scott IU, O’Brien JM, Murray TG. Retinoblastoma: a review degeneration with verteporfin: one year results of 2 randomized
emphasizing genetics and management strategies. Semin clinical trials—TAP report. Arch Ophthalmol. 1999;117:1329–45.
Ophthalmol. 1997;12:59–71. 47. Verteporfin Photodynamic Therapy (VIP) Study Group. Verteporfin
26. Smith BJ, O’Brien JM. The genetics of retinoblastoma and current therapy of subfoveal choroidal neovascularization in age-related
diagnostic testing. J Pediatr Ophthalmol Strabismus. 1996;33:120–3. macular degeneration: two-year results of a randomized clinical
27. Tamboli A, Podgor MJ, Horm JW. The incidence of retinoblastoma trial including lesions with occult with no classic choroidal neovas-
in the United States: 1974 through 1985. Arch Ophthalmol. 1990; cularization—Verteporfin in Photodynamic Therapy Report 2. Am
108:128–32. J Ophthalmol. 2001;131:541–60.
48. Brown DM, Kaiser PK, Michels M, Heler JS, Sy JP, ANCHOR 57. Schlote T, Kreutzer B, Kriegerowski M, et al. Diode laser cyclopho-
Study Group, et al. Ranibizumab vs verteporfin for neovascular tocoagulation in treatment of refractory glaucoma. Klin Monbl
age-related macular degeneration [1 year results of the ANCHOR Augenheilkd. 1997;211(4):250–6.
study]. N Engl J Med. 2006;355:1432–44. 58. Schlote T, Derse M, Rassman K. Efficacy and safety of contact
49. Rosenfeld PJ, Brown DM, Heier JS, Boyer DS, Kaiser PK, et al. transscleral diode laser cyclophotocoagulation for advanced glau-
Ranibizumab for neovascular age-related macular degeneration [2 year coma. J Glaucoma. 2001;10(4):294–301.
results of the MARINA study]. N Engl J Med. 2006;355:1419–31. 59. Rootman DB, Rootman J, Gregory S, Feldman K, Ma R. Stereotactic
50. CATT Research Group. Ranibizumab and bevacizumab for neovas- fractionated radiotherapy for cavernous venous malformations
cular age-related macular degeneration. N Engl J Med. 2011;364: (hemangioma) of the orbit. Ophthal Plast Reconstr Surg. 2012;
1897–908. 28:96–102.
51. CATT Research Group. Ranibizumab and Bevacizumab for treat- 60. Bloch O, Sun M, Kaur G, Barani IJ, Parsa AT. Fractionated radio-
ment of neovascular age-related macular degeneration: two-year therapy for optic nerve sheath meningiomas. J Clin Neurosci.
results. Ophthalmology. 2012;199:1388–98. 2012;19:1210–5.
52. Giansanti F, Eandi CM, Virgili G. Submacular surgery for choroidal 61. Kim CY, Lee SY. Orbtial paraganglioma: Gamma knife surgery as
neovascularization secondary to age-related macular degeneration. a therapeutic option. J Craniofac Surg. 2012;23:1127–8.
Cochrane Database Syst Rev. 2009;15(2), CD006931. 62. Kim MS, Park K, Kim JH, Kim YD, Lee JI. Gamma knife radiosur-
53. Schwartz SD, Hubschman JP, Heilwell G, Franco-Cardenas V, Pan gery for orbital tumors. Clin Neurol Neurosurg. 2008;110:1003–7.
CK, et al. Embryonic stem cell trials for macular degeneration: a 63. Hirschbein MJ, Collins S, Jean WC, Chang SD, Adler Jr
preliminary report. Lancet. 2012;379:713–20. JR. Treatment of intraorbital lesions using the accuracy CyberKnife
54. Singer MA, Amir N, Herro A, Porbandarwalla SS, Polard system. Orbit. 2008;27:97–105.
J. Improving quality of life in patients with end-stage age-related 64. Adler JR, Gibbs IC, Puataweepong P, Chang SD. Visual field pres-
macular degeneration: focus on miniature ocular implants. Clin ervation after multisession CyberKnife radiosurgery for perioptic
Ophthalmol. 2012;6:33–9. lesions. Neurosurgery. 2006;59:244–54.
55. The American Academy of Ophthalmology. Basic clinical science 65. Kim JH, Choi SY, Cho CK, Yang KM, Noh WC, Kim MS. Bilateral
course: glaucoma, sect. 10. San Francisco: The American Academy orbital metastases from breast cancer: a case report of successful
of Ophthalmology; 2006. palliation using stereotactic radiotherapy. Breast J. 2011;17:
56. Vladyka V, Liscak R, Simonova G, et al. Progress in glaucoma 669–71.
treatment research: a nonrandomized prospective study of 102 66. Bianciotto C, Shields CL, Lally SE, Freire J, Shields JA. CyberKnife
patients with advanced refractory glaucoma treated by Leksell radiosurgery for the treatment of intraocular and periocular lym-
gamma knife radiation. J Neurosurg. 2005;102(Suppl):214–9. phoma. Arch Ophthalmol. 2010;128:1561–7.
Ocular and Orbital Tumors: Viewpoint-
Fractionated Radiation 64
Nicolas Girard, Françoise Mornex, and Alain Vighetto
Fractionated radiotherapy is a standard treatment of ocular secondary to optic nerve vasculature damage led to reconsider
and orbital tumors, including optic nerve sheath meningioma this approach. Surgery remains the treatment of choice in
(ONSM). ONSM is a rare benign tumor that derivates from case of a mass effect requiring urgent decompression, or in
the arachnoid tissue. ONSM represents only 1 % of all secondary ONSM with intracranial extent.
meningiomas while accounting for up to 15 % of intraorbital The technical improvements in radiation planning and
tumors [1]. ONSM originating from the intraorbital or the delivery also contributed to this paradigm switch toward
intracanalicular portion of the optic nerve is referred to as radiotherapy. In most reports, radiotherapy has been deliv-
primary ONSM; secondary ONSM originates from the ered using conformal or fractionated stereotactic techniques
medial cerebral fossa and spreads to the optic nerve through- that may control the tumor growth while preserving sensitive
out the optic canal [1]. While the World Health Organization structures including the optic nerve and the retina, to ulti-
distinguishes three histologic grades—grade I, benign; grade mately limit the risks of visual impairment.
II, atypical; and grade III/IV, malignant [2]—and three histo- Here we describe the modalities, the risks, and the results
logic types, meningothelial, fibroblastic, and transitional [2], of definite radiotherapy treatment for ONSM as a model for
the main challenge of the treatment is actually related to the all orbital and ocular tumors, based on evidence available in
anatomic extent of the tumor that predicts the visual out- the literature, including our own experience. We also discuss
come, as patients progressively develop blindness due to the possible ways to optimize radiation delivery through the
optic nerve atrophy. Such evolution has also been reported use of next-generation techniques.
for other intraorbital tumors, such as uveal melanoma or reti-
noblastoma, as discussed in Chap. 63 of this book.
ONSM represents a model of treatment of orbital and Modalities of Radiotherapy
ocular tumors using fractionated radiotherapy, as the chal-
lenge of optic nerve preservation is amplified given the loca- Timeframe
tion of the tumor. Radiotherapy has emerged as the treatment
of choice for ONSM over the last two decades [1]. While The appropriate timeframe to treat ONSM has to be deter-
surgical resection had been seen as a standard in earlier mined taking into account the following points: first, observ-
reports, allowing the control of the intracranial extension of ing a patient with the aim of starting treatment when the
the tumor, as well as a definite pathological diagnosis, the patient ultimately becomes blind is inappropriate; second,
high risks of functional impairment precipitating blindness any treatment will lead to a risk of acute and late side effects
that may precipitate visual impairment; and third, a time
N. Girard, M.D. interval after diagnosis is of importance to assess the tumor
Department of Respiratory Medicine Oncology, Hospices Civils growth, especially in pauci-symptomatic patients. In our
de Lyon, 28 avenue Doyen Lépine, 69677 Lyon, France practice, radiotherapy is proposed after a documented dete-
F. Mornex, M.D., Ph.D. (*) rioration of visual acuity and/or visual field or when symp-
Department of Radiation Oncology, Centre Hospitalier Lyon Sud, toms become incapacitating for patient’s everyday life [3].
Hospices Civils de Lyon, Université Claude Bernard Lyon 1, In our series, the median elapsed time between ONSM diag-
EMR 3738, 69310 Lyon-Pierre Benite, France
e-mail: francoise.mornex@chu-lyon.fr nosis and radiotherapy was 23 months [3].
While no prospective cohort of patients has actually been
A. Vighetto
Neurology Service, Hospital neurologique, Université Lyon 1 et reported, our data, as well as those from other groups, indi-
Hospices Civils de Lyon, 59 boulevard Pinel, 69677 Bron, France cate that patients who experienced vision improvement or
774 N. Girard et al.
stabilization with radiotherapy tended to have been treated (Table 64.1) [3–6, 10, 12–36]. Given the need for sparing of
earlier after diagnostic than patients who had a visual wors- normal tissue through rapid dose fall-off and conformality of
ening [3–6]. Similarly, visual acuity was better at time of the irradiated volume to the lesion, stereotactic conditions
radiotherapy for patients who experienced visual improve- are preferred. As radiation is usually delivered over several
ment [3–6]. A standard recommendation is to start treating weeks (see below), the fixation of a stereotactic head ring, as
ONSM when visual acuity is below 20/40, after documented used in stereotactic radiosurgery, is not possible, as these
deterioration [7, 8]. devices may not be left in place for a period of more than
24–48 h. Several devices that provide a relocatable coordi-
nate system have been developed. Generally, these are based
Volumes on mask and/or bite block systems.
In a series of 34 patients from Amsterdam University, the
ONSM typically grows circumferentially around the optic radiotherapy technique was not a predictor for visual out-
nerve, between the nerve and its blood supply [9]. Treatment come [6, 19]. However, this study relied primarily on visual
planning is usually based on computed tomography scan acuity measured by Snellen eye chart testing to assess visual
with the patient in the treatment on a lying down position. outcomes and did not utilize more sensitive measures for
The gross tumor volume (GTV) is defined as the primary visual acuity or visual fields. Type of radiotherapy did not
tumor mass. Delineation of GTV may benefit from integrat- appear as a predictor of efficacy in other series, provided that
ing magnetic resonance imaging (MRI) data which have 54Gy is delivered to the tumor [17, 20].
become the gold standard for visualizing and diagnosing
intraorbital tumors [9]; gadolinium-enhanced, fat-suppressed
T1-weighted sequences are the best sequences in this setting. Doses, Constraints, and Fractionation Schemes
In our series of 10 patients, MRI was used to define the GTV
that corresponds to the area of contrast enhancement on the For ONSM, the high sensitivity of critical organs to radia-
axial and coronal images, in eight cases. tion, including the optic chiasm and the retina, favors radia-
Additional margins, ranging from 3 to 7 mm, are usually tion delivery using fractionated schemes. As shown in
added to the GTV to define the planning target volume Table 64.1, most studies reported the use of total doses rang-
(PTV). Dose-volume histograms for critical organs, including from 45 to 54 Gy that are delivered using a standard frac-
ing the eye, the retina, the lens, and the chiasm, are used to tionation scheme (1.8–2 Gy fractions, 5 days/week) with
establish treatment planning and beam orientation. In our high-voltage beams. Limiting the maximal dose below 50 Gy
practice, the 100 % isodose line is defined as the isocenter of does not appear to substantially reduce the risk of complica-
the treatment plan, and the dose is prescribed at this point. tions compared to a total dose of 50–54 Gy, and no differ-
Beam-eye-view display (in case of conformal treatment) is ence in risk is observed with fractional doses lower than
used to ensure optimal target volume coverage and normal 2 Gy. Although higher doses have been used treating menin-
tissue sparing. giomas by radiotherapy, there is little chance, based on the
The volume of nerve to be treated is critical. Patterns of excellent results of current protocols, that a better outcome in
failure are always the proximal and not the distal compo- terms of local control would be obtained (Table 64.1).
nent of the nerve [3–6, 10]. In the surgical series, intraocu- Besides total dose and fractionation, dosimetry has to
lar extension was rare, and enucleation was exceptionally integrate mean eye dose and maximal eye dose received,
required [11]. Therefore, significant components of the which are considered the most significant predictors of radi-
retina do not need to be included, reducing the risk of retinal ation efficacy and toxicity.
deficit. Proximally the whole length of the nerve needs to
be treated. In cases of secondary meningioma with intracra-
nial extension, part of the chiasm also needs to be included Results of Radiotherapy
in the PTV. In this setting, the use of stereotactic techniques
facilitates the radiation delivery through easier focusing of Endpoints
the beams.
Assessing the results of radiotherapy for ONSM remains
challenging and has to include both ophthalmologic and
Techniques oncologic endpoints while taking into account the benign
nature of the tumor. Long-term follow-up is mandatory to
While historical studies demonstrated the radiosensitivity of accurately evaluate these endpoints. We recommend that
ONSM using two-dimensional treatment planning, most all patients have a systematic follow-up schedule with at
series have now integrated the use of conformal radiotherapy least a biannual neuro-ophthalmologic examination and an
Table 64.1 Selected series reporting on fractionated radiotherapy for ONSM
Visual outcome Tumor outcome Radiation-induced toxicities
Total Dose per
dose fraction Improved Stable Worse Control Progression Re-progression
Author Year Reference n Technique (Gy) (Gy) (%) (%) (%) (%) (%) (%) Acute (%) Late (%)
Smith et al. 1981 [12] 5 2DRT 53 2 80 20 0 100 0 0 None Dry eye (20 %)
Cataract (20 %)
Kupersmith et al. 1987 [13] 4 2DRT 55 1.8 75 25 0 100 0 0 NR NR
Sarkies et al. 1987 [14] 4 2DRT 40 3.3 0 50 50 100 0 0 NR NR
Kennerdell et al. 1988 [7] 6 2DRT 55 1.8 100 0 0 100 0 0 None None
Becker et al.a 2002 [4] 42 FSRT 54 1.8 31 67 2 100 0 0 Erythema (25 %) None
Alopecia (100 %)
Turbin et al. 2002 [15] 18 3DCRT 40–55 1.4–1.8 44 NR NR 100 0 0 NR NR
Andrews t al. 2002 [16] 33 FSRT 51 1.8 42 50 8 100 0 0 NR Optic neuritis (3 %)
Pitz et al.a 2002 [17] 16 FSRT 54 1.9 40 60 0 100 0 0 Erythema (30 %) Pituitary disorder (7 %)
Alopecia (68 %)
Liu et al. 2002 [18] 5 FSRT 45–54 1.8 80 20 0 100 0 0 None None
64 Ocular and Orbital Tumors: Viewpoint-Fractionated Radiation
Saeed et al. 2003 [19] 6 FSRT 14 14 83 0 17 100 0 0 NR NR

Narayan et al. 2003 [20] 14 3DCRT 45–54 1.8–2.0 36 50 14 100 0 0 Eye pain (7 %) Dry eye (7 %)
Retinopathy (7 %)
Baumert et al. 2004 [21] 23 FSRT 43–45 1.8–2.0 70 22 8 100 0 0 Eye pain (4 %) Retinopathy (7 %)
Richards et al. 2005 [22] 4 FSRT 50–54 1.7 100 0 0 100 0 0 NR NR
Landert et al. 2005 [23] 7 FSRT 52–59 1.7–1.8 86 0 14 100 0 0 NR NR
Sitathanee et al. 2006 [24] 12 FSRT 45 1.8 33 58 9 100 0 0 None None
Litré et al. 2007 [25] 8 FSRT 50 1.8 100 0 0 100 0 0 None Vitreous hemorrhage (13 %)
Arvold et al. 2009 [10] 22 FSRT 50–58 1.8 63 32 5 100 0 1 None Retinopathy (14 %)
Milker-Zabel 2009 [26] 32 FSRT 50 1.8 38 59 3 100 0 0 NR Optic neuritis (3 %)
et al.
Smee et al. 2009 [27] 12 3DCRT 50–54 1.8–2.0 92 0 8 100 0 0 None None
Saeed et al. 2010 [6] 34 3DCRT/ 45–54 1.8 41 50 9 100 0 0 Erythema (47 %) Dry eye (15 %)
FSRT Alopecia (10 %) Cataract (9 %)
Retinopathy (9 %)
Lesser et al. 2010 [6] 11 3DCRT 54 1.8 91 9 100 0 0 NR NR
Liu et al. 2010 [28] 30 SRS 10–17 NR 36 44 20 100 0 0 None None
Kurt et al. 2010 [29] 2 SRS 10–17 3.5–6.5 100 0 0 100 0 0 None None
Metellus et al. 2011 [30] 9 3DCRT 58 1.8 78 22 0 100 0 0 NR NR
(continued)
775
Table 64.1 (continued)
776
Visual outcome Tumor outcome Radiation-induced toxicities

Total Dose per
dose fraction Improved Stable Worse Control Progression Re-progression
Author Year Reference n Technique (Gy) (Gy) (%) (%) (%) (%) (%) (%) Acute (%) Late (%)
Abouaf et al. 2011 [3] 10 2DRT/ 54 1.8–2.0 60 10 30 100 0 0 Conjunctivitis Dry eye (20 %)
3DCRT/ (10 %)
FSRT/ Alopecia (10 %) Cataract (40 %)
IMRT
Retinopathy (20 %)
Adeberg et al. 2011 [31] 19 FSRT 54 1.8 47 48 5 100 0 0 None None
Marchetti et al. 2011 [32] 21 SRS 25 5 35 65 0 100 0 0 None Optic neuritis (5 %)
Pacelli et al. 2011 [33] 5 FSRT 54 1.8 100 0 0 100 0 0 NR NR
Romanelli et al. 2011 [34] 3 SRS 20 5 100 0 0 100 0 0 None None
Solda et al. 2012 [35] 51 FSRT 50 1.5–1.7 32 58 10 None Retinopathy (10 %)
Paulsen et al. 2012 [5] 109 FSRT 54 1.8 13 75 12 96 4 NR Erythema (2 %) None
Eye pain (3 %)
Alopecia (21 %)
Most cases correspond to primary ONSM
2DRT two-dimensional radiotherapy, 3DCRT three-dimensional conformal radiotherapy, IMRT intensity-modulated radiotherapy, FSRT fractionated stereotactic radiotherapy, SRS stereotactic
radiosurgery, NR not reported
a
Overlap among patients
N. Girard et al.
64 Ocular and Orbital Tumors: Viewpoint-Fractionated Radiation 777
annual brain imaging. Visual acuity and visual field testing had improved both visual acuity and visual field, two patients
must be performed. had only improved visual acuity, two patients had only
As reported by other groups, our primary evaluation crite- improved visual field, one patient had stabilized visual field
rion is the visual acuity at the latest available examination but had deteriorated visual acuity, and one patient had dete-
after radiotherapy is compared to the baseline data [3]. When riorated both.
visual acuity is stable, visual fields are a major criterion to With regard to other treatment modalities, Turbin et al.
assess. The gain of more than two lines on Snellen chart reported a series of 64 patients with ONSM treated by vari-
equivalent or the reduction of more than 3 dB of corrected ous modalities [15]; of the 59 patients with vision greater
mean defect (cMD) on static perimetry is considered a sig- than no light perception at diagnosis, 13 patients were
nificant improvement. The loss of two lines or more on observed only, 12 had surgery only including 8 complete
Snellen chart or the increase of 3 dB or more of cMD on resection, 18 received radiation alone, and 16 had surgery—
static perimetry is considered as deterioration. Other cases mostly consisting of a biopsy—and radiation. Visual acuity
are considered as showing a stabilized function. The measurements at diagnosis among the four groups were not
Goldmann perimetry tests may also be used, with improve- different (p = 0.186). Visual acuity fell significantly for the
ment of more than 50 % of the score with V4 object (the observed group only (p = 0.002), surgery only (p = 0.019),
larger and more intense object) considered as an improve- and the surgery with radiation groups (p = 0.030). The
ment. Secondary criteria include the evolution of visual radiation-only group showed a decrease in visual acuity that
symptoms and other clinical symptoms. was not significant (p = 0.301) [15]. In another study, the
visual improvement with radiotherapy was estimated to be as
high as 150 % compared to a historical group of observed
Visual Outcome patients [16].
Our review of the literature indicates that about 60–80 % of

patients receiving radiotherapy for ONSM experience visual Ophthalmologic Examination Outcome
improvement; 20–30 % of patients have a visual stabiliza-
tion, while 10 % experience visual impairment, in most At ophthalmologic examination, the classical signs of
cases because of radiation-induced toxicities (see below; ONSM, referred to as the Walsh triad, include optic atrophy
Table 64.1). Most studies actually reported data only for and optociliary vessel shunting [7, 9]. These signs are not
visual acuity. specific, as these may be found in other disorders—
In our series, median elapsed time from radiation therapy inflammatory, compressive, and ischemic—that lead to optic
to the best visual acuity evaluation was 17.5 months (range, nerve dysfunction. After radiotherapy, shrinkage of optocili-
2–55 months). At best evaluation, median maximal gain of ary shunt vessels and resolution of optic disk swelling and
visual acuity was 0.50 logMAR. We actually defined two proptosis may be immediate [20, 22]. Papilledema disap-
groups of patients: one was composed of patients who expe- pears, and the relative afferent pupillary defect may resolve.
rienced visual acuity improvement or stabilization, and the Patients then do experience rapid symptomatic improve-
other was composed of patients for whom visual acuity ment, including relief of pain or regression of proptosis, che-
worsened at final endpoint. At the latest examination, visual mosis, or diplopia. Furthermore, ipsilateral ocular motility
acuity continued to improve in six cases, remained stable at improves in about 20 % of patients [3, 5].
pretreatment level in one case, and became worse in three
cases compared to the pretreatment evaluation. Visual out-
come was clearly linked with the visual state at time of radio- Oncologic Outcome
therapy: in the first group, the median visual acuity was 1.3
logMAR before radiotherapy (Snellen equivalent, 20/400; In the reported literature, radiotherapy leads to tumor control
range, 20/2,000–20/40). In contrast, patients in the second in virtually all patients (Table 64.1). This actually mostly
group had a median visual acuity of 0.3 logMAR (=20/40; consists of tumor size stabilization, when radiological objec-
range, 20/28–20/50), which was significantly better than in tive response to radiotherapy is observed in 20–30 % of
group 1 (p = 0.037). Regarding visual field, cMD values had cases. In the global series reported by Turbin et al., 21
significantly improved after radiotherapy (p = 0.018). At the patients (33 % of the cohort) experienced radiographic pro-
best evaluation, the median visual field gain (i.e., a decrease gression, including four patients who were observed, seven
of the deficit) was 3.3 dB (median cMD at best = 7.6 dB). At patients who had surgery alone, eight patients who had
the latest examination, visual fields improved in six patients, biopsy and radiation, and only two patients who received
stabilized in two, and worsened in two. Overall, four patients radiation alone [15].
Safety of Radiotherapy Optimization of Radiotherapy
Historically, the major fear with the use of radiotherapy for Given the efficacy of conformal stereotactic radiotherapy,
the treatment of ONSM, at a time when evidence of efficacy with high rates of visual improvement and preservation with
was limited, was the risk of acute and—of more impor- steadily decreasing complications, the role of new radiother-
tance—late toxicity and ultimately acceleration of visual apy techniques is not established. Intensity-modulated radio-
impairment. In irradiating orbit tumors, the main difficulty is therapy (IMRT) consists of a real-time modeling of the
due to the fact that the optic nerve passes through the target contours and the amount of photons delivered within the
itself. Radiation-induced toxicity in patients with ONSM radiation beam, using a programmed movement of the blades
treated with radiotherapy is actually hard to evaluate from of the collimator. This allows the administration of beams of
the existing literature, as (1) these data may not be reported, variable shapes during a single sequence, possibly proving to
(2) acute vs. late toxicities may not be differentiated, (3) be helpful to target tumors that are close to critical tissues.
radiation-induced toxicities on the optic nerve may produce IMRT may be useful to optimize the focalization of the bal-
similar symptoms as the tumor, and (4) follow-up of patient listics, as reported by our group and others [3, 7].
may not be sufficiently long to capture these data. Stereotactic radiosurgery corresponds to the administra-
Historical data reported the occurrence of eye necrosis in tion of high daily doses of radiation (theoretically higher
up to 5 % of cases treated to a total dose of 54 Gy [4]. In our than 3 Gy, but usually ranging from 10 to 17 Gy) in a low
experience, acute toxicity occurred in 40 % of patients but number of fractions (1–4 on average), for a total dose equiva-
was moderate and transient in all cases. In the reported litera- lent to 50–54 Gy using standard fractionation schemes.
ture, late toxicities (more than 2 years after the end of radio- Although this technique may be less toxic as it uses highly
therapy) are observed in 10 % of patients and include focalized radiation beam, it has not been generally used for
retinopathy, optic neuritis, but also alopecia, dry eye, and ONSM because of the high toxicity to the optic nerve when
cataracts. Overall, retinopathy represents about 30 % of late high-dose single fraction radiation is given [36, 38]. As
toxicities and ranges from retinal pigment epithelium atro- shown in Table 64.2, several series of patients with ONSM
phy and minor retinal hemorrhages with macular edema to treated with stereotactic radiosurgery have been reported
extensive formation of epiretinal membranes, polypoidal [28, 29, 32, 34]. Radiosurgery was delivered using various
choroidal vasculopathy, and even peripapillary serous retinal modalities, from one fraction of 10–17 Gy to 5 fractions of
detachment. Photodynamic therapy may be required. With 5 Gy, respectively. Collectively, visual improvement was
regard to neuritis, cranial nerves, optic chiasm, and contralat- observed in 35 % of patients and deterioration in 20 %; these
eral optic nerve within the target are usually spared by ste- rates are higher than that observed after fractionated stereo-
reotactic technique. This is even better accomplished with tactic radiotherapy. Similarly, tumor response rates are
the use of a low fraction size: if the daily fraction size is kept higher, ranging from 40 to 60 %. No serious acute or late side
below 2 Gy with a total dose to the GTV of 50–54 Gy, the effect was observed.
risk of optic nerve injury is estimated to be as low as 3 % at Proton beam radiotherapy delivers protons instead of
10 years [36]. Of note, another organ at risk is the pituitary photons and has been in development since the 1950s.
gland, but a minimum dose of 50 Gy to the anterior part of Protons do not scatter significantly in tissues and allow the
the gland has been described to produce a significant risk of maximum dose to be delivered at a precise depth. Tissues
an endocrinopathy [37]. situated beyond the maximum intensity peak receive no radi-
A major dosimetric predictor of radiation-induced toxicity ation. As for uveal melanoma, proton beam radiotherapy
is the median mean eye dose; in our study, this dose was would then represent an alternative, especially for ONSM
significantly lower in the group of patients experiencing that is the closest to organs at risk, especially the retina [36].
visual improvement after radiotherapy, with 6.1 Gy vs. In a dosimetric study comparing treatment planning for
39.4 Gy in the group of patients experiencing deterioration ONSM using IMRT or proton beam radiotherapy, both tech-
(p = 0.025) [3]. Other radiation characteristic differences did niques achieved a similar optimal dose homogeneity in the
not reach significance, including median total dose and max- target volume; dose-volume histograms favored in every
imal eye dose. Total dose to the GTV has been reported by case of the use of protons due to an inferior low-to-medium
several authors as a predictor of toxicity, mostly in patients dose range in the organs at risk, but the clinical significance
treated before the conformal treatment planning era. of these results remains elusive [36].
Additionally, radiation retinopathy tends to be associated To conclude, fractionated stereotactic radiotherapy is the
with more anterior lesions where higher doses are delivered standard treatment of ONSM, allowing visual improvement
at close proximity to the retina [17, 20]. in 80 % of patients, tumor control in all cases, with limited
Table 64.2 Selected series reporting on stereotactic radiosurgery for ONSM
Radiation-induced
Visual outcome Tumor outcome toxicities
Total dose Dose per
Author Year Reference n (Gy) fraction (Gy) Improved (%) Stable (%) Worse (%) Control (%) Progression (%) Re-progression (%) Acute (%) Late (%)
64 Ocular and Orbital Tumors: Viewpoint-Fractionated Radiation
Liu et al. 2010 [28] 30 10–17 NR 36 44 20 100 0 0 None None

Kurt et al. 2010 [29] 2 10–17 3.5–6.5 100 0 0 100 0 0 None None
Marchetti et al. 2011 [32] 21 25 5 35 65 0 100 0 0 None Optic
neuritis
(5 %)
Romanelli et al. 2011 [34] 3 20 5 100 0 0 100 0 0 None None
Most cases correspond to primary ONSM
NR not reported
779
risk of radiation-induced effects. The mean eye dose has to Simeone FA, Goldman HW, Curran Jr WJ. Fractionated stereotactic
be considered as a limiting constraint in treatment planning, radiotherapy for the treatment of optic nerve sheath meningiomas:
preliminary observations of 33 optic nerves in 30 patients with his-
and it is related to the tumor size and location. Treatment has torical comparison to observation with or without prior surgery.
to be done early when visual impairment occurs. Further Neurosurgery. 2002;51:890–902.
evaluation of the role of new techniques is mandatory. 17. Pitz S, Becker G, Schiefer U, Wilhelm H, Jeremic B, Bamberg M,
Collaborative studies have to be encouraged in the field given Zrenner E. Stereotactic fractionated irradiation of optic nerve
sheath meningioma: a new treatment alternative. Br J Ophthalmol.
the rarity of the tumor. These points may also be considered 2002;86:1265–8.
for other ocular and orbital tumors. 18. Liu JK, Forman S, Hershewe GL, Moorthy CR, Benzil DL. Optic
nerve sheath meningiomas: visual improvement after stereotactic
radiotherapy. Neurosurgery. 2002;50:950–5.
19. Saeed P, Rootman J, Nugent RA, White VA, Mackenzie IR,
References Koornneef L. Optic nerve sheath meningiomas. Ophthalmology.
2003;110:2019–30.
1. Bloch O, Sun M, Kaur G, Barani IJ, Parsa AT. Fractionated radio- 20. Narayan S, Cornblath WT, Sandler HM, Elner V, Hayman
therapy for optic nerve sheath meningiomas. J Clin Neurosci. JA. Preliminary visual outcomes after three-dimensional conformal
2012;19:1210–5. radiation therapy for optic nerve sheath meningioma. Int J Radiat
2. Jain D, Ebrahimi KB, Miller NR, Eberhart CG. Intraorbital menin- Oncol Biol Phys. 2003;56:537–43.
giomas. A pathologic review using current World Health 21. Baumert BG, Villa S, Studer G, Mirimanoff RO, Davis JB, Landau
Organization criteria. Arch Pathol Lab Med. 2010;134:766–70. K, Ducrey N, Arruga J, Lambin P, Pica A. Early improvements in
3. Abouaf L, Girard N, Lefort T, D’hombres A, Tilikete C, Vighetto A, vision after fractionated stereotactic radiotherapy for primary optic
Mornex F. Standard-fractionated radiotherapy for optic nerve nerve sheath meningioma. Radiother Oncol. 2004;72:169–1674.
sheath meningioma: visual outcome is predicted by mean eye dose. 22. Richards JC, Roden D, Harper CS. Management of sight-
Int J Radiat Oncol Biol Phys. 2012;82:1268–77. threatening optic nerve sheath meningioma with fractionated ste-
4. Becker G, Jeremic B, Pitz S, Buchgeister M, Wilhelm H, Schiefer reotactic radiotherapy. Clin Experiment Ophthalmol. 2005;33:
U, Paulsen F, Zrenner E, Bamberg M. Stereotactic fractionated 137–41.
radiotherapy in patients with optic nerve sheath meningioma. Int J 23. Landert M, Baumert BG, Bosch MM, Lütolf UM, Landau K. The
Radiat Oncol Biol Phys. 2002;54:1422–9. visual impact of fractionated stereotactic conformal radiotherapy
5. Paulsen F, Doerr S, Wilhelm H, Becker G, Bamberg M, Classen on seven eyes with optic nerve sheath meningiomas. J Neuro-
J. Fractionated stereotactic radiotherapy in patients with optic nerve ophthalmol. 2005;25:86–91.
sheath meningioma. Int J Radiat Oncol Biol Phys. 2012;82:773–8. 24. Sitathanee C, Dhanachai M, Poonyathalang A, Tuntiyatorn L,
6. Saeed P, Blank L, Selva D, Wolbers JG, Nowak PJ, Geskus RB, Theerapancharoen V. Stereotactic radiation therapy for optic nerve
Weis E, Mourits MP, Rootman J. Primary radiotherapy in progres- sheath meningioma; an experience at Ramathibodi Hospital. J Med
sive optic nerve sheath meningiomas: a long-term follow-up study. Assoc Thai. 2006;89:1665–9.
Br J Ophthalmol. 2010;94:564–8. 25. Litré CF, Noudel R, Colin P, Peruzzi P, Bazin A, Sherpereel B,
7. Kennerdell JS, Maroon JC, Malton M, Warren FA. The manage- Bernard MH, Rousseaux P. Fractionated stereotactic radiotherapy
ment of optic nerve sheath meningiomas. Am J Ophthalmol. for optic nerve sheath meningioma: eight cases. Neurochirurgie.
1988;106:450–7. 2007;53:333–8.
8. Eddleman CS, Liu JK. Optic nerve sheath meningioma: current 26. Milker-Zabel S, Huber P, Schlegel W, Debus J, Zabel-du BA.
diagnosis and treatment. Neurosurg Focus. 2007;23:E4. Fractionated stereotactic radiation therapy in the management of
9. Roy AA, Davagnanam I, Evanson J. Abnormalities of the globe. primary optic nerve sheath meningiomas. J Neurooncol. 2009;94:
Clin Radiol. 2012;67:1011–22. 419–24.
10. Arvold ND, Lessell S, Bussiere M, Beaudette K, Rizzo JF, Loeffler 27. Smee RI, Schneider M, Williams JR. Optic nerve sheath meningio-
JS, Shih HA. Visual outcome and tumor control after conformal mas–nonsurgical treatment. Clin Oncol (R Coll Radiol). 2009;21:
radiotherapy for patients with optic nerve sheath meningioma. Int J 8–13.
Radiat Oncol Biol Phys. 2009;75:1166–72. 28. Liu D, Xu D, Zhang Z, Zhang Y, Li Y, Liu X, Jia Q, Zheng L, Song
11. Alper MG. Management of primary optic nerve meningiomas. G. Long-term results of Gamma Knife surgery for optic nerve
Current status–therapy in controversy. J Clin Neuroophthalmol. sheath meningioma. J Neurosurg. 2010;113:28–33.
1981;1:101–17. 29. Kurt G, Tonge M, Börcek AO, Karahacioglu E, Gurel O, Baykaner K,
12. Smith JL, Vuksanovic MM, Yates BM, Bienfang DC. Radiation ther- Ceviker N, Aykol S. Fractionated gamma knife radiosurgery for optic
apy for primary optic nerve meningiomas. J Clin Neuroophthalmol. nerve tumors: a technical report. Turk Neurosurg. 2010;20:241–6.
1981;1:85–99. 30. Metellus P, Kapoor S, Kharkar S, Batra S, Jackson JF, Kleinberg L,
13. Kupersmith MJ, Warren FA, Newall J, Ransohoff J. Irradiation of Miller NR, Rigamonti D. Fractionated conformal radiotherapy for
meningiomas of the intracranial anterior visual pathway. Ann management of optic nerve sheath meningiomas: long-term out-
Neurol. 1987;21:131–7. comes of tumor control and visual function at a single institution.
14. Sarkies NJ. Optic nerve sheath meningioma: diagnostic features Int J Radiat Oncol Biol Phys. 2011;80:185–92.
and therapeutic alternatives. Eye (Lond). 1987;1:597–602. 31. Adeberg S, Welzel T, Rieken S, Debus J, Combs SE. Prior surgical
15. Turbin RE, Thompson CR, Kennerdell JS, Cockerham KP, intervention and tumor size impact clinical outcome after precision
Kupersmith MJ. A long-term visual outcome comparison in patients radiotherapy for the treatment of optic nerve sheath meningiomas
with optic nerve sheath meningioma managed with observation, (ONSM). Radiat Oncol. 2011;6:117.
surgery, radiotherapy, or surgery and radiotherapy. Ophthalmology. 32. Marchetti M, Bianchi S, Milanesi I, Bergantin A, Bianchi L, Broggi
2002;109:890–9. G, Fariselli L. Multisession radiosurgery for optic nerve sheath
16. Andrews DW, Faroozan R, Yang BP, Hudes RS, Werner-Wasik M, meningiomas: an effective option: preliminary results from a
Kim SM, Sergott RC, Savino PJ, Shields J, Shields C, Downes MB, monoinstitutional experience. Neurosurgery. 2011;69:1116–22.
64 Ocular and Orbital Tumors: Viewpoint-Fractionated Radiation 781
33. Pacelli R, Cella L, Conson M, Tranfa F, Strianese D, Liuzzi R, Solla 36. Miralbell R, Cella L, Weber D, Lomax A. Optimizing radiotherapy
R, Farella A, Salvatore M, Bonavolontà G. Fractionated stereotactic of orbital and paraorbital tumors: intensity-modulated X-ray beams
radiation therapy for orbital optic nerve sheath meningioma—a vs. intensity-modulated proton beams. Int J Radiat Oncol Biol Phys.
single institution experience and a short review of the literature. 2000;47:1111–9.
J Radiat Res. 2011;52:82–7. 37. Kocher M, Treuer H, Hoevels M, Semrau R, Sturm V, Mueller
34. Romanelli P, Wowra B, Muacevic A. Multisession CyberKnife RP. Endocrine and visual function after fractionated stereotactic radio-
radiosurgery for optic nerve sheath meningiomas. Neurosurg Focus. therapy of perioptic tumors. Strahlenther Onkol. 2013;189:137–41.
2007;23:E11. 38. Kwon Y, Bae JS, Kim JM, Lee DH, Kim SY, Ahn JS, Kim JH, Kim
35. Soldà F, Wharram B, Gunapala R, Brada M. Fractionated stereotac- CJ, Kwun BD, Lee JK. Visual changes after gamma knife surgery
tic conformal radiotherapy for optic nerve sheath meningiomas. for optic nerve tumors. Report of three cases. J Neurosurg. 2005;
Clin Oncol (R Coll Radiol). 2012;24:e106–12. 102:143–6.
Part V
Patient Care and Socio-Economic Issues
Complications and Management
in Radiosurgery 65
Igor J. Barani and Minesh P. Mehta
using a variety of fiducial systems that are referenced to a

Introduction component of the treatment device (e.g., head-frame + its
corresponding adapter). SRS is usually performed in a single
The central nervous system (CNS) is exposed to ionizing session, using a rigidly attached stereotactic guiding device,
radiation in a number of clinical situations, predominantly or other mask-based immobilization systems that are often
those involving cancer treatment. Radiotherapy remains a combined with stereotactic image guidance system, but can
major treatment modality for primary and metastatic neo- be performed in a limited number of sessions (≤5). Various
plasms located in the CNS, and exposure of the brain and technologies can be used to perform SRS, including linear
spinal cord is often unavoidable in the radiotherapeutic man- accelerator specifically modified for this purpose, particle
agement of tumors located close to the CNS such as head and beam accelerators, and multisource 60Co units (e.g., Gamma
neck cancers. In addition, there is the increasing application Knife).
of radiation in the eradication of arteriovenous malforma- SRS techniques can safely deliver high doses of radiation
tions (AVMs) and other benign disorders of the brain, such to a precisely defined target by exploiting multiple beamlets,
as trigeminal neuralgia and epilepsy [1, 2]. In each of these each of which does not transport much energy, that converge
situations, the radiation dose that can be administered safely on the target. With this approach, the dose within tumor is
is limited by the potential for injury to the normal CNS tis- much higher than the dose in the surrounding normal brain
sue. Because of the potential severity of CNS damage tissue, as a result of a sharp dose gradient (“dose falloff”)
induced by radiation, the response of normal brain and spinal achieved by the multiple intersecting beams of radiation.
cord to radiation exposure has been a subject of considerable Since normal tissue interface is larger and dose falloff is less
research. As a result, this type of injury has been well rapid for larger lesions, increasing the exposure of the sur-
described in terms of histological and functional criteria as rounding normal brain tissue, the prescribed dose for brain
well as radiobiologic parameters [3–5]. In contrast, the cel- metastases is inversely proportional to the maximum tumor
lular and biochemical processes responsible for the expres- diameter, generally resulting in better control rates of smaller
sion of radiation-induced CNS injury remain poorly defined. lesions. This physical phenomenon largely limits the appli-
Stereotactic radiosurgery (SRS) is a treatment technique cation of SRS to lesions ≤3–4 cm in diameter.
that utilizes very focal ionizing radiation to inactivate or Side effects from SRS are dependent on target location,
eradicate a defined target(s) (AANS/CNS/ASTRO Definition dose of radiation delivered, volume of treatment, and prox-
of Stereotactic Radiosurgery, March 20, 2006). The target is imity to critical neurologic structures [6, 7]. Larger target
defined by high-resolution imaging, either magnetic reso- volumes have also been associated with higher rates of radia-
nance (MR) or computed tomography (CT) and is registered tion toxicity [8]. Comorbidities such as smoking, diabetes
to a stereotactic coordinate space. This space can be defined mellitus, hypertension, certain autoimmune conditions, cer-
tain somatic genetic perturbations, and immunosuppression
increase the risk of complications from conventionally frac-
I.J. Barani, M.D. tionated radiation therapy [9, 10], and they also appear to
Department of Radiation Oncology, University of California,
increase the risks of radiosurgery-related complications [11].
San Francisco, San Francisco, CA, USA
Radiation-induced toxicity can be minimized or prevented
M.P. Mehta, M.B.Ch.B., F.A.S.T.R.O. (*)
by judicious application of SRS principles, techniques, and
Department of Radiation Oncology, University of Maryland,
22 South Greene Street, Baltimore, MD 21201, USA dose–volume planning constraints. This chapter will primarily
e-mail: mineshpmehta@gmail.com; mmehta@ummm.edu focus on radiation necrosis, arguably the most troublesome
786 I.J. Barani and M.P. Mehta
and dreaded complication of SRS, but specific complication radiation necrosis are recognized. In diffuse necrosis,
management related to SRS treatment of brain metastases, widespread periventricular white matter abnormalities are
AVMs, and meningiomas will also be discussed. noted. In the focal pattern, circumscribed lesions are present.
Focal radiation necrosis may be peritumoral or distant, as
well as uni- or multifocal [12].
Definitions of Radiation Toxicity
Based on time of expression, therapeutic radiation-induced Historical Perspective

CNS injury has been classically divided into three reactions:
acute, early delayed and late delayed [3–5]. Acute injury Brain radiation necrosis has been recognized as a potential
(acute radiation encephalopathy) is expressed in days to complication of radiation therapy for at least five decades.
weeks after irradiation and is fairly uncommon under current Some of the earliest experiments demonstrating radiation
radiotherapy protocols. Early-delayed injury commonly necrosis in the CNS were performed in primates and small
occurs from 1 to 6 months and can involve transient demye- animals [13, 14]. In 1950, Lowenberg-Scharenberg and
lination with somnolence and Lhermitte’s syndrome after Bassett described amyloid degeneration of the human brain
brain and spinal irradiation, respectively. Although acute and following x-ray therapy [15]. Subsequently, multiple
early-delayed injuries can result in severe reactions, they are instances of cerebral radiation necrosis were reported [16–
normally reversible and resolve spontaneously. Recently, it 19]. It soon became apparent that this pathology could simu-
has been recognized that changes in delayed memory recall late the disease originally treated with radiation, a dilemma
possibly also represent an early-delayed response, becoming that is still a common clinical problem [20, 21]. In many
manifest within 3–4 months of cranial radiotherapy and cases, pathologic examination of surgical or autopsy speci-
putatively linked to a radiosensitive stem cell compartment. mens was necessary for definitive diagnosis.
The phenomenon of pseudoprogression, best described in Interestingly, MR imaging abnormalities of radiation
the context of concurrent chemoradiotherapy in the manage- necrosis do not always correlate with clinical symptoms such
ment of glioblastoma is yet another example of an early- as cognitive decline, seizures, and weakness. The converse is
delayed reaction, generally occurring within 1–6 months of also true; some patients with symptoms felt to be treatment
completion of therapy, and putatively a consequence of a related may have minimal changes on imaging. A major
dysregulated blood–brain barrier either through endothelial challenge of current neuro-oncology practice is to achieve
damage to the abnormal vasculature surrounding a tumor or the best possible balance between providing the benefits of
through a robust inflammatory/immunogenic response to treatment while minimizing its long-term deleterious effects
tumor cell death. In contrast, late-delayed effects, which on normal brain.
occur at times greater than 6 months after irradiation, are
usually irreversible and progressive. Late-delayed injury is
characterized by demyelination, stem cell depletion, vascu- Dose and Volume Dependence
lar abnormalities, sclerosing phenomena, and ultimately
necrosis that is generally restricted to the white matter. A The radiation tolerance of normal tissues depends on total
consequence of such white matter injury is the clinical/radio- dose, dose per fraction, total time of exposure, volume, radi-
graphic phenomenon of leukoencephalopathy and nonob- ation quality, and adjunctive therapies such as chemotherapy,
structive hydrocephalus. as well as several host parameters. In 1980, Sheline et al. [5]
This classification also helps make the important distinc- published a pioneering study on brain radiation tolerance
tion between radiation injury and radiation necrosis. The based on dose and fraction size. They created a log–log plot
term radiation “necrosis” has been loosely used to refer to of megavoltage dose versus number of treatment fractions
any changes noted on MR imaging that are felt to represent from 80 published cases of brain radionecrosis and deter-
“treatment effect” rather than recurrent tumor. These imag- mined an isoeffect line below which few cases of necrosis
ing changes are noted to reverse partially or completely in were observed (See Fig. 65.1). They then modified the exist-
some patients, raising the possibility that rather than repre- ing equivalent dose formula of the era, the Ellis formula for
senting dead tissue, they represent reversible vascular and nominal standard dose in “ret,” to create a formula for brain
parenchymal abnormalities. Radiation necrosis in this dis- tolerance in “neuret” where
cussion relates primarily to “late-delayed injury” with the
neuret = D ´ N -0.44 ´ T -0.06
morphologic correlates of necrosis. It is this late-delayed
injury that will be addressed below in detail. with D, total radiation dose in rad (cGy); N, number of radia-
Radiation necrosis may be classified further based on the tion fractions; and T, total time in days [5]. Of note, the nega-
pattern of brain involvement. Diffuse and focal patterns of tive exponents indicate that a delivery of radiation dose in
65 Complications and Management in Radiosurgery 787
Fig. 65.1 Megavoltage dose equivalent in rads (cGy) versus number of 0.44. From Sheline GE, Wara WM, Smith V. Therapeutic Irradiation
treatment fractions for cases of brain radionecrosis. These data were and Brain Injury. Int J Radiat Oncol Biol Phys. 1980;6(9):1215–1228;
used to derive the neuret formula. Most cases of radiation necrosis used with permission
occurred for dose/fraction combinations above the line with a slope of
fewer fractions and a shorter time course is more deleterious and β are expressed in units of dose in Gy and the ratio tends
and that fraction size (a function of N) is much more impor- to be in the range of 2–3 for late effects such as lung fibrosis
tant to brain radiation tolerance than is treatment time. The or CNS injury versus around 8–10 for acute-responding tis-
threshold for necrosis was about 1,000–1,100 neuret. sues such as hair follicles, mucosa, and rapidly dividing
The influence of fraction size was also assessed by Lee tumor cells, although a considerable range exists. As an
et al. [22] using a product of total dose (D) and fraction size example, 50 Gy total dose prescription given at 2 Gy per
(d) in Gy2. The product (Dd) was found to be the most sig- daily fraction will have a BED value of 100, assuming α/β = 2
nificant predictive factor for necrosis in a multivariate analy- (a ratio often used for late effects). A biologically equivalent
sis of seven treatment parameters [22, 23]. A study by Ruben dose for late effects given at 3 Gy/fraction can be determined
et al. [24] confirms the significance of Dd and also demon- by solving the formula 100 = D(1 + 3/2), which gives D = 100/
strates that average fraction size ≥2.25 Gy is a significant (1 + 3/2) = 40. Thus, 39 Gy at 3 Gy/fraction is approximately
independent risk factor. Shorter overall treatment time from equivalent to 50 Gy at 2 Gy/fraction in terms of late effects.
giving twice daily radiation treatment was also found to be a As fraction size increases, comparable values for BED for
significant risk factor in the study by Lee et al. [23] either tumor or late effects can be computed. However, data
In the 1970s, a linear quadratic (LQ) model was devel- surrounding the use of this approach with very few fractions
oped to describe cell survival after exposure to ionizing radi- or very large dose per fraction are sparse and unreliable.
ation, where the surviving fraction of cells S after exposure Useful tolerance parameters are the 5 and 50 % probabili-
to a dose D is represented by ties of brain injury at 5 years (TD5/5 and TD50/5, respectively).
At “standard” fractionation of 1.8–2.0 Gy per fraction, the
(
- a D + b D2 )
S =e . TD5/5 is estimated to be 50 ± 10 Gy for the whole brain,
60 ± 10 Gy for a portion of the brain, and 45–50 Gy for a
The constant α describes the initial slope of the survival 10-cm segment of spinal cord [27, 28]. Thresholds of 54 [29]
curve at low doses and β describes the quadratic component and 57.6 Gy[30] have been reported for cerebral radiation
of cell killing [25]. A common related formula first intro- necrosis, but it is apparent that radionecrosis may occur
duced in the early 1980s[26] is for biologically effective below these limits with conventional therapeutic radiation
dose (BED): [5, 31]. In the North Central Cancer Treatment Group low-
grade glioma trial, patients who received 64.8 Gy with
é d ù
BED = D ê1 + , conventionally fractionated radiation had a significantly
ë a / b úû higher actuarial incidence of necrosis (6 patients out of 102)
than those who received 50.4 Gy, and only one patient (out
where D is the total dose and d is the size of a single radiation of 101) treated with 50.4 Gy in 28 fractions developed radio-
fraction in Gy. The ratio of α/β has no units, because both α necrosis [31]. In a report by Marks et al. [29], only 1 patient
out of 139 developed necrosis in an area irradiated to <50 Gy

in 27 fractions. In the study by Sheline et al. [5], only 3 of 80 Incidence
cases of brain radionecrosis were in patients receiving doses
≤50 Gy, in fraction sizes <2.5 Gy. In contrast, 17/20 cases of The incidence of cerebral radiation necrosis after conven-
radionecrosis occurring at ≤50 Gy were associated with tional radiotherapy for primary brain tumors in modern radia-
fraction sizes ≥2.5 Gy. This implies the relative safety of tion oncology practice is poorly defined. This is partially due
doses in 50 Gy range, but only so long as fraction sizes are to limited data on radiation dose-fractionation schedules and
maintained under 2.5 Gy. lack of both the numerator and the denominator for patient
In single-fraction radiosurgery, the risk of brain necrosis populations within which cases of radiation necrosis are
increases rapidly above 12 or 13 Gy, and there is a higher described [5]. Additional barriers relate to the difficulty of
risk with larger nontarget volume encompassed by the distinguishing necrosis from tumor recurrence by neuroimag-
12-Gy isodose line, particularly over 10 mL [32, 33]. ing and the low rates of reoperation and autopsy in these
Although the LQ model and BED concept are presumed to patients [35] and to limited survival time in patients with
“break down” in the context of single fraction radiosurgery, malignant brain tumors. Despite advances in magnetic reso-
it is noteworthy that a single fraction of 13 Gy would yield nance imaging (MRI) and functional imaging, necrosis or
a BED of 97.5, assuming an α/β ratio of 2, very similar to necrosis combined with tumor is still probably frequently
the 100 Gy BED value obtained with 50 Gy in 2 Gy frac- mistaken for tumor progression alone, likely resulting in
tions, with both representing lower, relatively “safe” thresh- underestimation of the true incidence of brain radionecrosis.
olds. A dose-finding protocol, Radiation Therapy Oncology A recent review of 426 patients treated for glioma and
Group 90–05, found a similar relationship between higher followed until death or for at least 3 years identified 21
risk of CNS toxicity from single-fraction radiosurgery and patients with cerebral radionecrosis [24]. Most patients were
larger tumor volume, with lower “maximum tolerated doses” treated with conventional radiotherapy along with
for larger tumors [34]. chemotherapy, but ten of the patients who developed radio-
necrosis had additional boost or salvage stereotactic radia-
tion, usually 18 Gy in three fractions. The crude radionecrosis
Latency incidence was 4.9 % and the actuarial incidence was 2.9, 5.1,
9.3, and 13.3 % at 6, 12, 24, and 36 months post-treatment,
The mean interval from the end of radiotherapy until onset respectively [24]. When the analysis was restricted to patients
of necrosis was 11.6 months in a recent study [24], though treated with a radiation dose biologically equivalent to at
radiation necrosis has been reported to manifest as early as least 45 Gy in 25 fractions, the crude incidence of radione-
3 months [35] and as late as 47 years after radiotherapy crosis increased to 6 %.
[36]. Peritumoral parenchyma seems to be more vulnerable The use of radiosurgery or interstitial radiation after con-
to damage from radiation, and the potential role of chemo- ventional radiotherapy clearly increases the risk of radione-
therapy in enhancing this effect was suggested by the crosis, though reporting rigor, fidelity of follow-up, technique
observation that the latency period for the development of variability and patient population heterogeneity complicate
radionecrosis in glioma patients (typically treated with estimation of the rate of radionecrosis, which ranges from
chemoradiation) appears to be approximately five times 2.6 to 56 % in reported series [29, 31–34, 40–44].
shorter than in other patients receiving equivalent radiation It is important to note that many studies of brain radione-
dose, such as those with nasopharyngeal carcinoma, often crosis include patients treated with chemotherapy in addition
treated with radiotherapy alone [5, 23, 31, 37, 38]. An alter- to radiotherapy. Although some investigators have suggested
native explanation for this comes from the observation that that injury is more often related to radiation than chemother-
higher rates of radiation necrosis from radiosurgery are apy [45], others reported changes similar to “radiation”
observed when treating recurrent malignant gliomas as necrosis from treatment with chemotherapy only for extra-
opposed to brain metastases, and the putative hypothesis is cranial malignancies [46]. It is now widely accepted that
that malignant gliomas are endowed with a significantly both treatment modalities are injurious to normal brain, with
more disrupted and leaky peritumoral vasculature, with the risk of necrosis increasing when they are used together.
endothelium that is more susceptible to damage from large-
fraction radiation. Indirect support for this stems from
studies of aggressively hypofractionated chemoradiotherapy Histopathologic Features
for glioblastoma in which the concomitant use of bevaci-
zumab, known to stabilize the endothelium, significantly The end point of irreversible radiation injury in the brain is
reduced observed rates of necrosis (relative to historic necrosis that may develop months to years after the comple-
expectations) [39]. tion of radiation therapy [3, 16]. Histopathologically, radiation
necrosis occurs most commonly in the white matter around (b) contrast enhancement, (c) progressive enlargement, at
blood vessels. The most frequent histopathologic feature is least over several months, (d) surrounding edema, and (e)
fibrinoid necrosis of blood vessel walls with surrounding exertion of mass effect [52]. On MRI, brain necrosis usually
perivascular parenchymal coagulative necrosis with or with- consists of a somewhat irregular rim-enhancing mass with a
out an inflammatory response (sometimes referred to as central area of low signal on T1-weighted images (Fig. 65.2,
“dirty necrosis”) [47]. Focal radiation necrosis may be peri- upper center panel). The contrast enhancement of these
tumoral or distant, as well as unifocal or multifocal [12]. lesions is thought to be secondary to radiation-induced endo-
Extension and confluence of multiple perivascular foci of thelial damage resulting in breakdown of the blood–brain
necrosis result in large serpentine or confluent zones of barrier. On T2-weighted images, the surrounding edema has
parenchymal necrosis [3, 48]. Over time, dystrophic calcifi- high signal intensity, the solid portion of the radionecrotic
cations can develop in these necrotic areas. An additional lesion has low signal intensity, and the central necrotic com-
vascular lesion that is often observed consists of clusters of ponent may show either increased or decreased signal inten-
abnormally dilated, thin-walled telangiectasias. Late vascu- sity (Fig. 65.2, lower center panel). Later, focal brain atrophy
lar changes include vessel wall thickening caused by hyalin- develops, and sometimes large cysts develop. The arcuate
ization, with resultant luminal narrowing, and ultimately, in fibers in white matter are relatively resistant to radiation
some cases, aneurysmal dilation. White matter changes may necrosis and are usually involved late in the disease process.
also be focal or diffuse; in the diffuse pattern, widespread The predilection for periventricular white matter involvement
periventricular demyelination is seen [16]. in radiation necrosis may, on some occasions, mimic multiple
sclerosis. Multiple lesions can resemble multiple metastases,
and radiation-induced necrotic lesions can also spread subep-
Diagnosis endymally and mimic subependymal tumor spread.
Despite the many commonalities in radiographic appear-
Radiation necrosis should be in the differential diagnosis for ance between tumor and necrotic tissue, several investigators
any patient with progressive neurological symptoms or a have attempted to define patterns of radiation necrosis on
new or enlarging enhancing brain lesion with a history of conventional MRI with the aim of distinguishing tumor from
prior intracranial radiation therapy to a dose of at least 12 Gy necrosis. A recent study of 59 patients treated with radiosur-
in a single dose (Fig. 65.2) or 50 Gy with at 1.8–2.0 Gy per gery for metastatic brain tumors who underwent subsequent
daily fraction. A major dilemma in the diagnosis of radiation craniotomy for symptomatic lesion enlargement analyzed
necrosis is differentiating it from recurrent or persistent the following features: (a) arteriovenous shunting, (b) gyri-
tumor. Tumor recurrence and radiation necrosis often have a form lesion or edema distribution, (c) perilesional edema, (d)
similar appearance on conventional contrast-enhanced MRI cyst formation, and (e) pattern of enhancement [53]. The
[49, 50]. Because of this, it is often difficult to distinguish authors also defined a radiographic feature, the lesion quo-
which entity is responsible for the appearance of an enlarg- tient, which is the ratio of the nodule as seen on T2 imaging
ing lesion. Traditionally, a patient’s clinical course, biopsy, to the total enhancing area on T1 imaging, and suggested that
and serial imaging for several months have been used to try a lesion quotient ≥0.6 correlates with recurrent tumor,
to distinguish between tumor recurrence and radionecrosis whereas a quotient of ≤0.3 correlates with radionecrosis.
[49–51]. Any investigation short of extensive tissue exami- The examined radiographic features, taken together, achieved
nation lacks diagnostic sensitivity and specificity. Anatomic 80 % or greater predictive value but had either low sensitivity
and physiologic MRI, positron emission tomography (PET), or low specificity [53]. In an older study, Kumar et al. [12]
and thallium single photon emission computed tomography cite two radiographic features that are consistent with radia-
(SPECT) imaging have all been used in an attempt to define tion necrosis, the “Swiss cheese” pattern and the “soap bub-
the best noninvasive method to diagnose radiation necrosis, ble” pattern. The first pattern is characterized by diffuse
though there is no evidence to date that any of these investi- areas of enhancement affecting the cortex and white matter
gations is clearly superior to the other modalities. and intermixed with necrotic areas. The soap bubble appear-
ance refers to smaller lesions with heterogeneously enhanc-
ing necrotic cores.
MRI The key point is that radiation necrosis should be included
high in the differential diagnosis if some of the described
Radiation necrosis can closely resemble recurrent tumor on imaging features or patterns are recognized. However, no
anatomic (“conventional”) MRI because of the following radiographic pattern is pathognomonic for radiation necrosis
shared features: (a) focus at or close to the original tumor site, and as such, supplementary functional imaging is often nec-
which is commonly the site of maximum radiation dose, essary and even then, surgical excision may still be required.
Fig. 65.2 T1-weighted, post-contrast MRI (upper panels) and FLAIR after the onset of necrosis, the patient underwent a 2-month long treat-
images (lower panels) for a patient with a breast cancer brain metastasis ment with dexamethasone resulting in significant clinical and radio-
treated with Gamma Knife radiosurgery with a single fraction of 18 Gy, graphic improvement (right panels). Symptoms resolved rapidly after
prescribed to 50 % isodose line (upper, left panel). Symptomatic radia- initiation of dexamethasone therapy and did not recur after discontinu-
tion necrosis developed 10 months later (center panels). Five months ation of steroids
MRS In practice, MRS has been used alone in an attempt to

reduce the need to biopsy new enhancing lesions on MRI
Magnetic resonance spectroscopy (MRS) has been explored [62, 63] and in combination with biopsy to help guide sam-
as a noninvasive means of diagnosing radiation necrosis and pling and improve diagnostic yield [64, 65].
distinguishing it from tumor recurrence. The technique
allows metabolites from an area of interest to be quantified
and compared with those of surrounding normal tissue [54]. DWI
Metabolites such as N-acetyl aspartate (NAA), phosphoryl-
choline/glycerophosphorylcholine (Cho), creatine (Cr), lac- Diffusion-weighted imaging (DWI) is another interesting
tate (Lac), and mobile lipids (Lip) provide different spectra approach for utilizing MR technology to diagnose radiation
in tumors compared with normal brain or radiation necrosis. necrosis. The technique depends on the diffusion of water
Elevated levels of choline compounds have been repeatedly molecules within the tissue under study and generates an
reported in the “peritumoral region” for gliomas [55–59] but apparent diffusion coefficient (ADC) map [54]. Areas with
not metastases [56]. high ADC appear dark on DWI whereas low ADC areas pro-
In general, tumor is characterized by NAA/Cho ratio <1 and duce increased signal. The differences in ADC in various tis-
Lip/Cho ratio <3. In contrast, Lip/Cho ratio >3 or significant sues are thought to result from both changes in the balance
decrease of all metabolite levels within the lesion or surround- between intracellular and extracellular water and changes in
ing brain are considered characteristic of radiation-induced the structure of the two compartments. In theory, areas of
necrosis [60]. The technique has evolved significantly from radiation necrosis should have higher (or better) diffusion
single-voxel, low-resolution studies to multi-voxel acquisitions coefficients than areas of hypercellular tumor so that the
that have a much higher resolution [60, 61]. technique could be used to differentiate the two conditions.
This explanation represents a gross simplification of the or higher indicated tumor recurrence and a ratio of 0.6 or less
biologic milieu of both tumor and necrosis, but high diffu- was suggestive of nonneoplastic contrast-enhancing tissue.
sion on ADC has been reported in cerebral radiation necrosis A more recent approach, described by Barajas et al. [80],
caused by treatment of brain tumors [66–69]. uses three hemodynamic variables to describe the shape of
Previous studies applied this technique to evaluate cellu- the signal intensity–time curve obtained from dynamic sus-
larity in gliomas with the aim of predicting tumor grade [70]. ceptibility contrast pMRI: rCBV, relative peak height, and
Signal intensity and ADCs in neoplastic tissue, peritumoral percentage of signal intensity recovery; among 27 patients
edema, and normal brain tissue have been reported for low- with necrosis or tumor recurrence or progression after single-
and high-grade gliomas, meningiomas, and brain metastases fraction radiosurgery for brain metastases, rCBV and relative
[71]. Necrotic brain tissue in the temporal lobe after radia- peak height were significantly higher and the percentage of
tion therapy for nasopharyngeal carcinoma has been charac- signal intensity recovery was significantly lower for tumor
terized using DWI as well [72]. Previously reported ADC versus necrosis [80, 81]. Relative CBV is the most widely
values for high-grade glial tumors vary from 1.1 to used hemodynamic variable derived from pMRI and has been
1.37 × 10−3 mm/s2 [73–75]. ADCs in peritumoral edema or shown to correlate with tumor grade and tumor microvascular
temporal lobe necrosis were reported as 1.29 and attenuation [56, 79, 82–88]. Relative peak height, which rep-
2.88 × 10−3 mm/s2 in two notable studies [72, 76]. Although resents the maximal change in signal intensity during the pas-
ADC maps can provide useful information that may help dis- sage of contrast agent, has been shown to correlate with rCBV
tinguish tumor from necrosis, DWI has not been studied suf- measurements and tumor capillary blood volume. Finally,
ficiently to permit unequivocal diagnosis. Presently, DWI is percentage of signal intensity recovery, an indicator of the
primarily used as an adjunct to other studies. blood–brain barrier integrity, reflects the degree of contrast
agent leakage through tumor microvasculature and provides
insight into the alteration of capillary permeability.
Perfusion MRI Several important limitations are associated with dynamic,
contrast-enhanced pMRI. Because the technique is suscepti-
Dynamic, contrast-enhanced perfusion MRI (pMRI) of the bility weighted, it is extremely sensitive to structures or
brain provides hemodynamic information that complements lesions that induce strong magnetic field inhomogeneities
the anatomic information available from conventional such as blood products, calcium, melanin, metals, or lesions
MRI. Contrast-enhanced pMRI methods exploit signal near brain–bone–air interfaces such as the skull base. When
changes that accompany the passage of a paramagnetic con- there is a large amount of susceptibility artifact, pMRI fails
trast agent (gadolinium) through the cerebrovascular system to provide any meaningful information [89]. Also, the calcu-
to derive relative blood volume and flow information [77, 78]. lation of rCBV can be grossly inaccurate in lesions with a
Analysis of dynamic data from pMRI yields quantitative severe breakdown or absence of the blood–brain barrier.
estimates of the relative cerebral blood volume (rCBV) of
the lesion versus other regions of interest such as surrounding
tissue or contralateral normal tissue. For CBV measurements, FDG-PET
a series of images is acquired at intervals of approximately
1 s before, during, and after bolus injection of the contrast Positron emission tomography (PET) can be used to measure
agent. Rapid gradient-echo or echo-planar imaging is gener- the metabolism of glucose labeled with an 18 F analog.
ally used to acquire multiple T1-weighted slices at each time Because radiation necrosis should be characterized by low
point during the imaging interval. When a paramagnetic glucose consumption, 18 F-fluorodeoxyglucose (FDG) is
contrast agent passes through the cerebral vascular system, often used in PET studies to provide contrast with areas of
it induces differences in local magnetic susceptibility tumor, which should have higher uptake. Early studies
between the intravascular space and the surrounding normal reported that 18 F-FDG-PET had a sensitivity of 81–86 % and
tissue, with a transient and differential loss of relaxation signal. a specificity of 40–94 % for distinguishing between radiation
This enables calculation of the contrast concentration–time necrosis and tumor [90].
curve, and rCBV can then be estimated from the area under There are multiple limitations to the usefulness of PET in
this curve and corrected for contrast leakage using a refer- the brain. Tumor glucose metabolism is highly variable, and
ence measurement in the contralateral, unaffected centrum the overlap of 18 F-FDG uptake between radiation necrosis
semiovale white matter. and tumor can be considerable [91–93]. In addition, tumor
Sugahara et al. [79] were among the first to propose that glucose metabolism is frequently lower than that of normal
this technique should be utilized when it is necessary to dif- brain tissue; thus, tumor uptake may be difficult to appreciate
ferentiate tumor recurrence from radiation necrosis. In a series amid the background of highly metabolically active normal
of 20 cases, these authors found that a normalized rCBV of 2.6 brain tissue or of treated brain tissue, which tends to have
lower glucose metabolism than untreated brain and may also SPECT
have a wide range of metabolic activity. Also, high 18 F-FDG
uptake can occur from inflammation that may be associated SPECT relies on the injection of a radionuclide such as 201Tl
with necrosis, resulting in misinterpretation. In general, it (Thallium-201) and 99mTc (Technetium-99) and the use of a
has not been possible to define a reliable cutoff of standard- gamma camera and computed tomography (CT) to detect its
ized uptake values to distinguish between tumor and necro- distribution in the brain. Uptake of the radionuclide is greater
sis. Strategies to improve predictive value include comparing in tumor cells than in normal tissue and inflammatory cells
the ratio of the lesion to contralateral normal white or gray and depends on Na/K-ATPase pump activity [107]. Because
matter or determining whether lesion activity is above the uptake requires a viable cell, it is correspondingly low in
expected background activity in adjacent brain tissue [94– necrotic areas. This technique is much less expensive and
97]; in either case, it can be helpful to have the anatomic more widely available than PET but has disadvantages with
information from coregistration with MRI [97, 98]. regard to limited breadth of molecular processes that are
accessible, quantitation accuracy, and spatial resolution.
High-grade tumors have a higher thallium index than
Amino Acid PET low-grade tumors. In a series of 15 biopsy-proven cases,
Schwartz et al. [108] found excellent correlation between
Amino acid PET tracers and amino acid analog PET tracers imaging and histology (14/15 cases) using dual-isotope
constitute another class of tumor imaging agents [99, 100]. SPECT scanning with 201Tl and 99mTc-hexamethylpropylene
These are particularly attractive for imaging brain tumors amine oxime (HMPAO).161 A subsequent study revealed
because of the high uptake in tumor and low uptake in nor- high 201Tl uptake in recurrent tumor and low uptake in radia-
mal brain tissue; amino acids are transported into the cell via tion necrosis. In cases with intermediate uptake, 99mTc-
carrier-mediated processes [100], and amino acid transport is HMPAO uptake helped to differentiate tumor recurrences
generally increased in malignant cells [101, 102]. The most from radiation necrosis by showing increased perfusion in
extensively studied amino acid tracer is 11C-methionine tumor [108]. After these initial exciting reports of sensitivi-
[103]. Because of the short half-life of 11C (approximately ties in excess of 90 % for detecting tumor recurrence with
20 min), 18 F-labeled aromatic amino acid analogs (with a specificity approximately 60 % [109], other reports citing
half-life of approximately 109 min) have been developed for concerns about the diagnostic accuracy of the technique for
tumor imaging [104]. distinguishing recurrent tumor from radiation necrosis began
Because amino acid tracers appear more sensitive than to surface, including cases of thallium-avid radiation necro-
18
F-FDG-PET in visualizing tumor, they also have poten- sis [110–112].
tially better diagnostic accuracy in evaluating radiation More recently, Alexiou et al. [113] substantiated potential
necrosis. However, the degree of amino acid uptake in radia- superiority of 99mTc-tetrofosmin (99mTc-TF) over 99mTc-sestamibi,
tion necrosis is not known. 11C-methionine has been avail- a commonly used imaging agent, for brain tumor imaging,
able for about two decades, but few studies have been owing to the fact that 99mTc-TF accumulation is independent
published on its performance in differentiating radiation of the multidrug-resistance phenotype of some glioma cells
necrosis from brain tumor recurrence.147 In a study of 21 (167,168). 99mTc-TF is a tumor-seeking diphosphine that
patients with brain metastases treated with SRS, does not cross the intact blood–brain barrier, and uptake is
11
C-methionine correctly identified seven of nine recurrences dependent on regional blood flow and cell membrane integ-
and 10 of 12 radiation injury lesions [95]. rity, thus reflecting cellular metabolic status and viability.
In rats, uptake of 18F-FET (18 F-fluoroethyl-L-tyrosine), Alexiou et al. also reported that 99mTc-TF could successfully
18
F-FDG, and 18 F-choline has been compared in acute differentiate tumor recurrence from radiation injury [113,
lesions caused by cerebral irradiation and cryotherapy [105]. 114]. These results are yet to be confirmed by other groups.
Both 18 F-FDG and 18 F-choline accumulated in macrophages,
a common inflammatory cell type in radiation necrosis, but
18
F-FET uptake was absent from macrophages. Moreover, Histological Examination
the ratio of 18 F-FET uptake in radiation necrosis to that in
normal cortex was much lower than the corresponding ratios Given the limitations of the various imaging techniques, a
for 18 F-FDG and 18 F-choline, suggesting that 18 F-FET may definitive diagnosis of radiation necrosis may require patho-
be able to differentiate radiation necrosis from tumor recur- logic examination of surgical specimens. Because of the
rence. Absence of 18 F-FET uptake in a case of radiation potential for sampling error with biopsy alone, total resec-
necrosis has been reported [106]. Although these results are tion is preferred. Resection also improves edema and mass
promising, larger systematic studies are needed to evaluate effect associated with radiation necrosis and/or active tumor.
the diagnostic performance of these tracers. If resection is not safe or feasible, the imaging modalities
described above may be used to guide biopsies, so that the is poorly tolerated or when radiation necrosis becomes pro-
region most suspicious for active tumor can be sampled to gressively larger and symptomatic despite the use of ste-
maximize diagnostic yield. roids, surgical excision of the necrotic mass is indicated [3,
117, 119]. In general, surgical treatment is an effective
therapy for well-circumscribed areas of necrosis and much
Treatment less effective for diffusely necrotic lesions [120]. It is pre-
sumed that removal of the necrotic nidus, a source of
While the finding of radiation necrosis, or “treatment effect,” inflammation, reduces the associated edema and thereby
is often met with relief that the changes seen on imaging do secondary mass effect. Surgery thus provides immediate
not represent tumor growth, patients and their families and relief of symptoms and frequently enables patients to be
health care providers need to understand the implication of rapidly tapered off steroids.
this condition. Radiation necrosis can produce symptoms that
are as disabling as tumor recurrence. Most patients with radia- Hyperbaric Oxygen (HBO). HBO therapy is the inhalation of
tion necrosis present with focal neurologic deficits related to 100 % oxygen at an elevated pressure, usually over 1.5 times
the necrotic lesion as well as headaches as a result of mass the atmospheric pressure, in a controlled and monitored envi-
effect from reactive vasogenic edema. Some cases stabilize or ronment. When used after radiation therapy, HBO has been
improve with the treatments described below or with the pas- shown to increase tissue oxygenation and angiogenesis and
sage of time, but some cases may be relentlessly progressive. improve capillary bed function [121]. HBO has been used in
the treatment and prevention of late complications after radio-
Steroids. Steroids are frequently used to treat radiation- therapy in different sites, and many papers have been pub-
induced edema and necrosis. Animal models of radiation- lished reporting its use [119]. Hulshof et al. [122] prospectively
induced CNS necrosis demonstrate that edema precedes the evaluated HBO in patients with cognitive problems following
development of necrosis and that the anti-inflammatory effect cerebral radiotherapy and found that only one out of seven
of dexamethasone, when provided early, can modify subse- patients experienced significant improvement. Leber et al.
quent vascular and inflammatory changes to reduce and delay [123] reported two cases of successful HBO therapy for radi-
the development of subsequent necrosis [115–117]. Thus, ation damage following Gamma Knife radiosurgery for
controlling edema after irradiation or in the early stages of AVMs, with marked improvement in necrotic lesions on
necrosis may limit the subsequent development or course of imaging without the use of corticosteroids. Given the scarcity
necrosis. Among patients with temporal lobe necrosis after of evidence and lack of controlled trials, there is considerable
radiotherapy for nasopharyngeal cancer, dexamethasone uncertainty on the place of HBO in the management of late
given at the onset of radiation necrosis, during which there neurological sequelae of radiation therapy.
was marked reactive edema, resulted in significant objective
improvement of necrosis in 38 % of patients, whereas none of Anticoagulants. Because it is thought that radiation necrosis
the patients with cystic necrosis improved [118]. results mainly from vascular changes associated with isch-
The optimal dose and timing of the corticosteroids have emia, anticoagulants (heparin and warfarin [Coumadin]) and
not been prospectively studied; however, it appears that antiplatelet medications such as pentoxifylline (Trental),
administration early in the course of radionecrosis is more aspirin, and ticlopidine (Ticlid) have been used to attempt to
effective than one that is delayed, and the doses given must halt its progression. Five of eight patients with cerebral radi-
be adequate to control clinical symptoms. After symptoms ation necrosis unresponsive to steroids had clinical improve-
improve, the steroid dose should be tapered as tolerated. ment on anticoagulation with heparin (Heparin) and warfarin
Some patients become symptomatic when steroid treatment (Coumadin) [124]. The potential risk of bleeding from these
is discontinued and thus require long-term treatment over agents must be weighed against any expected benefit, and
many weeks or months. Attention must then be directed these patients must be closely monitored for consumptive
toward minimizing the medical complications associated coagulopathies, especially in the setting of concurrent che-
with steroid treatment, such as steroid myopathy, glucose motherapy. Definitive conclusions are not available owing to
intolerance, secondary infections, gastric irritation, neuro- the small number of patients studied. However, in cases
psychologic issues, and osteopenia. where radiation necrosis is progressive and symptomatic,
full anticoagulation is reasonable assuming there are no con-
Surgical Resection. Radiation-induced necrosis is clearly a traindications. There are no published comparisons for anti-
dynamic process. Cases have been documented in which platelet versus anticoagulant agents.
necrosis spontaneously resolves, remains stable, or pro-
gressively enlarges to produce a space-occupying lesion Antioxidant Therapy. Oral pentoxifylline (Trental) and
that becomes increasingly symptomatic. When steroid therapy tocopherol (vitamin E) combination therapy has been reported
to benefit patients with skin and chest wall radionecrosis as in volume of contrast enhancement was 79 % and the mean
well as mandibular osteoradionecrosis [125]. Pentoxifylline decrease in T2-FLAIR volume was 49 %. Similar results
is thought to reverse radiation damage primarily through its were published in a recent case report of bevacizumab at
ability to increase locoregional blood flow; it improves red 5 mg/kg every other week for temporal lobe necrosis that
blood cell deformability and promotes laminar blood flow by developed 2 years after aggressive radiotherapy with chemo-
inhibiting intercellular adhesion molecule expression, result- therapy for nasopharyngeal carcinoma [131] and in three of
ing in decreased adhesion between endothelial cells and leu- four children with pontine gliomas suspected of having radi-
kocytes. Pentoxifylline also decreases plasma fibrinogen ation necrosis [132]. More recently, Gutin et al. [39] reported
while increasing fibrinolytic activity [126], and it is a nonspe- no cases of radionecrosis among 25 patients with recurrent
cific inhibitor of many inflammatory cytokines [127]. malignant gliomas treated with fractionated stereotactic reir-
Tocopherol is an antioxidant compound that scavenges ROS radiation to 30 Gy in six fractions along with bevacizumab,
generated during oxidative stress and protects lipid mem- 10 mg/kg every 2 weeks. Optimal dosing, duration of treat-
branes against lipid peroxidation [125]. Williamson et al. ment, and efficacy of bevacizumab (or similar agents) need
[128] reported on 11 patients with delayed adverse radiation further evaluation in prospective trials.
effects after treatment with Gamma Knife radiosurgery for
various benign and malignant disorders who were treated Rehabilitation. Depending on the symptoms and clinical
with pentoxifylline (400 mg orally twice daily) and tocoph- deficits caused by radiation necrosis, patients may benefit
erol (400 IU orally twice daily). The edema volume measured from physical, occupational, or speech therapy.
on serial FLAIR-MRI improved in 10 of 11 patients, with an Brain necrosis is associated with a significant degree of
average volume decrease of 72.3 mL. The only patient with- morbidity and sometimes mortality. It has been recognized
out improvement was found to have tumor recurrence. The as a potential complication of brain radiation therapy for at
authors concluded that pentoxifylline and tocopherol may be least five decades, with higher risk for larger volumes, higher
of benefit for treatment of adverse radiation effects. Further total dose, and higher dose per fraction, especially above
studies are needed to determine the efficacy of this regimen about 50 Gy at 1.8–2.0 Gy per daily fraction or 12 Gy in one
for the treatment of brain radionecrosis. fraction (three- and five-fraction radionecrosis thresholds are
not well know at this time) (see Table 65.1). The latency
Bevacizumab. The premise for the use of bevacizumab for period is highly variable, but the minimum latency is about 3
radiation necrosis came from observations of radionecrosis- months after radiation exposure. The pathophysiology is not
associated endothelial cell dysfunction and hypoxia with lib- completely understood; contributing factors appear to
eration of vasoactive compounds, including vascular include injury to vasculature, oligodendrocytes, and/or neu-
endothelial growth factor (VEGF). Bevacizumab is a human- ral stem cells and chronic inflammation. It may be difficult to
ized murine monoclonal antibody against the VEGF mole- distinguish radiation necrosis from progressive or recurrent
cule. It blocks VEGF from reaching its capillary targets and tumor following treatment for primary or metastatic brain
is currently being used in the treatment of several solid tumors. Techniques that have been used include MRI, MRS,
tumors. Gonzalez et al. [129] retrospectively evaluated 15 PET, and SPECT, short of surgical biopsy or resection for a
patients treated with bevacizumab or bevacizumab combined histopathologic diagnosis. The symptoms of radiation necro-
with chemotherapy after radiation for malignant brain sis may be mild or reversible, but are sometimes significant
tumors. Radiation necrosis was diagnosed in eight patients, or progressive, warranting treatment with steroids and/or
all of whom underwent therapy with bevacizumab on surgical resection. Less well-proven but possibly effective
either a 5 mg/kg/2-weeks or 7.5 mg/kg/3-weeks schedule. interventions include HBO therapy, anticoagulation, antioxi-
Posttreatment MRI scans, performed approximately 8 weeks dant therapy, and therapy with VEGF inhibitors.
after therapy, demonstrated measurable improvements in all
patients. The imaging findings correlated with clinical
improvement as evaluated by the average reduction in daily Complications for Common Radiosurgery
dexamethasone requirement by 8.6 mg. The authors of this Indications
study concluded that bevacizumab, alone or in combination
with other agents, can reduce radiation necrosis by decreas- Brain Metastases. Metastases to the brain represent the most
ing capillary leakage and the associated brain edema. common indication for radiosurgery [6]. In a series evaluating
Torcuator et al. [130] reported on six glioma patients with radiosurgery for brain metastases, Shaw et al. report that the
biopsy-proven brain radionecrosis treated with bevacizumab most common complication was symptomatic cerebral edema
at 10 mg/kg every other week, combined with irinotecan in associated with radiosurgery [8]. Four patients in this series
two cases of suspected tumor cells in addition to necrosis. experienced severe cerebral edema, two of whom responded
Compared with pre-bevacizumab imaging, the mean decrease to corticosteroids. Six percent of patients in this series
Table 65.1 Summary of suggested dose constraints for selected critical brain structures
One fraction Three fractions Five fractions

Max critical
volume above Threshold Max. point Threshold Max. point Threshold Max point End point
Tissue threshold dose (Gy) dose (Gy) dose (Gy) dose (Gy) dose (Gy) dose (Gy) (>-Grade 3)
Optic pathway <0.2 cm3 8 10 15.3 (5.1 Gy/fx) 17.4 (5.8 Gy/fx) 23 (4.6 Gy/fx) 25 (5 Gy/fx) Neuritis
Cochlea 9 17.1 (5.7 Gy/fx) 25 (5 Gy/fx) Hearing loss
Brainstem <0.5 cm3 10 15 18 (6 Gy/fx) 23.1 (7.7 Gy/fx) 23 (4.6 Gy/fx) 31 (6.2 Gy/fx) Cranial
(not medulla) neuropathy
Spinal cord and <0.35 cm3 10 14 18 (6 Gy/fx) 21.9 (7.3 Gy/fx) 23 (4.6 Gy/fx) 30 (6 Gy/fx) Myelitis
medulla <1.2 cm3 7 12.3 (4.1 Gy/fx) 14.5 (2.9 Gy/fx)
Spinal cord <10 % of 10 14 18 (6 Gy/fx) 21.9 (7.3 Gy/fx) 23 (4.6 Gy/fx) 30 (6 Gy/fx) Myelitis
subvolume subvolume
(5–6 mm above
and below level
treated per Ryu)
Adapted from Benedict SH, Yenice KM, Followill D, Galvin JM, Hinson W, Kavanagh B, et al. Stereotactic body radiation therapy: The report of
AAPM task group 101. Med Phys 2010, Aug;37(8):4078-101
required surgical resection for persistent and symptomatic metabolized through the hepatic system (e.g., levetiracetam).
radionecrosis within 1 year of SRS. These investigators noted If the time from radiosurgery is weeks to months, then
that larger target volume >8.2 cm3 was significantly associ- edema, necrosis, tumor recurrence, and/or hemorrhage as
ated with increased acute and delayed complications due to causes of symptoms should be considered. Evaluation in this
radiosurgery [8]. In a different series of 40 patients with intra- time period almost always necessitates evaluation with a CT
cranial metastasis, Mehta et al. [133] reported a 10 % radia- scan (if acute onset and suspicion of cerebral hemorrhage is
tion toxicity rate associated with radiosurgery. high; melanoma, renal cell carcinoma, and uterine carci-
Since most patients with brain metastases are exposed to noma metastases have higher propensity to bleed) or MRI. If
various systemic therapies, often involving novel targeted symptoms are relatively mild, then outpatient treatment with
agents, radiosensitizing effects of these new agents are not corticosteroids (or increasing the dose) is a reasonable first
well known and understood. There are some case reports sug- step. If there is no improvement in symptoms within several
gesting enhanced cutaneous toxicity with the combination of days despite increased dose of steroids, then repeat MR
BRAF inhibitors and radiotherapy but whether these effects imaging is indicated to help distinguish radionecrosis from
translate into greater neurotoxicity is not known [134]. tumor or rule out other causes of symptoms. If symptoms
Symptomatic cerebral edema is the most commonly resolve with steroids and the medication can be tapered and
reported complication of SRS for intracranial metastases withdrawn without symptomatic recurrence, then no further
with a range of 3–18 % [8, 135]. Typically, these complica- intervention is required. Because long-term high-dose ste-
tions occur over a broad time period. Focal symptoms are roids can cause a host of problems including hyperglycemia,
usually closely related to the location of the injury and/or to muscle weakness, Cushingoid facies, insomnia, irritability
increased intracranial pressure. Whenever a patient becomes and fatigue, other alternatives need to be investigated depend-
symptomatic after radiosurgery with a headache, seizure, or ing on the cause of symptoms (see Fig. 65.3).
progressive neurologic deficit, the interval since radiosur- Intracranial Meningiomas. Meningiomas are most com-
gery and the clinical setting during which radiosurgery was mon benign intracranial tumors and account for approxi-
performed must first be considered. For example, was the mately 20 % of all primary brain tumors [136, 137].
patient treatment with radiosurgery alone or after whole- Radiosurgery is often used as an alternative to surgery for
brain radiotherapy (WBRT)? Apart from radiation side some skull base meningiomas or those in difficult locations
effects, tumor recurrence and growth must also be consid- where complete surgical resection would not be possible or
ered as possible causes of symptoms. Acute side effects in subtotal resection would be associated with unacceptable
the first few hours or days after radiosurgery can be treated neurologic morbidity [138, 139]. This generally includes
symptomatically for complaints of pain, nausea, and fatigue. petroclival, parasellar, and cavernous sinus locations.
Seizures are quite rare in the acute setting, but if they occur, Radiosurgery is also an excellent treatment option for many
anticonvulsant levels should be checked (if applicable), dose elderly or those with significant comorbidities that make sur-
adjusted (an additional half-loading dose given [orally or IV] gical resection less appealing [7, 140]. In many cases, radio-
and the maintenance dose increased pending the results), or surgery is pursued as a therapeutic option even without a
the patients should be converted to anticonvulsants not tissue diagnosis [7]. Single session radiosurgery is generally
Brain Metastases
SRS
Acute Phase Early-Delayed Phase Late-Delayed Phase

(days to weeks) (1-6 months) (>6 months)
Symptomatic Treatment: Symptomatic Treatment: Symptomatic Treatment:

1. Seizures - trial of steroids, if persistent - trial of steroids, if persistent
- check drug levels
- give 1/2 loading dose
- recheck levels in 7-10 days
- consider adding or changing to other drug
2. Headaches
- OTC medications
- trial of steroids, if persistent
Improvement No Improvement
Consider slow steroid taper

after 10-14 days of therapy
CT scan to rule out hemorrhage MRI to rule out recurrence

- renal cell carcinoma (RCC),
melanoma, choriocarcinoma
Suspected Radionecrosis: Tumor Recurrence:

1. additional imaging (pMRI, DWI, AA-PET) 1. Consider salvage therapy
2. institute empiric treatment - surgery (if symptomatic)
- start/continue steroids (adjust dose) - repeat radiosurgery
- pentoxyfilline, vitamin E (tocopherol) - whole-brain RT (WBRT)
- bevacizumab (Avastin) - other
- surgical resection
Fig. 65.3 Complications management algorithm in patients with brain metastases treated with stereotactic radiosurgery (SRS)
limited to lesions <3.5 cm in size and more than 2–5 mm The optic apparatus is quite sensitive to radiation injury
distant from the anterior optic structures [141–143]. Lesions [150]. For single-session radiosurgery, the studies have indi-
>3.5 cm in size may still be accessible to multisession radio- cated the relative incidence of radiation-induced optic neu-
surgery (three or five fractions) [144]. ropathy to be rare for maximal doses to anterior optic
Radiosurgical complications after treatment for meningi- structures of <8 Gy. The risk increases in the range of
oma include peritumoral edema (which can become symp- 8–12 Gy and exceeds 10 % in the range of 12–15 Gy. One
tomatic), seizures, and cranial and optic neuropathies major study indicates a low risk with maximal dose of
[145–147]. Symptomatic peritumoral edema tends to occur <12 Gy to the anterior optic structures [151], but there is
at a higher frequency at parasagittal and cerebellopontine general agreement that Dmax ≤10 Gy is safe [152].
angle locations [148]. Cranial neuropathies, including optic Other cranial nerves appear to be more resistant to radia-
nerve injury, are the highest risk complications of cavernous tion than the optic apparatus, and doses up to 20 Gy can often
sinus meningioma radiosurgery [139, 149]. In a report by be tolerated [7]. Kondziolka et al. [153] reported on a series
Singh et al. on early complications of Gamma Knife radio- of 50 patients treated for meningiomas with a mean marginal
surgery for intracranial meningiomas in 77 patients, seizures dose of 17 Gy. In this series, the actuarial 2-year local control
and headaches occurred in about 10 % of patients, and a sig- rate was 96 % and only three patients exhibited delayed
nificant number (22 %) of parasagittal meningiomas resulted radiosurgical complications of trigeminal hyperesthesia,
in symptomatic edema. For this reason, treatment of larger hemiparesis, and oculomotor neuropathy between 3 and 12
tumors in the parasagittal location should generally be months after treatment. All of these complications slowly
avoided and fractionated techniques considered. improved over time [153].
Meningioma
SRS

Symptomatic Treatment: Symptomatic Treatment: MRI scan

1. Seizures - trial of steroids, if persistent
- check drug levels
- recheck levels in 7-10 days T2 change only Necrosis Tumor Recurrence
2. Headaches
- OTC medications
3. Nausea/vomiting
- anti-emetics Improvement No Improvement
Consider slow steroid taper

after 10-14 days of therapy
MRI scan MRI scan Consider salvage surgery

T2 change only, no symptoms T2 change and symptoms
Continue Observation Continue Steroids

(adjust dose as needed)
Suspected Radionecrosis:
1. additional imaging (pMRI, DWI, AA-PET)
2. institute empiric treatment
- start/continue steroids (adjust dose)
- pentoxyfilline, vitamin E (tocopherol)
- bevacizumab (Avastin)
Fig. 65.4 Complications management algorithm in patients with meningioma(s) treated with stereotactic radiosurgery (SRS)
Most cranial neuropathies after radiosurgery are mild and maximal dose to the anterior optic structures (chiasm, optic
improve with time, but uncommon severe complications nerves, and optic radiations) to <10 Gy. Many practioners
such as hemiparesis can develop after treatment [149, 154, chose to limit the dose to ≤8 Gy. It is reasonable to manage
155]. For example, Chang et al. reported on a series of 179 peritumoral edema (and associated symptoms) as well as
patients treated with Gamma Knife radiosurgery for menin- cranial neuropathies with a trial of glucocorticoids and then
giomas and noted complications in nearly 25 %, mostly follow basic clinical principles for complication manage-
asymptomatic edema found on follow-up imaging, and about ment as outlined in Fig. 65.4.
2 % cranial neuropathy and radionecrosis rate. Another study
by Hakim et al. [156] reported on 127 patients treated with Vestibular Schwannomas. Vestibular schwannomas are com-
linac radiosurgery for meningiomas (benign, atypical, and mon benign tumors that are frequently managed with micro-
malignant). They reported 2 deaths (1.6 %) due to treatment- surgery and/or radiosurgery. After radiosurgery or radiation
related edema and subsequent herniation as well as relatively therapy, neurologic structures adjacent to vestibular schwan-
high rates of hemiparesis. This study emphasizes the impor- nomas are at risk for radiation toxicity. These structures
tance of patient selection for radiosurgery and the impor- include the brainstem and trigeminal, facial, and vestibuloco-
tance of not treating large tumor volumes with single-session chlear nerves. It has been suggested that oligodendrocytes of
radiosurgery, and especially not in the retreatment setting. the brainstem are more sensitive to radiation injury than
In general, skull base meningiomas can be effectively peripheral Schwann cells that myelinate the cranial nerves in
treated (local control rates >95 %) with radiosurgery with a the transition zone and distally [7, 157, 158]. Because of this,
low complication profile (≤10 %). These complications typi- it has been suggested that vestibular schwannomas may be at
cally involve cranial neuropathies, which are mild and often an increased risk of mild acute complications from SRS [159].
recover over time, except for injury to the anterior optic Miller et al. [160] reported on a series of 82 patients with
structures. Optic neuropathy may be a permanent vestibular schwannomas treated with a single-session high-
complication of radiosurgery but is preventable by limiting the versus low-dose SRS. This report suggested that a marginal
dose in excess of 16 Gy is significantly associated with per- same study, audiometric data demonstrated that useful hear-
manent facial neuropathy. Median onset to neuropathy was 6 ing was preserved in only 33 % of patients; 67 % of the
months in this series. Interestingly, even though higher mar- treated cohort lost functional hearing after SRS [162]. Other
ginal dose appears to be a risk factor for trigeminal and facial single institution reports noted 40–66 % hearing preserva-
neuropathy, nearly half of the patients in this series with neu- tion rates [163–165].
ropathies experience symptomatic improvement over time Other less common late-delayed complications reported
[160]. Other factors that affect neuropathy rates are the length after SRS for vestibular schwannomas include facial spasm
of the cranial nerve irradiated and brainstem dose near the and hydrocephalus [166, 167]. In some cases, shunting may
root entry zone. Foote et al. [161] reported these correlations be required to address the hydrocephalus.
in their study of 149 patients with vestibular schwannomas Overall, the complication risk from radiosurgery using
who were treated definitively with SRS. They found that a dose of 12 or 13 Gy for vestibular schwannomas is quite
maximal brainstem dose (Dmax) in excess of 17.5 Gy was low. Most modern series report trigeminal and facial neu-
found to closely correlate with trigeminal and facial neuropa- ropathy risks at less than 8 %, and in most cases these neu-
thies in their series. For this reason, they recommended (and ropathies improve with time. Because the number of studies
it is also our institutional practice) to treat with a dose of that rigorously assess hearing preservation after radiosur-
12.5 Gy at the tumor margin. This dose balances efficacy of gery is relatively small with limited sample sizes and retro-
treatment with risks of complications. Surgical resection prior spective design, the hearing preservation rate after
to SRS was also identified as a significant risk factor for cra- radiosurgery is not well defined, with more recent reports
nial neuropathy in this study. being in the 50–70 % range (with variable follow-up). To
Andrews et al. [162] confirmed some of these findings in address the problem of inconsistent reporting, a consensus
a report of 69 patients treated with SRS for vestibular meeting on reporting systems for vestibular schwannoma
schwannomas with a 12 Gy at the tumor margin. Facial and was convened in 2003 and created reporting definitions;
trigeminal neuropathies occurred in only 2–5 % of treated however, these have not been widely adopted by investiga-
patients with a median follow-up greater than 2 years. In this tors (Fig. 65.5) [168].
Vestibular Schwannoma
SRS

Symptomatic Treatment: MRI scan Symptomatic Treatment: Motor & Sensory Symptoms
1. Nausea/vomiting T2 change only, no symptoms 1. Trigeminal pain
- anti-emetics - carbamazepine, gabapentin
- trial of steroids, if persistent - trial of steroids, if persistent
2. Headaches 2. Hearing loss or facial weakness
- OTC medications - trial of steroids, if persistent
3. Exacerbation of existing symptoms
Improvement No Improvement MRI scan
Continue Observation Consider slow steroid taper Continue Steroids

after 10-14 days of therapy (adjust dose as needed)
T2 change only and symptoms: Suspected Radionecrosis:

1. short-course steroids; consider 1. additional imaging (pMRI, DWI, AA-PET)
slow taper after 10-14 days of therapy 2. institute empiric treatment
if improvement - start/continue steroids (adjust dose)
2. if no improvement, consider surgery - pentoxyfilline, vitamin E (tocopherol)
- bevacizumab (Avastin)
Fig. 65.5 Complications management algorithm in patients with vestibular schwannoma treated with stereotactic radiosurgery (SRS)
AVMs. Radiosurgery is an effective treatment of AVMs in The management of complications in AVM patients is
any brain location for both children and adults, with oblitera- more involved than that of brain metastases after SRS since
tion rates depending primarily on the AVM size and prescrip- it is important to rule out intracranial hemorrhage via CT
tion dose. Complications of treatment also depend on the urgently in patients who present with a new, severe head-
AVM location, size, prescription dose, 12-Gy isodose vol- ache, nausea, vomiting, and/or new neurologic deficit. For
ume, age, radiographic method used for target definition, those patients with history of prior hemorrhage and who are
perforating artery supply, and era of treatment. Complications not candidates for surgical intervention and have intranidal
of AVM treatment are highly variable and inconsistently aneurysm, evaluation for endovascular treatment of the
reported. aneurysm is recommended (but not of the entire AVM).
The treatment of pediatric AVMs poses unique challenges Patients with known feeding artery aneurysms should first
because the effects of radiation on developing brain and the undergo surgical or endovascular treatment of these aneu-
risk of second malignancy need to be considered. Balanced rysms before radiosurgery. Early-delayed effects of SRS
against these concerns are the mortality and morbidity of such as edema are not uncommon but majority of patients
hemorrhagic events over the number of remaining years of tend to remain asymptomatic. Cyst formation, as a late com-
life (assuming 2 % annual incidence of hemorrhage). For plication of SRS, is rare and can be treated surgically if it is
example, Kaido et al. [169] reported on a development of a symptomatic (see Fig. 65.6).
glioblastoma 6.5 years in a boy who was treated for a peri-
ventricular AVM 6.5 years earlier with a marginal dose of
20 Gy. Levy et al. [170] reported on a series of 53 consecu- Conclusion
tive children who underwent at least 3 years of imaging fol-
low up after radiosurgery of intracranial AVMs. They Radiosurgery is a very effective and popular treatment option
grouped patients into three groups based on the volume of for several common intracranial diagnoses and represents a
the treated AVM: (1) group 1, ≤3 mL, (2) group 2, >3 mL but major development in neurosurgery and radiation oncology
≤10 mL, and (3) group 3, >10 mL. Twenty-eight patients over the last 30 years. For some treatment indications, such
(80 %) in group 1 and 11 (64.7 %) in group 2 achieved comas small vestibular schwannomas and meningiomas, it is
plete obliteration during the follow-up period, and only one becoming the de facto treatment of choice given its effective-
patient in group 3 did not achieve obliteration. This study ness and favorable side effect profile. As with any other form
demonstrated effectiveness of radiosurgical treatment for of therapy, there remains the risk of treatment-associated
pediatric AVMs. Brainstem edema was the primary intracra- complications, although these risks are small and have been
nial complication, reported in 1 patient in the series. Four reduced as experience with various radiosurgical techniques
patients experienced hemorrhagic events between 30 and 98 has grown. The key to management of radiosurgical compli-
months after SRS. In this study, small target volume dose cations is prevention (to the extent of possible) of known or
were associated with likelihood of obliteration. expected side effects by judicious and thoughtful application
A comprehensive report by Flickinger et al. [171] did of radiosurgical principles and available dose–volume con-
much to solidify our understanding of the risk of complica- straints to guide treatment planning. Robust and rigorous
tions after AVM radiosurgery. In an analysis of 85 patients quality assurance and safety programs need to be in place to
with a median follow-up of 45 months, the volume of brain maximize patient safety and achieve the desired clinical out-
receiving ≥12 Gy was used to construct a post-radiosurgery comes. This requires a multidisciplinary team and constant
expression score (PIE). In addition to 12-Gy volume, ana- evaluation of outcomes data. Since most of the radiosurgical
tomic location was also associated with increased risk of data are generally of limited quality (retrospective, single-
complication (from low to high risk: frontal, temporal, institutional studies), it is imperative that further/future
intraventricular, parietal, cerebellar, corpus callosum, occipi- research, including toxicity reporting and management, is
tal, medulla, thalamus, basal ganglia, and pons/medulla). standardized and consistently reported to aid in interpreta-
Larger AVMs are associated with lower obliteration rates, tion of published outcomes and toxicity data. Use of consen-
longer time to obliteration after treatment, and an increased sus outcomes definitions needs to be made part of daily
incidence of complications. Two prevailing approaches were clinical practice and use of anecdotal data to justify treat-
used to try to improve outcomes in these patients: (1) multi- ment decisions needs to be discouraged. This robust and pro-
session radiosurgery and (2) volume-staged treatments. spective approach must be applied to treatment-related
Single-institution reports of these various approaches report toxicity management, especially as new treatment approaches
marginal success in achieving obliteration rates and there is (e.g., bevacizumab for radiation necrosis) become increas-
no convincing study to suggest that one approach should be ingly available and applications of radiosurgery continue to
preferred over the other [172–176]. be extended.
Arteriovenous Malformation
SRS

Symptomatic Treatment: Severe headaches Mild worsening symptoms

1. Seizures New focal deficit Increased seizures
- check drug levels
- recheck levels in 7-10 days
2. Headaches MRI scan MRI scan MRI scan
- OTC medications T2 change only, no symptoms T2 change and symptoms Edema, necrosis, or cyst
3. Nausea/vomiting
- anti-emetics Continue Observation Start/continue Steroids
- trial of steroids, if persistent (adjust dose as needed)
CT to rule out hemorrhage Suspected Radionecrosis:

1. additional imaging (pMRI, DWI, AA-PET)
2. institute empiric treatment
- start/continue steroids (adjust dose)
- pentoxyfilline, vitamin E (tocopherol)
Hemorrhage No Hemorrhage
3. AVM resection
- Re-evaluate for surgery
Fig. 65.6 Complications management algorithm in patients with arteriovenous malformations (AVMs) treated with stereotactic radiosurgery
(SRS)
11. Quigg M, Yen CP, Chatman M, Quigg AH, Macneill IT,

References Przybylowski CJ, Yan G, Sheehan JP. Risks of history of diabetes
mellitus, hypertension, and other factors related to radiation-
1. Quigg M, Barbaro NM. Stereotactic radiosurgery for treatment of induced changes following Gamma Knife surgery for cerebral arte-
epilepsy. Arch Neurol. 2008;65(2):177–83. riovenous malformations. J Neurosurg. 2012;117(Suppl):144–9.
2. Starke RM, Komotar RJ, Hwang BY, Fischer LE, Otten ML, 12. Kumar AJ, Leeds NE, Fuller GN, Van Tassel P, Maor MH, Sawaya
Merkow MB, Garrett MC, Isaacson SR, Connolly ESJ. A compre- RE, Levin VA. Malignant gliomas: MR imaging spectrum of radi-
hensive review of radiosurgery for cerebral arteriovenous malfor- ation therapy- and chemotherapy-induced necrosis of the brain
mations: outcomes, predictive factors, and grading scales. after treatment. Radiology. 2000;217(2):377–84.
Stereotact Funct Neurosurg. 2008;86(3):191–9. 13. Arnold A, Bailey P. Alterations in the glial cells following irradiation
3. Gutin PH, Leibel SA, Sheline GE. Radiation injury to the nervous of the brain in primates. AMA Arch Pathol. 1954;57(5):383–91.
system. New York: Raven Press; 1991. xiv, p. 482. 14. Russell DS, Wilson CW, Tansley K. Experimental radio-necrosis
4. Hopewell JW, Wright EA. The nature of latent cerebral irradiation of the brain in rabbits. J Neurol Neurosurg Psychiatry. 1949;
damage and its modification by hypertension. Br J Radiol. 12(3):187–95.
1970;43(507):161–7. 15. Lowenberg-Scharenberg K, Bassett RC. Amyloid degeneration of
5. Sheline GE, Wara WM, Smith V. Therapeutic irradiation and brain the human brain following X-ray therapy. J Neuropathol Exp
injury. Int J Radiat Oncol Biol Phys. 1980;6(9):1215–28. Neurol. 1950;9(1):93–102, illust.
6. Hong TS, Tomé WA, Hayes L, Yuan Z, Badie B, Rao R, Mehta 16. Lampert P, Tom MI, Rider WD. Disseminated demyelination of
MP. Acute sequelae of stereotactic radiosurgery. 2004. the brain following Co60 (gamma) radiation. Arch Pathol. 1959;
7. Smith MC, Ryken TC, Buatti JM. Radiotoxicity after conformal 68:322–30.
radiation therapy for benign intracranial tumors. Neurosurg Clin 17. Nielsen SL, Kjellberg RN, Asbury AK, Koehler AM.
N Am. 2006;17(2):169–80. Neuropathologic effects of proton-beam irradiation in man. II.
8. Shaw E, Scott C, Souhami L, Dinapoli R, Bahary J-P, Kline R, Evaluation after pituitary irradiation. Acta Neuropathol. 1972;
Wharam M, Schultz C, Davey P, Loeffler J. Radiosurgery for the 21(1):76–82.
treatment of previously irradiated recurrent primary brain tumors and 18. Nielsen SL, Kjellberg RN, Asbury AK, Koehler AM.
brain metastases: initial report of radiation therapy oncology group Neuropathologic effects of proton-beam irradiation in man. I.
protocol 90–05. Int J Radiat Oncol Biol Phys. 1996;34(3):647–54. Dose-response relationships after treatment of intracranial neo-
9. Erbetta A, Salmaggi A, Sghirlanzoni A, Silvani A, Potepan P, plasms. Acta Neuropathol. 1972;20(4):348–56.
Botturi A, Ciceri E, Bruzzone MG. Clinical and radiological fea- 19. Pennybacker J, Russell DS. Necrosis of the brain due to radiation
tures of brain neurotoxicity caused by antitumor and immunosup- therapy; clinical and pathological observations. J Neurol
pressant treatments. Neurol Sci. 2008;29(3):131–7. Neurosurg Psychiatry. 1948;11(3):183–98.
10. Lawrence YR, Li XA, el Naqa I, Hahn CA, Marks LB, Merchant 20. Eyster EF, Nielsen SL, Sheline GE, Wilson CB. Cerebral radiation
TE, Dicker AP. Radiation dose-volume effects in the brain. Int J necrosis simulating a brain tumor. Case report. J Neurosurg.
Radiat Oncol Biol Phys. 2010;76(3 Suppl):S20–7. 1974;40(2):267–71.
21. Ghatak NR, White BE. Delayed radiation necrosis of the hypo- bevacizumab with hypofractionated stereotactic irradiation for
thalamus. Report of a case simulating recurrent craniopharyngi- recurrent malignant gliomas. Int J Radiat Oncol Biol Phys. 2009;
oma. Arch Neurol. 1969;21(4):425–30. 75(1):156–63.
22. Lee AW, Foo W, Chappell R, Fowler JF, Sze WM, Poon YF, Law 40. Kreth FW, Faist M, Warnke PC, Rossner R, Volk B, Ostertag CB.
SC, Ng SH, O SK, Tung SY, Lau WH, Ho JH. Effect of time, dose, Interstitial radiosurgery of low-grade gliomas. J Neurosurg. 1995;
and fractionation on temporal lobe necrosis following radiother- 82(3):418–29.
apy for nasopharyngeal carcinoma. Int J Radiat Oncol Biol Phys. 41. Scharfen CO, Sneed PK, Wara WM, Larson DA, Phillips TL,
1998;40(1):35–42. Prados MD, Weaver KA, Malec M, Acord P, Lamborn KR, et al.
23. Lee AW, Kwong DL, Leung SF, Tung SY, Sze WM, Sham JS, Teo High activity iodine-125 interstitial implant for gliomas. Int J
PM, Leung TW, Wu PM, Chappell R, Peters LJ, Fowler JF. Factors Radiat Oncol Biol Phys. 1992;24(4):583–91.
affecting risk of symptomatic temporal lobe necrosis: significance 42. Shrieve DC, Er A, Black PM, Wen PY, Fine HA, Kooy HM,
of fractional dose and treatment time. Int J Radiat Oncol Biol Loeffler JS. Treatment of patients with primary glioblastoma mul-
Phys. 2002;53(1):75–85. tiforme with standard postoperative radiotherapy and radiosurgi-
24. Ruben JD, Dally M, Bailey M, Smith R, McLean CA, Fedele cal boost: prognostic factors and long-term outcome. J Neurosurg.
P. Cerebral radiation necrosis: incidence, outcomes, and risk fac- 1999;90(1):72–7.
tors with emphasis on radiation parameters and chemotherapy. Int 43. Sneed PK, McDermott MW, Gutin PH. Interstitial brachytherapy
J Radiat Oncol Biol Phys. 2006;65(2):499–508. procedures for brain tumors. Semin Surg Oncol. 1997;13(3):
25. Douglas BG, Fowler JF. Letter: fractionation schedules and a qua- 157–66.
dratic dose-effect relationship. Br J Radiol. 1975;48(570):502–4. 44. Wowra B, Schmitt HP, Sturm V. Incidence of late radiation necro-
26. Barendsen GW. Dose fractionation, dose rate and iso-effect rela- sis with transient mass effect after interstitial low dose rate radio-
tionships for normal tissue responses. Int J Radiat Oncol Biol therapy for cerebral gliomas. Acta Neurochir (Wien).
Phys. 1982;8(11):1981–97. 1989;99(3–4):104–8.
27. Hall EJ, Giaccia AJ. Radiobiology for the radiologist. Philadelphia: 45. Imperato JP, Paleologos NA, Vick NA. Effects of treatment on
Lippincott Williams & Wilkins; 2006. ix, p. 546. long-term survivors with malignant astrocytomas. Ann Neurol.
28. Milano MT, Constine LS, Okunieff P. Normal tissue tolerance 1990;28(6):818–22.
dose metrics for radiation therapy of major organs. Semin Radiat 46. Burger PC, Kamenar E, Schold C, Fay JW, Phillips GL, Herzig
Oncol. 2007;17(2):131–40. GP. Encephalomyelopathy following high-dose BCNU therapy.
29. Marks JE, Bagĺan RJ, Prassad SC, Blank WF. Cerebral radione- Cancer. 1981;48:1318–27.
crosis: incidence and risk in relation to dose, time, fractionation 47. Schultheiss TE, Stephens LC, Maor MH. Analysis of the histopa-
and volume. Int J Radiat Oncol Biol Phys. 1981;7(2):243–52. thology of radiation myelopathy. Int J Radiat Oncol Biol Phys.
30. Gutin PH, Leibel SA, Sneline GE. Radiation injury to the nervous 1988;14(1):27–32.
system. 1991. 48. Yoshii Y. Pathological review of late cerebral radionecrosis. Brain
31. Shaw E, Arusell R, Scheithauer B, O’Fallon J, O’Neill B, Dinapoli Tumor Pathol. 2008;25(2):51–8.
R, Nelson D, Earle J, Jones C, Cascino T. Prospective randomized 49. Tsuruda JS, Kortman KE, Bradley WG, Wheeler DC, Van Dalsem
trial of low-versus high-dose radiation therapy in adults with W, Bradley TP. Radiation effects on cerebral white matter: MR
supratentorial low-grade glioma: initial report of a North Central evaluation. AJR Am J Roentgenol. 1987;149(1):165–71.
Cancer Treatment Group/Radiation Therapy Oncology Group/ 50. Dooms GC, Hecht S, Brant-Zawadzki M, Berthiaume Y, Norman
Eastern Cooperative Oncology Group study. J Clin Oncol. 2002; D, Newton TH. Brain radiation lesions: MR imaging. Radiology.
20(9):2267–76. 1986;158(1):149–55.
32. Flickinger JC, Kondziolka D, Pollock BE, Maitz AH, Lunsford 51. Mintz A, Perry J, Spithoff K, Chambers A, Laperriere
LD. Complications from arteriovenous malformation radiosur- N. Management of single brain metastasis: a practice guideline.
gery: multivariate analysis and risk modeling. Int J Radiat Oncol Curr Oncol. 2007;14(4):131–43.
Biol Phys. 1997;38(3):485–90. 52. Verma N, Cowperthwaite MC, Burnett MG, Markey
33. Korytko T, Radivoyevitch T, Colussi V, Wessels BW, Pillai K, MK. Differentiating tumor recurrence from treatment necrosis: a
Maciunas RJ, Einstein DB. 12 Gy gamma knife radiosurgical review of neuro-oncologic imaging strategies. Neuro Oncol.
volume is a predictor for radiation necrosis in non-AVM intracra- 2013;15(5):515–34.
nial tumors. Int J Radiat Oncol Biol Phys. 2006;64(2):419–24. 53. Dequesada IM, Quisling RG, Yachnis A, Friedman WA. Can stan-
34. Shaw E, Scott C, Souhami L, Dinapoli R, Kline R, Loeffler J, dard magnetic resonance imaging reliably distinguish recurrent
Farnan N. Single dose radiosurgical treatment of recurrent previ- tumor from radiation necrosis after radiosurgery for brain metasta-
ously irradiated primary brain tumors and brain metastases: Final ses? A radiographic-pathological study. Neurosurgery. 2008;63(5):
report of RTOG protocol 90–05. Int J Radiat Oncol Biol Phys. 898–903. discussion 904.
2000;47(2):291–8. 54. Kang TW, Kim ST, Byun HS, Jeon P, Kim K, Kim H, Lee JI.
35. Gilbert HA, Kagan AR. Radiation damage to the nervous system: Morphological and functional MRI, MRS, perfusion and diffusion
a delayed therapeutic hazard. New York: Raven Press; 1980. xiv, changes after radiosurgery of brain metastasis. Eur J Radiol.
p. 209. 2009;72(3):370–80.
36. Babu R, Huang PP, Epstein F, Budzilovich GN. Late radiation 55. Fan G, Sun B, Wu Z, Guo Q, Guo Y. In vivo single-voxel proton
necrosis of the brain: case report. J Neurooncol. 1993;17(1):37–42. MR spectroscopy in the differentiation of high-grade gliomas and
37. Soffietti R, Sciolla R, Giordana MT, Vasario E, Schiffer D. Delayed solitary metastases. Clin Radiol. 2004;59(1):77–85.
adverse effects after irradiation of gliomas: clinicopathological 56. Law M, Cha S, Knopp EA, Johnson G, Arnett J, Litt AW. High-
analysis. J Neurooncol. 1985;3(2):187–92. grade gliomas and solitary metastases: differentiation by using
38. Woo E, Lam K, Yu YL, Ma J, Wang C, Yeung RT. Temporal lobe perfusion and proton spectroscopic MR imaging. Radiology.
and hypothalamic-pituitary dysfunctions after radiotherapy for 2002;222(3):715–21.
nasopharyngeal carcinoma: a distinct clinical syndrome. J Neurol 57. Moller-Hartmann W, Herminghaus S, Krings T, Marquardt G,
Neurosurg Psychiatry. 1988;51(10):1302–7. Lanfermann H, Pilatus U, Zanella FE. Clinical application of pro-
39. Gutin PH, Iwamoto FM, Beal K, Mohile NA, Karimi S, Hou BL, ton magnetic resonance spectroscopy in the diagnosis of intracra-
Lymberis S, Yamada Y, Chang J, Abrey LE. Safety and efficacy of nial mass lesions. Neuroradiology. 2002;44(5):371–81.
58. Nelson SJ, McKnight TR, Henry RG. Characterization of 74. Sugahara T, Korogi Y, Kochi M, Ikushima I, Shigematu Y, Hirai T,
untreated gliomas by magnetic resonance spectroscopic imaging. Okuda T, Liang L, Ge Y, Komohara Y, Ushio Y, Takahashi
Neuroimaging Clin N Am. 2002;12(4):599–613. M. Usefulness of diffusion-weighted MRI with echo-planar tech-
59. Sabatier J, Gilard V, Malet-Martino M, Ranjeva JP, Terral C, Breil nique in the evaluation of cellularity in gliomas. J Magn Reson
S, Delisle MB, Manelfe C, Tremoulet M, Berry I. Characterization Imaging. 1999;9(1):53–60.
of choline compounds with in vitro 1H magnetic resonance spec- 75. Krabbe K, Gideon P, Wagn P, Hansen U, Thomsen C, Madsen
troscopy for the discrimination of primary brain tumors. Invest F. MR diffusion imaging of human intracranial tumours.
Radiol. 1999;34(3):230–5. Neuroradiology. 1997;39(7):483–9.
60. Chernov MF, Hayashi M, Izawa M, Usukura M, Yoshida S, Ono Y, 76. Castillo M, Smith JK, Kwock L, Wilber K. Apparent diffusion
Muragaki Y, Kubo O, Hori T, Takakura K. Multivoxel proton MRS coefficients in the evaluation of high-grade cerebral gliomas.
for differentiation of radiation-induced necrosis and tumor recur- AJNR Am J Neuroradiol. 2001;22(1):60–4.
rence after gamma knife radiosurgery for brain metastases. Brain 77. Rosen BR, Belliveau JW, Vevea JM, Brady TJ. Perfusion imaging
Tumor Pathol. 2006;23(1):19–27. with NMR contrast agents. Magn Reson Med. 1990;14(2):249–65.
61. Fayed N, Davila J, Medrano J, Olmos S. Malignancy assessment 78. Potchen EJ. Magnetic resonance angiography: concepts & appli-
of brain tumours with magnetic resonance spectroscopy and cations. St. Louis: Mosby-Year Book; 1993. xiv, p. 650.
dynamic susceptibility contrast MRI. Eur J Radiol. 2008;67(3): 79. Sugahara T, Korogi Y, Tomiguchi S, Shigematsu Y, Ikushima I,
427–33. Kira T, Liang L, Ushio Y, Takahashi M. Posttherapeutic intraaxial
62. Wald LL, Nelson SJ, Day MR, Noworolski SE, Henry RG, Huhn brain tumor: the value of perfusion-sensitive contrast-enhanced
SL, Chang S, Prados MD, Sneed PK, Larson DA, Wara WM, MR imaging for differentiating tumor recurrence from nonneo-
McDermott M, Dillon WP, Gutin PH, Vigneron DB. Serial proton plastic contrast-enhancing tissue. AJNR Am J Neuroradiol.
magnetic resonance spectroscopy imaging of glioblastoma multi- 2000;21(5):901–9.
forme after brachytherapy. J Neurosurg. 1997;87(4):525–34. 80. Barajas RF, Chang JS, Sneed PK, Segal MR, McDermott MW,
63. Lin A, Bluml S, Mamelak AN. Efficacy of proton magnetic reso- Cha S. Distinguishing recurrent intra-axial metastatic tumor from
nance spectroscopy in clinical decision making for patients with radiation necrosis following gamma knife radiosurgery using
suspected malignant brain tumors. J Neurooncol. 1999;45(1): dynamic susceptibility-weighted contrast-enhanced perfusion MR
69–81. imaging. AJNR Am J Neuroradiol. 2009;30(2):367–72.
64. Dowling C, Bollen AW, Noworolski SM, McDermott MW, 81. Cha S, Johnson G, Wadghiri YZ, Jin O, Babb J, Zagzag D,
Barbaro NM, Day MR, Henry RG, Chang SM, Dillon WP, Nelson Turnbull DH. Dynamic, contrast-enhanced perfusion MRI in
SJ, Vigneron DB. Preoperative proton MR spectroscopic imaging mouse gliomas: correlation with histopathology. Magn Reson
of brain tumors: correlation with histopathologic analysis of Med. 2003;49(5):848–55.
resection specimens. AJNR Am J Neuroradiol. 2001;22(4): 82. Cha S. Dynamic susceptibility-weighted contrast-enhanced perfu-
604–12. sion MR imaging in pediatric patients. Neuroimaging Clin N Am.
65. Hall WA, Martin A, Liu H, Truwit CL. Improving diagnostic yield 2006;16(1):137–47, ix.
in brain biopsy: coupling spectroscopic targeting with real-time 83. Cha S, Lupo JM, Chen MH, Lamborn KR, McDermott MW,
needle placement. J Magn Reson Imaging. 2001;13(1):12–5. Berger MS, Nelson SJ, Dillon WP. Differentiation of glioblastoma
66. Tsui EY, Chan JH, Leung TW, Yuen MK, Cheung YK, Luk SH, multiforme and single brain metastasis by peak height and per-
Tung SY. Radionecrosis of the temporal lobe: dynamic suscepti- centage of signal intensity recovery derived from dynamic
bility contrast MRI. Neuroradiology. 2000;42(2):149–52. susceptibility-weighted contrast-enhanced perfusion MR imag-
67. Asao C, Korogi Y, Kitajima M, Hirai T, Baba Y, Makino K, Kochi ing. AJNR Am J Neuroradiol. 2007;28(6):1078–84.
M, Morishita S, Yamashita Y. Diffusion-weighted imaging of 84. Lupo JM, Cha S, Chang SM, Nelson SJ. Dynamic susceptibility-
radiation-induced brain injury for differentiation from tumor weighted perfusion imaging of high-grade gliomas: characteriza-
recurrence. AJNR Am J Neuroradiol. 2005;26(6):1455–60. tion of spatial heterogeneity. AJNR Am J Neuroradiol.
68. Nowinski WL, Prakash B, Volkau I, Ananthasubramaniam A, 2005;26(6):1446–54.
Beauchamp NJJ. Rapid and automatic calculation of the 85. Essig M, Waschkies M, Wenz F, Debus J, Hentrich HR, Knopp
midsagittal plane in magnetic resonance diffusion and perfusion MV. Assessment of brain metastases with dynamic susceptibility-
images. Acad Radiol. 2006;13(5):652–63. weighted contrast-enhanced MR imaging: initial results. Radiology.
69. Park SH, Chang KH, Song IC, Kim YJ, Kim SH, Han 2003;228(1):193–9.
MH. Diffusion-weighted MRI in cystic or necrotic intracranial 86. Sugahara T, Korogi Y, Kochi M, Ikushima I, Hirai T, Okuda T,
lesions. Neuroradiology. 2000;42(10):716–21. Shigematsu Y, Liang L, Ge Y, Ushio Y, Takahashi M. Correlation
70. Lam WW, Poon WS, Metreweli C. Diffusion MR imaging in gli- of MR imaging-determined cerebral blood volume maps with his-
oma: does it have any role in the pre-operation determination of tologic and angiographic determination of vascularity of gliomas.
grading of glioma? Clin Radiol. 2002;57(3):219–25. AJR Am J Roentgenol. 1998;171(6):1479–86.
71. Kono K, Inoue Y, Nakayama K, Shakudo M, Morino M, Ohata K, 87. Lev MH, Ozsunar Y, Henson JW, Rasheed AA, Barest GD, Harsh
Wakasa K, Yamada R. The role of diffusion-weighted imaging in GRT, Fitzek MM, Chiocca EA, Rabinov JD, Csavoy AN, Rosen
patients with brain tumors. AJNR Am J Neuroradiol. 2001;22(6): BR, Hochberg FH, Schaefer PW, Gonzalez RG. Glial tumor grad-
1081–8. ing and outcome prediction using dynamic spin-echo MR suscep-
72. Tsui EY, Chan JH, Ramsey RG, Leung TW, Cheung YK, Luk SH, tibility mapping compared with conventional contrast-enhanced
Lai KF, Wong KP, Fong D, Yuen MK. Late temporal lobe necrosis MR: confounding effect of elevated RCBV of oligodendrogliomas
in patients with nasopharyngeal carcinoma: evaluation with com- [corrected]. AJNR Am J Neuroradiol. 2004;25(2):214–21.
bined multi-section diffusion weighted and perfusion weighted 88. Aronen HJ, Gazit IE, Louis DN, Buchbinder BR, Pardo FS,
MR imaging. Eur J Radiol. 2001;39(3):133–8. Weisskoff RM, Harsh GR, Cosgrove GR, Halpern EF, Hochberg
73. Chapman CH, Nagesh V, Sundgren PC, Buchtel H, Chenevert TL, FH, et al. Cerebral blood volume maps of gliomas: comparison
Junck L, Lawrence TS, Tsien CI, Cao Y. Diffusion tensor imaging with tumor grade and histologic findings. Radiology. 1994;191(1):
of normal-appearing white matter as biomarker for radiation- 41–51.
induced late delayed cognitive decline. Int J Radiat Oncol Biol 89. Young IR, Cox IJ, Coutts GA, Bydder GM. Some considerations
Phys. 2012;82(5):2033–40. concerning susceptibility, longitudinal relaxation time constants
and motion artifacts in vivo human spectroscopy. NMR Biomed. 108. Schwartz RB, Carvalho PA, Alexander ER, Loeffler JS, Folkerth
1989;2(5–6):329–39. R, Holman BL. Radiation necrosis vs high-grade recurrent gli-
90. Langleben DD, Segall GM. PET in differentiation of recurrent oma: differentiation by using dual-isotope SPECT with 201ti and
brain tumor from radiation injury. J Nucl Med. 2000;41(11): 99mtc-hmpao. AJNR Am J Neuroradiol. 1991;12(6):1187–92.
1861–7. 109. Kline JL, Noto RB, Glantz M. Single-photon emission Ct in the
91. Hustinx R, Smith RJ, Benard F, Bhatnagar A, Alavi A. Can the evaluation of recurrent brain tumor in patients treated with gamma
standardized uptake value characterize primary brain tumors on knife radiosurgery or conventional radiation therapy. AJNR Am J
FDG-PET? Eur J Nucl Med. 1999;26(11):1501–9. Neuroradiol. 1996;17(9):1681–6.
92. Spence AM, Mankoff DA, Muzi M. Positron emission tomography 110. Moody EB, Hodes JE, Walsh JW, Thornsberry S. Thallium-avid
imaging of brain tumors. Neuroimaging Clin N Am. 2003;13(4): cerebral radiation necrosis. Clin Nucl Med. 1994;19(7):611–3.
717–39. 111. Yoshii Y, Moritake T, Suzuki K, Fujita K, Nose T, Satou M.
93. Wong TZ, van der Westhuizen GJ, Coleman RE. Positron emis- Cerebral radiation necrosis with accumulation of thallium 201 on
sion tomography imaging of brain tumors. Neuroimaging Clin N single-photon emission Ct. AJNR Am J Neuroradiol. 1996;17(9):
Am. 2002;12(4):615–26. 1773–6.
94. Henze M, Mohammed A, Schlemmer HP, Herfarth KK, Hoffner 112. de Vries B, Taphoorn MJ, van Isselt JW, Terhaard CH, Jansen GH,
S, Haufe S, Mier W, Eisenhut M, Debus J, Haberkorn U. PET and Elsenburg PH. Bilateral temporal lobe necrosis after radiotherapy:
SPECT for detection of tumor progression in irradiated low-grade confounding SPECT results. Neurology. 1998;51(4):1183–4.
astrocytoma: a receiver-operating-characteristic analysis. J Nucl 113. Alexiou GA, Fotopoulos AD, Papadopoulos A, Kyritsis AP,
Med. 2004;45(4):579–86. Polyzoidis KS, Tsiouris S. Evaluation of brain tumor recurrence
95. Hustinx R, Pourdehnad M, Kaschten B, Alavi A. PET imaging for By (99m)tc-tetrofosmin SPECT: a prospective pilot study. Ann
differentiating recurrent brain tumor from radiation necrosis. Nucl Med. 2007;21(5):293–8.
Radiol Clin North Am. 2005;43(1):35–47. 114. Alexiou GA, Tsiouris S, Kyritsis AP, Polyzoidis KS, Fotopoulos
96. Ricci PE, Karis JP, Heiserman JE, Fram EK, Bice AN, Drayer AD. Brain SPECT By 99mtc-tetrofosmin for the differentiation of
BP. Differentiating recurrent tumor from radiation necrosis: time tumor recurrence from radiation injury. J Nucl Med.
for re-evaluation of positron emission tomography? AJNR Am J 2008;49(10):1733–4; author reply 1734.
Neuroradiol. 1998;19(3):407–13. 115. Delattre JY, Rosenblum MK, Thaler HT, Mandell L, Shapiro WR,
97. Chao ST, Suh JH, Raja S, Lee SY, Barnett G. The sensitivity and Posner JB. A model of radiation myelopathy in the rat. Pathology,
specificity of FDG PET in distinguishing recurrent brain tumor regional capillary permeability changes and treatment with dexa-
from radionecrosis in patients treated with stereotactic radiosur- methasone. Brain. 1988;111(Pt 6):1319–36.
gery. Int J Cancer. 2001;96(3):191–7. 116. Tada E, Matsumoto K, Kinoshita K, Furuta T, Ohmoto T. The pro-
98. Wang SX, Boethius J, Ericson K. FDG-PET on irradiated brain tective effect of dexamethasone against radiation damage induced
tumor: ten years’ summary. Acta Radiol. 2006;47(1):85–90. by interstitial irradiation in normal monkey brain. Neurosurgery.
99. Ishiwata K, Kubota K, Murakami M, Kubota R, Sasaki T, Ishii S, 1997;41(1):209–17. discussion 217–9.
Senda M. Re-evaluation of amino acid PET studies: can the pro- 117. Martins AN, Severance RE, Henry JM, Doyle TF. Experimental
tein synthesis rates in brain and tumor tissues be measured in vivo? delayed radiation necrosis of the brain. Part 1: effect of early dexa-
J Nucl Med. 1993;34(11):1936–43. methasone treatment. J Neurosurg. 1979;51(5):587–96.
100. Jager PL, Vaalburg W, Pruim J, de Vries EG, Langen KJ, Piers 118. Lee AW, Ng SH, Ho JH, Tse VK, Poon YF, Tse CC, Au GK, O SK,
DA. Radiolabeled amino acids: basic aspects and clinical applica- Lau WH, Foo WW. Clinical diagnosis of late temporal lobe necro-
tions in oncology. J Nucl Med. 2001;42(3):432–45. sis following radiation therapy for nasopharyngeal carcinoma.
101. Isselbacher KJ. Sugar and amino acid transport by cells in cul- Cancer. 1988;61(8):1535–42.
ture—differences between normal and malignant cells. N Engl J 119. Pasquier D, Hoelscher T, Schmutz J, Dische S, Mathieu D,
Med. 1972;286(17):929–33. Baumann M, Lartigau E. Hyperbaric oxygen therapy in the treat-
102. Busch H, Davis JR, Honig GR, Anderson DC, Nair PV, Nyhan ment of radio-induced lesions in normal tissues: a literature
WL. The uptake of a variety of amino acids into nuclear proteins review. Radiother Oncol. 2004;72(1):1–13.
of tumors and other tissues. Cancer Res. 1959;19:1030–9. 120. Lorenzo ND, Nolletti A, Palma L. Late cerebral radionecrosis.
103. Herholz K, Holzer T, Bauer B, Schroder R, Voges J, Ernestus RI, Surg Neurol. 1978;10(5):281–90.
Mendoza G, Weber-Luxenburger G, Lottgen J, Thiel A, Wienhard 121. Marx RE, Ehler WJ, Tayapongsak P, Pierce LW. Relationship of
K, Heiss WD. 11c-methionine PET for differential diagnosis of oxygen dose to angiogenesis induction in irradiated tissue. Am J
low-grade gliomas. Neurology. 1998;50(5):1316–22. Surg. 1990;160(5):519–24.
104. Laverman P, Boerman OC, Corstens FH, Oyen WJ. Fluorinated 122. Hulshof MCCM, Stark NM, van der Kleij A, Sminia P, Smeding
amino acids for tumour imaging with positron emission tomogra- HMM, Gonzalez DG. Hyperbaric oxygen therapy for cognitive
phy. Eur J Nucl Med Mol Imaging. 2002;29(5):681–90. disorders after irradiation of the brain. Strahlenther Onkol.
105. Spaeth N, Wyss MT, Weber B, Scheidegger S, Lutz A, Verwey J, 2002;178(4):192–8.
Radovanovic I, Pahnke J, Wild D, Westera G, Weishaupt D, 123. Leber KA, Eder HG, Kovac H, Anegg U, Pendl G. Treatment of
Hermann DM, Kaser-Hotz B, Aguzzi A, Buck A. Uptake of cerebral radionecrosis by hyperbaric oxygen therapy. Stereotact
18f-fluorocholine, 18f-fluoroethyl-l-tyrosine, and 18f-fdg in acute Funct Neurosurg. 1998;70 Suppl 1:229–36.
cerebral radiation injury in the rat: implications for separation of 124. Glantz MJ, Burger PC, Friedman AH, Radtke RA, Massey EW,
radiation necrosis from tumor recurrence. J Nucl Med. Schold SCJ. Treatment of radiation-induced nervous system injury
2004;45(11):1931–8. with heparin and warfarin. Neurology. 1994;44(11):2020–7.
106. Weber WA, Wester HJ, Grosu AL, Herz M, Dzewas B, Feldmann 125. Delanian S, Porcher R, Balla-Mekias S, Lefaix JL. Randomized,
HJ, Molls M, Stocklin G, Schwaiger M. O-(2-[18f]fluoroethyl)-l- placebo-controlled trial of combined pentoxifylline and tocoph-
tyrosine and L-[methyl-11c]methionine uptake in brain tumours: erol for regression of superficial radiation-induced fibrosis. J Clin
initial results of a comparative study. Eur J Nucl Med. Oncol. 2003;21(13):2545–50.
2000;27(5):542–9. 126. Okunieff P, Augustine E, Hicks JE, Cornelison TL, Altemus RM,
107. Schillaci O, Filippi L, Manni C, Santoni R. Single-photon emis- Naydich BG, Ding I, Huser AK, Abraham EH, Smith JJ, Coleman
sion computed tomography/computed tomography in brain N, Gerber LH. Pentoxifylline in the treatment of radiation-induced
tumors. Semin Nucl Med. 2007;37(1):34–47. fibrosis. J Clin Oncol. 2004;22(11):2207–13.
127. Delanian S, Lefaix JL. Current management for late normal tissue 147. Pollock BE, Stafford SL, Link MJ. Gamma knife radiosurgery for
injury: radiation-induced fibrosis and necrosis. Semin Radiat skull base meningiomas. Neurosurg Clin N Am. 2000;11(4):
Oncol. 2007;17(2):99–107. 659–66.
128. Williamson R, Kondziolka D, Kanaan H, Lunsford LD, Flickinger 148. Singh VP, Kansal S, Vaishya S, Julka PK, Mehta VS. Early com-
JC.Adverse radiation effects after radiosurgery may benefit from plications following gamma knife radiosurgery for intracranial
oral vitamin e and pentoxifylline therapy: a Pilot Study. Stereotact meningiomas. J Neurosurg. 2000;93:57–61.
Funct Neurosurg. 2008;86(6):359–66. 149. Kobayashi T, Kida Y, Mori Y. Long-term results of stereotactic
129. Gonzalez J, Kumar AJ, Conrad CA, Levin VA. Effect of bevaci- gamma radiosurgery of meningiomas. Surg Neurol. 2001;55(6):
zumab on radiation necrosis of the brain. Int J Radiat Oncol Biol 325–31.
Phys. 2007;67(2):323–6. 150. Farooq O, Lincoff NS, Saikali N, Prasad D, Miletich RS, Mechtler
130. Torcuator R, Zuniga R, Mohan YS, Rock J, Doyle T, Anderson J, LL. Novel treatment for radiation optic neuropathy with intrave-
Gutierrez J, Ryu S, Jain R, Rosenblum M, Mikkelsen T. Initial nous bevacizumab. J Neuroophthalmol. 2012;32(4):321–4.
experience with bevacizumab treatment for biopsy confirmed 151. Stafford SL, Pollock BE, Leavitt JA, Foote RL, Brown PD, Link
cerebral radiation necrosis. J Neurooncol. 2009;94(1):63–8. MJ, Gorman DA, Schomberg PJ. A study on the radiation toler-
131. Wong ET, Huberman M, Lu XQ, Mahadevan A. Bevacizumab ance of the optic nerves and chiasm after stereotactic radiosurgery.
reverses cerebral radiation necrosis. J Clin Oncol. 2008;26(34): Int J Radiat Oncol Biol Phys. 2003;55(5):1177.
5649–50. 152. Mayo C, Martel MK, Marks LB, Flickinger J, Nam J, Kirkpatrick
132. Liu AK, Macy ME, Foreman NK. Bevacizumab as therapy for J. Radiation dose–volume effects of optic nerves and chiasm. Int J
radiation necrosis in four children with pontine gliomas. Int J Radiat Oncol Biol Phys. 2010;76(3):S28–35.
Radiat Oncol Biol Phys. 2009;75(4):1148–54. 153. Kondziolka D, Lunsford LD, Coffey RJ, Flickinger JC. Stereotactic
133. Mehta MP, Rozental JM, Levin AB, Mackie TR, Kubsad SS, radiosurgery of meningiomas. J Neurosurg. 1991;74(4):552–9.
Gehring MA, Kinsella TJ. Defining the role of radiosurgery in the 154. Hudgins WR, Barker JL, Schwartz DE, Nichols TD. Gamma knife
management of brain metastases. Int J Radiat Oncol Biol Phys. treatment of 100 consecutive meningiomas. Stereotact Funct
1992;24(4):619–25. Neurosurg. 1996;66 Suppl 1:121–8.
134. Satzger I, Degen A, Asper H, Kapp A, Hauschild A, Gutzmer 155. Chang JH, Chang JW, Choi JY, Park YG, Chung SS. Complications
R. Serious skin toxicity with the combination of BRAF inhibitors after gamma knife radiosurgery for benign meningiomas. J Neurol
and radiotherapy. J Clin Oncol. 2013;31(13):e220–2. Neurosurg Psychiatry. 2003;74(2):226–30.
135. Adler JR, Cox RS, Kaplan I, Martin DP. Stereotactic radiosurgical 156. Hakim R, Alexander ER, Loeffler JS, Shrieve DC, Wen P, Fallon
treatment of brain metastases. J Neurosurg. 1992;76(3):444–9. MP, Stieg PE, Black PM. Results of linear accelerator-based
136. Hoffman S, Propp JM, McCarthy BJ. Temporal trends in inci- radiosurgery for intracranial meningiomas. Neurosurgery.
dence of primary brain tumors in the United States, 1985–1999. 1998;42(3):446–53; discussion 453–4.
Neuro Oncol. 2006;8(1):27–37. 157. Linskey ME, de Lunsford DL, Flickinger JC. Radiosurgery for
137. Claus EB, Bondy ML, Schildkraut JM, Wiemels JL, Wrensch M, acoustic neurinomas: early experience. Neurosurgery. 1990;26(5):736.
Black PM. Epidemiology of intracranial meningioma. 158. Linskey ME, Johnstone PAS. Radiation tolerance of normal tem-
Neurosurgery. 2005;57(6):1088–95. discussion 1088–95. poral bone structures: implications for gamma knife stereotactic
138. Lee JY, Kondziolka D, Flickinger JC, Lunsford LD. Radiosurgery radiosurgery. Int J Radiat Oncol Biol Phys. 2003;57(1):196–200.
for intracranial meningiomas. Prog Neurol Surg. 2007;20:142–9. 159. Werner-Wasik M, Rudoler S, Preston PE, Hauck WW, Downes
139. Lee JY, Niranjan A, McInerney J, Kondziolka D, Flickinger JC, BM, Leeper D, Andrews D, Corn BW, Curran WJ. Immediate side
Lunsford LD. Stereotactic radiosurgery providing long-term effects of stereotactic radiotherapy and radiosurgery. Int J Radiat
tumor control of cavernous sinus meningiomas. J Neurosurg. Oncol Biol Phys. 1999;43(2):299–304.
2002;97(1):65–72. 160. Miller RC, Foote RL, Coffey RJ, Sargent DJ, Gorman DA,
140. Muthukumar N, Kondziolka D, Lunsford LD, Flickinger JC. Schomberg PJ, Kline RW. Decrease in cranial nerve complica-
Stereotactic radiosurgery for anterior foramen magnum meningio- tions after radiosurgery for acoustic neuromas: a prospective study
mas. Surg Neurol. 1999;51(3):268. of dose and volume. Int J Radiat Oncol Biol Phys. 1999;43(2):
141. Chin LS, Szerlip NJ, Regine WF. Stereotactic radiosurgery for 305–11.
meningiomas. Neurosurg Focus. 2003;14(5):1–6. 161. Foote KD, Friedman WA, Buatti JM, Meeks SL, Bova FJ, Kubilis
142. Pollock BE, Stafford SL, Utter A, Giannini C, Schreiner SA. PS. Analysis of risk factors associated with radiosurgery for ves-
Stereotactic radiosurgery provides equivalent tumor control to tibular schwannoma. J Neurosurg. 2001;95(3):440–9.
Simpson Grade 1 resection for patients with small-to medium-size 162. Andrews DW, Suarez O, Goldman HW, Downes MB, Bednarz G,
meningiomas. Int J Radiat Oncol Biol Phys. 2003;55(4):1000–5. Corn BW, Werner-Wasik M, Rosenstock J, Curran Jr
143. Stafford SL, Pollock BE, Foote RL, Link MJ, Gorman DA, WJ. Stereotactic radiosurgery and fractionated stereotactic radio-
Schomberg PJ, Leavitt JA. Meningioma radiosurgery: tumor con- therapy for the treatment of acoustic schwannomas: comparative
trol, outcomes, and complications among 190 consecutive observations of 125 patients treated at one institution. Int J Radiat
patients. Neurosurgery. 2001;49(5):1029–38. Oncol Biol Phys. 2001;50(5):1265.
144. Coppa ND, Raper DM, Zhang Y, Collins BT, Harter KW, Gagnon 163. Rowe JG, Radatz M, Walton L, Kemeny AA. Stereotactic radio-
GJ, Collins SP, Jean WC. Treatment of malignant tumors of the surgery for type 2 neurofibromatosis acoustic neuromas: patient
skull base with multi-session radiosurgery. J Hematol Oncol. selection and tumour size. Stereotact Funct Neurosurg. 2003;79(2):
2009;2:16. 107–16.
145. Roche PH, Regis J, Dufour H, Fournier HD, Delsanti C, Pellet W, 164. Hasegawa T, Kida Y, Kobayashi T, Yoshimoto M, Mori Y, Yoshida
Grisoli F, Peragut JC. Gamma knife radiosurgery in the manage- J. Long-term outcomes in patients with vestibular schwannomas
ment of cavernous sinus meningiomas. J Neurosurg. 2000;93 treated using gamma knife surgery: 10-year follow up. J Neurosurg.
Suppl 3:68–73. 2005;102(1):10–6.
146. Nicolato A, Foroni R, Alessandrini F, Maluta S, Bricolo A, Gerosa 165. Koh E, Millar B, Ménard C, Michaels H, Heydarian M, Ladak S,
M. The role of gamma knife radiosurgery in the management of McKinnon S, Rutka JA, Guha A, Pond GR. Fractionated stereo-
cavernous sinus meningiomas. Int J Radiat Oncol Biol Phys. tactic radiotherapy for acoustic neuroma. Cancer. 2007;109(6):
2002;53(4):992–1000. 1203–10.
166. Kondziolka D, Lunsford LD, McLaughlin MR, Flickinger JC. 172. Lindvall P, Bergström P, Löfroth P-O, Hariz MI, Henriksson R,
Long-term outcomes after radiosurgery for acoustic neuromas. N Jonasson P, Bergenheim AT. Hypofractionated conformal stereo-
Engl J Med. 1998;339(20):1426–33. tactic radiotherapy for arteriovenous malformations. Neurosurgery.
167. Kondziolka D, Nathoo N, Flickinger JC, Niranjan A, Maitz AH, 2003;53(5):1036–43.
Lunsford LD. Long-term results after radiosurgery for benign 173. Veznedaroglu E, Andrews DW, Benitez RP, Downes MB, Werner-
intracranial tumors. Neurosurgery. 2003;53(4):815. Wasik M, Rosenstock J, Curran Jr WJ, Rosenwasser RH.
168. Kanzaki J, Tos M, Sanna M, Moffat DA. New and modified Fractionated stereotactic radiotherapy for the treatment of large
reporting systems from the consensus meeting on systems for arteriovenous malformations with or without previous partial
reporting results in vestibular schwannoma. Otol Neurotol. embolization. Neurosurgery. 2004;55(3):519.
2003;24(4):642–9. 174. Chang T-C, Shirato H, Aoyama H, Ushikoshi S, Kato N, Kuroda
169. Kaido T, Hoshida T, Uranishi R, Akita N, Kotani A, Nishi N, S, Ishikawa T, Houkin K, Iwasaki Y, Miyasaka K. Stereotactic
Sakaki T. Radiosurgery-induced brain tumor: case report. J irradiation for intracranial arteriovenous malformation using ste-
Neurosurg. 2001;95(4):710–3. reotactic radiosurgery or hypofractionated stereotactic radiother-
170. Levy EI, Niranjan A, Thompson TP, Scarrow AM, Kondziolka D, apy. Int J Radiat Oncol Biol Phys. 2004;60(3):861.
Flickinger JC, Lunsford LD. Radiosurgery for childhood intracranial 175. Sirin S, Kondziolka D, Niranjan A, Flickinger JC, Maitz AH,
arteriovenous malformations. Neurosurgery. 2000;47(4):834–42. Lunsford LD. Prospective staged volume radiosurgery for large
171. Flickinger JC, Kondziolka D, Lunsford LD, Pollock BE, arteriovenous malformations: indications and outcomes in other-
Yamamoto M, Gorman DA, Schomberg PJ, Sneed P, Larson D, wise untreatable patients. Neurosurgery. 2006;58(1):17.
Smith V. A multi-institutional analysis of complication outcomes 176. Pollock BE, Lunsford LD, Kondziolka D, Maitz A, Flickinger
after arteriovenous malformation radiosurgery. Int J Radiat Oncol JC. Patient outcomes after stereotactic radiosurgery for “operable”
Biol Phys. 1999;44(1):67–74. arteriovenous malformations. Neurosurgery. 1994;35(1):1–8.
Building a Radiosurgery Practice
66
Ali Farooqui and N. Scott Litofsky
radiosurgical procedures into their practice. They must

Introduction determine how to interface with the various disciplines
involved in order to be part of a cohesive working group.
One of the major trends in medicine over the last decade is They must let physicians and patients inside and outside the
the development of minimally invasive techniques for the community know about radiosurgery as an available treat-
performance of surgical procedures. Both physicians and ment option. This chapter discusses from a neurosurgery
patients prefer these techniques because they are typically perspective many of the issues that one should consider
easier to perform and result in less pain and a quicker recov- when building a radiosurgery practice from the ground up.
ery. Additionally, insurance companies also prefer these pro-
cedures because patients experience shorter hospital stays
due to the fast recovery period, which ultimately results in Current Utilization and Potential
fewer costs for insurance companies. In order to best develop Future Growth
minimally invasive procedures, the medical community has
had to address several key issues. These include the develop- The current neurosurgeon (and radiosurgeon) should be
ment of the necessary technologies to perform the proce- aware of the tremendous growth of the field since its incep-
dures, identification of appropriate patient profiles and tion. As potential new indications for treatment develop,
disease processes that lend themselves to these treatments, patients who would previously not have been treated become
and education of physicians—both those that treat the dis- candidates for radiosurgery. In addition, the growth of radio-
eases and those who refer patients for such treatments— surgery represents a paradigm shift in treatment, which
about the procedures. means that patients who previously have been treated by
In neurosurgery, one of the most prevalent minimally other means, such as open craniotomy, may now defer those
invasive techniques is stereotactic radiosurgery. Stereotactic treatments in favor of radiosurgery.
radiosurgery is a technique in which a single, high dose of Lars Leksell of the Karolinska Institute introduced the
ionizing radiation is focused on a three-dimensional target term stereotactic radiosurgery in 1951, and he treated his first
volume using stereotactic guidance. A variety of radiation patient with the Gamma Knife, a cobalt-60 source of ioniz-
sources, imaging technologies, and stereotactic localizing ing radiation, in Sweden in 1967. Dade Lunsford at University
devices have been developed to make this procedure read- of Pittsburgh introduced the Gamma Knife to the USA, treat-
ily available to patients throughout the world. Neurosurgeons ing his first patient in 1987 making the Pittsburgh unit the
beginning their own practice, joining an existing practice, fifth in the world. By 1999, 123 Gamma Knife units were in
or becoming a hospital employee must determine if they use worldwide [1], and by 2012, that number grew to 257
will perform radiosurgery and how they will incorporate [2]. Linear accelerator (LINAC) radiosurgery programs
developed later; the first unit was established in Barcia-
Solario, Spain, in 1982. Lutz and Winston started the first US
A. Farooqui, M.D. LINAC radiosurgery program in Boston in 1986, and its sub-
N.S. Litofsky, M.D., F.A.A.N.S., F.A.C.S. (*) sequent growth has exceeded that of the Gamma Knife. In
Division of Neurological Surgery, Department of Surgery, 1999, more than 200 LINAC programs existed in the USA
University of Missouri School of Medicine, One Hospital Drive,
alone [1]; today linear accelerators are now found in almost
DC048.00, MC326, Columbia, MO 65212, USA
e-mail: farooquia@health.missouri.edu; every major medical center in the USA [3]. A third form of
litofskyn@health.missouri.edu delivery of ionized radiation for radiosurgery—heavy
808 A. Farooqui and N.S. Litofsky
charged particles—has also been available since the 1950s As part of a current and continuing national trend to
(initially at Lawrence Livermore in Berkeley). A handful of develop centers for neuroscience care, radiosurgery pro-
programs are available in the USA and throughout the world grams are and will be a “hot” topic for the next several years.
(41 in 2012) [4], but this technology has been limited by the Radiosurgery is also an important component of neurologi-
cost of building a cyclotron to produce the protons with the cal surgery residency training. The Residency Review
final price of a proton therapy center currently on the order of Committee (RRC) for neurological surgery has expanded the
between 120 and 300 million dollars [5]. specialty definition of neurological surgery to include treat-
Clinical indications for treatment with stereotactic radio- ment of diseases of the brain and spinal cord and their cover-
surgery have expanded as availability has increased. The ings and blood supply with stereotactic radiosurgery. This
largest group of patients is those with metastatic tumors to expanded definition presupposes that neurosurgeons receive
the brain, for which several studies have shown radiosurgery clinical and didactic training in radiosurgery during resi-
to be more cost-effective than open surgery [6, 7]. Patients dency training. Currently, the RRC requires neurosurgery
with gliomas who have been treated with radiosurgery have residents to complete and participate in a minimum of ten
improved survival compared to others [8]. Benign tumors, adult cranial radiosurgery cases [13].
such as meningiomas, particularly those involving the skull Concerns have been raised about potential market satura-
base or dural sinuses are frequent targets, as are schwanno- tion as new programs are built, which can limit individual
mas and pituitary adenomas. Arteriovenous malformations program patient volumes [11]. Despite these concerns, the
can be effectively treated with radiosurgery. Additionally, role of stereotactic radiosurgery as a tool to treat diseases of
epilepsy, trigeminal neuralgia, pain, and movement disorders the nervous system is quite high and should remain so for the
are all treatable with radiosurgery [1, 9]. Although Leksell foreseeable future.
initially developed radiosurgery to create lesions for func-
tional neurosurgery, such use has become routine only in the
last few years. Radiosurgery in Residency Training
As mentioned above, exposure to radiosurgery is an integral

Future Growth of Radiosurgery part of residency training. The Residency Review Committee
does not specify how that training should be accomplished,
Current clinical indications create a theoretical patient base leaving those details to each individual program.
of about 171 patients per million in the USA who could be Each training program tends to tailor the radiosurgery
candidates for stereotactic radiosurgery (Table 66.1) [10]. experience for its residents to the unique circumstances of
The number actually treated is not nearly this high because the program. Didactic education may come in the form of
other forms of therapy are often utilized due to lesion loca- lectures, either as part of curriculum conference or as Grand
tion, size, or number. In actuality, according to data from the Rounds. Individualized reading will also likely be required.
Innovations Center, about 72,000 patients could be treated Textbooks, such as this one, are an excellent source to expand
with radiosurgery each year [11]. Because of increased cost- one’s breadth of knowledge. Recently, Kondziolka detailed
effectiveness, lower procedural morbidity, and patient the 100 most cited radiosurgery papers written; a number of
choice, radiosurgery is used increasingly as the first line these form the foundation of the principles of stereotactic
treatment for its indications. As such, craniotomy as a treat- radiosurgery outlined in texts, including indications, out-
ment is expected to slightly decline over the next decade, come, dose consideration, and complications [14]. Reading
with radiosurgery volumes increasing by almost double. For the source literature, while perhaps more labor intensive, can
instance, radiosurgery cases for neuro-oncologic indications help crystallize one’s thinking about the information offered.
are expected to increase from about 18,000 in 2004 to Recently, the University of Florida posted an online com-
32,000 in 2014 [11]. Furthermore, many patients previously puter module that was made available to all training pro-
thought to have inoperable lesions due to location or patient grams to enhance didactic radiosurgery education [15].
health are now candidates for stereotactic radiosurgery. Knowledge of the various radiosurgery systems in use is
Population aging and population growth will also increase helpful, as one does not always know what device will be
potential patient volumes. Lastly, as patient becomes more available in one’s practice setting.
informed about health-care options, they are more and more Radiosurgical experience in the form of surgical cases is
frequently requesting treatment with minimally invasive paramount to incorporating stereotactic radiosurgery one’s
techniques like radiosurgery [12]. Clearly, the demand for clinical practice. Although the RRC for neurologic surgery
stereotactic radiosurgery treatment will encourage the did mandate that each resident participate in a minimum of
involvement of new providers where availability is low to ten radiosurgery cases before the completion of residency
improve patient access. training [13], mechanisms by which the resident will obtain
66 Building a Radiosurgery Practice 809
Table 66.1 Population- Market Size

based market size for National
Incremental Total SRS
Annual Treatable with
diseases treatable by Lesion Treatable Incremental
Incidence per SRS
stereotactic radiosurgery Patients per Treatable
Million
Million Patients
Arteriovenous
19 70% $13 8
Malformation
Benign Neoplasms 20 40% $8 5
Neoplasms of
4 25% $1 1
Uncertain Behavior
Malignant
40 25% $10 6
Neoplams
Brain Metastases 425 25% $106 59
Meningioma 8 25% $2 2
Pituitary Adenoma 21 20% $4 3
Trigeminal Neuralgia 43 50% $21 12
Vestibular
9 70% $6 4
Schwannoma
Total 589 39% $171 100
From: Innovations Center, Health Care Advisory Board. Future of Neurosciences;
Strategic Forecast and Investment Blueprint. The Advisory Board Company 2008;
used with permission.
that case experience is not so certain. Almost always, radio- other instances, a neurosurgeon may have subspecialty inter-
surgery cases are performed as outpatient procedures and ests for which radiosurgery can be a treatment option for
frequently off-site from the hospital in which the residents patients. Given the current indications for radiosurgery, such
rotate. To solve this problem, some programs, including subspecialties include neuro-oncology, functional neurosur-
Johns Hopkins, University of Pittsburg, and University of gery (pain or movement disorders), epilepsy surgery, and
Missouri-Columbia, have established elective or selective cerebrovascular surgery. Neurosurgeons focused on these
rotations to ensure that residents receive adequate exposure. subspecialties may also chose to treat patients with other
At Johns Hopkins, for example, a 4-month block during the “non-subspecialty” disease processes with radiosurgery;
PGY-4 year is devoted to functional and stereotactic radio- while the targets and doses may be different, many of the
surgery. In other programs, the resident may need to make technical principles are similar.
some additional effort to be able to participate; understand- A general neurosurgeon in private practice or academic
ing that minimum case numbers are required for residency practice may also include radiosurgery as part of his/her clin-
completion may provide the necessary impetus. ical offerings. Clinical volume may be, but is not necessarily
less than in the previous circumstances, depending on one’s
local environment. For instance, in a medical community
Practice Development with a 500,000 catchment area, if four neurosurgeons are
each offering radiosurgery, one’s case volume may be less
Decision to Proceed than five cases per year; the neurosurgeon may decide that
for that volume, she/he would rather refer appropriate radio-
The first step in developing a radiosurgery practice is making surgery cases to another colleague.
the decision to include radiosurgery in one’s personal proce- The neurosurgeon should consider a number of factors in
dural offerings to patients. Not all neurosurgeons perform all making the decision of whether or not to include radiosur-
procedures. Even neurosurgeons in general practice have the gery as part of one’s practice. Perhaps the first factor is “what
option to chose the types of cases they wish to do and refer is the potential volume of patients available to treat?” By
other cases to other neurosurgeons, either within the same extrapolating data of potential case volume and population,
group, to another group, to local academic center, or to a perhaps one radiosurgery case could be performed for every
distant academic center. 4,000 people each year. Volume, however, is not usually this
If one is in an academic neurosurgical practice or a large high. One needs to assess the extent of out-migration from
group private practice, individuals in the group may have his/her catchment area and how many other practitioners are
subspecialty interests. Sometimes, a neurosurgeon can spe- doing radiosurgery locally, as referral patterns may be diffi-
cialize in radiosurgery only. Colleagues within the group, cult to break. Another factor is the proximity of radiosurgery
neurosurgeons and radiation oncologists who do not do facilities to one’s office or hospital(s). Many planning
radiosurgery, and other physicians will refer potential cases systems can permit remote planning, which makes distance
to that particular individual for evaluation and treatment. In less of an issue; in contrast, travel time is not usually being
$12,000
$10,910 $10,910
$10,543 $10,543
$9,671
$10,000 $9,261
$7,752
$8,000 $7,262
$6,000 2007
2008
$4,000
$2,000
$0
Gamma Knife LINAC CyberKnife Robotic LINAC
Fig. 66.1 Changes in CMS reimbursement for outpatient SRS by Medicare over 1 year. From: Innovations Center, Health Care Advisory Board.
Future of Neurosciences; Strategic Forecast and Investment Blueprint. The Advisory Board Company 2008; used with permission
Table 66.2 Changes in physician charges and RVUs at a single institution for various CPT codes
2008 2009 2010 2011 2012
CPT Code Lesion
Charge RVU Charge RVU Charge RVU Charge RVU Charge RVU
61793 1 or more $4,825 17.75 N/A N/A N/A N/A N/A N/A N/A N/A
61796 Under 3.5 N/A N/A $1,717 10.79 $2,100 13.93 $2,100 13.93 $2,100 13.93
61797 Additional N/A N/A $478 3.48 $478 3.48 $500 3.48 $500 3.48
61798 Over 3.5 N/A N/A $1,717 10.79 $1,717 19.85 $2,800 19.85 $2,800 19.85
61799 Additional N/A N/A $616 4.81 $661 4.81 $1,152 4.81 $1,152 4.81
63620 Spinal N/A N/A $1,717 10.79 $1,717 15.6 $1,717 15.6 $1,717 15.6
63621 Additional N/A N/A $549 4 $549 4 $556 4 $556 4
used for other productivity. The expectations of one’s [10]. Figure 66.1 shows the change in total radiosurgery
employer may be an issue. A hospital-employed neurosur- reimbursement per case by Medicare between 2007 and
geon or one joining an existing group (academic or private 2008. Table 66.2 illustrates the changes in physician billing
practice) may be expected to include radiosurgery to meet per case at the University of Missouri from 2008 to 2012,
service needs. The type of radiosurgery system(s) available is which shows the dynamic change in the billing for radiosur-
also important to consider. The device may not be the same gery. Over the span of one year the RVU for radiosurgery
as used during residency or other practice venues, and pur- dropped by seven units resulting in a significant decrease in
chase of a familiar device or system is not practical for an reimbursement. However, over the following years, the total
individual and unlikely for an institution unless its present charges and RVUs have steadily increased. In addition, more
system is ready for upgrade or replacement. Therefore, a CPT codes were added to allow for better characterization of
neurosurgeon may need to obtain additional training in the treatments being performed. Although there has been an
anticipation of including radiosurgery in practice. A vendor- overall decrease in the reimbursement of stereotactic radio-
sponsored course or a trip to a center with the particular surgery, the earning potential may be equivalent due to the
radiosurgery platform can go a long way to increasing com- unbundling of the CPT codes. The future of radiosurgery
fort with the existing system. reimbursement is difficult to predict, but the reality of the
Current and potential future levels of reimbursement are situation is that it will depend heavily on the political and
also an issue. The National Health Care Advisory Board had socioeconomic condition of the country and its effects on the
suggested a potential decrease in reimbursement in 2008 health-care system at large.
Time spent on radiosurgery means time not spent on other A few credentialing protocols have been described for
types of cases. Compensation for these other cases may neurosurgeons [16, 17]. Some centers may incorporate all, a
exceed reimbursement from radiosurgery; therefore, personal portion, or none of these protocols. Completing the require-
economics also enters the equation. ments for credentialing is usually not particularly arduous,
One should also keep in mind that radiosurgery is just one but should be anticipated. Most credentialing requirements
of several options for treating particular disorders. A physi- will probably include at least some of following features:
cian’s selection of radiosurgery versus open surgery for a • Neurosurgeons should complete a radiosurgery course,
procedure depends on the individual situation and patient, as ideally sponsored by the vendor of the software/hardware
one method may be preferred or superior to the other for being used at the Center. The course should include simu-
various reasons. However, offering radiosurgery increases lated cases. Most vendors provide such a course for phy-
the scope of one’s practice, thereby increasing procedure sicians as part of the purchase of their devices. The
volume and thus profits. For example, some patients have neurosurgeon may need to pay for the course; such costs
absolutely no interest in an open operation if a minimally can be negotiated as part of the contractual process for
invasive option is available, even if open surgery has some hospital or group-employed neurosurgeons.
clear advantages. By offering radiosurgery, one’s practice • Neurosurgeons should have experience in stereotactic
volume can benefit from those patients who may otherwise surgery. Such experience demonstrates competence and
choose to leave the area to receive radiosurgery elsewhere understanding of volumetric localization of targets in
when it is not locally available. Lastly, as indications for three-dimensional space and insures that Halo ring immo-
radiosurgery change, additional patient volumes may accrue. bilization, which many systems require, will be well
Two examples are worth mentioning. A patient with multiple accomplished. Cases providing such experience, in addi-
brain metastases, one of which is very large and symptom- tion to radiosurgery, include deep brain stimulation and
atic, may be a candidate for radiosurgery for the smaller stereotactic brain biopsy.
lesions after surgical resection of the large one. In another • The neurosurgeon should perform a number of radiosur-
case, a patient with a single metastasis may be a candidate gery cases under the direct supervision of an experienced
for radiosurgery to the postoperative tumor bed instead of neuro/radiosurgeon. This supervised experience may
whole-brain radiotherapy. In both circumstances, being able occur during residency training or as an attending
to provide radiosurgery generates additional case volume to neurosurgeon.
the neurosurgeon that, in the absence of radiosurgery capa- Documentation of the experience is important as well.
bilities, would not have been available. Case number may vary. Cleveland Clinic specified that either
the neurosurgeon or radiation oncologist involved in a case
has residency experience of 50–100 cases or attending expe-
Becoming Credentialed rience of 25–50 cases [17]. Initial supervision by a senior
colleague may also be necessary [16].
Institutions and patients both want to be certain that physi-
cians who care for patients are qualified to do so. Radiosurgery
is no different from any other treatment in that regard. For Building a Referral Base
neurosurgeons, radiosurgery is not as technically demanding
as other procedures (microvascular surgery, for instance), but Once a neurosurgeon decides to perform radiosurgery, she/
it does require a significant cognitive skill set. The neurosur- he should identify potential referring physicians. These phy-
geon needs to know which patients and lesions are appropri- sicians may be from within one’s hospital community or
ate for treatment, what structures are potentially in harm’s from the surrounding area. Again, consideration of the indi-
way during treatment of a particular patient, what is an cations for radiosurgery can help identify to which physi-
appropriate target volume, etc. Knowledge of radiation cians the neurosurgeon should reach out to explain the
dosimetry is helpful when interfacing with the radiation available services.
oncologist, who must frequently incorporate past radiation One key group of potential referring physicians should be
treatments into a radiosurgery prescription to achieve an radiation oncologists. Radiation oncology collaborates with
effective dose without undue long-term sequela. Therefore, neurosurgery for patient treatment in most centers. The radi-
neurosurgeons who provide radiosurgery need to be creden- ation oncologist is usually responsible for the radiation dose,
tialed. Such credentialing is particularly important if the as well as supervising the physics personnel involved with
radiosurgery center is open to multiple physicians from out- developing the treatment plan. Radiation oncology is often,
side the hospital/clinic in order for the center to maintain but not always, the responsible party for upkeep and quality
quality control. assurance of the devices used to generate and deliver the ion-
izing radiation. In some centers, a single radiation oncolo- gery. Additionally, direct phone calls from physicians per-
gist may have radiosurgery responsibilities; in others, those forming radiosurgery to referring physicians can be very
duties may be shared. Identification and dialogue with radia- effective. Finally, promptly communicating back to referring
tion oncologists outside the locale or who would not be physicians after assessing patients enhances the process.
participating in radiosurgery procedures is also helpful. Marketing to patients directly is also important. Because
Discussion with these radiation oncologists is appropriate to patients are using the Internet more often as their source of
define groups of patients for to treat with radiosurgery. medical information, physicians who do not utilize Web sites
Updating them on evolving radiosurgery indications, such as are at a distinct disadvantage as patients look for options for
radiosurgery to multiple lesions, radiosurgery to the tumor treatment of their ailments. Therefore, information on a Web
bed after resection, fractionated radiosurgery for large site should be written at a level that patients can compre-
lesions, and radiosurgery for spinal lesions, is appropriate hend. For example, reduction or simplification of medical
[18–21]. Sharing references can be helpful, as some indica- vocabulary is helpful to make the information more under-
tions represent a paradigm shift for radiation oncology, as standable. Television and radio spots, although more expen-
well as for neurosurgery. sive, can also be used to enhance one’s marketing to patients.
Although medical oncologists do not directly participate The physician can augment marketing also by partnering
in stereotactic radiosurgery treatment, their participation as with the institution housing the radiosurgery center. Two dif-
referring physicians is essential, since patients with meta- ferent strategies for marketing are available [11]. In the first,
static carcinoma represent the largest group of patients to marketing revolves around the radiation delivery device as
treat. Better patient referral can be expected when the medi- the key point of focus. The institution’s advertisements indi-
cal oncologists understand the issues related to stereotactic cate that it has the capability of treating patients with a
radiosurgery treatment. Gamma Knife, a Cyberknife, a Trilogy, or any other delivery
While radiation oncologists and medical oncologists will system that has been purchased for the program. This type of
likely account for the majority of referrals, other physicians product marketing appeals to patients by name-recognition
should not be discounted. Endocrinologists should become of a particular device. An alternative strategy focuses on the
aware of the potential for radiosurgery to treat pituitary particular disease process—like trigeminal neuralgia, or
tumors refractory to surgical resection. Otolaryngologists AVMs, or brain tumors—and notifies the public that a nonin-
often refer patients with vestibular schwannomas. vasive treatment modality is available for that disease. Both
Neurologists may be a source of patients with trigeminal strategies, especially if used together, can be very effective.
neuralgia, epilepsy, movement disorders, or arteriovenous Even though marketing is important to increase physician
malformations. Dentists and oral surgeons may also refer and patient awareness about radiosurgery as a tool to treat
patients with trigeminal neuralgia [11]. Because pediatric diseases, it is important to avoid exaggeration and misrepre-
patients can account for 10–15 % of patients in some centers sentation in information provided, particularly on the
[22], pediatricians and pediatric oncologists are also a poten- Internet, where there are no truth-in-advertising regulations,
tial source of referrals that can be cultivated. and where “data” can be promulgated directly to patients
without scientific peer review [23]. Brada and Cruickshank
[24] raised such concerns in an editorial in the British
Marketing to the Community Medical Journal. Unfortunately, these concerns were lost in
the authors’ own misrepresentations [25–28]. As centers and
Success in building a radiosurgery program can be related to physicians market their expertise and capabilities, they must
marketing to the medical community as well as to patients. provide accurate information to maintain their credibility.
Potential referral sources as described above need to know
that radiosurgery expertise is available. Helpful information
includes a list of the clinical indications for radiosurgery, Practical Practice Issues
inclusion and exclusion criteria, the sequence of events
involved in treatment of a patient, and how to refer patients Common Radiosurgery Systems
for treatment. A variety of means of communicating this
information are available. Presentations at medical staff In most circumstances, a neurosurgeon will utilize a radiosur-
meetings, mailed brochures or letters, Grand Rounds presen- gery system already in existence in a hospital or free-standing
tations, and conferences are all useful for marketing pur- clinic. While a detailed description of each potentially avail-
poses, particularly since they also provide appropriate able system is beyond the scope of this chapter, an under-
education about radiosurgery. The Internet has also increased standing of the potential pros and cons of each system in
physician and patient access to information about radiosur- terms of the neurosurgeon and patient interface with the sys-
tem is important. System cost, while important for the pur- immobilization with a frame; a few institutions across the
chaser, has less importance to the individual physician unless world, however, have begun to utilize the Extend program
that physician is involved with initial purchase in an advisory for the Gamma Knife that relies on a mouthpiece and a head-
capacity or as an investor. For purchase, factors such as rest, thereby eliminating the use of a head ring.
patient population base, institutional financial resources, and CyberKnife is a robotic LINAC equivalent used for
physician preference all play a role in determining which stereotactic radiosurgery. It is a relatively new delivery system,
device may be most appropriate. Generally speaking, while developed in 1997 at Stanford University, and approved by
each device has its advantages and disadvantages, the treat- the FDA in 2001 [33, 34]. It does not require rigid head
ment efficacy of each is relatively equivalent [29]. immobilization, and it can be used for intracranial and extra-
Proton beam systems are the least common systems avail- cranial indications, but only for stereotactic or IMRT indica-
able. The few that are present are usually associated with tions; it cannot be used for standard LINAC applications. For
academic medical centers. Their scarcity likely relates to the ease of the provider, it has the capability to accommodate
fact that proton beam systems are by far more expensive. multiple remote planning stations. Cost is about $3.5 mil-
Since proton beam systems are so uncommon, their issues lion. Annual hardware and software upgrades may run
will not be discussed further here. $225,000–$450,000; an upgraded vault is also required.
Gamma Knife, which has the longest track record for Most facilities will breakeven financially treating 109
radiosurgical treatment, is probably the best technology from patients per year [11]. Because Cyberknife is a small linear
a marketing perspective. As a radiosurgical tool, Gamma accelerator, treatment times are somewhat longer than those
Knife has the most widespread name recognition. When systems using a standard LINAC. Cyberknife also is devel-
patients inquire about radiosurgery, they usually ask, “Is that oping name recognition, which helps marketing to physi-
the Gamma Knife?” Many physicians also generically refer cians and patients. Furthermore, because a rigid head frame
to radiosurgery as “Gamma Knife.” Gamma Knife planning is not required for treatment, some patients specifically
times are fairly short, so more than one patient can be treated request treatment with Cyberknife, as opposed to Gamma
in one day. It has fewer moving parts than other systems, so Knife or other frame-based LINAC treatments.
maintenance and downtime is less; more patients per year A number of LINAC-based systems are available. Vendors
can theoretically be treated [30, 31]. Purchase of a Gamma include Varian (Trilogy), Integra, and Brainlab. Each system
Knife has allowed some programs to grow greatly, exceeding has somewhat different features, but all the systems have
expectations of programs’ directors [17]. For greater than remote planning capabilities. The Integra system requires a
200 radiosurgery patients per year, Gamma Knife has lower head ring for treatment purposes. A new system cost about
overall costs than a LINAC [31]. In a setting in which a large $2.5–3.2 million in the early 2000s. Currently, the cost of
treatment population is available, the Gamma Knife is a fre- maintenance and cost of a new system are somewhat guarded
quent choice because of its ease of operation, ability to pro- by the vendors and not open for publication due to various
vide conformal dose plans for a wide variety of target shapes financing options that each vendor allows. If used solely for
and sizes, and general familiarity of the device [32]. radiosurgery, breakeven patient volume for LINAC systems
Despite its advantages, Gamma Knife has several limita- has been 122 patients per year [10]. But because the LINAC
tions. Initial costs for Gamma Knife are higher than for can be used for other radiation indications, radiosurgery vol-
LINAC systems [30]; it costs approximately $3.2 million. ume does not need to be nearly that high. Retrofitting an
Also, because of the radiation decay half-life of 5.25 years, existing LINAC to perform radiosurgery with the usual cone
the treatment time doubles after 5 years. The cobalt sources collimators is probably the most inexpensive means of begin-
therefore must be replaced every 5 years at considerable ning a radiosurgery program. Software planning systems are
expense (estimated at $750,000). An upgraded radiation available for several hundred thousand dollars.
vault is also required, running about $700,000 when pur- One feature of the LINAC system is a mini-multileaf col-
chased new. Maintenance contracts, which all systems limator (MMLC) to enhance treatment options. MMLCs allow
require, are about $145,000 [11]. Another drawback of the shaping of X-ray beams to conform more precisely to the
Gamma Knife is that it can only be used for intracranial shape of the lesion in the beam’s eye view. They permit more
applications. Because breakeven volume is estimated at 86 homogenous radiation dosing to the lesion than can be
patients per year [11], small volume centers may have diffi- obtained with either the Gamma Knife or circular collimators
culty recouping their investment. Gamma Knife also requires using noncoplanar arcs [35]. Using an MMLC, each lesion
rigid head immobilization; some new technologies can be can be treated using a single isocenter, or target, which permits
used without application of a Halo ring (required by most faster treatment time. Frequently, an MMLC can be added to
Gamma Knife units), which improves patient comfort. Most an existing LINAC. MMLCs tend to be a bit bulkier than cir-
institutions utilizing Gamma Knife still require rigid head cular cone collimators, so physics staff and radiation therapists
must be aware of the possibility of collision between the work on the plan between other patient responsibilities and
LINAC and the patient/couch [36]. An MMLC and its accom- discuss the plan with the faculty as they review the work.
panying software can be added to an existing LINAC. Less flexibility is present with treatment space since treatment
Novalis, made by Brainlab, is an example of a standard obviously needs to occur where the radiosurgery device is
LINAC-based device. With its software and hardware addi- located.
tions, it can be used for intracranial and extracranial radio-
surgery. An additional feature is that it can also be used for
standard radiation therapy and IMRT. Therefore, it can func- Staff Responsibilities
tion as a backup linear accelerator. Novalis, like CyberKnife,
does not require rigid head immobilization for most of its Patients being treated with stereotactic radiosurgery experi-
radiosurgery indications; such immobilization is available ence a significant amount of anxiety about the procedure
for functional indications such as trigeminal neuralgia to [37]. Therefore, efforts to reduce patient anxiety are war-
improve accuracy. Novalis also includes an MMLC within ranted. Since less patient anxiety is associated with higher
its platform. Beams can be shaped in either static or arcing satisfaction, addressing anxiety can lead to practice growth.
fields to increase flexibility. Treatment times with Novalis Clear identification of staff responsibilities before, during,
(7 Gy/min) are faster than for CyberKnife (4 Gy/min) and after treatment can be helpful to this end.
because it is similar to a standard LINAC. For programs with Prior to planning and treatment, patients need education
small anticipated radiosurgery volumes, a LINAC-based about the procedure. The neurosurgeon and his/her staff need to
program may make the most sense [32]. provide a thorough explanation of the indications for the proce-
dure, its risks, and its alternatives, as with any surgical proce-
dure. Any required medications should be discussed. Plans for
Identification of Space follow-up after the procedure should be outlined. If a Halo ring
immobilization is planned, issues regarding treatment of pain
Radiosurgery is a fluid surgical procedure. Components of and sedation should be included in the preparation.
the procedure include patient assessment, treatment plan- The radiation oncologist and his/her staff should discuss
ning, and patient treatment. When building his/her radiosur- the issues related to positioning the patient during planning
gery practice, the neurosurgeon needs to consider where she/ and treatment. They need to assess the potential need for
he will perform these tasks. Consideration for efficient use of sedation or anxiolytics for the patient while immobilized on
time is essential. the gantry. The timeline for planning and treatment should be
Patient evaluation usually will occur during clinic, or clearly outlined. In addition, considerations of dosing in
office, time. One option to consider for a physician with a relation to surrounding critical structures, review of previous
heavy radiosurgery practice is to see patients in the radiation radiation history, and risk of side effects should be reviewed
facility. This option has a number of advantages. First, the in the context of these considerations.
neurosurgeon has easy access to his radiation oncology col- Nursing staff is an integral part for the success of the pro-
league to discuss cases. Second, the neurosurgeon can also gram [38]. It is best to designate a program coordinator who
do radiosurgery procedures while seeing patients, so that not only guides the patients’ through the procedure but also
she/he is readily available for the key portion of the proce- can assist in the procedures, prescribe medications, discharge
dure. The neurosurgeon can do his/her portion of treatment patients from services, and provide follow-up care. This indi-
planning in between patients and can readily review dosim- vidual ultimately allows the neurosurgeon and radiation
etry. Last, radiosurgery patients will have some familiarity oncologist to focus on planning and treatment, ultimately
with the facility prior to treatment and can see the radiation allowing for more efficient use of their time. While an
oncologist and neurosurgeon in close proximity in space and advanced practice nurse, or nurse practitioner (NP), is prob-
time, making “one-stop shopping” possible, and thereby ably the best option to serve as the program coordinator
improving patient satisfaction. because he/she can write prescriptions if needed, a clinical
Treatment planning can easily occur in the radiation nurse specialist can also effectively function in this role. The
dosimetry section. This location may not be convenient for coordinator should keep a checklist on each patient to be
the neurosurgeon to review plans if his/her office is not in sure that all pretreating planning and education has been
the same facility. One alternative is a remote planning sta- completed and that the patient is ready on the day of treat-
tion, available with many systems. A computer workstation ment. Nursing assistants, if available, can help with patient
can be set up in the neurosurgeon’s office where plans can transportation. Radiation oncology nursing staff are usually
be developed and reviewed whenever time permits. This involved with patient care when the patient is being treated
setup is particularly helpful for frameless treatments, when [38], particularly if the patient needs medication or assis-
planning can be accomplished days in advance of the treat- tance with using toilet facilities while wearing the head ring
ment. It is also ideal for resident training; the resident can in the department before being placed on the gurney.
Participating staff should have a general idea of the poten- outcome, and complications. It can also include routes of
tial benefits of radiosurgery to patients. Individuals who have referral. Regular review can reveal areas for quality
a better understanding of these treatment issues will feel more improvement for patient care and communication with
like a part of the team. They will be better able to provide referring physicians.
emotional and knowledgeable support to minimize patient
fear and anxiety [6]; they will also be less likely to make unin-
formed remarks that may send the wrong message to patients. Conclusions
Radiosurgery is an effective treatment option for many

Patient Follow-up patients; neurosurgeons should be able to provide and per-
form radiosurgical procedures. Preparation is the key to suc-
Follow-up visits after radiosurgery serve several purposes. cess. Identification of the issues involving radiosurgery and
Treatment response and post-procedural complications can effective marketing techniques that are outlined in this chap-
be assessed. Patient evaluation can guide further interven- ter are one way to ensure success in this evolving field. A
tion, change the treatment plan, and monitor or treat radia- thorough understanding of reimbursement and anticipation
tion side effects. The appropriate timing of follow-up visits of the changes in radiosurgery billing trends will allow a
or their necessity, however, has not been well defined. practitioner to maximally expand his/her practice.
Numerous reports have described symptoms associated with
SRS, including seizure, new neurological focal deficits, Acknowledgement This chapter is adapted, in part, from Litofsky NS,
headache, nausea/vomiting, vertigo, and death [39–42]. D’Agostino-Demers A: Building a radiosurgery program. In Chin LS,
Regine WF (ed): Priniciples and Practice of Stereotactic Radiosurgery.
However, studies guiding timing of clinical follow-up of
New York, Springer, 2007. Pp 691–698.
SRS patients based upon these symptoms are sparse.
New symptoms following SRS are common. Complication
rates range from 2.2 [39] to 64 % [41]. New symptoms in the
first 3 months after radiosurgery may occur in 24 [42]–35 % References
of patients (personal data, unpublished). The majority of
these symptoms are self-limited, usually requiring reassur- 1. Niranjan A, Lunsford LD. Radiosurgery: where we were, are, and
may be in the third millennium. Neurosurgery. 2000;46:531–43.
ance. Patients treated with a head ring for immobilization 2. Leksell Gamma Knife Treatment Centers. [Internet] 2012. http://
may require analgesia for mild headache. www.elekta.com/patients/treatment-information/treatment-centers/
Telephone-based follow-up in the first several weeks fol- leksell-gamma-knife.html 2012 Elektra AB.
lowing radiosurgery may be appropriate [43, 44]. Patient 3. Rahman M, Murad GJ, Bova F, Friedman WA, Mocco J. Stereotactic
radiosurgery and the linear accelerator: accelerating electrons in
interaction via telehealth is another means of evaluating neurosurgery. Neurosurg Focus. 2009;27(3):E13.
patients to avoid the costs in time and money for travel. 4. Particle Therapy Cooperative Group. Facilities in operation.
Video evaluation of patients near their home and electronic [Internet] 2012. Updated 2012 June 26. http://ptcog.web.psi.ch/
review of any imaging are possible. Medication can be pre- ptcentres.html
5. Nafziger B. Proton therapy: Can you afford it? DOTmed business
scribed and patients can receive appropriate physician news. 2011.
instruction and guidance. Patients generally experience vary- 6. Ott K. A comparison of craniotomy and gamma knife charges in a
ing rates of satisfaction with telehealth in other specialties, community-based gamma knife center. Stereotact Funct Neurosurg.
including surgical subspecialties and neurology [45–49]. 1996;66 Suppl 1:357–64.
7. Rutigliano MJ, Lunsford LD, Kondziolka D, Strauss MJ, Khanna V,
Telehealth evaluation has been shown to be more useful than Green M. The cost effectiveness of stereotactic radiosurgery versus
telephone sessions [48]. surgical resection in the treatment of solitary metastatic brain
Subsequent follow-up remains necessary to assess patients tumors. Neurosurgery. 1995;37:445–55.
for radiation-induced changes and/or tumor recurrence. Every 8. Loeffler JS, Alexander E, Shea WM, Wen PY, Fine HA, Kooy HM,
Black PM. Radiosurgery as part of the initial management of patients
3 months seems appropriate for malignant lesions, while every with malignant gliomas. J Clin Oncol. 1992;10(9):1379–85.
6–12 months makes sense for benign lesions. 9. Niranjan A, Jawahar A, Lunsford LD, Kondziolka D, Flickinger JC.
Radiosurgery: future directions and new frontiers. In: Germano
IM (ed). LINAC and Gamma Knife radiosurgery. USA:AANS,
2000:3–10.
Data Management 10. Innovations Center, Health Care Advisory Board. Future of neuro-
sciences; strategic forecast and investment blueprint. Washington,
Data management is essential in building a radiosurgery D.C.: The Advisory Board Company; 2008.
practice. A checklist on each patient helps ensure that 11. Innovations Center, Health Care Advisory Board. 2004–2005
National Member meeting: Future of neurosciences. Strategic fore-
the process is consistent and no gaps in the process occur. cast and investment blueprint. Washington, D.C.: The Advisory
A patient database can facilitate tracking of disease processes, Board Company; 2004.
12. Ward-Smith P. Stereotactic radiosurgery for malignant brain 31. Konigsmaier H, de Pauli-Ferch B, Hackl A, Pendi G. The costs of
tumors: the patients perspective. J Neurosci Nurs. 1997;29: radiosurgical treatment: comparison between gamma knife and
117–22. linear accelerator. Acta Neurochirurgica (Wein). 1998;140:
13. News RRC. Neurological surgery newsletter. April: ACGME; 1101–11.
2012. 32. Yu C, Jozsef G, Apuzzo ML, Petrovich Z. Dosimetric comparison
14. Kondziolka D. Citation measures in stereotactic radiosurgery: pub- of CyberKnife with other radiosurgical modalities for an elipsodal
lication across a discipline. Stereotact Funct Neurosurg. 2011;89(1): target. Neurosurgery. 2003;53:1163.
55–61. 33. Adler JR, Chang SD, Murphy MJ, Doty J, Geis P, Hancock SL. The
15. Radiosurgery course of the Neurosurgery department of the Cyberknife: a frameless robotic system for radiosurgery. Stereotact
University of Florida. [Internet] Updated 2012 Aug 10. http:// Funct Neurosurg. 1997;69:124–8.
online-course.neurosurgery.ufl.edu/ 34. Kuo JS, Yu PZ, Apuzzo ML. The CyberKnife stereotactic radiosur-
16. Groetsch SJ, Hardy T, Hodgens D, Lizarraras A, Ott K, Scharfen C, gery system: description, installation, and an initial evaluation of
Tung H. The open Gamma Knife center concept. Stereotact Funct use and functionality. Neurosurgery. 2003;53:1235–9.
Neurosurg. 1996;66 Suppl 1:296–301. 35. Shiu AS, Kooy HM, Ewton JR, Tung SS, Wong J, Antes K, Maor
17. Suh JH, Barnett GH, Miller DW, Crownover RL, Willoughby TR, MH. Comparison of miniature multileaf collimation (MMLC) with
Weinhous MS, Barrett PA, Walsh JG, Macklis RM. Successful con- circular collimation for stereotactic treatment. Int J Radiat Oncol
version from a linear accelerator-based program of a Gamma Knife Biol Phys. 1997;37:679–88.
radiosurgery program: the Cleveland Clinic experience. Stereotact 36. Urie MM, Lo YC, Litofsky S, Fitzgerald TJ. Miniature multileaf
Funct Neurosurg. 1999;72 Suppl 1:159–67. collimator as an alternative to traditional circular collimators for
18. Andrews DW, Scott CB, Sperduto PW, Flanders AE, Gaspar LE, stereotactic radiosurgery and stereotactic radiotherapy. Stereotact
Schell MC, Werner-Wasik M, Demas W, Ryu J, Bahary JP, Souhami Funct Neurosurg. 2001;76:47–62.
L, Rotman M, Mehta MP, Curran Jr WJ. Whole brain radiation with 37. Clifford W, Sharpe H, Khu KJ, Cusimano M, Knifed E, Bernstein
or without stereotactic radiosurgery boost for patients with one to M. Gamma knife patients’ experience: lessons learned from a quali-
three brain metastases: phase III results of the RTOG 9508 ran- tative study. J Neurooncol. 2009;92(3):387–92.
domised trial. Lancet. 2004;363:1665–72. 38. Browner CM, Hendrickson K. A nursing perspective of gamma
19. Jensen CA, Cahn MD, McCoy TP, Bourland JD, de Guzman AF, knife treatment. Barrow Neurological Institute Quarterly. 1997;
Ellis TL, Ekstrand KE, McMullen KP, Munley MT, Shaw EG, 13(1):1–6.
Urbanic JJ, Tatter SB. Cavity-directed radiosurgery as adjuvant 39. Chin LS, Lazio BE, Biggins T, Amin P. Acute complications fol-
therapy after resection of a brain metastasis. J Neurosurg. lowing gamma knife radiosurgery are rare. Surg Neurol. 2000;53(5):
2010;114(6):1585–91. doi:10.3171/2010.11 JNS10939. 498–502.
20. Roberage D, Souhami L. Tumor bed radiosurgery following resec- 40. St. George EJ, Kudhail J, Perks J, Plowman PN. Acute symptoms
tion of brain metastases: a review. Technol Cancer Res Treat. after gamma knife rasiosurgery. J Neurosurg. 2002;97:631–4.
2010;9(6):597–602. 41. Majhail N, Chander S, Mehta V, Julka P, Ganesh T, Rath G. Factors
21. Ryu S, Yin FF, Rock J, Zhu J, Chu A, Kagan E, Rogers L, Ajlouni influencing early complications following Gamma Knife radiosur-
M, Rosenblum M, Kim JH. Image-guided and intensity-modulated gery. A prospective trial. Stereotact Funct Neurosurg. 2001;76(1):
radiosurgery for patients with spinal metastasis. Cancer. 2003;97(8): 36–46.
2013–8. 42. Singh VP, Kansai S, Vaishya S, Julka PK, Mehta VS. Early compli-
22. Barnett GH. Evolution and organization of a regional Gamma cations following Gamma Knife radiosurgery for intracranial
Knife center. Stereotact Funct Neurosurg. 1996;66 Suppl 1:365–9. meningiomas. J Neurosurg. 2000;3:57–61.
23. Linskey ME. Stereotactic radiosurgery versus stereotactic radio- 43. Pal B. Following up outpatients by telephone: pilot study. Br Med
therapy for patients with vestibular schwannoma: a Leksell Gamma J. 1998;316(7145):1647.
Knife Society 2000, debate. J Neurosurg. 2000;93(Supple 3):90–5. 44. McVay MR, Karen KR, Mathews DL, Jackson RJ, Kokoska ER,
24. Brada M, Cruickshank G. Radiosurgery for brain tumors. Triumph Smith SD. Postoperative follow-up: is a phone call enough? J
of marketing over evidence based medicine [editorial]. Br Med Pediatr Surg. 2008;43(1):83–6.
J. 1999;318:411–2. 45. Sevean P, Dampier S, Spandoni M, Strickland S, Pilatzke
25. Adams CBT. Editorial was wrong to denigrate radiosurgery so S. Patients and families experiences with video telehealth in rural/
strongly [letter]. Br Med J. 1999;318:1489. remote communities in Northern Canada. J Clin Nurs. 2009;
26. Ganz JC. Radiosurgery for brain tumours. Not all practitioners of 18:2573–9.
this technique can have succumbed to marketing [letter]. Br Med 46. Abrams DJ, Geier MR. A comparison of patient satisfaction with
J. 1999;318:1490. telehealth and on-site consultations: a pilot study for prenatal
27. Ganz JC: Presentation based on extremely selective use of references genetic counseling. J Genet Couns. 2006;15:199–205.
[letter]. http://bmj.bmjjournals.com/cgi/eletters/318/7181/411. 47. Miller GG, Levesque K. Telehealth provides effective pediatric
28. Loeffler J, Lindquist C. Radiosurgery for brain tumors [letter]. Br surgery care to remote locations. J Pediatr Surg. 2002;37:
Med J. 1999;318:7181. 752–4.
29. Luxton G, Petrovich Z, Joszef G, Nedzi LA, Apuzzo 48. Fincher L, Ward C, Dawkins V, Magee V, Willson P. Using tele-
ML. Stereotactic radiosurgery: principles and comparison of treat- health to educate Parkison’s disease patients about complicated
ment methods. Neurosurgery. 1993;32:241–59. medication regimes. J Gerontol Nurs. 2009;35:16–24.
30. Stieber VW, Bourland JD, Tome WL, Meht MP. Gentleman (and 49. Heinzelmann PJ, Williams CM, Lugn NE, Kvedar JC. Clinical out-
ladies), choose your weapons: Gamma Knife vs. linear accelerator comes associated with telemedicine/telehealth. Telemed J E Health.
radiosurgery. Technol Cancer Res Treat. 2003;2:79–86. 2005;11:329–47.
Patient Care in Stereotactic
Radiosurgery 67
Terri F. Biggins
It is difficult to imagine what patient care must have

Introduction entailed in 1968. Nurses were limited in what role they
played and the duties they were allowed to perform. Short-
The intention of this chapter is to provide basic guidelines acting intravenous sedation and portable monitoring for
surrounding the patient process in relation to stereotactic patient travel were not readily available. Although patients
radiosurgery, with the goal of providing a seamless and today generally live longer and require much care, patients
pleasant experience for the patient and staff. Although the in the 1960s presented their own challenges. In 1969, Dr.
role of the registered nurse (RN) is highlighted, additional Leksell delivered stereotactic radiosurgery to many people
disciplines may gain insight into the patient perspective. suffering with severe pain. Parkinson disease was another
Expectantly, valuable information will be gained not only for disorder treated in those early years [3]. Maintaining an
those opening a new center but for the established site as acceptable comfort level while undergoing a lengthy pro-
well. Various methods of stereotactic radiosurgery have been cedure must have been trying. Unlike the frames in use
developed over the past several years, but essentially the today, the first patients treated by Dr. Leksell and his staff
mechanics are similar. This author’s experience is with the needed to endure the placement of a plaster helmet [3].
Leksell Gamma Knife, and therefore this will be the equip- The plaster had to dry before it could be used. All of the
ment referenced in this chapter. dose planning needed to be completed by drawing on the
pneumoencephalogram. A trip across town in an ambu-
lance was required in some cases in order to deliver the
Historical Perspective radiation [3]. One patient treatment could typically take
many hours to complete.
More than 50 years ago, a Swedish neurosurgeon pioneered Stereotactic radiosurgery recipients benefit greatly from
a method of treating brain disorders without opening the today’s advancements in technology. Radiographic imaging
skull. Approximately 17 years later in 1968, this same neu- in association with the use of computed tomography (CT)
rosurgeon, Dr. Lars Leksell, installed the first Gamma Knife and magnetic resonance imaging (MRI), as well as updated
using cobalt-60 at the Karolinska Institute in Stockholm. computerized treatment planning, has greatly reduced patient
This invention was considered the first stereotactic device. waiting time. Advancements in patient monitoring make the
Stereotactic radiosurgery delivers a single high dose of ion- delivery of sedation drugs safer. Access to education through
izing radiation to a radiographically well-defined, small home computers enhances a patient’s knowledge base.
intracranial target without delivering a significant portion of Today’s focus on pain control allows for easier monitoring
the prescribed dose to the surrounding brain tissue [1]. Some and control of discomfort. “Frameless” systems are begin-
of the first patients treated were afflicted with pain, move- ning to emerge, promising even greater tolerance of treat-
ment, or behavioral disorders [1]. The first Gamma Knife ment in the near future.
center in the United States opened in 1987 [2]. Clearly, the life expectancy in the United States and many
regions throughout the world has increased dramatically. As
the world’s population lives longer and technology advances,
T.F. Biggins, R.N., B.S.N., C.N.R.N. (*)
one looks for less invasive and debilitating healthcare treat-
Department of Nursing, University of Maryland Medical Systems,
22 S. Greene St, Baltimore, MD 21201, USA ments. Dr. Lars Leksell’s vision some 50 years ago has
e-mail: tbiggins@unm.edu helped achieve this very goal [4].
818 T.F. Biggins
new staff can be conducted in this medium. Documentation

Training and Staffing of minutes and attendance will ensure all staff have been
properly included in the quality assurance process.
The training required to perform radiosurgery can be very
specialized. In the case of Gamma Knife therapy, there are
less than 100 machines in the United States. Therefore, if an The Patient Flow Process
institution is planning to open a new radiosurgery practice,
training will likely occur at one of the existing sites in the Intake
United States. It is highly recommended that new staff attend
a course at an accredited center. Ultimately, each indepen- Patients will enter the radiosurgery program either by physi-
dent radiosurgery center with input from the state radiation cian or self-referral. In the age of the Internet, more indi-
safety department will set rules for professional staff train- viduals are educating themselves about healthcare options
ing. Certificates of education completion will be presented to in the comfort of their own homes in front of computers.
professional staff at the conclusion of training. Physician Eventually, a phone call will be made to set up consultation
staff must be credentialed before independently using the appointments.
radiosurgery tool. The nurse should have access to this list All the usual information regarding the patient’s demo-
and ensure its strict adherence. graphic data is collected at this time. Referring and primary
The registered nurse is a valuable member of the radiosur- care physician information is gathered as well. Past hospi-
gery team and must not be forgotten during the training talizations and locations of treatment records, particularly
phase. A background in various nursing specialties prepares pertaining to the diagnosis that is relevant to the radiosur-
one to adequately care for the individual undergoing stereo- gery treatment, should be gathered. Radiographic imaging
tactic radiosurgery. A background in neurosurgery, neurol- such as MRI, CT, and angiography prints as well as reports
ogy, radiation-oncology, medical oncology, and pain is an absolute must for the consultation process. A preprinted
management can be great assets. Certainly, nursing certifica- comprehensive intake form will ensure consistent data col-
tion in any one or more of the above specialties would be lection. A follow-up letter should be sent to each patient
advantageous. after this phone call. The letter should contain directions,
The volume of patients treated at each center will help appointment dates and times, and concise instructions on
determine the number of nurses and supporting staff needed what materials the individual is responsible for bringing to
to staff the center at any one time. If one nurse typically the appointment.
staffs the area, it is strongly recommended that at least two Many patients are eager to ask questions about radiosur-
other nurses are trained and ready to step in secondary to gery and what it entails during this initial intake call. It is
planned or unplanned absences or for an increase in patient never too early to begin the education process. Having a
acuity. These “back-up” nurses should complete an orienta- clinical person available to begin dialogue at the time of
tion program with the “permanent” nurse and reorient them- intake is invaluable. Inappropriate referrals can be redirected
selves by assisting during a patient treatment day every 6–8 at this time as well. The RN is an excellent resource for this
weeks (Table 67.1). Supporting staff such as nurse extenders role. The earlier the involvement of the nursing staff, particu-
should also receive a thorough orientation period. larly the staff who will be present at the time of treatment,
the better the patient satisfaction and outcome with the entire
experience.
Quality Assurance
Every radiosurgery center needs a quality assurance pro- Preprocedure Consultation

gram in place. All patients should be interviewed after
treatment, possibly during a follow-up phone call from the At the time of initial intake, the patient will be given
nurse. Questions regarding patient satisfaction and out- appointment(s) with all the team members necessary to
come expectations are essential. Quarterly meetings involv- properly evaluate the appropriateness of radiosurgery.
ing the entire professional staff should be held to discuss This may include consultations with neurosurgery,
the results of the patient interviews. Patient satisfaction goals radiation-oncology, neuroradiology, medical oncology,
can be set and constant reevaluation conducted to ensure and nursing. In that most patients who need to travel a
quality improvement. distance and also rely on others for transportation attempt
Meetings can also be utilized to discuss any additions or to combine as many appointments as possible into one
changes to current protocol. Radiation safety classes and visit. The neurosurgery-oncology multidisciplinary clinic
review of emergency patient procedures for all existing and is ideal for this type of referral. Prior to rendering a definite
Table 67.1 Sample orientation checklist for nursing staff
1. Introduction to facility Date: Initial:
a. Physical layout
b. Supply areas
c. Location of offices
d. Location of hospital and unit manuals
e. Medication stock and narcotics
f. ECG, crash cart, respiratory supplies
2. Flow of patients: introduction to areas
a. Admitting area
b. Minor surgery suite: frame placement
c. Radiology: MRI, CT, angiography suite
d. Gamma Knife: waiting area, treatment room, family waiting
e. Transfer to recovery suite
f. Discharge
3. Introduction to interdisciplinary staff
a. Physicians: neurosurgeons, radiation oncologist, physicist, medical director
b. Gamma Knife secretary
c. Radiology technicians and scheduling assistant
d. Neurosurgery anesthesiologist for pediatric cases
4. Equipment
a. G-frame, tray, and contents
b. Emergency equipment
c. Gamma Knife couch, video and audio monitoring
d. Physiologic monitoring
e. Fiducial indicators: MRI, CT, angiography, instructions for use
f. Filling of indicators
g. Assisting with “skull measurement”
5. Charting
a. Admission note
b. Special Procedure Nurses Note
c. Time-out signatures
d. Physician orders
e. Patient treatment log
f. Discharge documentation
g. Patient teaching: pre-and post-procedure
6. Policies, protocols, and procedures
a. Unit safety manual
b. Credentialing manual
c. Nursing procedure/policy manual (i.e., conscious sedation policy)
d. Radiation dosimetry badges and monthly reports
e. Routine patient care: adult
f. Patient care: pediatric
g. Radiation emergency procedures
7. Manage patient flow-through procedure
a. Readies necessary equipment in treatment and prep room
b. Greets patient, preps for procedure
c. Assist with frame placement
d. Assist with radiology procedures
e. Assist with skull measurements
f. Reunite family with patient
g. Provide ongoing education as well as post-procedure instructions
h. Administer medications (pre-procedure such as steroid, anticonvulsant, pain)
i. Assist with positioning during treatment
j. Monitor vital signs, chart treatment records in appropriate book
k. Assist with frame removal, apply dressings
l. Ready chart and transport patient to recovery
m. Clean, reassemble, and store G-frame, ready fixations pins for sterilization
n. Complete discharge instructions
Date and initial each item after completion during orientation
820 T.F. Biggins
treatment plan, the patient’s case should be presented at a weekly • Anesthesia consult (select patient’s receiving deep or
neurosurgery-oncology team conference for final approval. general anesthesia)
It is imperative that the patient records are gathered prior • ECG
to the appointment so as to expedite the scheduling of the • Serum results (electrolytes, complete blood count,
appropriate treatments. After consulting with the appropriate coagulation)
medical personnel, the patient should meet with a represen- • Pregnancy testing on women of child-bearing age
tative of the nursing staff. Viewing of a patient education • Anticonvulsant serum levels (Dilantin, Tegretol, pheno-
video, tours of the radiosurgery facility, and the answering of barbital, etc.)
questions can all help to ease anxiety. Appropriate consent • All consents complete (including angiography, MRI,
forms should be obtained at this time as well. The patient CT, etc.)
should leave at the end of this consultation period with a firm • Acquisition of imaging and reports necessary for
treatment plan in place. localization
Completing the above list will minimize the risk of post-
ponement, cancellation, or a negative outcome on the day of
Preauthorization and Collection the procedure.
of Preoperative Data
The preauthorization process should be started when the Patient Arrival and Preparation
patient makes the decision and is scheduled for the radiosur-
gery treatment. Designated personnel from the treating insti- The majority of radiosurgery centers require individuals to
tution contacts the patient’s insurance company. Information arrive early on the morning of the procedure. Patient registra-
to have handy at the time of contact may include: tion should be complete prior to arrival to minimize any
• Diagnosis added anxiety. Patients should first change into comfortable
• Procedure code clothing as outlined in pre-procedure instructions. Intravenous
• Outpatient versus inpatient visit access is obtained, as well as any last-minute blood work.
• Date of procedure Diabetic patients should be checked for baseline serum glu-
• Referring physician consultation note cose to determine any intervention. Vital signs are obtained,
• Number of lesions to be treated and a baseline neurologic assessment is completed by nursing
• List of treatments that have been completed related to the staff. Medications taken in the past 12 h are reviewed, and a
diagnosis final check of consent forms occurs.
Allow several days for preauthorization to be finalized.
Each institution has a policy regarding what preoperative
evaluations are required prior to procedures. The following Frame Application
list should be considered as a guide regarding data collection
and chart completeness: After all pertinent paperwork is completed and approxi-
• Preauthorization obtained from insurance company mately 60 min prior to frame placement, LMX-4 (Ferndale
• Preoperative check list completed by RN (Table 67.2) IP, Inc., Ferndale, MI) cream (lidocaine 4 %) should be
• History and physical complete applied to the forehead and to the back of the head in the
estimated region of insertion of the local anesthetic. Cover
the LMX-4 cream with a clear adhesive covering to keep
from dripping, particularly into the patient’s eyes (Fig. 67.1).
Table 67.2 Preoperative nursing check list Early application of LMX-4 cream will yield the best patient
• Review patient allergy history satisfaction during frame placement. Any oral sedation
• Assess any contraindication to imaging studies (contrast medications should be administered 30 min prior to frame
allergies, MRI metal concerns) placement.
• NPO status
After directing family members to the appropriate waiting
• Review of appropriate medications patient is to take prior to
procedure
area, the patient is moved to a procedure room for applica-
• What to bring to hospital (music, appropriate clothing, snack,
tion of the stereotactic frame. The room should contain mon-
medications) itoring capabilities (ECG, pulse oximetry, blood pressure),
• Assess anxiety level (help plan for proper sedation) as well as access to oxygen. The room lighting should be
• Describe flow of the procedure and when and where family adequate and room temperature comfortable for the patient.
members will be allowed to wait and visit Environmental stimuli should be kept to a minimum. When
• Review post-procedure course the treating physician is present, a “time out” should occur,
67 Patient Care in Stereotactic Radiosurgery 821
verifying correct patient, correct diagnosis, and appropriate As a rule, 2–5 mL of anesthetic is injected at each site to
site of treatment verification. achieve satisfactory comfort (Fig. 67.4). Provided there is
While sitting erect on a gurney the stereotactic frame is proper physician credentialing, the use of moderate sedation
suspended on the patient’s head by use of a Velcro strap such as IV Versed (1–2 mg) may be helpful for relieving pain
(Fig. 67.2). One may opt to use the magnetic resonance and anxiety during the procedure. There is also the added
(MR) indicator box during frame fitting to ensure a comfort- benefit of an amnestic effect after the frame has been placed.
able, accurate fit during imaging (Fig. 67.3). If conscious All individuals in the room need to be versed with emer-
sedation is to be used, it should be administered just prior to gency procedures in case of respiratory depression.
frame application with strict adherence to policy surround- The majority of patients tolerate frame application with
ing this competency. Anesthesia staff may be utilized to minor discomfort (Fig. 67.5). Pre-procedure education is a
administer deeper sedation for appropriate individuals. key component. Describing what to anticipate in regard to
Before the fixation screws are advanced, local anesthetic is sensations such as pressure during tightening of fixation
injected. A small (25-gauge) needle and 10-mL syringe is screws will assist in managing anxiety. Allow individuals to
filled with lidocaine HCL 1 % and epinephrine 1:1,000,000. view photos of previous patients wearing the device, or coor-
Two milliliters of 8.4 % sodium bicarbonate is combined dinate patient-to-patient contact with previous radiosurgery
with 20 mL of the lidocaine solution to help decrease the candidates. Time required for frame application rarely
“burning” sensation commonly associated with lidocaine. requires longer than 15 min.
Skull measurements will be conducted after the frame is
securely in place (Fig. 67.6). The patient should be sitting
erect with support behind the back allowing for easy access
to the head. There is absolutely no discomfort associated
with this process.
Radiographic Imaging
After the frame is in place, the patient is moved via gurney to

the appropriate imaging machine(s). Individuals will undergo
MRI, CT, and/or angiography imaging (Fig. 67.7). Each of
these machines requires use of individualized fiducial local-
ization equipment. Prior screening regarding allergies, metal
risk factors, and claustrophobia will expedite safe, efficient
imaging. Care is taken to place the appropriate fiducial box
securely onto the frame seating all connections and to ensure
Fig. 67.1 Apply LMX-4 cream at least 1 h prior to frame placement the side clips are tightly fastened (Fig. 67.8). Providing the
Fig. 67.2 (a) Apply Velcro head holder to frame to create a sling. (b) Sit frame with holder easily onto the head
822 T.F. Biggins
patient with comfort measures such as warm blankets and

knee support makes this a more pleasant experience. Pressure
points, particularly at the shoulders, should be properly pad-
ded to reduce risk of impaired skin integrity. Administration
of conscious sedation is encouraged in the claustrophobic
individual. The MR fiducial should be filled with copper sul-
fate 0.25 L solution (Fig. 67.9). Air bubbles that collect
between the plastic panels could make image acquisition
impossible and may require a second trip to the imaging cen-
ter and a delay in the treatment. The imaging fiducials are
very fragile and costly and should be carefully stored and
cared for by a designee of the radiosurgery team.
Educate the patient regarding length of imaging and what
sensations may be experienced. The patient requiring angi-
ography will need explicit instructions specifically concern-
Fig. 67.3 Checking fit of indicator box during frame placement ing post-procedure monitoring and care. Radiology staff also
Fig. 67.4 Neurosurgeon injects lidocaine Fig. 67.6 The taking of skull measurements prior to treatment planning
Fig. 67.5 Patient immediately after completion of frame placement Fig. 67.7 Patient within the head coil awaiting start of MRI
need to be educated about quick removal of the stereotactic the fiducial in that sound is magnified via the frame attached
equipment subsequent to a medical emergency. to the skull. After imaging, the patient is transported to the
Before imaging is complete and prior to patient removal stereotactic center to await the treatment.
from the table, the radiation physicist should ascertain that
the images have been completed correctly and that the trans-
fer to the planning computer can be successfully achieved. Treatment Planning
Neuroradiology should be monitoring and reading all images
prior to patient treatment. Carefully remove the patient from Treatment planning times vary according to the complexity
the imaging table. Great care is taken when fitting or removing of the case. Typically, the patient will be waiting 1–2 h before
treatment begins. Allowing family to visit during this time
promotes a decrease in patient anxiety. Providing a light
snack is appropriate if conscious sedation is not expected
during the treatment phase. Ongoing monitoring of vital
signs, neurologic assessment, and the reinforcement of edu-
cation is completed at this time. Assess and treat pain prior to
the initiation of treatment.
The Stereotactic Treatment
Comfortable positioning during the treatment phase can be

achieved by using a combination of pillows and blankets to
provide support. The supine position is typically used for
treatment so elevating the knees will support the lower back.
Shoulder and head support may be used as long as it does not
interfere with machine operation. Meet individual patient
Fig. 67.8 Check that MRI fiducial is without air bubbles and correctly needs as necessary. ECG/BP/oxygen monitoring should be
attached to frame prior to imaging used as appropriate. The availability of suction equipment
Fig. 67.9 (a) A spinal needle, syringe, and copper is gathered. (b) Extracting the sulfate. (c) Spinal needle is narrow and makes for easy filling of
fiducial
824 T.F. Biggins
Table 67.3 Treatment room supplies

• Oxygen set-up including ambu bag
• Suction set-up
• Crash cart
• Emergency patient removal equipment supplied by manufacturer
• Suture material/suture tray for uncontrolled bleeding from pin sites
• Pillows, blankets, etc., for patient support, comfort
• Gauge, Band-Aids, and all supplies for frame removal
• IV supplies, gloves
• Bedpans, urinals, emesis basins, tissues
• ECG, BP, oxygen monitoring
will help ensure patient safety. Assess all supplies needed the
day prior to treatment (Table 67.3).
Educate the patient prior to each step. Ensure that close
contact is maintained via audio and video monitoring. Fig. 67.10 Fixation screw sites immediately after frame removal
Describe the number of exposures to anticipate and the
amount of time each exposure will last. The use of two-way
microphones during treatment will assist in warning when
movement will occur. Most patients choose to rest or sleep
during treatment, and the playing of music into the room has
proved to be very relaxing. Keep the patient and family
informed of treatment progression. During helmet or equip-
ment changes, allow for bathroom breaks or change of posi-
tion. Total treatment time can vary greatly depending on
diagnosis.
Postoperative Care
Immediately upon completion of the procedure, the stereo-

tactic frame should be removed. Light massaging with
manipulation of the pin sites will help close the skin and
decrease scarring (Figs. 67.10 and 67.11). Antibiotic oint- Fig. 67.11 Fixation screw sites after the manipulation of skin
ment and some sort of dressing may be applied to the pin
sites (Fig. 67.12). Typically, patients experience discomfort
at the pin sites within 15 min of frame removal. Encourage
pain medications either for prophylaxis or at initial onset of
discomfort. Have anti-emetics on hand for the occasional
nausea. The convalescence period varies depending on
patient general condition and length of procedure. The
patient requiring cerebral angiography prior to treatment
may require a slightly longer recovery time.
Characteristically, patients who have undergone stereotactic
radiosurgery are prepared to return home within an hour or
two of procedure completion. Seizure-prone individuals
may require an overnight stay. Remind families that driving
is not recommended the day of treatment and prohibited if
any conscious sedation has been administered. Encourage
the consumption of food and drink prior to discharge.
Follow-up instructions should be carefully reviewed with the
patient and family. If follow-up appointment times are not Fig. 67.12 Band-Aids are commonly used to cover pin sites
assigned on discharge, provide appropriate contact individuals overnight
and phone numbers. Ensure prescriptions are made avail- the nursing staff should make a thorough assessment of the
able as needed, although over-the-counter pain relievers patient and family needs.
almost always are sufficient for any discomfort related to Preoperative blood work is generally not necessary in the
the procedure. Patients taking anticonvulsant therapy should young pediatric patient. The application of LMX-4 cream is
be reminded of the importance of frequent monitoring of recommended before any needle sticks, such as an intrave-
serum levels of these drugs. Ensure that patients who are nous (IV) start. Whenever possible, have the child in an age-
taking steroid therapy understand which physician office appropriate environment. The use of pediatric specialty
they are to contact regarding any concerns with these drugs. pre-and postoperative suites will help make transitions more
Check to ensure that a gastrointestinal support drug has fluent. Classically, all children below the age of 13 years
been prescribed for anyone on steroid therapy. Stress the require general anesthesia. Most adolescent children are able
importance of providing radiology imaging (MRI, CT, etc.) to tolerate radiosurgery with some light to moderate seda-
at the time of follow-up with the radiosurgery team. Assist tion. Nurses can help determine the maturity level of the ado-
with referrals so imaging can be completed in a timely man- lescent patient through patient and parent interview. Pediatric
ner prior to the visit. Provide the imaging centers with phy- anesthesia physicians can help determine the optimal seda-
sician phone and fax numbers. tion plan for the individual child. Clearly, the younger child
will require parent involvement until anesthesia has occurred.
Encourage the young child to take a favorite tool or stuffed
Home Follow-up animal and keep it with the child until they are anesthetized.
Allow family members to be with a child of any age during
Phone contact should be made within 24 h of discharge. The appropriate intervals. As with any patient, provide frequent
RN is the best individual to complete this call. Reaffirm pre- updates to parents throughout the procedure to decrease
operative teaching as necessary. If home care or therapy anxiety.
(nursing, PT, OT, etc.) was established at the time of the ste- The identical stereotactic frame that is used for an adult
reotactic treatment, assess that these referrals have been can be used for the child. Certainly, the longer fixation
completed and the patient has been contacted by the appro- screws will need to be used to compensate for a smaller head
priate persons. Assess pain status and make recommenda- circumference. The use of lidocaine at the pin sites may or
tions as appropriate. Assist with scheduling follow-up visits may not be used for the anesthetized patient. Supporting the
as needed. The follow-up phone call is the ideal time to frame during application of a patient under general anesthe-
assess patient satisfaction regarding the entire stereotactic sia can be difficult and may require two people, one to sup-
process. port the frame and the second to support the head. In that the
child may be immobile for several hours during planning and
treatment, monitor pressure points on the skin, and a Foley
Care of the Pediatric Patient catheter may be practical. Monitor body temperature and
keep warm blankets near.
A small percentage of all stereotactic candidates fall under During the delivery of radiation, the anesthesiologist will
the category of the pediatric population. Typically this group need to have visualization of key components. Certainly the
is categorized as the following: cardiac, BP, and oxygen monitor should be in plain view, and
Newborn: birth to 3 months the ability to change monitoring parameters accessible. If pos-
Infant: 3 months to 1 year sible, two cameras should be available in the treatment room.
Toddler: 1–4 years One camera should be focused on the patient for visualization
Preschool: 4–5 years of respirations and the other focused on the ventilator, specifi-
School age: 6–12 years cally the CO2 waveform. Prior to the start of treatment, evalu-
Adolescent: 13–18 years ate the length of all tubing and account for movement into the
Brain tumors and arteriovenous malformations are some machine. Ensuring that these features are in place will pro-
of the conditions in this population that can be treated via mote a safe atmosphere. Offering the adolescent music of their
radiosurgery. The central nervous system continues to choosing can decrease anxiety during treatment.
develop throughout childhood, therefore careful consider- Removal of the stereotactic frame prior to reversal of anes-
ation is given to the appropriateness of the delivery of radia- thesia is recommended. Moving the patient to a pediatric
tion. Consultations are scheduled with the patient and parents PACU (Post-anesthesia care unit) setting will allow for safe
involving all the pertinent members of the radiosurgery team. recovery. Allow family to visit in the PACU as soon as pos-
Risks and benefits of the procedure are discussed at length. sible. Observe the very young child for signs of discomfort
Once a decision has been made to proceed with radiosurgery, specifically immediately after frame removal and administer
826 T.F. Biggins
Table 67.4 Tricks of the trade

Pre-procedure
• Arrange initial consultations with all physicians/nurse for the same day
• When scheduling patients for the procedure, attempt to schedule the simplest treatment for the first case of the day progressing to the most
complex treatment for the last patient of the day
Frame placement
• Use Velcro strap instead of ear bars. The strap can be positioned to fit each patient individually. This increases comfort for patient and
individual supporting the frame
• Use LMX-4 cream on forehead prior to lidocaine injection. Strive for application at least 1 h prior to frame placement
• Add sodium bicarbonate to lidocaine to decrease “sting” at injection site
• If using CT scan, place fixation screws as far from the targeted area as possible to decrease artifact
• Always have patient films available for frame placement
• Use “curved” front posts for extreme lateral or very low (cerebellar) lesions
• Place front posts low on forehead to decrease risk of collision with treatment helmet
• Test all fiducials during frame placement. This will avoid the need to adjust the frame later in the process
• Know your patient’s surgical history, specifically past craniotomies, shunt placements, etc. Fixation screws need to be carefully placed to
avoid these areas
• Place fixation screws as “flush” as possible with the post to avoid collision
• While on a gurney, prop a pillow(s) behind patient’s back to make access to the back of the head adequate for frame placement
• Follow correct site protocol by marking side of treatment involving unilateral site, such as trigeminal neuralgia. Ask the patient to verbalize
side of pain prior to premedication
Radiographic imaging
• Place earplugs in patient prior to MRI
• Elevate knees during scans
• Provide padding anywhere pressure is a concern due to the frame or fiducial cradle touching shoulders, etc. (foam pads, 4 × 4 s)
• Do not dismiss a patient’s complaint of heating at a fixation screw site. Immediately check the screw for any magnetic components and the
patient’s skin for any burning
• Attempt to ensure that the imaging is acceptable to all involved (radiation physicist, neurosurgeon, radiologist) before leaving the imaging
area
• Check that MRI fiducial is filled with solution and free of air bubbles the day before needed
Treatment
• Suggest music be played while undergoing treatment
• Offer support for knees
• Make sure the mattress is at an acceptable height to ensure patient comfort throughout the entire procedure
• If plug patterns will be needed on different helmets, roll the subsequent helmet out of the room and prepare plug pattern while the patient is
being treated
Frame removal
• Massage or pinch screw sites closed immediately after frame removal to assist in the quick closure of screw openings and to decrease
scarring
Patient-specific issues
• CT cisternogram can be successfully used to visualize the trigeminal nerve when MRI is contraindicated
• Anticoagulant therapy will need to be stopped as per physician recommendation when angiogram or cisternogram are to be performed
• Use hardware that will cause the least amount of artifact when using CT imaging (i.e., carbon-filled posts, nonmetallic fixation screws)
analgesics as needed. Encourage the older child to request

pain medications as well. Admit to age-appropriate setting Tricks of the Trade (Table 67.4)
until discharge.
Educate parents and the child regarding pain management, Recognized colleagues have all learned the radiosurgery pro-
pin site care, and follow-up appointments. Provide phone cess through some trial and error. Beginning with consulta-
numbers including emergency contact. After discharge, it is tion, preoperative analysis, frame fitting, gathering of
recommended that nursing staff make phone contact with equipment, proper imaging techniques, much can be learned
families within 24 h. from the established center.
the follow-up patient visit will assist in making the experi-

Conclusion ence gratifying for patients and clinicians.
Care of the radiosurgery patient can initially be challenging

for the clinician. Proper training, communication with estab- References
lished centers and clinical experts, as well as hands-on expe-
rience will support a thriving center. A support group for 1. Mughaw SB. An overview of methods in stereotactic radiosurgery.
Radiol Technol. 1992;63:402–5.
radiosurgery nurses allows for the constant flow of ideas and
2. Browner CM, Hendrickson K. A nursing perspective of Gamma
solutions to challenges that only this population can appreci- Knife treatment. Barrow Quarterly. 1997;13:41–8.
ate. A strong quality assurance program with ongoing patient 3. Lindquist C, Kihlstrom L, Hellstrand E. Functional neurosurgery—
satisfaction evaluation is crucial. Establishing concise guide- a future for the Gamma Knife? Stereotact Funct Neurosurg.
1991;57:72–88.
lines beginning with the intake process and flowing through
4. A tribute to Dr. Lars Leksell. Another perspective—The Publication
for the International Radiosurgery Support Association. 1998:2–3.
Index
A trigeminal nerve function, 343

Accelerated partial breast irradiation (APBI), 163–164 tumor control, 341–343
Acoustic neurinoma, 5 Acromegaly
Acoustic neuromas fractionated radiotherapy, 397–398
biologically effective dose, 357–359 SRS and, 371–373
derivation of hearing RET formula, 359, 364, 365 Active eye-immobilization techniques, 744
fractionated stereotactic radiotherapy Adenocarcinoma, 227–228
clinical outcomes, 363 Age-related macular degeneration (ARMD)
principles of late-responding tissues, 356 diagnosis, 759
treatment, 360 pathophysiology, 758
vs. Gamma Knife, 361, 362 proton beam therapy and photocoagulation, 759
vs. SRS, 363, 364 radiotherapy, 759
gadolinium-enhanced, 361–363 surgical approach, 768–769
microsurgical resection, 359, 360 Amino acid positron emission tomography, 792
middle fossa approach, 360 Angiofibroma
optic nerve sheath meningiomas, 356, 357 pathophysiology, 486
optimal treatment outcome, 365 treatment, 486–487
RET rosigmoid approach, 360 Anti-angiogenic therapy, 283
stereotactic radiosurgery, 17–18, 142, 360 Anxiety, 734
translabyrinthine approach, 360 Arteriovenous malformations (AVMs)
Acoustic schwannomas arterial aneurysms, 605
acoustic tumors, 343–344 clinical presentation, 606–607
brainstem tolerance, 342 dose selection, 130
cochlear tolerance, 342–343 endovascular therapy
complications, 341–343 embolization, 618–621
facial nerve preservation, 343 natural history, 617–618
G-R classification, 340 follow-up, 130
hearing preservation, 342–343 management, 589–590
House–Brackmann classification, 340 natural history, 605–606
hydrocephalus, 343 patient selection, 130
Koos classification, 340 presurgical embolization, 609–611
management, 340–341 presurgical planning
neurofibromatosis type 2, 339 diffusion tensor imaging, 609
patient assessment, 339–340 functional MRI, 608–609
radiosurgery failures, 345 grading systems, 607–608
surgery imaging studies, 608
age and tumor size, 350 radiosurgery, 564
conservative management, 347 case studies, 598–600
facial nerve preservation, 348 complications, 594–595
hearing preservation, 349–351 large AVMs, 596–597
management, 348 obliteration, 592–593
middle fossa approach, 348, 349 patient selection, 591
porus acousticus, 348, 351 post-radiosurgical hemorrhage, 593–594
radiosurgery vs. excision, 350 repeat SRS, 595–596
repeat retrosigmoid craniotomy, 351 Spetzler–Martin grading system, 597–598
retrosigmoid approach, 348, 349, 351 technique, 591–592
translabyrinthine approach, 349, 351 stereotactic image acquisition, 130
treatment method, 350 stereotactic radiation therapy, 61–62
treatment planning issues, 344–345 stereotactic radiosurgery, 18, 142, 410, 611
DOI 10.1007/978-1-4614-8363-2, © Springer Science+Business Media New York 2015
830 Index
Arteriovenous malformations (AVMs) (cont.) brainstem, 645–646

surgical considerations clinical presentation, 639–640
anesthesia, 611 diagnostic imaging, 640–645
intraoperative angiography, 613 epidemiology, 637–638
tenets of surgical resection, 611–612 genetic basis, 638
University of Florida experience, 130–131 microsurgical treatment, 641–644
Astrocytic tumors, 249 molecular basis, 638
Atypical facial pain, 655–656 natural history, 638–639
prevalence, 637
radiosurgical treatment, 644–645
B treatment, 640–641
Baclofen, 650, 668 definition, 637
Beam’s-eye-view technique, 305 Gamma Knife radiosurgery, 628–630
Benign meningiomas intracranial
external beam radiation therapy, 325–328 clinical presentation, 624
grading criteria, 325 deep-seated, resection of, 627
gross total resection, 325 dural arteriovenous fistulas, 631–633
IMRT, 328 epidemiology, 623–624
radiotherapy, 327–328 management options, 625
Simpson grades, 326 molecular basis, 624
Benign pineal region tumors, 463 natural history, 626
Bevacizumab pathophysiology, 623
head and neck tumors, 552 radiobiological considerations, 631
malignant glioma, 283 radiologic features, 624–625
radiation necrosis, 794 radiosurgery, 625–631
Bilateral acoustic neurofibromatosis, 417–418 vein of Galen malformations, 633–634
Bilateral adrenalectomy, 405 Cavernous sinus, 379, 381, 383–386
Body tumor immobilization, 182 Central nervous system (CNS) tumors, 142
Brachytherapy Cerebral angiography, 18–19
ocular metastases, 757 Cerebral cavernous malformations (CCMs)
orbital and ocular metastases, 757 anatomical distribution, 638
in radioactive plaque, 767 brainstem cavernous malformations, 645–646
radionecrosis and, 251 clinical presentation, 639–640
surgical interventions, 755 diagnostic imaging, 640–645
use of, 274 epidemiology, 637–638
using intracavitary intubation, 530 genetic basis, 638
uveal melanomas, 748, 750 microsurgical treatment
Brain metastases, 14–15, 794–795. See also Metastatic infratentorial lesions, 642–644
brain tumors supratentorial lesions, 641–642
Brainstem astrocytomas, 463 molecular basis, 638
Brainstem cavernous malformations, 645–646 natural history, 638–639
Brain Tumor Cooperative Group (BTCG), 250 prevalence, 637
Breast immobilization apparatus, 166–167 radiosurgical treatment, 644–645
Bromocriptine treatment, 640–641
pituitary tumors, 380 Cerebral metastases, 128–129
prolactinomas, 404 Cetuximab, 552
somatotroph adenomas, 405 Charged particles
Bruch membrane, 748 acoustic neuroma, 142
BTCG. See Brain Tumor Cooperative Group (BTCG) arteriovenous malformations, 142
central nervous system tumors, 142
challenges, 143
C dose-distribution curves, 135–136
Cabergoline energy transfer pattern, 135
prolactinomas, 404 hepatocellular carcinoma
somatotroph adenomas, 405 clinical outcomes, 141–142
Capsaicin, 668–669 rationale, 141
Carbamazepine meningioma, 142
epilepsy, 725 non-small cell lung cancer
trigeminal neuralgia, 649–650, 666 advantages, 141
Carmustine (BCNU), 281, 282 anatomy and dose, 138–139
Cavernomas. See Cerebral cavernous malformations (CCMs) highly hypofractionated regimens, 139
Cavernous angiomas. See Cerebral cavernous malformations (CCMs) modestly hypofractionated regimens, 139–140
Cavernous hemangioma, 758 photon radiation, 135
Cavernous malformations physical and technical considerations, 135–137
CCMs pituitary tumors, 142
anatomical distribution, 638 radiobiologic advantages, 137
Index 831
uveal melanomas, 750 role of, 395

Chemodectoma. See Glomus jugulare tumors simulation, 395–397
Chemo-radiation therapy, 273 staging system for, 394
Chemotherapy therapeutic approaches, 394
head and neck tumors gross total resection, 509
altered fractionation, 554–555 hypothalamic involvement, 507
bevacizumab, 552 papillary, 506–507
definitive radiotherapy, 551 pediatric brain tumors, 416–417
human papillomavirus, 552–553 retrochiasmatic lesions, 507–508
hyperfractionated radiotherapy, 554 skull base tumors
IMRT, 553 pathophysiology, 488
locoregional control and survival, 549–550 surgical result, 489
reirradiation, 555–558 treatment, 488–489
stereotactic radiotherapy, 558–559 stereotactic radiosurgery, 412–413
neurofibromatosis-1, 417 Craniotomy, 461–463
pediatric brain tumors, 439–441 Cryptic vascular malformations. See Cerebral cavernous
pineal region tumors malformations (CCMs)
germ cell tumors, 477–479 Cumulative Ocular Melanoma Study (COMS), 766
pineal parenchymal tumors, 479–480 Cushing’s disease, 399
Children’s Oncology Group trial, 430 diagnosis, 404
Chondromas fractionated radiotherapy, 399
pathophysiology, 489 medical therapy, 404–405
radiosurgery, 491 SRS and, 370–371
treatment, 490–491 CyberKnife
Chondrosarcomas beam characteristics, 85
pathophysiology, 489–490 clinical indications
radiosurgery, 491 glioblastoma multiforme, 153–154
treatment, 490–491 intracranial, 153
Chordomas intrathoracic tumors, 155–156
features, 517 liver tumors, 156–157
fractionated stereotactic radiotherapy pancreatic tumors, 156–157
outcomes, 520–521 perioptic tumors, 154
patient selection, 519–520 spinal indications, 155
techniques, 520 trigeminal neuralgia, 154–155
histological appearance, 505 vestibular schwannomas, 154
malignant tumor of spine, 565 2D-3D image registration, 157
radiation therapy, 505–506 digitally reconstructed radiographs, 147–148
skull base tumors image guided tracking system, 151–153
pathophysiology, 489 imaging requirements, 85–86
radiosurgery, 491 intracranial lesions, 147
treatment, 490–491 IRIS collimator, 149
treatment, 518–519 isocentric beams, 148–149
Choroidal melanomas, 748 non-neurologic disorders, 157
Clonazepam, 668 overview, 84–85
Compton scattering, 39–41 planning process, 87–88
Computed tomography (CT) robotic manipulator, 147
meningioma, 316, 317 Synchrony™ respiratory tracking system, 147
paragangliomas, 522 treatment delivery, 148, 150–151
pineal region tumors, 470 treatment planning, 149–150
Contrast-enhanced perfusion MRI, 791 uses, 49, 100, 813
Conventional fractionated radiation therapy, 489, 596 Cyclophosphamide, 478
Convexity meningiomas, 317–319
Corpus callosotomy, 716
Corticobasilar degeneration (CBD), 695 D
Corticotroph adenomas, 404–405 DAT scan, essential tremor, 697
Costotransversectomy, 576 Deep brain stimulation (DBS)
Craniopharyngiomas pallidal, 686–687
adamantinomatous, 506 Parkinson disease, 697
chemotherapy, 441 subthalamic, 688–689
endoscope, 509 thalamic, 681–683
fractionated radiotherapy Depression, 734, 738
clinical presentation, 393 Diffusion tensor imaging (DTI), 7, 609
diagnostic evaluation, 393–394 Diffusion-weighted imaging (DWI), 790–791
epidemiology and histology, 392 Digital reconstruction radiography (DRR), 28–29
preparatory steps, 395–397 Dirty necrosis, 789
results, 399–400 Dopamine agonist therapy, 404
832 Index
Dose and dosimetry, SRS, 41–42 presurgical evaluation, 711–712

isodose line and dose-volume histogram, 44–45 primary generalized, 711
measurement-based dose calculation, 42 seizure types and treatment considerations
model-based dose calculation, 43 antiepileptic therapy, 723–724
output, 43 carbamazepine, 725
pencil-beam kernel, 44 cavernous malformations, 724
percent depth dose, 42–43 focal structural lesions, 724
point kernel, 43–44 levetiracetam, 726
profile/off-axis ratio, 43 low-grade gliomas, 724
Drug-resistant epilepsy phenytoin, 725
differential effect concept, 699 primary motor cortex, 723
radiosurgery surgery
dose issue, 703–704 corpus callosotomy, 716
hypothalamic hamartomas, 700–703 hemispherectomy, 715–716
indications, 706–707 hemispherotomy, 715–716
mesial temporal lobe epilepsy, 703 hypothalamus hamartomas, 714–715
patient selection, 705 medial temporal lobe epilepsy
potential concerns, 705–706 medial temporal lobe epilepsy:
potential role, 700 corroborative testing, 712
target definition, 704–705 medial temporal lobe epilepsy: outcomes, 713–714
Dural arteriovenous fistulas (AVFs), 631–633 medial temporal lobe epilepsy: procedures, 712–713
Dynamic contrast-enhanced MRI, 13 multiple subpial transections, 716
Dynamic susceptibility contrast MRI, 13 treatment implications, 726–728
Dystonia, 686 Epithelial membrane antigen (EMA), 295, 505, 518
Essential tremor (ET)
Gamma Knife thalamotomy, 671–672
E complications, 675
EBRT. See External beam radiation therapy (EBRT) dose for, 673
Electrical stimulation for epilepsy indications, 673
brain stimulation, 717 microelectrode recording, 675
vagal nerve stimulation, 716–717 radiation efficacy, 673–675
Ellis formula, 786 target coordination, 673
EMA. See Epithelial membrane antigen (EMA) tolerance of internal capsule, 673
Embolization vs. conventional thalamic surgery, 675
angiofibroma, 486 medical therapy, 697
parasagittal meningiomas, 317 Esthesioneuroblastomas
petroclival meningiomas, 318 craniofacial surgical resection, 506
sphenoid wing meningiomas, 318 endoscopy, 506
Embryonal tumors kadish staging, 506
chemotherapy, 440–441 skull base tumors, 487–488
stereotactic radiosurgery, 411 Etomidate, 405
En bloc spondylectomy, 576–577 Evisceration, 766
Endoscopic biopsy, 461 External beam radiation therapy (EBRT)
Endovascular embolization, AVMs anaplastic/WHO grade III, 332
curative strategies, 618–619 arget volumes and dose, 333
palliative, 621 atypical/WHO grade II meningiomas, 330
preoperative, 618 benign meningiomas/WHO grade I
pre-radiosurgery, 619–620 FSRT, 327, 328
targeted, 620–621 gross total resection, 325–326
Enucleation, 766–767 IMRT, 328, 329
Ependymal tumors, 249 radiotherapy, 327–328
Ependymoblastomas, 421 simpson grades, 325
Ependymoma cognitive morbidity, 334
chemotherapy, 440 fractionated, 326, 327
fractionated radiotherapy meningiomas, 296
cognitive effects, 435–436 ONSMs, 328–330
endocrine effects, 436 recurrent meningiomas, 330
side effects, 436–437 technical factors, 333
pediatric brain tumors, 418–420 toxicity, 333–334
stereotactic radiosurgery, 410–411 Extracranial SRS, 7
Epilepsy Extracranial stereotactic localization, 29–30. See also Stereotactic
acute/subacute effects of radiosurgery, 726 body radiation therapy (SBRT)
electrical stimulation
brain stimulation, 717
vagal nerve stimulation, 716–717 F
intracranial lesions, 723 Fentanyl patch, 669
localization-related, 711 Fotemustine, 281
Index 833
Fractionated radiotherapy tolerance of internal capsule, 673

ONSM vs. conventional thalamic surgery, 675
modalities movement disorders, 683–685
modalities: doses, constraints and fractionation schemes, 774–776 GammaPod
modalities: techniques, 774–776 accelerated partial breast irradiation, 163–164
modalities: timeframe, 773–774 advantages, 175
modalities: volumes, 774 breast immobilization apparatus, 166–167
optimization, 778–780 continuous trajectory patient positioning system, 167
results dose painting, 173–174
results: endpoints, 774, 777 dosimetry, 168–170
results: oncologic outcome, 777 hypofractionated whole breast irradiation, 163–164
results: ophthalmologic examination outcome, 777 irradiation device, 165–166
results: visual outcome, 777 lumpectomy, 163–164
safety, 778 post-operative treatment planning, 169–172
vs. stereotactic radiosurgery, 473–474 preoperative RT role, 165
Fractionated stereotactic radiotherapy (FSRT) preoperative treatment planning, 172–173
benign meningiomas, 327, 328 technologic advances, 164–165
clinical outcomes, 363 treatment planning system, 167–168
pituitary adenomas, 396 Gardner-Robertson hearing (G-R) classification, 340
skull base tumors Germ cell tumors (GCTs), 441, 477–479
chondrosarcomas, 519–521 Germinomas
chordomas, 519–521 chemotherapy, 477–478
paragangliomas, 523–525 pineal region tumors, 465
Fractionation, 6–7 Glaucoma, 760–761, 769
Functional MRI, 608–609 Glial neoplasms, 262–263
Glioblastoma multiforme (GBM), 153–154
Glioblastomas, 418
G Glioma
Gabapentin, 667 low-grade, 415–416
Gamma Knife (GK) stereotactic brachytherapy
in adults, 409–410 benefits and risks, 272, 274–275
automatic positioning system, 112–113 brachytherapy (125I) advanced treatment, 273–274
brain disorders, 111–112 GliaSite brachytherapy, 272
cavernous malformations, 628–630 high-grade gliomas, 273
cobalt-60, 111 indications, 273
components and functions, 96–97 vs. EBRT, 273–274
conformal dose planning, 118 surgery
4C unit, 79–82, 113–114 benefits, 272, 274–275
daily quality assurance, 115–116 GliaSite radiation therapy, 273
docking mechanism, 114–115 high-grade, 269–270
drug-resistant epilepsies, 700 low-grade, 270–271
limitations, 813 management, 271, 273
meningiomas, 299–300 WHO grade I and II, 412
movement disorders, 690 WHO grade III and IV, 411–412
number and location of active sites, 111–112 Glomus jugulare tumors
pallidotomy, 675–677 microsurgical resection, 501–502
Perfexion, 81–82 skull base tumors
PERFEXION unit, 113–115 pathophysiology, 491
photon beam delivery systems, 45–47 radiosurgery, 493–494
pineal region tumors, 449, 452 side effects, 493
quality assurance of images, 117 surgical resection, 491–493
schematic diagram, 111, 113 treatment plan, 493, 494
stereotactic angiography, 117 SRS, 503
stereotactic CT imaging, 117 Goldmann perimetry tests, ONSM, 777
stereotactic guiding device, 116 Gonadotroph adenomas, 406–407
stereotactic MRI, 116–117
stereotactic radiation delivery, 118–119
target determination, 117–118 H
vs.MVD, 662 Head and neck tumors
Gamma Knife thalamotomy (GKT) chemotherapy/fractionated radiation therapy
intractable tremor, 671–672 altered fractionation, 554–555
complications, 675 bevacizumab, 552
dose for, 673 definitive radiotherapy, 551
indications, 673 human papillomavirus, 552–553
microelectrode recording, 675 hyperfractionated radiotherapy, 554
radiation efficacy, 673–675 IMRT, 553
target coordination, 673 locoregional control and survival, 549–550
834 Index
Head and neck tumors (cont.) magnetic resonance imaging, 11–12

reirradiation, 555–558 metabolic PET imaging, 14
stereotactic radiotherapy, 558–559 MR diffusion and diffusion tensor imaging, 13–14
nasopharyngeal carcinoma (see Nasopharyngeal carcinoma (NPC)) physiologic MR imaging and metabolic PET imaging, 12–13
risk factors, 543 post AVM follow-up imaging, 20
stereotactic radiotherapy radiologic and anatomic considerations, 12
basal cell carcinoma, 539 recurrent gliomas, 16–17
bilateral paragangliomas, 539 trigeminal neuralgia, 18
CyberKnife system, 540 IMSRS. See Intensity-modulated stereotactic radiosurgery
Gamma Knife/linac-based gantry system, 539–540 IMSRS)
hurtle cell carcinoma, 539 Intensity-modulated radiation therapy (IMRT)
squamous cell carcinoma, 539 benign meningiomas, 328, 329
surgery head and neck tumors, 553
evolution, 544–545 meningiomas, 301–302
stereotactic radiosurgery, 545–546 pituitary adenomas, 395, 397
treatment, 543–544 Intensity-modulated stereotactic radiosurgery (IMSRS), 301–302
Helium ion irradiation, 750 International Headache Society, 649
Hemangioblastomas, 563–564 Intraconal space, 765
Hepatocellular carcinoma, 141–142 Intracranial cavernous malformations
High-grade gliomas (HGG) clinical presentation, 624
chemotherapy, 440 epidemiology, 623–624
F-SRT, 258–259 management options, 625
pediatric SRS, 411–412 molecular basis, 624
perspective, 249–251 pathophysiology, 623
prognostic factors, 249 radiobiological considerations, 631
RTOG classes, 250 radiologic features, 624–625
stereotactic radiosurgery radiosurgery, 627–631
before and after EBRT, 257 dose planning technique, 625–627
bevacizumab, 257, 258, 262 dural arteriovenous fistulas, 631–633
comparison of treatment, 252 role, 625
EBRT, 252 vein of Galen malformations, 633–634
F-SRT, 252 Intracranial meningiomas, 795–797
glial neoplasms, 262–263 Intraoperative magnetic resonance imaging, 690
LINAC vs. Gamma Knife, 252 Intrathoracic tumors, 155–156
lumustine, 257, 258 Isocentric linear accelerator, 47–49
NovoTTF-100A, 257, 258
pre-EBRT boost, 257
primary management, 254–257 J
rationale, 251–252 Juvenile pilocytic astrocytomas (JPAs), 415–416
recurrent, 257–258
results, 260
temozolomide, 262 K
vandetanib, 262 Ketoconazole, 404–405
strategies, 262–263 Koos classification, 340
symptoms, 249
toxicities of treatment, 261–262
High-intensity focused ultrasound (HIFU), 690 L
Hospital-employed neurosurgeon, 810 Laminectomy, 575
House-Brackmann classification, 340 Lamotrigine, 667
Human papillomavirus (HPV), 544, 552–553 Leksell Gamma Knife, 745
Hydrocephalus, 343 Leukemia pathologic infiltrates, 758
Hyperfractionated radiotherapy, 554 Levodopa, 696
Hyperthyroidism, 380 LGK PERFEXION system, 745
Hypofractionated whole breast irradiation (HWBI), 163–164 Linear accelerators (LINACs)
Hypopituitarism, 375, 386 with circular cones, 82
Hypothalamus hamartomas (HH), 700–702, 714–715 microwave technology, 122
with multileaf collimators, 83
orbital and ocular diseases, 745–746
I photon beams, 122
Imaging techniques pituitary adenomas, 370
acoustic neuroma, 17–18 radiosurgery
arteriovenous malformations, 18 arteriovenous malformations, 130–131
assessment of treatment response, 15–16 cerebral metastases, 128–129
AVM radiosurgery treatment planning, 20 collimated X-ray beam, 123
brain metastases, 14–15 conformal dose, 124
cerebral angiography, 18–19 fundamental concepts, 121
dynamic susceptibility contrast MRI, 13 intensity modulated radiation therapy, 124
Index 835
malignant gliomas, 129–130 oligoastrocytoma, 286

meningiomas, 126–128 oligodendroglial tumor, 286–288
sphere packing technique, 124 procarbazine, 282
stralkniven, 123 temozolomide, 281–283
Talairach stereotactic localization system, 123 toxicity and quality of life, 288
traveling-wave technology, 123 vs. stereotactic radiosurgery, 288–290
TrilogyT system, 124–125 radiosurgery, 129–130
vestibular schwannomas, 125–126 Mask fixation technique, 744
rhumbatron, 122 McGill pain score, 653
SRS units, 97–100 Medial temporal lobe epilepsy (MTLE)
vs. Gamma Knife, 252–254 corroborative testing, 712
Linear-quadratic (LQ) model, 787 outcomes, 713–714
formalism, 63–64 procedures, 712–713
lethal lesions, 66 Medulloblastoma
mechanistic basis, 62–63 fractionated radiotherapy
pairwise production of chromosome aberrations, 65–66 cognitive effects, 428–431
repair mechanisms, 66 endocrine effects, 431
single doses, 66 neurovascular effects, 431
in vitro, 64–65 treatment evolution, 428–431
in vivo, 64–65 young children, 431–432
Liver tumors, SBRT, 156–157 pediatric brain tumors, 420–421
clinical outcomes and patient selection, 196–199 Meningiomas, 126–128, 142
complications, 200 atypical and malignant, 295, 302–303
treatment planning, 195–196 case study, 306–310
Localization-related epilepsy, 711 characteristics, 295
Low-grade gliomas clinical follow-up results, 297
chemotherapy, 439–440 complication avoidance, 305
pediatric disorders, 415–416 fractionated radiotherapy
surgical management, 270–271 classification, 324
Lumustine, 257, 258 clinical presentation, 393
Lung tumors, SBRT diagnostic evaluation, 393–394
clinical outcomes, 192–194 EBRT (see External beam radiation therapy)
complications, 193–195 epidemiology and histology, 392
fractionation, 191–192 etiology, 324–325
gross tumor volume, 190 histopathology, 325
patient selection, 190 preparatory steps, 395–397
radiation-induced meningioma, 324
results, 399
M roles, 395
Magnetic resonance imaging (MRI), 11–12 simulation, 395–397
arteriovenous malformation, 608–609 staging system for, 394
dynamic contrast-enhanced, 13 therapeutic approaches, 394
dynamic susceptibility contrast, 13 Gamma Knife radiosurgery, 299–300
meningiomas, 298–300, 302, 303, 316 intensity-modulated radiation therapy, 301–302
pineal region tumors, 447 MRI scan, 298–300, 302, 303
pituitary adenomas, 393 optic nerve sheath, 302
Malignant gliomas patient algorithm, 303–305
chemotherapy pediatric brain tumors, 418
advantages, 288–290 procedure, 305
anaplastic astrocytoma, 285–286 progesterone and estrogen receptors, 323
anti-angiogenic therapy, 283 radiation therapy, 296
antitumor activity, 280–281 radiographic diagnosis, 316
bevacizumab, 281, 283 signs and symptoms, 316
brain barrier blood, 279 SRS, 297–299
carmustine, 281, 282 SRT, 300–301
complexity of glioma biology, 279–280 stereotactic radiation, 296–297, 302–303
concomitant medications, 280 surgery
EORTC/NCIC trial, 287, 288 case reports, 318–321
fotemustine, 281 considerations, 295–296
of glioblastoma, 283–284 convexity region, 317
irinotecan, 281 decision-making and techniques, 316
lomustine, 281, 282 parasagittal region, 317
methylguanine methyltransferase, 284–285 petroclival region, 318
miscellaneous agents, 283 and radiosurgery, 303
nimustine, 281 sphenoid wing, 317–318
nitrosoureas, 281 treatment planning, 305
obstacles, 279–281 WHO grade classification, 315–316
836 Index
Mesial temporal lobe epilepsy, 703 radiosurgical pallidotomy, 687–688

Metabolic PET imaging, 14 radiosurgical subthalamotomy, 689
Metastatic brain tumors radiosurgical thalamotomy
adenocarcinoma, 227–228 complications, 685
complications, 236–237 Gamma Knife thalamotomy, 683–685
cost-effectiveness, 237–238 merits, 683
patient selection, 233–235 MR spectroscopy imaging, 13
prognostic factors, 211–212 MR venography, 317
quality of life, 238–239 Multicenter Collaborative Ocular Melanoma Study
renal cell carcinoma, 227 (COMS), 755
single brain metastasis, 211 Multileaf collimators (MLC)
stereotactic radiosurgery 3D conformal delivery, 83
acute complications, 223 dynamic conformal arcs, 84
biologically effective dose, 215 fixed-field intensity-modulated fields, 84
brainstem metastases, 220–221 volumetric modulated arc therapy, 84
chemotherapy and radiosensitizers, 225 Multiple sclerosis, 655–656
chronic complications, 223–225
costs, 225
ionizing radiation, 214 N
large brain metastases, 222–223 Narcotics, 669
large institutional/multi-institutional results, 218–219 Nasopharyngeal carcinoma (NPC)
morbidity, 218 local control and survival rates, 529–530
multiple brain metastases, 220 local failures
neurocognitive testing, 218 case study, 531
phase II/III trials, 219–220 complications, 534–536
quality-of-life issues, 217 conventional salvage treatment, 530–531
recurrent metastatic disease, 216 radiosurgery planning and treatment, 533–534
surgical resection cavity, 222–223 target localization, 532
vs. surgery, 215–216, 233–234 tumor control, 534, 535
treatment outcome, 235 outcomes, 538
tumor local control and recurrence, 235–237 treatment results
whole-brain radiation therapy external reirradiation, 536
chemotherapy and radiosensitizers, 225 fractionated stereotactic radiation, 536–537
fractionation schemes, 212 gamma knife radiosurgery, 536
phase II/III trials, SRS, 219–220 prognostic scoring system, 537
Radiation Therapy Oncology Group, 212–213 radiotherapy for, 537
radiosurgery, 216–217 National Health Care Advisory Board, 810
and surgery, 213–214 Neurofibromatosis-1 (NF-1), 417
Metastatic tumors Neurofibromatosis-2 (NF-2), 417–418
orbit, 766 NGGCT. See Nongerminomatous germ cell tumor (NGGCT)
spine, 195 Nimustine (ACNU), 281
Metyrapone, 405 Nitrosoureas, 281
Microelectrode recording (MER), 682 Nongerminomatous germ cell tumor (NGGCT)
Microvascular decompression (MVD), trigeminal neuralgia, 650 chemotherapy, 478–479
economical and patient’s perspective, 662 GKRS, 452
in elder patients, 662 pineal region tumors, 446–447, 465
procedure, 660–661 platinum-based multiagent chemotherapy, 472
radiofrequency ablation, 662 stereotactic radiosurgery, 465
results, 661–662 teratomas, 479
vs. Gamma Knife, 662 Non-small cell lung cancer
vs. glycerol rhizotomy, 662 advantages, 141
Middle fossa approach, 348, 349, 360 anatomy and dose, 138–139
Mini-multileaf collimator (MMLC), 813–814 highly hypofractionated regimens, 139
Mitotane, 405 modestly hypofractionated regimens, 139–140
Monoamine oxidase (MAO) type B inhibitors, 696 Novalis, 814
Movement disorders
DBS
pallidal, 686–687 O
subthalamic, 688–689 Obsessive-compulsive disorder (OCD), 734, 737–738
thalamic, 681–683 Ocular and orbital disease
Gamma Knife radiosurgery, 690 fractionated radiotherapy (see Optic nerve sheath meningioma
high-intensity focused ultrasound, 690 (ONSM))
intraoperative magnetic resonance imaging, 690 surgeon’s application, SRT, 770
medical therapy surgical approach
essential tremor, 697 ARMD, 768–769
Parkinson’s disease, 695–697 glaucoma, 769
Index 837
retinoblastoma, 767–768 Pancreatic tumors, 156–157

uveal melanoma, 766–767 Papillary tumor of pineal region (PTPR), 464
Olfactory neuroblastomas. See Esthesioneuroblastomas Paragangliomas
Oligoastrocytoma, 286 features, 521–522
Oligodendroglioma, 249, 286 fractionated stereotactic radiotherapy
Onyx, 610 outcomes, 524–525
Optic nerve sheath meningioma (ONSM), 302, 328–330 patient selection, 523–524
fractionated radiotherapy techniques, 524
modalities skull base tumors
modalities: doses, constraints and fractionation schemes, pathophysiology, 491
774–776 radiosurgery, 493–494
modalities: techniques, 774–776 side effects, 493
modalities: timeframe, 773–774 surgical resection, 491–493
modalities: volumes, 774 treatment plan, 493, 494
optimization, 778–780 treatment, 522–523
results Parasagittal meningiomas, 317, 319
results: endpoints, 774, 777 Parkinson’s disease
results: oncologic outcome, 777 medical therapy, 695–697
results: ophthalmologic examination outcome, 777 posteroventral pallidotomy, 675–677
results: visual outcome, 777 Partial lamellar sclerouvectomy, 767
safety, 778 Pediatric brain tumors
origin, 773 chemotherapy
Orbital and ocular diseases craniopharyngiomas, 441
psychosocial aspect, conservative treatment, 755 embryonal tumors, 440–441
stereotactic radiosurgery/radiotherapy ependymoma, 440
ARMD, 758–760 germ cell tumors, 441
dosimetry, 746, 747 high-grade gliomas, 440
eye fixation, 744 low-grade gliomas, 439–440
glaucoma, 760–761 fractionated radiotherapy
Leksell Gamma Knife, 745 ependymoma, 435–437
leukemia pathologic infiltrates, 758 fractionated irradiation, 428
LINAC, 745–746 medulloblastoma, 428–432
malignant lymphomas, 758 primary brain tumors, 432–435
metastatic lesions, 757 radiosurgery, 428
retinoblastoma, 756 RT-1 protocol, 433–435
side effects and complications, 761–762 target volume, 433
uveal melanomas (see Uveal melanomas) surgery
vascular lesions, 758 case studies, 423
Orbital lymphoproliferative disease, 766 high-grade tumors, 418–421
Organized radiosurgery, 8 low-grade tumors, 415–418
Otolaryngologists, 812 surgical management, 415
Oxcarbazepine, 666 vascular malformations, 421–422
Pegvisomant, 405
Percutaneous balloon compression, 660
P Percutaneous glycerol rhizotomy (PGR)
Pain and psychiatric disorders trigeminal neuralgia, 659–660
pathophysiology, 732 vs. MVD, 662
pharmacological treatment, 735 Percutaneous radiofrequency ablation
stereotactic radiosurgery trigeminal neuralgia, 660
anxiety, 734–735 vs. MVD, 662
arguments, 735 Perioptic tumors, 154
chronic pain, 733 Petroclival meningiomas, 318
complications, 738–739 Phenytoin, trigeminal neuralgia, 668
depression, 734 Photon beam delivery systems
dosimetry, 736 CyberKnife, 49
obsessive-compulsive disorder, 734 Gamma Knife, 45–47
outcomes, 737–738 isocentric linear accelerator, 47–49
somatic pain, 731 Pineal astrocytomas, 463
trigeminal neuralgia, 731–733 Pineal parenchymal tumor (PPT)
Pallidal deep brain stimulation, 686–687 chemotherapy, 479–480
Pancreatic ductal adenocarcinoma (PDAC) fractionated radiation therapy, 470, 473
definitive treatment, 201–203 pineal region, 461–463
neoadjuvant therapy, 201 Pineal region tumors
palliative setting, 201 anatomy, 445–446
resectability, 203 case studies, 453–456
treatment planning considerations, 203 chemotherapy
838 Index
Pineal region tumors (cont.) gonadotroph adenomas, 406–407

germ cell tumors, 477–479 nonfunctioning adenoma, 406–407
pineal parenchymal tumors, 479–480 prolactinomas, 403–404
fractionated radiation therapy somatotroph adenomas, 405
benign cyst, 469 thyrotropin adenomas, 405–406
computed tomography, 470 meningiomas (see Meningiomas)
NGGCTs, 469, 473 surgery
pineoblastomas, 469–470 ACTH-secreting, 386, 387
pineocytomas, 469 anesthetic, 380–381
platinum-based multiagent chemotherapy, 472 bromocriptine therapy, 380
radiotherapy, 470–471 cavernous sinus, 379, 381, 383–386
spinal failure, 471–472 endocrinologic, 380–381
spinal seeding, 470 endoscopic advancement, 382–383
symptoms, 470 GH-secreting, 379, 381, 387
teratomas, 469 hyperthyroidism, 380
GKRS for, 452 hypopituitarism, 386
imaging, 447 indications for, 379–380
management, 447–448 microscopic endonasal, 382
pathology, 446–447 modifications of transsphenoidal approch, 384–386
presentation and natural history, 446 and planning, 380–381
radiosurgery in, 449–452 radiation therapy, 386–388
surgical approaches, 448–449 and radiographic considerations, 380–381
benign tumors, 463 transcranial, 386
craniotomy, 461–463 transsphenoidal approach, 381, 382
endoscopic biopsy, 461 Platinum-based multiagent chemotherapy, 472
glial tumors, 463–465 Polymethylmethacrylate (PMMA), 574, 576
radiographic appearance, 459–460 Post-anesthesia care unit (PACU), 825
results, 463 Posteroventral pallidotomy (PVP), 675–677
stereotactic biopsy, 460–461 Post-herpetic neuralgia, 655–656
surgical management, 460 Pregabalin, 667–668
Pineoblastomas (PBs) Primary generalized epilepsy, 711
chemotherapy, 479–480 Primitive neuroectodermal tumors (PNET), 420
pediatric brain tumors, 421, 447 Procarbazine, 282
Pineocytomas Prolactinomas
chemotherapy, 479 fractionated radiotherapy, 398–399
pineal region tumors, 447 medical therapy, 403–404
Pituitary adenomas Proton beam systems, 813
fractionated radiotherapy Proton beam therapy (PBT), 5–6, 750
classification, 392 Proton SRS
clinical presentation, 392–393 clinical treatment planning, 90–92
diagnostic evaluation, 393–394 considerations, 90
epidemiology and histology, 391–392 dose modeling, 88–89
results, 395–397 pre-planning, 89
role of, 395 proton delivery systems, 88
simulation, 395–397
staging system for, 394
therapeutic approaches, 394 R
stereotactic radiosurgery Radiation necrosis
and acromegaly, 371–372 complications
and cushing’s disease, 370–372 brain metastases, 794–795
CyberKnife, 370 intracranial meningiomas, 795–797
devices, 370 management algorithm, 800
endocrine remission and late recurrence, 373–375 vestibular schwannomas, 797–799
Gamma Knife, 370 diagnosis
hypopituitarism, 375 amino acid PET, 792
LINAC, 370 diffusion-weighted imaging, 790–791
and nonfunctioning, 370, 371 FDG-PET, 791–792
and prolactinomas, 370–374 histological examination, 792–793
proton beam unit, 370 magnetic resonance spectroscopy, 790
transsphenoidal approach, 370 MRI, 789, 790
Pituitary tumors, 142 perfusion MRI, 791
adenomas (see Pituitary adenomas) SPECT, 792
craniopharyngiomas (see Craniopharyngiomas) dose and volume dependence, 786–788
medical therapy histopathologic features, 788–789
corticotroph adenomas, 404–405 historical perspective, 786
diagnosis and classification, 403 incidence, 788
Index 839
latency, 788 chondromas, 489–491

radiation toxicity, 786 chondrosarcomas, 489–491
treatment chordomas, 489–491
anticoagulants, 793 clinical entities, 486
antioxidant therapy, 793–794 craniopharyngiomas, 488–489
bevacizumab, 794 dose limitations, 484–485
hyperbaric oxygen, 793 esthesioneuroblastomas, 487–488
rehabilitation, 794 fractionated/stereotactic radiotherapy
steroids, 793 chondrosarcomas, 517–521
surgical resection, 793 chordomas, 517–521
Radiation safety program, 103 paragangliomas, 521–525
Radiation Therapy Oncology Group (RTOG), 254, 256 glomus jugulare tumors/chemodectoma/paraganglioma, 491–494
Radiobiological principles radiosurgical tolerance, 484–485
accelerated repopulation, 60 re-irradiation, 486
applications surgery
AVMs, 61–62 chordomas, 503–504
benign tumors, 62 considerations, 500
malignancies, 60–61 craniopharyngiomas, 506–509
dose distributions, 66–67 esthesioneuroblastomas, 506, 507
linear-quadratic model glomus jugulare tumors, 501–503
formalism, 63–64 meningiomas, 509–511
lethal lesions, 66 trigeminal schwannomas, 500–501
mechanistic basis, 62–63 technique evolution, 484
pairwise production of chromosome aberrations, 65–66 Somatic pain, 731
repair mechanisms, 66 Somatostatin analog (SSA)
single doses, 66 somatotroph adenomas, 405
in vitro, 64–65 thyrotropin adenomas, 406
in vivo, 64–65 Somatotroph adenomas, 405
malignant tumors, 67–68 Spetzler–Martin grading system
reoxygenation, 57–59 arteriovenous malformations, 597–598
repair, 59–60 AVMs, 607
Radiosurgical pallidotomy, 687–688 Sphenoid wing meningiomas, 317–318
Radiosurgical subthalamotomy, 689 Spiegel–Wycis frame, 4
Radiosurgical thalamotomy Spinal metastases
complications, 685 adjuvant postoperative, 566–568
Gamma Knife thalamotomy, 683–685 brachy P32 plaques, 568–569
merits, 683 fractionated radiation therapy
Recurrent gliomas, 16–17 clinical results, 584–587
Registered nurse, 818 single-fraction SBRT, 583–584
Renal cell carcinoma, 227 hypofractionation, 568
Residency Review Committee (RRC), 808 primary irradiation, 565–566
Retinoblastoma salvage reirradiation dose, 568, 569
diagnosis and treatment, 756 Spinal tumors
pathophysiology and radiosurgery, 756 benign tumors
surgical approach, 767–768 arteriovenous malformation, 564
Retrosigmoid approach, 348, 349, 351 hemangioblastomas, 563–564
RTOG 0539 study, 331, 333 combined therapy timing, 575
curative resection, 574
diagnosis, 574
S immobilization, 182
SBRT. See Stereotactic body radiation therapy (SBRT) malignant tumors, 565
Scandinavian Glioblastoma Study Group (SGSG), 250 metastases
Simpson grades system, 325, 326 adjuvant postoperative, 566–568
Single-fraction radiosurgery, 788 brachy P32 plaques, 568–569
SJMB03 trial, 429 hypofractionation, 568
SJMB96 trial, 429 primary irradiation, 565–566
Skull base meningiomas salvage reirradiation dose, 568, 569
cavernous sinus, 509 neural decompression, 573
cerebellopontine angle, 511 pain control, 574
petroclival meningiomas, 510 stabilization, 574
sphenoid wing, 509 surgical approach
stereotactic radiosurgery, 300 anterior approaches, 576
stereotactic radiotherapy, 301 case studies, 578–579
tuberculumsellae, 511 decompressive posterior laminectomy, 575
Skull base tumors en bloc spondylectomy, 576–577
angiofibroma, 486–487 location, 571–572
840 Index
Spinal tumors (cont.) Stereotactic brachytherapy (SBT)

posterolateral approaches, 576 advent of GliaSite brachytherapy, 272
spinal cord decompression, 575–576 benefits and risks, 272, 274–275
vertebroplasty and kyphoplasty, 577 brachytherapy (125I) advanced treatment, 273–274
surgical decision making algorithm, 573 high-grade gliomas, 273
treatment modality, 572 indications, 273
Spongioblastoma multiforme, 249 vs. EBRT, 273–274
Squamous cell carcinomas of head and neck (HNSCC) Stereotactic instrument, 4–5
human papillomavirus in, 552–553 Stereotactic localization techniques
IMRT, 553 digital reconstruction radiography, 28–29
recurrent, 553, 559 extracranial, 29–30
SRS. See Stereotactic radiosurgery (SRS) frame-based to frameless system, 28
SRT. See Stereotactic radiotherapy (SRT) history of stereotaxis, 25–27
Stereotactic biopsy, pineal region tumors, 460–461 image-guided, 26–28
Stereotactic body radiation therapy (SBRT), 164 immobilization and target tracking, 30–31
high-dose fractions, 177 robotic tools, 30
historical perspective, 178 Stereotactic radiosurgery (SRS), 355–356
image guidance accuracy, 101
equipment quality assurance, 189–190 in children
in-room CT, 187 arteriovenous malformations, 410
kV cone beam CT, 187–188 craniopharyngiomas, 412–413
MV cone beam CT, 188 embryonal tumors, 411
MV TomoTherapy, 189 ependymomas, 410–411
overview, 186 Gamma Knife, 409–410
patient safety, 189 goal of, 409
real-time tumor tracking, 189 high-grade gliomas, 411–412
strategies, 186–187 low-grade astrocytomas, 412
three-dimensional volumetric, 187 clinical indications, 808
two-dimensional equipment, 187 components and functions
imaging response, 180–181 Cyberknife, 100
immobilization Gamma Knife, 96–97
body tumor, 182 linear accelerator-based units, 97–100
breathing motion assessment, 182–183 Tomotherapy unit, 101
overview, 182 Compton scattering, 39–41
spinal tumor, 182 cost and budget, 106–107
immune response, 179–180 design principles, 95–96
liver tumors dose and dosimetry, 41–42
clinical outcomes and patient selection, 196–199 isodose line and dose-volume histogram, 44–45
complications, 200 measurement-based dose calculation, 42
treatment planning, 195–196 model-based dose calculation, 43
lung tumors output, 43
clinical outcomes, 192–194 pencil-beam kernel, 44
complications, 193–195 percent depth dose, 42–43
fractionation, 191–192 point kernel, 43–44
gross tumor volume, 190 profile/off-axis ratio, 43
patient selection, 190 dose characteristics, 51
mechanism of action, 179 Gamma Knife SRS, 302
metastatic tumors, spine, 195 head and neck tumors, 545–546, 558–559
normal tissue response, 181–182 installation and acceptance, 103–105
pancreatic ductal adenocarcinoma meningiomas, 296–299
definitive treatment, 201–203 movement disorders
neoadjuvant therapy, 201 application, 671
palliative setting, 201 gamma knife pallidotomy, 675–677
resectability, 203 gamma knife subthalamotomy, 677
treatment planning considerations, 203 gamma knife thalamotomy, 671–675
planning target volume, 177 pair production, 40–41
radiobiologic models, 178–179 patient process, 826
spinal metastases frame application, 820–821
clinical results, 584–587 home follow-up, 825
single-fraction SBRT, 583–584 intake, 818
treatment planning patient arrival and preparation, 820
heterogeneity corrections, 185–186 pediatric patient care, 825–826
imaging, 183–184 postoperative care, 824–825
overview, 183 preauthorization process, 820
plan evaluations, 185 preprocedure consultation, 818, 820
target volumes, 184 radiographic imaging, 821–823
Index 841
stereotactic treatment, 823–824 T

treatment planning, 823 Tardive dyskinesia, 695
pediatric brain tumors, 439–441 Temozolomide (TMZ)
personnel training and qualifications, 105–106 high-grade gliomas, 262
photoelectric absorption, 39–40 malignant glioma, 281–283
photon beam delivery systems Teratomas, 469–471
CyberKnife, 49 Thalamic deep brain stimulation, 681–683
Gamma Knife, 45–47 Thyroid eye disease, 766
isocentric linear accelerator, 47–49 Thyrotropin adenomas, 405–406
pituitary adenoma Tomotherapy unit, 101
and acromegaly, 371–372 Topiramate
and cushing’s disease, 370–372 epilepsy, 725, 726
CyberKnife, 370 trigeminal neuralgia, 667
devices, 370 Translabyrinthine approach, 349, 351, 360
endocrine remission and late recurrence, 373–375 Transoral robotic surgery (TORS), 544–545
Gamma Knife, 370 Transpupillary thermotherapy, 767
hypopituitarism, 375 Transsphenoidal approach, 381
LINAC, 370 Treatment planning, SRS
and nonfunctioning, 370, 371 contouring structures, 77
and prolactinomas, 370–374 CyberKnife
proton beam unit, 370 beam characteristics, 85
transsphenoidal approach, 370 imaging requirements, 85–86
population-based market size, 808, 809 overview, 84–85
potential issues and challenges, 102 planning process, 87–88
practice development design, 77–78
credentialing protocols, 811 dose calculation, 78
decision making, 809–811 forward vs. inverse planning, 78
marketing, 812 Gamma Knife, 73
referral base, 811–812 4C unit, 79–82
practice issues Perfexion, 81–82
data management, 815 image registration, 75–76
identification of space, 814 linac-based SRS
patient follow-up, 815 with circular cones, 82
radiosurgery systems, 812–814 with multileaf collimators, 83
staff responsibilities, 814–815 mini multileaf collimators, 74
proton’s interaction with matter, 50–51 patient imaging, 75–76
proton therapy, 50 prescription, 77
quality control, assurance and management, 51–53, 105–106, 818 proton SRS
radiation safety program, 103 clinical treatment planning, 90–92
radiation source considerations, 90
accelerator tube, 37–38 dose modeling, 88–89
bending magnet, 37–38 pre-planning, 89
cobalt-60, 35–36 proton delivery systems, 88
electromagnetic spectrum, 35–36 quality evaluation, 78–79
electron gun, 36 sphere packing technique, 74
flattening filter free mode, 39–40 stereotactic localization, 74–75
modulator, 36 technical improvements, 73
power supply, 36 three-dimensional image data, 73
radiofrequency source, 37 Trigeminal neuralgia (TN), 18, 154–155
treatment head, 38–39 ablative percutaneous, 650
waveguide system, 37 atypical facial pain, 655–656
X-ray production, 36–37 Barrow Neurological Institute, 653
residency training, 808–809 case studies, 652
service and maintenance, 104–105 complications, 654
shielding design, 102–103 dose and dose rate, 651–652
training and staffing, 818, 819 dose selection, 736
vs. fractionated radiation therapy, 473–474 GK-SRS effect, 650, 653–655
Stereotactic radiotherapy (SRT) See also Radiobiological principles linac and CyberKnife SRS, 650
applications, 302–303 McGill pain score, 653
atypical meningiomas, 297 medical management, 735
benign meningiomas, 297 antiepileptic drugs, 666–668
meningiomas, 296–297, 300–301 clinical features, 665
ONSM, 302 GABA-related agents, 668
Subperiosteal space, 765 Trigeminal neuralgia (TN) (cont.)
Subthalamic deep brain stimulation, 688–689 local anesthetics, 668–669
Synchrony™ respiratory tracking system, 147 narcotic agents, 669
842 Index
neuropathic pain, 665–666 V

pathophysiology, 665 Vagal nerve stimulation (VNS), 716–717
medical therapy, 649–650 Vandetanib, 262
microvascular decompression, 736 Vascular endothelial growth factor-A (VEGF)
multiple sclerosis, 655–656 ARMD, 768
MVD, 650 bevacizumab, 262
oxcarbazepine, 666 head and neck tumors, 552
pathophysiology, 733 Vascular lesions, 758
post-herpetic neuralgia, 655–656 Vein of Galen aneurysms, 633–634
and radiosurgery, 731–732 Vein of Galen malformations (VGMs), 633–634
secondary, 655 Vertebroplasty, spinal metastases, 577
SRS, 650 Vestibular schwannomas, 125–126, 154, 797–799.
surgery See also Acoustic neuromas
balloon micro-compression, 660 VGMs. See Vein of Galen malformations (VGMs)
microvascular Volumetric modulated arc therapy (VMAT), 84
decompression, 660–662 Von Recklinghausen neurofibromatosis, 417
radiofrequency ablation, 660
trigeminal neuralgia, 659–660
Trigeminal schwannomas W
gross total resection, 500–501 Wada test, 711
SRS, 501 Whole-brain radiation therapy (WBRT)
TrilogyT system, 124–125 chemotherapy and radiosensitizers, 225
complications, 242
effectiveness, 242
U efficacy, 241
Uveal melanomas focal treatment of brain metastases
anatomic stage/prognostic Aoyama trial, 244
groups, 750 cranial radiation therapy, 246
brachytherapy, 748, 750 long-term neuropsychological side effects, 243
charged particle therapy, 750 low-grade glioma, 245
clinical trials, 755 mini-mental state examinations, 246
MRI, 751–752 neurocognitive functioning, 245
pathophysiology, 746–747 neuroimaging, 243
prognosis, 754–755 Patchell study, 244
proton beam therapy, 750 postoperative radiotherapy, 244
radiosurgery, 750–751 fractionation schemes, 212
radiotherapy, 747 issues, 241–242
stereotactic radiotherapy, 750 phase II/III trials, SRS, 219–220
TNM staging system, 748 Radiation Therapy Oncology Group, 212–213
toxicity scoring system, 751 radiosurgery, 216–217
treatment, 752, 755–756 recurrent brain metastases, 246
tumor location, 747–748 and surgery, 213–214
tumor response, 752–753 with/without SRS, 242–243

Editors: Lawrence S. Chin William F. Regine

Uploaded by

Copyright:

Available Formats

Editors: Lawrence S. Chin William F. Regine

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Editors: Lawrence S. Chin William F. Regine

Uploaded by

Copyright:

Available Formats

Lawrence S.

Principles and Practice

ISBN 978-1-4614-8362-5 ISBN 978-1-4614-8363-2 (eBook)

Library of Congress Control Number: 2014952886

© Springer Science+Business Media New York 2015

Printed on acid-free paper

Springer is part of Springer Science+Business Media (www.springer.com)

1 The History of Stereotactic Radiosurgery ............................................................. 3

Part II Radiation Biology and Physics

4 The Physics of Stereotactic Radiosurgery ............................................................. 35

Part III Stereotactic Radiosurgery Techniques

8 Gamma Knife Radiosurgery ................................................................................... 111

13 Stereotactic Body Radiation Therapy .................................................................... 177

Part IV Treatment of Disease Types

14 Metastatic Brain Tumors......................................................................................... 211

31 Pediatric Disorders: Viewpoint—Surgery ............................................................. 415

Part V Patient Care and Socio-Economic Issues

65 Complications and Management in Radiosurgery ............................................... 785

Index .................................................................................................................................. 829

J. Gregory Cairncross, M.D. Southern Alberta Cancer Research Institute, University of

Christopher Dardis, M.D. Barrow Neurological Clinics, Phoenix, AZ, USA

Krisztian Homicsko, M.D., Ph.D. Department of Oncology, Centre Hospitalier Universitaire

Gregory J.A. Murad, M.D. Department of Neurosurgery, University of Florida, Gainesville,

Maryam Rahman, M.D., M.S. Department of Neurosurgery, University of Florida,

Shervin M. Shirvani, M.D. Department of Radiation Oncology, Banner MD Anderson

David M. Wildrick, Ph.D. Department of Neurosurgery, The University of Texas

surgery and fractionated stereotactic radiotherapy (FSRT)

Emergence of Radiosurgical Practice

Lars Leksell conducted his earliest work with SRS while

gamma knife (GK), containing 179 cobalt sources in a hemi-

electrons to almost the speed of light, then directing the elec-

These and other articles have fostered a useful debate regard-

With frameless LINAC-based SRS and SRT, it became

Table 1.1 SRS landmarks Table 1.2 Peer-reviewed publication on SRS

10. Hirsch O. Uber methoden der operativen behandlung von hypophy-

Introduction Imaging and Treatment Planning

Assessing tumor response after SRS can be difﬁcult due

Dynamic susceptibility contrast (DSC) T2*-weighted imag-

Recurrent Gliomas shown to potentially improve target delineation. Areas of con-

Fig. 2.5 (a) University of

AVM Radiosurgery Treatment Planning

In the 1930s, visualization of intracranial structures by

Fig. 3.2 The incorporation of

As the speed and the computation power of modern com-

adjacent CT slices to generate an effective transmission

Fig. 3.5 The use of cone beam

2. Zernov DN. L´encephalometrie. Rev Gen Clin Ther. 1890;19:302.

Protons, deﬁned as positively charged particles found in

X-Ray Production in a Linear Accelerator Electron Gun

Radiofrequency Source basically microwave resonators. Resonance is a very common

Waveguide System Bending Magnet

Fig. 4.4 A typical cylindrically symmetric cavity

Fig. 4.6 A schematic of the treatment head of a medical linear accel-

and lose its energy as bremsstrahlung radiation (i.e., an X-ray).

A Photon’s Interaction with Matter

The effectiveness of radiation treatment is based on how

the interaction. The energy and momentum are conserved

Fig. 4.12 A photon enters a volume of a medium and transfers a por-

Radiation therapy’s aim is to kill the target tissue (or cell)