(Molecular and Translational Medicine) Alan H.B. Wu, Kiang-Teck J. Yeo-Pharmacogenomic Testing in Current Clinical Practice_ Implementation in the Clinical Laboratory (Molecular and Translational Medi

Molecular and Translational Medicine
Series Editors
William B. Coleman
Gregory J. Tsongalis
For other titles published in this series, go to

www.springer.com/series/8176
Alan H.B. Wu Kiang-Teck J. Yeo
Editors
Pharmacogenomic Testing
in Current Clinical Practice
Implementation in the Clinical Laboratory
Editors
Alan H.B. Wu Kiang-Teck J. Yeo
University of California, San Francisco University of Chicago
San Francisco General Hospital Department of Pathology
San Francisco, CA Clinical Pharmacogenomics Program
USA Chicago, IL
wualan@labmed2.ucsf.edu USA
jyeo@bsd.uchicago.edu
ISBN 978-1-60761-282-7 e-ISBN 978-1-60761-283-4

DOI 10.1007/978-1-60761-283-4
Springer New York Dordrecht Heidelberg London
Springer Science+Business Media, LLC 2011

All rights reserved. This work may not be translated or copied in whole or in part without the written
permission of the publisher (Humana Press, c/o Springer Science+Business Media, LLC, 233 Spring
Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly
analysis. Use in connection with any form of information storage and retrieval, electronic adaptation,
computer software, or by similar or dissimilar methodology now known or hereafter developed is
forbidden.
The use in this publication of trade names, trademarks, service marks, and similar terms, even if they
are not identified as such, is not to be taken as an expression of opinion as to whether or not they are
subject to proprietary rights.
While the advice and information in this book are believed to be true and accurate at the date of going
to press, 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
Humana Press is part of Springer Science+Business Media (www.springer.com)

Preface
The personalization of therapeutics promises to deliver a more efficacious and

cost-effective approach towards treating patients with chronic diseases. For many
medications, the pharmaceutical industrys objective of one drug fits all individu-
als has proven to be incorrect. Pharmacogenomics is an emerging discipline that
will be essential for implementing personalized medicine. While there have been
dozens of contemporary review articles on the science and specific application of
pharmacogenomics to particular drugs, this book, Pharmacogenomic Testing in
Current Clinical Practice: Implementation in the Clinical Laboratory is the first
compilation of the tests currently in routine clinical use. In this rapidly changing
field, we recognize that a text of this type will be quickly outdated. Nevertheless, we
have assembled chapters from the key authorities and investigators who have con-
ducted the essential clinical trials necessary to justify pharmacogenomic testing
today.
This book is designed as a reference to clinical laboratory directors who are
contemplating or assigned the task of establishing a pharmacogenomics labora-
tory, and pharmacologists and clinicians who must interpret results of testing.
Each author has given a pharmacologic background on the target drug, the need
for pharmacogenomic testing, and how results can be translated into clinical deci-
sions. Where appropriate, case studies are given to illustrate typical clinical sce-
narios. An extensive bibliography is cited so that the reader can refer to the
original studies.
In planning for this book, we made a distinction between pharmacogenomic
tests for genes that alter drug metabolism, transport, and excretion from com-
panion diagnostic tests that are performed on tumor tissues. Pharmacogenomic
tests are conducted to determine germ line mutations using DNA extracted from
blood or buccal swabs. Companion diagnostics are conducted to determine
somatic mutations using DNA and RNA from tissue biopsies. The US Food and
Drug Administrations Center for Drug Evaluation and Research have made a
similar distinction. The CDER have made recommendations on the use of phar-
macogenomic tests in conjunction with specific therapeutics such as warfarin
or clopidogrel. For many companion tests, the FDA has coapproved a drug
(e.g., trastusmab) requiring a positive test result (Her2/neu) for its labeled
v
vi Preface
therapeutic use. We have determined that companion diagnostic tests are

outside of the scope of this book.
San Francisco, CA Alan H.B. Wu
Chicago, IL Kiang-Teck J. Yeo
May, 2010
Acknowledgements
We like to thank all of our contributors for making the development and publica-
tion of this book possible. We also wish to thank the various companies, journals
who have granted us permission to use the figures in this book.
This book is dedicated to our wives, Pamela and Tet-Kin for their understanding
and selfless support of our careers and giving us the time to work on this book.
vii
Contents
Part I Basic Concepts
1 Issues in Translation of Pharmacogenomics

into Clinical Practice.................................................................................. 3
Kiang-Teck J. Yeo, Nikolina Babic, and Alan H.B. Wu
2 Molecular Diagnostic Methods in Pharmacogenomics.......................... 15

Nikolina Babic, Loren J. Joseph, and Kiang-Teck J. Yeo
3 Economics of Pharmacogenomic Testing in Clinical Practice............... 35

Alan H.B. Wu
4 From Personalized Medicine to Personalized Justice:

The Promises of Translational Pharmacogenomics
in the Justice System.................................................................................. 47
Steven H.Y. Wong, Christopher Happy, Daniel Blinka, Susan Goch,
Jeffrey M. Jensen, Joseph M. Donald, Howard Coleman,
Saeed A. Jortani, Yolanda Lurie, Cynthia L. Morris-Kukoski,
Manuela G. Newman, Paul J. Orsulak, Tara Sander, Michael A. Wagner,
Jennifer R. Wynn, Alan H.B. Wu, and Kiang-Teck J. Yeo
Part II Chemotherapeutics
5 Irinotecan.................................................................................................... 59
R. Stephanie Huang, Federico Innocenti, and Mark J. Ratain
6 Pharmacogenomics of Tamoxifen............................................................. 77
Christine L.H. Snozek, Alicia Algeciras-Schimnich,
Matthew P. Goetz, and Loralie J. Langman
7 Thiopurines................................................................................................. 91
Terreia S. Jones and Mary V. Relling
ix
x Contents
Part III Specific Pharmacogenomic Targets: Cardiovascular Drugs
8 The Pharmacogenetics of Vitamin K Antagonist

Anticoagulation Drugs............................................................................... 117
Charles Eby
9 Clopidogrel and Salicylates....................................................................... 139

Janice Y. Chyou and Marc S. Sabatine
10 Genotype-Guided Statin Therapy............................................................ 155

Richard L. Seip, Jorge Duconge, and Gualberto Ruao
11 The Statin Response Gene: KIF6............................................................. 175

H. Robert Superko, Tom White, James Forrester,
and Spencer King III
Part IV Drugs that Cause Delayed Hypersensitivity
12 Abacavir...................................................................................................... 201
Elizabeth J. Phillips and Simon A. Mallal
13 Allopurinol.................................................................................................. 213
Pei Chen, Shuen-Iu Hung, Shih-Yang Chen,
and Yuan-Tsong Chen
14 Carbamazepine and Its Structurally-Related Antiepileptics................. 225

Shuen-Iu Hung, Wen-Hung Chung, Jing-Jane Tsai,
and Yuan-Tsong Chen
Part V Miscellaneous Drugs
15 Pharmacogenetics of Flucloxacillin and Amoxicillin-Clavulanate

Associated Hepatic Dysfunction/Injury................................................... 239
Hong-Kee Lee and Lionel D. Lewis
16 Immunosuppressants Pharmacogenomics............................................... 249

Ping Wang
Index.................................................................................................................. 267
Contributors
Alicia Algeciras-Schimnich
Mayo Clinic, Rochester, MN, USA
Nikolina Babic
Department of Pathology, Pritzker School of Medicine, University of Chicago,
5841 South Maryland Avenue, Chicago, IL 60637, USA
Daniel Blinka
Marquette Law School, Milwaukee, WI, USA
Pei Chen
Institute of Biomedical Sciences, Academia Sinica, Taipei, Taiwan
Shih-Yang Chen
Center of Gout, Country Hospital, Taiwan
Yuan-Tsong Chen
Institute of Biomedical Sciences, Academia Sinica, Taipei 11529, Taiwan;
Department of Pediatrics, Duke University Medical Center, Durham, NC, USA
Wen-Hung Chung
Department of Dermatology, Chang Gung Memorial Hospital,
Chang Gung University College of Medicine, Taipei 10507, Taiwan
Janice Y. Chyou
Department of Medicine, Brigham and Womens Hospital, Harvard University,
Boston, MA, USA
Howard Coleman
Genelex Corporation, Seattle, WA, USA
Hon. Joseph M. Donald
Milwaukee County Circuit Court, Milwaukee, WI, USA
Jorge Duconge
Department of Pharmaceutical Sciences, School of Pharmacy,
University of Puerto Rico, San Juan, PR 00936-5067, USA
xi
xii Contributors
Charles Eby
Washington University in St. Louis, 660 South Euclid Avenue,
St. Louis, MO 63110, USA
James Forrester
Celera Inc., 1401 Harbor Bay Parkway, Alameda, CA 94502, USA
Susan Gock
Medical College of Wisconsin, Milwaukee, WI 53326, USA
Matthew P. Goetz
Christopher Happy
Medical College of Wisconsin, Milwaukee, WI 53326, USA
R. Stephanie Huang
Department of Medicine, Committee on Clinical Pharmacology
and Pharmacogenomics, and Cancer Research Center,
The University of Chicago, 900 E. 57th Street, KCBD Room 7148,
Chicago, IL 60637, USA
Shuen-Iu Hung
Institute of Pharmacology, National Yang-Ming University, Taipei 11221, Taiwan
Federico Innocenti
Department of Medicine, Committee on Clinical Pharmacology;
Pharmacogenomics, and Cancer Research Center, University of Chicago,
Chicago, IL 60637, USA
Jeffrey M Jentzen
University of Michigan, Ann Arbor, MI, USA
Terreia S. Jones
St. Jude Childrens Research Hospital, Memphis, TN, USA;
University of Tennessee Health Science Center, Memphis, TN, USA
Saeed A. Jortani
University of Louisville, Louisville, KY, USA
Loren J. Joseph
Department of Pathology, Pritzker School of Medicine,
University of Chicago, 5841 South Maryland Avenue, Chicago, IL 60637, USA
Spencer King III
Loralie J. Langman
Hong-Kee Lee
Dartmouth Medical School & Dartmouth-Hitchcock Medical Center, Lebanon,
NH 03756, USA
Contributors xiii
Lionel D. Lewis
Dartmouth Medical School & Dartmouth-Hitchcock Medical Center, Lebanon,
NH 03756, USA
Yolanda Lucire
Forensic and Medico-Legal Psychiatry, NSW 2027, Australia
Simon A. Mallal
Institute for Immunology & Infectious Diseases,
Murdoch University, Murdoch, WA, Australia;
Sir Charles Gairdner Hospital, Nedlands, Perth, WA, Australia;
Royal Perth Hospital, Perth, WA, Australia
Cynthia L. Morris-Kukoski
Federal Bureau of Investigation Laboratory, Quantico, VA, USA
Manuella G. Neuman
University of Toronto, ON, Canada
Paul J. Orsulak
Randox Corporation, Minneapolis, MN, USA
Elizabeth J. Phillips
Institute for Immunology & Infectious Diseases,
Murdoch University, Murdoch, WA, Australia;
Sir Charles Gairdner Hospital, Nedlands, Perth, WA, Australia;
Royal Perth Hospital, Perth, WA, Australia
Mark J. Ratain
Department of Medicine, Committee on Clinical Pharmacology;
Pharmacogenomics, and Cancer Research Center, University of Chicago,
5841 S. Maryland Ave., MC 2115 (Room I-217), Chicago, IL 60637, USA
Mary V. Relling
St. Jude Childrens Research Hospital, Memphis, TN, USA
and University of Tennessee Health Science Center, Memphis, TN, USA
Gualberto Ruao
Genetics Research Center, Hartford Hospital,
Hartford, CT 06106, USA
Marc S. Sabatine
Brigham and Womens Hospital, Harvard University, Boston, MA, USA
Tara Sander
Medical College of Wisconsin, PO Box 26509, Milwaukee, WI 53326, USA
Richard L. Seip
Genomas, Inc., Hartford Hospital, 67 Jefferson Street, Hartford, CT 06106, USA;
Division of Cardiology, Hartford Hospital, Hartford, CT 06102-5037, USA;
Genetics Research Center, Hartford Hospital, Hartford, CT 06102-5037, USA
xiv Contributors
Christine L.H. Snozek

H. Robert Superko
Jing-Jane Tsai
Department of Neurology, College of Medicine, National Cheng Kung University,
Tainan, Taiwan
Michael A. Wagner
Indiana University School of Medicine, Indianapolis, IN, USA
Ping Wang
The Methodist Hospital, Houston, TX, USA
Tom White
Steven H.Y. Wong
Wake Forest University School of Medicine, Department of Pathology,
Winston-Salem, NC 27157, USA
Alan H.B. Wu
Department of Laboratory Medicine, University of California,
San Francisco, CA 94110, USA
Jennifer R. Wynn
LaGuardia Community College and John Jay College
of Criminal Justice, New York, NY, USA
Kiang-Teck J. Yeo
Part I
Basic Concepts
Chapter 1
Issues in Translation of Pharmacogenomics
into Clinical Practice
Kiang-Teck J. Yeo, Nikolina Babic, and Alan H.B. Wu
Keywords Pharmacokinetics Pharmacodynamics Pharmacogenomics Barriers

Clinical laboratory improvement act
1.1Clinical Laboratory Tests
Clinical laboratory tests are commonly used for supporting the diagnosis and
prognosis of a disease, monitoring the efficacy progress of therapeutic manage-
ment decisions, and measuring the foreign toxins. Recent advances in genomics
have resulted in the generation of new assay panels to support personalized drug
therapy this offers new opportunities for clinical laboratories to make important
contributions to healthcare. Pharmacogenomics (PGx) is a recognized discipline
within pharmacology that involves testing relevant human genes, whose products
are involved with the inter-individual variability of a drugs pharmacokinetics,
pharmacodynamics, and human leukocyte antigen (HLA) system profile. The
effective use of PGx testing promises to improve the therapeutic efficacy of drugs
while reducing the incidence and severity of adverse drug effects, and drive the
optimum drug selection for therapy.
The desire to improve the therapeutic index of drugs stems from the knowledge
that the efficacy of the most widely used FDA-approved drugs averages around
50% with a range of 25% for chemotherapeutics and up to 80% for analgesics [1],
and incidence of serious adverse drug reactions (ADRs) estimated to be two
million per year in the US and cause 100,000 deaths [2]. In a prospective study of
hospital admissions caused by ADRs at two large UK hospitals, Pirmohamed etal.
found a prevalence of 6.5% with a projected annual cost of $847 million [3].
Parts of this chapter was published in Wu etal. [12] with permission from Future Medicine Ltd.,
http://www.futuremedicine.com/loi/pgs.
K.-T.J. Yeo()
e-mail: jyeo@bsd.uchicago.edu
A.H.B. Wu and K.-T.J. Yeo (eds.), Pharmacogenomic Testing in Current Clinical Practice, 3
Molecular and Translational Medicine, DOI 10.1007/978-1-60761-283-4_1,
4 K.-T.J. Yeo et al.
The Institute of Medicine was instructed by the US Congress to perform a

comprehensive study of drug safety and to find means to prevent or reduce medi-
cation errors [4]. The degree by which ADRs can be reduced through PGx testing
is currently unknown. Errors associated with incorrect scripting or dispensing of
prescriptions, or by patients not following a physicians dosing recommendations
will not be corrected by PGx testing. However, adverse events due to an incorrect
prescribed dose or the use of a therapeutic agent that is not optimal for a particular
individual may be minimized by PGx testing. For example, the implementation of
pharmacogenomic testing for abacavir [5] or carbamazepine [6] promises to
greatly reduce the incidence of potentially life-threatening, delayed cutaneous
hypersensitivity reactions.
There are major ongoing research efforts conducted by academia and the phar-
maceutical industry in discovering new gene associations that better predict indi-
vidual drug handling and the effect (therapeutic and toxic) of existing and novel
therapeutics. The US Food and Drug Administration (FDA), in its commitment to
advancing personalized medicine, is taking a leadership role in drafting guidance
documentation for companion diagnostics. In some cases, the need for a companion
laboratory test (i.e., the need for a certain test result before a drug is given) will be
seen as an impediment to the implementation of that drug, as the pharmaceutical
industry will naturally opt to develop safer and more efficacious alternatives that do
not require companion testing. In other cases, PGx tests may well help salvage
drugs that might not normally receive FDA approval or perhaps even those that
have been removed from the market, if the toxicity profile can be minimized with
targeted testing (Table 1.1). In the end, perhaps only a relatively small cadre of cur-
rent and future drugs will require PGx testing just as today, only a few drugs require
regular therapeutic drug monitoring. Therefore, important studies and decisions lie
Table 1.1 Examples of drugs undergoing pharmacogenomics rescue

Therapeutic Potential PGx
Drug/company Class indications Goal biomarker FDA status
Bucindolol/ b Blocker Heart failure Efficacy b1-adrenergic Rejected in June
Arca receptor poly 2009, resubmis
Biopharma morphism sion in progress
Lumiracoxib/ COX-2 Arthritis Safety MHC Class II Never approved in
Novartis analgesics genes US, pulled from
foreign markets
in 2007
Tremelimumab/ Cancer Melanoma Safety Genome studies Conducting a
Pfizer in progress second drug trial
after finding
a biomarker
that identifies
patients who
benefit from the
drug
Adapted from http://personalizedmedicinecoalition.org/membership/newsletter/PMC-Newsletter-
Spring10.pdf Accessed 5/4/2010
1 Issues in Translation of Pharmacogenomics into Clinical Practice 5
ahead with regard to which drugs are most appropriate for such testing, how testing
is to be conducted and reimbursed, how PGx tests results are reported and inter-
preted, and demonstrating the medical and/or economic benefits of testing.
1.2Terminology and Definitions
Pharmacogenetics is not a new discipline, and the importance of inheritance in the

role of drug response was known since the 1950s as demonstrated by genetic
variation of pseudocholinesterase on succinylcholine effects [7] and N-acetyl trans-
ferase 2 (NAT2) polymorphisms on isoniazid [8] metabolism as reviewed by
Weinshilboum [9]. However, most PGx studies in the past employed phenotypic
markers of genetic variation (e.g., drug/metabolite concentrations in blood or urine,
direct enzyme activities measurements), so advances in this field were somewhat
limited and slow. With the completion of the Human Genome Project in 2001 [10],
there was a re-birth and accelerated discovery in the field of pharmacogenetics
and pharmacogenomics, since we now have a better and more comprehensive
genetic knowledge base and new molecular technologies to study precise genotypic
variations and correlate them to phenotype. Vogel [11] introduced the term phar-
macogenetics, which can be defined as the study of the role of a small number of
genetic variations that are relevant to a drugs disposition or effect. Pharmacogenomics
involves the study of a larger collection of genomic factors that contribute to the
individual variability of drug responses. This may include genes that regulate phase
I oxidative drug metabolism (especially the cytochrome P450 family of enzymes),
phase II drug conjugation enzymes (e.g., glucuronosyltransferases and
N-acetyltransferases), drug transporter proteins (e.g., organic anion transporters),
and drug targets-enzymes or receptors [9]. In the past, the variations observed for
most drugs in terms of utilization and adverse events were limited to only a few
SNPs within a gene (monogenic traits) and pharmacogenetics may be the pre-
ferred term. However, for warfarin and clopidogrel, multiple SNPs within several
genes (polygenic traits) have been determined to influence the drugs therapeutic
effects, and therefore the term pharmacogenomics may be preferred. Despite
these definitions, these terms are often used interchangeably and the acronym
PGx is used commonly.
1.3Current Barriers
The term personalized medicine was coined in the late 1990s and relates to
providing the right dose to the right person at the right time using genetic markers
of drug metabolism, transport, and receptor interactions. While this is a very attrac-
tive concept that was widely covered in the lay press, the implementation of
pharmacogenomics at the bedside or in the physicians office has been slow. The
barriers to wider adoption and implementation include (a) lack of education and
understanding by prescribing physicians regarding available pharmacogenetictests;

(b) lack of consensus guidelines on interpretation and the use of pharmacogenetic
results; (c) paucity of randomized controlled trials demonstrating the clinical utility
of such testing; (d) lack of studies demonstrating the cost-effectiveness of using
pharmacogenetic markers; (e) lack of specific reimbursement codes; (f) lack of com-
mercial assays; (g) lack of standardization in a given PGx panel, reporting, and
interpretation of results; and (h) tension between patients awareness of media/web-
based pharmacogenetic testing and physicians readiness to adopt patient-driven
testing (reviewed extensively by Wu et al. [12]). However, there are additional
current challenges like the Center for Medicaid and Medicare Services (CMS)
recent ruling of not supporting reimbursement of warfarin sensitivity genotyping due
to a lack of current available evidence for improved health outcomes for Medicare
beneficiaries [13], which has resulted in a temporary setback in adoption.
1.4Potential Solutions
1.4.1FDA Initiatives
Since physicians and patients find no value in the genotyping per se, PGx testing
results must be put into the context of actionable decisions. The FDA continues
to lead the efforts to translate recent advances in PGx into clinical practice. In this
respect, the FDA has approved revised labeling requirements for selected
drugswhere polymorphisms have been linked to either a reduction in drug efficacy
or an increased incidence of adverse events (Table 1.2). This drug re-labeling initia-
tive now raises a medico-legal implication in cases where an ADR has occurred.
For example, if a patient suffers an adverse reaction that is directly linked to the use
of a drug, and the prescribing physician did not know that the individual was
predisposed to that adverse event as the result of a genetic polymorphism, that
physician may find it difficult to defend a medical malpractice lawsuit, given that a
warning for that event was present in the drugs label [14]. These recent drug
re-labeling recommendations are likely to drive the need to utilize PGx testing for
optimal prescribing to a defensive practice of medicine. Irrespective of the driver,
when implemented, proper utilization of PGx promises to incrementally, and in
some cases, substantially improve the safe use therapeutics.
1.4.2Clinical Trials
Because of the ongoing debate about the clinical usefulness of warfarin genotyping
a poster child case for PGx a recent multi-centered, double-blind, randomized
control trial the Clarification of Optimal Anticoagulation through Genetics (COAG)
was launched. This study seeks to evaluate the efficacy in the use of clinical and
Table 1.2 US Food and Drug Administration relabeling initiative

Drug Enzyme Goala Year Status
6-MP b
TPMT c
Safety 2003 Completed
Azathioprine TPMT Safety 2003 Completed
Atomoxetine CYPd 2D6 Safety 2004 Completed
Irinotecan UGT1A1e Safety 2004 Completed
Warfarin CYP 2C9, VKORC1f Safety 2007 Completed
Abacavir HLA-B*5701g Safety 2007 Completedh
Allopurinol HLA-B*5801 Safety 2007 Completed
Carbamazepine HLA-B*1502 Safety 2007 Completedi
Phenytoin and Fosphenytoin HLA-B*1502 Safety 2008 Completedi
Clopidogrel CYP 2C19 Efficacy 2009 Completed
Tamoxifen CYP 2D6 Efficacy 2006 Pending
a
Safety goals indicate a potential for reducing drug-induced adverse events. Efficacy goals
refer to the potential for improving the effectiveness of the drug
b
6-Mercaptopurine
c
Thiopurine methyltransferase
d
Cytochrome
e
UDP-glucuronosyltransferase
f
Vitamin K epoxide reductase complex 1
g
Human lymphocyte antigen
h
Department of Health and Human Services recommendation made to genotype Caucasian
patients prior to Abacavir therapy
FDA recommendation made to genotype Asian patients prior to carbamzepine or phenytoin
i
therapy
genetic information to guide warfarin therapy initiation and improve anticoagulation

control for patients. The trial is sponsored by the National Heart, Lung, and Blood
Institute and will recruit 1,200 patients initiated on warfarin among 12 different
medical centers in the US and is estimated to take up to 4 years to complete [15].
At the University of Chicago, we are currently performing The Clinical and Economic
Implications of Genetic Testing for Warfarin Management Trial funded by the Agency
for Healthcare Research and Quality. The aims of this study are (a) to set up a genetic
registry of a racially diverse set of patients undergoing warfarin therapy; (b) to deter-
mine the efficacy, costs and cost-effectiveness of existing pharmacogenetic algo-
rithms for the management of warfarin therapy among hospitalized patients; and
(c)to develop a clinical pharmacogenetic algorithm for the management of warfarin
therapy among hospitalized African American patients [16].
With regard to another widely used antithrombotic the antiplatelet drug, clopi-
dogrel, two recent large clinical studies showed that individuals carrying CYP2C19
loss-of-function alleles (e.g., *2,*3,*4, or *5) had significantly lower concentra-
tions of the active drug metabolite, reduced platelet inhibition, resulting in threefold
increase in risk of stent restenosis, and a 3.6-fold increase in rate of cardiovascular
events [17, 18]. Because of these accumulating evidences, the FDA has recently
issued a black-box warning to health-care professionals to consider CYP 2C19
genotyping so as to identify patients who are poor metabolizers of clopidogrel; this
allows providers to act on this information to consider alternative antiplatelet
edications or alternative dosing strategies for clopidogrel [19]. These and several
m
ongoing clinical trials when completed may contribute significantly to the accumu-
lating evidence that relevant PGx testing can lead to effective tailoring of target
drug therapies to avoid ADRs, while maximizing efficacy.
1.4.3Cost-Effectiveness Considerations
A key solution for uptake of PGx is to have studies demonstrating the cost-
effectiveness of such testing. Detailed description of this aspect is covered in
Chap. 3. A recent collaborative study by Mayo and Medco, involving 900 patients
on warfarin, showed that hospitalization rates decreased by 30% when genotyping
information was used in warfarin dosing versus the usual clinical dosing [20].
This should translate to cost-savings that would easily justify the costs of phar-
macogenotyping, which continues to decrease with technological advance.
1.4.4PGx Algorithms, Reports, and Interpretations
Complex genotyping results need to be readily translated into actionable decisions

by providers to encourage wider use of testing. For example, warfarin sensitivity
genotyping results need to be integrated and made somewhat user-friendly by using
a PGx report as shown in Fig. 1.1. To further make it actionable, one needs to con-
solidate these results into a therapeutic output, which in this case, is the calculated
dose based on the measured polymorphisms and relevant clinical factors. Thus, it
would be important to have reliable and validated algorithms that can be employed
for this purpose. With regard to warfarin, recently the International Warfarin
Pharmacogenetics Consortium developed a warfarin dosing algorithm based on
genetic and clinical information and showed that in a validation cohort of 1,009
patients, the combined algorithm identified a larger fraction of patients requiring
extreme doses (49 mg/week and 21 mg/week) than the clinical algorithm alone
[21]. A commonly used web-based warfarin dosing algorithm that employs clinical,
co-medications, and genetic polymorphisms information (CYP 2C9, CYP 4F2, and
VKORC1) to calculate the appropriate warfarin dose can be found at the following
URL: http://warfarindosing.org/Source/InitialDose.aspx (accessed May 15, 2010).
1.4.5Laboratory Aspects and Guidelines
All clinical laboratory tests, including PGx tests, are regulated by the Clinical
Laboratory Improvement Act (CLIA) of 1988 and have to be performed in
certified laboratories. CLIA certification can be obtained by a laboratory getting
accreditation from an organization like College of American Pathologists, which
Patient Name Medical Record Age Sex
Number
Ordering Physician DOB
Attending Physician Report Notes
Collected Printed
Estimated
Test Allele Result Interpretation
Warfarin Dose*
Loading dose
(mg/d):________
430 C>T (*2) C/T

Reduced warfarin
CYP450 2C9 Therapeutic dose
metabolism expected
1075 A>C (*3) A/C (mg/d):________
Reviewed by:
Normal warfarin _______________

VKORC1 -1639 G>A G/A
sensitivity expected Pharm. D.
*Estimated warfarin dose is calculated and reviewed by the UCMC Clinical Pharmacist.
The dose is calculated using the algorithm published on http://www.warfarindosing.org website.
Fig. 1.1 Sample report for warfarin sensitivity genotyping
has been given deemed status by the Centers for Medicare and Medicaid Services.
Documentation of personnel qualifications and training, test proficiency, satisfactory
quality control, and validation of assays are critical components for accreditation.
PGx assays are classified as high-complexity tests since all of them require
extraction of DNA from blood (or oral fluids), and for many a requirement of
amplification by polymerase chain reaction followed by detection on beads, silicon
chips or solution-based systems. Before a PGx test can be released for clinical
use, each laboratory must demonstrate genotyping accuracy by validating against
a predicate assay cleared by the FDA or confirm with some definitive techniques
such as bi-directional sequencing method. For less common variants, the
existence of validated DNA sources containing such polymorphisms will greatly
facilitate the implementation of these methods, increasing the accessibility of
such technologies in the clinical laboratories. In addition, the clinical laboratory
Table 1.3 Current FDA-cleared pharmacogenomics assays

Test Manufacturer Testing platform Approval date Target drug
CYPa 2D6 Roche Amplichip 1/2005 Not specified
CYP 2C19 Roche Amplichip 1/2005 Not specified
UGT1A1b Hologic Invader 9/2005 Irinotecan
CYP2C9/VKORC1c Nanosphere Verigene 9/2007 Warfarin
(6484 allele)
CYP2C9/VKORC1 Autogenomics Infiniti 1/2008 Warfarin
(3673 allele)
CYP2C9/VKORC1 Paragon RT-PCR 5/2008 Warfarin
(6484 allele)
CYP2C9/VKORC1 GenMark eSensor 7/2008 Warfarin
(3673 allele)
cytochrome
a
b
UDP-glucuronosyltransferase
Vitamin K epoxide reductase complex 1
c
must also be ready to provide reliable PGx results within the required turnaround
time for efficacious use of the information. Inter-disciplinary collaboration
between the laboratory, clinical pharmacologist/pharmacist, and the provider is
essential to insure proper interpretations and integration of PGx results for timely
decision-making regarding dosing or alternative drug regimen. There are now
several commercial PGx multiplex assay platforms, several of which are FDA-
cleared for CYP 2D6, CYP 2C19, UGT1A1, CYP 2C9/VKORC1 (Table 1.3).
However more extensive PGx panels for some like CYP 2C19, CYP 2C9/VKORC1
are usually available as research use only or investigational use only products in
this case individual laboratory may have to further optimize the assay and validate
them independently, much like laboratory developed tests [12, 22].
The National Academy of Clinical Biochemistry has recently published the
2010 PGx guidelines that specify the requirements for test validation, quality con-
trol, and proficiency testing [23]. Furthermore, the GeT-RM collaborative project
sponsored by Center of Disease Control has recently validated a series of Coriell
DNA from human cell lines that contained relevant PGx variants that can serve as
a ready source of quality control materials (Table 1.4) [24]. Collectively, these
multi-center collaborative efforts will go a long way in facilitating the uptake of
PGx by the clinical laboratory, and by extension the clinical community, by making
it easy to order, interpret, and personalize drug therapy according to the relevant
genetic and clinical variables of an individual patient.
1.4.6Reimbursements
At present, there are no specific American Medical Association Current Procedural

Terminology (CPT) reimbursement codes available for any PGx test. Currently,
many laboratories bill according to the individual procedures required to produce a
Table 1.4 Genetic testing reference material coordination programa

Coriell cell line VKORC1
number CYP2C19 CYP2C9 c.-1639G>A CYP2D6
GM17289 *2/*2 *1/*1 AA *2/*4
GM17203 (*1/*2) *2/*17 *1/*1 AA *4/*35
GM17272 (*1/*1) *17/*17 *1/*1 AA *4/*10
GM17246 (*1/*8) *8/*17 *1/*2 GA *4/*35
GM17115 *1/*1 (*1/*1) *9/*9 GG *1/*2
GM10005 (1/*1) *1/*17 (*1/*1) *1/*9 GG *17/*29
GM07439 (*2/*2) *2/*10 (*1/*1) *1/*9 GG *4xN/*41
GM 12244 *1/*1 *2/*3 GG *35/*41
GM17052 *1/*3 *1/*1 AA *1/*1
GM02016 *1/*2 *1/*1 GG *2xN/*17
GM17296 (*1/*1) *17/*17 *1/*1 GA *1/*9
Pratt etal. Partial data from manuscript submitted to Journal of Molecular Diagnostics [24]
a
test result (DNA extraction, amplification, analysis of the DNA by probes, mutation
scanning, and interpretation). Each of the specific procedure codes can be multi-
plied by the number of alleles interrogated, e.g., CPT 83914 (for allele-specific
primer extension) 6. However, the lack of specific reimbursement CPT codes is
a real deterrent for clinical laboratory to implement such PGx tests as hospitals are
concerned about the uncertainties surrounding the reimbursement of these new
tests. The recent passage of the landmark healthcare reform, which contained prin-
ciples of personalized medicine, will hopefully speed up the adoption of pharma-
cogenomics by fostering better exchanges between the various governmental
agencies (CMS, FDA, NIH) to address barriers and offer solutions to translate these
discoveries to clinical practice.
Various efforts are underway by nongovernment, nonprofit organization like the
Personalized Medicine Coalition (PMC) to engage CMS to offer guidance on the
scientific evidence needed for CMS to approve reimbursement for a given genetic
test. PMC is a broad coalition of academic, industrial, patient, provider, and payer
communities with a mission to advance the understanding and adoption of person-
alized medicine for the benefit of the patient [25].
1.4.7Education, Training, and Pharmacogenomics Resources
A very important solution to the adoption of PGx testing is the education of all
health-care professionals regarding the basic principles of pharmacogenomics and
its emerging role in Personalized Medicine. There are many recent reviews on the
various aspects of pharmacogenomics. Weinshilboum etal. described the evolution
of pharmacogenomics as a science and summarized several of the scientific
advances that are keys to this field [9]. Roden etal. assembled a glossary of terms
and detailed the approaches toward discovery of new gene associations [26].
Because drug metabolism is a critical aspect of pharmacogenomics, Wilkinson
reviewed the CYP-450 system and how polymorphisms interact with the
pharmacokinetics of therapeutic drugs [27]. Bolonna etal. examined the potential
for pharmacogenomic testing focusing on psychiatric drugs [28], Yong et al.
focused on the cancer drugs [29], and Carlquist on cardiovascular disease drugs
[30]. For research, there are tools and consortia available to assist in the discovery
of association of drug response with genetic variation. The Pharmacogenetics
Research Network is a collaborative group of investigators who have been orga-
nized into five major clinical areas: asthma, depression, cardiovascular disease,
drug addiction, and cancer [31]. For the rapid dissemination of data, there is also a
publicly available knowledge base that collates information describing the relation-
ship between drugs, diseases, and genetic variation; this is the Pharmacogenomics
Knowledge Base [32]. Medical and graduate schools are beginning to incorporate
pharmacogenomics in their pharmacology curricula. Even more encouraging is the
recent availability of a web-based tool, called DNA Twist, to educate middle and
high school students about the world of pharmacogenomics [33, 34]. These educa-
tional efforts should produce a new generation of students who are facile with the
use of genomics, proteomics, and metabolomics (and other future -omics) in
future applications of personalized medicine.
1.5Conclusion
Personalized medicine is currently attracting a lot of attention in the media, and the
general public is becoming more aware of the promises of the genomic revolution.
Web-based direct-to-consumer DNA testing services abound that promises a wealth
of genetic information (costing less than $1,000) and are now widely marketed to
the average consumer as tools of empowering personal choices about your health
and well-being. Pharmacogenomics is one aspect of personalized medicine that will
continue to gain ground and make inroads into clinical practice. Ultimately, a large
part for the success of wider adoption of pharmacogenomics will depend on a
multidisciplinary team approach. The stakeholders in this process include depart-
ments of clinical pharmacology/pharmacy, medical genetics, laboratory medicine,
and researchers in the field of genetics and pharmacogenomics. Medical specialties
must also be invested in this where appropriate, e.g., oncology, cardiology, and
psychiatry. A major component of a successful program will be the education of
physicians, who use the medications for which pharmacogenomics data will be
relevant, and the incorporation of clinical pharmacogenomics in the physician
training curricula both during and after medical school.
Due to rapid advances in genomics, proteomics, and metabolomics, we predict
that there will be an increasing need for a clinical pharmacogenomics service the
clinical laboratory should take the lead in making these tests available when
sufficient evidence is established for clinical usefulness. Most importantly, these
complex results need to be made user-friendly to the clinical community (e.g., via
improved information technology) before widespread adoption can happen.
With this revolution underway, new opportunities are now available for medical
students, graduate students, medical technologists, residents, and fellows in patho
logy and laboratory medicine to become specialist in this area of personalized
medicine.
References
1. Spear, B. B., Heath-Chiozzi, M., & Huff, J. (2001). Clinical application of pharmacogenetics.
Trends in Molecular Medicine, 7, 201204.
2. Lazarou, J., Pomeranz, B. H., & Corey, P. N. (1998). Incidence of adverse drug reactions in
hospitalized patients: A meta-analysis of prospective studies. The Journal of the American
Medical Association, 279, 12001205.
3. Pirmohamed, M., James, S., Meakin, S., Green, C., Scott, A. K., Walley T. J., et al. (2004).
Adverse drug reactions as cause of admission to hospital: Prospective analysis of 18 820
patients. British Medical Journal, 329, 1519.
4. Aspden, P., Wolcott, J., Bootman, J. L., & Cronenwett, R. L., (Eds.). (2007). Preventing medi-
cation errors: Committee on identifying and preventing medication errors. Washington:
National Academic Press.
5. Mallal, S., Phillips, E., Carosi, G., Molina, J.-M., Workman, C., Tomazic, J., et al. (2008).
HLA-B*5701 screening for hypersensitivity to abacavir. The New England Journal of
Medicine, 358, 568579.
6. Chung, W. H., Hung S. I., Hong, H. S., Hsih, M. S., Yang, L. C., Ho, H. C., et al. (2004).
Medical genetics: A marker for Stevens-Johnson syndrome. Nature, 428, 486486.
7. Kalow, W. (1990). The pennsylvania state university college of medicine 1990 Bernard B.
Brodie lecture: Pharmacogenetics: Past and future. Life Sciences, 47, 13851397.
8. Evans, D., Manley, K., & McKusick, V. (1960). Genetic control of isoniazid metabolism in
man. British Medical Journal, 2, 485491.
9. Weinshilboum, R. M., & Wang, L. (2006). Pharmacogenetics and pharmacogenomics:
Development, science, and translation. Annual Review of Genomics and Human Genetics,
7,223245.
10. Collins, F. S., Morgan, M., & Patrinos, A. (2003). The human genome project: Lessons from
large-scale biology. Science, 300, 286290.
11. Vogel, F., & Jager, P. (1969). The genetic load of a human population due to cystostatic agents.
Humangenetik, 7, 287304.
12. Wu, A. H. B., Babic, N., & Yeo, K. T. J. (2009). Implementation of pharmacogenomics into
the clinical practice of therapeutics: Issues for the clinician and the laboratorian. Personalized
Medicine, 6, 315327.
13. Jensen, T. S., Jacques, L. B., Ciccanti, M., Long, K., Eggleston, L., & Roche, J. (2010).
Proposed decision memo for pharmacogenomic testing to predict warfarin responsiveness.
Accessed May 17, 2010, from http://www.cms.gov/mcd/viewdraftdecisionmemo.asp?from2=
viewdraftdecisionmemo.asp&id=224&
14. Wong, S. H., Happy, C., Blinka, D., Gock, S., Jentzen, J. M., Donald, Hon. J., et al. (2010).
From personalized medicine to personalized justice: The promises of translational pharmaco
genomics in the justice system. Pharmacogenomics, 11(6), 731737.
15. Rosenberg, Y., & Schron, E. (2010). Clarification of optimal anticoagulation through genetics
(COAG); A randomized, multicenter, double-blind clinical trial to evaluate efficacy in the use of
clinical plus genetic information to guide warfarin therapy initiation and improve anticoagula-
tion control for patients. Accessed May 15, 2010, from http://coagstudy.org
16. Meltzer, D. O., & Marlow E. (2010). Clinical and economic implications of genetic testing
for warfarin management. Accessed May 15, 2010, from http://www.clinicaltrialssearch.org/
clinical-and-economic-implications-of-genetic-testing-for-warfarin-management-nct00964353.
html
17. Simon, T., Verstuyft, C., Mary-Krause, M., Quteineh, L., Drouet, E., Meneveau, N., et al.
(2009). Genetic determinants of response to clopidogrel and cardiovascular events. The New
England Journal of Medicine, 360, 363375.
18. Mega, J. L., Close, S. L., Wiviott, S. D., Shen, L., Hockett, R. D., Brandt, J. T., et al. (2009).
Cytochrome P-450 polymorphisms and response to clopidogrel. The New England Journal of
Medicine, 360, 354362.
19. U.S. Food and Drug Administration (2010). Reduced effectiveness of plavix (clopidogrel) in
patients who are poor metabolizers of the drug. Accessed May 15, 2010, from http://www.
fda.gov/drugs/drugsafety/PostmarketDrugSafetyInformationforPatientsandProviders/
ucm203888.htm
20. Epstein, R. S., Moyer, T. P., Aubert, R. E., OKane, D. J., Xia, F., Verbrugge, R.R., et al.
(2010). Warfarin genotyping reduces hospitalization rates. Journal of the American College of
Cardiology, 55, 28042812
21. The International Warfarin Pharmacogenetics C. (2009). Estimation of the warfarin dose with
clinical and pharmacogenetic data. The New England Journal of Medicine, 360, 753764.
22. Babic, N., Haverfield, E. V., Burrus, J. A., Lozada, A., Das, S., & Yeo, K. T. J. (2009).
Comparison of performance of three commercial platforms for warfarin sensitivity genotyping.
Clinica Chimica Acta, 406, 143147.
23. Valdes, R., Payne, D., & Linder, M. (2010). Laboratory Medicine Practice Guidelines. Laboratory
analysis and application of pharmacogenetics to clinical practice. Accessed November 26,
from http://www.aacc.org/members/nacb/LMPG/OnlineGuide/PublishedGuidelines/LAACP/
Pages/default.aspx
24. Centers for Disease Control and Prevention (2010). Genetic testing reference materials
coordination program (GeT-RM). Accessed November 26, http://wwwn.cdc.gov/dls/genetics/
rmmaterials/MaterialsAvailability.aspx
2 5. Personalized Medicine Coalition. Accessed May 15, 2010, from http://www.personal-
izedmedicinecoalition.org/
26. Roden, D. M., Altman, R. B., Benowitz, N. L., Giacomini, K. M., Johnson, J. A., Krauss, R. M.,
et al. (2006). Pharmacogenomics: Challenges and opportunities. Annals of Internal Medicine,
145, 749757.
27. Wilkinson, G. R. (2005). Drug metabolism and variability among patients in drug response.
The New England Journal of Medicine, 352, 22112221.
28. Bolonna, A. A., Arranz, M. J., Mancama, D., & Kerwin, R. W. (2004). Pharmacogenomics
can genetics help in the care of psychiatric patients? International Review of Psychology, 16,
311319.
29. Yong, W. P., & Innocenti, F. (2007). Translation of pharmacogenetic knowledge into cancer
therapeutics. Clinical Advances in Hematology & Oncology, 5, 698706.
30. Carlquist, J. F. (2007). Cardiovascular therapy and pharmacogenetics in 2007.
Pharmacogenomics, 8, 2123.
31. Giacomini, K. M., Brett, C. M., Altman, R. B., Benowitz, N. L., Dolan, M. E., Flockhart, D. A.,
et al. (2007). The Pharmacogenetics Research Network: From SNP discovery to clinical drug
response. Clinical Pharmacology and Therapeutics, i 81, 328345.
32. Pharmacogenomics knowledge base. Accessed May 17, 2010, from http://www.pharmgkb.org/
33. Altman, R., Schwartz, D., Berlin, D., Mittal, A., Person, M., Oppezzo, M., et al. (2010). DNA
twist A web-based tool of pharmacogenomics. Accessed May 17, 2010, from http://www.
dnatwist.org/BDE/index.html
34. Berlin, D. S., Person, M. G., Mittal, A., Oppezzo, M. A., Chin, D. B., Starr, B., et al. (2010).
DNATwist: a Web-based tool for teaching middle and high school students about pharmaco
genomics. Clinical Pharmacology and Therapeutics, 87, 393395.
Chapter 2
Molecular Diagnostic Methods
in Pharmacogenomics
Nikolina Babic, Loren J. Joseph, and Kiang-Teck J. Yeo
Keywords Real-time polymerase chain reaction Pyrosequencing Mass

spectrometry Microarray Allele-specific primer extension
2.1Introduction
Rapid advances in pharmacogenomics research have facilitated the transfer of

pharmacogenomics testing into clinical laboratories. In the past several years, the
US Food and Drug Administration (FDA) has begun to recognize the importance
of genetic information and has required advisories on drug labels seeking to inform
physicians and patients about the availability of genetic tests to guide drug dosing
and prescriptions [1]. This has furthered the desire to adopt this testing into clinical
laboratories. At this time, FDA has required manufacturers of 11 drugs to modify
their labels to include information on pharmacogenomics testing [2]. With the
advances in pharmacogenomics and information technology fields, a new concept
of genomic testing has also emerged on the market a concept of direct-to-consumer
DNA testing. Web-based companies are in the market with the goal of enabling a
patient to take control of his/her own genetic testing and results. One such company,
23 & Me (www.23andme.com, accessed 04/08/10), offers a variety of genetic tests,
including pharmacogenomic tests for warfarin sensitivity, 5-FU toxicity, clopi-
dogrel efficacy, and abacavir hypersenstivity. Traditional clinical laboratories are
also expanding their repertoire of pharmacogenomic testing. Therefore, it is clear
that there is an increased need for rapid, reliable, and cost-effective genotyping
methodologies amenable to easy adoption by the clinical laboratory community.
At present, the most frequently genotyped targets are the cytochrome P450 (CYP
450) super-family and phase II drug-metabolizing enzymes such as uridine diphos
N. Babic(*)
Department of Pathology, Pritzker School of Medicine, The University of Chicago,
e-mail: nbabic@bsd.uchicago.edu
16 N. Babic et al.
Table 2.1 Different SNP analysis approaches

PCR
Method Mechanism requirement Detection
Hybridization
Microarray Solid phase-based Yes Fluorescence;
electrochemical signal
Verigene Signal amplification No End point or real-time
Invader technology fluorescence detection
Real-time PCR Hybridization probes Yes Fluorescence
TaqMan probes
Molecular beacons
FRET probes
Sequencing
Sequencing Chain terminating sequencing Yes Capillary electrophoresis
(Sanger method) fluorescence
Pyrosequencing Sequencing by synthesis Yes Light
Mass spectrometry Mass differentiation of single Yes Mass spectrometry
or multiple analytes
Adapted from Wang etal. [3]
phate-glucyronosyltransferase 1A1 (UGT1A1), thiopurine S-methyltransferase

(TPMT), and N-acetyltransferase (NAT).
A number of commercial platforms are now available for pharmacogenomics
testing (Table 2.1). As shown in Table 2.1, most approaches require PCR
amplification of DNA template with different single nucleotide polymorphism
(SNP) detection and identification strategies. The exceptions are Verigene
(Nanosphere, Northbrook, IL) and Invader (Hologic, Madison, WI) technologies,
which are based on the principle of signal amplification. The genetic variants can
be detected using hybridization, real-time PCR, sequencing, or mass spectrometry
methods that involve extension of specific primers designed to detect SNPs of
interest. Hybridization is generally utilized in microarray platforms, and it involves
capture oligonucleotides immobilized on a solid surface that hybridize to the
labeled primer extension products. In real-time PCR, as the reaction progresses,
the fluorescence is generated and detected in real time. Sequencing utilizes labeled
chain-terminating nucleotides that are incorporated during primer extension.
Labeled fragments are then separated by electrophoresis and identified based on
size. In mass spectrometry, primer extension products are detected based on dif-
ferential mass-to-charge ratios between wild type and variant strands. The most
frequently used approaches currently employed are microarrays and real-time
PCR.
This chapter provides an overview of different platforms and technologies avail-
able, specifically focusing on their applications to pharmacogenomics testing.
Virtually all methodologies described here have other applications in genetic and
microbiology testing.
2 Molecular Diagnostic Methods in Pharmacogenomics 17
2.2DNA Extraction and Quantitation
DNA quality is critical for successful genotyping regardless of the platform or

methodology used. The classical method of DNA extraction involves cell lysis, followed
by organic (usually chloroform/isoamyl alcohol) extraction and ethanol precipitation.
Although this method is reliable, it is also time-consuming and cumbersome. Many
satisfactory DNA isolation kits are available on the market today. Some examples of
commercially available kits include QIAamp DNA Mini Kit (Qiagen Inc., Valencia,
CA), based on solid-phase extraction technology, MagMax kit (Applied Biosystems,
Foster City, CA) and Maxwell 16 system (Promega Corporation, Madison, WI), which
employ magnetic bead technology to extract DNA from the sample. These methods are
straight-forward and can be used either for manual or automated DNA extraction.
Manufacturers generally provide specific, easy to follow protocols for each kit.
The most common method for DNA quantitation is absorbance measurement at
260 nm, since DNA absorbs UV light at this wavelength. For a pure double-
stranded (ds)DNA, an absorbance of 1.0 at 260 nm corresponds to a DNA concen-
tration of 50 mg/ml. Although this method is sensitive to interferences from
nucleotides, single-stranded (ss)DNA, RNA, and proteins, the amount of these
potential interferents is minimal in most DNA preparation methods and does not,
therefore, present a significant problem.
DNA purity is assessed by determining the ratio of DNA absorbance (260 nm)
to protein absorbance (280 nm). DNA with A260/280 ratios between 1.7 and 1.9
should be sufficiently pure and free of protein interferences.
A more specific method for DNA quantification is the picogreen assay.
Picogreen is a fluorescent intercalator dye that is specific for dsDNA and will fluo-
resce when bound to DNA. The system is capable of detecting dsDNA at concentra-
tions of 25 pg/ml. The picogreen assay is more labor-intensive than the UV
spectrometric method; however, the enhanced sensitivity is usually not required for
the majority of DNA preparations from whole blood.
2.3Microarray-Based Genotyping Platforms
Genomic microarrays utilize solid surfaces, such as a small chip, where a large number
of probes are arrayed, allowing for many SNPs to be interrogated simultaneously.
Table 2.2 shows microarray-based platforms currently available on the market.
Virtually, all of the assays offered on these platforms require multiplex PCR amplifi-
cation prior to genotyping. One exception is Verigene, which works on a basis of
signal amplification and does not require a PCR step. At present, only a handful of
genotyping tests are available for invitro diagnostic (IVD) use. Most of the IVD tests
on the market today are for warfarin sensitivity testing, a result of the recent FDA
decision to approve updated labeling for Coumadin, the brand name of warfarin
(http://packageinserts.bms.com/pi/pi_coumadin.pdf., accessed 04/07/10), requiring
the inclusion of genotyping considerations regarding patients response to the drug.
18 N. Babic et al.
Table 2.2 Genotyping platforms and respective pharmacogenomics tests

Platform Manufacturer Test Target drug
Amplichip Roche (Plesanton, CA) CYP2D6 (IVD) Various drugs
CYP2C19 (IVD) Various drugs
eSensor GenMark (Carlsbad, CYP2C9/VKORC1 (IVD) Warfarin
CA ) CYP 2C19 (RUO) Clopidogrel
INFINITI AutoGenomics CYP2C9/VKORC1 (IVD) Warfarin
(Carlsbad, CA ) CYP2D6, 2C19, 3A4, 3A5 Various drugs
UGT1A1 Irinotecan
MDR-1 Various
5-FU 5-Fluorouracil
NAT-2 Various drugs
Tag-It Luminex (Austin, TX) Open system Various
Verigene Nanosphere CYP2C9/VKORC1 (IVD) Warfarin
(Northbrook, IL)
Real-time PCR ParagonDx (Morisville, CYP2C9/VKORC1 (IVD) Warfarin
NC)
eQ-PCR Light TrimGen (Sparks, MD) CYP2C9/VKORC1 (IVD) Warfarin
Cycler
Invader Hologic (Madison, WI) UGT1A1 (IVD) Irinotecan
IVD invitro diagnostics; RUO research use only
The remainder of this section describes different methodologies utilized by the

current microarray genotyping platforms.
2.3.1Affymetrix Technology
The Affymetrix technology is utilized in the manufacture of DNA chips for geno-
typing in the Roche CYP450 system (AmpliChip 450 Test, shown in Fig. 2.1) that
provides detection of CYP2D6 and CYP2C19 gene variants. This test is FDA-
cleared for IVD use and is currently offered by several major reference laboratories.
Affimetrix chips are oligonucleotide arrays that contain hundreds of thousands of
oligonucleotide probes anchored on glass support at very high density. The basis of
this technology is hybridization of the target DNA to the microarray and generation
of fluorescent signal. A workstation provided by Affymetrix is necessary to collect
data and perform the analysis.
2.3.2eSensor Technology
GenMark eSensor utilizes electrochemical detection technology (Fig. 2.2). The sin-
gle-stranded capture probes are immobilized on the working electrode surface.
Following PCR amplification, the target DNA is digested with exonuclease into sin-
Fig. 2.1 Roche AmpliChip.

The amplified target DNA is
fluorescently labeled, frag-
mented, and added to the chip
where a laser scans and ana-
lyzes the intensity of fluores-
cence. (Courtesy of Roche
Diagnostics)
gle strands and combined with single-stranded signal and capture probes. Signal
probes contain the polymorphism of interest and an electrochemically active ferrocene
dye unique for that SNP. Once all the reagents are combined, the target amplicon
hybridizes simultaneously to signal and capture probes, generating current. If there is
no match between the signal probe and the DNA target, no current will be generated.
Of the systems currently available, this is the only true random-access analyzer,
where each sample can be loaded and unloaded independently. Upon completion of
PCR and exonuclease digestion steps (~3.5 h), the analysis time on the instrument is
30 min [4]. Individual analyzers can run 18 samples and can be configured in such
a way that a maximum of 24 samples could be analyzed at once. At present, the only
commercially available IVD test is warfarin sensitivity genotyping test.
2.3.3INFINITI and Luminex xTAG Allele-Specific

Primer Extension
The allele-specific primer extension (ASPE) method is utilized by the INFINITI

(AutoGenomics Inc., Carlsbad, CA) and Luminex xTAG (Luminex Corporation,
Austin, TX) platforms. The ASPE method involves two phases following PCR: enzy-
matically driven sequence specific primer extension and a capture on a solid surface,
such as chip (INFINITI) or bead (Luminex), for detection. Figure 2.3 illustrates this
methodology, using the INFINITI assay as an example. In general, an amplified PCR
20 N. Babic et al.
Fig. 2.2 GenMark eSensorTM Warfarin Sensitivity assay genotyping principle. (a) Three-step pro-
cess involved in genotyping on eSensor platform: DNA is extracted from the whole blood, ampli-
fied by PCR, digested into single strands, and loaded onto a cartridge provided by manufacturer.
Each cartridge contains specific capture probes on particular electrodes. Cartridges are then read
by fully automated analyzer. (b) A close-up view of one electrode within a cartridge. The sample
is heterozygous in this example, so both the wild type and mutant target molecules will hybridize
to the capture probe and to the appropriate signal probe. This means that two different types of
ferrocene are present on the electrodes surface. Thus, when the instrument checks for hybridiza-
tion complexes, it will read current coming from both types and interpret the sample as heterozy-
gous for this mutation. (Courtesy of GenMark)
fragment serves as a template for ASPE reaction. ASPE occurs on board the instru-
ment where PCR amplified genomic DNA is combined with the allele-specific prim-
ers. These primers are then extended by the activity of polymerases. Extension occurs
only if the 3 end of the allele-specific primer is bound to the homologous sequence
on the amplicon. A short segment at the 5 end of the allele-specific primer remains
free. This free 5 segment is used to hybridize ASPE product to a complementary
oligonucleotide sequence (called Zipcode in the INIFINITY methodology), that is
immobilized on a chip or a bead. The signal is generated and detected using the
fluorescent labels. In the INFINITI assay, the fluorescent label is directly incorporated
into a growing primer strand as fluorescently labeled deoxycytidine triphosphate
(dCTP). In the case of xMap Luminex assay, biotinylated dCTP is incorporated into
5 SNP 3
Singlestrand DNA Target
Antizipcode A Antizipcode B
Primer A Primer B
Hybridization and Primer Extension

nd DNA Target
(Cy5) dCTP
PrimerTarget Denaturation and Detection

BioFilmChip
Zipcode A
Zipcode B
Fig. 2.3 Automated micro array method: INFINITITM analyzer method combines allele specific
primer extension (ASPE) and fluorescence detection. The primer extension occurs only when a
matching primer hybridizes to the target sequence. If there is a mismatch (e.g., Primer B), no
extension occurs. During the ASPE step, fluorescently labeled deoxycytidine triphosphate, (Cy5)
dCTP, is incorporated into the growing strand. Unique, complementary nonhuman oligonucleotide
sequences (zipcode and antizipcode) are used to capture fluorescent signal. Figure adapted from
INFINITI Analyzer Customer Training manual and used with permission.
a growing primer strand. The fluorescent dye is conjugated to streptavidin, which

then binds to biotinylated dCTP and the hybridized beads are read by the Luminex
system. Luminex xMAP is a fluidic system based on the principles of flow cytometry
where the stream of suspended microspheres passes through the detection chamber
single file. Once in detection chamber, each microparticle is excited by two different
lasers and the fluorescence is detected.
Presently, the INFINITI platform offers the most comprehensive test menu of all
the automated microarray platforms available (Table 2.2); however, the only IVD
test panel currently available on this platform is warfarin sensitivity genotyping
panel that includes CYP2C9 *2 (c.681G>A), *3 (c.636G>A) and VKORC1
(c.1639G>A). The other tests are for research use only (RUO) at this time.
It is worth noting that for IVD genotyping panels, all the primer sets and multi-
plex PCR parameters are provided by the manufacturer. In case of RUO, investiga-
tional use only (IUO) or analyte specific reagents (ASR) tests, the individual
laboratory has to optimize the PCR conditions and occasionally design their own
primer sets and independently validate these assays before they can be offered for
clinical service. Luminex currently offers only IUO genotyping tests, including
CYP450 2C19, 2D6 and warfarin sensitivity (2C9/VKORC1) tests. Although not
all the reagents and parameters are optimized by the manufacturer, in this case the
manufacturer does provide a sample protocol for ASPE and hybridization to xTAG
22 N. Babic et al.
beads (http://www.luminexcorp.com/support/protocols/xtag_protocols.html,
accessed 10/20/10). Due to the unique design of beads, one can perform up to 100
reactions within a single well. Therefore, up to 100 different SNPs could conceiv-
ably be interrogated for a single patient.
2.3.4Verigene Technology
As mentioned previously, the Verigene platform (shown in Fig. 2.4) does not
require PCR amplification to detect polymorphisms in target DNA. Here, capture
oligonucleotide probes attached to a solid phase support hybridize to the target
DNA. Following capture, target sequences are further hybridized to complemen-
tary oligonucleotides bound to nanometer-sized gold particles. The final step is the
catalytic deposition of silver onto the gold nanoparticle resulting in the enhance-
ment of the signal by six orders of magnitude, thus making this a highly sensitive
method.
Verigene is another example of a fully automated, user friendly instrument. An
additional benefit of this platform is the capability of the system to do DNA extrac-
tion on-board. The user process for Verigene System is as follows: the user places
25 mL of the DNA sample and an equal volume of the test-specific sample buffer
into the test cartridge and then inserts the cartridge into the Verigene Processor unit
(Fig. 2.4), or, the user inserts the test cartridge, an extraction tray, and pipette tips
into the Verigene Processor SP unit and adds 1 ml of the whole blood sample into
the extraction tray. The processor will then execute either DNA hybridization or
extraction and hybridization, followed by the genotyping test. Currently, an IVD
test for warfarin sensitivity and an RUO test for ten different potentially clinically
relevant 2C19 SNPs are available on Verigene.
Microarrays are optimally suited for detection of high number of allelic variants,
for example, genotyping the numerous CYP 2D variants. This technology allows
for large-scale simultaneous screening and detection of thousands of SNPs that
might be clinically relevant.
However, most of the microarray platforms currently available are not conducive
to high-throughput testing in terms of number of patient samples that could be
genotyped. For example, INIFINITI takes approximately 8 h to result a batch of 24
samples, while eSensor can provide results for 824 patients within 4 h, depending
on the number of analytical towers used. Thus, the platforms that are designed such
that the operator can potentially walk-away once all the samples and reagents are
loaded, such as INIFINITI and eSensor, are a good fit for an average clinical labora-
tory that does not have staff with extensive molecular genetics background. The
platforms capable of analyzing large number of samples that offer mostly RUO or
ASR tests are geared more toward the labs that have some level of expertise in
molecular genetic testing, since the appropriate primers and assay parameters have
to be established and optimized by the user. The manufacturers do, in general,
provide basic guidelines to aid the user in establishing these parameters.
Fig. 2.4 The VerigeneTM genotyping platform. (a) The Verigene Reader is barcode scanning and
network communication-enabled central control unit and main user interface. (b) The Verigene
Processor is comprised of four independent test modules controlled by the Verigene Reader. It
facilitates the wet chemistry inside the test cartridge (c). Test processing in this module relies on
offline nucleic acid extraction. (d) The Verigene Processor SP possesses all of the same test pro-
cessing functionality and components as the regular processor but additionally provides capabili-
ties to extract DNA directly from the specimen. These additional functions require the following
consumables: a Verigene Extraction Tray and Verigene Tip Holder Assembly (not shown).
(Courtesy of Nanosphere)
2.3.5Real-Time PCR
In the application of real-time PCR for SNP genotyping, fluorescently labeled

probes are designed to detect SNPs of interest. As the PCR progresses, the genomic
DNA is amplified resulting in higher fluorescence with each consecutive cycle.
Real-time PCR is performed on thermal cyclers that can detect fluorescence at
different wavelengths. The most commonly used types of signal detection probes
in use today for real-time PCR include hydrolysis (TaqMan) probes and hybridiza-
tion probes. Both types of probes utilize fluorescence resonance energy transfer
(FRET) principle, where the fluorescence is generated by the energy transfer from
one probe to another. The difference is in probe design and the principle used to
achieve specificity. The mechanism of each probe is described below.
2.3.5.1Hydrolysis Probes
Hydrolysis probes are the most widely used real-time PCR probes. In hydrolysis
probes, specificity lies in probe design, and the nucleotide sequence of each probe
is complimentary to the target DNA region that contains a SNP of interest. Typical
assay design is shown using ParagonDx real-time PCR with the TaqMan probe
as an example (Fig. 2.5a). The probe consists of a reporter dye and a quencher. If
the primer and the target amplicon match, they hybridize and Taq DNA polymerase
24 N. Babic et al.
initiates primer extension. At the same time, the appropriate probe also hybridizes
to the target. During the extension step, the exonuclease activity of polymerase
cleaves the reporter dye from the probe. Once the dye and quencher are separated,
the dye is free to emit its characteristic fluorescence. If there is no match between
the probe and the target, the probe does not hybridize and the reporter dye
fluorescence remains quenched. The change in fluorescence emitted is plotted with
each PCR cycle and the amplification plot is generated (Fig. 2.5b). The genotype is
determined based on the end-point fluorescence, where the fluorescence emitted by
the reporter dye bound to variant allele is compared to the fluorescence emitted
bythe reporter dye bound to the wild-type allele. The genotype is determined based
on the relative ratio of the two. Currently, a maximum of two sets of dyes can be
used per reaction tube, which limits the extent of multiplexing in assays using
classic hybridization probes.
An assay marketed by Fluidigm Corporation (San Francisco, CA) consists of
dynamic arrays for SNP genotyping (Fig. 2.6) offering a potential for multiplexing
through automation. This technology utilizes microfluidics-based devices called
integrated fluidic circuits (IFCs) where networks of interconnected microchannels
are fabricated onto miniature devices. Up to 96 patients and 96 different SNPs can
be applied to a single device (96.96 dynamic array). Thus, 96 SNPs can be geno-
typed for each patient. This technology enables a laboratory to use the existing
a b
Reporter Dye Quencher Amplification Plots
Primer Probe
8433
7433
Primer annealing 6433
Fluorescence (dR)
Probe hybridization 5433

Polymerase 4433
3433
2433
Extension 1433
433
567
1567
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35
Cycles
Fig. 2.5 Real-time PCR using Taqman probe method. (a) A schematic of PCR using Taqman
probe methodology. Taqman probe consists of a reporter dye and a quencher. When the probe is
intact, the reporter dye emission is quenched. The matching probe hybridizes to the target and Taq
DNA polymerase initiates primer extension. During the primer extension step, the exonuclease
activity of polymerase cleaves the reporter dye from the probe. The dye is then free to emit its
characteristic fluorescence. If there is no match between the probe and the target, the probe does
not hybridize and the reporter dye fluorescence remains quenched. (b) Representative amplification
plots that show increasing reporter dye fluorescence with each PCR cycle. The data was generated
on a Stratagene Mx3005P real-time PCR platform (Stratagene, La Jolla, CA). Figure reproduced
from Personalized Medicine, Babic etal. [5] with permission of Future Medicine Ltd
Fig. 2.6 A 48.48 dynamic array chip is shown on the left. The center of the chip is the integrated
fluidic circuit (IFC), a network of fluid channels, valves and reaction chambers. The blow-up
portion of the IFC shows one of the 2,304 individual reaction chambers and its associated contain-
ment and interface valves. Following the sample loading, the digital array is thermocycled,
imaged, and analyzed on a BioMark system (right) (Courtesy of Fluidigm Corporation)
TaqMan SNP genotyping assays in a microarray configuration requiring only 192

pipetting steps within 3 h-time frame. This is in sharp contrast to a conventional
384-well microplate real-time PCR assay that would require 18,432 pipetting steps
and 8 days to complete (http://fluidigm.com/applications/genotype-profiling.html,
accessed 04/04/10).
2.3.5.2Hybridization Probes
Hybridization probes are a pair of fluorescent probes of different lengths placed in

close proximity (see Fig. 2.7 for illustration). In these probes, specificity is achieved
by designing the shorter probe to be specific for variant allele, while the longer
probe is common. In such a case, melting curves recorded by real-time PCR will be
different. Primary condition for the FRET to occur is that there is a significant
overlap between the emission spectrum of one fluorophore (termed donor) with the
excitation spectrum of the other (termed acceptor). Thus in real-time PCR design,
one oligonucleotide probe (the sensor probe) is specific for a target and contains a
5 donor molecule, while the other (an anchor probe) is common and contains a 3
acceptor molecule. Only when these are placed within 15 bp of each other, energy
can be transferred from the donor to the acceptor, resulting in the emission of
fluorescence.
The main advantage of real-time PCR is the speed with which samples can be
analyzed, since the signal is read in real-time and there are no post-PCR processing
steps. PCR can be performed in 96- or 384-well format, thus allowing a high-through-
put design. Furthermore, since this is a closed-tube method of analysis, the risk of
26 N. Babic et al.
Fig. 2.7 Mechanism of action of hybridization probes. Shown are the two fluorescent dyes with
overlapping spectra where the emission wavelength of one dye overlaps with the excitation wave-
length of the other. When the dyes are apart green dye will fluoresce. When the dye are placed
within 15 bp of each other, such as upon binding to target DNA, the emission wavelength of the
green dye will be absorbed by the red dye, causing only red fluorescence to be detected
sample contamination, errors from mistakes in tube transfers, or amplicons escaping

into the laboratory environment is minimal. Until recently, one of the major disadvan-
tages of real-time PCR was the limitation in multiplexing capabilities. Since it is often
necessary to test several alleles for the same patient, it would be desirable to stream-
line the processes such that all of the necessary genotype information could be
obtained from minimal user intervention. Several manufacturers have developed
sophisticated and streamlined real-time PCR instruments. One such instrument,
already described at the beginning of this section, is marketed by Fluidigm (see Sect.
2.3.5.1). Another example is a BD MAX System (HandyLab Inc., Ann Arbor, MI),
shown in Fig. 2.8. This platform combines fully automated DNA extraction with
microfluidics-based real-time PCR performed in disposable microfluidics cartridges.
BD MAX technology is another example of a true random-access analyzer, besides
GenMark eSensor, where each cartridge can be controlled separately. Although the
entire genotyping process on BD MAX instrument is very simple and elegant, unlike
Fluidigm, this platform offers only limited multiplexing capabilities.
2.4Invader Assay
The typical workflow of the Invader assay is shown in Fig. 2.9a. The Invader
chemistry is composed of two simultaneous isothermal reactions: a first reaction that
detects polymorphism of interest and a second reaction used to generate and amplify
Fig. 2.8 BD MAXTM system. a. Unitized Reagent Strip (URS). This strip contains all the reagents
and consumables required for lysis, nucleic acid extraction and PCR set-up. b. Disposable micro-
fluidic cartridges, showing a total of twelve reaction chambers. Each chamber is individually con-
trolled to enable the user to perform different PCR protocols simultaneously within a cartridge.
Each chamber can be sealed off with a set of micro-valves, prior to PCR initiation, thus preventing
evaporation and minimizing the risk of contamination. (Courtesy and Becton, Dickinson and
Company)
the signal. In the first reaction (Fig. 2.9b), two oligonucleotides, a primary probe,
and an invading oligonucleotide are used. Both oligonucleotides contain a nucle-
otide complimentary to the polymorphism of interest. Therefore, both the probe and
an invading oligonucleotide will bind to the variant allele, overlapping at the SNP
position. The 5 flap of the probe, including the overlapping nucleotide (i.e., SNP),
is then enzymatically cleaved. Released flaps from the primary reaction serve as
invading oligonucleotides for a hairpin FRET probe in a second reaction. Once the
5 flap is bound to the FRET probe, the probe is cleaved and fluorescent signal is
generated (Fig. 2.9c). Each released 5 flap can cycle between cleaved and non-
cleaved FRET probes, thereby amplifying the signal.
The advantages of Invader technology include streamlined workflow and high
throughput. Since this technology does not require PCR, it only takes 30 min to geno-
type a 96-well plate (Fig. 2.9a). Since the only instrumentation required is a UV plate
reader, the technology is not cost prohibitive. Finally, while the methodology is ame-
nable to be automation, it could not be easily multiplexed because a maximum of two
different primary probes and their complementary FRET probes can be used in a
single well. Therefore, only a single SNP can be detected in each well. Since all the
reagents have to be manually added to the plate, this technology is not very practical
for analysis of genes that have a large number of variants, such as CYP2D6 alleles.
2.5Sequencing
Sequencing is a definitive method of determining a DNA sequence and identify-

ing variant(s). The DNA sequence is determined by amplification of target
sequence, followed by fragmenting the genome, typically from a PCR product,
28 N. Babic et al.
Fig. 2.9 Invader assay. a. Overall workflow of the Invader assay. b. The scheme of the Invader
primary reaction where two oligonucleotides, a primary probe and an invading oligonucleotide,
bind to the target DNA simultaneously overlapping at the SNP position and subsequent cleavage
of the primary probe 5 flap. c. Secondary reaction involving binding of 5 flap to the FRET probe,
subsequent cleavage and signal generation. (Courtesy of Hologic)
into short fragments. In chain terminator sequencing (Sanger sequencing),

primer extension is initiated by DNA polymerase. A low concentration of a
chain-terminating nucleotide (most commonly a di-deoxynucleotide) is used
along with the four deoxynucleotide bases. Each of the dideoxynucleotide chain
terminators is labeled with a separate fluorescent dye. Limited incorporation of
the chain-terminating nucleotide by the DNA polymerase results in a series of
related DNA fragments that are terminated only at positions where that particu-
lar nucleotide is used. The fragments are then size-separated by capillary elec-
trophoresis and different nucleotides within a strand are visualized as different
color peaks (Fig. 2.10).
2.6Pyrosequencing
Pyrosequencing, also known as a sequencing-by-synthesis, is a technique based on

real-time detection of DNA synthesis (Fig. 2.11). The pyrosequencing platform
PSQ 96MA (Qiagen Inc., Valencia, CA) is capable of reading 50100 bases and
genotyping virtually any SNP with high accuracy. In addition, this technology can
also be used quantitatively to determine the amount of each allele in the DNA
Fig. 2.10 DNA sequencing detected by fluorescence capillary electrophoresis. This figure shows
986C>T polymorphism. Modified from http://www.clcbio.com/index.php?id=785, accessed on
4/22/10
Fig. 2.11 Pyrosequencing. Target DNA ius first PCR-amplified and then reacted with deoxyribo-
nucleotide triphosphates (dNTPs), liberating pyrophosphates (PPi) in equimolar quantities (A). In
(B), each liberated PPi is reacted with a series of enzymes generating light. Each fluorescent signal
is seen as a peak in a raw data output (Pyrogram), shown in (C). QIAGEN, all rights reserved.
sample and thus estimate allelic frequency of a SNP in a population [5]. Following
the PCR reaction, DNA amplicons are separated into single strands and hybridized
with a sequencing primer. During the sequencing reaction, dNTPs are added one
at a time, up to 50 bases until the desired region containing the polymorphism has
30 N. Babic et al.
been sequenced. The pyrosequencing reaction also contains a mixture of enzymes,

DNA polymerase, ATP sulfurylase, luciferase, and apyrase. In a coupled enzymatic
reaction with ATP sulfurylase and luciferase, the pyrophosphate released from an
incorporated dNTP generates a light signal that is proportional to the amount of
dNTP incorporated. Apyrase is required for the removal of unincorporated dNTP
and generation of ATP prior to the addition of next nucleotide. The light signal is
detected with a CCD camera and is recognized as a Pyrogram. Pyrosequencing
has been successfully applied to CYP2C9 and CYP3A5 genotyping [6, 7]. Seatki
et al. [8] used pyrosequencing to provide comprehensive genotyping of six
UGT1A1 polymorphisms in the Japanese population, while Odeberg etal. studied
the prevalence of different UGT1A1 polymorphisms in a Swedish cohort, encom-
passing 14 different ethnic groups [9]. Pyrosequencing was also used to study the
role of UGT1A7 and UGT1A9 polymorphisms in prediction of response and tox-
icity in colorectal patients treated with irinotecan [10] and in the study of the
importance of ABCB1 gene polymorphism in ovarian cancer resistance to pacli-
taxel [11].
2.7Mass Spectrometry
Mass spectrometry is a widely used technology for the identification of a variety

of compounds, ranging from small molecules, such as drugs, to biomolecules,
such as proteins, peptides, and oligonucleotides. In this approach, the analyte is
detected as a peak with a specific mass-to-charge ratio. Mass spectrometry has
been successfully applied to genotyping by several groups [1218]. The most
frequently used configurations are matrix-assisted laser desorption ionization
time of flight (MALDI-TOF) due to the fact that MALDI allows efficient ioniza-
tion and fast analysis times, while TOF mass analyzers are not mass limited and
do not require compound fragmentation prior to detection.
One of the implementations of the MALDI TOF MS is the MassARRAY
IPLEX Gold SNP Genotyping system (SEQUENOM, Sand Diego, CA),
shown in Fig. 2.12. Gabriel et al. described a protocol for SNP genotyping
using this platform [19]. The general workflow of MassARRAY assay is
fairly simple, requiring initial PCR target amplification, followed by a spe-
cifically designed ASPE process. The resultant extension product will have
allele-specific difference in masses, which is a base of detection and SNP
identification by mass spectrometry. MassARRAY system has a capability of
processing up to 384 samples in parallel and allows for multiplexing of up to
40 different SNPs in a single well. It takes approximately 4560 min of
instrument time to process and result for a 384 p osition chip. Someexamples
of MassARRAY technology applications include CYP2D6 g enotyping and
study of genetic variants in prostate cancer [14, 15]. Recently Yang etal. have
reported the use of similar mass spectrometry technology called
surface-enhanced laser desorption and ionization mass spectrometer (SELDI-
TOF) for warfarin g enotyping [12].
The advantages of using mass spectrometry for genotyping include high sensi-
tivity and specificity, high throughput, and cost savings required for genotyping.
However, the equipment itself is fairly expensive and highly specialized requiring
highly trained personnel.
Fig. 2.12 MassARRAY iPLEX gold reaction. Following the PCR amplification of target DNA,
all the unincorporated dNTPs are treatedTM with shrimp alkaline phosphatase (SAP) to render
them unavailable to future reaction. The iPLEX Gold cocktail is then added and the mixture is
thermocycled to extend the primer by only one nucleotide (Note: The primers should be designed
such that the 3 end of each primer is immediately adjacent to the polymorphic site). The extended
primers are then analyzed by MALDI-TOF and potential variants differentiated by differences in
mass of terminating nucleotides. (Courtesy of Sequenom, Inc.)
32 N. Babic et al.
2.8Practical Considerations
Overall, a number of very good genotyping methodologies are available on the

market today. The variety in platform designs and test menus allows the application
of genotyping testing to a wide range of clinical laboratories, from small to large
hospitals and specialized genetic centers. Despite a large number of platforms on
the market today, there are only a few characterized DNA reference materials avail-
able for PGx testing. Therefore, running the appropriate QCs often presents addi-
tional challenge for the PGx laboratory. In the absence of reference materials,
laboratories often resort to the use of unconfirmed and nonrenewable sources of
DNA material, such as residual patient samples. Recently our laboratory, in
collaboration with CDC, characterized a panel of 107 genomic DNA reference
materials available from the Coriell Cell Repostiories for five loci (CYP2D6,
CY2C19, CYP2C9, VKORC1, and UGT1A1) commonly included in PGx testing
panels (http://wwwn.cdc.gov/dls/genetics/rmmaterials/MaterialsAvailability.aspx,
accessed 4/20/2010). This will enable laboratories performing PGx testing to pur-
chase already characterized quality control materials. Another consideration that
should not be overlooked by any laboratory performing the PGx testing is the vari-
ability in allele variant definition between different assay platforms. This can lead
to the discrepant results, especially when multiple SNPs are used to identify variant
alleles, for example, *4 vs. *10 variant in CYP2D6. CYP2D6*4 has no enzymatic
function, while CYP2D6*10 has decreased enzymatic function. The major SNP in
*4 is c.1846G>A with the minor SNPs c.100C>T, c.974C>A, c.984A>G, and
c.4180G>C. In *10 haplotype, the major defining SNP is c.100C>T and the minor
SNP is c.4180G>C. If the assay does not define multiple SNPs of variant alleles,
the potential exists, therefore, that patient could be identified as *4/*10 where the
true genotype may be *1/*4. This is important because *4/*10 is assigned an inter-
mediate metabolizer phenotype, while *1/*4 is assigned an extensive metabolizer
phenotype. This discrepancy could potentially affect the clinical management of
drug therapy.
Other practical aspects each laboratory should consider when selecting a geno-
typing platform include testing volume, specific tests, and the available personnel.
If a clinical chemistry laboratory, staffed with traditionally trained medical
technologists, is looking to implement warfarin genotyping test, for example, the
optimal choice for such laboratory would probably be one of the fully automated
microarry platforms (e.g., eSensor or INFINITI) that offers IVD products. Since
these platforms allow multiplex PCR, only one PCR tube needs to be prepared for
each patient. Furthermore, the manufacturers provide detailed instructions, opti-
mized PCR parameters, and all the reagents necessary. The operator can follow the
directions, load the samples onto the analyzer, and walk away. Virtually, all the
automated platforms contain the software that interprets the readings and displays
final genotype result for each sample. On the other hand, the laboratory that is
interested in establishing their own genetic testing for various applications would
benefit from the open platforms, such as Luminex or Beckman microarrays that
give the user freedom to design customized test menu. If one is interested in testing
limited number of alleles, real-time PCR should be considered. This technology
offers theshortest analysis time, and most of the real-time PCR instruments func-
tion asopen systems allowing user-specific testing. However, one should keep in
mind that open system microarrays and real-time PCR demand some previous
training and experience in design of molecular diagnostic tests, which is in contrast
to the fully automated platforms that could easily be operated by any medical tech-
nologists regardless of their background.
2.9Conclusion
While it is clear that technology in pharmacogenomics field has come a long way,
the data on clinical utility of pharmacogenomics testing is still lagging. According
to the Secretarys Advisory Committee on Genetics, Health and Society (SACGHS
[NIH]) (http://oba.od.nih.gov/oba/SACGHS/reports/SACGHS_PGx_report.pdf,
accessed 04/28/10), the successful implementation of pharmacogenomic testing
into a clinical laboratory requires tests that would predict a specific clinical out-
come and lead to informed decision making by clinicians. There are numerous
barriers currently preventing wider adoption of pharmacogenomics in clinical prac-
tice (dealt extensively by Yeo etal., Chap. 1), but solutions are beginning to appear.
One prediction is certain: before these pharmacogenomics tests can become the
standard of care in personalizing drug therapies, clinical trials showing utility and
cost-effectiveness of such testing are necessary.
References
1. Trent, R. J. (2010). Pathology practice and pharmacogenomics. Pharmacogenomics,

11,105111.
2. Flockhart, D. A., et al. (2009). Clinically available pharmacogenomics tests. Clinical
Pharmacology and Therapeutics, 86, 109113.
3. Wang, L., Luhum, R., & Lei, M. (2007). SNP and mutation analysis, in microarray technology
and cancer gene profiling. In: S. Mocellin (Ed.). Advances in experimental medicine and biology
(pp. 105116). Landes Bioscience and Springer Science+Business Media, LLC.
4. Babic, N., Haverfield, E. V., Burrus, J. A., Lozada, A., Das, S., & Yeo, K. T. (2009). Comparison
of performance of three commercial platforms for warfarin sensitivity genotyping. Clinica
Chimica Acta, 406, 143147.
5. Lavebratt, C., & Sengul, S. (2006). Single nucleotide polymorphism (SNP) allele frequency
estimation in DNA pools using Pyrosequencing. Nature Protocols, 1, 25732582.
6. Aquilante, C. L., Lobmeyer, M. T., Langaee, T. Y., & Johnson, J. A. (2004). Comparison of
cytochrome P450 2C9 genotyping methods and implications for the clinical laboratory.
Pharmacotherapy, 24, 720726.
7. Aquilante, C. L., Langaee, T. Y., Anderson, P. L., Zineh, I., & Fletcher, C. V. (2006). Multiplex
PCR-pyrosequencing assay for genotyping CYP3A5 polymorphisms. Clinica Chimica Acta,
372, 195198.
34 N. Babic et al.
8. Saeki, M., Saito, Y., Jinno, H., Tohkin, M., Kurose, K., & Kaniwa, N. (2003). Comprehensive
UGT1A1 genotyping in a Japanese population by pyrosequencing. Clinical Chemistry, 49,
11821185.
9. Mercke Odeberg, J., Andrade, J., Holmberg, K., Hoglund, P., Malmqvist, U., & Odeberg, J.
(2006). UGT1A polymorphisms in a Swedish cohort and a human diversity panel, and the
relation to bilirubin plasma levels in males and females. European Journal of Clinical
Pharmacology, 62, 829837.
10. Carlini, L., Meropol, N., Bever, J., Andria, M. L., Hill, T., & Gold, P. (2005). UGT1A7 and
UGT1A9 polymorphisms predict response and toxicity in colorectal cancer patients treated
with capecitabine/irinotecan. Clinical Cancer Research, 11, 12261236.
11. Green, H., Sokerkvist, P., Rosenberg, P., Horvath, G., & Peterson, C. (2008). ABCB1 G1199A
polymorphism and ovarian cancer response to paclitaxel. Journal of Pharmaceutical Sciences,
97, 20452048.
12. Yang, S., Xu, L. H., & Wu, H. M. (2010). Rapid genotyping of single nucleotide polymor-
phisms influencing warfarin drug response by surface-enhanced laser desorption and ioniza-
tion time-of-flight mass spectrometry. Journal of Molecular Diagnostics, 12, 17.
13. Tost, J., & Gut, I. G. (2005). Genotyping single nucleotide polymorphisms by MALDI mass
spectromtery in clinical applications. Clinical Biochemistry, 38, 335350.
14. Zheng, S. L., Sun, J., Wiklund, F., Smith, S., Stattin, P., & Li, G. (2008). Cumulative associa-
tion of five genetic variants with prostate cancer. The New England Journal of Medicine, 358,
910919.
15. Kohlrausch, F. B., Gama, C. S., Lobato, M. I., Belmonte-de-Abreu, P., Gesteira, A., &
Barros, F. (2009). Molecular diversity at the CYP2D6 locus in healthy and schizophrenic
southern Brazilians. Pharmacogenomics, 10, 14571466.
16. Misra, A., Hong, J. Y., & Kim, S. (2007). Multiplex genotyping of cytochrome P450 single-
nucleotide polymorphisms by use of MALDI-TOF mass spectrometry. Clinical Chemistry,
52, 933939.
17. Tindall, E. A., Speight, G., Petersen, D. C., Padilla, E. J. D., & Hayes, V. M. (2007). Novel
Plexor SNP genotyping technology: comparisons with TaqMan and homogenous
MassEXTEND MALDI-TOF mass spectrometry. Human Mutation, 28, 922927.
18. Blievernicht[AU4], J. K., Klein, K., Eichelbaum, M., Schwab, M., & Zanger, U. M. (2007).
MALDI-TOF mass spectrometry for multiplex genotyping of CYP2B6 single-nucleotide
polymorphisms. Clinical Chemistry, 53, 2433.
19. Gabriel, S., Ziaugra, L., & Tabbaa, D. (2009). SNP genotyping using the Sequenom
MassARRAY iPLEX platform. Current Protocols in Human Genetics, 60, 2.12.12.12.18.
Chapter 3
Economics of Pharmacogenomic Testing
in Clinical Practice
Alan H.B. Wu
Keywords Quality-adjusted life years Incremental cost-effectiveness ratio

Markov model
3.1Introduction
The realization of pharmacogenomic testing or any new clinical laboratory program

from the research bench into routine practice requires economic justification. It is
insufficient for clinical laboratories to simply report a genotype for a particular
individual. Test results must be interpreted by qualified individuals taking consid-
eration of the clinical context of the tested patient. Ultimately, these interpretations
must have the potential to lead to clinically actionable decisions. Ideally, these deci-
sions lead to better clinical outcomes thereby justifying the cost for testing. If testing
never leads to changes in patient management, pharmacogenomics will be for
research purposes only. In general, there are three types of medical questions that
can be answered by pharmacogenomic testing:
1. Is this the right drug for my patient?
There are many types of drugs that can be used to treat a patient with a particular
disease. The efficacy of these drugs may depend on a number of environmental,
demographic, and genetic factors. For many drugs, selection of the best drug for
a patient is a matter of trial and error.
A patient with a particular genotype might be linked to others who have previ-
ously shown to not benefit from the selected drug. While this does not guarantee
nonefficacy for the patient in question, the medical decision might be to consider
a different medication. For example, patients with breast cancer may be treated
with an aromatase inhibitor instead of tamoxifen if they have a null genotype for
A.H.B. Wu(*)
Department of Laboratory Medicine, University of California, San Francisco, CA 94110, USA
e-mail: wualan@labmed2.ucsf.edu
36 A.H.B. Wu
CYP2D6. If testing identifies the patient to have a genotype where the majority
of others have shown clinical efficacy, the patient and physician can get some
satisfaction and assurance that this drug will be successful. While the economic
benefit cannot be quantitated, these patients may be more compliant with their
medications if they know that therapeutic selection was based on the individuals
genetic makeup.
2. Will this drug produce a rare side effect?
The U.S. Food and Drug Administration (FDA) and equivalent regulatory bodies
in other countries are responsible for ensuring the safety of therapeutic drugs.
A drug will not receive regulatory clearance or will be removed by the FDA
from routine practice if a significant rate of adverse events is observed.
Nevertheless, unexpected catastrophic side effect can occur with use of medica-
tions at therapeutic doses. Pharmacogenomic testing has the promise of reducing
the incidence of rare adverse events for selected medications. The best example
is the link between particular human lymphocyte antigen (HLA) polymorphism
and prediction of Stevens Johnson syndrome (SJS) and Toxic Epidermal
Necrolysis (TEN) for drugs such as abacavir, carbamazepine, phenytoin, and
allopurinol. Where there are many other drugs that can produce SJS/TEN, their
genetic associations have not been uncovered. As with testing for drug efficacy,
the medical decision for patients determined at high risk for an adverse event
would be the use of an alternative drug. The economic impact of pharmacog-
enomic testing would be avoidance of the medical costs needed to treat patients
suffering from these side effects.
3. What is the best dosage for the drug that has been selected?
Most medications have a narrow therapeutic window for efficacy and toxicity
avoidance. Pharmacogenomic testing can be useful in predicting the dosage that
results in a blood concentration that falls within that window. Predictive dosage
algorithms have been established using a combination of specific patient demo-
graphic information, the history of other medications and diseases, and relevant
genotypes. The best example is for warfarin where a multiethnic dosing algo-
rithm has been created [1]. There are several areas where a dosing algorithm
based in part on pharmacogenomics might have an economic impact. The big-
gest impact is avoidance of disease progression if the drug is underdosed, or
adverse events if the drug is overdosed. Other benefits include the reducing or
eliminating the costs associated with adjusting dosages (e.g., new prescription
costs or reduced need for clinic visits and monitoring tests).
3.2Economic Outcome Measures
The cost justification of pharmacogenomic tests is a major area of interest among

policy makers and insurance companies and research among health economists. Like
genetic diseases, pharmacogenomic testing requires testing many dozens or even
hundreds of patients to identify one who will benefit from the test result [2]. The higher
3 Economics of Pharmacogenomic Testing in Clinical Practice 37
the incidence of a significant polymorphism and/or the more severe the clinical
consequence of an adverse drug event, the easier it is to justify pharmacogenomic
testing.
Economic measures for pharmacogenomics or any medical intervention can be
assessed by several means. One cost analysis approach is to estimate the actual
medication expenses for providing equivalent medical care for patients with and
without pharmacogenomic testing. For example, clopidogrel (Plavix) is a prodrug
that is effective in blocking platelet function and is used to prevent stenosis in
patients who are treated with percutaneous coronary angioplasty. Patients who are
carriers for the reduced metabolism genotype of CYP2C19 (*2 or *3) have poorer
outcomes than wildtypes. Prasugrel (Efient) is a drug that is not affected by
CYP2C19 polymorphism and is an alternative to clopidogrel. Table 3.1 shows a
hypothetical economic analysis comparing prasugrel drug costs, which are higher
than clopidogrel, to pharmacogenomic testing for clopidogrel and use of prasugrel
for carriers of the null gene.
In another economic model, the costs of providing testing for a population can
be compared to the savings achieved by avoiding an adverse event using an
incidence rate from published reports. The costs of an event can be estimated from
reimbursements given for specific Diagnosis Related Groups (AHRQ) [3]. Table3.2
shows a different hypothetical example for clopidogrel, whereby pharmacogenomic
testing can be justified if the medical intervention of increasing the dosage from the
standard 75 to 150 mg for CYP2C19 carriers can reduce the rate of adverse
outcomes to that of the wildtype [4]. Clinical trials are being conducted to test this
hypothesis. Both of these examples show that under the assumptions made, phar-
macogenomic testing is economically justified. The advantages of pharmacog-
enomic testing will be further enhanced with the availability of generic formula-
tions, as the patent for Plavix expires in November 2011.
The more complete economic model calculates the costs associated for producing
one quality-adjusted life-year (QALY) for a given medical intervention as the main
criteria for cost-effectiveness. The extreme limits of QALY are 0 for death and 1.0
for an individual who is in perfect health. Individuals in varying degrees of ill
Table 3.1 Hypothetical cost-effective pharmacogenomic (PGx) models following angioplasty

with stent placement: Drug expenses model for clopidogrel vs. prasugrel
PGx (prasugrel and
Measuring parameter No PGx (prasugrel only) clopidogrel)
Number of subjects 100 100
CYP2C19 carrier rate NA 25%
PGx testing ($150 ea) NA $15,000
Prasugrel annual drug costsa $200,000 (100 patients) $50,000 (25 patients)
Clopidogrel annual drug costs NA $125,000
Total $200,000 $190,000
Savings $10,000
a
Estimated drug costs of $5.45/day for prasugrel and $4.62/day for clopidogrel, from Cohen DJ.
http://www.cardiovascularbusiness.com/index.php?option=com_articles&view=article&id=18875:
tct-prasugrel-costs-hospitals-less-than-clopidogrel-due-to-less-repeat-pci&division=cvb
38 A.H.B. Wu

with stent placement: Clinical outcomes model for standard vs. high-dose clopidogrel
Measuring parameter No PGx (75 mg dosage) PGx (150 mg dosage)
Number of subjects 100 100
CYP2C19 carrier rate NA 25%
PGx testing ($150 ea) NA $15,000
Adverse event ratea 12% (12 patients) 8% (8 patients)
Annual DRG #122 for AMIb $240,000 $160,000
Clopidogrel annual drug costs $168,000 $210,000c
Total $408,000 $385,000
Savings $23,000
a
Rates taken from Mega etal. [4]
b
AMI DRG: $20,000 each, from Agency for Healthcare Research and Quality. http://www.ahrq.
gov/data/hcup/
c
Double drug costs for 25 patients on 150 mg dosage
health would have fractions between these limits (e.g., a bedridden individual may
have a QALY of 0.5). Table 3.4 shows examples of the impact of medical interven-
tions on QALY [7]. A calculation of the incremental cost-effectiveness ratio (ICER)
is the dollar amount necessary to achieve complete health benefit for a particular
intervention. It is calculated by:
(Costsstandard approach Costsproposed intervention )
ICER =
(QALYstandard QALYproposed intervention )
Health economists in the United States have determined that society in general is
willing to pay an ICER of up to $50,000/QALY for a proposed change in medical
practices [5], although figures as high as $100,000/QALY have been cited [3].
This threshold might be higher in countries where health and family values or their
willingness to pay for these services are higher than in the U.S., and lower in
regions where the population is less economically developed.
There are a few studies that have calculated the additional financial resources
necessary for pharmacogenomic testing relative to standard medical practices in
achieving a QALY. These studies use a Markov model where the medical costs are
calculated for a hypothetical patient who has a disease that can be treated with a
medication whose dosage or selection is based on a pharmacogenomic test [6]. This
base case is meant to be representative of the resources needed to treat the popu-
lation as a whole. Examples of pharmacogenomic Markov decision models for a
chemotherapeutic drug are shown in Fig. 3.1. Table 3.3 lists a hypothetical example
of the pharmacogenomic testing for clopidogrel using the same carrier frequency
rates and genotyping costs as Table 3.1and 3.2 and estimating QALY at 0.80 for a
poststent patient without restenosis or complications and 0.25 for a combination of
death, acute myocardial infarction, and stroke. Table 3.4 lists the ICER for other
representative medical interventions that have been studied [7].
In all of these economic models, the costs for performing genotyping are presumed
at the time of the analysis. In reality, such costs are not static and continue to decrease

with stent placement: Calculation of quality-adjusted life years (QALY) for clopidogrel
pharmacogenomics
Measuring parameter No PGx PGx
Fraction free of events (estimated QALY = 0.80)a 0.88 0.92
Fraction with events (estimated QALY = 0.25)a 0.12 0.08
QALYb 0.736 0.756
Additive QALY for PGx NA 0.020
CYP2C19 carrier ratec NA 25%
Cost for PGx testing per patient NA $150
Cost to identify 1 carrier patient (4 patients tested) NA $600
Expense/QALY ($600/0.022) NA $27,270
a
Estimated QALY for a patient without cardiac disease who has had a successful stent placement without
(QALY = 0.80) and with (QALY = 0 25) an adverse event (death, stroke, acute myocardial infarction)
b
QALY for no PGx: (0.88 0.80) + (0.12 0.25). PGx: (0.92 0.80) + (0.08 0.25)
c
Rates taken from Mega etal. [4]
Table 3.4 Representative QALY

Intervention Disease QALY range
Mammography screening Breast cancer 10,00025,000
Medications Hypertension 10,00060,000
Dialysis End-stage renal disease 50,000100,000
Implantable defibrillators Myocardial infarction and heart failure 30,00070,000
Data from reference Neumann etal. [7]
with technology improvements and commercial competition. Even minor changes in

expenses can have a major economic impact when multiplied by large numbers.
3.3Cost-Effectiveness Studies for Specific Pharmacogenomic

Testing Applications
Pharmacoeconomic studies applied to pharmacogenomic testing is a new field that

will be the focus of many future studies and be of extreme interest to clinicians, labo-
ratorians, policy makers, and payers of medical practices. In the following sections,
specific published studies related to specific pharmacogenomic tests are reviewed.
3.3.1Pharmacogenomic Testing for Thiopurine

Methyltransferase
Economic assessment of thiopurine methyltransferase (TPMT) testing for the

prevention of hematopoietic toxicity for patients treated with azathioprine has been
40 A.H.B. Wu
a Disease-free survival
Drug A: wildtype
Therapeutic selection M
Relapse-free time
PGx test
cancer
Disease-free survival
Drug B: null gene

M
Relapse-free time
b Disease-free survival
PGx-adjusted dose
M
Dosage adjustment
Relapse-free time
cancer
Disease-free survival
Standard dose
M
Relapse-free time
Fig. 3.1 Markov decision model for a hypothetical pharmacogenomic test to chemotherapeutic
efficacy. (a) Pharmacogenomics to determine therapeutic selection. (b) Pharmacogenomics to
determine dosage. M Markov modeling
studied by several investigators. According to Gurwitz, based on the TPMT

deficiency rate, the average genotyping cost to identify one deficient individual was
about $10,000 [8]. This figure must be weighed against the rate of adverse events
and mortality. Priest etal. showed a cost savings of $7 and $78/patient for using a
genotype vs. phenotype testing, respectively, of azathioprine for management of
inflammatory bowel disease [9]. Two pharmacoeconomic studies used a hypotheti-
cal decision analysis model, comparing direct medical costs for conventional
weight-based azathioprine dosing for patients with rheumatological conditions vs. a
dose derived from genotyping for TPMT [10, 11]. In both the Canadian and Korean
studies, the costs and drug drop-out rates were lower for the genotype-dosed model.
These outcomes were largely achieved by the avoidance of dose-related toxicities.
Two European studies have reported similar ICERs of $1,300 and $3,000 and for use
of TPMT testing in patients with inflammatory bowel disease and acute lymphoblas-
tic leukemia, respectively [12, 13] (Table 3.5). These reports suggest that even
though the incidence of TPMT variances is low, physicians planning on using aza-
thioprine should consider testing prior to drug administration and to use alternative
medications for patients at high risk for hematopoietic complications.
Table 3.5 Summary of pharmacogenomic cost-effective studies

Lab test
Drug Genotype costs ICERa References
Azathioprine TPMT b
$50 $1,300 Winter etal. [12]
Azathioprine TPMT $220 $3,000 van den Akker-van Marle etal. [13]
Warfarin 2C9 and VKORC1c $400 $170,000 Eckman etal. [16]
Warfarin 2C9 and VKORC1 $200 $357,000 You etal. [18]
Abacavir HLA*B-5701d $63 $1030,000 Hughes etal. [21]
Abacavir HLA*B-5701 $68 $36,700 Schackman etal. [22]
Clozapine 6-panel $500 $47,000 Perlis etal. [25]
Citalopram HT2Ae $500 $93,000 Perlis etal. [26]
Bupropion Dopamine receptor $300 $3,000 Heitjan etal. [30]
Tobramycin Mit 12s ribosomef $338 $79,300 Veenstra etal. [31]
a
ICER incremental cost-effectiveness ratio
b
TPMT thiopurine methyltransferase.
c
VKORC1 vitamin K epoxide reductase complex 1
d
HLA human lymphocyte antigen
e
HT serotonin receptor
f
Mit mitochondria
3.3.2Pharmacogenomic Testing for Warfarin Dosing
Several hypothetical models have been established to determine if a warfarin dose

based on the pharmacogenomic testing for CYP2C9 and vitamin K epoxide
reductase complex 1 is cost-effective. The Brookings Joint Center for Regulatory
Studies conducted an analysis based on a genotype cost of $350 and a 15 and 50%
reduction in bleeding events and stroke, respectively [14]. Based on these assump-
tions, they concluded that warfarin pharmacogenomics will save nearly $2 billion
dollars in the US per year as a nation. Unfortunately, the few outcome studies that
have been published have not suggested that dosages determined by pharmacog-
enomic testing will result in these outcome improvements.
Two pharmacoeconomic studies used atrial fibrillation as the test case. Leey
etal. concluded that a reduction in the incidence of warfarin complication by 0.1%
would be associated with a cost benefit for pharmacogenomic testing [15]. This was
based on a genotyping cost of $250 and that the use of pharmacogenomic testing
would not lead to potentially harmful modification to an anticoagulation regimen.
In contrast, using a genotyping cost of $400 and an assay turnaround time of 3 days,
Eckman etal. determined an overall ICER of $170,000 for warfarin pharmacog-
enomic testing and concluded that testing under these conditions was not
cost-effective (Table 3.5) [16]. In a subanalysis, these investigators concluded that
an ICER of <$50,000 could be achieved if testing was restricted to patients at high
risk for hemorrhage or if results were available within 24 h for under $200. The
faster reporting and enactment of pharmacogenomic test results could reduce the
incidence of bleeding during the immediate dosing period. The Eckman et al.
models were based on three existing randomized trials of standard vs. pharmacog-
enomic genotyping data (total 429 patients with 11 adverse events). You etal. used
the results from a single randomized trial [17] and computed a higher ICER of
42 A.H.B. Wu
$357,000 [18]. They conclude that cost-effectiveness can only be achieved if the
cost per test was under $47. While the economic conclusions of these latter two
trials are difficult to challenge, they were based on very small data set and the con-
clusions that pharmacogenomic testing is not indicated are premature [19].
Repeated cost-effectiveness estimates will be necessary with the publication of
additional randomized trials, such as the one sponsored by the National Heart, Lung
and Blood Institute [20].
3.3.3Pharmacogenomic Testing for Abacavir
There have only been a few reports examining the cost-effectiveness of pharmacog-
enomic testing for HLA-B*5701 to avoid delayed hypersensitivity reactions (HSR)
in patients taking abacavir. Hughes et al. compared the medical costs of treating
patients with HSR induced by abacavir and abacavir substitutes without pharma-
cogenomic testing vs. testing all subjects and the costs for abacavir substitutes for
those who were positive for HLA-B*5701 and the costs for treating HSR patients
who were negative for HLA-B*5701 and abacavir substitutes [21]. Using an ICER
of dollars per HSR avoided and a test cost of $63, these investigators found values
ranging from $10,000 to $30,000 depending on variability in alternative medica-
tions and medical treatments for HSR (Table 3.5). Similar results were reported by
Schackman etal. [22], who used a genetic test cost of $68 and reported an ICE of
$36,700. In both of these models, pharmacogenomic testing is only cost-viable for
the Caucasian population as other ethnicities have a low or absence incidence of the
HLA-B*5701 genotype. Moreover, the need for pharmacogenomic testing will
decline with the availability of other antiretroviral drugs for treating patients with
human immunodeficiency virus that are less expensive, do not produce HSR, or are
more efficacious than abacavir.
3.3.4Pharmacogenomic Testing for Antipsychotic Drugs
The major economic motivation for pharmacogenomic testing of antipsychotic

drugs involves avoidance of side effects, particularly for drugs that are metabolized
by CYP2D6, and therapeutic selection to maximize drug efficacy. There have been
no cost-effective models studied for CYP2D6 pharmacogenomic testing. One pilot
study showed that psychiatric patients seen at one facility who had a poor 2D6
metabolizer genotype had higher numbers of adverse drug events, higher costs
associated with treatment, and longer duration of stay than in those who were
extensive or intermediate metabolizers [23]. Although the sample sizes were small,
these investigators concluded that individuals who are poor metabolizers will ben-
efit most from genetic testing in terms of therapeutic decisions. Rodriquez-Antona
etal. did a rough estimate of the additional costs needed to treat patients who are
poor metabolizers of drugs used for psychiatry [24]. At $250 for each test, they
estimated a genotyping cost of $3,500 to test 14 patients in order to identify one
who is a poor metabolizer. If the length of stay was on average 7 days per patient
per year longer for the poor vs. extensive metabolizers, as suggested by previous
studies, there would be a cost savings of $4,900, exceeding that of the genotyping
costs. Such a model has not been prospectively tested.
There were two studies that performed economic modeling analysis for a spe-
cific antischizophrenic medication. Perlis etal. examined the impact of pharmacog-
enomic testing of neurotransmitter-receptor related genes for use of clozapine
among schizophrenic patients [25]. The model compared the use of conventional
first- and second-line antipsychotic drugs with no pharmacogenomic testing (and
use of clozapine as the third-line) vs. pharmacogenomic testing and use of clozap-
ine as the first-line drug for positive patients only. These investigators calculated a
cost of $47,000 per QALY for pharmacogenomic testing (Table 3.5). In a more
recent study, Perlis et al. constructed another model for comparing pharmacog-
enomic testing for response to serotonin-selective reuptake inhibitors (SSRIs) [26].
Positive responders are given citalopram vs. bupropion for negative responders.
The ICER vs. conventional strategy without testing was $93,500 (Table 3.5). In both
of these studies a relatively high genotyping cost of $500 was included in the
model, as none of these tests are commercially available. For clozapine response, a
panel of single nucleotide polymorphisms was proposed [27] in the genes for sero-
tonin receptor (5-HT2A and 5-HT2C), serotonin transporter promoter (5-HTT), and
histamine H2. For citalopram, the model was based on genotyping for 5-HT2A.
3.3.5Miscellaneous Other Pharmacoeconomic Studies
There are a number of economic studies that have examined other less widely stud-
ied pharmacogenomic tests. Furuta et al. established a dosing regimen based on
CYP2C19 genotyping for the use of proton pump inhibitor regimens for the eradi-
cation of Helicobacter pylori [28]. They found a higher rate of microbiological
eradication in the tailored vs. standard regimen group (96% vs. 70%, respectively)
with no significant increase in total costs ($669 vs. $657, respectively). The cost of
genotyping for 2C19 was set at $83. In patients with nephropathies, Costa-
Scharplatz etal. examined the cost-effectiveness of pharmacogenomic testing for
an insertion/deletion variance in the angiotensin converting enzyme (ACE) gene for
selection of ACE inhibitors vs. angiotensin II receptor blockers [29]. Using a very
modest genotyping cost of $30, they concluded that the addition of testing resulted
in a reduction of costs with the avoidance of the expenses for chronic hemodialysis.
An assessment of QALY was not made in this study.
For smoking cessation, Heitjan et al. performed an economic model for the
pharmacogenetic testing of an insertion/deletion variance in a dopamine receptor to
select bupropion or transdermal nicotine vs. no testing for all patients for bupro-
pion, transdermal nicotine, varenicline, or no drug at all [30]. The ICER was very
44 A.H.B. Wu
favorable at $3,000 for a genotype cost of $307. Veenstra etal. evaluated the use of
a genetic test for a mutation in mitochondrial 12S ribosomal rRNA to predict hear-
ing loss among cystic fibrosis patients treated with aminoglycosides [31]. With an
ICER of $79,300, they concluded that pharmacogenomic testing was not cost-ef-
fective and could lead to worse patient outcomes with the avoidance of antibiotic
treatment in falsely positive pharmacogenomic test results.
3.4Summary
Novel therapeutic, intervention, and medical practice decisions are becoming

increasingly linked to adherence to clinical practice guidelines, documentation of
medical evidence, and justification by economics. Reimbursement for new services
will be denied without data to support the medical and/or economic advantages.
Pharmacogenomic testing is especially being scrutinized because of the higher
costs associated with genetic testing and the promise and notoriety that this can
enable personalized medicine. Producing accurate effective economic analyses for
pharmacogenomic tests is dependent on the quality of the assumptions made.
Unfortunately, there are very few randomized studies available to fuel the analyses
that have been conducted. Therefore, some of the conclusions made are premature.
Nevertheless, such studies have led to policy changes regarding test utilization. The
announcement that the US Centers for Medicare and Medicaid will not reimburse
laboratories for warfarin pharmacogenomic testing unless such testing is part of a
clinical trial may be an example of a rush to judgment [32]. These important deci-
sions may inhibit the impetus and funding to conduct the randomized studies
needed to document efficacy of pharmacogenomic testing, and such attitudes may
become a self-fulfilling prophecies. A second problem for implementing pharma-
cogenomics is that the majority of those tested have the expected genotype and
therefore no therapeutic alterations are needed. The expenses for the identification
of one affected patient where a medical decision is needed must be justified by the
testing of all others. These economic realities have been the burden of tests for
genetic diseases as well.
References
1. Limdi, N. A., Wadelius, M., Cavallari, L., Eriksson, N., Crawford, D. C., Lee, M. T. M., etal.
(2009). The International Warfarin Pharmacogenetics Consortium. (2009). Estimation of the
warfarin dose with clinical and pharmacogenetic data. The New England Journal of Medicine,
360, 753764.
2. Woodcock, J., & Lesko, L. J. (2009). Pharmacogenetics: Tailoring treatment for the outliers.
3. Agency for Healthcare Research and Quality. Accessed October 15, 2010, http://www.ahrq.gov/
data/hcup.
4. Mega, J. L., Close, S. L., Wiviott, S. D., etal. (2009). Cytochrome P-450 polymorphisms and
response to clopidogrel. The New England Journal of Medicine, 360, 354362.
5. Gold, M. R., Siegel, J. E., Russell, L. B., etal. (1996). Cost-effectiveness in health and medicine
(1st ed.). New York: Oxford University Press.
6. Sonnenberg, F. A., & Beck, J. R. (1993). Markov models in medical decision making: A practical
guide. Medical Decision Making, 13, 322328.
7. Neumann, P. J., Rosen, A. B., & Weinstein, M. C. (2005). Medicare and cost-effectiveness
analysis. The New England Journal of Medicine, 353, 15161522.
8. Gurwitz, D., Rodriquez-Antona, C., Payne, K., etal. (2009). Improving pharmacoviligance in
Europe: TPMT genotyping and phenotyping in the UK and Spain. European Journal of
Human Genetics, 17, 991998.
9. Priest, V. L., Begg, E. J., Gardiner, S. J., etal. (2006). Pharmacoeconomic analyses of azathio-
prine, methotrexate and prospective pharmacogenetic testing for the management of inflam-
matory bowel disease. Pharmacoeconomics, 24, 767781.
10. Marra, C. A., Esdail, J. M., & Anis, A. S. (2002). Practical pharmacogenetics: The cost-effec-
tiveness of screening for thiopurine s-methyltransferase polymorphism in patients with rheuma-
tological conditions treated with azothioprine. The Journal of Rheumatology, 29, 25072512.
11. Oh, K. T., Anis, A. H., & Bae, S. C. (2004). Pharmacoeconomic analysis of thiopurine meth-
yltransferase polymorphism screening by polymerase chain reaction for treatment with aza-
thioprine in Korea. Rheumatology, 43, 156163.
12. Winter, J., Walker, A., Shapiro, D., etal. (2004). Cost-effectiveness of thiopurine methyltrans-
ferase genotype screening in patients about to commence azathioprine therapy for treatment
of inflammatory bowel disease. Alimentary Pharmacology & Therapeutics, 20, 593599.
13. van den Akker-van Marle, M. E., Gurwitz, D., & Detmar, S. B. (2006). Cost-effectiveness of
pharmacogenomics in clinical practice: A case rstudy of thioprine methyltransferase genotyping
in acute lymphoblastic leukemia in Europe. Pharmacogenomics, 7, 783792.
14. McWilliams, A., Lutter, R., & Nardinelli, C. Health care savings from personalized medicine
using genetic testing: The case of warfarin. AEI-Brookings Joint Center for Regulatory
Studies. Accessed October 15, 2010, http://reg-markets.org/admin/authorpdfs/redirect-safely.
php%3Fframe=../pdffiles/wp06-23_topost.pdf
15. Leey, J. A., McCabe, S., Koch, J. A., et al. (2009). Cost-effectiveness of genotype-guided
warfarin therapy for anticoagulation in elderly patients with atrial fibrillation. The American
Journal of Geriatric Pharmacotherapy, 7, 197203.
16. Eckman, M. H., Rosand, J., Greenberg, S. M., etal. (2009). Cost-effectiveness of using phar-
macogenetic information in warfarin dosing for patients with nonvalvular atrial fibrillation.
Annals of Internal Medicine, 150, 7383.
17. Anderson, J. L., Horne, B. D., Stevens, S. M., etal. (2007). Randomized trial of genotype-
guided versus standard warfarin dosing in patients initiating oral anticoagulation. Circulation,
116, 26632570.
18. You, J. H. S., Wong, R. S. M., & Cheng, G. (2009). Potential clinical and economic outcomes
of CYP2C9 and VKORC1 genotype-guided dosing in patients starting warfarin therapy.
Clinical Pharmacology and Therapeutics, 86, 540547.
19. Wu, A. H. B. (2009). Pharmacogenomic testing for warfarin dosing: We are ready now. Expert
Review of Cardiovascular Therapy, 12, 14831485.
20. Clarification of optimum anticoagulation through genetics (COAG) trial. Accessed October 15,
2010, http://coagstudy.org/.
21. Hughes, A. R., Mosteller, M., Bansal, A. T., et al. (2004). Cost-effectiveness analysis of
HLA-B*5701 genotying in abacavir hypersensitivity. Pharmcogenomics, 14, 335341.
22. Schackman, B. R., Scott, C. A., Walensky, R. P, et al. (2008). The cost-effectiveness of
HLA-B*5701 genetic screening to guide initial antiretrovial therapy for HIV. AIDS, 22,
20252037.
23. Chou, W. H., Yan, F. X., de Leon, J., etal. (2000). Extension of a pilot study: Impact from the
cytochrome P450 2D6 polymorphism on outcome and coast associated with severe mental
illness. Journal of Clinical Psychopharmacology, 20, 246251.
46 A.H.B. Wu
24. Rodriquez-Antona, C., Gurwitz, D., & de Leon, J. (2009). CYP2D6 genotyping for psychiatric
patients treated with risperidone: Considerations for cost-effectiveness studies.
25. Perlis, R. H., Ganz, D. A., Avorn, J., et al. (2005). Pharmacogenetic testing in the clinical
management of schizophrenia. A decision-analytic model. Journal of Clinical
Psychopharmacology, 25, 427434.
26. Perlis, R. H., Patrick, A., Smoller, J. W., etal. (2009). When is pharmacogenetic testing for
antidepressant response ready for the clinic? A cost-effectiveness analysis based on data from
the STAR*D study. Neuropsychopharmacology, 34, 22272236.
27. Arranz, M. J., Munro, J., Birkett, J., etal. (2000). Pharmacogenetic prediction of clozapine
response. Lancet, 355, 16151616.
28. Furuta, T., Shirai, N., Kodaira, M., et al. (2007). Pharmacogenomic-based tailored versus
standard therapeutic regimen for eradication of H. pylori. Clinical Pharmacology and
Therapeutics, 81, 521528.
29. Costa-Scharplatz, M., van Asselt, A. D. I., Bachmann, L. M., etal. (2007). Cost-effectiveness
of pharmacogenetic testing to predict treatment response to angiotensin-converting enzyme
inhibitor. Pharmacogenetics and Genomics, 17, 359368.
30. Heitjan, D. F., Asch, D. A., Ray, R., etal. (2008). Cost-effectiveness of pharmacogenetic testing
to tailor smoking-cessation treatment. The Pharmacogenomics Journal, 8, 319399.
31. Veenstra, D. L., Harris, J., Gibson, R. L., etal. (2007). Pharmacogenomic testing to prevent
aminoglycoside-induced hearing loss in cystic fibrosis patients: Potential impact on clinical,
patient, and economic outcomes. Genetics in Medicine, 9, 695704.
32. Centers for Medicare & Medicaid Services. (2009). Decision memo for pharmacogenomic
testing for warfarin response (CAG-00400N). Retrieved August 3, 2009, from https://www
.cms.hhs.gov/mcd/viewdecisionmemo.asp?from2=viewdecisionmemo.asp&id=224&.
Further Reading
Payne, K. (2008). Towards an economic evidence base for pharmacogenetics: Consideration of

outcomes is key. Pharmacogenomics, 9, 14.
Ginsburg, G. S., Konstance, R. P., Allsbrook, J. S., & Schulman, K. A. (2005). Implications of
pharmacogenomics for drug development and clinical practice. Archives of Internal Medicine,
165, 23312336.
Devieru, T., & Vala, M. V. (2006). Overview of the pharmacoeconomics of pharmacogenetics.
Chapter 4
From Personalized Medicine to Personalized
Justice: The Promises of Translational
Pharmacogenomics in the Justice System
Steven H.Y. Wong, Christopher Happy, Daniel Blinka, Suson Goch,

Jeffrey M. Jensen, Joseph M. Donald, Howard Coleman, Saeed A. Jortani,
Yolanda Lurie, Cynthia L. Morris-Kukoski, Manuela G. Newman,
Paul J. Orsulak, Tara Sander, Michael A. Wagner, Jennifer R. Wynn,
Alan H.B. Wu, and Kiang-Teck J. Yeo
Keywords Personalized justice Daubert standard
4.1Introduction
Enabled by Pharmacogenomics (PGx), molecular imaging, and other molecular

biomarkers, personalized medicine (PM) promises to optimize therapy while mini-
mizing side effects. It may also dramatically impact the justice system in ways we
are only beginning to understand.
Personalized medicine has already entered the curricula of well-regarded medical
schools such as that of Johns Hopkins [1], but law schools offer no analog. Although
clinical acceptance of PM has proved slow even with FDA support [2, 3], PMs legal
ramifications are evident. Recently, for example, the FDA relabeled some drugs with
companion PGx [2] such as warfarin with cytochrome P450 (CYP) 2C9 and vitamin
K epoxide reductase complex 1 to reduce bleeding [35]. If PGx retrospectively
reveals that the warfarin patient was at high risk and testing was not initially
*This editorial reflects the consensus of the coauthors and not the official position of opinion of
their respective employers.
This chapter was published in its entirety in: Wong SHY, etal.: Wong SH, Happy C, Blilnka D,
Gock S, Jentzen JM, Donald JM, Coleman H, Jortani SA, Lucire Y, Morris-Kukoski C, Neuman
M, Sander TL, Wagner MA, Wynn JR, Wu AHB, Yeo J. From personalized medicine to personal-
ized justice assembling the framework with pharmacogenomics and assessing its applicability
to the justice system. Pharmacogen 2010;11:731737.
S.H.Y. Wong(*)
Wake Forest University School of Medicine, Department of Pathology,
Winston-Salem, NC 27157, USA
e-mail: shwong@wfubmc.edu
48 S.H.Y. Wong et al.
p erformed, litigation may follow. Indeed, some lawyers advertise on the internet for
cases involving warfarin-related errors [6]. Consequently, PGx may become part of
defensive medicine.
Personalized Justice (PJ) complements PM and the overlapping practice of
translational medicine [710], which hold that individual differences are caused
primarily by genetic and environmental factors. The acronym TSPB captures its
elements in relation to adverse drug reaction (ADR): Toxicity, Sensitivity, impaired
Performance (e.g., driving under the influence of drugs), and Behavioral changes.
Future legal applications may include molecular imaging and analyses genomic,
proteomic, metabolomics, and epigenetics/imprintomics. By comparison, molecular
DNA fingerprinting for identity testing is well accepted [11]. Conceptually, Fig.4.1
proposes a social balance relationship for PM and PJ [10]. In assessing PJ, consider
two index scenarios:
4.1.1Drug Toxicity
A 9-year-old boy, diagnosed with ADHD, obsessive-compulsive disorder, and

Tourettes syndrome, was treated with methylphenidate, clonidine, and fluox-
etine [12]. Over a 10-month period, he developed GI toxicity, incoordination
and disorientation, and seizures. He died from a cardiac arrest. Postmortem
toxicology showed high fluoxetine and norfluoxetine concentrations, and PGx
revealed a poor CYP 2D6 metabolizer, resulting in fluoxetine accumulation and
Personalized
Personalized
Justice
Medicine
PGx Toxicity
Efficacy Sensitivity
Behavior
Performance
Fig. 4.1 Complementary relationship of PM and PJ. (Reproduced and modified with permission
from ref. [10]. In press)
4 From Personalized Medicine to Personalized Justice 49
toxicity. Subsequently, the boys parent was absolved from involvement in flu-
oxetine intoxication. Another example is genotyping uridine 5-diphosphate
(UDP)-glucuronyltransferase 1A1 for patients medicated with irinotecan to
avoid hematopoietic toxicity [13].
4.1.2Drug Sensitivity
In addition to warfarin, one should genotype HLA-B*5701 [14] and HLA-B*1502

[15] for patients medicated with abacavir [14] and some antiepileptics [15], respec-
tively, to avoid Stevens Johnson Syndrome (SSJ). Lawyers use internet advertising
to reach persons so affected [16].
4.1.3Evidence Base for Personalized Justice
In establishing PJ, a firm foundation should be based on sound legal principles as well
as reliable and valid evidence-based studies, not on junk science and unsubstanti-
ated case reports. This lesson resonates in the deficiencies that beset various forensic
sciences recently reported by the National Academy of Science [17, 18]. The
American Academy of Forensic Sciences supports the National Academy of Sciences
13 recommendations and the following principles: The need for strong scientific
foundations; laboratory accreditation; certification of technicians; the standardization
of terminology; ethical protocols; governmental oversight; and the education of legal
professionals, including judges, in forensic scientific methods and principles [1921].
It is imperative that PJ heeds these recommendations, including the study of the rela-
tionship of PGx biomarkers to TSPB and the education of interested parties including
forensic pathologists and toxicologists, those engaged in molecular diagnostics, and
of course, the legal community. Based on the aforementioned assessment, this edito-
rial ushers in the practice of PJ by: differentiating between science and myth, propos-
ing a legal framework, updating the reader on rapidly developing technological
advances, and illustrating scenarios and published cases.
4.2Legal Framework
While personalized medicine is rapidly taking root among the medical sciences,
one may reasonably expect a slower, more begrudging acceptance by the legal
profession. Law is innately conservative and reluctant to accommodate dramatic
change. Cutting-edge developments of all sorts often take decades to gain a
foothold [22]. It will be important, then, to educate judges, lawyers, and legal
academics about the explanatory power of PJ and PM. The laws incredibly rich
experience with DNA developments may, however, facilitate this task [23].
One set of barriers consists of evidence rules, particularly those involving expert
opinion testimony. The vaunted Daubert standard pioneered by the federal courts
and adopted by many states demands that judges serve as gatekeepers who will
ensure that only reliable science is admitted [24]. Although courts have been
distressingly inconsistent in how they scrutinize most sciences [25], DNA evidence
has become the gold standard for forensic sciences. And the DNA channel may
provide a helpful port of entry for PJ.
The prime questions, though, will relate to the role PJ will play in the legal sys-
tem. DNA evidence thus far is narrowly confined to trace evidence: Was this bio-
logical evidence left by the defendant or someone else? A thornier problem occurs
when we attempt to apply biological evidence to moral culpability, which pertains
to an accuseds personal blameworthiness. The Supreme Court recognized in Penry
v. Lynaugh [26] that punishment for a criminal offense should be directly related to
the defendants personal culpability. The concept of personal culpability acknowl-
edges that human choices are shaped by many factors: genetic, neurological, intel-
lectual, educational, social, and environmental. It follows, then, that an individuals
blameworthiness for criminal conduct may vary depending on the factors that
shaped his moral development or compromised his choices.
Thus, from a PJ perspective, the question becomes something like this: Should
courts consider identifiable biological conditions that predispose a person to crimi-
nal behavior in weighing moral culpability? Legal precedent suggests that it should.
Consider Roper v. Simmons [27] where the Supreme Court held that persons under
the age of 18 years could not be subjected to the death penalty because their brains
were not yet fully developed. MRIs and neuro-imaging showed that neuronal
changes in the brain continued into the early twenties. Because the brain affects
behavior, the justices ruled that punishing a person for behavior caused by an
underdeveloped brain (of which the defendant had no choice) violated the prohibi-
tion against cruel and unusual punishment. Similar logic was applied in Atkins v.
Virginia [28], which prohibited subjecting the mentally retarded to the death pen-
alty. Roper and Atkins illustrate the principle that criminal defendants with brain-
based deficits are not as morally culpable as those without. As such, they deserve a
lesser penalty. This is a legal springboard for PJ.
4.3Forensic Pathology Perspectives
For several medical examiner/coroner offices in US and Europe, PGx has served as
an adjunct for drug death certification an emerging practice of molecular autopsy
[9, 10]. Previous studies showed a higher prevalence of CYP 2D6 genetic variations,
corresponding to intermediate and slow metabolizers with decreased or without
enzymatic activity, in the decedents intoxicated with methadone, oxycodone, and
antidepressants [9, 10, 29, 30]. Thus, PGx might aid to interpret the effect of
impaired drug metabolism due to genetic variations. If potentially lethal medications
are identified at the scene with correspondingly toxic drug concentrations of the
decedent and subsequent PGx testing confirms an extensive (normal) metabolizer,

death is certified as suicide. If the deceaseds genotype is a variant resulting in
decreased drug metabolism, death is certified as accident. Recent indications of PJ
for forensic pathology include a PGx section in forensic toxicology texts for medical
examiners by Molina [31] and Karch [32]. Future molecular diagnostics biomarkers
of interest might include epigenetics/imprintomics and gene expression in under-
standing suicide [33, 34], metabolomics, and proteomics [10].
4.4Molecular Diagnostics
The detection of individual genetic variants is at the heart of PM and PJ. Single
nucleotide polymorphisms (SNPs), the most common type of genetic variation,
might affect drug metabolism [35]. Several SNP genotyping technologies facilitate
rapid PGx testing in clinical laboratories. The three main steps DNA extraction,
amplification, and detection may be performed by automated platforms. Biotech
companies offering PGx testing platforms, some with FDA-approval, include:
Luminex xTag, Roche Amplichip, Affymetric DMET chip, Autogenomics
INFINITI, Osmetech eSensors, ParagonDx, and ABI SNaPshot and Taqman assays.
Thus, the laboratory can rapidly develop, validate, and perform PGx testing in-
house within months, further enhanced by readily available quality control products
and survey programs. The limitations include: existing evidence to demonstrate
significant and medically relevant correlations for many disease-causing genes and
variants, limited detection of genetic variants within the context of each testing
platform, clinical interpretation of genotype results including environmental fac-
tors, and transplanted organs interfering with testing.
4.4.1Drug Hypersensitivity In Vitro Diagnostics
In vitro lymphocyte toxicity assay (LTA) compares peripheral blood lymphocytes of

patients with history of ADR to control individuals who take the same drug in the
same dose and do not present any ADR [36]. LTA is based upon the dysfunction of
mitochondria in people hypersensitive to certain drugs such as sulfonamides, non-
steroidal anti-inflammatory, protease inhibitors, and antiepileptics. This test can also
detect possible drugdrug interactions. Dysfunction of mitochondria has severe cel-
lular consequences and is linked to lack of detoxification of drugs in human. Several
surveillance strategies have evolved that limit mitochondrial damage and ensure cel-
lular integrity. Intraorganellar proteases conduct protein quality control and exert
regulatory functions, allowing mitochondria to protect against apoptosis. LTA can be
used in PJ when several drugs are incriminated in an ADR in order to enable distin-
guishing between the drug that produced the reaction and the other drugs, which
have been taken in the same period of time, but did not contribute to the ADR.
4.5Illustrative Cases and Scenarios
4.5.1Alcohol
Alcoholism, with up to 3040% inheritability, is a complex and controversial disease

having both environmental and genetic components. Genetic variations influence
pharmacokinetics/metabolism and pharmacodynamics. Alcohol dehydrogenase and
acetaldehyde dehydrogenase are two main polymorphic enzymes involved in alcohol
metabolism, with minor contribution by CYP 2E1. Pharmacodynamic systems
influenced by PGx are: gamma-aminobutyric acid A/B receptors, glutamate
(N-methyl-D-aspartic acid and a-amino-3-hydroxyl-5-methyl-isoxazole-propionate),
serotonin, voltage-activated calcium channels, dopamine/norepinepherine/acetylcho-
line, and opioid systems. For example, naltrexone, used for detoxification, binds to
opioid receptor m1 and the variants of the candidate gene of this receptor may affect
addiction treatment [37, 38].
4.5.2Antidepressants and Antipsychotics
Personalized justice might address the effect of antidepressant and/or antipsychotics

on behavioral changes. A recent review examined the relationship of violent behavior
to the antidepressants: paroxetine, sertraline, and fluoxetine. Different verdicts in a
series of medicolegal cases reflected the different judicial processes without
considering drug-induced violence [39]. Incidentally, Lucire (Lucire, Y. (2009)
Personal communication) studied patients medicated with antidepressants and
antipsychotics metabolized by polymorphic CYPs and assessed the development of
akathisic, suicidal and/or homicidal ideations, and relationship to CYP genes
variations. The validity of these preliminary observations is pending publication in
peer-reviewed journals and validation by other investigators.
4.5.3Warfarin
Oral warfarin anticoagulation is widely used to prevent thromboembolic events.

Dosing selection is due to a narrow therapeutic range with a large interindividual
variation (20-fold) affected by genetic and nongenetic factors [40, 41]. About
1017% of patients experience bleeding [40]. Genotyping of CYP2C9, CYP
4F2, VKORC1, and relevant clinical factors account up to ~56% of dosing vari-
ability [42, 43] In 2007, the FDA relabeled warfarin to suggest genotyping, fol-
lowed by the January 22, 2010 relabeling [5]: The patients CYP2C9 and
VKORC1 genotype information, when available, can assist in selection of the
starting dose. Previously, in May 2009, the Centers for Medicare & Medicaid
Services recommended against reimbursement [44]. Potential legal culpability
was addressed in the introduction.
4.5.4Pain Management
In addressing pain management with safety, Woodcock of the FDA discussed the
balance of providing patient with efficacious analgesics and the associated risks
[45]. For example, in ultrarapid metabolizers, greater CYP2D6 activity can lead
to poisoning after opioid administration. A 2007 case report detailed a breast-
feeding infant who suffered respiratory depression and died as a result of toxic
amounts of morphine being present in the milk [46]. The mother, later identified
with multiple copies of CYP2D6 genes corresponding to an ultrarapid metabo-
lizer, over-converted codeine to a high amount of morphine. This was
excreted into breast milk, resulting in babys high morphine concentrations identi-
fied in postmortem analysis. Consequently, guidelines were developed for breast-
feeding mother medicated with codeine.
4.6Conclusions
In recognizing the complementary, check and balance relationship of PM and PJ,

translational PGx may serve the promising role of an ADJUNCT biomarker for
interpreting drug-related toxicity and sensitivity. Currently, robust, scientific, and
clinical studies are lacking to substantiate the relationship of PGx and behavioral/
performance changes [39]. These desired PJ studies are challenging to perform due
to ethical and legal consideration and lack of funding. Consequently, interpretations
may be extrapolated from case reports and clinical behavioral and performance
studies, e.g., studies related to driving under the influence of drug. Other advances
include automated platforms and potential use of oral fluid for toxicology and PGx.
Oral fluid, currently being evaluated for forensic drug testing [4750] and
therapeutic drug monitoring, is easily collected for PM and PJ pending on outcome
studies to demonstrate efficacy comparable to blood samples. In considering PGx
for PJ, pros and cons are listed in Table 4.1.
In ushering in PJ practice with PGx, a working group should consist of col-
leagues from interrelated disciplines in order to probe and keep abreast of recent
developments, to grade evidence of case reports and outcome studies, and to
develop inclusion and exclusion criteria. With sound scientific and legal principles
and correct interpretation, a firm and lasting foundation supports the emerging
concept of PJ becoming a reality to enhance patient safety and to maintain social
justice.
Table 4.1 Pros and Cons of using PGx as an adjunct biomarker in personalized justice
Pros
Stability of DNA in postmortem settings
Personalized approach for assessing drug response
Assist in interpretation of drug concentrations in postmortem toxicology and drug death
certification
Assess patient compliance
Turn-around-time suitable for medicolegal/forensic applications
PGx cost is low in comparison to the legal settlement
Might differentiate chronic vs. acute toxicity
Cons
Data available in clinical cases, but limited in postmortem cases
Legal interpretation challenging due to complexity
Drug inhibitors and inducers of enzymes, and environmental factors complicating interpretation
Does not account for posttranslational modifications
Multiple enzyme systems involved in the metabolism
Acknowledgements We acknowledge permission Future Medicine Ltd., http://www.futuremedi-

cine.com/loi/pgs.
Reference
1. Miller, E. D. (2009). A Bold Leap into the Future Personalized medicine is key to the new
Genes to Society curriculum. In S. E. Pasquale (Ed.), Hopkins medicine (p. 48). Baltimore:
Johns Hopkins Medicine.
2. Wu, A. H. B., Babic, N., & Yeo, K. T. T. (2009). Implementation of pharmacogenomics into
the clinical practice of therapeutics: Issues for the clinician and the laboratorian. Personalized
3. Coumadin Tablets (Warfarin Solution Tablets, USP) Crystalline Coumadin for injection
(Warfarin solution for injection, USP). http://www.accessdata.fda.gov/drugsatfda_docs/
label/2010/009218s108lbl.pdf
4. Wadelius, M. & Pirmohamed, M. (2007). Pharmacogenetics of warfarin: Current status and
future challenges. The Pharmacogenomics Journal, 8, 99111.
5. Ray, T. (2010). FDA updates warfarin labeling with PGx-guided dosing ranges.
Pharmacogenomics Report, Feb. 3.
6. Coumadin overdose warfarin overdose, law examples. http://www.spanglaw.com/medical-
malpractice/medication-errors/coumadin/
7. Wong, S. H. (2007). Pharmacogenomics and personalized medicine for drug addiction and
toxicology: Towards personalized justice? 11th Asian Pacific Congress of Clinical
Biochemistry, Beijing, China.
8. Wong, S. H. Y. & Happy, C. (2009). Personalized justice, translational pharmacogenomics
and personalized medicine relevant to the forensic sciences? Tox Talk, 33, 2223.
9. Wong, S. H., Jentzen, J. M., Shi, R. Z., & The Forensic Pathology/Toxicology Methadone
Pharmacogenomics Study Group (FPTMPGxSG). (2008). Personalized medicine enabling
personalized justice: Methadone pharmacogenomics as an adjunct for molecular autopsy,
and for addiction and driving under the influence of drugs (DUID). Clinical Chemistry and
Laboratory Medicine: CCLM/FESCC, 46, A118.
10. Wong, S. H. Y. (2011). Pharmacogenomics as molecular autopsy an adjunct to forensic

pathology/toxicology: From Gregor Mendel to personalized medicine and personalized
justice. In D. Moffat Osselton & B. Widdop (Eds.), Clarkes analysis of drugs and poisons
(4th ed.). London: Royal Pharmaceutical Society Publishing.
11. Jeffreys, A. J., Wilson, V., & Thein, S. L. (1985). Individual-specific fingerprints of human
DNA. Nature, 316, 7679.
12. Sallee, F. R., DeVane, C. L., & Ferrell, R. E. (2000). Fluoxetine-related death in a child with
cytochrome P-450 2D6 genetic deficiency. Journal of Child and Adolescent
13. Cote, J. F., Kirzin, S., Kramar, A., Mosnier, J. F., Diebold, M. D., Soubeyran, I., etal. (2007).
UGT1A1 polymorphism can predict hematologic toxicigy inpatients treated with irinotecan.
Clinical Cancer Research: An Official Journal of the American Association for Cancer
Research, 13, 32693275.
14. Mallal, S., Phillips, E., Carosi, G., Molina, J. M., Workman, C., Tomazic, J., et al. (2008).
HLA-B*5701 Screening for Hypersensitivity to Abacavir. The New England Journal of
Medicine, 358, 568579.
15. Chung, W. H., Hung, S. I., Hong, H. S., Hsih, M. S., Yang, L. C., Ho, H. C., etal. (2004).
Medical Genetics: A marker for Stevens-Johnson Syndrome. Nature, 428, 486.
16. SSJ Law. http://ssjlaw.mobi/
17. Committee on Identifying the Needs of the Forensic Science Community National Research
Council of the National Academies. (2009). Strengthening Forensic Science in the United
States: A Path Forward, National Academy of Sciences (pp. 1352). Washington: The
National Academies Press.
18. Strengthening forensic science in the United States: a path forward http://www.nap.edu/cata-
log/12589.html.
19. American Academy of Forensic Sciences. (2009). The American Academy of Forensic
Sciences approved position statement in response to National Academy of Sciences Forensic
Needs Report. Sept. 4, 2009.
20. Bohan, T. L. (2009). Presidents message. Academy News, 39(1), 3435.
21. The American Academy of Forensic Sciences approves position statement in response to the
National Academy of Sciences Forensic Needs report. http://www.aafs.org/pdf/AAFS_
Position_Statement_for_Press_Distribution_090409.pd
22. Gianelli, P. (2003). Understanding evidence (3rd ed.) (p. 323). Albany: LexisNexis Matthew
Bender.
23. Federal Judicial Center. (2000). Reference manual on scientific evidence (2nd ed.) (p. 487).
Washington.
24. Fed. R. Evid. 702.
25. Kassirer, J. P. & Cecil, J. S. (2002). Inconsistency in evidentiary standards for medical
testimony: Disorder in the courts. The Journal of the American Medical Association, 288,
13821387.
26. 429 U.S. 302 (1989).
27. 543 U.S. 551 (2005).
28. 536 U.S. 304 (2002).
29. Wong, S. H. Y., Wagner, M. A., Jentzen, J. M., Schur, C., Bjerke, J., Gock, S. B., etal. (2003).
Pharmacogenomics as an adjunct of molecular autopsy for forensic pathology/toxicology:
Does genotyping CYP 2D6 serve as an adjunct for certifying methadone toxicity? Journal of
Forensic Sciences, 48, 14061415.
30. Wong, S. H. Y., Gock, S. B., Shi, R. Z., Jin, M., Wagner, M. A., Schur, B. C., etal. (2006).
Pharmacogenomics as an aspect of molecular autopsy for forensic pathology/toxicology.
In S. H. Y. Wong, M. Linder, & R. Jr. Valdes (Eds.), Pharmacogenomics and proteomics:
Enabling the practice of personalized medicine (pp. 311320). Washington: AACC Press.
31. Molina, D. K. (2010). Handbook of forensic toxicology for medical examiners (pp. 1370).
Boca Raton: CRC Press. (2010) (Appendix C PGXs, 3437.)
32. Karch, S. B. (2009). Karchs pathology of drug abuse (4th ed.) (pp. 1709). Boca Raton: CRC
Press.
33. Ernst, C., Mechawar, N., & Turecki, G. (2009). Suicide neurobiology. Progress in Neurobiology,
89, 315333.
34. Dolgin, E. (2009). Epigenetic suicide note. The Scientist, http://www.the-scientist.com/blog/
print/55843/.
35. Kim, S. & Misra, A. (2007). SNP genotyping technologies and biomedical applications.
Annual Review of Biomedical Engineering, 9, 289320.
36. Neuman, M. G., Malkiewicz, I. M., & Shear, N. H. (2000). A novel lymphocyte toxicity assay
to assess drug hypersensitivity syndromes. Clinical Biochemistry, 33, 517524.
37. King, A. C., Volpicelli, J. R., Frazer, A., & oBrien, C. P. (1997). Effect of naltrexone on
subjective alcohol response in subjects at high and low risk for future alcohol dependence.
38. Wang, J. B., Imai, Y., Eppler, C. M., Gregor, P., Spivak, C. E., & Uhl, G. R. (1993). m opiate
receptor: cDNA cloning and expression. Proceedings of the National Academy of Sciences of
the United States of America, 90, 1023010234.
39. Healy, D., Herxheimer, A., & Menkes, D. B. (2006). Antidepressants and violence: Problems
at the Interface of Medicine and Law. PloS, 3, 14781487.
40. Stehle, S., Kirchheiner, u., Lazar, A., & Fuhr, U. (2008). PGXs of oral anticoagulants. Clinical
Pharmacokinet, 47, 565594.
41. Kangelaris, K. N., Bent, S., Nussbaum, R. L., Garcia, D. A., & Tice, J. A. (2009). Genetic
testing before anticoagulation? A systematic review of PGX dosing of warfarin. Journal of
General Internal Medicine, 24, 656664.
42. Caldwell, M. D., Awad, T., Johnson, J. A., Gage, B. F., Falkowski, M., Gardina, P., et al.
(2008). CYP4F2 genetic variant alters required warfarin dose. Blood, 111, 41064112.
43. Gage, B. F., Eby, C., Johnson, J. A., Deych, E., Rieder, M. J., Ridker, P. M., etal. (2008). Use
of PGX and clinical factors to predict the therapeutic dose of warfarin. Clinical Pharmacology
and Therapeutics, 84, 326331.
44. Centers for Medicare & Medicaid Services. https://www.cms.hhs.gov/mcd/viewdraftdecisionmemo.
asp?from2=viewdraftdecisionmemo.asp&id=224&
45. Woodcock, J. (2009). A difficult balance pain management, drug safety and the FDA. The
New England Journal of Medicine, 361, 21052107.
46. Madadi, P., Koren, G., Cairns, J., Chitayat, D., Gaedigk, A., Leeder, J. S., et al. (2007).
Safety of codeine during breastfeeding: Fatal morphine poisoning in the breastfed neonate
of a mother prescribed codeine. Canadian Family Physician Mdecin de famille canadien,
53, 3335.
47. Bosker, W. M. & Huestis, M. A. (2009). Oral fluid testing for drugs of abuse. Clinical
Chemistry, 55, 19101931.
48. Cone, E. J. & Huestis, M. A. (2007). Interpretation of oral fluid tests for drugs of abuse.
Annals of the New York Academy of Sciences, 1098, 51103.
49. Pil, K. & Verstraete, A. Z. (2008). Current developments in drug testing in oral fluids.
Therapeutic Drug Monitoring, 30, 196202.
50. Bush, D. M. (2008). The US mandatory guidelines for federal workplace drug testing programs:
Current status and future considerations. Forensic Science International, 174, 111119.
Part II
Chemotherapeutics
Chapter 5
Irinotecan
R. Stephanie Huang, Federico Innocenti, and Mark J. Ratain
Keywords Irinotecan UGT1A1 Pharmacogenetics
5.1Pharmacology
Irinotecan (Fig. 5.1 also called CPT-11 or 7-ethyl-10-[4-(1-piperidino)-1-piperidino]

carbonyloxycamptothecin) is a semisynthetic analog of the natural alkaloid camp-
tothecin. It was first approved in Japan in 1994 for small-cell lung cancer and
hematologic malignancies, followed by approvals in France for the treatment of
advanced colorectal cancer in 1995 [1]. Irinotecan was introduced in the US in
1996 and received a full U.S. Food and Drug Administration (FDA) approval for
the treatment of colorectal cancer in 1998. Currently in the US, irinotecan is
primarily utilized for patients with metastatic colorectal carcinoma whose disease
has recurred or progressed following initial fluorouracil-based therapy. However,
the drug has demonstrated efficacy in treating a wide variety of neoplasms and was
approved for the use of treating lung and breast cancer in Japan [2].
The development of irinotecan started in the late 1980s in Japan. In an anticancer
drug activity screen conducted in a HST-1 human squamous carcinoma cell line,
SN-38, a main active metabolite of irinotecan, was found to augment anticancer
activity in combination with cisplatin, mitomycin C, 5-fluorouracil (5-FU), and
etoposide [3]. The invitro assay also showed greater activity in colon and hepato-
cellular carcinoma cell lines when comparing SN-38 to cisplatin, mitomycin C,
doxorubicin and 5-FU [4]. Furthermore, irinotecan demonstrated therapeutic
R.S. Huang(*)
Department of Medicine, Committee on Clinical Pharmacology and Pharmacogenomics,
and Cancer Research Center, The University of Chicago, 900 E. 57th Street,
KCBD Room 7148, Chicago, IL 60637, USA
e-mail: rhuang@medicine.bsd.uchicago.edu
60 R.S. Huang et al.
Fig. 5.1 Structure of CPT-11
e fficacy when tested in a panel of human tumor xenografts derived from adult and
pediatric central nervous system malignancies [5], from human testicular embryonal
carcinomas [6], and from human ovarian cancer and soft-tissue sarcoma lines
grown in nude mice [7].
In a series of phase II single-agent irinotecan trials conducted in Japan in the
early 1990s, promising antitumor activity was observed in non-small cell lung [8],
small cell lung cancer [9], uterine, cervical and ovarian cancer [10], gastric cancer
[11], metastatic colorectal cancer [12], pancreatic cancer [13], breast cancer [14]
as well as refractory leukemia and lymphoma [15, 16]. These results were subse-
quently confirmed outside of Japan [17, 18].
The primary use of irinotecan (outside of Japan) is in treating advanced
colorectal cancer. The response rate to irinotecan as a single agent has ranged from
17 to 27% [19]. Efficacy was demonstrated both in chemotherapy-nave patients
and those who progressed after 5-FU-based chemotherapy [2023]. As a single
agent, irinotecan can be given at 125 mg/m2 weekly (with intermittent breaks), or
as a single 350 mg/m2 dose every 3 weeks [24]. Combinations of irinotecan with
5-FU and leucovorin (FOLFIRI) resulted in significant improvements in objective
tumor response rates, time to tumor progression, and survival when compared with
5-FU/LV alone [2528]. Guichard etal. showed that the schedule of irinotecan and
5-FU combinations administration is a critical parameter for chemotherapeutic
efficacy both in vitro and in vivo [29]. In addition to FOLFIRI, other irinotecan
containing therapies, including FOLFOXIRI (irinotecan, oxaliplatin, 5-FU, and
leucovorin) and the cetuximab irinotecan regimen, are also recommended by the
National Comprehensive Cancer Network (NCCN) practice guidelines in treating
advanced colon and rectal cancer [24, 30].
5.2Pharmacokinetics
After intravenous infusion, irinotecan plasma concentrations decline in a multi

exponential manner, with a mean terminal elimination half-life of 612 h. The mean
terminal elimination half-life of the active metabolite SN-38 is 1020 h [3133].
5 Irinotecan 61
Large interindividual variability in the pharmacokinetics of irinotecan was observed.

Over the recommended dose range of 50350 mg/m2, the plasma area under the
curve concentration (AUC) of irinotecan increases linearly with dose [3134].
Maximum concentrations of the active metabolite SN-38 are generally seen within
2-h following the end of a 90-min infusion of irinotecan [33]. SN-38 rebound con-
centrations were observed in many courses at about 0.51 h following the end of the
i.v. infusion, which is suggestive of enterohepatic recycling [35]. Irinotecan exhibits
moderate plasma protein binding (3068% bound) while SN-38 is highly bound to
human plasma proteins (approximately 95% bound). The plasma protein to which
irinotecan and SN-38 predominantly bind is albumin.
Irinotecan is metabolized to 7-ethyl-10-[4-N-(5-aminopentanoic acid)-1-
piperidino] carbonyloxycamptothecin (APC) [36, 37] or 7-ethyl-10-(4-amino-1-
piperidino) carbonyloxycamptothecin (NPC) [38] and potential other intermediate
metabolites [39] via a cytochrome P450 mediated process [40]. CYP2A6, CYP2C9,
and CYP3A4/5 show roles in irinotecan metabolism invitro [41]; while CYP3A4
is the most relevant CYP isoform in irinotecan metabolism invivo. Neither APC
nor NPC contributes directly to irinotecan invivo activity. NPC is further converted
to SN-38 (7-ethyl-10-hydroxy-camptothecin) by carboxylesterase [42, 43]. SN-38
is subsequently conjugated predominantly by the enzyme UDP-glucuronosyl trans-
ferase 1A1 (UGT1A1) to form a glucuronide metabolite (SN-38G) [44] (http://
www.pharmgkb.org/search/pathway/irinotecan/liver.jsp). Rivory etal. showed that
the transformation of SN-38 to the glucuronide is the rate-limiting step in the elimi-
nation of SN-38 [45].
The disposition of irinotecan has not been fully elucidated in humans. The
urinary excretion of irinotecan is 1120% of the initial dose; SN-38, <1%; and
SN-38G, 3%. The cumulative biliary and urinary excretion of irinotecan and its
metabolites (SN-38 and SN-38G) over a period of 48 h following administration
of irinotecan in two patients ranged from approximately 25% (100 mg/m2) to
50% (300 mg/m2) of the initial dose. The biliary excretion of the carboxylate
forms of irinotecan and SN-38 and SN-38G is mediated by multiple transporters,
including ABCB1 [46], ABCC2 [46], and ABCG2 [47]. The contribution of each
transporter differs greatly [48, 49]. This was further supported by increased sen-
sitivity to irinotecan and SN-38 after antisense cMOAT cDNA transfection in the
HepG2 cell line [50]. Total clearance of irinotecan is about 14.6 6.4 L/h/m2, and
it does not vary with increased dosage [51]. It has been reported that body mass
index, age, and sex may be independent predictors of pharmacokinetic parameters
of irinotecan [52].
5.3Pharmacodynamics
Irinotecan and its active metabolite SN-38 bind to the topoisomerase I-DNA
complex and prevent religation of these single-strand breaks. Research suggests
that the cytotoxicity of irinotecan is due to double-strand DNA damage pro-
duced during DNA synthesis when replication enzymes interact with the ternary
complex formed by topoisomerase I (TopoI), DNA, and either irinotecan or

SN-38 [53]. Mammalian cells cannot efficiently repair these double-strand
breaks.
SN-38 is approximately 1,000 times as potent as irinotecan as an inhibitor of
topoisomerase I purified from human and rodent tumor cell lines. In vitro cytotoxic-
ity assays show that the potency of SN-38 relative to irinotecan varies from 2- to
2,000-fold. SN-38G had 1/50 to 1/100 the activity of SN-38 in cytotoxicity assays
using two cell lines invitro.
5.4Toxicity
Expected irinotecan toxic effects include gastrointestinal and hematological

complications, such as nausea, vomiting, diarrhea, and infection. The more severe
toxicities, namely severe early and late onset diarrhea and neutropenia, occurred in
approximately 30% of patients depending on dose and schedule [21]. Fuchs etal.
reported 36 and 19% of grade 3/4 diarrhea occurred in patients treated with irino-
tecan weekly or once every 3 weeks, respectively. Furthermore, grade 3/4 neutro-
penia was found to occur at a similar rate in both the weekly and once every
3weeks regimen groups (29 and 34%, respectively) [54]. In treatment combination,
the rates of grade 3 and 4 diarrhea have been reported to be approximately 10 and
15% for FOLFIRI and the combination of bevacizumab and FOLFIRI, respectively
[55]. Febrile neutropenia, a severe form of neutropenia that requires immediate
medical attention, is observed in 311% of colorectal cancer patients who under-
went FOLFIRI treatment [26, 56].
Diarrhea associated with irinotecan administration is probably the result of
the enterocolitis caused by high levels of SN-38 retained for a long period in
the intestine [57]. Gupta etal. observed high correlation between glucuronida-
tion of SN-38 and the severity of diarrhea and suggest that modulation of
glucuronidation may affect irinotecan therapeutic outcomes [58]. Furthermore,
Takasuna etal. demonstrated that the inhibition of the beta-glucuronidase activity
in the intestinal microflora may ameliorate the diarrhea caused by irinotecan in
rats [59].
Given the potential adverse effects, careful monitoring of the white blood cell
(WBC) count with differential, hemoglobin, and platelet count is recommended
before each dose of irinotecan. Concomitant medications such as antiemetics, atro-
pine, and loperamide [60] can be given to patients for prophylaxis and/or manage-
ment of symptoms from treatment. The ASCO guideline for the management of
treatment-induced diarrhea can be found in Engs review [61]. Furthermore, empir-
ical antimicrobial therapy can be given both in case of diarrhea and febrile neutro-
penia [62]. According to the most recent US guidelines, treatment with
colony-stimulating factors is appropriate for patients with a greater than 20% risk
of febrile neutropenia [63].
5 Irinotecan 63
5.5Pharmacogenetics
Tumor-specific somatic mutations and abnormal gene expression as well as germline

genetic variations have been reported to be associated with irinotecan therapeutic
efficacy and toxicity [6468]. However, to date, the role of somatic mutations has
not been confirmed to be significant in irinotecan therapeutic outcomes.
Germline DNA variations may affect both pharmacokinetics and pharmacody-
namics of irinotecan. The most well-known example is between UGT1A1 genetic
variation and irinotecan-induced toxicity. The UGT1A gene locus has been mapped
to chromosome 2q37. The entire UGT1A locus spans approximately 200 kb. To date,
at least ten functional UGT1A proteins are known to be produced from this single
gene locus composed of alternative first exons shared with four common exons [69].
The genetic organization of the UGT1A gene locus enables a tissue-specific gene
expression of hepatic and extrahepatic UGTs and most likely ensures that a broad
array of differing substrates can undergo glucuronidation in humans [70]. Genetic
variations within the UGT1A gene locus are common with over 100 single-nucleotide
polymorphisms (SNPs) within the promoter regions and the UGT1A coding
sequence, many of which are found to be in linkage disequilibrium with each other
[70]. A detailed UGT1A1 allele nomenclature can be found at the following website
http://www.pharmacogenomics.pha.ulaval.ca/webdav/site/pharmacogenomics/
shared/Nomenclature/UGT1A/UGT1A1.htm (accessed on January 2010).
A case report showed that individuals with Gilberts syndrome, a benign form of
familial hyperbilirubinemia, have an enhanced risk for irinotecan toxicity [71]. This
suggests a potential genetic basis for irinotecan-related toxicity in the UGT*1.1 gene.
Iyer etal. demonstrated that UGT1A1 is the isoform responsible for the glucuroni-
dation (inactivation) of the active metabolite of irinotecan, SN-38. These findings
indicate a genetic predisposition to the metabolism of irinotecan, suggesting that
patients with low UGT1A1 activity, such as those with Gilberts syndrome, may be
at an increased risk for irinotecan toxicity [44]. A small study conducted in Japan
supports this hypothesis that UGT1A1 homozygote 7/7 TATA box genotype is asso-
ciated with high SN-38/SN-38G metabolic ratio, which leads to impaired SN-38
glucuronidation [72]. Bosma etal. showed that the insertion of an additional repeat
(TA7) is associated with a decrease in UGT1A1 expression and consequently
decreased glucuronidation of its targets [73]. Further investigation showed that this
common insertion/deletion in the UGT1A1 promoter TATA box reduces the tran-
scriptional efficiency of the gene, with an inverse correlation between the number
of TA repeats (5, 6, 7, 8 alleles) and transcriptional efficiency [74]. Among these
TATA box variations, the 6 allele is classified as *1, while the 7 allele is classified
as *28. Homozygosity for the *28 allele is the most common variant associated
with Gilberts syndrome in Caucasians [73]. The 5 allele (*36) and the 8 allele
(*37) repeats are rare in Caucasians, but they appear to be more common in
African-Americans. Several UGT1A1 variants are almost exclusively found in
Asians. For example, UGT1A1*6 allele is a nonsynonymous coding variant
211G>A missense in UGT1A1 exon 1, resulting in glycine 71 alteration to arginine
(G71R) [75]. UGT1A1*27 (P229Q) has a frequency of ~3% in Asians [76], and
UGT1A1*7 (Tyr486Asp) is very rare [77].
It has been shown that UGT1A1*28 and *6 polymorphisms correlate with
reduced glucuronidation activity toward SN-38 and bilirubin [77, 78]. Individuals
who are homozygous for the UGT1A1*28 allele commonly suffer dose-limiting
neutropenia through decreased degradation and clearance of SN-38 [79]. This
genotypephenotype association has been confirmed by multiple studies with various
significance and effect size [80]. For example, a prospective study of adult cancer
patients who received irinotecan monotherapy demonstrated that patients who car-
ried two (TA7) alleles had a 50% incidence of grade 4 neutropenia, while those who
were heterozygous for (TA7) or carried no (TA7) alleles had a 12.5 and 0% incidence,
respectively [81]. A recent meta-analysis further established that the incidence of
toxicity in UGT1A1*28 patients was positively correlated with the dose used, as
genotypephenotype association was only significant at medium or high irinotecan
treatment doses (>250 mg/m2, every 3 weeks) [82]. This is further supported by
pediatric studies of low-dose irinotecan, which show little association between the
*28 allele and toxicity [83, 84]. A recent prospective trial conducted in Italy found
that the maximum tolerated dose (MTD) of irinotecan in FOLFIRI is 310 mg/m2 in
patients with the *1/*28 genotype and 370 mg/m2 in those with the *1/*1 genotype
[85]. The relationship between higher irinotecan dose and better treatment effi-
ciency remains to be evaluated; however, data seem to point toward the utility of
UGT1A1*28 genotype as a potential therapeutic guideline in optimizing irinotecan
treatment efficacy and minimizing toxicity.
Despite the significant correlation observed between UGT1A1 genetic variation
and irinotecan induced toxicity, the UGT1A1*28 genetic test has median positive
predictive value of 0.5 and median negative predictive value of 0.85 [80]. Despite
a high rate of false positives, this test can be useful to reach informed decisions
about how to select among alternative effective therapies, like the ones available for
metastatic colorectal cancer. Further research efforts have involved the studies of
other candidate genes and more recently combination of genetic and nongenetic
factors in order to improve the predictive value of the test. For example, genetic
variations in other glucuronosyltransferases (e.g., UGT1A7, UGT1A9) [8688] and
transporters (e.g., ABCB1, ABCC2, ABCG2) [8992] have also been suggested to
contribute to variability in irinotecan pharmacokinetics and toxicity. A principal
component analysis to estimate irinotecan pharmacokinetic variation confirmed the
role of polymorphisms in UGT1A1, 1A7, and 1A9 in the irinotecan SN-38 pathway,
which involves the conversion from irinotecan to SN-38 as well as the enterohe-
patic recirculation of SN-38 [93]. More recently, the functional significance of
SLCO1B1 variations has been demonstrated. SLCO1B1 plays a significant role in
SN-38 transportation [94]. A haplotype variation in SLCO1B1 (*15) exhibited
decreased transport activities for SN-38 invitro [94]. This genetic effect was seen
in Asian cancer patients who carry SLCO1B1*15, with decreased irinotecan clear-
ance and subsequently increased exposure to irinotecan [95]. Increased irinotecan
related toxicities were also observed in lung cancer patients who carry SLCO1B1
variations [96]. A case report demonstrated a 66-year-old Japanese male with
5 Irinotecan 65
pharyngeal carcinoma who carried UGT1A1*6/*28 and SLCO1B1*15/15 genotypes

had extensive SN-38 accumulation and severe irinotecan induced toxicity after
one-cycle of treatment [97]. A multivariate analysis showed SLCO1B1 521TC or
CC and UGT1A1*6/*6 genotypes were independently predictive for grade 4
neutropenia in non-small cell lung cancer patients treated with irinotecan [98]. This
was further supported by a combined genetic predictors and patient characteristics
model in predicting neutropenia and pharmacokinetics of irinotecan, in which
genetic variations in drug transporters (e.g., ABCC1 and SLCO1B1) appear to have
a significant impact [99]. Although the additive effect of drug transporter genetic
polymorphisms and UGT1A1 on irinotecan pharmacokinetics and neutropenia has
been reported [100], the additional effects of these transporter polymorphisms
remain small [101]. Furthermore, the genetic polymorphisms in irinotecan target
genes have been extensively evaluated. Recently, a UBC9 10920CG genotype was
reported to be associated with increased irinotecan sensitivity in non-small cell lung
cancer patients although these results have not yet been replicated [102]. Variations
in other pharmacodynamic genes (TOP1, PARP1, TDP1, and XRCC1) were not
associated with irinotecan-induced neutropenia, despite initial positive findings
[103, 104]. Larger studies, either candidate gene or genome-wide association, are
required for both hypothesis generation and replication.
5.6Pharmacoethnicity
The frequency of the UGT1A1 *1/*1, *1/*28, and *28/*28 genotypes varies greatly
among different ethnic groups [80]. In Caucasians, genotype frequencies are 56,
2836 and 917%, respectively. In Africans, they are 1330, 3850, and 1733%,
respectively [105]. In Asians, they are 6584, 1531, and 14%, respectively; while
the UGT1A1*6 allele is almost exclusively found in Asians (~20%) [75]. Han etal.
suggested that UGT1A1*28 testing alone may not be sufficient to predict
irinotecan-induced toxicity in patients of Asian origin. Instead, a combined *28 and
*6 test is more appropriate [87, 106]. In addition, it is well established that not only
allele frequencies but also the composition of haplotype blocks is influenced by the
population structure [107]. In Japanese patients, genetic linkage has been reported
between UGT1A1*6 and UGT1A7 and 1A9 polymorphisms [108]. This linkage
disequilibrium seems common in Asians in general. High variability of alleles and
haplotypes of UGT1A1, 1A6, and 1A7 was observed in the So Miguel Island popu-
lation with a strong interaction between functional polymorphisms related to the
alteration of the UGT enzyme activity [109]. In fact, UGT1A haplotype-based
approach has been shown to be an efficacious strategy to predict FOLFIRI treat-
ment outcomes [110].
Racial disparities have been observed in tumor response rate and severe adverse
events in Caucasian and African-American colorectal cancer patients after adjusting
for age, sex, performance status, and dose-intensity [111]. In a NCI-sponsored trial
(N9741), significant racial differences in the distribution of polymorphisms in key
candidate genes were also observed between races, suggesting that this disparity
may be in part due to varied genetic frequency in different ethnic groups.
Interestingly, grade 3 toxicity was higher in whites (48%) than in blacks (34%),
largely due to the higher rate of grade 3 diarrhea in the white patients, despite
thefact that the UGT1A1*28 genotype is more common in blacks than whites. The
authors concluded that a single genotypic difference is unlikely to account for the
observed racial variation in adverse effects and response rate; rather, if these differ-
ences are genetically determined, they are likely mediated by a complex interplay
of genotypes. In contrast to their findings, Gupta etal. reported the lack of race and
sex effects on the plasma availability of irinotecan, SN-38, and SN-38G as well as
in the incidence and severity of toxicity when treating cancer patients with single-
agent irinotecan [51]. Given the variation in genetic composition of UGTs in dif-
ferent ethnic groups, it is plausible that similar phenotypic outcomes may be
produced by different genetic profiles. Nonetheless, it would be ideal to incorporate
a wide range of ethnic groups to address unequally distributed alleles.
5.7Clinical Utility of UGT1A1 Genotyping
Based on pharmacogenetic evidence, the FDA approved the addition of a warning

to the irinotecan label (June 2005) (http://www.fda.gov/medwatch/SAFETY/2005/
Jun_PI/Camptosar_PI.pdf) [112]. The label warns that homozygosity for
UGT1A1*28 is a risk factor for severe neutropenia and that patients with this
genotype should be treated with a reduced starting dose of irinotecan. A com-
mercial genetic test (the Invader UGT1A1 Molecular Assay [Third Wave
Technologies]) was also approved by the FDA in 2005 for the detection of
UGT1A1*28, a first for any chemotherapy agent [113115]. However, other
UGT1A1 variants (e.g., *6) that also result in reduced enzyme activity in Asians
have not been included in the label warning or in the UGT1A1 Invader assay.
Baseline serum bilirubin level has been evaluated in predicting toxicity or effi-
cacy among patients receiving irinotecan for metastatic colorectal cancer.
Although modest elevations of bilirubin are associated with increased grade 34
neutropenia in patients treated with weekly irinotecan, baseline serum bilirubin
does not reliably predict overall irinotecan-related toxicity or efficacy [116].
Later additional methods, including DNA sequencing and fragment analysis,
have been compared to the Invader assay. All three methods were valuable for
genotyping the UGT1A1 (TA)n repeat, with the sequencing and size-based assays
having the fewest drawbacks [117]. ODwyer et al. have proposed a practical
approach in the utility of UGT1A1*28 allele testing in the clinic [118]. Currently,
the National Comprehensive Cancer Network (NCCN) practice guideline in
oncology for colon and rectal cancers states, Irinotecan should be used with cau-
tion and with decreased doses in patients with Gilberts disease or elevated serum
bilirubin. There is a commercially available test for UGT1A1. Guidelines for use
in clinical practice have not been established [24].
5 Irinotecan 67
The discovery and validation of UGT1A1 genetic variants and the establishment
of useful genotyping methods in predicting irinotecan-related toxicity were a para-
digmatic success in pharmacogenetic research and translational work. However,
UGT1A1 genotyping is not routinely performed in predicting irinotecan toxicity in
current clinical practice. For example, Gardiner surveyed Australian and New
Zealand laboratories that offered genetic testing and found very few performing
pharmacogenomic testing [119]. Corkindale etal. looked at reasons why pharma-
cogenetic tests are not used and listed a number of factors [120]. Among them were
lack of a clinical authority to use for interpretation, lack of peer recognition of the
tests, no understanding of the cost implications of the test, and difficulty in getting
practical information about the tests. In a more recent review, issues for the clinical
laboratories were also pointed out, including the availability of FDA-cleared tests,
the absence of reimbursement codes, the need for genotyping accuracy, and the
need to find clinical expertise to interpret laboratory results [121].
Low allelic penetrance, heterogeneity in patient populations and treatment
regimens, unaccounted geneenvironment interactions, and differences in outcome
measures across studies hamper the precise assessment of the clinical performance
of the UGT1A1 genotyping test [80]. Furthermore, a complex trait like drug
response is likely a result of many factors, with the combination of genetic and
environmental contributions. There are many appropriate explanations for the low
usage of the UGT1A1 genotyping test [115]. For example, although genotyping has
been consistently associated with hematological toxicity induced by higher doses
of irinotecan, the risk is reduced at the lower doses used in combination regimens
[82]. The difficulty in interpreting different genetic markers in different ethnic
groups has limited the utility of UGT1A1 genetic testing in irinotecan toxicity pre-
diction. Furthermore, the irinotecan package insert advises that a reduction in the
starting dose by at least one level should be considered for patients known to be
homozygous for the UGT1A1*28 allele. However, the precise optimal dose reduc-
tion in this patient population is not known. In fact, the prognostic impact of
UGT1A1*28 has not been established [122]. To date, no studies have been per-
formed to demonstrate preserved efficacy in irinotecan dose reduction. Another
concern is the lack of established reimbursement and the lack of education of clini-
cians about the potential value (and limitations) of testing [112]. In addition, to
answer the question of how UGT1A1 testing could add to the safety or efficacy of
irinotecan as compared to the current protocol of adjusting drug dose on the basis
of standard clinical tests such as WBC counts [123], pharmacoeconomic evaluation
is needed.
5.8Conclusion
Understanding the reasons for treatment failure and developing an ability to predict
those patients who would benefit the most (and least) remain important aims in
medicine [124]. Germline variations in the UGT1A1 gene locus have an impact on
irinotecan therapy induced toxicity, and this information has been acknowledged by
the FDA on the irinotecan label. However, the results of a genetic test need to be
integrated in the context of the clinical picture of each individual patient. The devel-
opment of pharmacogenetic models of drug toxicity risk must include independent
variables related to the characteristics of the patients, the disease, concomitant
medications, and other variables. The data accumulated so far and the information
added to the revised label suggest that the UGT1A1*28 testing will not be a manda-
tory test in the clinic but, rather, its use will be at the discretion of the treating physi-
cian, at least in the short term [80]. The future lies in the discovery of additional
genetic and nongenetic markers through genome-wide association studies and
combination of germline and cancer pharmacogenetics to establish more precise
irinotecan toxicity and efficacy prediction models.
Acknowledgments The authors would like to thank Dr. Ryan Munoz for critical review of this
chapter.
References
1. PRNewswire. Campto(R) (irinotecan/CPT-11) receives regulatory approval in France for

treatment of advanced colorectal cancer May 19, 1995. http://www.thefreelibrary.com/
CAMPTO(R) (irinotecan/CPT-11) receives regulatory approval in France...-a016884603
2. Shigeoka, Y., Itoh, K., Igarashi, T., etal. (2001). Clinical effect of irinotecan in advanced and
metastatic breast cancer patients previously treated with doxorubicin- and docetaxel-containing
regimens. Japanese Journal of Clinical Oncology, 31, 370374.
3. Masumoto, N., Nakano, S., Esaki, T., etal. (1995). Sequence-dependent modulation of anti-
cancer drug activities by 7-ethyl-10-hydroxycamptothecin in an HST-1 human squamous
carcinoma cell line. Anticancer Research, 15, 405409.
4. Matsuoka, H., Yano, K., Seo, Y., etal. (1995). Cytotoxicity of CPT-11 for gastrointestinal
cancer cells cultured on fixed-contact-sensitive plates. Anti-Cancer Drugs, 6, 413418.
5. Hare, C., Elion, G., Houghton, P., etal. (1997). Therapeutic efficacy of the topoisomerase I
inhibitor 7-ethyl-10-(4-[1-piperidino]-1-piperidino)-carbonyloxy-camptothecin against pedi-
atric and adult central nervous system tumor xenografts. Cancer Chemotherapy and
6. Miki, T., Sawada, M., Nonomura, N., etal. (1997). Antitumor effect of CPT-11, a camptoth-
ecin derivative, on human testicular tumor xenografts in nude mice. European Urology, 31,
9296.
7. Jansen, W., Kolfschoten, G., Erkelens, C., Van Ark-Otte, J., Pinedo, H., & Boven, E. (1997).
Anti-tumor activity of CPT-11 in experimental human ovarian cancer and human soft-tissue
sarcoma. International Journal of Cancer, 73, 891896.
8. Fukuoka, M., Niitani, H., Suzuki, A., etal. (1992). A phase II study of CPT-11, a new deriva-
tive of camptothecin, for previously untreated non-small-cell lung cancer. Journal of Clinical
Oncology, 10, 1620.
9. Masuda, N., Fukuoka, M., Kusunoki, Y., etal. (1992). CPT-11: A new derivative of camp-
tothecin for the treatment of refractory or relapsed smallcell lung cancer. Journal of Clinical
Oncology, 10, 12251229.
5 Irinotecan 69
10. Takeuchi, S., Dobashi, K., Fujimoto, S., etal. (1991). A late phase II study of CPT-11 on
uterine cervical cancer and ovarian cancer. Research Groups of CPT-11 in gynecologic cancers.
Gan To Kagaku Ryoho, 18, 16811689.
11. Futatsuki, K., Wakui, A., Nakao, I., etal. (1994). Late phase II study of irinotecan hydrochloride
(CPT-11) in advanced gastric cancer. CPT-11 Gastrointestinal Cancer Study Group. Gan To
Kagaku Ryoho, 21, 10331038.
12. Shimada, Y., Yoshino, M., Wakui, A., etal. (1993). Phase II study of CPT-11, a new camp-
tothecin derivative, in metastatic colorectal cancer. CPT-11 Gastrointestinal Cancer Study
Group. Journal of Clinical Oncology, 11, 909913.
13. Sakata, Y., Shimada, Y., Yoshino, M., etal. (1994). A late phase II study of CPT-11, irinote-
can hydrochloride, in patients with advanced pancreatic cancer. CPT-11 Study Group on
Gastrointestinal Cancer. Gan To Kagaku Ryoho, 21, 10391046.
14. Taguchi, T., Tominaga, T., Ogawa, M., Ishida, T., Morimoto, K., & Ogawa, N. (1994). A late
phase II study of CPT-11 (irinotecan) in advanced breast cancer. CPT-11 Study Group on
Breast Cancer. Gan To Kagaku Ryoho, 21, 10171024.
15. Ohno, R., Okada, K., Masaoka, T., etal. (1990). An early phase II study of CPT-11: A new
derivative of camptothecin, for the treatment of leukemia and lymphoma. Journal of Clinical
Oncology, 8, 19071912.
16. Ota, K., Ohno, R., Shirakawa, S., etal. (1994). Late phase II clinical study of irinotecan hydro-
chloride (CPT-11) in the treatment of malignant lymphoma and acute leukemia. The CPT-11
Research Group for Hematological Malignancies. Gan To Kagaku Ryoho, 21, 10471055.
17. Wagener, D., Verdonk, H., Dirix, L., etal. (1995). Phase II trial of CPT-11 in patients with
advanced pancreatic cancer, an EORTC early clinical trials group study. Annals of Oncology,
6, 129132.
18. Verschraegen, C. F., Levy, T., Kudelka, A. P., etal. (1997). Phase II study of irinotecan in
prior chemotherapy-treated squamous cell carcinoma of the cervix. Journal of Clinical
Oncology, 15, 625631.
19. Shimada, Y., Rougier, P., & Pitot, H. (1996). Efficacy of CPT-11 (irinotecan) as a single
agent in metastatic colorectal cancer. European Journal of Cancer, 32A(Suppl. 3),
S13S17.
20. Pitot, H. C., Wender, D. B., OConnell, M. J., etal. (1997). Phase II trial of irinotecan in
patients with metastatic colorectal carcinoma. Journal of Clinical Oncology, 15,
29102919.
21. Rothenberg, M. (1998). Efficacy and toxicity of irinotecan in patients with colorectal cancer.
Seminars in Oncology, 25(5 Suppl. 11), 3946.
22. Rothenberg, M. L., Eckardt, J. R., Kuhn, J. G., etal. (1996). Phase II trial of irinotecan in
patients with progressive or rapidly recurrent colorectal cancer. Journal of Clinical Oncology,
14, 11281135.
23. Rougier, P., Bugat, R., Douillard, J. Y., etal. (1997). Phase II study of irinotecan in the treat-
ment of advanced colorectal cancer in chemotherapy-naive patients and patients pretreated
with fluorouracil-based chemotherapy. Journal of Clinical Oncology, 15, 251260.
24. Engstrom, P., Arnoletti, J. P., & Benson, A., et al.(2010). NCCN colon cancer practice
guidelines in oncology v.1.2010. Oncology (Williston Park), 1. http://www.nccn.org/
professionals/physician_gls/PDF/colon.pdf
25. Douillard, J. Y., Cunningham, D., Roth, A. D., etal. (2000). Irinotecan combined with fluo-
rouracil compared with fluorouracil alone as first-line treatment for metastatic colorectal
cancer: A multicentre randomised trial. Lancet, 355, 1041.
26. Ducreux, M., Ychou, M., Seitz, J. F., etal. (1999). Irinotecan combined with bolus fluorou-
racil, continuous infusion fluorouracil, and high-dose leucovorin every two weeks (LV5FU2
regimen): A clinical dose-finding and pharmacokinetic study in patients with pretreated
metastatic colorectal cancer. Journal of Clinical Oncology, 17, 2901.
27. Mitry, E., Douillard, J. Y., Van Cutsem, E., et al. (2004). Predictive factors of survival in
patients with advanced colorectal cancer: An individual data analysis of 602 patients
included in irinotecan phase III trials. Annals of Oncology, 15, 10131017.
28. Saltz, L. B., Cox, J. V., Blanke, C., etal. (2000). Irinotecan plus fluorouracil and leucovorin
for metastatic colorectal cancer. The New England Journal of Medicine, 343, 905914.
29. Guichard, S., Cussac, D., Hennebelle, I., Bugat, R., & Canal, P. (1997). Sequence-dependent
activity of the irinotecan-5FU combination in human colon-cancer model HT-29 invitro and
invivo. International Journal of Cancer, 73, 729734.
30. Cunningham, D., Humblet, Y., Siena, S., etal. (2004). Cetuximab monotherapy and cetux-
imab plus irinotecan in irinotecan-refractory metastatic colorectal cancer. The New England
Journal of Medicine, 351, 337345.
31. Masuda, N., Fukuoka, M., Kudoh, S., et al. (1993). Phase I and pharmacologic study of
irinotecan in combination with cisplatin for advanced lung cancer. British Journal of Cancer,
68, 777782.
32. Rothenberg, M. L., Kuhn, J. G., Burris, H. A. III, etal. (1993). Phase I and pharmacokinetic
trial of weekly CPT-11. Journal of Clinical Oncology, 11, 21942204.
33. Rowinsky, E. K., Grochow, L. B., Ettinger, D. S., etal. (1994). Phase I and pharmacological
study of the novel topoisomerase I inhibitor 7-Ethyl-10-[4-(1-piperidino)-1-piperidino]car-
bonyloxycamptothecin (CPT-11) administered as a ninety-minute infusion every 3 weeks.
Cancer Research, 54, 427436.
34. de Forni, M., Bugat, R., Chabot, G. G., etal. (1994). Phase I and pharmacokinetic study of
the camptothecin derivative irinotecan, administered on a weekly schedule in cancer patients.
Cancer Research, 54, 43474354.
35. Catimel, G., Chabot, G., Guastalla, J., etal. (1995). Phase I and pharmacokinetic study of
irinotecan (CPT-11) administered daily for three consecutive days every three weeks in
patients with advanced solid tumors. Annals of Oncology, 6, 133140.
36. Haaz, M. C., Rivory, L., Riche, C., Vernillet, L., & Robert, J. (1998). Metabolism of irinote-
can (CPT-11) by human hepatic microsomes: Participation of cytochrome P-450 3A and
drug interactions. Cancer Research, 58, 468472.
37. Rivory, L. P., Riou, J. F., Haaz, M. C., etal. (1996). Identification and properties of a major
plasma metabolite of irinotecan (CPT-11) isolated from the plasma of patients. Cancer
Research, 56, 36893694.
38. Dodds, H. M., Haaz, M. C., Riou, J. F., Robert, J., & Rivory, L. P. (1998). Identification of
a new metabolite of CPT-11 (irinotecan): Pharmacological properties and activation to
SN-38. The Journal of Pharmacology and Experimental Therapeutics, 286, 578583.
39. Sai, K., Kaniwa, N., Ozawa, S., & Sawada, J. I. (2001). A new metabolite of irinotecan in
which formation is mediated by human hepatic cytochrome P-450 3a4. Drug Metabolism
and Disposition, 29, 15051513.
40. Lokiec, F., du Sorbier, B. M., & Sanderink, G. J. (1996). Irinotecan (CPT-11) metabolites in
human bile and urine. Clinical Cancer Research, 2, 19431949.
41. Hanioka, N., Ozawa, S., Jinno, H., etal. (2002). Interaction of irinotecan (CPT-11) and its
active metabolite 7-ethyl-10-hydroxycamptothecin (SN-38) with human cytochrome P450
enzymes. Drug Metabolism and Disposition, 30, 391396.
42. Danks, M. K., Morton, C. L., Pawlik, C. A., & Potter, P. M. (1998). Overexpression of a
rabbit liver carboxylesterase sensitizes human tumor cells to CPT-11. Cancer Research, 58,
2022.
43. Rivory, L. P., Bowles, M. R., Robert, J., & Pond, S. M. (1996). Conversion of irinotecan
(CPT-11) to its active metabolite, 7-ethyl-10-hydroxycamptothecin (SN-38), by human liver
carboxylesterase. Biochemical Pharmacology, 52, 1103.
44. Iyer, L., King, C., Whitington, P., etal. (1998). Genetic predisposition to the metabolism of
irinotecan (CPT-11). Role of uridine diphosphate glucuronosyltransferase isoform 1A1 in
the glucuronidation of its active metabolite (SN-38) in human liver microsomes. The Journal
of Clinical Investigation, 101, 847854.
5 Irinotecan 71
45. Rivory, L., & Robert, J. (1995). Identification and kinetics of a beta-glucuronide metabolite
of SN-38 in human plasma after administration of the camptothecin derivative irinotecan.
Cancer Chemotherapy and Pharmacology, 36, 176179.
46. Luo, F. R., Paranjpe, P. V., Guo, A., Rubin, E., & Sinko, P. (2002). Intestinal transport of
irinotecan in Caco-2 cells and MDCK II cells overexpressing efflux transporters Pgp,
cMOAT, and MRP1. Drug Metabolism and Disposition, 30, 763770.
47. Katayama, R., Koike, S., Sato, S., Sugimoto, Y., Tsuruo, T., & Fujita, N. (2009). Dofequidar
fumarate sensitizes cancer stem-like side population cells to chemotherapeutic drugs by
inhibiting ABCG2/BCRP-mediated drug export. Cancer Science, 100, 20602068.
48. Chu, X. Y., Kato, Y., Niinuma, K., Sudo, K. I., Hakusui, H., & Sugiyama, Y. (1997).
Multispecific organic anion transporter is responsible for the biliary excretion of the camp-
tothecin derivative iinrotecan and its metabolites in rats. The Journal of Pharmacology and
Experimental Therapeutics, 281, 304314.
49. Chu, X. Y., Kato, Y., & Sugiyama, Y. (1997). Multiplicity of biliary excretion mechanisms
for irinotecan, CPT-11, and its metabolites in rats. Cancer Research, 57, 19341938.
50. Koike, K., Kawabe, T., Tanaka, T., etal. (1997). A canalicular multispecific organic anion
transporter (cMOAT) antisense cDNA enhances drug sensitivity in human hepatic cancer
cells. Cancer Research, 57, 54755479.
51. Gupta, E., Mick, R., Ramirez, J., et al. (1997). Pharmacokinetic and pharmacodynamic
evaluation of the topoisomerase inhibitor irinotecan in cancer patients. Journal of Clinical
Oncology, 15, 15021510.
52. Miya, T., Goya, T., Fujii, H., etal. (2001). Factors affecting the pharmacokinetics of CPT-11:
The body mass index, age and sex are independent predictors of pharmacokinetic parameters
of CPT-11. Investigational New Drugs, 19, 6167.
53. Slichenmyer, W., & Von Hoff, D. (1990). New natural products in cancer chemotherapy.
Journal of Clinical Pharmacology, 30, 770788.
54. Fuchs, C. S., Moore, M. R., Harker, G., Villa, L., Rinaldi, D., & Hecht, J. R. (2003). Phase
III comparison of two irinotecan dosing regimens in second-line therapy of metastatic col-
orectal cancer. Journal of Clinical Oncology, 21, 807814.
55. Tam, V., Rask, S., Koru-Sengul, T., & Dhesy-Thind, S. (2009). Generalizability of toxicity
data from oncology clinical trials to clinical practice: Toxicity of irinotecan-based regimens
in patients with metastatic colorectal cancer. Current Oncology, 16, 1320.
56. Falcone, A., Ricci, S., Brunetti, I., et al. (2007). Phase III trial of infusional fluorouracil,
leucovorin, oxaliplatin, and irinotecan (FOLFOXIRI) compared with infusional fluorouracil,
leucovorin, and irinotecan (FOLFIRI) as first-line treatment for metastatic colorectal cancer:
The Gruppo Oncologico Nord Ovest. Journal of Clinical Oncology, 25, 16701676.
57. Araki, E., Ishikawa, M., Ligo, M., Koide, T., Itabashi, M., & Hoshi, A. (1993). Relationship
between development of diarrhea and the concentration of SN-38, an active metabolite of
CPT-11, in the intestine and the blood plasma of athymic mice following intraperitoneal
administration of CPT-11. Japanese Journal of Cancer Research, 84, 697702.
58. Gupta, E., Lestingi, T. M., Mick, R., Ramirez, J., Vokes, E. E., & Ratain, M. J. (1994).
Metabolic fate of irinotecan in humans: Correlation of glucuronidation with diarrhea. Cancer
Research, 54, 37233725.
59. Takasuna, K., Hagiwara, T., Hirohashi, M., etal. (1996). Involvement of {beta}-glucuronidase
in intestinal microflora in the intestinal toxicity of the antitumor camptothecin derivative
irinotecan hydrochloride (CPT-11) in rats. Cancer Research, 56, 37523757.
60. Abigerges, D., Armand, J. P., Chabot, G. G., etal. (1994). Irinotecan (CPT-11) high-dose
escalation using intensive high-dose loperamide to control diarrhea. Journal of the National
Cancer Institute, 86, 446449.
61. Eng, C. (2009). Toxic effects and their management: Daily clinical challenges in the treat-
ment of colorectal cancer. Nature Reviews. Clinical Oncology, 6, 207.
62. Sipsas, N., Bodey, G., & Kontoyiannis, D. (2005). Perspectives for the management of
febrile neutropenic patients with cancer in the 21st century. Cancer, 103, 11031113.
63. Crawford, J., Allen, J., Armitage, J., etal. (2010). NCCN Myeloid growth factors practice
guidelines in oncology v.1.2010. Oncology (Williston Park), 1. http://www.nccn.org/
professionals/physician_gls/PDF/myeloid_growth.pdf
64. Nemunaitis, J., Cox, J., Meyer, W., Courtney, A., & Mues, G. (1997). Irinotecan hydrochloride
(CPT-11) resistance identified by K-ras mutation in patients with progressive colon cancer after
treatment with 5-fluorouracil (5-FU). American Journal of Clinical Oncology, 20, 527529.
65. McLeod, H. L., & Keith, W. (1996). Variation in topoisomerase I gene copy number as a
mechanism for intrinsic drug sensitivity. British Journal of Cancer, 74, 508512.
66. Fukuoka, K., Adachi, J., Nishio, K., etal. (1997). p16INK4 expression is associated with the
increased sensitivity of human non-small cell lung cancer cells to DNA topoisomerase
I inhibitors. Japanese Journal of Cancer Research, 88, 10091016.
67. Singh, A., Tong, A., Ognoskie, N., Meyer, W., & Nemunaitis, J. (1998). Improved survival
in patients with advanced colorectal carcinoma failing 5-fluorouracil who received irinote-
can hydrochloride and have high intratumor C-fos expression. American Journal of Clinical
Oncology, 21, 466469.
68. Jansen, W., Hulscher, T., van Ark-Otte, J., Giaccone, G., Pinedo, H., & Boven, E. (1998).
CPT-11 sensitivity in relation to the expression of P170-glycoprotein and multidrug resis-
tance-associated protein. British Journal of Cancer, 77, 359365.
69. Girard, H., Lvesque, E., Bellemare, J., Journault, K., Caillier, B., & Guillemette, C. (2007).
Genetic diversity at the UGT1 locus is amplified by a novel 3 alternative splicing mecha-
nism leading to nine additional UGT1A proteins that act as regulators of glucuronidation
activity. Pharmacogenetics and Genomics, 17, 10771089.
70. Strassburg, C., Kalthoff, S., & Ehmer, U. (2008). Variability and function of family 1
uridine-5-diphosphate glucuronosyltransferases (UGT1A). Critical Reviews in Clinical
Laboratory Sciences, 45, 485530.
71. Wasserman, E., Myara, A., Lokiec, F., etal. (1997). Severe CPT-11 toxicity in patients with
Gilberts syndrome: Two case reports. Annals of Oncology, 8, 10491051.
72. Ando, Y., Saka, H., Asai, G., Sugiura, S., Shimokata, K., & Kamataki, T. (1998). UGT1A1
genotypes and glucuronidation of SN-38, the active metabolite of irinotecan. Annals of
Oncology, 9, 845847.
73. Bosma, P. J., Chowdhury, J. R., Bakker, C., etal. (1995). The genetic basis of the reduced
expression of bilirubin UDP-glucuronosyltransferase 1 in Gilberts syndrome. The New
74. Beutler, E., Gelbart, T., & Demina, A. (1998). Racial variability in the UDP-
glucuronosyltransferase 1 (UGT1A1) promoter: A balanced polymorphism for regulation of
bilirubin metabolism. Proceedings of the National Academy of Sciences of the United States
of America, 95, 81708174.
75. Akaba, K., Kimura, T., Sasaki, A., etal. (1998). Neonatal hyperbilirubinemia and mutation
of the bilirubin uridine diphosphate-glucuronosyltransferase gene: A common missense
mutation among Japanese, Koreans and Chinese. Biochemistry and Molecular Biology
International, 46, 2126.
76. Ando, Y., Fujita, K., Sasaki, Y., & Hasegawa, Y. (2007). UGT1AI*6 and UGT1A1*27 for
individualized irinotecan chemotherapy. Current Opinion in Molecular Therapeutics, 9,
258262.
77. Jinno, H., Tanaka-Kagawa, T., Hanioka, N., et al. (2003). Glucuronidation of 7-Ethyl-10-
hydroxycamptothecin (SN-38), an active metabolite of irinotecan (CPT-11), by human UGT1A1
variants, G71R, P229Q, and Y486D. Drug Metabolism and Disposition, 31, 108113.
78. Iyer, L., Hall, D., Das, S., etal. (1999). Phenotype-genotype correlation of invitro SN-38
(active metabolite of irinotecan) and bilirubin glucuronidation in human liver tissue with
UGT1A1 promoter polymorphism. Clinical Pharmacology and Therapeutics, 65, 576582.
79. Iyer, L., Das, S., Janisch, L., etal. (2002). UGT1A1*28 polymorphism as a determinant of
irinotecan disposition and toxicity. The Pharmacogenomics Journal, 2, 4347.
80. Innocenti, F., & Ratain, M. J. (2006). Pharmacogenetics of irinotecan: Clinical perspectives
on the utility of genotyping. Pharmacogenomics, 7, 12111221.
5 Irinotecan 73
81. Innocenti, F., Undevia, S., Iyer, L., et al. (2004). Genetic variants in the UDP-
glucuronosyltransferase 1A1 gene predict the risk of severe neutropenia of irinotecan.
Journal of Clinical Oncology, 22, 13821388.
82. Hoskins, J., Goldberg, R., Qu, P., Ibrahim, J., & McLeod, H. (2007). UGT1A1*28 genotype
and irinotecan-induced neutropenia: Dose matters. Journal of the National Cancer Institute,
99, 12901295.
83. Bomgaars, L. R., Bernstein, M., Krailo, M., etal. (2007). Phase II trial of irinotecan in chil-
dren with refractory solid tumors: A childrens oncology group study. Journal of Clinical
Oncology, 25, 46224627.
84. Stewart, C. F., Panetta, J. C., OShaughnessy, M. A., etal. (2007). UGT1A1 Promoter genotype
correlates with SN-38 pharmacokinetics, but not severe toxicity in patients receiving low-
dose irinotecan. Journal of Clinical Oncology, 25, 25942600.
85. Toffoli, G., Cecchin, E., Gasparini, G., etal. (2010). Genotype-driven phase I study of irino-
tecan administered in combination with fluorouracil/leucovorin in patients with metastatic
colorectal cancer. Journal of Clinical Oncology, 28, 866871.
86. Carlini, L., Meropol, N., Bever, J., et al. (2005). UGT1A7 and UGT1A9 polymorphisms
predict response and toxicity in colorectal cancer patients treated with capecitabine/irinote-
can. Clinical Cancer Research, 11, 12261236.
87. Han, J., Lim, H., Shin, E., etal. (2006). Comprehensive analysis of UGT1A polymorphisms
predictive for pharmacokinetics and treatment outcome in patients with non-small-cell lung
cancer treated with irinotecan and cisplatin. Journal of Clinical Oncology, 24, 22372244.
88. Innocenti, F., Liu, W., Chen, P., Desai, A., Das, S., & Ratain, M. (2005). Haplotypes of vari-
ants in the UDP-glucuronosyltransferase1A9 and 1A1 genes. Pharmacogenetics and
Genomics, 15, 295301.
89. de Jong, F., Scott-Horton, T., Kroetz, D., et al. (2007). Irinotecan-induced diarrhea:
Functional significance of the polymorphic ABCC2 transporter protein. Clinical
90. Sai, K., Kaniwa, N., Itoda, M., etal. (2003). Haplotype analysis of ABCB1/MDR1 blocks in
a Japanese population reveals genotype-dependent renal clearance of irinotecan.
Pharmacogenetics, 13, 741757.
91. Zhou, Q., Sparreboom, A., Tan, E., etal. (2005). Pharmacogenetic profiling across the irino-
tecan pathway in Asian patients with cancer. British Journal of Clinical Pharmacology, 59,
415424.
92. Mathijssen, R. H. J., Marsh, S., Karlsson, M. O., etal. (2003). Irinotecan pathway genotype
analysis to predict pharmacokinetics. Clinical Cancer Research, 9, 32463253.
93. Rosner, G. L., Panetta, J. C., Innocenti, F., & Ratain, M. J. (2008). Pharmacogenetic pathway
analysis of irinotecan. Clinical Pharmacology and Therapeutics, 84, 393.
94. Nozawa, T., Minami, H., Sugiura, S., Tsuji, A., & Tamai, I. (2005). Role of organic anion
transporter OATP1B1 (OATP-C) in hepatic uptake of irinotecan and its active metabolite,
7-ethyl-10-hydroxycamptothecin: In vitro evidence and effect of single nucleotide polymor-
phisms. Drug Metabolism and Disposition, 33, 434439.
95. Xiang, X., Jada, S., Li, H., etal. (2006). Pharmacogenetics of SLCO1B1 gene and the impact
of *1b and *15 haplotypes on irinotecan disposition in Asian cancer patients. Pharmacogenetics
and Genomics, 16, 683691.
96. Han, J., Lim, H., Shin, E., etal. (2008). Influence of the organic anion-transporting polypep-
tide 1B1 (OATP1B1) polymorphisms on irinotecan-pharmacokinetics and clinical outcome
of patients with advanced non-small cell lung cancer. Lung Cancer (Amsterdam, Netherlands),
69, 6975.
97. Takane, H., Kawamoto, K., Sasaki, T., etal. (2009). Life-threatening toxicities in a patient
with UGT1A1*6/*28 and SLCO1B1*15/*15 genotypes after irinotecan-based chemother-
apy. Cancer Chemotherapy and Pharmacology, 63, 11651169.
98. Han, J., Lim, H., Park, Y., Lee, S., & Lee, J. (2009). Integrated pharmacogenetic prediction
of irinotecan pharmacokinetics and toxicity in patients with advanced non-small cell lung
cancer. Lung Cancer (Amsterdam, Netherlands), 63, 115120.
99. Innocenti, F., Kroetz, D. L., Schuetz, E., et al. (2009). Comprehensive pharmacogenetic
analysis of irinotecan neutropenia and pharmacokinetics. Journal of Clinical Oncology, 27,
26042614.
100. Sai, K., Saito, Y., Maekawa, K., etal. (2010). Additive effects of drug transporter genetic
polymorphisms on irinotecan pharmacokinetics/pharmacodynamics in Japanese cancer
patients. Cancer Chemotherapy and Pharmacology, 66, 95105.
101. Onoue, M., Terada, T., Kobayashi, M., etal. (2009). UGT1A1*6 polymorphism is most pre-
dictive of severe neutropenia induced by irinotecan in Japanese cancer patients. International
Journal of Clinical Oncology/Japan Society of Clinical Oncology, 14, 136142.
102. Han, J. Y., Lee, G. K., Yoo, S. Y., etal. (2009). Association of SUMO1 and UBC9 genotypes
with tumor response in non-small-cell lung cancer treated with irinotecan-based chemo-
therapy. The Pharmacogenomics Journal, 10, 8693.
103. Hoskins, J., Rosner, G., Ratain, M. J., McLeod, H. L., & Innocenti, F. (2009). Pharmacodynamic
genes do not influence risk of neutropenia in cancer patients treated with moderately high-
dose irinotecan. Pharmacogenomics, 10, 11391146.
104. Hoskins, J. M., Marcuello, E., Altes, A., et al. (2008). Irinotecan pharmacogenetics:
Influence of pharmacodynamic genes. Clinical Cancer Research, 14, 17881796.
105. Hall, D., Ybazeta, G., Destro-Bisol, G., Petzl-Erler, M., & Di Rienzo, A. (1999). Variability
at the uridine diphosphate glucuronosyltransferase 1A1 promoter in human populations and
primates. Pharmacogenetics, 9, 591599.
106. Innocenti, F., Vokes, E. E., & Ratain, M. J. (2006). Irinogenetics: What is the right star?
Journal of Clinical Oncology, 24, 22212224.
107. Innocenti, F., Grimsley, C., Das, S., et al. (2002). Haplotype structure of the UDP-
glucuronosyltransferase 1A1 promoter in different ethnic groups. Pharmacogenetics, 12,
725733.
108. Fujita, K., Ando, Y., Nagashima, F., etal. (2007). Genetic linkage of UGT1A7 and UGT1A9
polymorphisms to UGT1A1*6 is associated with reduced activity for SN-38 in Japanese
patients with cancer. Cancer Chemotherapy and Pharmacology, 60, 515522.
109. Pacheco, P., Brilhante, M., Ballart, C., et al. (2009). UGT1A1, UGT1A6 and UGT1A7
genetic analysis: Repercussion for irinotecan pharmacogenetics in the So Miguel Island
Population (Azores, Portugal). Molecular Diagnosis & Therapy, 13, 261268.
110. Cecchin, E., Innocenti, F., DAndrea, M., et al. (2009). Predictive role of the UGT1A1,
UGT1A7, and UGT1A9 genetic variants and their haplotypes on the outcome of metastatic
colorectal cancer patients treated with fluorouracil, leucovorin, and irinotecan. Journal of
Clinical Oncology, 27, 24572465.
111. Sanoff, H. K., Sargent, D. J., Green, E. M., McLeod, H. L., & Goldberg, R. M. (2009). Racial
differences in advanced colorectal cancer outcomes and pharmacogenetics: A subgroup
analysis of a large randomized clinical trial. Journal of Clinical Oncology, 27, 41094115.
112. Huang, R. S., & Ratain, M. J. (2009). Pharmacogenetics and pharmacogenomics of antican-
cer agents. CA: A Cancer Journal for Clinicians, 59, 4255.
113. Hasegawa, Y., Sarashina, T., Ando, M., et al. (2004). Rapid detection of UGT1A1 gene
polymorphisms by newly developed Invader assay. Clinical Chemistry, 50, 14791480.
114. Maitland, M., Vasisht, K., & Ratain, M. (2006). TPMT, UGT1A1 and DPYD: Genotyping
to ensure safer cancer therapy? Trends in Pharmacological Sciences, 27, 432437.
115. Ratain, M. J. (2006). From bedside to bench to bedside to clinical practice: An odyssey with
irinotecan. Clinical Cancer Research, 12, 16581660.
116. Meyerhardt, J., Kwok, A., Ratain, M., McGovren, J., & Fuchs, C. (2004). Relationship of
baseline serum bilirubin to efficacy and toxicity of single-agent irinotecan in patients with
metastatic colorectal cancer. Journal of Clinical Oncology, 22, 14391446.
117. Baudhuin, L., Highsmith, W., Skierka, J., Holtegaard, L., Moore, B., & OKane, D. (2007).
Comparison of three methods for genotyping the UGT1A1 (TA)n repeat polymorphism.
Clinical Biochemistry, 40, 710717.
5 Irinotecan 75
118. ODwyer, P. J., & Catalano, R. B. (2006). Uridine diphosphate glucuronosyltransferase

(UGT) 1A1 and irinotecan: Practical pharmacogenomics arrives in cancer therapy. Journal
of Clinical Oncology, 24, 45344538.
119. Gardiner, S., & Begg, E. (2005). Pharmacogenetic testing for drug metabolizing enzymes:
Isit happening in practice? Pharmacogenetics and Genomics, 15, 365369.
120. Corkindale, D., Ward, H., & McKinnon, R. (2007). Low adoption of pharmacogenetic testing:
An exploration and explanation of the reasons in Australia. Personalized Medicine, 4,
191199.
121. Wu, A. H., Babic, N., & Yeo, K. T. J. (2009). Implementation of pharmacogenomics into the
clinical practice of therapeutics: Issues for the clinician and the laboratorian. Personalized
122. Liu, C., Chen, P., Chiou, T., et al. (2008). UGT1A1*28 polymorphism predicts
irinotecan-induced severe toxicities without affecting treatment outcome and survival in
patients with metastatic colorectal carcinoma. Cancer, 112, 19321940.
123. Haga, S., Thummel, K., & Burke, W. (2006). Adding pharmacogenetics information to drug
labels: Lessons learned. Pharmacogenetics and Genomics, 16, 847854.
124. Walther, A., Johnstone, E., Swanton, C., Midgley, R., Tomlinson, I., & Kerr, D. (2009). Genetic
prognostic and predictive markers in colorectal cancer. Nature Reviews. Cancer, 9, 489.
Chapter 6
Pharmacogenomics of Tamoxifen
Christine L.H. Snozek, Alicia Algeciras-Schimnich, Matthew P. Goetz,

and Loralie J. Langman
Keywords Estrogen receptors Adjuvant therapy Breast cancer Endoxifen
6.1Introduction
Tamoxifen is used in the treatment of estrogen receptor (ER) positive breast cancers
and in breast cancer prophylaxis for high-risk women [1, 2]. The benefits of tamox-
ifen are apparent, as the drug successfully reduces rates of recurrence and mortality
in patients with ER-positive breast cancer [3]. However, these benefits are not with-
out risk; adverse effects range from hot flashes to endometrial cancer and life-
threatening thromboembolism [2]. The use of genetic information to predict
response to tamoxifen therapy holds the potential to improve therapeutic outcome
while minimizing toxicity. This chapter will discuss aspects of tamoxifen pharma-
cology and pharmacogenetics, with case studies to explore the clinical utility of
genotype testing in tamoxifen therapy.
6.2Tamoxifen Pharmacology
Tamoxifen is the best-known selective estrogen receptor modulator (SERM), a family

of drugs, which also includes toremifene and raloxifene. SERMs affect ER activa-
tion to alter gene transcription and other sequelae of ER function [4]. Tamoxifen is
FDA-approved for the treatment of breast cancer in both men and women, from
carcinoma in situ to metastatic disease [5]. It is also approved for breast cancer
L.J. Langman(*)
e-mail: langman.loralie@mayo.edu
78 C.L.H. Snozek et al.
prophylaxis in high-risk women. Off-label uses include treatment of infertility from

anovulation or oligospermia, prevention of osteoporosis, and therapy for gyneco-
mastia and precocious puberty.
Tamoxifen is a competitive inhibitor of estrogen binding to the ER [6], but its
pharmacological activity varies between tissues and is thought to depend on the
profile of transcriptional coactivators expressed [7]. In the breast, tamoxifen acts as
an ER antagonist, disrupting estrogen binding and turning off ER-mediated prolif-
eration and survival signals. In contrast, both bone and the endometrium respond to
tamoxifen as an ER agonist, resulting in increased bone density and endometrial
proliferation [8]. Tamoxifen also has estrogen-like effects on serum lipid profiles
although no long-term cardiovascular benefit has been reported [9].
Tamoxifen is given orally and is well absorbed, with a concentration peak
occurring 37 h after dose administration [4]. Common side effects include hot
flashes, nausea, and breakthrough vaginal bleeding in postmenopausal women;
these rarely necessitate cessation of tamoxifen therapy. More serious adverse
responses have been reported, however, such as increased incidence of thromboem-
bolic events and endometrial cancer. Such adverse responses may contribute to the
finding that therapy with tamoxifen beyond 5 years fails to provide further benefit
to breast cancer patients and may worsen outcome with additional use [7]. For these
reasons, long-term tamoxifen use is generally limited to patients in whom the
benefits outweigh the risks, such as those diagnosed with ER-positive breast cancer,
or women at high risk of developing breast cancer.
The drug is extensively metabolized, with some metabolites possessing equal or
greater antiestrogenic activity compared to the parent drug. The hepatic cytochrome
P450 (CYP) family is largely responsible for tamoxifen biotransformation; the
major enzymes involved include CYP2D6 and CYP3A4/5, with lesser contribu-
tions from CYP2C9 and CYP2C19 (Fig. 6.1) [1014]. Removal of a methyl group
by CYP3A4/5 creates the major metabolite N-desmethyltamoxifen (NDT),
whereas hydroxylation by CYP2D6 results in formation of 4-hydroxytamoxifen
(4-OH-TAM). Sequential metabolism by both pathways creates the highly active
metabolite 4-hydroxy-N-desmethyltamoxifen, commonly known as endoxifen.
Conversion to NDT accounts for roughly 90% of a tamoxifen dose and accumulates
to higher concentrations than the parent drug, while 4-OH-TAM and endoxifen are
present in much lower quantities, less than 10% of total tamoxifen metabolism [14,
15]. NDT displays approximately the same activity as the parent drug; in contrast,
both 4-OH-TAM and endoxifen have much greater affinity for the ER and are 30- to
100-fold more potent inhibitors [1618]. In most individuals, endoxifen concentra-
tions are six- to tenfold greater than 4-OH-TAM levels, thus it is now considered
the primary active metabolite of tamoxifen [13, 19, 20].
The existence of several active metabolites and potential for metabolic variabil-
ity suggest that tamoxifen would be an excellent candidate drug for therapeutic
drug monitoring (TDM). However, due in part to the fact that few laboratories are
currently capable of measuring tamoxifen, NDT, 4-OH-TAM, and endoxifen, TDM
applications remain limited and therapeutic ranges are not well established [21, 22].
Given the multi-year administration protocols found to be optimal for tamoxifen
6 Pharmacogenomics of Tamoxifen 79
Fig. 6.1 Metabolism of tamoxifen to endoxifen
therapy, it is expected that expansion of TDM to routine practice would permit

more accurate assessment of long-term compliance and might improve patient
response.
For patients with poor response or contraindications to tamoxifen therapy, alter-
native hormonal therapy does exist. Raloxifene, for example, is another SERM that
has been shown to be effective, or better than tamoxifen for use in breast cancer
prophylaxis and prevention of osteoporosis [7]. For breast cancer therapy, the use
of aromatase inhibitors is recommended for patients who cannot tolerate tamoxifen,
or who have already received tamoxifen for 5 years. Aromatase inhibitors include
anastrozole, exemestane, and letrozole, and function by disrupting endogenous
synthesis of estrogen [4]. These agents are generally associated with lower inci-
dence of adverse responses, with therapeutic efficacy that is comparable to or, in
some settings, better than that of tamoxifen.
Tamoxifen is likely to remain a mainstay of hormonal therapy for breast cancer,
but there remains much opportunity for optimization of its use. The metabolic path-
ways responsible for conversion of the drug to its active metabolites are highly
variable, and it is becoming apparent that differences in an individuals ability to
form endoxifen contribute significantly to therapeutic response and patient outcome.
Tamoxifen pharmacogenetics is therefore an area of intense study, with the goals of

using germline polymorphisms in genes encoding metabolic enzymes to optimize
dosing and to select individuals most likely to respond to tamoxifen therapy.
6.3Tamoxifen Pharmacogenetics
6.3.1CYP2D6 Polymorphisms
The enzymatic activity of CYP2D6 varies greatly between individuals, due in part
to the high frequency of polymorphisms in the CYP2D6 gene. Over 100 CYP2D6
variant alleles have been described (www.cypalleles.ki.se, accessed October 20,
2009). The resulting enzymatic activity allows individuals to be categorized into
poor (PM), intermediate (IM), extensive (EM), and ultrarapid (UM) metabolizers.
There are significant ethnic differences in the frequency of CYP2D6 variants; one
of the most important PM alleles, CYP2D6*4, is present in 1221% of individuals
of Northern European descent, but is found in only 12% of Asians and Black
Africans [23]. Variants can be quite common, as seen with CYP2D6*10, which
confers an IM phenotype and is present in 57% of Han Chinese [24, 25]. The major
CYP2D6 alleles include fully functional CYP2D6*1; null alleles with essentially
no residual activity (*3-*8, *11-*16, *18-*20, *38, *40, *42, *44); reduced-function
alleles (*9, *10, *17, *29, *36, *37, *41); and amplified alleles comprised of
multiple copies of the gene (*1XN, *2XN, *35XN, and *41XN).
The influence of CYP2D6 genotype in tamoxifen metabolism has been shown in
a number of studies. During tamoxifen treatment, women with two or more fully
functional copies of CYP2D6 have higher plasma endoxifen concentrations than
patients with at least one null allele (*3-*6), or those taking known CYP2D6 inhibi-
tors [26], which suggests that CYP2D6 function is essential for optimal conversion
of tamoxifen to its highly active metabolites. Analysis of CYP2D6 genotype and
endoxifen plasma concentration in 158 patients from multiple ethnicities demon-
strated that patients with IM genotypes, e.g., *10 (reduced activity) or *4 (null)
heterozygotes, had endoxifen concentrations similar to PM. Similarly, a Chinese
study demonstrated that patients homozygous for CYP2D6*10 had lower serum
concentrations of 4-hydroxytamoxifen [27]. All these findings support the utility of
genotyping CYP2D6 to predict formation of highly active tamoxifen metabolites in
patients considering or undergoing therapy with tamoxifen.
The influence of CYP2D6 variant alleles has also been documented in terms of
treatment outcome. A retrospective study of 223 postmenopausal women examined
*4 (the most common null allele associated with PM status) and *6 (a low-frequency
PM variant) [28]. Women homozygous for CYP2D6*4 had poorer outcomes than
women with *4/*1 or *1/*1 genotypes, showing shorter time to relapse and worse
disease-free survival. In addition, despite 20% incidence of moderate to severe hot
flashes in women with zero or one *4 allele, no CYP2D6*4/*4 patients experienced
this side effect. Borges etal. confirmed these results and expanded the analysis to
include 33 different CYP2D6 variants [29]. Patients with reduced-activity CYP2D6
alleles (*4, *5, *10, *41) had significantly poorer outcome than carriers of func-
tional alleles, as documented by higher recurrence rates, shorter times to relapse,
and worse event-free survival. Studies looking at the association of CYP2D6*10
with clinical outcomes in Asian patients showed that patients with this IM variant
have a shorter recurrence-free survival period [27, 30].
Recently, the findings of a multicenter study that included 1,325 women treated
with tamoxifen for early stage breast cancer were published [31]. The study
included 609 women with EM, 637 women with IM, and 79 women with PM
CYP2D6 genotypes. The recurrence rates were 15% for EM, 21% for EM/IM
heterozygotes, and 29.0% for PM. This is the largest published study so far that
provides sufficiently powered evidence for an association between CYP2D6 genetics
and clinical outcome of tamoxifen. This data indicates that individuals with
CYP2D6 variants conferring PM status have a substantially higher risk of tamox-
ifen treatment failure, and in these patients alternative forms of adjuvant endocrine
therapy should be considered.
6.3.2CYP2D6 Inhibitors
Hot flashes, a common side effect of tamoxifen treatment, are often treated with
antidepressants such as the selective serotonin reuptake inhibitors (SSRI) [2].
Coadministration of the SSRIs paroxetine or fluoxetine, both of which potently
inhibit CYP2D6 activity, affects metabolism in tamoxifen-treated patients: CYP2D6
EM individuals show a significant reduction of endoxifen levels when these SSRIs
are added to tamoxifen therapy [26, 29]. In contrast, other SSRIs that are only weak
inhibitors of CYP2D6, such as venlafaxine, did not significantly affect endoxifen
levels, and thus may be preferable for the treatment of hot flashes in breast cancer
patients [26, 32]. However, coadministration of any CYP2D6 inhibitors, be they
strong (e.g., fluoxetine, paroxetine, bupropion, quinidine) or weak (e.g., sertraline,
duloxetine, cimetidine, terbinafine, amiodarone), has the potential to lower endoxifen
plasma concentrations [26, 29, 33], and may render tamoxifen less effective
[34,35]. In fact, strong CYP2D6 inhibitors have been shown to reduce endoxifen
concentrations in CYP2D6 EM to drug levels comparable to those seen in CYP2D6
PM [26]. This is referred to as a phenocopy, i.e., a phenotype induced by environ-
mental factors (in this case, enzymatic activity reduced by comedications), which
mimics a different genotype.
It has been shown that the phenocopying due to the coprescription of CYP2D6
inhibitors was an independent predictor of breast cancer outcome in postmeno-
pausal women taking tamoxifen [36] and of reduction in endoxifen plasma con-
centrations [37]. A recent study assessed the combined effect of genetic variation
and drug-induced inhibition of CYP2D6 on breast cancer outcomes [35]. In this
analysis, patients were segregated according to whether potent or weak/moderate,
CYP2D6 inhibitors were coprescribed with tamoxifen. Based on the genotype

and medication history, patients were classified as having an extensive (normal)
or decreased CYP2D6 metabolism. Patients with decreased metabolism had sig-
nificantly shorter time to recurrence and worse relapse-free survival than patients
with extensive metabolism. CYP2D6 genotype and concomitant potent CYP2D6
inhibitors are highly associated with endoxifen plasma concentration and may
have an impact on the response to tamoxifen therapy [29]. All together, these
studies indicate that coadministration of potent CYP2D6 inhibitors should be
avoided in patients taking tamoxifen as they might jeopardize the success of the
treatment.
Although SSRIs are frequently coadministered, many other drugs have been
reported to inhibit the CYP2D6 enzyme system (http://medicine.iupui.edu/
clinpharm/ddis/table.asp). The vast majority of theses drugs are prescription
drugs, but at least two CYP2D6 inhibitors are also know for their potential as
drugs of abuse, namely methadone and cocaine. Indeed, cocaine appears to
have a lower inhibition constant [38] than do paroxetine and fluoxetine [39, 40],
suggesting it is a more potent inhibitor of CYP2D6. It has also been suggested
that 3,4-methylenedioxymethamphetamine (MDMA, Ecstasy) has a similar
inhibition constant to fluoxetine [41].
The use of drugs of abuse, prescribed and nonprescribed medications, herbal
supplements, and some foods can inhibit or induce enzymatic activity manyfold, to
the extent that a PM phenocopy may be induced, regardless of genotype. The
complexity of tamoxifen metabolism precludes the use of a probe drug for pheno-
typing studies, thus measurement of the parent drug and its highly active metabo-
lites may prove a critical means of addressing both the interindividual variability
and the efficacy of tamoxifen therapy in breast cancer patients.
6.3.3Role of Other Polymorphisms in Tamoxifen

Treatment Outcome
Variants in genes encoding other metabolic enzymes have also been studied for
associations to tamoxifen metabolism and outcome, including CYP3A, CYP2C9,
CYP2C19, and SULT1A1. Conversion of tamoxifen to NDT is primarily mediated
by CYP3A4/5 [14], but the significance of polymorphisms in the genes encoding
these isoforms remains unclear. The null allele variant CYP3A5*3 was not associ-
ated with any statistically significant differences in plasma concentrations of
tamoxifen or its metabolites [26, 42]. Similarly, no differences in time to relapse,
disease-free survival or overall survival have been observed in CYP3A5*3
individuals, suggesting minimal influence on patient outcome [28]. However,
Wegman etal. reported that breast cancer patients with a CYP3A5*3/*3 genotype
show improved recurrence-free survival, an unexpected result given that this null
genotype should reduce formation of NDT, the precursor of endoxifen [43].
CYP2C9 and CYP2C19 are also capable of catalyzing formation of NDT,

though to a lesser extent [12, 14, 15]. An investigation of CYP2C19 variants *2, *3
(both null alleles), and *17 (which confers UM status) examined the association of
this gene with tamoxifen treatment outcome. Compared to the reference CYP2C19*1
allele, patients with the UM *17 allele had lower risk of relapse and prolonged time
to relapse, whereas the two null variants *2 and *3 were not associated with differ-
ences in treatment outcome [44]. Similarly, the presence of low-activity CYP2C9
variants (*2 and *3) affected neither patient outcome nor tamoxifen metabolite
concentrations [26, 44].
The Phase II enzyme SULT1A1 causes formation of sulfated tamoxifen
metabolites [45]. The role of this enzyme in enhancing drug clearance would
suggest that reduced SULT1A1 activity would slow clearance of active tamox-
ifen metabolites, and thus improve treatment efficacy [46]. Despite this, no sig-
nificant differences in plasma concentrations of tamoxifen and its metabolites
were found in patients with the low-activity SULT1A1*2 allele [26]; indeed,
some reports indicate that tamoxifen is less effective in patients with the *2 vari-
ant, with increased risk of cancer recurrence and death [4648]. Despite this, a
study of 677 tamoxifen-treated postmenopausal women found no association
between SULT1A1 genotype and either improved or worsened treatment out-
come [43, 48].
6.4CYP2D6 Genotyping and Tamoxifen: Clinical Practice

and Case Studies
Current evidence strongly suggests that knowledge of the CYP2D6 genotype may
be beneficial when selecting a breast cancer treatment. An FDA advisory panel
has suggested that the tamoxifen package insert should alert physicians of the
following concerns: first, that CYP2D6 PM patients are at increased risk for
recurrence of their breast cancers if treated with tamoxifen, and second, that
coadministration of certain SSRIs known to inhibit CYP2D6 can affect the
metabolism of tamoxifen [49].
Knowledge of patient genotype can guide selection of alternate therapies: for
example, in postmenopausal PM individuals diagnosed with breast cancer, aro-
matase inhibitors are a reasonable alternative to tamoxifen treatment. Randomized
clinical trials in postmenopausal breast cancer patients have demonstrated superior
efficacy and better overall safety for aromatase inhibitors as compared with
tamoxifen [50]. For breast cancer prevention, raloxifene appears to be a viable
alternative for CYP2D6 PM; raloxifene and tamoxifen have been shown to be
equally effective in reducing breast cancer incidence in high-risk postmenopausal
women [51, 52].
The following case studies highlight representative examples of the clinical util-
ity of genotyping CYP2D6 in patients receiving or considering tamoxifen therapy.
6.4.1Case 1
6.4.1.1Presentation
A 69-year-old female presented after palpating a mass within her right breast,
which was biopsied and diagnosed as infiltrating ductal carcinoma. She underwent
right wide local excision. Findings confirmed an infiltrating ductal carcinoma,
Nottingham grade III, forming a 1.8-cm mass in greatest dimension. Angiolymphatic
invasion was negative, and all margins were negative. A single right axillary sentinel
lymph node was negative for metastatic disease. The tumor cells were strongly
positive for ER (greater than 75%) and PR (5175%), and were HER-2 negative.
6.4.1.2Therapy
Following surgery, the patient received adjuvant radiation but declined adjuvant
chemotherapy. Five years treatment with anastrozole was recommended; however,
due to financial constraints, the patient requested tamoxifen. CYP2D6 testing
determined the patients genotype to be CYP2D6 *1/*1, and she was placed on
tamoxifen, 20 mg daily. After 3 months, she returned with significant vasomotor
symptoms (hot flashes) leading to insomnia and decreased quality of life. She
requested to go off tamoxifen. She was counseled against discontinuation and
instead was prescribed Venlafaxine XR, 75 mg once daily, which improved her hot
flashes substantially.
6.4.1.3Discussion
Prospective genotyping of this patient confirmed her EM status, indicating that she
is likely to be able to convert tamoxifen to its highly active metabolites. Activation
of tamoxifen is suggested by the patients presentation with hot flashes: this symptom
is less common in patients with PM alleles [35] and may be associated with clinical
outcome [53], although it is not recommended for use as a predictor of tamoxifen
response [54]. Venlafaxine has been shown to be effective in relief of tamoxifen-
induced hot flashes [55].
6.4.2Case 2
6.4.2.1Presentation
A 48-year-old, premenopausal female underwent a left skin-sparing mastectomy

for a 1.3-cm Nottingham grade II/III invasive ductal carcinoma. Angiolymphatic
invasion was negative, and all margins were negative. One sentinel lymph node was
found to be involved with metastatic carcinoma although an additional 27 axillary
nodes were dissected and shown to be negative. Estrogen and progesterone recep-
tors were strongly positive (greater than 75% nuclear staining), and HER-2 was
negative.
6.4.2.2Therapy
After surgery, the patient received doxorubicin and cyclophosphamide followed

by paclitaxel using the dose-dense schedule. Following the completion of che-
motherapy, the patient was rendered amenorrheic. A recommendation was given
for adjuvant tamoxifen, 20 mg daily, for 5 years. The patient experienced mini-
mal to no hot flashes while on tamoxifen. After 1.5 years on tamoxifen, she
requested a CYP2D6 genotype test. The results demonstrated CYP2D6 *4/*4,
consistent with a CYP2D6 poor metabolizer. She was then switched to anastro-
zole, and shortly thereafter developed moderate vasomotor symptoms and arth-
ralgias not requiring pharmacotherapy. She completed a total of 5 years of
adjuvant anastrozole.
6.4.2.3Discussion
This patients status as a PM was unknown at the time of tamoxifen initiation. In

the absence of CYP2D6 genotype testing, it is likely that she would have completed
the 5-year regimen as planned, thus increasing her likelihood of recurrence. The
availability of genotypic information permitted selection of a more appropriate
adjuvant endocrine therapy.
6.4.3Case 3
6.4.3.1Presentation
A 57-year-old female presented with new asymmetry in the right breast on a rou-
tine mammogram. A needle biopsy demonstrated invasive lobular carcinoma, for
which she underwent wide local excision and sentinel lymph node biopsy. The
sentinel node biopsy was positive, and she underwent a complete right axillary
lymph node dissection. Pathology demonstrated a 4.0 cm invasive lobular carci-
noma in greatest dimension, Nottingham grade I, with 8 of 13 lymph nodes posi-
tive for metastatic carcinoma. Angiolymphatic invasion was negative, and all
margins were negative. The tumor cells were ER and PR positive (5175% nuclear
staining) and HER-2 negative.
6.4.3.2Therapy
The patient received postoperative doxorubicin and cyclophosphamide followed by

paclitaxel using the dose-dense schedule, followed by adjuvant chest wall radiation.
A recommendation was given for adjuvant anastrozole; however, the patient refused
because of concerns regarding osteoporosis. A bone mineral density demonstrated
osteopenia, with a total hip T-score = 2.1. The patients CYP2D6 genotype was
obtained and determined to be CYP2D6*1/*2A. She began tamoxifen therapy, but
after 2.5 years could not tolerate the drug and was switched to anastrozole. After
6 months of anastrozole, the patient had developed disabling arthralgias, carpal
tunnel syndrome, and hair thinning; she thus discontinued anastrozole and went
back on tamoxifen. After 5 total years of adjuvant hormonal therapy, a recommen-
dation was given to switch to letrozole.
6.4.3.3Discussion
A heterozygous *1/*2A genotype suggests that this patients metabolic phenotype

would be in the range of extensive to ultra-rapid. Such individuals would have
successful conversion of tamoxifen to active metabolites, but may in fact be predis-
posed to adverse responses due to the increased efficiency of drug activation.
However, the use of tamoxifen is not contraindicated in such patients, so long as the
side effects are tolerable or can be managed with SSRIs or other therapeutics.
References
1. Colleoni M., Gelber S., Goldhirsch A., Aebi S., Castiglione-Gertsch M., Price KN., et al.
Tamoxifen after adjuvant chemotherapy for premenopausal women with lymph node-positive
breast cancer: International Breast Cancer Study Group Trial 13-93. Journal of Clinical
Oncology Mar 20 2006;24(9):13321341.
2. Fisher B., Costantino JP., Wickerham DL., Redmond CK., Kavanah M., Cronin WM., et al.
Tamoxifen for prevention of breast cancer: report of the National Surgical Adjuvant Breast
and Bowel Project P-1 Study. Journal of the National Cancer Institute Sep 16
1998;90(18):13711388.
3. Effects of chemotherapy and hormonal therapy for early breast cancer on recurrence and
15-year survival: an overview of the randomised trials. Lancet May 1420 2005;
365(9472):16871717.
4. Loose DS., Stancel GM. Estrogens and Progestins. In: Brunton LL, Lazo JS, Parker KL, eds.
Goodman & Gilmans The Pharmacological Basis of Therapeutics, 11th ed. New York:
McGraw-Hill; 2006:15541558.
5. Tamoxifen Citrate, in the Physicians Desk Reference online. Accessed 1 December, 2009.
http://www.thomsonhc.com/clinicalxpert/librarian.
6. Osborne CK. Tamoxifen in the treatment of breast cancer. The New England Journal of
Medicine Nov 26 1998;339(22):16091618.
7. Lee WL., Cheng MH., Chao HT., Wang PH. The role of selective estrogen receptor modulators on
breast cancer: from tamoxifen to raloxifene. Taiwanese Journal of Obstetrics & Gynecology Mar
2008;47(1):2431.
8. White IN. Tamoxifen: is it safe? Comparison of activation and detoxication mechanisms in

rodents and in humans. Current Drug Metabolism Jun 2003;4(3):223239.
9. Bird BR., Swain SM. Cardiac toxicity in breast cancer survivors: review of potential cardiac
problems. Clinical Cancer Research Jan 1 2008;14(1):1424.
10. Jordan VC., Collins MM., Rowsby L., Prestwich G. A monohydroxylated metabolite of
tamoxifen with potent antioestrogenic activity. Journal of Endocrinology Nov 1977;
75(2):305316.
11. Lien EA., Solheim E., Lea OA., Lundgren S., Kvinnsland S., Ueland PM. Distribution of
4-hydroxy-N-desmethyltamoxifen and other tamoxifen metabolites in human biological fluids
during tamoxifen treatment. Cancer Research Apr 15 1989;49(8):21752183.
12. Crewe HK., Notley LM., Wunsch RM., Lennard MS., Gillam EM. Metabolism of tamoxifen
by recombinant human cytochrome P450 enzymes: formation of the 4-hydroxy, 4-hydroxy
and N-desmethyl metabolites and isomerization of trans-4-hydroxytamoxifen. Drug
Metabolism and Disposition Aug 2002;30(8):869874.
13. Stearns V., Beebe KL., Iyengar M., Dube E. Paroxetine controlled release in the treatment of
menopausal hot flashes: a randomized controlled trial. Journal of the American Medical
Association Jun 4 2003;289(21):28272834.
14. Desta Z., Ward BA., Soukhova NV., Flockhart DA. Comprehensive evaluation of tamoxifen
sequential biotransformation by the human cytochrome P450 system in vitro: prominent roles
for CYP3A and CYP2D6. The Journal of Pharmacology and Experimental Therapeutics Sep
2004;310(3):10621075.
15. Coller JK., Krebsfaenger N., Klein K., Endrizzi K., Wolbold R., Lang T., et al. The influence
of CYP2B6, CYP2C9 and CYP2D6 genotypes on the formation of the potent antioestrogen
Z-4-hydroxy-tamoxifen in human liver. British Journal of Clinical Pharmacology Aug
2002;54(2):157167.
16. Borgna JL., Rochefort H. Hydroxylated metabolites of tamoxifen are formed in vivo and
bound to estrogen receptor in target tissues. The Journal of Biological Chemistry Jan 25
1981;256(2):859868.
17. Robertson DW., Katzenellenbogen JA., Long DJ., Rorke EA., Katzenellenbogen BS.
Tamoxifen antiestrogens. A comparison of the activity, pharmacokinetics, and metabolic acti-
vation of the cis and trans isomers of tamoxifen. Journal of Steroid Biochemistry Jan
1982;16(1):113.
18. Coezy E., Borgna JL., Rochefort H. Tamoxifen and metabolites in MCF7 cells: correlation
between binding to estrogen receptor and inhibition of cell growth. Cancer Research Jan
1982;42(1):317323.
19. Johnson MD., Zuo H., Lee KH., Trebley JP., Rae JM., Weatherman RV., et al. Pharmacological
characterization of 4-hydroxy-N-desmethyl tamoxifen, a novel active metabolite of tamoxifen.
Breast Cancer Research and Treatment May 2004;85(2):151159.
20. Lim YC., Desta Z., Flockhart DA., Skaar TC. Endoxifen (4-hydroxy-N-desmethyl-tamoxifen)
has anti-estrogenic effects in breast cancer cells with potency similar to 4-hydroxy-tamoxifen.
Cancer Chemotherapy and Pharmacology May 2005;55(5):471478.
21. Furlanut M., Franceschi L., Pasqual E., Bacchetti S, Poz D., Giorda G., et al. Tamoxifen and
its main metabolites serum and tissue concentrations in breast cancer women. Therapeutic
Drug Monitoring Jun 2007;29(3):349352.
22. Kisanga ER., Gjerde J., Guerrieri-Gonzaga A., Pigatto F., Pesci-Feltri A., Robertson C., etal.
Tamoxifen and metabolite concentrations in serum and breast cancer tissue during three
dose regimens in a randomized preoperative trial. Clinical Cancer Research Apr 1 2004;
10(7):23362343.
23. Ingelman-Sundberg M., Sim SC., Gomez A., Rodriguez-Antona C. Influence of cytochrome
P450 polymorphisms on drug therapies: pharmacogenetic, pharmacoepigenetic and clinical
aspects. Pharmacology and Therapeutics Dec 2007;116(3):496526.
24. Bradford LD. CYP2D6 allele frequency in European Caucasians, Asians, Africans and their
descendants. Pharmacogenomics Mar 2002;3(2):229243.
25. Garcia-Barcelo M, Chow LY., Chiu HF., Wing YK., Lee DT., Lam KL., et al. Genetic analysis
of the CYP2D6 locus in a Hong Kong Chinese population. Clinical Chemistry Jan 2000;
46(1):1823.
26. Jin Y., Desta Z., Stearns V., Ward B., Ho H., Lee KH., et al. CYP2D6 genotype, antidepressant
use, and tamoxifen metabolism during adjuvant breast cancer treatment. Journal of the
National Cancer Institute Jan 5 2005;97(1):3039.
27. Xu Y., Sun Y., Yao L., Shi L., Wu Y., Ouyang T., et al. Association between CYP2D6 *10
genotype and survival of breast cancer patients receiving tamoxifen treatment. Annals of
Oncology Aug 2008;19(8):14231429.
28. Goetz MP., Rae JM., Suman VJ., Safgren SL., Ames MM., Visscher DW., et al. Pharmacogenetics
of tamoxifen biotransformation is associated with clinical outcomes of efficacy and hot flashes.
Journal of Clinical Oncology Dec 20 2005;23(36):93129318.
29. Borges S., Desta Z., Li L., Skaar TC., Ward BA., Nguyen A., et al. Quantitative effect of
CYP2D6 genotype and inhibitors on tamoxifen metabolism: implication for optimization of
breast cancer treatment. Clinical Pharmacology and Therapeutics Jul 2006;80(1):6174.
30. Kiyotani K., Mushiroda T., Sasa M., Bando Y., Sumitomo I., Hosono N., et al. Impact of
CYP2D6*10 on recurrence-free survival in breast cancer patients receiving adjuvant tamox-
ifen therapy. Cancer Science May 2008;99(5):995999.
31. Schroth W., Goetz MP., Hamann U., Fasching PA., Schmidt M., Winter S., et al. Association
between CYP2D6 polymorphisms and outcomes among women with early stage breast cancer
treated with tamoxifen. The Journal of the American Medical Association Oct 7
2009;302(13):14291436.
32. Loibl S., Schwedler K., von Minckwitz G., Strohmeier R., Mehta KM., Kaufmann M.
Venlafaxine is superior to clonidine as treatment of hot flashes in breast cancer patientsa
double-blind, randomized study. Annals of Oncology Apr 2007;18(4):689693.
33. Lien EA., Solheim E., Kvinnsland S., Ueland PM. Identification of 4-hydroxy-N-desmethyl-
tamoxifen as a metabolite of tamoxifen in human bile. Cancer Research Apr 15 1988;48(8):
23042308.
34. Bijl MJ., van Schaik RH., Lammers LA., Hofman A., Vulto AG., van Gelder T., et al. The
CYP2D6*4 polymorphism affects breast cancer survival in tamoxifen users. Breast Cancer
Research and Treatment. Nov 2009;118(1):125130.
35. Goetz MP., Knox SK., Suman VJ., Rae JM., Safgren SL., Ames MM., et al. The impact of
cytochrome P450 2D6 metabolism in women receiving adjuvant tamoxifen. Breast Cancer
Research and Treatment Jan 2007;101(1):113121.
36. Goetz MP., Kamal A., Ames MM. Tamoxifen pharmacogenomics: the role of CYP2D6 as a
predictor of drug response. Clinical Pharmacology and Therapeutics. Jan 2008;83(1):160166.
37. Brauch H., Murdter TE., Eichelbaum M., Schwab M. Pharmacogenomics of tamoxifen therapy.
Clinical Chemistry Oct 2009;55(10):17701782.
38. Tyndale RF., Sunahara R., Inaba T., Kalow W., Gonzalez FJ., Niznik HB. Neuronal cyto-
chrome P450IID1 (debrisoquine/sparteine-type): potent inhibition of activity by (-)-cocaine
and nucleotide sequence identity to human hepatic P450 gene CYP2D6. Molecular
Pharmacology Jul 1991;40(1):6368.
39. Otton SV., Ball SE., Cheung SW., Inaba T., Rudolph RL., Sellers EM. Venlafaxine oxidation
in vitro is catalysed by CYP2D6. British Journal of Clinical Pharmacology Feb 1996;
41(2):149156.
40. Owen JR., Nemeroff CB. New antidepressants and the cytochrome P450 system: focus on
venlafaxine, nefazodone, and mirtazapine. Depression and Anxiety. 1998;7 (Supp l)
1:2432.
41. Ramamoorthy Y., Tyndale RF., Sellers EM. Cytochrome P450 2D6.1 and cytochrome P450
2D6.10 differ in catalytic activity for multiple substrates. Pharmacogenetics Aug
2001;11(6):477487.
42. Tucker AN., Tkaczuk KA., Lewis LM., Tomic D., Lim CK., Flaws JA. Polymorphisms in
cytochrome P4503A5 (CYP3A5) may be associated with race and tumor characteristics, but
not metabolism and side effects of tamoxifen in breast cancer patients. Cancer Letters Jan 10
2005;217(1):6172.
43. Wegman P., Elingarami S., Carstensen J., Stal O., Nordenskjold B., Wingren S. Genetic vari-
ants of CYP3A5, CYP2D6, SULT1A1, UGT2B15 and tamoxifen response in postmenopausal
patients with breast cancer. Breast Cancer Research 2007;9(1):R7.
44. Schroth W., Antoniadou L., Fritz P., Schwab M., Muerdter T., Zanger UM., et al. Breast cancer
treatment outcome with adjuvant tamoxifen relative to patient CYP2D6 and CYP2C19 geno-
types. Journal of Clinical Oncology Nov 20 2007;25(33):51875193.
45. Chen G., Yin S., Maiti S., Shao X. 4-Hydroxytamoxifen sulfation metabolism. Journal of
Biochemical and Molecular Toxicology 2002;16(6):279285.
46. Nowell S., Sweeney C., Winters M., Stone A., Lang NP., Hutchins LF., et al. Association
between sulfotransferase 1A1 genotype and survival of breast cancer patients receiving
tamoxifen therapy. Journal of the National Cancer Institute Nov 6 2002;94(21):16351640.
47. Nowell SA., Ahn J., Rae JM., Scheys JO., Trovato A., Sweeney C., et al. Association of
genetic variation in tamoxifen-metabolizing enzymes with overall survival and recurrence of
disease in breast cancer patients. Breast Cancer Research and Treatment Jun
2005;91(3):249258.
48. Gjerde J., Hauglid M., Breilid H., Lundgren S., Varhaug JE., Kisanga ER., et al. Effects of
CYP2D6 and SULT1A1 genotypes including SULT1A1 gene copy number on tamoxifen
metabolism. Annals of Oncology Jan 2008;19(1):5661.
49. Young D. Genetics examined in tamoxifens effectiveness: recurrence warning urged for
labeling. American Journal of Health-System Pharmacy Dec 1 2006;63(23):2286, 2296.
50. Howell A., Cuzick J., Baum M., Buzdar A., Dowsett M., Forbes JF., et al. Results of the ATAC
(Arimidex, Tamoxifen, Alone or in Combination) trial after completion of 5 years adjuvant
treatment for breast cancer. Lancet. Jan 1-7 2005;365(9453):6062.
51. Bevers TB. Raloxifene and the prevention of breast cancer. Expert Opinion on Pharmacotherapy
Nov 2006;7(16):23012307.
52. Vogel VG., Costantino JP., Wickerham DL., Cronin WM., Cecchini RS., Atkins JN., et al.
Effects of tamoxifen vs raloxifene on the risk of developing invasive breast cancer and other
disease outcomes: the NSABP Study of Tamoxifen and Raloxifene (STAR) P-2 trial. The
Journal of the American Medical Association Jun 21 2006;295(23):27272741.
53. Mortimer JE., Flatt SW., Parker BA., Gold EB., Wasserman L., Natarajan L., et al. Tamoxifen,
hot flashes and recurrence in breast cancer. Breast Cancer Research and Treatment Apr 2008;
108(3):421426.
54. Lynn Henry N., Rae JM., Li L., Azzouz F., Skaar TC., Desta Z., et al. Association between
CYP2D6 genotype and tamoxifen-induced hot flashes in a prospective cohort. Breast Cancer
Research and Treatment Oct 2009;117(3):571575.
55. Loprinzi CL., Barton DL., Sloan JA., Novotny PJ., Dakhil SR., Verdirame JD., et al. Mayo
Clinic and North Central Cancer Treatment Group hot flash studies: a 20-year experience.
Menopause Jul-Aug 2008;15(4 Pt 1):655660.
Chapter 7
Thiopurines
Terreia S. Jones and Mary V. Relling
Keywords Thiopurine methyltransferase Leukemia Inflammatory bowel

disease Azathioprine Mercaptopurine Thioguanine
7.1Introduction
Thiopurines are antimetabolite prodrugs used clinically as antineoplastics and as

immunosuppressants. The first clinically effective thiopurines, thioguanine and mer-
captopurine, were developed in the 1950s by Wellcome Research Laboratory scientists
Gertrude Elion and George Hitchings [1]. After the approval of mercaptopurine in
1953 by FDA for the treatment of leukemia, it was discovered that when it was com-
bined with other anticancer agents, childhood leukemia could be cured. The develop-
ment of these compounds contributed to Elion and Hitchings being awarded the Nobel
Prize in medicine in 1988. Today, there are three thiopurine drugs (mercaptopurine,
thioguanine, and azathioprine) (Fig. 7.1) that are widely used in the treatment of
leukemia, rheumatic diseases, inflammatory bowel disease, and solid organ transplant.
7.2Pharmacology of Thiopurine Drugs
7.2.1Clinical Indications and Target Patient Populations
For more than 50 years, mercaptopurine has been used as part antileukemic
maintenance therapy in the treatment of childhood acute lymphoblastic leukemia
(ALL) and has contributed to the high cure rates achieved (>80%) [2]. Thioguanine
M.V. Relling(*)
St. Jude Childrens Research Hospital, Memphis, TN, USA
and
University of Tennessee Health Science Center, Memphis, TN, USA
e-mail: mary.relling@stjude.org
92 T.S. Jones and M.V. Relling
Fig. 7.1 Chemical structures for the clinically used thiopurine drugs
is indicated in the treatment of myelogenous leukemia, while azathioprine is

indicated as adjunct therapy in solid organ transplant and in rheumatoid arthritis
(Table 7.1). Although the labeled indications for these agents are limited, there are
many unlabeled investigational uses that have proven to be beneficial.
Both azathioprine and mercaptopurine have shown great promise in the treat-
ment of inflammatory bowel disease (i.e., Crohns disease, ulcerative colitis) by
lowering steroid requirements and prolonging remission [3]. In Crohns disease,
azathioprine has been shown to prevent relapse as well as decrease the dependence
on and resistance to steroids [4, 5]. These agents have also proven to be beneficial
in ulcerative colitis therapy through effectively inducing remission in as many as
70% of patients [6]. In autoimmune hepatitis, when azathioprine is combined with
corticosteroids, the 20-year life expectancy increases to 80% and the incidence of
hepatic fibrosis can decrease by as much as 79% as compared to corticosteroids
alone [7]. Azathioprine has also been used to treat inflammatory eye conditions to
include uveitis and dysthyroid orbitopathy [8].
7.2.2Pharmacodynamics/Pharmacokinetics
Thiopurines are purine analogs (Fig. 7.1) requiring intracellular activation to exert
their cytotoxic effects. Mercaptopurine (a hypoxantine analog), azathioprine
(aprodrug of mercaptopurine), and thioguanine (a guanine analog) (Fig. 7.1) are all
subject to activation by hypoxanthine phosphoribosyl transferase (HPRT) and other
enzymes to form the cytotoxic thioguanine nucleotide (TGN) metabolites or
inactivation by thiopurine methyltransferase (TPMT) (Fig. 7.2).
TGN metabolite incorporation into DNA constitutes the primary mechanism of
thiopurine cytotoxicity. Thioguanine is more directly converted to TGNs bypassing
many of the enzymatic steps that are required for mercaptopurine activation
(Fig. 7.2). As the predominant inactivation pathway, TPMT catalyzes the
S-methylation of thiopurine drugs. Unlike thioguanine, mercaptopurine (and
azathioprine) can also undergo inactivation by xanthine oxidase.
Table 7.1 Labeled indications for clinically used thiopurines
Monitoring
Drug (brand name) Clinical indications Dosagea Place in therapy Toxicities considerations
Mercaptopurine Acute lymphoblastic Induction: Remission induction; Bone marrow CBC with differential
(Purinethol)b leukemia 70100 mg/m2/day maintenance therapy suppression, and platelet
7 Thiopurines
(children); hepatotoxicity, count, serum

100200 mg/m2/day (adults) renal toxicity, transaminases,
Acute myelogenous Maintenance: gastrointestinal alkaline
leukemia 5075 mg/m2/day (children); toxicity, phosphatase,
80100 mg/m2/day (adults) hypersensitivity bilirubin,
Thioguanine Acute myelogenous Children <3 years: Remission induction; reactions urinanalysis,
(Tabloid)c leukemias; chronic 3.3 mg/kg/day consolidation; hemoglobin, and
myelogenous Children >3 years and Adults: maintenance therapy hematocrit
leukemias; granulocytic 23 mg/kg/day
leukemias
Azathioprine Kidney transplant; solid Initial: 35 mg/kg/day Adjunct therapy
(Imuran)d organ transplant Maintenance: to prevent transplant
(non-renal) 13 mg/kg/day rejection
Rheumatoid arthritis Initial: 1 mg/kg/day 68 Disease modifying
weeks, increased by antirheumatic agent
0.5 mg/kg every 4 weeks
until response or up to
2.5 mg/kg/day
Maintenance: reduce by
0.5 mg/kg every 4 weeks until
lowest effective dose
is reached
a
Actual dosages may vary 23 fold from label doses. Many uses of thiopurines are off-label
b
Labeling for Purinethol as of August 2003
c
Labeling for Tabloid as of June 2009
d
Labeling for Imuran as of May 2008
93
Fig. 7.2 Thiopurine pathway, reproduced with permission www.pharmgkb.org. Azathioprine is a

prodrug that is metabolized to mercaptopurine; mercaptopurine is susceptible to direct inactivation
via methylation by the polymorphic enzyme TPMT, leaving more substrate available for anabo-
lism to the active metabolites, TGNs (thioguanine nucleotides). The secondary mercaptopurine
metabolite, thioinosine monophospate, can also be methylated by TPMT, leading to the formation
of methylmercaptopurine nucleotides (MeTIMP), which have some antitumor and immunosup-
pressant properties. Thioguanine is also directly methylated by TPMT to methylthioguanine and
is more directly metabolized to TGN, with no secondary methylated active metabolite analogous
to MeTIMP
7 Thiopurines 95
TPMT can form another toxic metabolite, methylthioinosine monophosphate

(meTIMP) or methylmercaptopurine nucleotides (MeMPN), which may exert
its cytotoxic effects through inhibition of de novo purine synthesis [9, 10]. It is
unclearto what extent meTIMP metabolites contribute to the overall cytotoxicity of
mercaptopurine; active methyl metabolites do not exist for thioguanine (Fig. 7.2).
TPMT is subject to a few well-studied deactivating genetic polymorphisms that
areresponsible for significant interpatient variability in response to thiopurine therapy
[11]. When TPMT enzymatic activity is low, excessively high TGN levels can result
and lead to life threatening toxicities (i.e., myelosuppression, second cancers).
7.2.3Dosing, Toxicity, and Monitoring Considerations
Thiopurines have a narrow therapeutic index requiring careful dosage and monitoring
considerations regardless of the disease state being treated. Although general dosing
recommendations are available for approved indications (Table 7.1), thiopurines are
typically dosed based on the track record of their use in treatment protocols, espe-
cially in the treatment of neoplasias.
The most serious acute toxicities associated with thiopurine therapy are myelo-
suppression and hepatotoxicity (Table 7.1). When myelosuppression is severe, life-
threatening complications can occur. When conventional therapy is administered,
patients who inherit dysfunctional TPMT can have excessively high concentrations
of the active TGNs in blood cells. Hence, routine monitoring of blood cell counts
is important to both ensure that adequate immunosuppression is achieved, as well
as to monitor for excessively low blood counts that could warrant dose reductions
or temporary withholding of therapy. Potentially life-threatening complications
include infection, anemia, and bleeding complications.
Thiopurine associated hepatotoxicity can present in many forms such as intrahe-
patic cholestasis, focal centralobular necrosis characterized by hyperbilirubinemia,
increased alkaline phosphatase and liver aminotransferases (aspartate and alanine),
jaundice, ascites, and encephalopathy [12, 13]. Hepatotoxicity is most often seen
after 2 months of therapy but can also occur very early in therapy (within 1 week)
or may be delayed for several years post therapy. Chronic thioguanine administra-
tion has been linked to veno-occlusive disease of the liver, which appears to be
dose-related [14]. The meTIMP metabolite contributes to toxicity caused by
mercaptopurine and azathioprine [15]. Routine monitoring of liver aminotransferases,
uric acid, and bilirubin is recommended for early detection of liver toxicity.
A serious delayed complication associated with thiopurine therapy is the risk of
developing a secondary cancer [1619]. This poses a real challenge for clinicians
because it is impractical to manage or monitor a disease that has not yet developed.
Second cancers that have been associated with thiopurine therapy are brain, skin,
and myelogenous leukemia [16, 20, 21]. Because secondary cancers may be associ-
ated with high levels of thiopurine active metabolites, thiopurine testing may be
indicated to minimize these high exposures (see below).
It is well established that polymorphisms in TPMT account for a significant

degree of interpatient variability in response to thiopurine therapy. At St. Jude
Childrens Research Hospital, all newly diagnosed ALL patients are assessed for
TPMT status prior to initiating therapy. The goal of this approach is to achieve
comparable active metabolite levels and toxicity profiles for all patients, regardless
of TPMT genotype or phenotype. However, determining a patients TPMT status
prior to initiating thiopurine therapy has not been universally employed. In a study
that assessed azathioprine efficacy and toxicity in patients with rheumatoid arthritis,
it was concluded that TPMT genotyping prior to initiating azathioprine therapy
could potentially allow for higher doses of azathioprine and ultimately greater treat-
ment efficacy [22]. In Crohns disease, thiopurines are discontinued in as many as
1/3 of patients due to poor or no response and up to 1/5 of patients due to adverse
drug effects [23]. In this patient population, poor response to thiopurines has
beenassociated with low levels of the active TGN metabolites while good response
was associated with higher levels [23]. Additionally, screening for TPMT status has
been suggested to be beneficial in the prevention of thiopurine-associated toxicity
in patients treated for inflammatory eye conditions [8].
In addition to understanding the importance of knowing a patients TPMT status
to prevent toxicity, there are a couple of drug interactions that can result in signifi-
cant toxicities if coadministered with thiopurines at standard doses. First, allopurinol
is a potent inhibitor of xanthine oxidase and when coadministered with mercap-
topurine or azathioprine, excessively high levels of the active TGNs can result [24].
In this situation, a dose reduction of the thiopurine drug would be warranted.
Thioguanine is not a substrate for this pathway so is not affected by allopurinol
coadministration. Second, aminosalicylates (i.e., mesalazine, olsalazine, and sul-
fasalazine) are known inhibitors of TPMT, and hence patients who receive these
drugs concurrently with thiopurines should be monitored closely as they are at an
increased risk of severe myelosuppression.
7.2.4Markers of Treatment Efficacy and Toxicity
Determining the best dose of thiopurines to administer is extremely important when

treating diseases that are life-threatening or that could result in significant morbidity.
Measuring thiopurine metabolites, TPMT activity, and TPMT genotypes can be
helpful to minimize the risk of treatment failure or life-threatening toxicities when
doses are too low or too high, respectively (Table 7.2).
When administering drugs with narrow therapeutic indices, it is important that
informative therapeutic monitoring parameters are measured and utilized to help
guide therapy. In the ALL patient population, the degree of myelosuppression can
be used to tailor mercaptopurine therapy. Indeed, a modest level of myelosuppres-
sion is expected, and in fact desired, as it gives an indication that the target cell
population is receiving adequate drug exposure.
Table 7.2 Thiopurine pharmacologic testing
7 Thiopurines
Affected by prior erythrocyte Utility for assessing

Test Relationship to dosing Relevant thiopurines transfusions compliance
Erythrocyte thioguanine Best after steady-state Mercaptopurine, thioguanine, Minimallyb Yes
nucleotide (TGN) dosinga azathioprine
metabolites
Erythrocyte methyl Best after steady-state Mercaptopurine, azathioprine Minimally Yes
mercaptopurine nucleotide dosinga
metabolites (MeTIMP)
MeTIMP/TGN ratio Best after steady-state Mercaptopurine, azathioprine Minimally Yes
dosinga
Erythrocyte thiopurine Not affected Mercaptopurine, thioguanine, Yes, especially if low activity No
methyltransferase (TPMT) azathioprine recipient receives blood
activity from high activity donorc
TPMT genotyped Not affected Mercaptopurine, thioguanine, No No
azathioprine
a
Most informative if patient has been receiving constant regular dosing for at least 24 weeks and obtained at 624 h from last dose
b
Given a starting hemoglobin of 8 g/dl and a TGN of 3,000 pmol/8 108 RBCs, if a patient is transfused with enough blood to raise the hemoglobin to 10 g/
dl, the erythrocyte TGN is projected to be diluted to 2,400 pmol/8 108 RBCs
c
Given a starting hemoglobin of 8 g/dl and a TPMT of 1 U/ml PRBCs, if a patient is transfused with enough blood to raise the hemoglobin to 10 g/dl, the
erythrocyte TPMT is projected to be artifactually elevated to 5.2 U/ml PRBCs, if transfused with blood cells from a donor whose TPMT status is 22 U/ml
PRBCs (consistent with homozygous wild-type90% chance)
d
TPMT genotype might not match host phenotype if patient has undergone allogeneic stem cell transplant, or if patient has a rare inactivating polymorphism
not interrogated by the genotyping test
97
Hematologic parameters are also used to monitor toxicity. Low platelets (<50
109/L), WBC (<1,000/mm3), and ANC (<300/mm3) could potentially lead to bleed-
ing complications and infection and often warrant withholding therapy to allow
levels to return to acceptable values. It is important to point out that when therapy
is withheld for any reason, the potential for treatment failure (i.e., disease progres-
sion or recurrence) can be increased. Thus, it is likely beneficial to avoid a period
of profound myelosuppression that may compromise therapy.
Thiopurine metabolite levels (i.e., MeTIMP and TGNs) can be used as param-
eters to assess and guide therapy (Table 7.2). As discussed later in this chapter,
medication compliance can be assessed in patients who are wild-type for TPMT
receiving mercaptopurine by the MeTIMP/TGN ratio. And finally, as mentioned
previously, elevated liver transaminases, total bilirubin, and uric acid are markers
of liver toxicity and should be monitored routinely while patients are receiving
thiopurines.
7.3Pharmacogenetics and Pharmacogenomics of Thiopurines
It is well known that interindividual differences in drug response exists, and

that genetic variation plays an integral role in the observed drug phenotypes.
TPMT is one of the best examples of the importance of pharmacogenetics in
providing optimal drug therapy. The genetic variations in TPMT result in a tri-
modal population distribution in TPMT activity [25]. TPMT protein activity is
directly correlated to the level of TPMT protein expressed, and inversely cor-
related to intracellular levels of TGNs (Fig. 7.3) [26]. Approximately 90% of
the population inherit homozygous wild-type TPMT alleles (TPMT*1/*1) and
have high levels of protein activity, ~10% are heterozygous and have
intermediate activity, and <1% are homozygous variant and have low to no
detectable activity (Fig. 7.3a, b). Individuals with low TPMT protein expression
are at risk for having high intracellular concentrations of active TGN metabo-
lites, and hence are at an increased risk of life threatening myelosuppression
and secondary cancers (Fig. 7.3c). Those who are TPMT deficient may require
as much as a 15-fold dose reduction in order to have comparable intracellular
TGNs and toxicity as patients with wild-type TPMT [27]. Low TPMT protein
activity caused by variant alleles is a result of increased protein degradation as
compared to wild-type TPMT (Fig. 7.3d).
7.3.1Relevant TPMT SNPs and Haplotypes
TPMT is located on chromosome 6 (locus 6p22.3), is 34 kilobases in length [28],

and consists of 10 exons (Fig. 7.4). There are 28 known variant alleles [29], many
of which have been associated with decreased protein levels in invitro studies [29].
7 Thiopurines 99
Fig. 7.3 Thiopurine methyltransferase and the clinical consequence of the genetic polymorphism,
reproduced with permission from Clinical Pharmacology and Therapeutics [26]. (a) The genetic
polymorphism in TPMT results in a trimodal population frequency distribution in TPMT activity
with deficient activity caused by inheritance of two variant (var) alleles, intermediate activity
caused by heterozygosity, and high activity associated with homozygous wild-type (*1/*1) geno-
types. (b) TPMT activity is directly proportional to the amount of TPMT protein; (c) and is
inversely related to intracellular concentrations of active TGN metabolites. Low TPMT is associ-
ated with toxicity and high TPMT may result in increased risk of relapse. (d) The biochemical
basis for low protein conferred by the most common variant polymorphism (*3A) is illustrated by
the longer half-life for invitro expressed TPMT*1 when compared to the variant protein
Fig. 7.4 TPMT genetic variants representing ~90% of the intermediate and low enzymatic activity.
Exonic locations are shown with the amino acid substitutions and SNP identifiers (from dbSNP)
listed
The level of TPMT enzymatic activity is currently the best predictor of how patients
will respond to thiopurine therapy. Although there are 28 known TPMT variants,
the level of TPMT protein activity is most often influenced by a few well-studied
deactivating genetic polymorphisms: TPMT*2, TPMT*3A, TPMT*3B, and
TPMT*3C (Fig. 7.4). Importantly, the frequency at which these variants occur will
vary based on race or ethnic background.
To date, the TPMT*3A variant, consisting of two nonsynonymous coding single
nucleotide polymorphisms (SNPs) located on exons 7 (ala154Thr) and 10
7 Thiopurines 101
(Tyr240Cys), is the major clinically relevant haplotype that is predictive of TPMT

activity and is the most common TPMT variant in whites [30]. The TPMT*3C allele
consists of a single nonsynonymous coding SNP located on exon 10 (Tyr240Cys)
and occurs most commonly in individuals of Eastern Asian and African decent
(Fig. 7.4). Although the TPMT*2 (exon 5) and TPMT*3B (exon 7) alleles have been
well studied; they are much less common than TPMT*3A or TPMT*3C variants.
Overall, these variants account for ~90% of the TPMT inactivating alleles. Although
TPMT genotype is predictive of TPMT activity and thiopurine tolerability in
homozygous variant individuals (these patients have no measurable protein activ-
ity), there is a great degree of variability in TPMT activity based on genotype in
wild-type and heterozygous individuals.
As discussed later in the chapter, there are three clinical tests used to assess TPMT
phenotype. The importance of using TPMT phenotypes to guide thiopurine therapy
is reflected in the fact that patients who receive standard therapy and have low intra-
cellular TGN metabolite levels have an increased risk of treatment failure, and
patients who have excessively high TPMT metabolites (i.e., TGNs) have an increased
risk of second cancers and life threatening myelosuppression [10, 25, 3134].
7.3.2Future Gene Associations and GWAS Studies
Although TPMT is well known to be important in thiopurine therapy, identifying

other genetic predictors of thiopurine toxicity and response to further optimize
therapy is the next step. Candidate gene or genome wide association studies
(GWAS) are the key to determining novel genes that can help further delineate
other predictors of thiopurine-associated toxicities as well as better predictors of
TPMT activity in individuals who have a wild-type TPMT genotype. Even TPMT
activity itself appears to be influenced by genetic variation outside of the TPMT
gene [26].
Recent studies have implicated inosine triphosphate pyrophophatase (ITPA) as
being predictive of the risk of thiopurine toxicity. ITPA is the enzyme responsible
for catalyzing the conversion of inosine triphosphate to inosine monophosphate (an
intermediate in purine metabolism) (Fig. 7.2). ITPA deactivating genetic polymor-
phisms result in decreased ITPA protein activity, which can lead to high levels of a
potentially cytotoxic metabolite, ITP, when thiopurines are given [35, 36]. Some
studies have found a correlation between ITPA and adverse events after thiopurines
in the treatment of inflammatory bowel disease [35] whereas others have failed to
show significant correlations in this patient population [37]. Interestingly, Stocco
etal. found that when the thiopurine dose is adjusted based on TPMT genotype in
ALL patients, individuals having the deactivating ITPA variant, rs41320251, were
at an increased risk of severe febrile neutropenia [36]. However, additional studies
in independent populations are needed to elucidate what role ITPA will play in
predicting thiopurine-associated toxicities and optimizing therapy.
7.4Clinical Utility of Pharmacogenetic Testing

for Personalized Medicine
There are only three clinically used thiopurines (mercaptopurine, thioguanine, and
azathioprine), and the metabolism of all three are affected by the TPMT polymorphism
(Fig. 7.1) [25, 31, 32, 38, 39]. The clinical utility of genetic testing for these agents is
perhaps more widely adopted than for any other group of medications or single genetic
test. The product labeling for all three thiopurines includes language on the impact of
pharmacogenetics on adverse effects and availability of testing (Table 7.3).
This group of drugs, paired with the genetic polymorphism in TPMT, represent a
perfect storm of findings that lend themselves to incorporation into clinical medi-
cine, such that the pharmacogenetics of thiopurines is often held up as one of the
most compelling examples in clinical pharmacology [33, 40]. First, the agents have
a narrow therapeutic index: in any one patient, the difference between a dose that is
effective and tolerated and a dose that is unacceptably toxic can be very small.
Second, the toxicities can be life-threatening (acute myelosuppression and secondary
cancers) [27, 39, 4143]. Third, some of the toxicities (e.g., leukemogenesis) have a
delay of years [1619, 44], and therefore cannot be monitored in real time and used
to adjust doses. Fourth, the diseases that are being treated with thiopurines can also
be life-threatening (e.g., leukemia) or extremely serious (e.g., Crohns disease) and
thus rapid introduction of the most effective dose of these agents is crucial to cure
the patient [34, 4548]. Fifth, these medications are often used in combination with
other agents that cause overlapping toxicities (e.g., other immunosuppressants or
anticancer drugs that can cause infection) and thus clinical monitoring parameters
(such as blood counts) may not differentiate which agent is the major culprit in caus-
ing acute toxicity. Sixth, a small number of polymorphisms account for the vast
majority of the inactivating alleles in the TPMT gene [11,4952], making genetic
testing feasible with a small number of interrogated polymorphisms. Seventh, a
single blood sample can be used to measure pharmacologic phenotypes of interest
(erythrocyte thiopurine metabolite levels and TPMT enzyme activity) that comple-
ment the genetic testing result. Finally, the TPMT polymorphism has no known
consequences in the absence of thiopurine exposure, thus, testing for it does not
carry any societal concerns outside the context of medication use, as is true for some
other polymorphisms that may also affect constitutive disease risk implications.
There are three types of tests available for thiopurine monitoring (Table 7.2):
erythrocyte TGN metabolites, erythrocyte TPMT activity, and TPMT genotype. For
azathioprine and mercaptopurine, erythrocyte MeTIMP can also be monitored.
Measurement of thiopurine metabolites allows for an assessment of patient
compliance or adherence with the oral thiopurine therapy [5355]. Patients who
have been taking thiopurines will have detectable thiopurine metabolite levels in
erythrocytes, even when plasma levels are not detectable (Fig. 7.5) [56]. Those with
wild-type TPMT will have MeTIMP/TGN ratios that are high, whereas if such
patients have been noncompliant, their ratios will be low and the absolute level of
TGN will be very low (e.g., <50 or 100 pmol/8 108 RBCs). There are data to
indicate that at least some patients who are wild-type for TPMT when they receive
7 Thiopurines
Table 7.3 Labeling related to pharmacogenetics of thiopurines as related to the TPMT polymorphism
Drug Label sections Dosing recommendations Testing
Mercaptopurinea Clinical pharmacology, warnings/ Substantial dose reductions for TPMT genotyping or phenotyping
precautions, laboratory tests, homozygous-TPMT deficient patients can identify patients who
dosage and administration to avoid life-threatening bone marrow have homozygous deficient or
suppression. Optimal starting dose for intermediate TPMT activity
heterozygotes not established
Thioguanineb Laboratory tests, dosage and Substantial dosage reductions may be required Some laboratories offer testing for
administration, warnings to avoid the life-threatening bone marrow TPMT deficiency
suppression in those with inherited
deficiency of the enzyme TPMT
Azathioprinec Clinical pharmacology, warnings/ Patients with low or absent TPMT activity are It is recommended that consideration
precautions, laboratory tests, at an increased risk of developing severe, be given to either genotype or
adverse reactions life-threatening myelotoxicity if receiving phenotype patients for TPMT
conventional doses
Dosage reduction is recommended in patients
with reduced TPMT activity
a
Labeling for Purinethol as of August 2003
b
Labeling for Tabloid as of June 2009
c
Labeling for Imuran as of May 2008
103
Fig. 7.5 Plasma concentrations of mercaptopurine (left y axis, dashed line) after daily doses of
mercaptopurine demonstrate a rapid plasma half-life and no accumulation. Erythrocyte (RBC)
concentrations of thioguanine nucleotide active metabolites (RBC TGN) indicate the slow accu-
mulation over a period of ~2 weeks of dosing to an eventual steady state concentration
intrapatient dosage escalations of mercaptopurine, shunt the medication toward

higher MeTIMP levels with little change in TGN concentrations [57, 58].
Patients receiving thioguanine have no MeTIMP levels to measure, as this
metabolite is formed from an intermediate of mercaptopurine that is not formed
after administration of thioguanine. Patients receiving thioguanine tolerate much
higher levels of TGN than do patients receiving mercaptopurine. Thus, any putative
target range of TGNs must account for the thiopurine being used. Moreover, as
most of the indications for thiopurines (e.g., inflammatory bowel disease, ALL) are
treated with a number of agents in addition to thiopurines, putative therapeutic
ranges are likely to be influenced by the doses of the concurrent medications.
TPMT activity is widely expressed, and thus activity in any one tissue (e.g.,
erythrocytes) reflects the activity in tissues more important for drug disposition (e.g.,
liver) and in drug action (e.g., bone marrow) [5962]. However, TPMT activity
appears to be higher in certain fractions of erythrocytes than others (and thus can be
affected by the proportion of immature erythrocytes in the sample) [63], by age [64],
and by time on therapy [34, 65], although the factors that contribute to such
intraindividual variability are poorly defined. Because patients with low TPMT
activity are at higher risk of receiving blood transfusions than those with wild-type
activity, and the majority of red cell donors will have wild-type TPMT status, eryth-
rocyte TPMT activity can be artifactually elevated by recent transfusions (Table 7.2) [66].
Nonetheless, TPMT activity is a useful adjunct, along with thiopurine metabolite
levels and TPMT genotype, to impute the TPMT status for any given patient.
TPMT genotypes are imputed based on the most likely haplotypes, given the
interrogated SNPs [50, 6769]. Complementary phenotype data are extremely helpful
7 Thiopurines 105
for several reasons. There is always a chance for one sample or another to be mislabeled
somewhere in the testing process, and this is particularly concerning when dealing
with genotypes, which are based on a single sample. A wild-type TPMT activity is
simply not consistent with a homozygous or even heterozygous low-activity genotype;
one sample or the other would be suspect and solved with repeat sampling. In addition,
most commercial genotyping assays test for only the three most common inactivating
SNPs; although rare, it is possible for a patient to have a rare inactivating SNP that
could result in a spurious wild-type genotype accompanied by low TPMT activity or
low MeTIMP/TGN ratio. Another rare possibility would be that two inactivating SNP
variants (e.g., Ala 154 Thr and Tyr 240 Cys, rs1800460 and rs1142345), each of which
is present in the heterozygous state, which normally are assumed to be in linkage dis-
equilibrium and thus allelic with each other could actually be present on opposite
alleles; this would lead to a genotype call of heterozygote (*1/*3A) but in fact repre-
sent a compound homozygote deficient genotype call (*3B/*3C). Whereas, when a
*1/*3A genotype would be consistent with heterozygous TPMT activity (and a low but
detectable MeTIMP/TGN ratio), and *3B/*3C genotype would be consistent with
undetectable or very low TPMT activity and undetectable MeTIMP levels pheno-
types that would be readily distinguished with one of the two phenotyping tests.
Multiple types of variant and wild-type alleles exist for every gene, and the
frequency of variant TPMT alleles differs substantially by racial or ancestral back-
ground [7073]. Because the cost of determining DNA sequence at every nucle-
otide is still prohibitively high, genotyping tests depend on technologies that survey
a patients DNA at particular target genetic sequences. One of the reasons that
TPMT genotyping has been more widely adopted than other pharmacogenetic tests
is that a relatively small number of variant alleles account for the vast majority of
TPMT inactivating variants (Figs. 7.4 and 7.6) [11, 67, 74], and accurately
categorize patients according to three major TPMT phenotypes: homozygous
variant or deficient, heterozygote, or homozygous wild-type. However, there is
Fig. 7.6 Distribution of TPMT genotypes from a large group of individuals of European ancestry.
Figure illustrates data from Shaeffeler etal. [51]
always the possibility for false negatives: a genotyping test cannot reveal any
information about areas of the gene not interrogated by the test (e.g., one can only
know that the patient is wild-type at the loci tested), and rare patients will have
rare or novel inactivating variants that will be missed by standardized genotyping
tests. The number of false negatives depends on the proportion of inactivating vari-
ants accounted for by the tested variants (which must be disclosed by the test).
Another complication is that most genotyping tests cannot experimentally deter-
mine haplotype, although such information is available using rather labor-intensive
long-range PCR tests [68, 75]. Thus, standard genotyping tests rely upon compar-
ing the results at individual loci with the population probabilities that polymor-
phisms are allelic with each other. If a patient is heterozygote at more than one
polymorphic site in a gene, and there is a high probability that the polymorphisms
are allelic (e.g., as is true for the Ala 154 Thr and Tyr 240 Cys variants, rs1800460
and rs1142345), the genotype for the patient is a likely heterozygote. If the geno-
type at two separate sites is heterozygote, but there is a high probability that the
polymorphisms are not allelic to each other (as would be true for the Ala 80 Pro
coupled with the Tyr 240 Cys variants, rs1800462 and rs1142345), then the patient
likely has a homozygous variant genotype.
7.4.1Dosage Adjustments
Although the importance of TPMT status to thiopurine tolerance is extremely well

documented, some clinicians do not test patients prior to starting thiopurine therapy.
Our group has long advocated testing for TPMT status prior to initiating thiopurine
therapy so that dosages can be adjusted accordingly. One reason to test everyone is
that even a short course of thiopurines given to the rare (1 in 300) homozygous
deficient individuals can result in serious myelosuppression [76, 77], which could be
avoided by starting with dramatically decreased doses (>10-fold lower) or choosing
an alternative therapy. Although it is true that only ~35% of heterozygous patients
receiving normal thiopurine doses experience myelosuppression so severe that doses
must be decreased, there are disadvantages of beginning thiopurine therapy and
adjusting doses downward only in those who experience toxicity. In treating cancer,
some regimens have so many myelosuppressive agents given relatively early in
therapy that profound myelosuppression early on may delay subsequent therapy
making thiopurine dose titration less feasible. Moreover, because some serious
long-term adverse effects (second tumors) have been associated with defective
TPMT activity [1619], even in patients who do not experience acute toxicity [78],
some advocate capping doses of thiopurine in heterozygotes at some dose lower than
recommended for homozygous wild-type patients (Fig. 7.5), although there are no
data to indicate that such a strategy decreases the risk of second cancers. However,
there are data to indicate that when TPMT status is used, in conjunction with clinical
myelosuppression, to adjust mercaptopurine doses, relapse rates in childhood ALL
are not adversely impacted [45].
7 Thiopurines 107
Whereas TPMT status has been convincingly linked to tolerance of thiopurines,

specific target levels of erythrocyte thiopurine metabolites have not been well
established. Although a precise threshold TGN level that is consistent with non-
compliance has not been established, our practice is to counsel patients regarding
the dangers of noncompliance to TGN levels <100 pmol. Moreover, although
MeTIMP and high TPMT activity have been clearly implicated as associated with
higher risk of hepatic toxicity [12, 79, 80], the tests have a poor predictive power,
as most patients with high MeTIMP do not develop hepatotoxicity. At least in ALL,
the presence of elevated serum aminotransferases has been associated with
improved cure rates [13], and thus it is not clear whether mild hepatic dysfunction
should be considered an indication to modify therapy. Some have advocated an
approach of combining allopurinol or thioguanine with mercaptopurine in order to
maximize formation of TGN and minimize formation of MeMPN [24], which may
obviate some of the hepatotoxicity that may be associated with MeMPN. It should
also be acknowledged that thioguanine (as opposed to mercaptopurine) has been
associated with a higher risk of veno-occlusive disease of the liver [8183] and thus
the use of thioguanine may be associated with its own set of liver toxicity.
Therefore, the algorithm we use for adjusting mercaptopurine doses in patients
with ALL at St. Jude (Fig. 7.7) uses TPMT status to institute lower-than-normal
starting doses of thiopurines in the rare homozygote-deficient patients and the
heterozygous patients but titrates final dosages based on tolerance to thiopurines,
without targeting specific therapeutic ranges for TGN or MeTIMP metabolites.
Although only about 35% of heterozygotes require a dosage reduction based on
acute myelosuppression, the median daily dose for those who do experience myelo-
suppression is ~40% lower (44 mg/m2) than that tolerated in wild-type patients
(75mg/m2) [79, 84]. After instituting lower starting dosages in those with at least
one variant TPMT allele, we then titrate up the thiopurine dose to achieve the
desired level of myelosuppression, generally leaving the nonthiopurine therapy
unadjusted. In heterozygotes, we generally cap the mercaptopurine dose somewhat
lower (60 mg/m2) than the normal mercaptopurine dose in wild-type patients
because of our concerns for secondary tumorigenesis. In those with wild-type
TPMT, we adjust the doses of mercaptopurine plus any other myelosuppressive
agents based on an algorithm that places no special emphasis on thiopurines as the
culprits (Fig. 7.7). Other polymorphisms, such as ITPA for thiopurines [36, 85] or
SLCO1B1 for methotrexate [86], may have an important role in acute toxicity of
therapy, and as these are validated, may be incorporated into dosage regimens.
7.5Case Report
JB is a 6-year-old child with high-risk ALL. A phase of intensification chemo-

therapy included cytarabine 50 mg/m2/day 5 days; mercaptopurine 50 mg/m2/day
PO 10 days; and cyclophosphamide 500 mg/m2 IV on days 1 and 5. JB was
neutropenic for 6 weeks whereas the median period neutropenia is only 10 days.
Fig. 7.7 Overview of algorithm for dosing of acute leukemia continuation therapy including mercap-
topurine among children at St. Jude. For the 1 in 300 patients with homozygous TPMT deficiency,
mercaptopurine doses are reduced to less than 10% of the standard dose (e.g., 10 mg/m2/day 3 days/
week instead of 75 mg/m2/day daily). For the ~10% of patients with heterozygote TPMT status, we
attempt to cap the mercaptopurine dose at 60 mg/m2/day; if myelosuppression is observed, the prefer-
ence is to reduce the mercaptopurine dose rather than that of other myelosuppressive agents (such as
methotrexate). For the ~90% of patients who are wild-type for TPMT, there is no reason (based on
TPMT status) to preferentially decrease the dose of mercaptopurine vs. any other myelosuppressive
agent. Doses of chemotherapy in ALL are generally titrated to achieve a target level of myelosuppres-
sion; in patients with a defect in TPMT activity, there is a pharmacologic basis for focusing on decreas-
ing the dose of thiopurine and attempting to give the other myelosuppressive agents at least full dose
Amaintenance phase of chemotherapy started with methotrexate 20 mg/m2 orally

every week and mercaptopurine orally at 75 mg/m2/day. On day 14 of maintenance,
the child was admitted for fever and neutropenia, with a platelet count of 20,000/l
and hemoglobin of 6 g/dl. Both mercaptopurine and methotrexate were stopped. He
has required multiple RBC transfusions over the last 8 weeks. On day 17 of main-
tenance, blood was drawn for thiopurine testing and showed a RBC TPMT activity
of 6 U/ml PRBCs (heterozygote range = 514 U/ml), RBC TGN = 800 pmol/8
108 RBCs, MeMPN was undetectable, and the genotype was *1/*3A.
7.5.1How Do We Interpret JBs Findings?
Only 10 days of thiopurines were given initially, along with other myelosuppressive
therapy, so it is possible that the 6 weeks of myelosuppression was caused at least
partly by agents other than thiopurines. However, the repeat of severe pancytopenia
7 Thiopurines 109
(platelets, red cells, and neutrophils all suppressed) after only 2 weeks of maintenance
raises suspicions that thiopurines are the culprits.
The RBC TPMT activity of 6 U/ml PRBCs would be consistent with a TPMT
heterozygote; however, because the patient has received multiple RBC transfusions
within the 90 days leading up to the test, the activity may be artifactually altered
(probably increased) by allogeneic transfusions, so perhaps the TPMT activity is
lower than 10 U/ml.
The fact that the RBC TGN is 800 pmol/8 108 RBCs (whereas after just 14
days of therapy, wild-type patients would generally have levels 100400 pmol/8
108 RBCs) [53, 87, 88] would be consistent with low TPMT activity. Because most
normal ranges are based on samples drawn <24 h from the last dose of a regimen
of at least 2 weeks of therapy (i.e., steady state) [89], and this patient actually
received no mercaptopurine in the 72 h prior to this sampling, this TGN may actu-
ally be lower than reflective of the patients true steady-state. The fact that the
MeMPN levels are undetectable, in the presence of relatively high TGN, is highly
indicative of absent (homozygous deficient) TPMT activity in this patient.
The genotype was *1/*3A, consistent with a heterozygote. Based on the fre-
quency of the *3A allele, this is the most likely genotype. However, because of the
very severe toxicity, the absent methyl metabolites, and the high TGN despite mod-
est dosing, in this case, the genotype interpretation of *3B/*3C (the low activity
genotypes on different alleles) is more likely. On further investigation, it turns out
that JB is of African ancestry, a group for whom *3B and *3C alleles are more
common than in those of European ancestry.
Thus, our ultimate interpretation is that this patient is a rare TPMT homozygous
deficient patient. When counts recover, mercaptopurine is restarted at 10 mg/m2/day
on 3 days per week, along with methotrexate at 40 mg/m2/week. After 8 weeks of
tolerating this regimen, no transfusions are required. A repeat TPMT activity measure
comes back below the limit of detection of 1 U/ml, and TGN are at 760 pmol/8 108
RBCs, and MeMPNs remain undetectable. Later in therapy, the mercaptopurine dose
is titrated (based only on the desired level of neutrophil count) up to 10 mg/m2/day
7 days a week, and no transfusions are required for the remainder of therapy.
Acknowledgments Supported by NCI grant CA 21765 and the NIH/NIGMS Pharmacogenetics

Research Network and Database (U01 GM61393, and by the American Lebanese Syrian
Associated Charities (ALSAC).
References
1. DeVita, V. T., Jr., & Chu, E. (2008). A history of cancer chemotherapy. Cancer Research, 68,
86438653.
2. Cheok, M. H., Lugthart, S., & Evans, W. E. (2006). Pharmacogenomics of acute leukemia.
Annual Review of Pharmacology and Toxicology, 46, 317353.
3. Gearry, R. B., & Barclay, M. L. (2005). Azathioprine and 6-mercaptopurine pharmacogenetics
and metabolite monitoring in inflammatory bowel disease. Journal of Gastroenterology and
Hepatology, 20(8), 11491157.
4. Sandborn, W., Sutherland, L., Pearson, D., May, G., Modigliani, R., & Prantera, C. (2000).
Azathioprine or 6-mercaptopurine for inducing remission of Crohns disease. Cochrane
Database of Systematic Reviews (Online), (2), CD000545.
5. Pearson, D. C., May, G. R., Fick, G., & Sutherland, L. R. (2000). Azathioprine for maintaining
remission of Crohns disease. Cochrane Database of Systematic Reviews (Online), (2),
CD000067.
6. Derijks, L. J., Gilissen, L. P., Hooymans, P. M., & Hommes, D. W. (2006). Review article:
Thiopurines in inflammatory bowel disease. Alimentary Pharmacology & Therapeutics, 24,
715729.
7. Czaja, A. J. (2009). Current and future treatments of autoimmune hepatitis. Expert Review of
Gastroenterology & Hepatology, 3, 269291.
8. Gauba, V., Saldanha, M., Vize, C., & Saleh, G. M. (2006). Thiopurine methyltransferase
screening before azathioprine therapy. The British Journal of Ophthalmology, 90, 923924.
9. Allan, P. W., & Bennett, L. L., Jr. (1971). 6-Methylthioguanylic acid, a metabolite of
6-thioguanine. Biochemical Pharmacology, 20, 847852.
10. Krynetski, E., & Evans, W. E. (2003). Drug methylation in cancer therapy: Lessons from the
TPMT polymorphism. Oncogene, 22, 74037413.
11. Yates, C. R., Krynetski, E. Y., Loennechen, T., Fessing, M. Y., Tai, H. L., & Pui, C. H., etal.
(1997). Molecular diagnosis of thiopurine S-methyltransferase deficiency: genetic basis for
azathioprine and mercaptopurine intolerance. Annals of Internal Medicine, 126, 608614.
12. Berkovitch, M., Matsui, D., Zipursky, A., Blanchette, V. S., Verjee, Z., & Giesbrecht, E., etal.
(1996). Hepatotoxicity of 6-mercaptopurine in childhood acute lymphocytic leukemia: phar-
macokinetic characteristics. Medical and Pediatric Oncology, 26, 8589.
13. Schmiegelow, K., & Pulczynska, M. (1990). Prognostic significance of hepatotoxicity during
maintenance chemotherapy for childhood acute lymphoblastic leukaemia. British Journal of
Cancer, 61, 767772.
14. Gisbert, J. P., Gonzalez-Lama, Y., & Mate, J. (2007). Thiopurine-induced liver injury in
patients with inflammatory bowel disease: A systematic review. The American Journal of
Gastroenterology, 102, 15181527.
15. de Boer, N. K., van Bodegraven, A. A., Jharap, B., de Graaf, P., & Mulder, C. J. (2007). Drug
insight: Pharmacology and toxicity of thiopurine therapy in patients with IBD. Nature Clinical
Practice. Gastroenterology & Hepatology, 4, 686694.
16. Relling, M. V., Rubnitz, J. E., Rivera, G. K., Boyett, J. M., Hancock, M. L., & Felix, C. A.,
etal. (1999). High incidence of secondary brain tumours after radiotherapy and antimetabo-
lites. Lancet, 354, 3439.
17. Yenson, P. R., Forrest, D., Schmiegelow, K., & Dalal, B. I. (2008). Azathioprine-associated
acute myeloid leukemia in a patient with Crohns disease and thiopurine S-methyltransferase
deficiency. American Journal of Hematology, 83(1), 8083.
18. Relling, M. V., Yanishevski, Y., Nemec, J., Evans, W. E., Boyett, J. M., & Behm, F. G., etal.
(1998). Etoposide and antimetabolite pharmacology in patients who develop secondary acute
myeloid leukemia. Leukemia, 12, 346352.
19. Thompsen, J., Schroder, H., Kristinsson, J., Madsen, B., Szumlanski, C., & Weinshilboum, R.,
etal. (1999). Possible carcinogenic effect of 6-mercaptopurine on bone marrow stem cells:
Relation to thiopurine metabolism. Cancer, 86, 10801086.
20. Hijiya, N., Hudson, M. M., Lensing, S., Zacher, M., Onciu, M., & Behm, F. G., etal. (2007).
Cumulative incidence of secondary neoplasms as a first event after childhood acute lympho-
blastic leukemia16. The Journal of the American Medical Association, 297, 12071215.
21. Karran, P., & Attard, N. (2008). Thiopurines in current medical practice: molecular mecha-
nisms and contributions to therapy-related cancer. Nature Reviews. Cancer, 8, 2436.
22. Corominas, H., Domenech, M., Laiz, A., Gich, I., Geli, C., & Diaz, C., etal. (2003). Is thio-
purine methyltransferase genetic polymorphism a major factor for withdrawal of azathioprine
in rheumatoid arthritis patients? Rheumatology (Oxford), 42, 4045.
23. Dubinsky, M. C., Lamothe, S., Yang, H. Y., Targan, S. R., Sinnett, D., & Theoret, Y., etal.
(2000). Pharmacogenomics and metabolite measurement for 6-mercaptopurine therapy in
inflammatory bowel disease. Gastroenterology, 118, 705713.
7 Thiopurines 111
24. Leong, R. W., Gearry, R. B., & Sparrow, M. P. (2008). Thiopurine hepatotoxicity in
inflammatorybowel disease: The role for adding allopurinol. Expert Opinion on Drug Safety,
7, 607616.
25. Weinshilboum, R. M., & Sladek, S. L. (1980). Mercaptopurine pharmacogenetics: monogenic
inheritance of erythrocyte thiopurine methyltransferase activity. American Journal of Human
Genetics, 32, 651662.
26. Jones, T. S., Yang, W., Evans, W. E., & Relling, M. V. (2007). Using HapMap tools in phar-
macogenomic discovery: The thiopurine methyltransferase polymorphism. Clinical
27. Evans, W. E., Horner, M., Chu, Y. Q., Kalwinsky, D., & Roberts, W. M. (1991). Altered mer-
captopurine metabolism, toxic effects, and dosage requirement in a thiopurine methyltrans-
ferase-deficient child with acute lymphocytic leukemia. The Journal of Pediatrics, 119,
985989.
28. Szumlanski, C., Otterness, D., Her, C., Lee, D., Brandriff, B., & Kelsell, D., et al. (1996).
Thiopurine methyltransferase pharmacogenetics: human gene cloning and characterization of
a common polymorphism. DNA and Cell Biology, 15, 1730.
29. Ujiie, S., Sasaki, T., Mizugaki, M., Ishikawa, M., & Hiratsuka, M. (2008). Functional charac-
terization of 23 allelic variants of thiopurine S-methyltransferase gene (TPMT*2 *24).
Pharmacogenetics and Genomics, 18, 887893.
30. Tai, H. L., Krynetski, E. Y., Yates, C. R., Loennechen, T., Fessing, M. Y., & Krynetskaia, N.
F., etal. (1996). Thiopurine S-methyltransferase deficiency: Two nucleotide transitions define
the most prevalent mutant allele associated with loss of catalytic activity in Caucasians.
American Journal of Human Genetics, 58, 694702.
31. Krynetski, E. Y., & Evans, W. E. (1998). Pharmacogenetics of cancer therapy: Getting per-
sonal. American Journal of Human Genetics, 63, 1116.
32. Evans, W. E., & McLeod, H. L. (2003). Pharmacogenomicsdrug disposition, drug targets,
and side effects. The New England Journal of Medicine, 348, 538549.
33. Weinshilboum, R. (2003). Inheritance and drug response. The New England Journal of
Medicine, 348, 529537.
34. Lennard, L., Lilleyman, J. S., Van Loon, J., & Weinshilboum, R. M. (1990). Genetic variation
in response to 6-mercaptopurine for childhood acute lymphoblastic leukaemia. Lancet, 336,
225229.
35. Marinaki, A. M., Ansari, A., Duley, J. A., Arenas, M., Sumi, S., & Lewis, C. M., etal. (2004).
Adverse drug reactions to azathioprine therapy are associated with polymorphism in the gene
encoding inosine triphosphate pyrophosphatase (ITPase). Pharmacogenetics, 14, 181187.
36. Stocco, G., Cheok, M. H., Crews, K. R., Dervieux, T., French, D., & Pei, D., etal. (2009).
Genetic polymorphism of inosine triphosphate pyrophosphatase is a determinant of mercap-
topurine metabolism and toxicity during treatment for acute lymphoblastic leukemia. Clinical
37. Van Dieren, J. M., Hansen, B. E., Kuipers, E. J., Nieuwenhuis, E. E., & Van der Woude, C. J.
(2007). Meta-analysis: Inosine triphosphate pyrophosphatase polymorphisms and thiopurine
toxicity in the treatment of inflammatory bowel disease. Alimentary Pharmacology &
38. Lennard, L., Van Loon, J. A., & Weinshilboum, R. M. (1989). Pharmacogenetics of acute
azathioprine toxicity: Relationship to thiopurine methyltransferase genetic polymorphism.
39. McBride, K. L., Gilchrist, G. S., Smithson, W. A., Weinshilboum, R. M., & Szumlanski, C.
L. (2000). Severe 6-thioguanine-induced marrow aplasia in a child with acute lymphoblastic
leukemia and inhibited thiopurine methyltransferase deficiency. Journal of Pediatric
Hematology/Oncology, 22(5), 441445.
40. Evans, W. E., & Relling, M. V. (1999). Pharmacogenomics: Translating functional genomics
into rational therapeutics. Science, 286, 487491.
41. Boonsrirat, U., Angsuthum, S., Vannaprasaht, S., Kongpunvijit, J., Hirankarn, N., &
Tassaneeyakul, W., et al. (2008). Azathioprine-induced fatal myelosuppression in systemic
lupus erythematosus patient carrying TPMT*3C polymorphism. Lupus, 17, 132134.
42. Tassaneeyakul, W., Srimarthpirom, S., Reungjui, S., Chansung, K., & Romphruk, A. (2003).
Azathioprine-induced fatal myelosuppression in a renal-transplant recipient who carried
heterozygous TPMT*1/*3C. Transplantation, 76, 265266.
43. Gisbert, J. P., Luna, M., Mate, J., Gonzalez-Guijarro, L., Cara, C., & Pajares, J. M. (2006).
Choice of azathioprine or 6-mercaptopurine dose based on thiopurine methyltransferase
(TPMT) activity to avoid myelosuppression. A prospective study. Hepatogastroenterology,
53, 399404.
44. Schmiegelow, K., Al-Modhwahi, I., Andersen, M. K., Behrendtz, M., Forestier, E., &
Hasle, H., etal. (2009). Methotrexate/6-mercaptopurine maintenance therapy influences the
risk of a second malignant neoplasm after childhood acute lymphoblastic leukemia: Results
from the NOPHO ALL-92 study. Blood, 113, 60776084.
45. Relling, M. V., Pui, C. H., Cheng, C., & Evans, W. E. (2006). Thiopurine methyltransferase
in acute lymphoblastic leukemia1. Blood, 107, 843844.
46. Ansari, A., Elliott, T., Baburajan, B., Mayhead, P., ODonohue, J., & Chocair, P., etal. (2008).
Long-term outcome of using allopurinol co-therapy as a strategy for overcoming thiopurine
hepatotoxicity in treating inflammatory bowel disease. Alimentary Pharmacology &
47. Gisbert, J. P., Nino, P., Rodrigo, L., Cara, C., & Guijarro, L. G. (2006). Thiopurine methyl-
transferase (TPMT) activity and adverse effects of azathioprine in inflammatory bowel dis-
ease: Long-term follow-up study of 394 patients. The American Journal of Gastroenterology,
101(12), 27692776.
48. Schmiegelow, K., Forestier, E., Kristinsson, J., Soderhall, S , Vettenranta, K., & Weinshilboum, R.,
etal. (2009). Thiopurine methyltransferase activity is related to the risk of relapse of childhood
acute lymphoblastic leukemia: Results from the NOPHO ALL-92 study. Leukemia, 23, 557564.
49. Kumagai, K., Hiyama, K., Ishioka, S., Sato, H., Yamanishi, Y., & McLeod, H. L., etal. (2001).
Allelotype frequency of the thiopurine methyltransferase (TPMT) gene in Japanese.
50. Evans, W. E. (2002). Comprehensive assessment of thiopurine S-methyltransferase (TPMT)
alleles in three ethnic populations. Journal of Pediatric Hematology/Oncology, 24, 335336.
51. Schaeffeler, E., Fischer, C., Brockmeier, D., Wernet, D., Moerike, K., & Eichelbaum, M.,
etal. (2004). Comprehensive analysis of thiopurine S-methyltransferase phenotype-genotype
correlation in a large population of German-Caucasians and identification of novel TPMT
variants. Pharmacogenetics, 14, 407417.
52. Otterness, D., Szumlanski, C., Lennard, L., Klemetsdal, B., Aarbakke, J., & Park-Hah, J. O.,
et al. (1997). Human thiopurine methyltransferase pharmacogenetics: gene sequence poly-
morphisms. Clinical Pharmacology and Therapeutics, 62, 6073.
53. Relling, M. V., Hancock, M. L., Boyett, J. M., Pui, C. H., & Evans, W. E. (1999). Prognostic
importance of 6-mercaptopurine dose intensity in acute lymphoblastic leukemia Blood, 93,
28172823.
54. Davies, H. A., Lennard, L., & Lilleyman, J. S. (1993). Variable mercaptopurine metabolism
in children with leukaemia: A problem of non-compliance? British Medical Journal (Clinical
Research Ed.), 306, 12391240.
55. Lennard, L., Welch, J., & Lilleyman, J. S. (1995). Intracellular metabolites of mercaptopurine
in children with lymphoblastic leukaemia: A possible indicator of non-compliance? British
Journal of Cancer, 72, 10041006.
56. Lennard, L., Keen, D., & Lilleyman, J. S. (1986). Oral 6-mercaptopurine in childhood leuke-
mia: parent drug pharmacokinetics and active metabolite concentrations. Clinical
57. Dervieux, T., Medard, Y., Verpillat, P., Guigonis, V., Duval, M., & Lescoeur, B., etal. (2001).
Possible implication of thiopurine S-methyltransferase in occurrence of infectious episodes
during maintenance therapy for childhood lymphoblastic leukemia with mercaptopurine.
Leukemia, 15, 17061712.
58. Lennard, L., Welch, J., & Lilleyman, J. S. (1996). Mercaptopurine in childhood leukaemia:
The effects of dose escalation on thioguanine nucleotide metabolites. British Journal of
Clinical Pharmacology, 42, 525527.
7 Thiopurines 113
59. McLeod, H. L., Relling, M. V., Liu, Q., Pui, C. H., & Evans, W. E. (1995). Polymorphic
thiopurine methyltransferase in erythrocytes is indicative of activity in leukemic blasts from
children with acute lymphoblastic leukemia. Blood, 85, 18971902.
60. Szumlanski, C. L., Honchel, R., Scott, M. C., & Weinshilboum, R. M. (1992). Human liver
thiopurine methyltransferase pharmacogenetics: Biochemical properties, liver-erythrocyte
correlation and presence of isozymes. Pharmacogenetics, 2, 148159.
61. Van Loon, J. A., Szumlanski, C. L., & Weinshilboum, R. M. (1992). Human kidney thiopurine
methyltransferase. Photoaffinity labeling with S-adenosyl-L-methionine. Biochemical
62. Van Loon, J. A., & Weinshilboum, R. M. (1982). Thiopurine methyltransferase biochemical
genetics: human lymphocyte activity. Biochemical Genetics, 20, 637658.
63. Lennard, L., Chew, T. S., & Lilleyman, J. S. (2001). Human thiopurine methyltransferase
activity varies with red blood cell age. Gut, 52, 539546.
64. Calvo, P. L., Canaparo, R., Baldi, M., Serpe, L., Giaccone, M., & Lerro, P., et al. (2008).
Impact of age in phenotype of thiopurine methyltransferase. Digestive and Liver Disease, 40,
A55A56.
65. McLeod, H. L., Krynetski, E. Y., Wilimas, J. A., & Evans, W. E. (1995). Higher activity of
polymorphic thiopurine S-methyltransferase in erythrocytes from neonates compared to
adults. Pharmacogenetics, 5, 281286.
66. Cheung, S. T., & Allan, R. N. (2003). Mistaken identity: misclassification of TPMT pheno-
type following blood transfusion. European Journal of Gastroenterology & Hepatology, 15,
12451247.
67. Stanulla, M., Schaeffeler, E., Flohr, T., Cario, G., Schrauder, A., & Zimmermann, M., etal.
(2005). Thiopurine methyltransferase (TPMT) genotype and early treatment response to mer-
captopurine in childhood acute lymphoblastic leukemia. The Journal of the American Medical
Association, 293, 14851489.
68. von Ahsen, N., Armstrong, V. W., & Oellerich, M. (2004). Rapid, long-range molecular hap-
lotyping of thiopurine S-methyltransferase (TPMT) *3A, *3B, and *3C. Clinical Chemistry,
50, 15281534.
69. Hamdan-Khalil, R., Gala, J. L., Allorge, D., Lo-Guidice, J. M., Horsmans, Y., & Houdret, N.,
etal. (2005). Identification and functional analysis of two rare allelic variants of the thiopurine
S-methyltransferase gene, TPMT*16 and TPMT*19. Biochemical Pharmacology, 69,
525529.
70. Hon, Y. Y., Fessing, M. Y., Pui, C. H., Relling, M. V., Krynetski, E. Y., & Evans, W. E. (1999).
Polymorphism of the thiopurine S-methyltransferase gene in African Americans. Human
Molecular Genetics, 8, 371376.
71. Ameyaw, M. M., Collie-Duguid, E. S., Powrie, R. H., Ofori-Adjei, D., & McLeod, H. L.
(1999). Thiopurine methyltransferase alleles in British and Ghanaian populations. Human
Molecular Genetics, 8, 367370.
72. Collie-Duguid, E. S., Pritchard, S. C., Powrie, R. H., Sludden, J., Li, T., & McLeod, H. L.
(1999). The frequency and distribution of thiopurine methyltransferase alleles in Caucasian
and Asian populations. Pharmacogenetics, 9, 3742.
73. Ando, M., Ando, Y., Hasegawa, Y., Sekido, Y., Shimokata, K., & Horibe, K. (2001). Genetic
polymorphisms of thiopurine S-methyltransferase and 6- mercaptopurine toxicity in Japanese
children with acute lymphoblastic leukaemia. Pharmacogenetics, 11, 269273.
74. Schutz, E., von Ahsen, N., & Oellerich, M. (2000). Genotyping of eight thiopurine methyl-
transferase mutations: three-color multiplexing, two-color/shared anchor, and fluorescence-
quenching hybridization probe assays based on thermodynamic nearest-neighbor probe
design. Clinical Chemistry, 46, 17281737.
75. McDonald, O. G., Krynetski, E. Y., & Evans, W. E. (2002). Molecular haplotyping of genomic
DNA for multiple single nucleotide polymorphisms located kilobases apart using long range
polymerase chain reaction and intramolecular ligation. Pharmacogenetics, 12, 9399.
76. Kaskas, B. A., Louis, E., Hindorf, U., Schaeffeler, E., Deflandre, J., & Graepler, F., et al.
(2003). Safe treatment of thiopurine S-methyltransferase deficient Crohns disease patients
with azathioprine. Gut, 52, 140142.
77. Colombel, J. F., Ferrari, N., Debuysere, H., Marteau, P., Gendre, J. P., & Bonaz, B., et al.
(2000). Genotypic analysis of thiopurine S-methyltransferase in patients with Crohns disease
and severe myelosuppression during azathioprine therapy. Gastroenterology, 118, 10251030.
78. Yang, J., Bogni, A., Cheng, C., Bleibel, W. K., Cai, X., & Fan, Y., etal. (2008). Etoposide
sensitivity does not predict MLL rearrangements or risk of therapy-related acute myeloid
leukemia. Clinical Pharmacology and Therapeutics, 84, 691697.
79. Relling, M. V., Hancock, M. L., Rivera, G. K., Sandlund, J. T., Ribeiro, R. C., & Krynetski,
E. Y., etal. (1999). Mercaptopurine therapy intolerance and heterozygosity at the thiopurine
S-methyltransferase gene locus. Journal of the National Cancer Institute, 91, 20012008.
80. Nygaard, U., Toft, N., & Schmiegelow, K. (2004). Methylated metabolites of 6-mercaptopu-
rine are associated with hepatotoxicity. Clinical Pharmacology and Therapeutics, 75(4),
274281.
81. Lennard, L., Richards, S., Cartwright, C. S., Mitchell, C., Lilleyman, J. S., & Vora, A., etal.
(2006). The thiopurine methyltransferase genetic polymorphism is associated with thiogua-
nine-related veno-occlusive disease of the liver in children with acute lymphoblastic leukemia.
82. Satti, M. B., Weinbren, K., & Gordon-Smith, E. C. (1982). 6-thioguanine as a cause of toxic
veno-occlusive disease of the liver. Journal of Clinical Pathology, 35, 10861091.
83. Gill, R. A., Onstad, G. R., Cardamone, J. M., Maneval, D. C., & Sumner, H. W. (1982). Hepatic
veno-occlusive disease caused by 6-thioguanine. Annals of Internal Medicine, 96, 5860.
84. Evans, W. E., Hon, Y. Y., Bomgaars, L., Coutre, S., Holdsworth, M., & Janco, R., etal. (2001).
Preponderance of thiopurine S-methyltransferase deficiency and heterozygosity among
patients intolerant to mercaptopurine or azathioprine. Journal of Clinical Oncology, 19,
22932301.
85. Duley, J. A., Marinaki, A. M., Arenas, M., & Florin, T. H. (2006). Do ITPA and TPMT geno-
types predict the development of side effects to AZA? Gut, 55, 1048; author reply 10481049.
86. Trevino, L. R., Shimasaki, N., Yang, W., Panetta, J. C., Cheng, C., & Pei, D., etal. (2009).
Germline genetic variation in an organic anion transporter polypeptide associated with metho-
trexate pharmacokinetics and clinical effects. Journal of Clinical Oncology.
87. Schmiegelow, K., & Bruunshuus, I. (1990). 6-Thioguanine nucleotide accumulation in red
blood cells during maintenance chemotherapy for childhood acute lymphoblastic leukemia,
and its relation to leukopenia. Cancer Chemotherapy and Pharmacology, 26, 288292.
88. Lennard, L., & Lilleyman, J. S. (1989). Variable mercaptopurine metabolism and treatment
outcome in childhood lymphoblastic leukemia. Journal of Clinical Oncology, 7, 18161823.
89. Lennard, L., & Lilleyman, J. S. (1996). Individualizing therapy with 6-mercaptopurine and
6-thioguanine related to the thiopurine methyltransferase genetic polymorphism. Therapeutic
Drug Monitoring, 18, 328334.
Part III
Specific Pharmacogenomic Targets:
Cardiovascular Drugs
Chapter 8
The Pharmacogenetics of Vitamin K
Antagonist Anticoagulation Drugs
Charles Eby
Abstract Since the first report on warfarin pharmacogenetics in 1999, genetic

v ariants have emerged as an important predictor of warfarin maintenance doses
before therapy is initiated, raising expectations of greatly improved clinical out-
comes. However, much of the information on warfarin sensitivity conveyed by
genetic variants is captured by early international normalized ratio values tradition-
ally used to guide dose titration. Thus, inclusion of early international normalized
ratios in prediction models reduces the contribution of genetics. Moreover, in large
population cohorts, genetics explained only 2030% of variance in warfarin doses.
Finally, even pharmacogenetic prediction models did not predict doses reliably in
the majority of at-risk patients with warfarin requirements at the low or high end of
the dose range. Currently, the clinical utility and costeffectiveness of pharmacoge-
netic based dosing are being assessed in large prospective trials in various settings.
In the interim, enthusiasm for warfarin pharmacogenetics should not supersede strict
adherence to traditional measures used to optimize coumarin anticoagulation.
Keywords Warfarin Vitamin K epoxide reductase complex Cytochrome P450 2C9
8.1Introduction
Until very recently, three derivatives of dicoumarol, warfarin, acenocoumarol, and

phenprocoumon, were the only available drugs for oral anticoagulation therapy.
While they differ in half-life and potency, all three coumarins have similar
pharmacogenetic, pharmacokinetic, and pharmacodynamic properties, and warfarin,
the drug of choice in most countries, will be the focus of this chapter. Warfarin
is effective for prevention of arterial and venous thromboses and emboli. Oral
C. Eby(*)
Washington University in St. Louis, 660 South Euclid Avenue, St. Louis, MO 63110, USA
e-mail: eby@pathology.wustl.edu
118 C. Eby
anticoagulationtherapy with warfarin reduces the relative risk of stroke in patients

with atrial fibrillation by 68% [1]. Extended warfarin therapy prevents recurrent
thromboses in patients with DVTs [2] and reduces DVT complications after hip
replacement surgery [3]. It is estimated that two million people start warfarin annu-
ally in the United States [4]. However, warfarin has a narrow therapeutic window
and unpredictable anticoagulant dose response, which is reflected in one of the
highest rates of reported adverse drug events [4].
8.2Pharmacology of Warfarin
Warfarins anticoagulant effect is due to inhibition of hepatic posttranslational

modifications to coagulation factors X, IX, VII, and II (prothrombin) (Fig. 8.1). Six
to thirteen glutamic acid residues near the N-terminus of these coagulation factors
undergo carboxylation of the gamma carbon atom, producing calcium ion binding
sites, increasing the affinity of the coagulation factors for negatively charged
phospholipid surfaces, and accelerating the rate of thrombin generation and formation
of fibrin clot. The gamma-glutamyl carboxylase (GGCX) enzyme requires a cofac-
tor, reduced vitamin K, which undergoes oxidation to vitamin K epoxide during
gamma carboxylation, and subsequent reduction by vitamin K epoxide reductase
(VKOR) to regenerate reduced vitamin K. Wafarin competitively inhibits VKOR
activity to deplete the availability of reduced vitamin K, diminish GGCX activity,
Fig. 8.1 Warfarin pharmacodynamics: Warfarin indirectly produces an anticoagulated state by

preventing posttranslational modifications to selected coagulation factors. The S-enantiomer
strongly blocks vitamin K oxide reductase (VKOR) conversion of vitamin K epoxide to reduced
vitamin K. Reduced vitamin K is a cofactor for g-carboxylation of glutamic acid residues on
factors X, IX, VII, and II (prothrombin). The Pharmacogenomics Journal 2004; 4, 224225,
reproduced with permission
8 The Pharmacogenetics of Vitamin K Antagonist Anticoagulation Drugs 119
and decrease the activity of circulating coagulation factors X, IX, VII, and II, thus
delaying the rate of blood clot formation [5].
Knowledge of warfarins unpredictable anticoagulant effect and bleeding risk
necessitated therapeutic monitoring with the prothrombin time (PT) clotting test
when warfarin was licensed in the United States in 1954 under the brand name
Coumadin. The PT initiates clotting of plasma by adding thromboplastin, a phos-
pholipid/protein extract from brain tissue. The rate of clot formation is dependent
upon the levels of coagulation factors V, X, VII, and II: the later three are vitamin K
dependent, and their activities are reduced by warfarin. A PT ratio (patient PT/mean
normal PT) of 1.52.0 produced acceptable efficacy and safety [6]. Over several
decades, clinical management and therapeutic monitoring of warfarin therapy
improved. Recognition of variable sensitivity of commercial thromboplastins to vita-
min K-dependent coagulation factor depletion led to standardization by calibrating
PT reagents to a World Health Organization reference thromboplastin to obtain an
International Sensitivity Index (ISI) factor, and subsequent conversion of PT ratios to
International Normalized Ratios (INR) (PT patient/PT mean normal)ISI [6]. For most
oral anticoagulation indications, an INR target range of 23 provides the best balance
of safety and efficacy [6]. However, to maintain a therapeutic INR, patients require
warfarin doses that vary approximately 20-fold. The graph in Fig8.2 shows the dis-
tribution of daily warfarin doses to obtain stable therapeutic INRs (23) in 5,701
patients from all over the world. The average therapeutic warfarin dose per week
(INR 23) also varies among racial and ethnic groups: people of African descent 40 mg,
Caucasians 31.5 mg, and East Asians 21 mg [7]. However, determining patients
Fig. 8.2 Distribution of therapeutic warfarin doses (target INR range of 23) from patients whose
data were pooled by the International Warfarin Pharmacogenetic Consortium investigators to derive
dosing algorithms [94]. Ref. [96], reproduced with permission
120 C. Eby
therapeutic warfarin dose by trial and error adjustments produces unpredictable

fluctuations in INRs, delays in obtaining a stable dose, and bleeding complications,
particularly during the first weeks and months of therapy [8].
During the last decade, discoveries of genetic variation impact on warfarin phar-
macokinetics and pharmacodynamics have been translated into improved predic-
tions of patients therapeutic warfarin dose. The results of preliminary clinical
applications of pharmacogenetic (PGx)-based warfarin dosing have produced a
range of responses from laboratory, practitioner, regulatory, and reimbursement
stakeholders, ranging from skepticism to conviction that genetic-guided personal-
ized medicine has found an ideal candidate in warfarin pharmacogenetics.
However, new oral anticoagulant drugs are poised to challenge warfarin [9].
Designed to inhibit the active form of coagulation factors X (direct factor Xa inhibi-
tors), or factor II (direct thrombin inhibitors), clinical trials with rivaroxaban and
dabigatran etexilate, respectively, have demonstrated similar efficacy and safety com-
pared to warfarin, while not requiring therapeutic drug monitoring. If approved by
regulatory agencies for prevention of thromboembolic events in patients, several
factors will likely effect how quickly and to what degree oral direct coagulation
factor inhibitors replace warfarin: their comparative costs and reimbursement rates,
convenience, and postmarketing efficacy and safety of fixed dosing without labora-
tory monitoring of anticoagulant effect. Continued utilization of warfarin would be
supported if ongoing clinical trials investigating PGx-based dosing demonstrate
improved safety.
8.3Pharmacogenetics of Warfarin
8.3.1CYP2C9
Warfarin consists of a racemic mixture of S and R enantomers. Both forms are

efficiently absorbed from the small intestines, circulate bound to albumin, and
are taken up by hepatocytes [5]. However, S-warfarin accounts for approximately
80% of the VKOR inhibitory effect. Therefore, a reduced rate of S-warfarin
metabolism has greater consequences than a similar change in R-warfarin pharmaco
kinetics. S-warfarin undergoes hydroxylation by cytochrome P450 (CYP) 2C9 and
eventual excretion in the bile [10]. R-warfarin metabolism is more complex, involv-
ing CYP1A1, CYP1A2, and CYP3A4 to produce an inactive alcohol which is
excreted in the urine [10]. The other coumarins, acenocoumarol and phenprocoumon,
are also racemic mixtures, with more potent S enantomers which undergo CYP2C9-
mediated metabolism. Despite some differences in half lives and metabolism, the
anticoagulant effect of all three coumarins is potentiated by single nucleotide poly-
morphisms which reduce CYP2C9 activity [11]. There are two notable CY2C9 single
nucleotide polymorphic (SNP) variant alleles: *2 (420C>T, exon 3, substituting
cysteine for arginine at amino acid 144) and *3 (1075A>C, exon 7, substituting leu-
cine for isoleucine at amino acid 359). Both invitro and invivo experiments confirm
Table 8.1 Variations in CYP2C9 *2/*3 and VKORC1 haplotype frequencies among different
populations. Adapted from Marsh etal. [13] with permission
EuropeanAmerican AfricanAmerican Asian
CYP2C9*2 0.14 0.02 0.0
CYP2C9*3 0.06 0.01 0.04
VKORC1 haplotype A group 0.42 0.21 0.85
VKORC1 haplotype B group 0.57 0.58 0.14
VKORC1 haplotype other 0.01 0.21 0.01
Low-dose group 0.55 0.22 0.86
Haplotype A and 2C9 variant 0.18 0.01 0.06
that CYP2C9*3 expression markedly reduces S-warfarin clearance, while CYP2C9*2

has a more modest impact [12]. The minor allelic frequencies of *2 and *3 SNPs are
~14 and 6%, respectively, in Caucasians [13] and much lower in African Americans
and Asians (Table 8.1). Retrospective cohort [14] and case control studies [15] associ-
ated lower therapeutic warfarin doses, delayed attainment of therapeutic doses, more
supratherapeutic INRs, and higher rates of major and minor bleeding complications
in patients with one or more *2 or *3 alleles.
8.3.2Vitamin K Epoxide Reductase Complex 1
In 2004, two groups independently discovered the gene for VKOR enzyme activity
[16, 17]. Located on chromosome 16, the VKOR Complex 1 (VKORC1) 11 kb
gene contains three exons and has no homology to other known genes. Rieder and
colleagues sequenced the VKORC1 gene, introns, and 5 and 3 flanking regions
from 186 Caucasians with known therapeutic warfarin dose requirements and iden-
tified ten common SNPs (frequency >5%) [18]. Based on linkage disequilibriums
between SNPs, Rieder inferred 9 VKORC1 haplotypes. Group A haplotypes (H1,
H2) were associated with lower therapeutic warfarin doses, and Group B haplo-
types (H7, H8,H9) were associated with higher warfarin doses, independent of
CYP2C9 *2/*3 status. Four informative SNPs were used to infer VKORC1 haplo-
types in an independent cohort of Caucasians with known therapeutic warfarin
doses, confirming a significant difference in maintenance warfarin dose: lowest for
haplotype AA, intermediate for haplotype AB, and highest for haplotype BB. In
addition, VKORC1 mRNA levels in liver tissue correlated with VKORC1 haplo-
type [18]. Subsequent investigations have shown that genotyping patients for one
of two SNPs in high linkage disequilibrium (1639G>A rs9923231 and 1173C/T
rs9934438) account for nearly identical percentages of warfarin dose variability and
can substitute for haplotype A [7]. In vitro expression experiments support SNP 1639
in the 5 promoter region as the likely functional SNP due to decreased gene tran-
scription [19], providing a mechanism for increased sensitivity to warfarin in
patients who inherit one or two 1639G>A alleles. Investigators have confirmed
the associations between VKORC1 haplotypes or tagSNPs and CYP2C9*2/*3
122 C. Eby
SNPs and therapeutic warfarin dose in various populations and clinical settings
[2023]. For example, patients who are extremely slow metabolizers (CYP2C9
*3/*3) and very sensitive to warfarin (VKORC1 1639AA) require therapeutic
warfarin doses in the range of 0.52.0 mg/day while patients whose genotype is
CYP2C9*1*1, VKORC1 1639GG require doses of 57 mg/day [24]. Collectively,
there is ample biochemical, molecular, and clinical evidence supporting a genetic
contribution to warfarin pharmacokinetic and pharmacodynamic interindividual
variation [25, 26]. Citing its mandate to promote personalized medicine and patient
safety [4], in August, 2007, the United States Food and Drug Administration (FDA)
revised the package insert for Coumadin to provide CYP2C9 and VKORC1 phar-
macogenetic information and to alert prescribers that patients with variations in
these genes may require lower doses compared to patients without them.
8.4Dosing Algorithms for Warfarin
8.4.1Initial Algorithms
By combining clinical and demographic data with CYP 2C9 and VKORC1 SNP
genotypes obtained from patients with known therapeutic warfarin doses, investigators
have determined the percent of warfarin dose variability attributable to different
variables and derived algorithms to predict the therapeutic dose for a warfarin-nave
patient initiating anticoagulation treatment [20, 22, 2734]. Despite heterogeneity
regarding sample size, therapeutic warfarin dose criteria, clinical, demographic,
and medication information ascertainment, statistical analysis, and subjects race
and ethnicity, combining clinical and pharmacogenetic data, i.e., PGx-based dosing
algorithms, accounts for 5060% of warfarin dosing variability. VKORC1 haplotype
or 1639G>A genotype consistently has the most effect, accounting for 2534%
[27, 28, 35] of dosing variability, followed by age, body size, CYP2C9*3, and
CYP2C9*2. Other minor, but statistically significant in some cohorts, variables
include amiodarone which inhibits CYP2C9 activity [27], smoking [27, 31], indica-
tion for anticoagulation [27, 31], INR target [27, 31], statin therapy [27], gender
[28], race [27], and enzyme-inducing drugs [28, 31]. However, most algorithms
were derived from small, homogeneous, predominantly Caucasian populations.
When applied to multiethnic [36] or African American patients [27, 37], these
models accounted for a lower percent of warfarin dose variability, although Wu and
colleagues reported comparable performances for five Caucasian-derived PGx
algorithms when applied to a multiethnic population in San Francisco [38].
8.4.2IWPC and Advanced Algorithms
To address these limitations, a group of investigators formed the International

Warfarin Pharmacogenetics Consortium (IWPC) to derive and validate a PGx
d osing algorithm based on clinical and pharmacogenetic data from 4,043 and 1,009
subjects, respectively, from many countries and continents and whose racial
makeup was 55% White, 30% Asian, 9% Black, and 6% mixed or unknown [39].
The IWPC PGx dosing algorithm included age, height, weight, VKORC1 1639
genotype, CYP2C9*2/*3 genotypes, Asian and African race, enzyme-inducing
drugs, and amiodarone and accounted for 47 and 43% of the dosing variability in
the derivation and validation cohorts, respectively.
To facilitate use of PGx genotyping and dosing algorithms, Gage and colleagues
created a nonprofit website: www.WarfarinDosing.org. A clinician can obtain an
estimated therapeutic warfarin dose for a patient based on clinical and demographic
information or a more accurate PGx-based dose estimate if CYP2C9*2/*3 and
VKORC1 1,639 genotypes are available. The PGx algorithm was derived by
Gages group at Washington University in St. Louis [27]. Alternatively, one can
select the IWPC PGx algorithm to calculate a dose; however, the two algorithms
provide nearly identical initial dosing estimates.
Including a patients INR response to the first few doses of warfarin improves the
accuracy of a PGx dosing algorithm. Based on data collected from patients prescribed
warfarin for VTE prophylaxis after hip and knee arthroplasties, Millican etal. derived
a dose refinement algorithm which explained 79% of the variability in therapeutic
warfarin dose [40] and included the following independent variables: INR after three
doses, first and second warfarin doses, postoperative blood loss, smoking, liver dis-
ease, CYP 2C9 *2/*3, and VKORC1 1639 SNPs. Coagulation factor levels fall with
peri-operative blood loss causing a temporary INR prolongation which is accounted
for in the algorithm. Slow metabolism SNPs (CYP 2C9 *2/*3) continued to have a
strong impact on the revised dose estimate, while warfarin sensitivity genotype
(VKORC1 1639G>A) decreased in importance since the INR response after three
doses incorporated most of patients warfarin sensitivity phenotype. Improved warfa-
rin dosing accuracy by applying a dose refinement PGx algorithm after 4 days of
anticoagulation therapy has been prospectively validated in orthopedic patients [41].
Investigators extended this approach further in a large, multicenter retrospective
cohort of patients starting warfarin for various indications by developing a PGx dose
refinement algorithm accounting for 4258% of therapeutic warfarin variability after
four or five doses, compared to 2643% for a clinical dose refinement algorithm [42].
Two important conclusions can be drawn from these studies. First, using a clinical
dosing algorithm for the first 34 warfarin doses followed by a PGx dose refinement
algorithm improves dosing accuracy and allows more time to perform genotyping.
Secondly, patients who have a genetic sensitivity to warfarin display it quickly in their
INR response while the effect of slow metabolizing genetic variants is delayed. In a
retrospective analysis of the participants in the PREVENT trial [43], Ferder et al.
determined the contribution of CYP2C9*2/*3 and VKORC1 1639G>A genotypes
to explaining therapeutic dose variation 0, 7, 14, and 21 days after starting warfarin
[44]. PREVENT was a prospective randomized trial comparing low intensity warfa-
rin therapy (INR 1.52.0) to placebo in patients with idiopathic venous thromboem-
bolic events who had completed 3 months of warfarin (INR 23) and who had
stopped anticoagulation therapy for 1 month. At enrollment, all patients took an
initial dose of 3 mg, which was adjusted, using a study nomogram, based on weekly
124 C. Eby
Fig. 8.3 Percentage of warfarin dose variability explained by PGx genotype (CYP2C9 *2/*3 and
VKORC11639G>A), clinical variables (age, body surface area, target INR, gender, race, smoking,
amiodarone, statin), most recent INR, and prior weeks average warfarin dose at weekly time
points. Ferder etal. [44], reproduced with permission
INRs until a therapeutic dose was achieved. Figure 8.3 shows the individual and total
contributions to explaining warfarin therapeutic dose variability for clinical data, PGx
genotype, INR, and prior warfarin dose history. Initially, PGx genotype accounted for
43% of the variability, declining to 12, 4, and 1.4% after 1, 2, and 3 weeks, respec-
tively, as warfarin dosing history captured more of the PGX genetics phenotype.
Similar findings were reported by Li etal. in a cohort who underwent more frequent
early INR monitoring with a therapeutic target of 23 [45]. We can conclude that PGx
testing can improve dosing accuracy, although with diminishing impact, for approxi-
mately a week after starting warfarin, while warfarin dosing history and INR response
steadily eclipse genetic information. And consequently, once a patients therapeutic
warfarin dose has been determined by trial and error empiric dosing, there is no
apparent utility in performing PGx testing.
8.4.3Limitations of Dosing Algorithms
Despite the advances in clinical research to date, approximately 50% of warfarin

dosing variability prior to starting therapy cannot be accounted for in Caucasians
and Asian, and even more variability is unexplained in African Americans [4648].
Potential sources of missing information include additional genetic variation, epi-
genetic factors, and patient behaviors including diet and compliance. Considerable
efforts have been directed at discovering polymorphisms in genes whose products

are involved in the vitamin K cycle and its inhibition, but the results have not been
dramatic. Comparisons of SNPs or other genetic variations in gamma-glutamyl
carboxylase (GGCX) [49, 50], calumenin, a GGCX regulatory gene [4951], vita-
min K-dependent clotting proteins (factors X, VII, II, and protein C [31, 49, 53,
54]), and CYP2C18 and CYP2C19 [49] to therapeutic warfarin doses have pro-
duced inconsistent, positive associations with very weak clinical impact, in small
retrospective studies. Other investigated candidate genes include apolipoprotein E,
ABCB1 (ATP-binding cassette transporter B1), EPHX1 (epoxide hydrolase 1
microsomal gene), and ORM1+2 (orosomucoid 1 and 2 genes) [49, 55]. Inconsistent
findings are likely due to differences in patient selection, SNP minor variant allele
frequencies, statistical methods, and therapeutic warfarin dose criteria. However,
when Wadelius and colleagues compared 183 SNPs in 29 candidate genes to thera-
peutic warfarin doses in 1,496 Swedes, only VKORC1 1639GA and 11173C>T,
and CYP2C9*2/*3 SNPs were significantly associated with warfarin dose [28].
Some SNPs do have a clinically important effect on warfarin dosing, but only in
ethnic and racial groups with high frequencies for the variant. For example, a
VKORC1 SNP, 5417G>T, replacing aspartic acid with tyrosine (D36Y), is associ-
ated with moderate warfarin resistance, but this SNP is only prevalent in Ashkenazi
(4%) and Ethiopian (14%) Jews [56, 57]. Inclusion of this mutation on a PGx panel
would improve warfarin dosing accuracy in communities including these ethnic
groups. Additional rare VKORC1 nonsynonymous nucleotide substitutions altering
amino acids in the cytoplasmic loop of VKOR have been discovered in patients
requiring very large maintenance warfarin doses [17, 58].
8.5Genome-Wide Association Studies
Investigators have performed several genome-wide association studies (GWAS) to

discover novel SNP candidates associated with therapeutic warfarin dose. Using a
drug metabolizing enzyme and transporter array chip, Caldwell and colleagues
identified a SNP in CYP4F2 (rs2108622; 1297G>A;V433M) in 951 Caucasians
with an allelic frequency of 30% which was associated with higher therapeutic
warfarin doses and accounted for 2% of warfarin dosing variation [59]. Subsequent
studies reported V433M minor allelic frequencies ranging from 30 to 45%, and the
contribution to explaining warfarin dosing variation ranged from no effect [60] or
17% in predominantly Caucasian samples [6164]. In vitro kinetic experiments
using recombinant supersomes expressing wild type or V433M CYP4F2 in liver
microsomes confirmed CYP4F2 V433M allele has a reduced capacity to metabo-
lize vitamin K which would produce higher hepatic vitamin K concentrations and
require higher warfarin doses to achieve an anticoagulant response [65]. A GWAS
using a chip with approximately 550 103 SNPs did not identify any SNPs other
than CYP2C9*2/*3 and known VKORC1 polymorphisms in a sample of 181
Caucasians [66]. Two larger GWAS projects did identify a few additional candidate
126 C. Eby
SNPs. Takeuchi and colleagues screened 1053 Swedes with a 326 103 SNP chip
and identified an association between CYP4F2 and warfarin dose, in addition to
CYP2C9 and VKORC1 SNP clusters [67]. Finally, Teichert and colleagues in
Rotterdam performed a GWAS with a 550 103 SNP chip on DNA from 1,451
patients who took acenocoumarol. They confirmed an association for CYP4F2
V433M and a SNP in CYP2C18 (rs1998591) with warfarin dose plus previously
noted SNP clusters in CYP2C9 and VKORC1 [68]. Based on these GWAS results,
it is unlikely that there are other undiscovered genetic variants in Caucasians with
the clinical impact of CYP2C9 *2/*3 and VKORC1 haplotype A. However, adding
SNPs which marginally improve dosing accuracy to PGx dosing algorithms is fea-
sible and has been done for CYP4F2 V433M at www.WarfarinDosing.org.
Improving the accuracy of warfarin PGx algorithms for patients of African
descent is a priority. While nongenetic factors may be involved, genetic differences
clearly are important. There are at least 12 common VKORC1 haplotypes in
populations of African descent compared to four in Caucasians. VKORC1 haplo-
type A, inferred from SNPs 1639G>A or 1173T>C, accounts for only 4.2% of
warfarin dosing variability in African Americans compared to 22.5% in Caucasians,
based on analysis of the IWPC data [7]. However, this is not due to other VKORC1
haplotypes or SNPs unique to African Americans, but to the low allelic frequency
of haplotype A in African Americans compared to Caucasians, since the impact of
a VKORC1 1639A SNP on lower warfarin dose requirement is the same for any
individual, regardless of race [7]. The impact of CYP2C9 *2/*3 SNPs on warfarin
dose in African American populations is negligible compared to Caucasians [69],
but this is also most likely due to the very low frequencies of these alleles in African
Americans [48]. On the other hand, there are several SNPs in CYP2C9 with higher
allele frequencies in African Americans compared to Caucasians which are associ-
ated with lower warfarin doses. CYP2C9*5 (1080C>G) substitutes glutamatic acid
for asparagine at amino acid 360 (G360R), which is next to the Ile359Leu substitu-
tion coded by CYP2C9*3, and markedly reduces S-warfarin metabolism [70].
CYP2C9*6 is a null mutation due to deletion of adenine at nucleotide 818 and is
associated with lower warfarin doses [71]. CYP2C9*11 introduces an arginine to
tryptophan substitution at amino acid position 335 (R335W) and is associated with
a 33% reduction in warfarin dose and invitro evidence of decreased enzyme stabil-
ity [72]. Recently, investigators reported an association between CYP2C9*8
(817A>G), coding for substitution of histidine for arginine at amino acid position
150 (R150H), and reduced warfarin dose [71, 73]. These four SNPs have frequen-
cies of <1% in Caucasians, while minor allele frequencies are higher in African
Americans: 0.71.5% for *5, 0.71.3% for *6, 4.76.5% for *8, and 1.31.8% for
*11 [46, 71, 73]. In a cohort of 226 African Americans, 52 (23%) had a CYP2C9
variant (*2, *3, *5, *6, *8, or *11), which accounted for 6% of warfarin dosing
variation, compared to 7% for VKORC1 1639G>A SNP [71]. Two commercial
genotype platforms, Autogenomics Infiniti and GenMark Dx (formerly Osmetech)
eSensor, currently offer extended warfarin PGx assays which include CYP2C9 *5,
*6, and *11. A GWAS with warfarin dose using DNA from people of African
descent is likely to identify additional genetic variants leading to more accurate
PGx dosing algorithms and more SNPs to add to genetic testing panels.
8.6Warfarin Genotype Testing Technologies
While PGx-based algorithms continue to evolve with the addition of more SNPs
and refinements based on early INR response, investigators have confirmed the
analytical validity of genotyping methods for the three core SNPs: CYP2C9
*2/*3 and VKORC1 1639G>A or 1173C>T. In their assessment of analytical
and clinical validity of warfarin PGx, McClain and colleagues summary of published
and unpublished genotyping methods showed analytical sensitivity and specificity
results of 100% for CYP2C9 *1, *2, and *3 genotype combinations when com-
pared to sequencing or PCR-RFLP reference methods [74]. Molecular diagnostic
manufacturers have responded to the interest in warfarin PGx, and to date, there
are five FDA 510K approved medical devices for CYP2C9 *2/*3 and VKOCR1
1639G>A or 1173C>T: Nanosphere (Verigene), Autogenomics (INFINITI),
GenMark Dx (eSENSOR), Paragon Dx reagents with Cepheid Smart Cycler,
and TrimGen reagents with Roche Light Cycler. Several groups have indepen-
dently confirmed commercial platforms analytical validity [7577]. Using
archived DNA from 112 previously genotyped patients, King and colleagues
evaluated INFINITI automated allele specific primer extension (ASPE) microar-
ray instrument, Invader cleavase-based fluorescence assay, and Luminex ASPE
and Tag-it bead hybridization platform. All methods were 100% accurate for
CYP2C9 *1,*2, and *3 genotypes. INFINITI was 100% accurate, and Invader
and Luminex 97% accurate for VKOCR1 SNP 1639G>A genotypes [75]. Babic
et al. compared INFINITI, eSENSOR, and ParagonDx reagents/Stratagene real
time PCR instrument platforms using 100 DNA samples. Once again, CYP2C9
*2 and *3 genotype concordance was 100%. VKORC1 SNP 1639G>A genotype
accuracy was 100 and 97% for eSENSOR and INFINITI instruments, respec-
tively, and 100% for VKORC1 SNP 1173C>T using Paragonx/Stratagene and
INFINITI platforms [76]. In addition to the FDA-approved warfarin PGx SNPs,
Autogenomics and GenMark offer partially overlapping extended genotype pan-
els for cytochrome P450 and VKORC1 SNPs. Babics comparison of the two
instruments CYP2C9 *5,*6, and *11 SNP genotypes was 100% concordant [76].
Only GenMarks eSENSOR extended panel includes CYP4F2 1297G>A SNP
(V433M), and compared to direct sequencing, eSENSOR CYP4F2 1297G>A
genotyping was 100% accurate.
While analytical accuracy is method independent, other factors such as technical
complexity, instrument and reagent costs, turnaround time (TAT), reliability, and
versatility for other molecular diagnostic testing will influence a laboratory direc-
tors decision when selecting a platform for wafarin PGx testing. While attention is
often focused on analytical TAT [75, 77], it is important to view this from a wider
perspective. First, molecular diagnostic laboratories are not staffed or organized to
perform STAT genotyping. At best, clinicians should expect warfarin PGx results
the next working day if testing is done locally, andwithin 23 days if performed at
a reference laboratory. All currently validated methods and platform TATs are 8 h
from DNA isolation to genotype results and would be adaptable to most laboratories
work flow. Complying with external proficiency requirements from accreditation
128 C. Eby
organizations can be challenging for molecular diagnostic laboratories, yet this

quality assurance indicator is an important component of analytical validity.
In2007, the College of American Pathologists addressed this need and introduced
a pharmacogenomics survey which includes CYP2C9*2/*3 and VKORC1 1639G>A
or 1173C>T SNPs.
8.7Clinical Trials on Pharmacogenetic Testing Efficacy
8.7.1Published Trials
While pharmacogenetic-based algorithms do improve warfarin dosing accuracy,

despite the previously mentioned shortcomings, this is insufficient evidence of
clinical utility to warrant routine use [78]. The next step in warfarin PGx research
has been prospective trials to determine if PGx-based warfarin dosing improves
clinically meaningful patient outcomes. Clearly, the most important outcomes are
bleeding and thromboembolic events. However, based on the low frequencies of
these events in previous trials of warfarin in patients with atrial fibrillation and
venous thromboembolic events [79], investigators have typically used INR response
as a surrogate endpoint for clinical utility. There is evidence to support associations
between bleeding complications and elevated INRs [79, 80], and thrombosis and
lower INRs [81].
To date, there are three completed, prospective randomized trials comparing
INR results between a PGx dosing group and a study-specific nomogram dosing
control group (Table 8.2). Major hemorrhages were uncommon among patients
enrolled in these studies. In the Caraco and colleagues trial, time to first therapeu-
tic INR, time to stable therapeutic dose, and percent of time INR was within thera-
peutic range (% INR TTR) significantly favored the PGx dosing arm even though
VKORC1 genotyping was not performed [82]. Subjects in the PGx arm with
CYP2C9 *1/*1 genotype (63%) were started on 1.25 the control warfarin dose,
which may account for the superior results in this group. Two other studies did not
demonstrate a difference in primary INR endpoints. Hillman performed CYP2C9
*2/*3 genotyping and started control subjects on 5 mg of warfarin [83], while the
Couma-Gen trial PGx algorithm included CYP2C9 and VKORC1 genotypes and
patients in both arms received loading warfarin doses on days 1 and 2 [32].
However, when outcomes in the Couma-Gen trial were analyzed based on subjects
genotype status, there was a significant reduction in out-of-target-range INRs in the
PGx-dosed subjects who were wild type or had more than one variant SNP. In addi-
tion, the difference between stable therapeutic warfarin dose and initial dose was
significantly smaller for the PGx arm compared to the control arm for subjects with
wild type or >1 variant SNP, which constituted 59% of the study population [32].
While these observations require independent confirmation, they suggest PGx-
based algorithms improve initial warfarin dosing accuracy in Caucasian patients
compared to a uniform, empiric starting dose by increasing the dose for wild type
Table 8.2 Summary of three small prospective, randomized trials of pharmacogenetic-based initial warfarin dosing
Initial Warfarin dosing INR time in therapeutic range (%)
Author (Reference) Subjects Genotyped SNPs PGx algorithm arm Control arm Genotype arm Control arm
Hillman etal. [83] 38 CYP2C9*2/*3 d.1: algorithm dose 5 mg 41.7 41.5
d.2: INR-based nomogram both arms
Caraco etal. [95] 191 CYP2C9*2/*3 d.1: *1/*1: 1.25 control dose. 5 mg 80.4 63.4
Proportionately lower doses for CYP2C9
variants based on % reduction in S-warfarin
metabolism
d.28: INR-based nomogram both arms
Anderson etal. [32] 206 CYP2C9*2/*3 d.l2: 2 algorithm dose 10 mg 49.8 51.9
VKORC1 1173CTad.34: algorithm dose 5 mg
d. 5: INR-based nomogram both arms
P < 0.001
a
In high linkage disequilibrium with VKORC1 promoter SNP 1639AG
8 The Pharmacogenetics of Vitamin K Antagonist Anticoagulation Drugs
129
130 C. Eby
patients and decreasing the dose for patients with multiple variant alleles, while
PGx-based algorithms do not improve dosing accuracy or % INR TTR for patients
with single variant SNPs. Therefore, all patients would undergo the cost of geno-
typing in order to identify a modest majority of patients who may potentially ben-
efit from the information.
8.7.2Regulatory and Reimbursement Issues for Warfarin

Pharmacogenomics
Based on the limited prospective PGx-based warfarin dosing data, it is not possible
to accurately estimate the economic cost-benefit of routine warfarin PGx genotyp-
ing [84], and clinical specialty societies [85] and laboratory science organizations
[86] advocate waiting for more definitive evidence. Reflecting this perspective, in
August, 2009, the Centers for Medicare and Medicaid Services (CMS) announced
it would not reimburse for warfarin pharmacogenetic testing of Medicare patients
except when done in a CMS approved randomized clinical trial setting. However,
in January, 2010, the FDA revised Coumadin labeling again, adding a table of
expected ranges for therapeutic warfarin dose based on CYP2C9 *2/*3 and
VKORC1 1639G>A SNP status [87] (Table 8.3). While the label change does not
explicitly require genetic testing, it does provide initial dosing guidelines for
patients with all combinations of these three SNPs, which may encourage more
clinicians to order warfarin PGx testing.
8.7.3Ongoing Clinical Trials
Meanwhile, three large multicenter trials are underway, comparing clinical

dosing algorithms to PGx dosing algorithms in patients starting a coumarin
anticoagulant: Clarification of Optimal Anticoagulation through Genetics (COAG)
trial (ClinicalTrials.gov: NCT00839157); European-Pharmacogenetics of
Anticoagulant Therapy (EU-PACT) trial [88], and the Genetics Informatics
Trial (GIFT) (ClinicalTrials.gov:NCT01006733)(Table 8.4). All three trials are
prospective, randomized, fully or partially blinded designs to minimize potential
Table 8.3 Range of expected therapeutic warfarin doses based on CYP2C9 and VKORC1
genotypes
CYP2C9
VKORC1 *1/*1 *1/*2 *1/*3 *2/*2 *2/*3 *3/*3
GG (mg) 57 57 34 34 34 0.52
AG (mg) 57 34 34 34 0.52 0.52
AA (mg) 34 34 0.52 0.52 0.52 0.52
Bristol-Myers Squib Coumadin prescribing information, revised January 2010
Table 8.4 Comparison of three large prospective randomized trials to evaluate efficacy of PGx
algorithms for initial warfarin dosing
COAG EU-PACT GIFT
Study population Newly dxd VTE or Newly dxd VTE or Hip or knee
AF AF arthroplasty DVT
prophylaxis
Recruitment target 1,238 2,955 1,600
Oral anticoagulant Warfarin Warfarin, Warfarin
acenocoumarol, or
phenprocoumon
Target INR 23 23 or 23.5 1.8 or 2.5
Design Double blinded Patient blinded Double blinded
Treatment arms Clinical algorithm vs. Clinical algorithm vs. Clinical algorithm vs.
PGx algorithm PGx algorithms PGx algorithm
Loading dose Day 1 Days 13 Days 12
Revision algorithm Yes Yes Yes
Primary outcome INR % TIR INR % TIR VTE, major bleed, death,
vascular death
Follow-up 4 weeks 3 months 46 weeks
biases and to determine whether adding CYP2C9 *2/*3 and VKORC1 1639G>A
genotype information to a clinical-based algorithm will affect clinically meaningful
outcomes. Despite recruitment goals of 1,238 and 2,955 for COAG and EU-PACT,
respectively, the primary endpoint will be % INR TTR rather than major hemor-
rhagic or thrombotic complications because the incidence of these events is low, and
in order to have statistical power to detect a difference for those outcomes, recruit-
ment targets would be many times higher and prohibitively expensive. Subjects
participating in GIFT are at high risk for DVTs after hip or knee arthroplasties, and
in this study, the efficacy of PGx-based warfarin dosing will in part be based on the
rates of both symptomatic and asymptomatic DVTs detected by Dopplar ultrasound
after 46 weeks of warfarin therapy. Despite numerous study design differences,
these trials use similar algorithms and collect uniform clinical, genetic, and outcome
data to permit analysis of pooled data in the future. However, it will be several years
from now before we can more accurately judge the efficacy and cost-effectiveness
of warfarin PGx based on the outcomes of COAG, EU-PACT, GIFT, and possibly
other prospective randomized trials.
In the interim, other factors will influence utilization of warfarin PGx tests
including new genetic discoveries leading to improved dosing accuracy, test costs,
laboratory charges, third party reimbursement patterns, recent FDA label changes,
and results from less well-controlled studies performed in general practice environ-
ments. For example, a Medco Research Institute and Mayo Clinical Laboratories
collaboration offered PGx testing to Medco insured outpatients beginning warfarin
and sent CYP2C9 *2/*3 and VKORC1 1639 SNP results with interpretive and
management guidelines to participating patients physician [89]. Compared to
2,688 historical controls from the same insured pool who started warfarin a year
earlier and were not offered PGx testing, 896 patients who underwent warfarin PGx
132 C. Eby
testing had a 28% reduction in hospitalization rate (25.5 v. 18.5%) and 27% reduction
in hospitalization rate for bleeding or clotting complications (8.1 v. 6.0%) during
the 6-month follow-up period after starting warfarin. While the real world man-
agement environment is a strength, there is potential for uncontrolled variables to
confound these results. Due to the logistical barriers of informed consent and
sample collection, the median time from patients starting warfarin to sending PGx
genotypes to their physicians was 32 days which would likely diminish the impact
of genetic information after 4 weeks of trial and error warfarin dosing. In addition,
the investigators did not compare INR results between the PGx and control popula-
tions or correlate INR control with hospitalization rates. Nevertheless, the prelimi-
nary findings are noteworthy and may convince more clinicians to order PGx
genetic testing instead of waiting several years for more definitive results from
prospective, randomized controlled trials.
8.8Pharmacologic Alternatives to Warfarin
For 56 years, there were no alternatives to warfarin for oral anticoagulation therapy,
until recently. Now, two oral direct coagulation factor inhibitors are poised to
challenge warfarin, and other compounds are in the pharmaceutical pipeline [9].
Rivaroxaban is a direct factor Xa inhibitor, and dabigatran etexilate is a direct
thrombin inhibitor. The large, prospective, randomized RE-LY trial compared war-
farin therapy (INR 23) and two doses of dabigatran in atrial fibrillation patients
and demonstrated similar rates of stroke and systemic emboli and lower major
bleeding rates for dabigatran 110 mg twice daily, and lower stroke and systemic
emboli rates with similar major bleeding rates for dabigatran 150 mg twice daily
[90]. The RE-COVER trial compared warfarin (INR 23) to dabigatran, 150 mg
twice daily in patients with acute VTEs, and demonstrated similar low rates of
recurrent thromboses and major bleeding [91]. In November 2010, the FDA
approved dabigatran 150 mg BID to prevent thromboembolic events in patients
with atrial fibrillation. In the RECORD series of randomized prospective trials,
rivaroxaban, 10 mg per day, was more effective than low molecular weight heparin
for DVT prophylaxis after hip or knee arthroplasty with similar bleeding complica-
tion rates [92]. Dabigatran and rivaroxaban are administered as fixed doses in
adults, do not require therapeutic monitoring of anticoagulant effect to ensure
efficacy and safety, and lack major pharmacogenetic or drug interactions [9]. The
safety and efficacy of direct inhibitors have not been investigated in patients with
prosthetic heart valves or children, or patients with moderate to severe renal insuf-
ficiency since both drugs are partially eliminated by the kidneys. It is anticipated
that direct coagulation factor inhibitors will be much more expensive than warfarin,
but the convenience of dispensing with periodic INR monitoring will make
them attractive to many patients and physicians. How warfarin and PGx testing
will be integrated into the approaching competitive oral anticoagulation drug era is
unclear.
8.9Case Report
The following description of our experience in managing initiation of warfarin ther-

apy demonstrates the challenges and successes one can achieve with PGx dosing
algorithms [93]. An internist referred a man to our anticoagulation service on a Friday
afternoon. The patient consented to participate in a PGx-based dosing algorithm trial,
and provided the following information: age 74, race white, weight 180 lbs, height
64, and medications amiodarone 400 mg and pravastatin. His baseline INR was 1.1,
and the target therapeutic INR was 23. Since CYP2C9 *2/*3 and VKORC1
1639G>A genotyping would not be performed until the following Monday, a phar-
macist entered the clinical and demographic data into an online dosing calculator at
WarfarinDosing.org and obtained a predicted dose of 3.0 mg/day, reflecting the nega-
tive impact of his advanced age and two medications on warfarin requirements.
Monday afternoon, the patient returned to clinic and after taking 3 mg of warfarin for
3 days his INR3 was 1.7. Using a trial and error approach to warfarin initiation,
most clinicians would continue the current dose, or increase it, and repeat an INR in
a few days. However, his genotype was CYP2C9 *3/*3 and VKORC1 1639AA,
indicating he was an extremely poor metabolizer of S-warfarin and very sensitive to
warfarin inhibition of VKOR. Adding the patients genotype, three doses of 3 mg and
INR3 of 1.7 to WarfarinDosing.org generated a revised estimated therapeutic warfa-
rin dose of 1.7 mg/day. With empiric dose adjustments during the subsequent 4
weeks, all INRs were therapeutic, there were no bleeding or embolic complications,
and the patients average warfarin dose was 0.7 mg/day. This case illustrates several
important features of PGx algorithm warfarin dosing. First, it is not necessary to have
warfarin PGx genotype results when determining an initial warfarin dose since the
impact of slow metabolizing genotypes on INR is delayed and can be incorporated
into a revised PGx algorithm dose estimate after three or 4 days of treatment.
Secondly, PGx algorithms do not always produce accurate estimates of a patients
eventual therapeutic dose (1.7 vs. 0.7 mg/day respectively in this case), but are a defi-
nite improvement over starting all patients on 5 or 10 mg/day followed by INR-based
trial and error dose adjustments. And finally, PGx dosing algorithms are an adjunct
to, not a substitute for, anticoagulation management expertise in order to safely and
effectively care for patients like the man in this case study.
References
1. Atrial Fibrillation Investigators. (1994). Risk factors for stroke and efficacy of antithrombotic
therapy in atrial fibrillation. Archives of Internal Medicine, 154, 14491457.
2. Lagerstedt, C. I., Olsson, C. G., Fagher, B. O., Oqvist, B. W., & Albrechtsson, U. (1985).
Need for long-term anticoagulant treatment in symptomatic calf-vein thrombosis. Lancet,
2(8454), 515518.
3. Prandoni, P., Bruchi, O., Sabbion, P., Tanduo, C., Scudeller, A., Sardella, C., et al. (2002).
Prolonged thromboprophylaxis with oral anticoagulants after total hip arthroplasty: A pro-
spective controlled randomized study. Archives of Internal Medicine, 162, 19661971.
134 C. Eby
4. Kim, M. J., Huang, S. M., Meyer, U. A., Rahman, A., & Lesko, L. J. (2009). A regulatory
science perspective on warfarin therapy: A pharmacogenetic opportunity. Journal of Clinical
5. Ansell, J., Hirsh, J., Poller, L., Bussey, H., Jacobson, A., & Hylek, E. (2004). The pharmacology
and management of the vitamin K antagonists: The Seventh ACCP Conference on
Antithrombotic and Thrombolytic Therapy. Chest, 126, 204S233S.
6. Ansell, J. E. (1993). Oral anticoagulant therapy 50 years later. Archives of Internal Medicine,
153, 586596.
7. Limdi, N. A., Wadelius, M., Cavallari, L., Eriksson, N., Crawford, D. C., Lee, M. T., et al.
(2010). Warfarin pharmacogenetics: A single VKORC1 polymorphism is predictive of dose
across three racial groups. Blood, 110, 38273834.
8. Linkins, L. A., Choi, P. T., & Douketis, J. D. (2003). Clinical impact of bleeding in patients
taking oral anticoagulant therapy for venous thromboembolism: A meta-analysis. Annals of
Internal Medicine, 139, 893900.
9. Garcia, D., Libby, E., & Crowther, M. A. (2010). The new oral anticoagulants. Blood, 7(115),
1520.
10. Kaminsky, L. S., & Zhang, Z. Y. (1997). Human P450 metabolism of warfarin. Pharmacology
& Therapeutics, 73, 6774.
11. Beinema, M., Brouwers, J. R., Schalekamp, T., & Wilffert, B. (2008). Pharmacogenetic
differences between warfarin, acenocoumarol and phenprocoumon. Thrombosis and
Haemostasis, 100, 10521057.
12. Takahashi, H., & Echizen, H. (2001). Pharmacogenetics of warfarin elimination and its clinical
implications. Clinical Pharmacokinetics, 40, 587603.
13. Marsh, S., King, C. R., Porche-Sorbet, R. M., Scott-Horton, T. J., & Eby, C. S. (2006). Population
variation in VKORC1 haplotype structure. Journal of Thrombosis and Haemostasis, 4, 473474.
14. Aithal, G. P., Day, C. P., Kesteven, P. J. L., & Daly, A. K. (1999). Association of polymor-
phisms in the cytochrome P450 CYP2C9 with warfarin dose requirement and risk of bleeding
complications. Lancet, 353(9154), 717719.
15. Higashi, M. K., Veenstra, D. L., Kondo, L. M., Wittkowsky, A. K., Srinouanprachanh, S. L.,
Farin, F. M., et al. (2002). Association between CYP2C9 genetic variants and anticoagulation-
related outcomes during warfarin therapy. The Journal of the American Medical Association,
287, 16901698.
16. Li, T., Chang, C. Y., Jin, D. Y., Lin, P. J., Khvorova, A., & Stafford, D. W. (2004). Identification
of the gene for vitamin K epoxide reductase. Nature, 427(6974), 541544.
17. Rost, S., Fregin, A., Ivaskevicius, V., Conzelmann, E., Hrtnagel, K., Pelz, H. J., et al. (2004).
Mutations in VKORC1 cause warfarin resistance and multiple coagulation factor deficiency
type 2. Nature, 427(6974), 537541.
18. Rieder, M. J., Reiner, A. P., Gage, B. F., Nickerson, D. A., Eby, C. S., McLeod, H. L., et al.
(2005). Effect of VKORC1 haplotypes on transcriptional regulation and warfarin dose.
19. Yuan, H. Y., Chen, J. J., Lee, M. T., Wung, J. C., Chen, Y. F., Charng, M. J., et al. (2005). A
novel functional VKORC1 promoter polymorphism is associated with inter-individual and
inter-ethnic differences in warfarin sensitivity. Human Molecular Genetics, 14, 17451751.
20. Sconce, E. A., Khan, T. I., Wynne, H. A., Avery, P., Monkhouse, L., King, B. P., et al. (2005).
The impact of CYP2C9 and VKORC1 genetic polymorphism and patient characteristics upon
warfarin dose requirements: Proposal for a new dosing regimen. Blood, 106, 23292333.
21. Wadelius, M., Chen, L. Y., Downes, K., Ghori, J., Hunt, S., Eriksson, N., et al. (2005).
Common VKORC1 and GGCX polymorphisms associated with warfarin dose. The
Pharmacogenomics Journal, 5, 262270.
22. Roper, N., Storer, B., Bona, R., & Fang, M. (2010). Validation and comparison of pharmaco-
genetics-based warfarin dosing algorithms for application of pharmacogenetic testing. The
Journal of Molecular Diagnostics, 12, 283291.
23. Limdi, N. A., Arnett, D. K., Goldstein, J. A., Beasley, T. M., McGwin, G., Adler, B. K., et al. (2008).
Influence of CYP2C9 and VKORC1 on warfarin dose, anticoagulation attainment and maintenance
among European-Americans and African-Americans. Pharmacogenenomics, 9, 511526.
24. Moyer, T. P., OKane, D. J., Baudhuin, L. M., Wiley, C. L., Fortini, A., Fisher, P. K., et al.
(2009). Warfarin sensitivity genotyping: A review of the literature and summary of patient
experience. Mayo Clinic Proceedings, 84, 10791094.
25. Wu, A. H. (2009). Pharmacogenomic testing for warfarin dosing: We are ready now. Expert
26. Wadelius, M. (2009). Point: Use of pharmacogenetics in guiding treatment with warfarin.
Clinical Chemistry, 55, 709711.
27. Gage, B. F., Eby, C., Johnson, J. A., Deych, E., Rieder, M. J., Ridker, P. M., et al. (2008). Use
of pharmacogenetic and clinical factors to predict the therapeutic dose of warfarin. Clinical
28. Wadelius, M., Chen, L. Y., Lindh, J. D., Eriksson, N., Ghori, M. J. R., Bumpstead, D., et al.
(2009). The largest prospective warfarin-treated cohort supports genetic forecasting. Blood,
113, 784792.
29. Zhu, Y., Shennan, M., Reynolds, K. K., Johnson, N. A., Herrnberger, M. R., Valdes, R., Jr.,
et al. (2007). Estimation of warfarin maintenance dose based on VKORC1 (1639 G>A) and
CYP2C9 genotypes. Clinical Chemistry, 53, 11991205.
30. Tham, L. S., Goh, B. C., Nafziger, A., Guo, J. Y., Wang, L. Z., Soong, R., et al. (2006). A
warfarin-dosing model in Asians that uses single-nucleotide polymorphisms in vitamin K
epoxide reductase complex and cytochrome P450 2C9. Clinical Pharmacology and
31. Aquilante, C. L., Langaee, T. Y., Lopez, L. M., Yarandi, H. N., Tromberg, J. S., Mohuczy, D.,
et al. (2006). Influence of coagulation factor, vitamin K epoxide reductase complex subunit 1,
and cytochrome P450 2C9 gene polymorphisms on warfarin dose requirements. Clinical
32. Anderson, J. L., Horne, B. D., Stevens, S. M., Grove, A. S., Barton, S., Nicholas, Z. P., et al.
(2007). Randomized trial of genotype-guided versus standard warfarin dosing in patients
initiating oral anticoagulation. Circulation, 116, 25632570.
33. Herman, D., Peternel, P., Stegnar, M., Breskvar, K., & Dolzan, V. (2006). The influence of
sequence variations in factor VII, gamma-glutamyl carboxylase and vitamin K epoxide reductase
complex genes on warfarin dose requirement. Thrombosis and Haemostasis, 95, 782787.
34. Takahashi, H., Wilkinson, G. R., Nutescu, E. A., Morita, T., Ritchie, M. D., Scordo, M. G.,
etal. (2006). Different contributions of polymorphisms in VKORC1 and CYP2C9 to intra-
and inter-population differences in maintenance dose of warfarin in Japanese, Caucasians and
African-Americans. Pharmacogenetics and Genomics, 16, 101110.
35. Meckley, L. M., Wittkowsky, A. K., Rieder, M. J., Rettie, A. E., & Veenstra, D. L. (2008). An
analysis of the relative effects of VKORC1 and CYP2C9 variants on anticoagulation related
outcomes in warfarin-treated patients. Thrombosis and Haemostasis, 100, 229239.
36. Lubitz, S. A., Scott, S. A., Rothlauf, E. B., Agarwal, A., Peter, I., Doheny, D., et al. (2010).
Comparative performance of gene-based warfarin dosing algorithms in a multiethnic popula-
tion. Journal of Thrombosis and Haemostasis, 8, 10181026.
37. Schelleman, H., Chen, J., Chen, Z., Christie, J., Newcomb, C. W., Brensinger, C. M., et al.
(2008). Dosing algorithms to predict warfarin maintenance dose in Caucasians and African
Americans. Clinical Pharmacology and Therapeutics, 84, 332339.
38. Wu, A. H., Wang, P., Smith, A., Haller, C., Drake, K., Linder, M., et al. (2008). Dosing algo-
rithm for warfarin using CYP2C9 and VKORC1 genotyping from a multi-ethnic population:
Comparison with other equations. Pharmacogenomics, 9, 169178.
39. Klein, T. E., Altman, R. B., Eriksson, N., Gage, B. F., Kimmel, S. E., Lee, M. T., et al. (2009).
Estimation of the warfarin dose with clinical and pharmacogenetic data. The New England
40. Millican, E. A., Lenzini, P. A., Milligan, P. E., Grosso, L., Eby, C., Deych, E., et al. (2007).
Genetic-based dosing in orthopaedic patients beginning warfarin therapy. Blood, 110,
15111515.
41. Lenzini, P. A., Grice, G. R., Milligan, P. E., Dowd, M. B., Subherwal, S., Deych, E., et al. (2008).
Laboratory and clinical outcomes of pharmacogenetic vs. clinical protocols for warfarin initia-
tion in orthopedic patients. Journal of Thrombosis and Haemostasis, 6, 16551662.
136 C. Eby
42. Lenzini, P., Wadelius, M., Kimmel, S., Anderson, J. L., Jorgensen, A. L., Pirmohamed, M.,
et al. (2010). Integration of genetic, clinical, and INR data to refine warfarin dosing. Clinical
43. Ridker, P. M., Goldhaber, S. Z., Danielson, E., Rosenberg, Y., Eby, C. S., Deitcher, S. R.,
et al. (2003). Long-term, low-intensity warfarin therapy for the prevention of recurrent venous
thromboembolism. The New England Journal of Medicine, 348, 14251434.
44. Ferder, N. S., Eby, C. S., Deych, E., Harris, J. K., Ridker, P. M., Milligan, P. E., et al. (2010).
Ability of VKORC1 and CYP2C9 to predict therapeutic warfarin dose during the initial weeks
of therapy. Journal of Thrombosis and Haemostasis, 8, 95100.
45. Li, C., Schwarz, U. I., Ritchie, M. D., Roden, D. M., Stein, C. M., & Kurnik, D. (2009).
Relative contribution of CYP2C9 and VKORC1 genotypes and early INR response to the
prediction of warfarin sensitivity during initiation of therapy. Blood, 113, 39253930.
46. Limdi, N. A., Beasley, T. M., Crowley, M. R., Goldstein, J. A., Rieder, M. J., Flockhart, D. A.,
et al. (2008). VKORC1 polymorphisms, haplotypes and haplotype groups on warfarin dose
among African-Americans and European-Americans. Pharmacogenomics, 9, 14451458.
47. Schelleman, H., Chen, Z., Kealey, C., Whitehead, A. S., Christie, J., Price, M., et al. (2007).
Warfarin response and vitamin K epoxide reductase complex 1 in African Americans and
Caucasians. Clinical Pharmacology and Therapeutics, 81, 742747.
48. Schelleman, H., Limdi, N. A., & Kimmel, S. E. (2008). Ethnic differences in warfarin mainte-
nance dose requirement and its relationship with genetics. Pharmacogenomics, 9, 13311346.
49. Wadelius, M., Chen, L. Y., Eriksson, N., Bumpstead, S., Ghori, J., Wadelius, C., et al. (2007).
Association of warfarin dose with genes involved in its action and metabolism. Human
Genetics, 121, 2334.
50. Rieder, M. J., Reiner, A. P., & Rettie, A. E. (2007). Gamma-glutamyl carboxylase (GGCX)
tagSNPs have limited utility for predicting warfarin maintenance dose. Journal of Thrombosis
and Haemostasis, 5, 22272234.
51. Voora, D., Koboldt, D. C., King, C. R., Lenzini, P. A., Eby, C. S., Porche-Sorbet, R., et al.
(2010). A polymorphism in the VKORC1 regulator calumenin predicts higher warfarin dose
requirements in African Americans. Clinical Pharmacology and Therapeutics, 87, 445451.
52. Gonzalez-Conejero, R., Roldan, V., & Ferrer, F. (2007). The genetic interaction between
VKORC1 c1173t and calumenin a29809g modulates the anticoagulant response of acenocou-
marol. Journal of Thrombosis and Haemostasis, 5, 17011706.
53. Shikata, E., Ieiri, I., Ishiguro, S., Aono, H., Inoue, K., Koide, T., et al. (2004). Association of
pharmacokinetic (CYP2C9) and pharmacodynamic (vitamin K-dependent protein-Factors II,
VII, IX, and X, proteins S and C, and {gamma}-glutamyl carboxylase) gene variants with
warfarin sensitivity. Blood, 103, 26302635.
54. DAmbrosio, R. L., DAndrea, G., Cappucci, F., Chetta, M., Di Perna, P., Brancaccio, V., etal.
(2004). Polymorphisms in factor II and factor VII genes modulate oral anticoagulation with
warfarin. Haematologica, 89, 15101516.
55. Loebstein, R., Vecsler, M., Kurnik, D., Austerweil, N., Gak, E., Halkin, H., et al. (2005). Common
genetic variants of microsomal epoxide hydrolase affect warfarin dose requirements beyond the
effect of cytochrome P450 2C9. Clinical Pharmacology and Therapeutics, 77, 365372.
56. Loebstein, R., Dvoskin, I., Halkin, H., Vecsler, M., Lubetsky, A., Rechavi, G., et al. (2007). A
coding VKORC1 Asp36Tyr polymorphism predisposes to warfarin resistance. Blood, 109,
24772480.
57. Scott, S., Edelmann, L., Kornreich, R., & Desnick, R. (2008). Warfarin pharmacogenetics:
CYP2C9 and VKORC1 genotypes predict different sensitivity and resistance frequencies in
the Ashkenazi and Sephardi Jewish populations. American Journal of Human Genetics, 82,
495500.
58. Harrington, D. J., Gorska, R., Wheeler, R., Davidson, S., Murden, S., Morse, C., et al. (2008).
Pharmacodynamic resistance to warfarin is associated with nucleotide substitutions in
VKORC1. Journal of Thrombosis and Haemostasis, 6, 16631670.
59. Caldwell, M. D., Awad, T., Johnson, J. A., Gage, B. F., Falkowski, M., Gardina, P., et al.
(2008). CYP4F2 genetic variant alters required warfarin dose. Blood, 111, 41064112.
60. Zhang, J. E., Jorgensen, A. L., Alfirevic, A., Williamson, P. R., Toh, C. H., Park, B. K., et al.
(2009). Effects of CYP4F2 genetic polymorphisms and haplotypes on clinical outcomes in
patients initiated on warfarin therapy. Pharmacogenetics and Genomics, 19, 781789.
61. Pautas, E., Moreau, C., Gouin-Thibault, I., Golmard, J. L., Mah, I., Legendre, C., et al. (2010).
Genetic factors (VKORC1, CYP2C9, EPHX1, and CYP4F2) are predictor variables for warfarin
response in very elderly, frail inpatients. Clinical Pharmacology and Therapeutics, 87, 5764.
62. Prez-Andreu, V., Roldn, V., Antn, A. I., Garca-Barber, N., Corral, J., Vicente, V., et al.
(2009). Pharmacogenetic relevance of CYP4F2 V433M polymorphism on acenocoumarol
therapy. Blood, 113, 49774979.
63. Borgiani, P., Ciccacci, C., Forte, V., Sirianni, E., Novelli, L., Bramanti, P., et al. (2009).
CYP4F2 genetic variant (rs2108622) significantly contributes to warfarin dosing variability in
the Italian population. Pharmacogenomics, 10, 261266.
64. Perini, J. A., Struchiner, C. J., Silva-Assuncao, E., & Suarez-Kurtz, G. (2010). Impact of
CYP4F2 rs2108622 on the stable warfarin dose in an admixed patient cohort. Clinical
65. McDonald, M. G., Rieder, M. J., Nakano, M., Hsia, C. K., & Rettie, A. E. (2009). CYP4F2 is
a vitamin K1 oxidase: An explanation for altered warfarin dose in carriers of the V433M
variant. Molecular Pharmacology, 75, 13371346.
66. Cooper, G. M., Johnson, J. A., Langaee, T. Y., Feng, H., Stanaway, I. B., Schwarz, U. I., et al.
(2008). A genome-wide scan for common genetic variants with a large influence on warfarin
maintenance dose. Blood, 112, 10221027.
67. Takeuchi, F., McGinnis, R., Bourgeois, S., Barnes, C., Eriksson, N., Soranzo, N., et al. (2009).
A genome-wide association study confirms VKORC1, CYP2C9, and CYP4F2 as principal
genetic determinants of warfarin dose. PLoS Genetics, 5, e1000433.
68. Teichert, M., Eijgelsheim, M., Rivadeneira, F., Uitterlinden, A. G., van Schaik, R. H.,
Hofman, A., et al. (2009). A genome-wide association study of acenocoumarol maintenance
dosage. Human Molecular Genetics, 18, 37583768.
69. Limdi, N., Goldstein, J., Blaisdell, J., Beasley, T., Rivers, C., & Acton, R. (2007). Influence
of CYP2C9 genotype on warfarin dose among African American and European Americans.
Personalized Medicine, 4, 157169.
70. Dickmann, L. J., Rettie, A. E., Kneller, M. B., Kim, R. B., Wood, A. J., Stein, C. M., et al.
(2001). Identification and functional characterization of a new CYP2C9 variant (CYP2C9*5)
expressed among African Americans. Molecular Pharmacology, 60, 382387.
71. Cavallari, L. H., Langaee, T. Y., Momary, K. M., Shapiro, N. L., Nutescu, E. A., Coty, W. A.,
et al. (2010). Genetic and clinical predictors of warfarin dose requirements in African
Americans. Clinical Pharmacology and Therapeutics, 87, 459464.
72. Tai, G., Farin, F., Rieder, M. J., Dreisbach, A. W., Veenstra, D. L., Verlinde, C. L., et al.
(2005). In-vitro and in-vivo effects of the CYP2C9*11 polymorphism on warfarin metabolism
and dose. Pharmacogenetics and Genomics, 15, 475481.
73. Scott, S. A., Jaremko, M., Lubitz, S. A., Kornreich, R., Halperin, J. L., & Desnick, R. J.
(2009). CYP2C9*8 is prevalent among African-Americans: Implications for pharmacogenetic
dosing. Pharmacogenomics, 10, 12431255.
74. McClain, M. R., Palomaki, G. E., Piper, M., & Haddow, J. E. (2008). A rapid-ACCE review
of CYP2C9 and VKORC1 alleles testing to inform warfarin dosing in adults at elevated risk
for thrombotic events to avoid serious bleeding. Genetics in Medicine, 10, 8998.
75. King, C. R., Porche-Sorbet, R. M., Gage, B. F., Ridker, P. M., Renaud, Y., Phillips, M. S.,
et al. (2008). Performance of commercial platforms for rapid genotyping of polymorphisms
affecting warfarin dose. American Journal of Clinical Pathology, 129, 876883.
76. Babic, N., Haverfield, E. V., Burrus, J. A., Lozada, A., Das, S., & Yeo, K. T. (2009).
Comparison of performance of three commercial platforms for warfarin sensitivity genotyping.
Clinica Chimica Acta; International Journal of Clinical Chemistry, 406, 143147.
77. Langley, M. R., Booker, J. K., Evans, J. P., McLeod, H. L., & Weck, K. E. (2009). Validation
of clinical testing for warfarin sensitivity: Comparison of CYP2C9-VKORC1 genotyping
assays and warfarin-dosing algorithms. The Journal of Molecular Diagnostics, 11, 216225.
138 C. Eby
78. Eby, C. S. (2009). Counterpoint: Pharmacogenetic-based initial dosing of warfarin: Not ready
for prime time. Clinical Chemistry, 55, 712714.
79. Landefeld, S. C., & Beyth, R. (1993). Anticoagulant-related bleeding: Clinical epidemiology,
prediction and prevention. The American Journal of Medicine, 95, 315328.
80. Long, A. L., Bendz, L., Horvath, M. M., Cozart, H., Eckstrand, J., Whitehurst, J., et al. (2010).
Characteristics of ambulatory anticoagulant adverse drug events: A descriptive study.
Thrombosis Journal, 8, 5.
81. Kearon, C., Ginsberg, J. S., Kovacs, M. J., Anderson, D. R., Wells, P., Julian, J. A., et al.
(2003). Comparison of low-intensity warfarin therapy with conventional-intensity warfarin
therapy for long-term prevention of recurrent venous thromboembolism. The New England
82. Caraco, Y., Blotnick, S., & Muszkat, M. (2007). CYP2C9 genotype-guided warfarin prescri
bing enhances the efficacy and safety of anticoagulation: A prospective randomized controlled
study. Clinical Pharmacology and Therapeutics, 83, 460470.
83. Hillman, M. A., Wilke, R. A., Yale, S. H., Vidaillet, H. J., Caldwell, M. D., Glurich, I., et al.
(2005). A prospective, randomized pilot trial of model-based warfarin dose initiation using
CYP2C9 genotype and clinical data. Clinical Medicine & Research, 3, 137145.
84. Eckman, M. H., Rosand, J., Greenberg, S. M., & Gage, B. F. (2009). Cost-effectiveness of
using pharmacogenetic information in warfarin dosing for patients with nonvalvular atrial
fibrillation. Annals of Internal Medicine, 150, 7383.
85. Ansell, J., Hirsch, J., Hylek, E., Jacobsn, A., Crowther, M., & Palaretti, G. (2008).
Pharmacology and management of the vitamin K antagonists. Chest, 133, 160S198S.
86. Flockhart, D. A., OKane, D., Williams, M. S., Watson, M. S., Flockhart, D. A., Gage, B.,
et al. (2008). Pharmacogenetic testing of CYP2C9 and VKORC1 alleles for warfarin. Genetics
in Medicine, 10, 139150.
87. Safety Labeling Changes Approved By FDA Center for Drug Evaluation and Research
(CDER). (2010, January). Safety. Retrieved March 10, 2010, from http://www.fda.gov/Safety/
MedWatch/SafetyInformation/ucm201100.htm
88. van Schie, R. M., Wadelius, M. I., Kamali, F., Daly, A. K., Manolopoulos, V. G., de Boer, A.,
et al. (2009). Genotype-guided dosing of coumarin derivatives: The European pharmacogenet-
ics of anticoagulant therapy (EU-PACT) trial design. Pharmacogenomics, 10, 16871695.
89. Epstein, R. S., Moyer, T. P., Aubert, R. E., Kane DJ, O., Xia, F., Verbrugge, R. R., et al.
(2010). Warfarin genotyping reduces hospitalization rates results from the MM-WES (Medco-
Mayo Warfarin Effectiveness Study). Journal of the American College of Cardiology, 55,
28042812.
90. Connolly, S. J., Ezekowitz, M. D., Yusuf, S., Eikelboom, J., Oldgren, J., Parekh, A., et al.
(2009). Dabigatran versus warfarin in patients with atrial fibrillation. The New England
91. Schulman, S., Kearon, C., Kakkar, A. K., Mismetti, P., Schellong, S., Eriksson, H., et al.
(2009). Dabigatran versus warfarin in the treatment of acute venous thromboembolism. The
92. Eriksson, B. I., Kakkar, A. K., Turpie, A. G., Gent, M., Bandel, T. J., Homering, M., et al. (2009).
Oral rivaroxaban for the prevention of symptomatic venous thromboembolism after elective hip
and knee replacement. The Journal of Bone and Joint Surgery. British Volume, 91, 636644.
93. Grice, G. R., Milligan, P. E., Eby, C., & Gage, B. F. (2008). Pharmacogenetic-based dose
refinement prevents warfarin overdose in a patient who is highly warfarin sensitive. Journal
of Thrombosis and Haemostasis, 6, 207209.
94. Klein, T. E., Altman, R. B., Eriksson, N., Gage, B. F., Kimmel, S. E., Lee, M. T., et al. (2009).
Estimation of the warfarin dose with clinical and pharmacogenetic data. The New England
95. Caraco, Y., Blotnick, S., & Muszkat, M. (2008). CYP2C9 genotype-guided warfarin prescrib-
ing enhances the efficacy and safety of anticoagulation: A prospective randomized controlled
study. Clinical Pharmacology and Therapeutics, 83, 460470.
96. Sterling, J. A. (2010). Hospital pharmacy pulse-recent publications on medications and phar-
macy. Hospital Pharmacy, 45(2), 167175.
Chapter 9
Clopidogrel and Salicylates
Janice Y. Chyou and Marc S. Sabatine
Keywords Antiplatelet therapy Cardiovascular disease Clopidogrel
Antiplatelet therapy is at the cornerstone of coronary disease management and is

also indicated for long-term management of peripheral vascular and cerebrovascular
diseases. Currently, the most widely used antiplatelet medications are aspirin and
clopidogrel with a combination of these two medications as dual therapy for
patients undergoing percutaneous coronary intervention with stent implantation.
9.1Clopidogrel
9.1.1Pharmacology and Clinical Efficacy
9.1.1.1Mechanism of Action
Clopidogrel is an oral thienopyridine. It is a prodrug, 85% of which is metabolized

by esterases to form an inactive carboxylic acid derivative; the remaining 15% is
activated by hepatic metabolism by the cytochrome 450 system [1]. The active thiol
metabolite irreversibly binds the P2Y12 component of the adenosine diphosphate
(ADP) receptor, leading to inhibition of ADP-dependent platelet activation and
aggregation. Peak plasma concentrations of the active metabolite can be reached
after several hours [13], with increased peak metabolite concentration and platelet
inhibition when a single loading dose of 600 vs. 300 mg is administered [4].
Inhibition of platelet aggregation and activation lasts for the lifetime of the platelet,
and recovery of platelet function requires approximately 5 days off therapy [5].
J.Y. Chyou(*)
Department of Medicine, Brigham and Womens Hospital,
Harvard Medical School, Boston, MA, USA
e-mail: jchyou@partners.org
140 J.Y. Chyou and M.S. Sabatine
9.1.1.2Clinical Efficacy
The addition of clopidogrel to aspirin as dual antiplatelet therapy has been shown
to prevent death and ischemic complications in patients with acute coronary
syndrome (ACS) [68] as well as in patients undergoing percutaneous coronary
interventions (PCI) [6, 7, 9]. In the management of stable coronary artery disease,
the addition of clopidogrel to aspirin in patient for primary prevention of ACS is
not indicated [10], but subgroup analysis suggests that patients with a prior
myocardial infarction may benefit from dual clopidogrel plus aspirin therapy for
secondary prevention of ACS [11].
In regard to stroke prevention, dual antiplatelet therapy combining aspirin and
clopidogrel is an effective (though inferior) alternative to warfarin for primary
stroke prevention in patients with atrial fibrillation who are indicated but not
candidates for warfarin therapy [12, 13]. In the secondary prevention of stroke,
clopidogrel monotherapy has been shown to be as effective as aspirin plus clopi-
dogrel [14] and as effective as Aggrenox (dipyridamole plus aspirin) [15], and
that clopidogrel monotherapy carries less bleeding risk than combination
therapies [14, 15].
In the management of mixed atherothrombosis (defined as recent ischemic
stroke, recent myocardial infarction, or symptomatic peripheral arterial disease),
clopidogrel monotherapy (at dose of 75mg daily) was modestly more effective than
aspirin (325mg daily) in reducing the combined risk of ischemic stroke, myocardial
infarction, or vascular death without significant difference in bleeding events [16].
9.1.1.3Dosing and Duration of Therapy
The PCI-CURE and PCI-CLARITY studies demonstrated a significant reduction in

ischemic events in patients who were pretreated with clopidogrel before PCI versus
those who were not [6, 7] The optimal loading and maintenance dose of clopidogrel
is a subject of active research. The CURRENT-OASIS 7 trial [17] reported that a
loading dose of 600mg with post-PCI maintenance dose of 150mg for the first
week followed by 75mg daily thereafter reduced stent thrombosis and CV events
(primarily nonfatal MI) compared to a loading dose of 300mg followed by main-
tenance dose of 75mg daily. The high-dose clopidogrel arm was associated with a
modest increase in CURRENT-defined major bleeds, but with no increase in TIMI
major bleeds, intracranial hemorrhage, fatal bleeds, or CABG-related bleeds.
The duration of dual antiplatelet therapy is another area of active research.
Currently, the ACC/AHA guideline recommends dual antiplatelet therapy with
aspirin and clopidogrel for at least 1 month for those with bare metal stent implan-
tation and for 12 months for those with drug-eluting stent (DES) implantation
[18]. The DES-LATE trial examined the benefit and risks of extending dual anti-
platelet therapy beyond 12 months, but was underpowered to draw definitive
conclusions [19]. The ongoing DAPT trial [20] will provide further insights into
the benefits and safety of long-term dual antiplatelet therapy after DES
implantation.
9 Clopidogrel and Salicylates 141
9.1.2Variable Response, DrugDrug Interactions,

and Pharmacogenetics of Clopidogrel
Despite established clinical efficacy, variability in clopidogrel response has been

observed: inhibition of platelet aggregation with clopidogrel therapy approximates a bell-
shaped distribution and individuals with the least degree of platelet inhibition have higher
rates of ischemic events including stent thrombosis [21]. These observations have led to
investigations of clopidogrels absorption, metabolism, and action pathway.
Investigators have examined drugs such as statins and proton-pump inhibitors
(PPIs) that are competitive substrates for or inhibitors of key CYP450 enzymes in
clopidogrel metabolism (Fig. 9.1) [22]. Statins are predominantly metabolized by
CYP3A4. Concurrent statin use was found to attenuate platelet inhibition by clopi-
dogrel in a dose-dependent manner but not to increase major cardiovascular events
[2325]. PPIs inhibit the CYP2C19 enzyme crucial for clopidogrel metabolism
[26]. Observational data initially raised concerns about the PPI omeprazole
decreasing clopidogrels clinical efficacy [27]. A subsequent analysis of large clini-
cal trial database noted concomitant PPI and clopidogrel to be associated with a
trend toward a modest attenuation of platelet aggregation inhibition, but not to be
associated with adverse clinical outcomes [28]. Data from the prospective, random-
ized-controlled COGENT trial reported no difference in the risk of cardiovascular
events or ischemic complications and a benefit in reduction of gastrointestinal
bleeding in patients concomitantly taking PPI and clopidogrel [29].
Fig. 9.1 Clopidogrel absorption, metabolism, and action pathway, adapted from Simon et al.
NEJM [22] and used with permission
Investigators have also examined genes encoding the CYP450 enzymes, which are
highly polymorphic with certain alleles conferring greatly reduced enzymatic function
[30]. CYP2C19 is involved in both steps of the CYP450 conversion of clopidogrel to
its active metabolite. Subgroup analyses of clinical trial and registry databases [31, 32]
and a genome-wide association study (GWAS) [33] have identified CYP2C19 poly-
morphism to be independently associated with diminished inhibition of ADP-induced
platelet aggregation upon clopidogrel administration. Of the reduced-function allelic
variants of the CYP2C19 enzyme (including *2, *3, *4, *5), the *2 variant is the most
common [31] and is carried by approximately 30% of Whites [31], 40% Blacks, and
55% of East Asians [34]. The *2 variant rs4244285 involves a single base-pair muta-
tion of G->A at position 681 [35] which creates an aberrant splice site, leading to the
synthesis of truncated nonfunctional CYP2C19 protein [36].
Compared to noncarriers, carriers of at least one copy of CYP2C19*2 allele have
approximately 30% lower levels of active clopidogrel metabolite and approximately
25% relatively less platelet inhibition [31]. Moreover, among patients with ACS and
planned PCI, carriers of at least one copy of CYP2C19*2 variant had a 50% increase
in the risk of cardiovascular death, MI, or stroke and a threefold increased risk of
stent thrombosis [31]. These observations have been seen in a total of nine clinical
studies presented or published to date [3133, 37, 38]. Based on the totality of the
data to date, both heterozygotes and homozygotes appear to be at increased risk.
On the other hand, the CYP2C19*17 allelic variant, involving a single base-pair
mutation of C->T at position 808, has been linked to increased transcriptional
activityof the CYP2C19 enzyme, leading to extensive clopidogrel metabolism with
enhanced production of active metabolites and subsequently exaggerated platelet
response to clopidogrel therapy. CYP2C19*17 variant carrier status has been asso-
ciated with greater inhibition of ADP-induced platelet aggregation, increased risk
of bleeding, but without significant influence on stent thrombosis [39].
Polymorphism of other genes involved in the clopidogrel absorption, metabolism,
and action pathway (Fig. 9.1) has also been explored. Specifically, polymorphism of
the ABCB1 gene (encoding for P-glycoprotein involved in intestinal absorption of
clopidogrel) [31], the CYP3A4 gene (encoding for eponymous protein involved in
clopidogrel metabolism) [40], and the P2RY12 gene (encoding for the receptor for
active metabolite of clopidogrel) [41] has generated interests. However, only
CYP2C19 polymorphism has been consistently replicated in studies and was the only
significant polymorphism noted in a GWAS investigating gene variants that influence
clopidogrel response [33].
9.1.3Therapeutic Implications of Pharmacogenomics

of Clopidogrel and Novel Antiplatelet Agents
The replicable associations between CYP2C19 polymorphism and decreased platelet

inhibition as well as worsened cardiovascular outcomes have led the U.S. Food and
Drug Administration (FDA) and the manufacturer to revise the clopidogrel prescribing
information in May 2009 to include mentions of the CYP2C19 genetic polymorphism.

Limited evidence is available to guide potential therapeutic modifications for individu-
als with CYP2C19*2 allele. Potential therapeutic modifications considered include
escalation of clopidogrel dosage or switching to an alternate agent.
While earlier studies comparing clopidogrel loading dose of 300, 600, and
900mg in patients undergoing PCI found no improved benefit in administration of
loading dose higher than 600mg [4], data from a study incorporating CYP2C19
polymorphism genotyping suggest that higher clopidogrel loading and maintenance
dose (of 1,200 mg as loading and 150 mg daily as maintenance) may improve
platelet inhibition in carriers of reduced-function CYP2C19 alleles [42].
Switching to an alternate antiplatelet agent may also be a logical modification for
carriers of CYP2C19*2 alleles indicated for clopidogrel therapy for ACS and PCI.
Prasugrel is a third-generation thienopyridine that also confers antiplatelet activity by
binding to the platelet P2Y12 receptor [43]. Like clopidogrel, it is an irreversible inhibi-
tor. Unlike clopidogrel, it achieves a much higher degree of platelet inhibition. The
clinical efficacy of prasugrel for ACS and PCI was established in the TRITON-TIMI
38 trial which found prasugrel to have superior efficacy compared with clopidogrel in
reducing all-cause mortality or vascular complications including stent thrombosis,
albeit with an increased risk of bleeding[44]. Prasugrel was approved for use in PCI for
ACS by the FDA in July 2009. Retrospective subgroup analysis of the TRITON-TIMI
38 trial found common functional CYP polymorphisms (including CYP2C19 polymor-
phism) not to affect active drug metabolism levels, platelet aggregation inhibition, or
clinical cardiovascular event rates in individuals treated with prasugrel [45]. The dif-
ferential roles of esterases and CYP metabolism in the activation of clopidogrel and
prasugrel likely mediate this differential impact of CYP polymorphism. Whereas
esterases shunt the majority of clopidogrel to a dead-end inactive pathway with the
remaining prodrugs requiring a 2-step CYP-dependent oxidation to produce active
clopidogrel metabolites, esterases are part of the activation pathway of prasugrel and
prasugrel is oxidized to its active metabolite in a single CYP-dependent step [45].
Another potential antiplatelet alternative, not yet approved by the FDA, is
ticagrelor, an oral reversible direct antagonist of platelet ADP PGY12. The PLATO
trial reported ticagrelor to be superior to clopidogrel (loading dosing 300600mg
with 75 mg daily maintenance dose) in reduction of vascular death, myocardial
infarction, or stroke, but with an increase in the rate of non-CABG-related bleeding
in ACS [46]. Ticagrelor is an active compound and not a prodrug. However, it is
metabolized to inactive compounds by CYP3A4/5 and the impact of variants in
those genes on safety and efficacy remains undefined.
9.1.4Pharmacogenetic Testing
Should individuals with planned clopidogrel therapy undergo pharmacogenetic test-

ing to guide therapeutic management, given the significantly increased morbidity
and mortality in CYP2C19*2 carriers? The most recent Plavix boxed warning
includes Tests are available to identify a patients CYP2C19 genotype and can be
used as an aid in determining therapeutic strategy. Consider alternative treatment or
treatment strategies in patients identified as CYP2C19 poor metabolizers [47].
Currently, pharmacogenetic testing in patients being treated with clopidogrel is not
part of the standard of care. Although genetic testing for CYP2C19 can be done in a
clinical pathology laboratory, the turnaround time for such testing would likely exceed
the length of stay for most cardiology patients, thereby precluding integration of such
information during the loading phase of clopidogrel therapy. To that end, several com-
panies are now working on a point-of-care genotyping platform using whole blood.
If and when genetic data did become available, it remains unclear how clinicians
would use such information. Small studies have suggested that CYP2C19 reduced-
function allele carriers have a greater response to increased loading and maintenance
doses of clopidogrel than do wildtype individuals [42,48]. However, the specific doses
of clopidogrel needed to achieve bioequivalence to the platelet inhibition seen with a
300-mg load and 75mg daily in wildtype individuals remains to be defined. Moreover,
some data suggest that it may require prasugrel rather than higher doses of clopidogrel
to reliably achieve adequate platelet inhibition in CYP2C19*2 carriers [49].
The relative clinical value of genetic testing and platelet function testing
remainsto be defined; some studies suggest that they offer complementary predic-
tive values [50]. Lastly, how such testing would be reimbursed by insurance
companies isunknown. In summary, prospective clinical trials are needed to further
evaluate if testing indeed improves outcome, if pharmacogenetic testing is superior
to platelet function testing, and if testing of all-comers versus only the high-risk
population is most feasible. These questions will need to be answered before phar-
macogenetic testing of CYP2C19 polymorphism can be recommended as part of
the standard of care for clopidogrel therapy.
9.2Salicylates
9.2.1Pharmacology and Clinical Efficacy
9.2.1.1Mechanism of Action
Aspirin irreversibly inhibits cyclooxygenase-1 and 2 (COX-1 and 2) enzymes. COX

enzymes catalyze the conversion of arachidonic acid to PGG2 and PGH2, resulting in
downstream synthesis of prostanoids and TXA2. Therefore, administration of aspirin
decreases downstream production of PGD2, PGE2, PGF2, PGI2, and TXA2. Key to
aspirins clinical utility as an antiplatelet agent is the observation that inhibition of
TXA2 predominates, whereas inhibition of PGI2 appears to be clinically irrelevant
[51]. Together, this balance contributes to aspirins net antithrombotic effects primar-
ily by inhibition of TXA2-mediated platelet aggregation which leads to not only its
clinical utility in prevention of thrombosis, but also its adverse effects of bleeding.
9.2.1.2Metabolism, Pharmacodynamics, and Pharmacokinetics
Absorption of enterally administered aspirin is rapid, with bioavailability of about

50%, reaching peak serum level in 12h and lasting 46h, and with dose-dependent
half-life of elimination of salicylates [51, 52]. Enterally ingested aspirin is activated
by hydrolysis of gastric esterase. In ACS, patients are often instructed to chew the
aspirin, allowing for sublingual absorption with subsequent activation via esterase
hydrolysis in the bloodstream and earlier peaking of serum salicylate levels.
Metabolism of salicylate occurs primarily by hepatic glucoronidation; the metabo-
lite of aspirin is ultimately renally excreted [53].
9.2.1.3Clinical Efficacy and Dosing
Aspirins broad array of clinical use can be divided into four general categories:
vascular-antiplatelet, anti-inflammatory, analgesic, and anti-pyretic. Cardiovascular
indications of aspirin as an antiplatelet agent are well-established in the treatment
[5460] and secondary prevention [61] of ACS. Aspirin therapy is paramount in the
setting of coronary revascularization with clinical efficacy in the peri- and postpro-
cedural management of PCI with [6, 7, 9, 44, 62] or without [63, 64] stent implanta-
tion and in patients who have undergone coronary artery bypass surgery [6567].
Although the use of aspirin for primary prevention of myocardial infarction is com-
mon, clinical efficacy has not been consistently demonstrated. Data from primary
prevention trials demonstrate efficacy in reduction in risk of myocardial infarction
in patients with chronic stable angina [68, 69]. Without selecting for patients with
chronic stable angina, aspirin primary prevention trials yielded gender-specific
results; while combined analyses of the Physicians Health Study and the British
Physicians Study noted reduction in myocardial infarction events in men [70, 71],
the Womens Health Study found aspirin to lower the risk of stroke without
affectingthe risk of myocardial infarction in women [72]. While controversial, low-
dose aspirin was found to be inefficacious in primary prevention of atherosclerotic
events in patients with type-2 diabetes [73].
In primary prevention of cardioembolic stroke in patients with atrial fibrillation,
high-dose aspirin alone is reasonable for lone atrial fibrillation [74] and low-dose
aspirin in combination with clopidogrel has been found efficacious for patients with
atrial fibrillation with moderate stroke risk who are not candidates for warfarin
therapy [12]. Aspirin is also indicated in the management of acute ischemic stroke
[75, 76], transient ischemic attack [7679], peripheral artery disease [77, 80], and
in peri- and postprocedural management of percutaneous stents to peripheral ves-
sels [81]. Additionally, peripheral artery disease patients on aspirin therapy were
found to have reduction in stroke event rates [80].
Therapeutic indications capitalizing on aspirins anti-inflammatory effects
include very high-dose therapy for pericarditis and acute rheumatic fever. Although
aspirin also has analgesic and antipyretic effects, its uses as the principal agent in
these areas are more limited. Due to potential association with Reyes syndrome,
aspirin should not be used in children for viral infections [82].
9.2.1.4Variable Response, Pharmacogenetics,

and Pharmacogenomics of Aspirin
Variability in aspirin response has been noted [8387] with potential associa-
tions with worsened cardiovascular outcomes [85]. Notable contributors include
noncompliance [88], drugdrug interaction especially with nonsteroidal anti-
inflammatory drugs [89], female gender [90], and diabetes mellitus [40, 91].
Genetic polymorphisms may play a role in aspirin response.
The PIA polymorphism involves a Leu (PIA1) to Pro (PIA2) substitution at position 33
of the GP IIIa subunit of the platelet GP IIb/IIIa receptor, which has been most com-
monly studied [92]. However, studies have yielded conflicting results as to whether there
is an association between PIA2 allele and impaired aspirin response [93] or not [94].
More recently, a large candidate gene study found SNP rs2768759 (located
10kb upstream of the platelet endothelial aggregation receptor-1 (PEAR1) gene) to
be associated with decreased inhibition of platelet aggregation with aspirin therapy.
Given that rs2768759 is quite far upstream from the PEAR1gene, the observed
association with rs2768759 may reflect the effect of a SNP in the PEAR1 gene in
linkage disequilibrium with rs2768759 or that rs2768759 may confer the observed
effects by being part of the immediately upstream NTRK1 (neurotrophic tyrosine
kinase receptor type 1) instead [95].
A subsequent genomic array study noted SNPs in PEAR1 to be associated with
increased platelet response to collagen-related peptide and that PEAR1 protein
expression increased after platelet degranulation. Neither rs2768759 (which was
felt to be part of the NTRK1 gene by this genomic array study) nor SNPs in linkage
disequilibrium with rs2768759 were included in the genotyping of the genomic
array study; subsequent genotype of rs2768759 by TaqMan revealed no association
with platelet response [96]. Thus, while genetic influences of aspirin response are
likely present, large well-replicated studies are needed to establish potential genetic
determinants of aspirin variability.
9.2.1.5Therapeutic Implications and Pharmacogenetic Testing
Until further large-scale studies are able to demonstrate reproducible links between
specific genetic determinants and biochemical evidence of aspirin variability and
effects on clinical endpoints, no clear therapeutic implications or strategies for
pharmacogenetic testing can be made based on current data.
9.3Case Study
A 45-year-old man with diabetes mellitus, hypertension, and dyslipidemia presented

with substernal chest pressure radiating to left jaw 3 months after implantation of a
DES to the circumflex artery for non-ST-segment elevation myocardial infarction.
Patient endorses compliance with dual antiplatelet therapy of 75 mg clopidogrel

daily and aspirin 81mg daily (after switching from 325mg daily 2 months ago).
The patient also takes glyburide, metoprolol, atorvastatin, and omeprazole, the last
of which was prescribed years ago upon diagnosis of gastritis. Posterior EKG found
patient to have ST segment elevation in posterior leads. Prompt cardiac catheteriza-
tion revealed stent thrombosis of the previously placed DES, which was well
deployed. Thrombus aspiration was performed and recanalization was achieved by
placement of a new DES.
In addition to providing the current standard of care for ST-segment elevation
myocardial infarction [97], should this patient undergo clopidogrel pharmaco-
genetic testing, and how should we proceed with this patients medication
regimen?
Given that the patient presented with stent thrombosis despite compliance
with clopidogrel therapy and despite a well-deployed stent, empirically increas-
ing the maintenance dose of clopidogrel to 150mg daily would be reasonable,
as would changing to prasugrel. Such recommendations would hold regardless
of the patients genotype, given the patients clinical presentation. More prob-
lematic is whether all patients should undergo genetic (or platelet function)
testing prospectively at the time of initiation of ADP receptor blockade therapy.
For now, such a recommendation cannot be made. Data from additional studies,
some of which are ongoing, will be needed to define the benefits of such an
approach.
References
1. Caplain, H., Donat, F., Gaud, C., etal. (1999). Pharmacokinetics of clopidogrel. Seminars in
Thrombosis and Hemostasis, 25, 2528.
2. Taubert, D., Kastrati, A., Harlfinger, S., etal. (2004). Pharmacokinetics of clopidogrel after
administration of a high loading dose. Thrombosis and Haemostasis, 92, 311316.
3. Thebault, J. J., Kieffer, G., & Cariou, R. (1999). Single-dose pharmacodynamics of clopidogrel.
Seminars in Thrombosis and Hemostasis, 25, 38.
4. von Beckerath, N., Taubert, D., Pogatsa-Murray, G., etal. (2005). Absorption, metaboliza-
tion, and antiplatelet effects of 300-, 600-, and 900-mg loading doses of clopidogrel:
Results of the ISAR-CHOICE (Intracoronary Stenting and Antithrombotic Regimen:
Choose Between 3 High Oral Doses for Immediate Clopidogrel Effect) Trial. Circulation,
112, 29462950.
5. Raju, N. C., Eikelboom, J. W., & Hirsh, J. (2008). Platelet ADP-receptor antagonists for car-
diovascular disease: Past, present and future. Nature Clinical Practice Cardiovascular
6. Sabatine, M. S., Cannon, C. P., Gibson, C. M., et al. (2005). Addition of clopidogrel to aspirin
and fibrinolytic therapy for myocardial infarction with ST-segment elevation. The New
7. Yusuf, S., Zhao, F., Mehta, S. R., et al. (2001). Effects of clopidogrel in addition to aspirin in
patients with acute coronary syndromes without ST-segment elevation. The New England Journal
of Medicine, 345, 494502.
8. Chen, Z. M., Jiang, L. X., Chen, Y. P, et al. (2005). Addition of clopidogrel to aspirin in
45,852 patients with acute myocardial infarction: randomised placebo-controlled trial.
Lancet, 366, 16071621.
9. Steinhubl, S. R., Berger, P. B., Mann, J. T., 3rd, etal. (2002). Early and sustained dual oral
antiplatelet therapy following percutaneous coronary intervention: A randomized controlled
trial. Journal of the American Medical Association, 288, 24112420.
10. Bhatt, D. L., Fox, K. A., Hacke, W., etal. (2006). Clopidogrel and aspirin versus aspirin alone
for the prevention of atherothrombotic events. The New England Journal of Medicine, 354,
17061717.
11. Bhatt, D. L., Flather, M. D., Hacke, W., etal. (2007). Patients with prior myocardial infarction,
stroke, or symptomatic peripheral arterial disease in the CHARISMA trial. Journal of the
American College of Cardiology, 49, 19821988.
12. ACTIVE Investigators, Connolly, S. J., Pogue, J., etal. (2009). Effect of clopidogrel added to
aspirin in patients with atrial fibrillation. The New England Journal of Medicine, 360,
20662078.
13. ACTIVE Writing Group of the ACTIVE Investigators, Connolly, S., Pogue, J., etal. (2006).
Clopidogrel plus aspirin versus oral anticoagulation for atrial fibrillation in the Atrial fibrilla-
tion Clopidogrel Trial with Irbesartan for prevention of Vascular Events (ACTIVE W):
A randomised controlled trial. Lancet, 367, 19031912.
14. Diener, H. C., Bogousslavsky, J., Brass, L. M., etal. (2004). Aspirin and clopidogrel com-
pared with clopidogrel alone after recent ischaemic stroke or transient ischaemic attack in
high-risk patients (MATCH): Randomised, double-blind, placebo-controlled trial. Lancet,
364, 331317.
15. Sacco, R. L., Diener, H. C., Yusuf, S., etal. (2008). Aspirin and extended-release dipyrida-
mole versus clopidogrel for recurrent stroke. The New England Journal of Medicine, 359,
12381251.
16. Committee CAPRIE Steering. (1996). A randomised, blinded, trial of clopidogrel versus
aspirin in patients at risk of ischaemic events (CAPRIE). Lancet, 348, 13291339.
17. CURRENT-OASIS 7 Investigators, Mehta, S.R., Bassand, J.P., et al. (2010). Dose compari-
sons of clopidogrel and aspirin in acute coronary syndromes. The New England Journal of
Medicine, 363, 930942.
18. King, S. B., 3rd, Smith, S. C., Jr, Hirshfeld, J. W., Jr, etal. (2008). 2007 Focused Update of
the ACC/AHA/SCAI 2005 Guideline Update for Percutaneous Coronary Intervention.
Journal of the American College Cardiology, 51, 172209.
19. Park, S. J., Park, D. W., Kim, Y. H., etal. (2010). Duration of dual antiplatelet therapy after
implantation of drug-eluting stents. The New England Journal of Medicine,
362(15):13741382.
20. Grines, C. L., Bonow, R. O., Casey, D. E., Jr, etal. (2007). Prevention of premature discon-
tinuation of dual antiplatelet therapy in patients with coronary artery stents: A science advi-
sory from the American Heart Association, American College of Cardiology, Society for
Cardiovascular Angiography and Interventions, American College of Surgeons, and American
Dental Association, with representation from the American College of Physicians. Journal of
the American College of Cardiology, 48, 734739.
21. Gurbel, P. A. & Tantry, U. S. (2006). Drug insight: Clopidogrel nonresponsiveness. Nature
Clinical Practice. Cardiovascular Medicine, 3, 387395.
22. Simon, T., Verstuyft, C., Mary-Krause, M., etal. (2009). Genetic determinants of response to
clopidogrel and cardiovascular events. The New England Journal of Medicine, 360, 363375.
23. Saw, J., Brennan, D. M., Steinhubl, S. R., etal. (2007). Lack of evidence of a clopidogrel-
statin interaction in the CHARISMA trial. Journal of the American College of Cardiology, 50,
291295.
24. Saw, J., Steinhubl, S. R., Berger, P. B., etal. (2003). Lack of adverse clopidogrel-atorvastatin
clinical interaction from secondary analysis of a randomized, placebo-controlled clopidogrel
trial. Circulation, 108, 921924.
25. Gorchakova, O., von Beckerath, N., Gawaz, M., etal. (2004). Antiplatelet effects of a 600 mg
loading dose of clopidogrel are not attenuated in patients receiving atorvastatin or simvastatin
for at least 4 weeks prior to coronary artery stenting. European Heart Journal, 25,
18981902.
26. Gilard, M., Arnaud, B., Cornily, J. C., etal. (2008). Influence of omeprazole on the antiplate-
let action of clopidogrel associated with aspirin: The randomized, double-blind OCLA
(Omeprazole CLopidogrel Aspirin) study. Journal of the American College of Cardiology, 51,
256260.
27. Ho, P. M., Maddox, T. M., Wang, L., etal. (2009). Risk of adverse outcomes associated with
concomitant use of clopidogrel and proton pump inhibitors following acute coronary
syndrome. Journal of the American Medical Association, 301, 937944.
28. ODonoghue, M. L., Braunwald, E., Antman, E. M., etal. (2009). Pharmacodynamic effect
and clinical efficacy of clopidogrel and prasugrel with or without a proton-pump inhibitor: An
analysis of two randomised trials. Lancet, 374, 989997.
29. Bhatt, D. L., Cryer, B. L., Contant, C. F., et al. (2010). Clopidogrel with or without
Omeprazole in Coronary Artery Disease. The New England Journal of Medicine, Oct 6.
[Epub ahead of print]
30. Ingelman-Sundberg, M., Sim, S. C., Gomez, A., etal. (2007). Influence of cytochrome P450
polymorphisms on drug therapies: Pharmacogenetic, pharmacoepigenetic and clinical aspects.
Pharmacology and Therapy, 116, 496526.
31. Mega, J. L., Simon, T., Anderson, J. L., et al. (2009). CYP2C19 genetic variants and
clinical outcomes with clopidogrel: A collaborative meta-analysis. Circulation, 120,
S598S599.
32. Collet, J. P., Hulot, J. S., Pena, A., etal. (2009). Cytochrome P450 2C19 polymorphism in
young patients treated with clopidogrel after myocardial infarction: A cohort study. Lancet,
373, 309317.
33. Shuldiner, A. R., OConnell, J. R., Bliden, K. P., et al. (2009). Association of cytochrome
P450 2C19 genotype with the antiplatelet effect and clinical efficacy of clopidogrel therapy.
Journal of the American Medical Association, 302, 849857.
34. Desta, Z., Zhao, X., Shin, J. G., etal. (2002). Clinical significance of the cytochrome P450
2C19 genetic polymorphism. Clinical Pharmacokinetics, 41, 913958.
35. Trenk, D., Hochholzer, W., Fromm, M. F., et al. (2008). Cytochrome P450 2C19 681G>A
polymorphism and high on-clopidogrel platelet reactivity associated with adverse 1-year clini-
cal outcome of elective percutaneous coronary intervention with drug-eluting or bare-metal
stents. Journal of the American College of Cardiology, 51, 19251934.
36. de Morais, S. M., Wilkinson, G. R., Blaisdell, J., et al. (1994). The major genetic defect
responsible for the polymorphism of S-mephenytoin metabolism in humans. The Journal of
Biological Chemistry, 269, 1541915422.
37. Hulot, J. S., Bura, A., Villard, E., etal. (2006). Cytochrome P450 2C19 loss-of-function poly-
morphism is a major determinant of clopidogrel responsiveness in healthy subjects. Blood,
108, 22442247.
38. Brandt, J. T., Close, S. L., Iturria, S. J., etal. (2007). Common polymorphisms of CYP2C19
and CYP2C9 affect the pharmacokinetic and pharmacodynamic response to clopidogrel but
not prasugrel. Journal of Thrombosis and Haemostasis: JTH, 5, 24292436.
39. Sibbing, D., Koch, W., Gebhard, D., etal. Cytochrome 2C19*17 allelic variant, platelet aggre-
gation, bleeding events, and stent thrombosis in clopidogrel-treated patients with coronary
stent placement. Circulation, 121, 512518.
40. Alfonso, F. & Angiolillo, D. J. (2006). Platelet function assessment to predict outcomes after
coronary interventions: Hype or hope? Journal of the American College of Cardiology, 48,
17511754.
41. Fontana, P., Dupont, A., Gandrille, S., etal. (2003). Adenosine diphosphate-induced platelet
aggregation is associated with P2Y12 gene sequence variations in healthy subjects.
Circulation, 108, 989995.
42. Gladding, P., Webster, M., Zeng, I., etal. (2008). The pharmacogenetics and pharmacodynamics
of clopidogrel response: An analysis from the PRINC (Plavix Response in Coronary
Intervention) trial. Journal of the American College of Cardiology Cardiovascular
Interventions, 1, 620627.
43. Niitsu, Y., Jakubowski, J. A., Sugidachi, A., etal. (2005). Pharmacology of CS-747 (prasugre,
LY640315), a novel, potent antiplatelet agent with invivo P2Y12 receptor antagonist activity.
Seminars in Thrombosis and Hemostasis, 31, 184194.
44. Wiviott, S. D., Braunwald, E., McCabe, C. H., etal. (2007). Prasugrel versus clopidogrel in
patients with acute coronary syndromes. The New England Journal of Medicine, 357,
20012015.
45. Mega, J. L., Close, S. L., Wiviott, S. D., et al. (2009). Cytochrome P450 genetic
polymorphisms and the response to prasugrel: Relationship to pharmacokinetic, pharmacody-
namic, and clinical outcomes. Circulation, 119, 25532560.
46. Wallentin, L., Becker, R. C., Budaj, A., etal. (2009). Ticagrelor versus clopidogrel in patients
with acute coronary syndromes. The New England Journal of Medicine, 361, 10451057.
47. Plavix (clopidogrel bisulfate) tablets Prescribing Information. Accessed at http://products.
sanofi-aventis.us/PLAVIX/PLAVIX.html.
48. Gladding, P., White, H., Voss, J., etal. (2009). Pharmacogenetic testing for clopidogrel using
the rapid INFINITI analyzer: A dose-escalation study. Journal of the American College of
Cardiology Cardiovascular Interventions, 2, 10951101.
49. Pena, A., Collet, J. P., Hulot, J. S., et al. (2009). Can we override clopidogrel resistance?
50. Giusti, B., Gori, A. M., Marcucci, R., etal. (2009). Relation of cytochrome P450 2C19 loss-
of-function polymorphism to occurrence of drug-eluting coronary stent thrombosis. The
American Journal of Cardiology, 103, 806811.
51. Patrono, C., Collar, B., Dalen, J., etal. (1998). Platelet-active drugs: The relationships among
dose, effectiveness, and side effects. Chest, 114, 470S488S.
52. Pedersen, A. K. & FitzGerald, G. A. (1984). Dose-related kinetics of aspirin: Presystemic
acetylation of platelet cyclo-oxygenase. The New England Journal of Medicine, 311, 12061211.
53. Needs, C. J. & Brooks, P. M. (1985). Clinical pharmacokinetics of the salicylates. Clinical
Pharmacokinetics, 10, 164177.
54. ISIS-2 Group. (1988). ISIS-2 (Second International Study of Infarct Survival) Collaborative.
Randomised trial of intravenous streptokinase, oral aspirin, both, or neither among 17,187
cases of suspected acute myocardial infarction: ISIS-2. ISIS-2 (Second International Study of
Infarct Survival) Collaborative Group. Lancet, 2, 349360.
55. Baigent, C., Collins, R., Appleby, P., etal. (1998). ISIS-2: 10 year survival among patients
with suspected acute myocardial infarction in randomised comparison of intravenous strepto-
kinase, oral aspirin, both, or neither. The ISIS-2 (Second International Study of Infarct
Survival) Collaborative Group. British Medical Journal (Clinical Research Ed.), 16,
13371343.
56. Roux, S., Christeller, S., & Ldin, E. (1992). Effects of aspirin on coronary reocclusion and
recurrent ischemia after thrombolysis: A meta-analysis. Journal of the American College of
Cardiology, 19, 671677.
57. RISC Group. (1990). Risk of myocardial infarction and death during treatment with low dose
aspirin and intravenous heparin in men with unstable coronary artery disease. Lancet, 336,
827830.
58. Lewis, H. D., Jr, Davis, J. W., Archibald, D. G., etal. (1983). Protective effects of aspirin against
acute myocardial infarction and death in men with unstable angina. Results of a Veterans
Administration Cooperative Study. The New England Journal of Medicine, 309, 396403.
59. Cairns, J. A., Gent, M., Singer, J., etal. (1985). Aspirin, sulfinpyrazone, or both in unstable
angina. Results of a Canadian multicenter trial. The New England Journal of Medicine, 313,
13691375.
60. Throux, P., Ouimet, H., McCans, J., et al. (1988). Aspirin, heparin, or both to treat acute
unstable angina. The New England Journal of Medicine, 319, 11051111.
61. Breddin, K., Loew, D., Lechner, K., et al. (1980). The German-Austrian aspirin trial: A
comparison of acetylsalicylic acid, placebo and phenprocoumon in secondary prevention of
myocardial infarction. On behalf of the German-Austrian Study Group. Circulation, 62,
V63V72.
62. Leon, M. B., Baim, D. S., Popma, J. J., etal. (1998). A clinical trial comparing three anti-
thrombotic-drug regimens after coronary-artery stenting. Stent Anticoagulation Restenosis
Study Investigators. The New England Journal of Medicine, 339, 16651671.
63. Schwartz, L., Bourassa, M. G., Lesprance, J., etal. (1988). Aspirin and dipyridamole in the
prevention of restenosis after percutaneous transluminal coronary angioplasty. The New
64. Barnathan, E. S., Schwartz, J. S., Taylor, L., etal. (1987). Aspirin and dipyridamole in the
prevention of acute coronary thrombosis complicating coronary angioplasty. Circulation, 76,
125134.
65. Gavaghan, T. P., Gebski, V., & Baron, D. W. (1991). Immediate postoperative aspirin improves
vein graft patency early and late after coronary artery bypass graft surgery. A placebo-
controlled, randomized study. Circulation, 83, 15261533.
66. Goldman, S., Copeland, J., Moritz, T., et al. (1988). Improvement in early saphenous vein
graft patency after coronary artery bypass surgery with antiplatelet therapy: Results of a
Veterans Administration Cooperative Study. Circulation, 77, 13421332.
67. Goldman, S., Copeland, J., Moritz, T., etal. (1994). Long-term graft patency (3 years) after
coronary artery surgery. Effects of aspirin: Results of a VA Cooperative study. Circulation, 89,
11381143.
68. Ridker, P. M., Manson, J. E., Gaziano, J. M., et al. (1991). Low-dose aspirin therapy for
chronic stable angina. A randomized, placebo-controlled clinical trial. Annals of Internal
Medicine, 114, 835839.
69. Juul-Mller, S., Edvardsson, N., Jahnmatz, B., etal. (1992). Double-blind trial of aspirin in
primary prevention of myocardial infarction in patients with stable chronic angina pectoris.
The Swedish Angina Pectoris Aspirin Trial (SAPAT) Group. Lancet, 340, 14211425.
70. Steering Committee of the Physicians Health Study Research Group. (1989). Final report on
the aspirin component of the ongoing Physicians Health Study. The New England Journal of
Medicine, 321, 129135.
71. Hennekens, C. H., Peto, R., Hutchison, G. B., etal. (1988). An overview of the British and
American aspirin studies. The New England Journal of Medicine, 318, 923924.
72. Ridker, P. M., Cook, N. R., Lee, I. M., etal. (2005). A randomized trial of low-dose aspirin in
the primary prevention of cardiovascular disease in women. The New England Journal of
Medicine, 352, 12931304.
73. Ogawa, H., Nakayama, M., Morimoto, T., etal. (2008). Low-dose aspirin for primary preven-
tion of atherosclerotic events in patients with type 2 diabetes: A randomized controlled trial.
The Journal of the American Medical Association, 300, 21342141.
74. The SPAF III Writing Committee for the Stroke Prevention in Atrial Fibrillation Investigators.
(1998). Patients with nonvalvular atrial fibrillation at low risk of stroke during treatment with
aspirin. The Journal of the American Medical Association, 279, 12731277.
75. CAST (Chinese Acute Stroke Trial) Collaborative Group. (1997). CAST: Randomised
placebo-controlled trial of early aspirin use in 20,000 patients with acute ischaemic stroke.
Lancet, 349, 16411649.
76. Chen, Z. M., Sandercock, P., Pan, H. C., etal. (2000). Indications for early aspirin use in acute
ischemic stroke: A combined analysis of 40 000 randomized patients from the Chinese Acute
Stroke Trial and the International Stroke Trial. Stroke, 31, 12401249.
77. Collaboration Antithrombotic Trialists. (2002). Collaborative meta-analysis of randomised
trials of antiplatelet therapy for prevention of death, myocardial infarction, and stroke in high
risk patients. British Medical Journal (Clinical Research Ed.), 324, 7186.
78. The Dutch TIA Trial Study Group. (1991). A comparison of two doses of aspirin (30 mg vs.
283 mg a day) in patients after a transient ischemic attack or minor ischemic stroke. The New
79. Farrell, B., Godwin, J., Richards, S., etal. (1991). The United Kingdom transient ischaemic
attack (UK-TIA) aspirin trial: Final results. Journal of Neurology, Neurosurgery, and
Psychiatry, 54, 10441054.
80. Berger, J. S., Krantz, M. J., Kittelson, J. M., etal. (2009). Aspirin for the prevention of car-
diovascular events in patients with peripheral artery disease: A meta-analysis of randomized
trials. Journal of the American Medical Association, 301, 19091919.
81. Vison, A., Tonello, D., Zalunardo, B., et al. (2009). Antithrombotic treatment before and
after peripheral artery percutaneous angioplasty. Blood Transfusion, 7, 1823.
82. Hurwitz, E. S., Barrett, M. J., Bregman, D., et al. (1985). Public Health Service study on
Reyes syndrome and medications. Report of the pilot phase. The New England Journal of
Medicine, 313, 849857.
83. Grotemeyer, K. H., Scharafinski, H. W., & Husstedt, I. W. (1993). Two-year follow-up of
aspirin responder and aspirin non responder. A pilot-study including 180 post-stroke patients.
Thrombosis Research, 71, 397403.
84. Eikelboom, J. W., Hirsh, J., Weitz, J. I., etal. (2002). Aspirin-resistant thromboxane biosyn-
thesis and the risk of myocardial infarction, stroke, or cardiovascular death in patients at high
risk for cardiovascular events. Circulation, 105, 16501655.
85. Chen, W. H., Cheng, X., Lee, P. Y., etal. (2007). Aspirin resistance and adverse clinical events
in patients with coronary artery diseas. The American Journal of Medicine, 120, 631635.
86. Poston, R. S., Gu, J., Brown, J. M., etal. (2006). Endothelial injury and acquired aspirin resis-
tance as promoters of regional thrombin formation and early vein graft failure after coronary
artery bypass grafting. The Journal of Thoracic and Cardiovascular Surgery, 131, 122130.
87. Ohmori, T., Yatomi, Y., Nonaka, T., etal. (2006). Aspirin resistance detected with aggregom-
etry cannot be explained by cyclooxygenase activity: Involvement of other signaling
pathway(s) in cardiovascular events of aspirin-treated patients. Journal of Thrombosis and
Haemostasis: JTH, 4, 12711278.
88. Cotter, G., Shemesh, E., Zehavi, M., etal. (2004). Lack of aspirin effect: Aspirin resistance or
resistance to taking aspirin? American Heart Journal, 147, 293300.
89. Catella-Lawson, F., Reilly, M. P., Kapoor, S. C., etal. (2001). Cyclooxygenase inhibitors and
the antiplatelet effects of aspirin. The New England Journal of Medicine, 345, 18091817.
90. Macchi, L., Christiaens, L., Brabant, S., etal. (2003). Resistance invitro to low-dose aspirin
is associated with platelet PlA1 (GP IIIa) polymorphism but not with C807T(GP Ia/IIa) and
C-5T Kozak (GP Ibalpha) polymorphisms. Journal of the American College of Cardiology,
42, 11151119.
91. Ferreira, I. A., Mocking, A. I., Feijge, M. A., etal. (2006). Platelet inhibition by insulin is absent
in type 2 diabetes mellitus. Arteriosclerosis, Thrombosis, and Vascular Biology, 26, 417422.
92. Weiss, E. J., Bray, P. F., Tayback, M., etal. (1996). A polymorphism of a platelet glycoprotein
receptor as an inherited risk factor for coronary thrombosis. The New England Journal of
Medicine, 334, 10901094.
93. Undas, A., Brummel, K., Musial, J., etal. (2001). Pl(A2) polymorphism of beta(3) integrins
is associated with enhanced thrombin generation and impaired antithrombotic action of
aspirinat the site of microvascular injury. Circulation, 104, 26662672.
94. Lev, E. I., Patel, R. T., Guthikonda, S., etal. (2007). Genetic polymorphisms of the platelet
receptors P2Y(12), P2Y(1) and GP IIIa and response to aspirin and clopidogrel. Thrombosis
Research, 119, 355360.
95. Herrera-Galeano, J. E., Becker, D. M., Wilson, A. F., et al. (2008). A novel variant in the
platelet endothelial aggregation receptor-1 gene is associated with increased platelet aggrega-
bility. Arteriosclerosis, Thrombosis, and Vascular Biology, 28, 14841490.
96. Jones, C. I., Bray, S., Garner, S. F., et al. (2009). A functional genomics approach reveals
novel quantitative trait loci associated with platelet signaling pathways. Blood, 114,
14051416.
97. Antman, E. M., Hand, M., Armstrong, P. W., etal. (2008) 2007 Focused Update of the ACC/
AHA 2004 Guidelines for the Management of Patients With ST-Elevation Myocardial
Infarction: A report of the American College of Cardiology/American Heart Association Task
Force on Practice Guidelines: Developed in collaboration With the Canadian Cardiovascular

Society endorsed by the American Academy of Family Physicians: 2007 Writing Group to
Review New Evidence and Update the ACC/AHA 2004 Guidelines for the Management of
Patients With ST-Elevation Myocardial Infarction, Writing on Behalf of the 2004 Writing
Committee. Circulation, 117, 296329.
98. Gurbel, P. A., Bliden, K. P., Hiatt, B. L., et al. (2003). Clopidogrel for coronary stenting:
Response variability, drug resistance, and the effect of pretreatment platelet reactivity.
99. Awtry, E. H. & Loscalzo, J. (2000). Aspirin. Circulation, 101, 12061218.
Chapter 10
Genotype-Guided Statin Therapy
Richard L. Seip, Jorge Duconge, and Gualberto Ruao
Keywords HMGCoA reductase Hydroxymethylglutaryl coenzyme A reductase

Atorvastatin Simvastatin Rosuvastatin Pravastatin Low-density lipoprotein
cholesterol Myalgia Myopathy Myositis Creatine kinase Neuromuscular
side effects Statin-induced neuromyopathy Physiogenomics SLCO1B1 KIF6
CLMN APOE HMGCR Major adverse cardiovascular event
10.1Introduction
One of the promises of the Human Genome Project is individualization of patient

care based on highly heterogeneous innate metabolic factors determined by DNA
typing of gene polymorphisms. Translation of such gene polymorphisms into
clinical decision support for personalized healthcare is the basis for DNA guided
medicine. Pharmacogenetics serves as the foundation for the most clinically
advanced application of DNA-guided medicine. Statin responsiveness is an area
of high research interest given the success of the drug class in the treatment of
hypercholesterolemia and in primary and secondary prevention of cardiovascular
disease. Ongoing genetic inquiry into response variability indicates probable
multigenic determinants for statin efficacy and safety. It is the vision that knowledge
of the patients genetic status for a number of common variants will soon guide
hyperlipidemic intervention.
R.L. Seip(*)
Genomas, Inc., Hartford Hospital, 67 Jefferson Street, Hartford, CT 06106, USA
and
Division of Cardiology, Hartford Hospital, Hartford, CT 06102-5037, USA
156 R.L. Seip et al.
10.2Genotype-Guided Drug Therapy
It is a worthy goal to predict variability in drug responsiveness. Such prediction has

the potential to reduce medication errors in general, trial and error on the part of
clinicians, risks of therapy, and the resources needed to treat. Statins belong to a
drug class that selectively and competitively inhibit the intracellular enzyme
hydroxymethylglutaryl Coenzyme A reductase (HMGCoA reductase) that is
expressed to different degrees in various tissues. In addition to the desirable inhibi-
tion of cholesterol synthesis, the inhibition of HMGCoA activity reduces synthesis
of geranyl and farnesyl products as shown in Figure 10.1, leading to decreased
isoprenylation of proteins and possible impairment of many varied cellular func-
tions. Statin entry into cells can be gated, and metabolic pathways for the drugs of
this class are varied and drug dependent.
The goal of identifying the genes modulating statin response is challenging.
In contrast to warfarin, for which variants in two genes alone determine 3040% of
the variability in effective drug dose, or clopidogrel, whose activation depends on
the CYP2C19 gene, many more genes have roles in statin pharmacogenetics discovery.
Fig. 10.1 Inhibition of cholesterol and isoprenoid synthesis by statins

10 Genotype-Guided Statin Therapy 157
Each will have a small contribution to the variability of the statin response. As we
will show, the story is evolving.
10.3Clinical Status of Statins
Statins are the most prescribed drugs in the United States [1] and the world [2]. In
the United States, prescriptions for atorvastatin (Lipitor), simvastatin (Zocor),
and rosuvastatin (Crestor) comprise 85% of the market share [1]. The HMG CoA
reductase inhibitors reduce cardiac events in coronary heart disease patients and in
previously healthy subjects [3]. This success has fostered increasingly aggressive
usage and dosing. Their main clinically relevant safety risk is statin-induced
neuromyopathy (SINM) evidenced as a constellation of neuromuscular side effects
(NMSEs). NMSEs are disabling to 320% of patients on statins, require alteration
of therapy, and reduce compliance [47]. NMSEs include myalgias (pain, weak-
ness, aches, cramps) and myositis (typically monitored by elevation of serum
creatine kinase [CK] activity) [4]. NMSEs vary in extent among drugs and from
patient to patient. Were there a system to predict the safety and efficacy of the
preeminent statin drugs according to the genome of each patient, a clinician could
optimize the selection from among atorvastatin, simvastatin, and rosuvastatin.
Alternatively, a patients genomic profile may prove incompatible with statins, and
the clinician could decide to avoid the drug class.
10.4Statin Effectiveness
10.4.1Low-Density Lipoprotein Cholesterol Lowering
Statins are the most effective medications for managing elevated concentrations of
low-density lipoprotein cholesterol (LDLc). As LDLc lowering coincides with
decreases in major cardiovascular events [3], it is a metric useful for evaluation of
statin efficacy in the short-term. Administered at maximum dosages, the most common
statins atorvastatin, simvastatin, and rosuvastatin lower LDLc by 4657% in
patients with primary hypercholesterolemia [810]. Pravastatin is less widely pre-
scribed and less potent, with a mean peak response of 34% [11]. The magnitude of
the LDLc response differs according to phenotypic, demographic, and as yet unex-
plained characteristics [12].
Approximately 50% of the variability in plasma LDLc is estimated to be attrib-
utable to inheritance [13]. Because baseline LDLc predicts the magnitude of LDLc
lowering with statins [14], there may be overlap in the genes that regulate LDLc
metabolism in the drug-free state and statin-mediated LDLc lowering. In the
absence of statin therapy, variants in genes such as (in approximate order of their
predictive strength) APOE cluster [15], APOB [16], LDLR [16, 17], SORT1/CELSR2/
PSRC1 [16], B4GALT4 [16], B3GALT4 [16], NCAN/CILP2 [16], NPC1L1 [18],
and PSCK9 [16, 19] affect LDLc. APOE cluster, LDLR, APOB, SORT1/CELSR2/
PSRC, B4GALT4, B3GALT4, PSCK9, and NCAN/CILP2 have been confirmed by
genome wide association studies (GWAS) in more than one population [16].
Pharmacogenetic studies of LDLc lowering associated with statin therapy have
focused on ~40 genes, mainly those in cholesterol synthetic, lipoprotein lipid trans-
port, and pharmacokinetic pathways. Single nucleotide polymorphisms (SNPs) in
genes of cholesterol metabolism such as HMGCR [2023] and lipoprotein transport
such as APOE [14, 24], CETP [25, 26], LIPC [27], APOA5 [28], LDLR rs1433099
[22], ABCA1 [24], and LEPR (223A>G, rs1137101) [29] can influence the statins
ability to lower LDLc levels. Individual variants in these genes explain significant
differences in statin response ranging up to about 10 mg/dL. Otherwise, only a
small number of common and multiple rare gene variants that contribute to the
phenotype are known [17, 3033]. The more heavily researched genes with the
greatest effects are highlighted below.
APOE harbors three common haplotypes, by convention labeled e2, e3, and e4,
which are the result of nonsynonymous SNPs rs429538 and rs7412 in exon 4.
Genotype frequencies are shown in Table 10.1.
The presence of e2 is significantly associated with a 47% greater decrease
(roughly 610mg/dL) in LDLc response in 1984 patients who received atorvastatin
80mg/day and were genotyped for 291,998 SNPs [14], and in 509 patients receiving
atorvastatin, simvastatin, or pravastatin and genotyped for 489 SNPs [24]. However,
studies demonstrating a null or opposite effect also have been reported [35, 36]. In
general, the e4 allele is associated with a greater need for statin treatment but poorer
response, while the e2 allele has been associated with greater response [37].
HMGCR polymorphisms have been shown to be responsible for up to an 8mg/dL
difference in LDL cholesterol-lowering through simvastatin [21]. The influence of
HMGCR haplotype 7 (Hap7), which is defined by three intronic SNPs, rs17244841,
rs3846662, and rs17238540, exerts an effect that is more prominent in combination
with a second haplotype (Hap2) that also includes rs3846662 [21]. The rs3846662
SNP G>A allele is also responsible for a splice variant in which exon 13 is omitted
and the resultant isoform has reduced statin sensitivity [38, 39]. The rs3846662 A
allele is less common in Africans (315%) [40] and African Americans (1617%)
compared to persons of European (5060%) and Asian (4350%) origin [39, 40].
Table 10.1 APOE genotypes

rs429538 (c.388 T>C, rs7412 (c.526 C>T,
APOE genotype p.C130R) p.R176C) Population frequencya
e2/e2 T/T T/T 0.007
e2/e3 T/T C/T 0.116
e2/e4 T/C C/T 0.022
e3/e3 T/T C/C 0.623
e3/e4 T/C C/C 0.213
e4/e4 C/C C/C 0.019
From a meta analysis of 86,000 patients [34]
a
Gene-specific contributions to statin response may be dependent on the interactions

of several haplotypes. For example, the combined presence of the LDLR L5 haplo-
type (presence of major allele at rs14548, rs1433099, rs7254521, rs5742911,and
rs2738467) and the HMGCR Hap2 further attenuates the apolipoprotein B response
to simvastatin in African Americans [41].
There are other examples of divergent effects on LDLc response ascribed to vari-
ants within a single gene. Of the multiple haplotypes in the LPL gene encoding lipo-
protein lipase, some were associated with increased statin-mediated lipid changes and
some with decreased changes [42]. Our work has shown that the ACACB gene har-
bors SNPs with dual effects on LDLc [43]. The rs2241220 SNP in exon 33 is a
marker for a greater decrease in LDLc in patients taking statins and rs34274 in intron
1 is a risk marker whose minor allele is associated with increased LDLc [43]. As
more systematic analyses that may include deep sequencing are completed, candidate
gene associations that include more complete genetic coverage will emerge.
Recent findings have extended the repertoire of gene variants associated with
statin efficacy to new mechanisms of drug action. A genome wide association
(GWA) study combining nearly 4,000 patients from three industry sponsored studies
(CAP, PRINCE, TNT) is the first GWA study to discover the candidate gene CLMN
(through the rs8014194 SNP in intron 1) that is associated with the change in LDLc
due to statin treatment [33]. CLMN is expressed in the liver and adipose tissue [44],
and encodes the protein calmin. Calmins function is not known [45]. However, it
contains a calponin-like binding domain that suggests actin-binding activity [45].
Another novel finding from a GWA study is KIF6 [46], which encodes a cytoskel-
etal protein involved in intracellular transport of protein complexes, membrane
organelles, and mRNA [47] (see Sect. 10.5, Pharmacogenetic dosing). Ultimately,
confirmation of genetic contributions to the cholesterol-lowering efficacy of statins
will require functional understanding of the genotype-induced alterations in lipo-
protein metabolism.
Pharmacokinetic genes. The hepatic transporter genes SLCO1B1, ABCG2, and
ABCB1 harbor genotypes known to affect the systemic exposure to various statins
[48]. The concentrations of rosuvastatin, atorvastatin, and simvastatin are elevated
1.8 to 3-fold in the presence of SLCO1B1 c.521CC (rs4149056) and 1.5 to 2.4-fold
in the case of ABCG2 c421AA (rs2231142) [48, 49]. Interest has focused on these
variants in relation to differential LDLc responses to statin therapy. In Chinese
patients treated with rosuvastatin, the ABCG2 c.421A genotype affected response
in a dose-wise manner with a 6.9% greater reduction in LDLc in AA vs. CC [50].
Simvastatin acid, but not lactone, and atorvastatin acid and lactone are substrates of
the intestinal P-glycoprotein efflux transporter, which pumps the drugs back into
the intestinal lumen during drug absorption and is the product of the ABCB1 gene.
Though ABCB1 haplotypes affect the pharmacokinetics of the active acid forms of
simvastatin and atorvastatin invivo [51], differential effects on LDLc lowering are
statistically significant but small (~4%) [52, 53] and by themselves not enough to
warrant pharmacogenetic prescribing.
The SLCO1B1 gene, which encodes organic anion transporter protein 1B1
(OATP1B1), harbors four haplotypes *1A, *1B, *5, and *15 that are relevant to drug
Table 10.2 Pharmacokinetic genes harboring variants that affect LDLc response to statins
Statins (in order Drug transporters Drug metabolizers
of lipophilicity) SLCO1B1 ABCB1 ABCG2 CYP3A4/5 CYP2C9 CYP2C19 CYP2D6
Simvastatin [14, 55, 59] [52] [52, 60] [36]
Atorvastatin [14] [53] [60, 61]
Rosuvastatin [62] [50] [62] [62] [62]
Pravastatin [5658] [63]
Fluvastatin [59, 64] [60] [36]
Significant effect according to current data; no significant effect according to current data;
results are conflicting according to current data; no symbol indicates data unavailable
transport [54]. The *1B haplotype carries the rs2306238 492A>G variant and *5
corresponds to rs4149056 625T>C on reference sequence NM_006446.4, and
iscommonly referred to as T521C, encoding OATP1B1:V174A [54]. The *15 variant
carries both SNP minor alleles [54]. The *5 variant product demonstrates deficient
transport activity and it raises risk of elevated CK activity [55] (discussed below, Sect.
10.6). However, its relationship to either statin-mediated LDL cholesterol lowering or
CVD risk reduction are conflicting; current evidence based on atorvastatin [14],
pravastatin [5658], and multiple statins [59] remains weak as shown in Table 10.2.
Pravastatin and fluvastatin are less lipophilic and expected to be affected by
transporting peptide variants. Heterozygous carriers of SLCO1B1*15 allele treated
with pravastatin for 8 weeks had poor LDLc reduction (-14.1%) relative to noncar-
riers (28.9%) [56] in a study of 33 patients. In 45 Chinese patients treated with
pravastatin for 30 days, those who were heterozygous for the SLCO1B1 521TC
(Val174Ala) functional genetic polymorphism, experienced total cholesterol lower-
ing of 14% compared to 22% in those who were homozygous wild type (521TT).
Results are consistent with decreased hepatic uptake due to the polymorphism. The
common *14 allele of SLCO1B1, which is distinguished by the presence of the
c.463C>A polymorphism, was associated with enhanced lipid-lowering efficacy
with hydrophilic fluvastatin in a study of 400 patients receiving the drug [64].
Simvastatin and atorvastatin are lipophilic and expected to be susceptible to hepatic
metabolism [65]. Variants in CYP3A4 and CYP3A5 genes affect atorvastatin and sim-
vastatin effectiveness [60]. The CYP3A5*3 variant is a poor metabolizer of simvastatin,
atorvastatin, and lovastatin compared to the wild type CYP3A5*1 resulting in a greater
statin concentration [66]. LDLc was 19% lower in CYP3A5*3 carriers receiving those
forms of therapy but no such effect is seen for pravastatin, which is hydrophilic, or
fluvastatin. In patients treated with atorvastatin 10mg/day, the CYP3A4 A-290G variant
(*1B, rs2740574) was significantly associated with higher levels of posttreatment LDL
cholesterol, whereas the *3 variant M445T (rs4986910) was associated with lower
levels of LDL cholesterol before and after treatment [61]. Homozygous carriers of the
CYP7A1 -204C allele or heterozygotes for both CYP7A1 -204C and APOE e4 alleles
showed significantly poorer LDLc reduction compared to that seen in other combinato-
rial genotype groups after 1 year of pravastatin treatment (-24.3 vs. -33.1%) [56].
10.5Mortality and Prevention of MACE
Long-term statin efficacy has been evaluated by determining mortality rate and/or
major adverse cardiovascular event rate (MACE). MACE may include fatal coro-
nary heart disease and nonfatal myocardial infarction and combined CHD.
Mechanistically, in addition to decreasing the circulating lipid substrate (LDLc) for
arterial plaque progression, SNPs may influence the ability of statins to (1) stabilize
plaques, (2) modulate coagulation, (3) reduce inflammation, and (4) correct
endothelial function [67]. So far, more than 30 genes have been examined, most
with ambiguous results [68]. The contribution of individual polymorphisms is
poorly understood.
CETP gene: In the REGRESS cohort, 812 patients with coronary artery disease
(CAD) were followed for 10 years, the last eight with pravastatin therapy [69]. The
efficacy (atherosclerotic disease death) of statin therapy was dependent on CETP
genotype and associated plasma CETP levels. Those carrying the TaqIB-B2 allele,
which is tagged with SNP rs17231506, had higher hazard ratios for MACE
(1.531.59) and all caused mortality (1.30). The B2 allele effect was confirmed in
another study [70]. Pravastatin therapy slowed the progression of coronary athero-
sclerosis in B1B1 carriers but not in B2B2 carriers, which comprised 16% of
patients taking pravastatin. This common DNA variant appears to predict whether
men with CAD will benefit from treatment with pravastatin to delay the progression
of CAD. Another CETP variant, the I405V (rs5882) polymorphism, can modify the
effect of simvastatin on TG reduction and HDL-C elevation with carriers of the I
allele responding better to treatment [25]. In a study of 82 SNPs in genes that were
selected for previously having polymorphisms associated to statins (ABCB1, CETP,
LDLR, LIPC, nitric oxide synthase [NOS], and HMGCR), two SNP-statin interac-
tions on MI were observed (one ABCB1, one LIPC) and five interactions on stroke
(one CETP, four LIPC). The strongest SNP-statin interaction was for synonymous
CETP SNP rs5883 on stroke (P=0.008) [71]. In a randomized study of 1,400 renal
transplant patients receiving fluvastatin or placebo, tests of reported associations
between CETP and CVD yielded varying results [72]. All of these findings need to
be confirmed in larger studies.
LDLR gene: The rs1433099 and rs2738466 SNPs were shown to significantly affect
the primary outcome of CHD death or nonfatal MI or fatal or nonfatal stroke in response
to high dose pravastatin [22]. The T allele for the C44857T rs1433099 was associated
with lower risk for CHD. In haplotype analysis, those carrying the C44857T[T]
A44964G[A] haplotype had a lower risk for primary endpoints (HR 0.69, CI 0.520.90)
and cardiovascular events (HR 0.74, CI 0.570.95) than the C44857T[C]A44964G[G]
haplotype carriers. This was not affected by pravastatin [22].
SNPs in the coagulation factors V (Arg506Gln G>A, rs6025) and VII (Arg353Gln
G>A, rs6046) influenced pravastatins ability to prevent fatal CAD and nonfatal
myocardial infarction in >9,000 patients randomized to pravastatin or placebo in the
GenHAT study [73]. The combined interaction (pravastatin-SNP) hazard ratios were
1.33 and 1.92, respectively. Polymorphisms in genes in the homocysteine pathway
(MTHFR 677 C>T and CBSins) appear to modify the efficacy of pravastatin in
reducing risk of cardiovascular events [74].
The stabilization of plaques by statins can be affected by variation in the matrix
metalloproteinases (MMP) secreted by inflammatory cells. These enzymes degrade
the extracellular matrix, undermining structural integrity and predisposing fibrous
caps to rupture [75]. The Ala227Pro polymorphism (rs428785, NM_006988:
+1134G>C) in the ADAMTS1 metalloproteinase gene is associated with signifi-
cantly increased risk of CAD or myocardial infarction [76]. Pravastatin decreased
the risk of fatal CAD/myocardial infarction to a threefold greater extent in patients
homozygous for 227Pro compared to those who were heterozygous [76].
Pharmacogenetic dosing: Selection of the optimal statin dose may also be
genetically determined, at least in part. The KIF6 gene encodes kinesin-like protein
6, a member of the molecular motor superfamily. Two prospective trials (CARE
and WOSCOPS) including more than 28,000 subjects [77, 78] have shown the
Trp719Arg variant (rs20455) in KIF6 to be associated with coronary events [77,
78]. Intensive compared to moderate statin therapy imparted greater protection
from coronary events in carriers of KIF6 719Arg compared to noncarriers (10.0 vs.
0.8% reduction, respectively, for absolute rates) [46]. The mechanism is unknown
but appears distinct from lipid or C-reactive protein lowering [46]. Functional studies
of the KIF6 kinesin are warranted, given the consistent association of Trp719Arg
with risk of coronary events and statin benefit [46].
In summary, evidence that statin-mediated reductions in CVD risk exceed levels
expected from LDLc lowering have resulted in examination of pleiotropic or
nonlipid lowering effects of statin treatment [67]. Large studies will be necessary
because the predictive effects of new biomarkers, the number of which will be
large, appear to be small.
10.6Statin Safety
Statins are well tolerated by the majority of patients at low starting dosages. However,
they can produce SINM and their usage is ultimately limited by toxicity. NMSEs
occur in >10% of patients [6], affecting compliance to statin therapy. These may
include myalgia (muscle aches, cramps), weakness, fatigue, heaviness, myositis,
neuropathy, and other forms of intolerance to statin therapy, as shown in Fig. 10.2.
There are serum and clinical surrogate markers for SINM that we have collec-
tively termed as NMSEs. Serologically, increased activity of serum CK provides
the predominant means for assessing the degree of myopathic severity. The elevation
of CK activity to tenfold ULN (upper limit of normal) has been suggested as indi-
cating severe SINM [79]. The pharmacokinetic gene SLCO1B1 *5 variant
rs4363657 SNP is strongly associated to the high degree of CK elevation [55],
likely through linkage disequilibrium with the nonsynonymous SNP rs4149056
(Val174Ala) or the rs4149080 SNP in the 13th intron [80], and there is independent
confirmation of the SLCO1B1*5 relationship to elevated CK [80]. Pharmacokinetic
Fig. 10.2 The nature and severity of statin-induced neuromyopathy (SINM) symptoms and
neuromuscular side effects
Table 10.3 Pharmacokinetic genes harboring variants that affect myalgia or CK response to
statins
Statins (in Drug transporters Drug metabolizers
order of
lipophilicity) SLCO1B1 ABCB1 ABCG2 ABCG8 CYP3A4/5 CYP2C9 CYP2C19 CYP2D6
Simvastatin [55, 80] [52] [52] [36] [36, 81]
Atorvastatin [82] [36] [36]
[36]
Rosuvastatin
Pravastatin [83]
Fluvastatin
Significant CK elevating effect according to current data; no significant CK elevating effect
according to current data; significant myalgia effect according to current data; no significant
myalgia effect according to current data; results are conflicting with respect to CK elevation and
myalgia according to current data; no symbol indicates data unavailable
gene-focused hypotheses are founded on the increased plasma concentrations

resulting from decreased hepatic entry (variation in drug transporters) and metabo-
lism (variation in cytochrome p450 and possibly glucuronidation pathways), as
mentioned (Sect. 10.4, Pharmacokinetic genes). Table 10.3 summarizes the
findings.
CK and myalgia: Myalgias occur in patients often with no or little CK elevation
and CK is not necessarily elevated in the presence of histopathological evidence of
statin associated muscle damage [84, 85]. Biopsy studies of the vastus lateralis in a
series of 44 statin myalgia patients revealed damage specific to the t-tubules and the
appearance of vacuoles within the muscle fibers in 60% of the patients (2% of
muscle fibers damaged). CK activity was within normal limits in 66% of the
patients with damage. Breakdown of the t-tubular system and subsarcolemmal
rupture has also been observed in asymptomatic patients taking statins and with
normal CK [85]. Only in the clinically rare condition of rhabdomyolysis is the
relationship between myopathy, extremely elevated CK, and clinical severity,
incontrovertible [86]. There is a need to identify novel surrogate markers that can
better predict high risk of myalgia in patients taking statins.
Pharmacodynamic genes: The pharmacodynamic-based mechanisms that have
been advanced have incorporated diverse and complex pathways and with a focus
mainly on myofibers. Thompson and colleagues [4, 87] have provided thorough
reviews of mechanisms of statin myopathy. Statin interactions with HMG-CoA
reductase homolog proteins may interfere with energy transduction processes [88,
89]. Some mechanisms find their basis in the possibility that statin inhibition of
nonsteroidal molecules such as ubiquinone and isoprenoids triggers disruption in
normal mitochondrial function, cell signaling, cell proliferation, and cell repair [4,
88, 9099]. Specific proposed myalgia etiologies include decreased sarcolemmal
[4] or sarcoplasmic reticular cholesterol [85], reduced production of ubiquinone or
coenzyme Q10 [100], decreased production of prenylated proteins [4], changes in
fat metabolism [101], increased uptake of cholesterol [102] or phytosterols [103],
failure to replace damaged muscle protein via the ubiquitin pathway [104], disrup-
tion of calcium metabolism in the skeletal muscle [84, 105], and inhibition of
selenoprotein synthesis [106]. Phenotypic expression is quite variable. As a result,
genetic markers are generally unknown though progress to identify candidate markers
has been made [107].
Statin exposure that produces myopathy sometimes unmasks a latent muscle
pathology with known genetic basis [107]. Severe statin myopathy is more frequent
in the presence of variants in muscle metabolic disease genes COQ2 (CoQ10 syn-
thetic pathway), CPT2 (carnitine transferase Type II deficiency), PYGM (McArdle
disease) [108], and AMPD1 (myoadenylate deaminase [MADA] deficiency) [109,
110]. As examples, the prevalence of heterozygosity for any of 20 mutations in
PYGM is 1/170 (0.6%) in the general population, but 12.3% in patients with statin
myopathy [109]. Similarly, the carrier frequency for CPT II deficiency is estimated
at 0.4% in the general population [109] and 3.8% in patients with statin myopathy
[109]. MADA deficiency is an autosomal recessive disorder traceable to the
AMPD1 gene that impairs conversion of 5-adenosine monophosphate (5AMP) to
inosine monophosphate (IMP). MADA deficiency caused by homozygosity for a
mutant allele in exon 3 of AMPD1 occurred in 6.5% of a series of patients with
severe statin myopathy compared to 2% estimated for the general population [111].
The increased carrier frequencies among patients with severe statin myopathy
suggest that these normally rare disorders are potentially common risk factors for
drug-induced myopathy when present in the carrier state [107]. Candidate genes for
deficiency of the muscle CoQ10 pathway [112] include PDSS1, PDSS2, COQ2,
ETFDH, and APTX1. Two genes overexpressed in damaged muscle of severe statin
myopathy regardless of CK status were SERCA3 (sarco endoplasmic reticulum

transporting Ca+2 ATPase type 3) and RYR3 (ryanodine receptor type 3) [86], both
of which participate in calcium homeostasis. Calcium homeostasis is also impaired
in malignant hyperthermia, an inherited condition that may be more common in
patients with severe statin myopathy, based on the contracture test results [105].
Muscle-based symptoms of statin myalgia may reflect origins outside of the
myofiber. Statins alter vascular function through decreased isoprenylation of
G-proteins and disturbance of the NOS system [113]. Cell culture studies have
shown that statins induced apoptosis in various cell types, including vascular
smooth muscle [114, 115]. Atogin-1 (encoded by FBXO32) is an ubiquitin ligase
that increases early in skeletal muscle atrophy and is also induced by statins [98].
The transfer of geranylgeranyl isoprene units but not farnesyl units prevents
atrogin-1 induction in cell culture [116]. There are also reports of neuropathy,
peripheral neuropathy, and polyneuropathy and in addition, two reports of aggrava-
tion of existing polyneuropathy associated with the use of HMG-CoA-reductase
inhibitors [117], suggesting a neural basis. Other new hypotheses support a link
between myopathy during statin therapy to genes affecting vascular smooth muscle
[7] and serotonergic transduction [118] and raise the possibility that statin side
effects arise through different pathways.
Without a unifying mechanistic understanding of the NMSE, the most relevant
clinical scenario is the percentage of patients who develop myalgias and muscle
weakness disabling enough to trigger a referral to a physician for neuromuscular
assessment and alternative drug therapy. Therefore, it has been our goal to model
clinical and genetic parameters for the analysis of statin response and the demarca-
tion of symptomatic and metabolic NMSEs including phenotype distribution, drug
regimen, sample size, predictive power, and allele frequency.
10.7Physiogenomic Research: Simultaneous Prediction

of Efficacy and Side Effects
To date, there exist only two published genome wide studies of statin response
[33,55]. Yet, based on emerging successes in predicting the balance of efficacy vs.
safety [119], genomic studies are expected to eventually guide drug selection.
Particularly at higher doses required for advanced disease, statins can induce myo-
pathy and their usage is ultimately limited by toxicity [120].
An upper limit of ~3% of variance in LDLc attributable to any single SNP
association has been hypothesized [41], based on findings in increasingly larger
study populations. In view of small contributions of many genes to the LDLc lowering,
and of the paucity of information on safety markers, our efforts are directed to
simultaneously derive SNP biomarkers of statin safety and efficacy. Physiogenomics
(PG) is a medical application of sensitivity analysis and systems engineering [121].
Sensitivity analysis is the study of the relationship between input and output from
a system as determined by system components. PG utilizes the genes as the
c omponents of the system. The gene variability, measured by SNPs, is correlated to

physiological responses, the output, of a diversely responsive human population
[122]. This approach determines how the SNP frequency varies among individuals
similarly responding to the input over the entire range of the response distribution.
Previously, PG has been utilized to identify genes relevant to dietary weight reduc-
tion [123, 124], exercise response [125], and drug-induced side effects [126].
We operationally defined statin efficacy as LDL cholesterol lowering [119] and
statin safety as the presence or absence of myalgia [118, 119] and the level of serum
CKs activity [7, 119]. In clinical populations of up to 466 patients receiving either
simvastatin, atorvastatin, or rosuvastatin, we employed genotyping using the PG
Array, which consists of 384 SNPs among 222 genes representing metabolic,
inflammatory, and cell regulatory genes [122].
In outpatients receiving statin therapy, SNPs rs2276307 and rs1935349 in the
serotonin receptor genes HTR3B and HTR7, respectively, significantly associate
with statin-induced myalgia [118], and SNPs rs12695902 and rs1799983, in
theAGTR1 (angiotensin receptor type I) and NOS genes, respectively, significantly
associate with myositis [7]. The physiogenomic approach has identified SNP markers
distributed across several gene pathways that suggest neural [118] and vascular [7]
components to augment the primarily myocyte-based hypotheses of statin induced
myopathy. With respect to LDLc lowering, we found a dual effect within the
ACACB gene [43] as discussed earlier.
The data indicate that SNP markers for myalgia, myositis (CK elevation), and
LDLc lowering are largely phenotype specific. Our working hypothesis is that statin
response phenotypes are modulated by separate gene pathways. Our findings and
others will permit construction of a prototype of a physiogenomic-based safety/effi-
cacy model that consolidates the myalgia, myositis, and LDL cholesterol reduction
components. Pending further validation, we predict the existence of genotypes to help
clinicians prescribe statins so as to minimize side effects and maximize efficacy.
10.8Conclusion
Genetic variants modulate the statins capacity to lower LDLc and induce myalgia
and myositis, all of which are factors that strongly influence drug selection in the
clinic. The strongest predictors of LDLc response are variants within the APOE and
HMGCR genes. Taken individually, these explain up to 7% of the variance in
change in LDLc but this figure is expected to decrease to 3% of the change in LDLc
with statin therapy as large diverse populations are studied. Other genes examined
have lesser effect sizes or have not been sufficiently studied to warrant mention.
Still, it is safe to predict that numerous markers taken together will have additive
effects in explaining the variance in LDLc response. With respect to CK elevation,
the SLCO1B1 gene harbors the single most important marker during simvastatin
therapy. Progress to identify other markers for myositis and myalgia is hindered
bythe absence of a unifying pathophysiological hypothesis for toxicity as well as
the hurdles inherent to assembling large patient databases with well characterized
phenotypic data.
In the long-term, statins decrease MACE and all cause mortality in part through
pleiotropic mechanisms. Pharmacogenomic studies have provided new insights
revealing that nonsynonymous SNPs in KIF6 (rs20455, Trp719Arg), ADAMTS1
(rs428785, Ala227Pro), and coagulation factors V (rs6025, Arg506Gln G>A) and
VII (rs6046, Arg353Gln G>A) affect patients risk of coronary events through
pravastatin [46, 73, 76, 77]. Whether other statins have the same effects is not
known. With relatively firm clinical endpoints serving as phenotypes, examination
of existing large clinical databases using pharmacogenomic approaches may be
useful.
References
1. Findlay, S. (2007). The statin drugs. Prescription and price trends October 2005 to December
2006 and potential cost savings to Medicare from increased use of lower cost statins.
Consumer Reports. Yonkers: Consumers Union.
2. Statins: The World Market, 20092024. 3-3-2009. London: Visiongain, Ltd. http://www.
reportlinker.com/p0148913/Statins-The-World-Market-2009-2024.html?request=news,
accessed May20, 2010.
3. Waters, D. D. (2006). What the statin trials have taught us. The American Journal of
4. Thompson, P. D., Clarkson, P., & Karas, R. H. (2003). Statin-associated myopathy. The
5. Pasternak, R. C., Smith, S. C., Jr., Bairey-Merz, C. N., Grundy, S. M., Cleeman, J. I., &
Lenfant, C. (2002). ACC/AHA/NHLBI clinical advisory on the use and safety of statins.
Stroke, 33, 23372341.
6. Bruckert, E., Hayem, G., Dejager, S., Yau, C., & Begaud, B. (2005). Mild to moderate mus-
cular symptoms with high-dosage statin therapy in hyperlipidemic patients the PRIMO
study. Cardiovascular Drugs and Therapy, 19, 403414.
7. Ruao, G., Thompson, P. D., Windemuth, A., Smith, A., Kocherla, M., Holford, T. R., etal.
(2005). Physiogenomic analysis links serum creatine kinase activities during statin therapy
to vascular smooth muscle homeostasis. Pharmacogenomics, 6, 865872.
8. Jones, P., Kafonek, S., Laurora, I., & Hunninghake, D. (1998). Comparative dose efficacy
study of atorvastatin versus simvastatin, pravastatin, lovastatin, and fluvastatin in patients
with hypercholesterolemia (the CURVES study). The American Journal of Cardiology, 81,
582587.
9. Stein, E. A., Davidson, M. H., Dobs, A. S., Schrott, H., Dujovne, C. A., Bays, H., et al.
(1998). Efficacy and safety of simvastatin 80 mg/day in hypercholesterolemic patients. The
expanded dose Simvastatin U.S. Study Group. The American Journal of Cardiology, 82,
311316.
10. Ballantyne, C. M., Weiss, R., Moccetti, T., Vogt, A., Eber, B., Sosef, F., et al. (2007).
Efficacy and safety of rosuvastatin 40 mg alone or in combination with ezetimibe in patients
at high risk of cardiovascular disease (results from the EXPLORER study). The American
Journal of Cardiology, 99, 673680.
11. Sacks, F. M., Moye, L. A., Davis, B. R., Cole, T. G., Rouleau, J. L., Nash, D. T., etal. (1998).
Relationship between plasma LDL concentrations during treatment with pravastatin and
recurrent coronary events in the Cholesterol and Recurrent Events trial. Circulation, 97,
14461452.
12. Simon, J. A., Lin, F., Hulley, S. B., Blanche, P. J., Waters, D., Shiboski, S., etal. (2006).
Phenotypic predictors of response to simvastatin therapy among African-Americans and
Caucasians: The Cholesterol and Pharmacogenetics (CAP) Study. The American Journal of
13. Heller, D. A., de Faire, U., Pedersen, N. L., Dahlen, G., & McClearn, G. E. (1993). Genetic
and environmental influences on serum lipid levels in twins. The New England Journal of
Medicine, 328, 11501156.
14. Thompson, J. F., Hyde, C. L., Wood, L. S., Pacige, S. A., Hinds, D. A., Cox, D. R., etal.
(2009). Comprehensive whole-genome and candidate gene analysis for response to statin
therapy in the treating to new targets (TNT) cohort. Circulation: Cardiovascular Genetics,
2, 173181.
15. Bennet, A. M., Di Angelantonio, E., Ye, Z., Wensley, F., Dahlin, A., Ahlbom, A., et al.
(2007). Association of apolipoprotein E genotypes with lipid levels and coronary risk. The
16. Willer, C. J., Sanna, S., Jackson, A. U., Scuteri, A., Bonnycastle, L. L., Clarke, R., et al.
(2008). Newly identified loci that influence lipid concentrations and risk of coronary artery
disease. Nature Genetics, 40, 161169.
17. Burnett, J. R., & Hooper, A. J. (2008). Common and rare gene variants affecting plasma LDL
cholesterol. The Clinical Biochemist Reviews, 29, 1126.
18. Carlezon, W. A., Jr., Beguin, C., DiNieri, J. A., Baumann, M. H., Richards, M. R.,
Todtenkopf, M., etal. (2006). Depressive-like effects of the kappa-opioid receptor agonist
salvinorin A on behavior and neurochemistry in rats. The Journal of Pharmacology and
Experimental Therapeutics, 316, 440447.
19. Cohen, J. C., Boerwinkle, E., Mosley, T. H., Jr., & Hobbs, H. H. (2006). Sequence variations
in PCSK9, low LDL, and protection against coronary heart disease. The New England
20. Chasman, D. I., Posada, D., Subrahmanyan, L., Cook, N. R., Stanton, V. P., Jr., & Ridker, P. M.
(2004). Pharmacogenetic study of statin therapy and cholesterol reduction. The Journal of the
American Medical Association, 291, 28212827.
21. Krauss, R. M., Mangravite, L. M., Smith, J. D., Medina, M. W., Wang, D., Guo, X., etal.
(2008). Variation in the 3-hydroxyl-3-methylglutaryl coenzyme a reductase gene is associ-
ated with racial differences in low-density lipoprotein cholesterol response to simvastatin
treatment. Circulation, 117, 15371544.
22. Polisecki, E., Muallem, H., Maeda, N., Peter, I., Robertson, M., McMahon, A. D., etal. (2008).
Genetic variation at the LDL receptor and HMG-CoA reductase gene loci, lipid levels, statin
response, and cardiovascular disease incidence in PROSPER. Atherosclerosis, 200, 109114.
23. Hiura, Y., Tabara, Y., Kokubo, Y., Okamura, T., Goto, Y., Nonogi, H., etal. (2010). Association
of the functional variant in the 3-hydroxy-3-methylglutaryl-coenzyme a reductase gene with
low-density lipoprotein-cholesterol in Japanese. Circulation Journal, 74, 518522.
24. Voora, D., Shah, S. H., Reed, C. R., Zhai, J., Crosslin, D. R., Messer, C., et al. (2008).
Pharmacogenetic predictors of statin-mediated low-density lipoprotein cholesterol reduction
and dose response. Circulation. Cardiovascular Genetics, 1, 100106.
25. Anagnostopoulou, K., Kolovou, G., Kostakou, P., Mihas, C., Mikhailidis, D., & Cokkinos, D. V.
(2007). Pharmacogenetic study of cholesteryl ester transfer protein gene and simvastatin
treatment in hypercholesterolaemic subjects. Expert Opinion on Pharmacotherapy, 8,
24592463.
26. Winkelmann, B. R., Hoffmann, M. M., Nauck, M., etal. (2003). Haplotypes of the choles-
teryl ester transfer protein gene predict lipid-modifying response to statin therapy. The
27. Zambon, A., Deeb, S. S., Brown, B. G., Hokanson, J. E., & Brunzell, J. D. (2001). Common
hepatic lipase gene promoter variant determines clinical response to intensive lipid-lowering
treatment. Circulation, 103, 792798.
28. Hubacek, J. A., Adamkova, V., Prusikova, M., Snejdrlova, M., Hirschfeldova, K., Lanska, V.,
et al. (2009). Impact of apolipoprotein A5 variants on statin treatment efficacy.
29. Sun, Y. M., Wang, L. F., Li, J., Li, Z. Q., & Pan, W. (2009). The 223A>G polymorphism of
the leptin receptor gene and lipid-lowering efficacy of simvastatin in Chinese patients with
coronary heart disease. European Journal of Clinical Pharmacology, 65, 157161.
30. Kathiresan, S., Melander, O., Guiducci, C., Surti, A., Burtt, N. P., Rieder, M. J., etal. (2008).
Six new loci associated with blood low-density lipoprotein cholesterol, high-density lipopro-
tein cholesterol or triglycerides in humans. Nature Genetics, 40, 189197.
31. Kathiresan, S., Melander, O., Anevski, D., Guiducci, C., Burtt, N. P., Roos, C., etal. (2008).
Polymorphisms associated with cholesterol and risk of cardiovascular events. The New
32. Lusis, A. J., & Pajukanta, P. (2008). A treasure trove for lipoprotein biology. Nature
Genetics, 40, 129130.
33. Barber, M. J., Mangravite, L. M., Hyde, C. L., Chasman, D. I., Smith, J. D., McCarty, C. A.,
et al. (2010). Genome-wide association of lipid-lowering response to statins in combined
study populations. PLoS ONE, 5, e9763.
34. Bennet, A. M., Di Angelantonio, A. E., Ye, Z., Wensley, F., Dahlin, A., Ahlbom, A., etal.
(2007). Association of apolipoprotein E genotypes with lipid levels and coronary risk. The
35. Tavintharan, S., Lim, S. C., Chan, Y. H., & Sum, C. F. (2007). Apolipoprotein E genotype
affects the response to lipid-lowering therapy in Chinese patients with type 2 diabetes mel-
litus. Diabetes, Obesity & Metabolism, 9, 8186.
36. Zuccaro, P., Mombelli, G., Calabresi, L., Baldassarre, D., Palmi, I., & Sirtori, C. R. (2007).
Tolerability of statins is not linked to CYP450 polymorphisms, but reduced CYP2D6
metabolism improves cholesteraemic response to simvastatin and fluvastatin.
Pharmacological Research, 55, 310317.
37. Saidi, S., Slamia, L. B., Ammou, S. B., Mahjoub, T., & Almawi, W. Y. (2007). Association of
apolipoprotein E gene polymorphism with ischemic stroke involving large-vessel disease and its
relation to serum lipid levels. Journal of Stroke and Cerebrovascular Diseases, 16, 160166.
38. Medina, M. W., Gao, F., Ruan, W., Rotter, J. I., & Krauss, R. M. (2008). Alternative splicing
of 3-hydroxy-3-methylglutaryl coenzyme A reductase is associated with plasma low-density
lipoprotein cholesterol response to simvastatin. Circulation, 118, 355362.
39. Medina, M. W., & Krauss, R. M. (2009). The role of HMGCR alternative splicing in statin
efficacy. Trends in Cardiovascular Medicine, 19, 173177.
40. The International HapMap Consortium. (2005). A haplotype map of the human genome.
Nature, 437, 12991320.
41. Mangravite, L. M., Medina, M. W., Cui, J., Pressman, S., Smith, J. D., Rieder, M. J., et al.
(2010). Combined influence of LDLR and HMGCR sequence variation on lipid-lowering
response to Simvastatin. Arteriosclerosis, Thrombosis, and Vascular Biology, 30, 14821492.
42. Goodarzi, M. O., Taylor, K. D., Scheuner, M. T., etal. (2007). Haplotypes in the lipoprotein
lipase gene influence high-density lipoprotein cholesterol response to statin therapy and
progression of atherosclerosis in coronary artery bypass grafts. The Pharmacogenomics
Journal, 7, 6673.
43. Ruao, G. R., Thompson, P. D., Kane, J. P., Pullinger, C. R., Windemuth, A., Seip, R. L.,
et al. (2010). Physiogenomic analysis of statin-treated patients: Domain specific counter
effects within the ACACB gene on LDL cholesterol? Pharmacogenomics, 11, 959971.
44. Hirosawa, M., Nagase, T., Ishikawa, K., Kikuno, R., Nomura, N., & Ohara, O. (1999).
Characterization of cDNA clones selected by the GeneMark analysis from size-fractionated
cDNA libraries from human brain. DNA Research, 6, 329336.
45. Ishisaki, Z., Takaishi, M., Furuta, I., Huh, N. (2001). Calmin, a protein with calponin
homology and transmembrane domains expressed in maturing spermatogenic cells.
Genomics, 74, 172179.
46. Iakoubova, O. A., Sabatine, M. S., Rowland, C. M., etal. (2008). Polymorphism in KIF6
gene and benefit from statins after acute coronary syndromes: Results from the PROVE
IT-TIMI 22 study. Journal of the American College of Cardiology, 51, 449455.
47. Hirokawa, N., & Noda, Y. (2008). Intracellular transport and kinesin superfamily proteins,
KIFs: Structure, function, and dynamics. Physiological Reviews, 88, 10891118.
48. Niemi, M. (2010). Transporter pharmacogenetics and statin toxicity. Clinical Pharmacology
49. Keskitalo, J. E., Zolk, O., Fromm, M. F., Kurkinen, K. J., Neuvonen, P. J., & Niemi, M.
(2009). ABCG2 polymorphism markedly affects the pharmacokinetics of atorvastatin and
rosuvastatin. Clinical Pharmacology and Therapeutics, 86, 197203.
50. Tomlinson, B., Hu, M., Lee, V. W., Lui, S. S., Chu, T. T., Poon, E. W., etal. (2010). ABCG2
polymorphism is associated with the low-density lipoprotein cholesterol response to rosuvas-
tatin. Clinical Pharmacology and Therapeutics, 87, 558562.
51. Keskitalo, J. E., Kurkinen, K. J., Neuvoneni, P. J., & Niemi, M. (2008). ABCB1 haplotypes
differentially affect the pharmacokinetics of the acid and lactone forms of simvastatin and
atorvastatin. Clinical Pharmacology and Therapeutics, 84, 457461.
52. Fiegenbaum, M., da Silveira, F. R., Van der Sand, C. R., Van der Sand, L. C., Ferreira, M. E.,
Pires, R. C., etal. (2005). The role of common variants of ABCB1, CYP3A4, and CYP3A5
genes in lipid-lowering efficacy and safety of simvastatin treatment. Clinical Pharmacology
53. Thompson, J. F., Man, M., Johnson, K. J., Wood, L. S., Lira, M. E., Lloyd, D. B., etal. (2005).
An association study of 43 SNPs in 16 candidate genes with atorvastatin response. The
54. Oshiro, C., Mangravite, L., Klein, T., & Altman, R. (2010). PharmGKB very important
pharmacogene: SLCO1B1. Pharmacogenetics and Genomics, 20, 211216.
55. Link, E., Parish, S., Armitage, J., Bowman, L., Heath, S., Matsuda, F., et al. (2008).
SLCO1B1 variants and statin-induced myopathy a genomewide study. The New England
56. Takane, H., Miyata, M., Burioka, N., Shigemasa, C., Shimizu, E., Otsubo, K., etal. (2006).
Pharmacogenetic determinants of variability in lipid-lowering response to pravastatin
therapy. Journal of Human Genetics, 51, 822826.
57. Igel, M., Arnold, K. A., Niemi, M., Hofmann, U., Schwab, M., Lutjohann, D., etal. (2006).
Impact of the SLCO1B1 polymorphism on the pharmacokinetics and lipid-lowering efficacy
of multiple-dose pravastatin. Clinical Pharmacology and Therapeutics, 79, 419426.
58. Zhang, W., Chen, B. L., Ozdemir, V., He, Y. J., Zhou, G., Peng, D. D., et al. (2007).
SLCO1B1 521TC functional genetic polymorphism and lipid-lowering efficacy of multi-
ple-dose pravastatin in Chinese coronary heart disease patients. British Journal of Clinical
59. Tachibana-Iimori, R., Tabara, Y., Kusuhara, H., Kohara, K., Kawamoto, R., Nakura, J., etal.
(2004). Effect of genetic polymorphism of OATP-C (SLCO1B1) on lipid-lowering response
to HMG-CoA reductase inhibitors. Drug Metabolism and Pharmacokinetics, 19, 375380.
60. Kivisto, K. T., Niemi, M., Schaeffeler, E., Pitkala, K., Tilvis, R., Fromm, M. F., etal. (2004).
Lipid-lowering response to statins is affected by CYP3A5 polymorphism. Pharmacogenetics,
14, 523525.
61. Kajinami, K., Brousseau, M. E., Ordovas, J. M., & Schaefer, E. J. (2004). CYP3A4 geno-
types and plasma lipoprotein levels before and after treatment with atorvastatin in primary
hypercholesterolemia. The American Journal of Cardiology, 93, 104107.
62. Bailey, K. M., Romaine, S. P., Jackson, B. M., Farrin, A. J., Efthymiou, M., Barth, J. H., etal.
(2010). Hepatic metabolism and transporter gene variants enhance response to Rosuvastatin
in patients with acute myocardial infarction The GEOSTAT-1 Study. Circulation.
Cardiovascular Genetics, 3, 276285.
63. Mega, J. L., Morrow, D. A., Brown, A., Cannon, C. P., & Sabatine, M. S. (2009).
Identification of genetic variants associated with response to statin therapy. Arteriosclerosis,
Thrombosis, and Vascular Biology, 29, 13101315.
64. Couvert, P., Giral, P., Dejager, S., Gu, J., Huby, T., Chapman, M. J., etal. (2008). Association
between a frequent allele of the gene encoding OATP1B1 and enhanced LDL-lowering
response to fluvastatin therapy. Pharmacogenomics, 9, 12171227.
65. McTaggart, F., Buckett, L., Davidson, R., etal. (2001). Preclinical and clinical pharmacology
of Rosuvastatin, a new 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitor. The
American Journal of Cardiology, 87, 28B32B.
66. Kim, K. A., Park, P. W., Lee, O. J., Kang, D. K., & Park, J. Y. (2007). Effect of polymorphic
CYP3A5 genotype on the single-dose simvastatin pharmacokinetics in healthy subjects.
Journal of Clinical Pharmacology, 47, 8793.
67. Wang, C. Y., Liu, P. Y., & Liao, J. K. (2008). Pleiotropic effects of statin therapy: Molecular
mechanisms and clinical results. Trends in Molecular Medicine, 14, 3744.
68. Nieminen, T., Kahonen, M., Viiri, L. E., Gronroos, P., & Lehtimaki, T. (2008). Pharma
cogenetics of apolipoprotein E gene during lipid-lowering therapy: Lipid levels and
prevention of coronary heart disease. Pharmacogenomics, 9, 14751486.
69. Regieli, J. J., Jukema, J. W., Grobbee, D. E., Kastelein, J. J., Kuivenhoven, J. A., Zwinderman,
A. H., etal. (2008). CETP genotype predicts increased mortality in statin-treated men with
proven cardiovascular disease: An adverse pharmacogenetic interaction. European Heart
Journal, 29, 27922799.
70. Kuivenhoven, J. A., Jukema, J. W., Zwinderman, A. H., de Knijff, P., McPherson, R.,
Bruschke, A. V., etal. (1998). The role of a common variant of the cholesteryl ester transfer
protein gene in the progression of coronary atherosclerosis. The Regression Growth
Evaluation Statin Study Group. The New England Journal of Medicine, 338, 8693.
71. Hindorff, L. A., Lemaitre, R. N., Smith, N. L., Bis, J. C., Marciante, K. D., Rice, K. M., etal.
(2008). Common genetic variation in six lipid-related and statin-related genes, statin use and
risk of incident nonfatal myocardial infarction and stroke. Pharmacogenetics and Genomics,
18, 677682.
72. Singer, J. B., Holdaas, H., Jardine, A. G., Fellstrom, B., Os, I., Bermann, G., et al. (2007).
Genetic analysis of fluvastatin response and dyslipidemia in renal transplant recipients. Journal
of Lipid Research, 48, 20722078.
73. Maitland-van der Zee, A. H., Peters, B. J., Lynch, A. I., etal. (2009). The effect of nine common
polymorphisms in coagulation factor genes (F2, F5, F7, F12 and F13) on the effectiveness
of statins: The GenHAT study. Pharmacogenetics and Genomics, 19, 338344.
74. Maitland-van der Zee, A. H., Lynch, A., Boerwinkle, E., etal. (2008). Interactions between
the single nucleotide polymorphisms in the homocysteine pathway (MTHFR 677C>T,
MTHFR 1298 A>C, and CBSins) and the efficacy of HMG-CoA reductase inhibitors in
preventing cardiovascular disease in high-risk patients of hypertension: The GenHAT study.
75. Libby, P. (2002). Inflammation in atherosclerosis. Nature, 420, 868874.
76. Sabatine, M. S., Ploughman, L., Simonsen, K. L., et al. (2008). Association between
ADAMTS1 matrix metalloproteinase gene variation, coronary heart disease, and benefit of
statin therapy. Arteriosclerosis, Thrombosis, and Vascular Biology, 28, 562567.
77. Iakoubova, O. A., Tong, C. H., Rowland, C. M., etal. (2008). Association of the Trp719Arg
polymorphism in kinesin-like protein 6 with myocardial infarction and coronary heart
disease in 2 prospective trials: The CARE and WOSCOPS trials. Journal of the American
College of Cardiology, 51, 435443.
78. Shiffman, D., Chasman, D. I., Zee, R. Y., Iakoubova, O. A., Louie, J. Z., Devlin, J. J., etal.
(2008). A kinesin family member 6 variant is associated with coronary heart disease in the
Womens Health Study. Journal of the American College of Cardiology, 51, 444448.
79. McKenney, J. M., Davidson, M. H., Jacobson, T. A., & Guyton, J. R. (2006). Final conclu-
sions and recommendations of the National Lipid Association Statin Safety Assessment Task
Force. The American Journal of Cardiology, 97, 89C94C.
80. Voora, D., Shah, S. H., Spasojevic, I., Ali, S., Reed, C. R., Salisbury, B. A., etal. (2009). The
SLCO1B1*5 genetic variant is associated with statin-induced side effects. Journal of the
American College of Cardiology, 54, 16091616.
81. Mulder, A. B., van Lijf, H. J., Bon, M. A., van den Bergh, F. A., Touw, D. J., Neef, C., etal.
(2001). Association of polymorphism in the cytochrome CYP2D6 and the efficacy and toler-
ability of simvastatin. Clinical Pharmacology and Therapeutics, 70, 546551.
82. Wilke, R. A., Moore, J. H., & Burmester, J. K. (2005). Relative impact of CYP3A genotype
and concomitant medication on the severity of atorvastatin-induced muscle damage.
83. Morimoto, K., Oishi, T., Ueda, S., Ueda, M., Hosokawa, M., & Chiba, K. (2004). A novel
variant allele of OATP-C (SLCO1B1) found in a Japanese patient with pravastatin-induced
myopathy. Drug Metabolism and Pharmacokinetics, 19, 453455.
84. Mohaupt, M. G., Karas, R. H., Babiychuk, E. B., etal. (2009). Association between statin-
associated myopathy and skeletal muscle damage. Canadian Medical Association Journal,
181, E11E18.
85. Draeger, A., Monastyrskaya, K., Mohaupt, M., Hoppeler, H., Savolainen, H., Allemann, C.,
et al. (2006). Statin therapy induces ultrastructural damage in skeletal muscle in patients
without myalgia. The Journal of Pathology, 210, 94102.
86. Staffa, J. A., Chang, J., & Green, L. (2002). Cerivastatin and reports of fatal rhabdomyolysis.
87. Ghatak, A., Faheem, O., & Thompson, P. D. (2010). The genetics of statin-induced myopathy.
Atherosclerosis, 210, 337343.
88. Rosenson, R. S. (2004). Current overview of statin-induced myopathy. The American
89. Wilke, R. A., Mareedu, R. K., & Moore, J. H. (2008). The pathway less traveled: Moving
from candidate genes to candidate pathways in the analysis of genome-wide data from large
scale Pharmacogenetic Association Studies. Current Pharmacogenomics and Personolized
90. Arora, R., Liebo, M., & Maldonado, F. (2006). Statin-induced myopathy: The two faces of
Janus. Journal of Cardiovascular Pharmacology and Therapeutics, 11, 105112.
91. Law, M., & Rudnicka, A. R. (2006). Statin safety: A systematic review. The American
Journal of Cardiology, 97, 52C60C.
92. Phillips, P. S., Haas, R. H., Bannykh, S., Hathaway, S., Gray, N. L., Kimura, B. J., et al.
(2002). Statin-associated myopathy with normal creatine kinase levels. Annals of Internal
Medicine, 137, 581585.
93. Moosmann, B., & Behl, C. (2004). Selenoprotein synthesis and side-effects of statins.
Lancet, 363, 892894.
94. Sinzinger, H., Schmid, P., & OGrady, J. (1999). Two different types of exercise-induced
muscle pain without myopathy and CK-elevation during HMG-Co-enzyme-A-reductase
inhibitor treatment. Atherosclerosis, 143, 459460.
95. Franc, S., Dejager, S., Bruckert, E., Chauvenet, M., Giral, P., & Turpin, G. (2003). A com-
prehensive description of muscle symptoms associated with lipid-lowering drugs.
Cardiovascular Drugs and Therapy, 17, 459465.
96. Sewright, K. A., Clarkson, P. M., & Thompson, P. D. (2007). Statin myopathy: Incidence,
risk factors, and pathophysiology. Current Atherosclerosis Reports, 9, 389396.
97. Dirks, A. J., & Jones, K. M. (2006). Statin-induced apoptosis and skeletal myopathy.
American Journal of Physiology. Cell Physiology, 291, C1208C1212.
98. Hanai, J., Cao, P., Tanksale, P., Imamura, S., Koshimizu, E., Zhao, J., et al. (2007). The
muscle-specific ubiquitin ligase atrogin-1/MAFbx mediates statin-induced muscle toxicity.
The Journal of Clinical Investigation, 117, 39403951.
99. Hansen, K. E., Hildebrand, J. P., Ferguson, E. E., & Stein, J. H. (2005). Outcomes in 45
patients with statin-associated myopathy. Archives of Internal Medicine, 165, 26712676.
100. Marcoff, L., & Thompson, P. D. (2007). The role of coenzyme Q10 in statin-associated myo-
pathy: A systematic review. Journal of the American College of Cardiology, 49, 22312237.
101. Phillips, P. S., & Haas, R. H. (2008). Statin myopathy as a metabolic muscle disease. Expert
102. Yokoyama, M., Seo, T., Park, T., Yagyu, H., Hu, Y., Son, N. H., et al. (2007). Effects of
lipoprotein lipase and statins on cholesterol uptake into heart and skeletal muscle. Journal of
Lipid Research, 48, 646655.
103. Paiva, H., Thelen, K. M., Van Coster, R., Smet, J., De Paepe, B., Mattila, K. M., etal. (2005).
High-dose statins and skeletal muscle metabolism in humans: A randomized, controlled trial.
104. Urso, M. L., Clarkson, P. M., Hittel, D., Hoffman, E. P., & Thompson, P. D. (2005). Changes
in ubiquitin proteasome pathway gene expression in skeletal muscle with exercise and statins.
Arteriosclerosis, Thrombosis, and Vascular Biology, 25, 25602566.
105. Guis, S., Figarella-Branger, D., Mattei, J. P., etal. (2006). In vivo and invitro characteriza-
tion of skeletal muscle metabolism in patients with statin-induced adverse effects. Arthritis
and Rheumatism, 55, 551557.
106. Moosmann, B., & Behl, C. (2004). Selenoproteins, cholesterol-lowering drugs, and the
consequences: Revisiting of the mevalonate pathway. Trends in Cardiovascular Medicine,
14, 273281.
107. Vladutiu, G. D. (2008). Genetic predisposition to statin myopathy. Current Opinion in
Rheumatology, 20, 648655.
108. Vieitez, I., Teijeira, S., Miranda, S., San, M. B., & Navarro, C. (2010). Novel human
pathological mutations. Gene symbol: PYGM. Disease: McArdle disease. Human Genetics,
127, 114115.
109. Vladutiu, G. D., Simmons, Z., Isackson, P. J., Tarnopolsky, M., Peltier, W. L., Barboi, A. C.,
etal. (2006). Genetic risk factors associated with lipid-lowering drug-induced myopathies.
Muscle and Nerve, 34, 153162.
110. Oh, J., Ban, M. R., Miskie, B. A., Pollex, R. L., & Hegele, R. A. (2007). Genetic determi-
nants of statin intolerance. Lipids in Health and Disease, 6, 7.
111. Morisaki, T., Gross, M., Morisaki, H., Pongratz, D., Zollner, N., & Holmes, E. W. (1992).
Molecular basis of AMP deaminase deficiency in skeletal muscle. Proceedings of the
National Academy of Sciences of the United States of America, 89, 64576461.
112. Caso, G., Kelly, P., McNurlan, M. A., & Lawson, W. E. (2007). Effect of coenzyme q10 on
myopathic symptoms in patients treated with statins. The American Journal of Cardiology,
99, 14091412.
113. McGown, C. C., & Brookes, Z. L. (2007). Beneficial effects of statins on the microcirculation
during sepsis: The role of nitric oxide. British Journal of Anaesthesia, 98, 163175.
114. Knapp, A. C., Huang, J., Starling, G., & Kiener, P. A. (2000). Inhibitors of HMG-CoA
reductase sensitize human smooth muscle cells to Fas-ligand and cytokine-induced cell
death. Atherosclerosis, 152, 217227.
115. Guijarro, C., Blanco-Colio, L. M., & Ortego, M., etal. (1998). 3-Hydroxy-3-methylglutaryl
coenzyme a reductase and isoprenylation inhibitors induce apoptosis of vascular smooth
muscle cells in culture. Circulation Research, 83, 490500.
116. Cao, P., Hanai, J., Tanksale, P., Imamura, S., Sukhatme, V. P., & Lecker, S. H. (2009). Statin-
induced muscle damage and atrogin-1 induction is the result of a geranylgeranylation defect.
The FASEB Journal, 23, 28442854.
117. de Langen, J. J., & van Puijenbroek, E. P. (2006). HMG-CoA-reductase inhibitors and
neuropathy: Reports to the Netherlands Pharmacovigilance Centre. The Netherlands Journal
of Medicine, 64, 334338.
118. Ruao, G., Thompson, P. D., Windemuth, A., Seip, R. L., Dande, A., Sorokin, A., et al.
(2007). Physiogenomic association of statin-related myalgia to serotonin receptors. Muscle
and Nerve, 36, 329335.
119. Ruao, G., Windemuth, A., Seip, R. L., Wu, A. H. B., & Thompson, P. D. (2007).
Physiogenomics of statin safety and efficacy. Journal of Lipidology, 1, 444.
120. Grundy, S. M., Cleeman, J. I., Daniels, S. R., Donato, K. A., Eckel, R. H., Franklin, B. A.,
etal. (2005). Diagnosis and management of the metabolic syndrome: An American Heart
Association/National Heart, Lung, and Blood Institute Scientific Statement. Circulation,
112, 27352752.
121. Ruao, G., Windemuth, A., & Holford, T. (2005). Physiogenomics: Integrating systems
engineering and nanotechnology for personalized medicine. In JD Bronzino (Ed.), The
biomedical engineering handbook (3rd ed., pp. 28-128-9). Boca Raton: CRC Press.
122. Ruao, G. (2006). Physiogenomic method for predicting clinical outcomes of treatments in
patients. Patent 20060278241. December 14, 2006.
123. Ruao, G., Windemuth, A., Kocherla, M., Holford, T., Fernandez, M. L., Forsythe, C. E.,
et al. (2006). Physiogenomic analysis of weight loss induced by dietary carbohydrate
restriction. Nutrition & Metabolism, 3, 20.
124. Seip, R. L., Volek, J. S., Windemuth, A., Kocherla, M., Fernandez, M. L., Kraemer, W. J.,
etal. (2008). Physiogenomic comparison of human fat loss in response to diets restrictive of
carbohydrate or fat. Nutrition & Metabolism, 5, 4.
125. Ruao, G., Seip, R. L., Windemuth, A., Zollner, S., Tsongalis, G. J., Ordovas, J., et al.
(2006). Apolipoprotein A1 genotype affects the change in high density lipoprotein choles-
terol subfractions with exercise training. Atherosclerosis, 185, 6569.
126. Ruao, G., Goethe, J. W., Caley, C., Woolley, S., Holford, T. R., Kocherla, M., etal. (2007)
Physiogenomic comparison of weight profiles of olanzapine- and risperidone-treated
patients. Molecular Psychiatry, 12, 474482.
Chapter 11
The Statin Response Gene: KIF6
H. Robert Superko, Tom White, James Forrester, and Spencer King III
It is not good to settle into a set of opinions. At first putting

forth great effort to be sure that you have grasped the basics,
then practicing so that they may come to fruition is something
that will never stop for your whole lifetime. Do not rely on
following the degree of understanding that you have discovered,
but simply think,
This is Not Enough
Tsunetomo [1]
Keywords Statins Cholesterol Lipid lowering Kinesin proteins
11.1Statin Therapy, Clinical Events and the Need

for a New Approach
For over two decades, substantial National efforts have been directed toward
reducing blood cholesterol levels of the American population with the intent of
reducing the burden of cardiovascular events. This initiative was initially prompted
by the successful results of the Lipid Research Clinic-Coronary Primary Prevention
Trial (LRC-CPPT) that first proved reducing LDL-C resulted in a statistically
significant reduction in cardiovascular events [2]. Since then an abundance of
monotherapy cholesterol-lowering drug trials have consistently reported an approx-
imate 25% relative risk reduction (RRR) for cardiovascular events [3].
While a 25% RRR in large population studies is laudable, a danger lies in the
assumption that blood cholesterol reduction has similar beneficial effects in all
people. Indeed, the cholesterol-lowering clinical trials left large groups of
H.R. Superko(*)
Celera Corp. 1401 Harbor Bay Parkway, Alameda, CA 94502, USA
e-mail: HighHDL@mac.com
176 H.R. Superko et al.
patients still experiencing clinical events despite successful LDL-C reduction [4].
This has now been termed, residual risk. A clue that approximately 50% of
statin treated patients may not receive benefit from LDL-C reduction was reported
in 2003 utilizing single photon emission computed tomography (SPECT) myocar-
dial perfusion imaging (MPI) [5]. In this investigation, patients achieved a 32%
statin induced LDL-C reduction and the mean stress perfusion defect improved in
48% but was unchanged in 43% and worsened in 9% and change in lipid levels did
not correlate with the change in perfusion defect. This chapter will review the find-
ings regarding the impact of a polymorphism in the kinesin-6 gene that helps to
identify a large group of patients that have greater than previously thought clinical
event reduction benefit from statin treatment, and, a second group of patients that
obtain significantly less benefit despite identical LDL-C reduction.
11.1.1Blood Cholesterol Lowering History
In the twenty-first century, atherosclerosis is well established as the leading cause of

death in most industrialized nations [6]. Large population based studies have identi-
fied several risk factors as targets of intervention that may reduce CV disease [7, 8].
This approach to identification of high risk populations was used in order to focus
efforts on a population subset that could derive the most benefit from a treatment
targeted to disorders such as hypertension or elevated blood cholesterol [3, 9, 10].
In the field of cholesterol and heart disease research, early efforts with thyroid,
estrogen, and clofibrate provided some evidence that cholesterol lowering was of
benefit but were marred by adverse drug interactions [11, 12]. The LRC-CPPT that
used a combination of moderately low-fat diet and cholestyramine in men with hyper-
cholesterolemia, first established that reducing elevated LDL-C resulted in fewer
coronary heart disease (CHD) events (significant by a one tailed t-test) [2]. Following
the success of the LRC-CPPT, HMGCoA reductase inhibitor medications (statins)
became available that achieved greater LDL-C reduction with fewer side effects [13].
These remarkable compounds made achieving reduced LDL-C values relatively easy.
Their success forms the basis of the paradox that easily achieved LDL-C reduction
may, in part, be responsible for the failure to substantially stem the tide of CHD.
Based on multiple clinical trials, the ATP-I recommended a LDL-C goal of
<130mg/dl to assist CHD risk reduction [14]. Since the publication of ATP-I in 1988,
multiple other clinical trials have reported that even greater LDL-C reduction in both
primary and secondary prevention populations can achieve even greater reductions in
CHD relative risk [15]. The success of these trials prompted ATP-II and ATP-III to
adjust LDL-C goals downward to the current recommended levels [3].
One danger of a medical therapy, deemed to be statistically significant in large
clinical trials, is the tendency to assume that almost all patients will benefit in a
similar manner from that single therapy, and thus ignore other potentially beneficial
treatments. Recently, the results of a series of statin-induced cholesterol-lowering
trials have been used to suggest that a new LDL-C goal of <70mg/dl should be
11 The Statin Response Gene: KIF6 177
Percent Clinical Events 30

% Control
25
% Treatment
20
15
10
5
0
T
D
SS
E
IT
PS
X
TN
AR
PI
LR
TE
VE
SS
O
LI
T-
S/
SC
O
PP
AP
PR
O
C
C
W
AF
Fig. 11.1 Percent clinical events reported in monotherapy statin clinical trials in the control group
compared to the treatment group. While an approximate 25% relative risk difference was obtained,
significant numbers of patients treated with statin therapy continued to have cardiovascular events
(Adapted from [4])
nationally embraced [16]. However, a large number of patients taking cholesterol-

lowering medications, and achieving lower LDL-C values, continued to have clini-
cal events (Fig. 11.1) [4]. This continues to be true even with the substantially
greater absolute LDL-C reduction achieved in recent trials. At this point in the his-
tory of cholesterol reduction, it is important to pause and discuss the possibility that
subgroups of patients can be identified that respond to a greater or lesser extent in
regard to clinical event reduction attributed to statin therapy.
11.1.2Clinical Relevance of RRR, ARR, and NNT
Many individuals in clinical trials have cardiovascular events despite statin treat-
ment that was successful in reducing LDL-C levels [4]. The effectiveness of a
treatment can be described by three concepts, relative risk reduction (RRR), absolute
risk reduction (ARR), and number needed to treat (NNT) (Table 11.1). RRR
assesses the reduction in risk in one group relative to another group. For example,
the risk reduction noted in the treatment group compared to the placebo group.
Therefore, if the total study size is 2,000 placebo and 2,000 treatment subjects, and
100 placebo patients have an event (5%) relative to 75 events in the treatment
group (3.8%), the RRR is 25% (25/100).
ARR assesses the absolute reduction difference in risk in one group compared to
the absolute reduction in another group. In the example above, the ARR would be
1.2% (5.03.8%). It is not uncommon for the public to misinterpret this 25% RRR as
meaning that 25% of the entire population was saved from an event due to the
treatment. In fact, if there were 1,000 subjects in the treatment group and 1,000
subjects in the placebo group, and 100 events were experienced in the placebo group
Table 11.1 Relative risk reduction (RRR), absolute risk reduction (ARR), and number needed to
treat (NNT) in statin-CHD trials in populations separated by KIF6 carrier status
Study RRR (%) ARR (%) NNT
CARE Original 24 3.0 34
CARE Genetic Substudy
KIF6 Carriers 37 4.9 20
KIF6 Noncarriers 20 1.4 72
WOSCOPS Original 31 3.5 46
WOSCOPS Genetic Substudy
KIF6 Carriers 50 5.5 18
KIF6 Noncarriers 9 0.1 >100
PROSPER Original 15 2.1 47.6
PROSPER Genetic Substudy
KIF6 Carriers 33.6 6.3 16
PROVE IT-TIMI 22 Original 16 3.9 26
PROVE IT Genetic Substudy
KIF6 Carriers 41 10 10
and 75 events in the treatment group, the difference between 100 and 75 is the 25%
RRR in events (n=25), not 25% of 1,000 subjects (n=250).
Numbers needed to treat (NNT) is a method that can assess the efficiency of
different therapies. NNT is the number of subjects that are needed to be treated in
order to prevent one event in a defined time period and is the inverse of the ARR. In
general, the higher the NNT, the less efficient and the lower the NNT, the greater the
efficiency of the treatment. For example, in the example above, it was necessary to
treat 2,000 subjects in order to prevent 25 events. Thus, the NNT is (100/1.2) or 83.
A lower NNT maximizes the benefits of the treatment, while lowering the overall
exposure of the population who derive little or no benefit from therapy. This has
implications for both population-based costs of therapy, risk: benefit, and sample
sizes in future trials.
11.1.3Atherosclerosis Pathophysiology
Not just obstruction. The impact of atherosclerosis on vascular disease can be quan-
titatively expressed either anatomically as the magnitude of vascular obstruction as
assessed by arteriography or, as the absolute number of clinical events. In the past,
it was thought that the amount of arterial obstruction correlated with the potential for
clinical events. This relationship between anatomic obstruction and clinical events is
tenuous, however, because in most individuals experiencing a coronary event the
culprit lesion caused less than a 50% stenosis [17]. Clinical events are caused by
plaque rupture with subsequent thrombosis, and rupture is largely independent of the
magnitude of stenosis [18, 19]. This important concept helps to explain why large
numbers of patients appearing in the emergency department with an acute coronary

event have no prior symptoms [20]. While severely stenotic lesions can rupture and
thrombose, approximately twice as many events are attributed to moderately stenotic
(<80%) lesions. In recognition of this, the process is now referred to as athero-
thrombosis and the plaque itself is referred to as a vulnerable plaque that is prone
to rupture or erosion that can result in a thrombotic event [21].
Classic well established CHD risk factors such as elevated blood cholesterol,
hypertension, and smoking are associated with the development of atheroma. The
factors that precipitate plaque rupture are less well defined [22]. Since the etiology
of plaque rupture may be related to factors other than the classic risk factors, it
should not be surprising that many patients can significantly improve a risk factor
such as elevated LDL-C and yet suffer a cardiovascular event. It is important to
distinguish the slow gradual process of atherosclerosis development and progres-
sion, from the factors that may contribute to an unstable plaque prone to rupture and
cause a cardiovascular event. Genetic polymorphisms may play a role in determining
risk for an event, and identifying response to a specific therapy.
11.2Kinesin Proteins
To sustain their specific functions and morphology, cells have an intracellular trans-
port mechanism controlled by microtubule-dependent motor proteins [23].
Microtubules are intracellular structures and play specialized roles in a variety of
cellular events including molecular transport which is dependent on proteins such
as kinesins [24]. People walk by alternating left and right movements of the legs.
Recent work shows that kinesins share the hallmarks of bipedal walking [25]. They
walk intracellular cargo along the microtubules with each alternating foot taking
an alternating 16nm step [26]. Recently, all KIF genes in the mammalian and
human genomes have been systematically identified [27]. The kinesin superfamily,
is composed of at least 14 families, based on phylogenetic analysis and each may
transport different cargoes. The most common cargo for N-kinesins are membra-
nous structures [28]. For example, Kinesin II transports a large protein complex that
contains the tumor suppressor adenomatous polyposis coli protein (APC). The APC
protein is commonly mutated in colon cancer, giving rise to versions of APC that
do not function properly.
KIF6 encodes a kinesin, a class of motor proteins involved in the intracellular
transport of cargo that includes membrane organelles, protein complexes, and
mRNAs [27, 29, 30]. The kinesin proteins consist of a conserved motor domain that
propels the kinesin along microtubules in an ATP-dependent manner and a noncon-
served tail domain that binds to either directly its cargo or to other cargo-binding
proteins. The tail domain of kinesin is responsible for cargo recognition and binding
(Fig. 11.2) [30]. Thus, this amino acid change might affect the cargo binding or
transporting activity of the kinesin protein encoded by KIF6. Interestingly, KIF6 is
expressed in coronary arteries and several other kinesins have been implicated in the
Cargo
Tail
KIF6 Trp719Arg Two necks
Stalk
Motor domains
Microtubule
Fig. 11.2 The kinesin 6 protein walks along microtubules dragging a cargo. The KIF6 Trp719Arg
polymorphism is located at the binding site of kinesin 6 protein
response of cardiac stroke volume to regular exercise and in the pathogenesis of

chronic diseases, such as neurodegenerative diseases, type 2 diabetes, and Alzheimers
disease [3133]. Functional biology studies of KIF6 in coronary artery disease, and
a potential role in vulnerable plaque, are underway using antibodies that can detect
the KIF6 protein in histopathological specimens. The polymorphism Trp719Arg in
the kinesin-like protein 6 has been associated with the presence of late outgrowth
endothelial progenitor cells in acute myocardial infarction [34]. Like the chromo-
some 9p21 single nucleotide polymorphisms (SNPs) which are also not yet charac-
terized in terms of functional biology, an understanding of the biological process(es)
involved promise to provide new insight into the pathology of CHD [35].
11.3Discovery of KIF6 and Relationship to CHD risk
The initial discovery of the KIF6 719Arg allele and the relationship to CHD was
first noted in a case-control study of myocardial infarction patients at the University
of California, San Francisco [36]. The allele frequencies of 11,053 SNPs were
determined in pooled DNA and the frequency of KIF6 719Arg was higher (p <0.05)
in cases compared to controls. This discovery study was then followed-up by
assessing in placebo groups of the two prospective randomized clinical trials CARE
and WOSCOPS the association with CHD of 10 SNPs in 9 candidate genes,
reported in the literature to be associated with CHD and twenty-five SNPs that
Celera had found to be associated with MI in the case control study. Of the 35 SNPs
tested, only KIF6 719Arg was significantly associated with increased risk for CHD
(p <0.05 after Bonferroni correction).
11.4KIF6 719Arg Polymorphism is Associated

with Coronary Heart Disease Event Risk
The association of the KIF6 719Arg polymorphism with cardiovascular risk has
been established in seven prospective studies, in over 50,000 subjects, including
the Atherosclerosis Risk in Communities (ARIC) Study, the Womens Health
Study (WHS), the Cardiovascular Health Study (CHS), and the placebo arms of
the West of Scotland Coronary Prevention Study (WOSCOPS), the Cholesterol
and Recurring Events (CARE) study and the PROspective Study of Pravastatin in
the Elderly at Risk (PROSPER) study. In the placebo arm of these clinical trials,
the effect of a polymorphism impacted by statin therapy could be fully appreciated
since randomization to placebo assured that statin use would not impact the analy-
sis. In the community free-living populations of ARIC, WHS, and CHD, statin use
in the community has the potential to blunt the ability of the KIF6 polymorphism
to identify risk due to the existing treatment of a portion of the population with a
statin medication.
11.4.1The KIF6 Polymorphism is Associated with Coronary

Heart Disease in the ARIC Study
The ARIC study is a prospective investigation of atherosclerosis and its clinical

sequelae involving 15,792 individuals aged 4564 years at recruitment [37, 38].
CHD cases (n = 1,452) were defined as participants with either myocardial
infarction (MI), CHD death, or coronary revascularization procedures. In this
study, carriers of two copies of the KIF6 719Arg risk variant were at increased
risk for incident CHD with a hazard ratio of 1.22 in a model adjusted for age and
sex [39].

Heart Disease in the Womens Health Study
The WHS is a large prospective, placebo controlled, 22 factorial designed trial

of aspirin and vitamin E in the prevention of cardiovascular disease [40]. The study
was conducted among 26,274 initially healthy women aged 45 years or older who
were followed for approximately 13 years for cardiovascular events including
c ardiovascular death, myocardial infarction (MI), ischemic stroke, and

revascularization procedures. In Caucasians, carriers of the KIF6 719Arg risk
variant were at greater risk of CHD with a hazard ratio of 1.24 and associated with
MI with a hazard ratio of 1.34. The risk for ischemic stroke did not differ between
carriers of the KIF6 719Arg risk variant and noncarriers. In addition to long follow-
up time, extensive baseline information was collected for all participants, including
traditional risk factors and emerging risk factors for cardiovascular disease [41]. No
substantial or significant difference was detected between KIF6 719Arg carriers
and noncarriers in regard to baseline values and traditional CHD risk factors, sug-
gesting that the risk conferred by the KIF6 risk variant is not only independent of
traditional risk factors, but also that of the other risk factors measured, such as
ApoA, ApoB, CRP, ICAM, and Fibrinogen.

Heart Disease in the Cardiovascular Health Study
The KIF6 719Arg risk variant was also associated with disease in the CHS, a pro-
spective study of risk factors for cardiovascular disease for men and women aged
65 years and older [42, 43]. CHD was defined as nonfatal MI or definite fatal MI,
angina pectoris, angioplasty or coronary artery bypass surgery or death due to
atherosclerotic CHD. In the genetic substudy of CHS, the CHD endpoints were
fatal and nonfatal MI. In Caucasian patients for whom genetic analysis was pos-
sible, 539 (12%) of 4,522 CHS participants had an MI during the 13 years of study
follow-up for a rate of 11.9 incident events per 1,000 person-years. The KIF6
719Arg risk variant was associated with events with a hazard ratio of 1.29 in a
model adjusted for age, sex, race, and traditional risk factors such as LDL-C and
hs-CRP [44].

Heart Disease in the Placebo Arms of the CARE
and WOSCOPS Studies
The prospective statin trials, CARE [45] and WOSCOPS [46], assessed the effect
of pravastatin in the prevention of myocardial infarction (MI) and CHD events. In
the placebo arm of both trials, Iakoubova et al. [47] assessed the association of
KIF6 with risk for coronary events. The genetic study of CARE comprised 2,913
Caucasian patients and assessed risk for fatal or nonfatal MI. The genetic study of
WOSCOPS, which was derived from a previously published (prospective) nested
case-control study [48], included the 481 on-trial CHD events as cases and the
1,086 controls for whom DNA was available for analysis. As previously reported
for the WOSCOPS nested study, the controls were matched to cases for age and
current smoking status and the CHD endpoint studied was death from CHD,
nonfatal MI, or revascularization procedures.
In the placebo arm of CARE, carriers of the KIF6 719Arg variant had a hazard
ratio of 1.50 for recurrent myocardial infarction in a model adjusted for (and inde-
pendent of) age, sex, smoking, history of hypertension, history of diabetes, body
mass index, LDL-C, and HDL-C. This hazard ratio for the KIF6 719Arg risk
variant was not only found to be independent of these traditional risk factors, but
also to predict risk of a similar magnitude compared to the traditional risk factors.
In the placebo arm of WOSCOPS, carriers of the KIF6 719Arg variant had a
similar odds ratio for CHD of 1.55 in a model adjusted for history of hypertension,
history of diabetes, body mass index, LDL-C, and HDL-C. The WOSCOPS and
CARE studies provide information in primary prevention patients with elevated
LDL-C (mean LDL-C = 192 mg/dl, WOSCOPS) and secondary prevention
patients with moderate LDL-C levels (mean LDL-C=139mg/dl, CARE). Thus,
an odds ratio of approximately 1.5 has been established and reproduced in the
placebo arms of two classic statin studies.

Heart Disease in the PROSPER Study
In the placebo arm of PROSPER, carriers of a single KIF6 719Arg risk variant,
with a prior history of vascular disease, were at increased risk for coronary events
with a hazard ratio of 1.36 in a model adjusted for sex, age, smoking, hyperten-
sion, diabetes, LDL-C, HDL-C, and country of recruitment [49]. Carriers of the
KIF6 719Arg variant, without prior vascular disease, were not associated with
increased risk of coronary events, a finding that is consistent with results from the
original trial.
11.4.6KIF6 Polymorphism Predicts Risk for Coronary Heart

Disease as Shown in Six Prospective Studies
The preponderance of published data strongly favors the conclusion that the KIF6
variant is associated with CHD risk as seen in the six prospective studies incor-
porating >50,000 subjects (Fig. 11.3). Three of the studies, CARE, WOSCOPS,
and PROSPER, were in the placebo arms of statin clinical trials, while the other
three studies included some subjects who were treated by their physician with a
statin or other CHD therapies. In the placebo arms of the statin trials, the KIF6
719Arg variant was associated with an approximate 50% increase in risk. Some
subjects in the free living population studies (CHS, ARIC and WHS), were
treated with statin medications by their physicians which may help to explain the
lower hazards ratios observed in the three population studies since, as described
CARE
PROSPER Untreated patients
WOSCOPS
CHS
ARIC Some treatment
WHS
0.5 1 1.5 2 2.5

Hazard Ratio*
Fig. 11.3 KIF6 variant predicts risk for coronary heart disease (CHD) in six prospective studies
with ~50,000 participants. KIF6 variant is associated with a significant increase in the hazard ratio
for MI, recurrent MI, and CHD in six large clinical investigations following adjustment for tradi-
tional risk factors. PROSPER patients with prior vascular disease
in the next section, the KIF6 719Arg variant predicts both risk for CHD and
response to statin therapy.
11.4.7KIF6 Risk Compared to Other Risk Markers
The clinical utility of KIF6 testing as a marker of CHD risk can be compared to the
utility of traditional risk factors in the placebo group of clinical trials. In the placebo
arm of CARE, following statistical adjustment for all the risk variables, the KIF6
variant was an independent and significant predictor of CHD risk and more
informative than age, HDL-C <40 mg/dl, hypertension, or LDL-C >130 mg/dl
(Fig. 11.4). In the placebo arm of WOSCOPS, following statistical adjustment for
all the risk variables, the KIF6 variant was an independent and significant predictor
of CHD risk and more informative than HDL-C <40mg/dl, LDL-C >189mg/dl,
and hypertension (Fig. 11.5). Both the CARE and WOSCOPS populations may be
considered at high risk for CHD events. CARE subjects had a prior history of CHD
and WOSCOPS subjects had elevated LDL-C.
11.5Lack of KIF6 polymorphism Associated

with CHD Risk in Some Studies
Some investigations have not found an association of the KIF6 polymorphism with
CHD risk and may be related to clinical trial design issues. In the Wellcome Trust
Case Control Consortium (WTCCC) study, KIF6 was not found to be associated with
increased CHD risk [50]. In the case control designed Ottawa Heart Study, KIF6 was
not associated with CHD as defined by an arteriographic endpoint [35, 51].
Smoking
Diabetes
KIF6 Variant
Age 55
HDL-C < 40
Hypertension
LDL-C 130
0.5 1 1.5 2 2.5 3

Adjusted Hazard Ratio
Fig. 11.4 Risk in CARE placebo arm KIF6 variant and traditional risk factors. KIF6 variant and
hazard ratio in relationship to classic cardiovascular risk factors. The hazard ratio is statistically
adjusted for the traditional risk factors and is independent of those factors
Diabetes
KIF6 Variant
HDL-C <40
LDL-C 189*
Hypertension
0.5 1 1.5 2 2.5 3

Adjusted Risk Ratio
Fig. 11.5 Risk of CHD in West of Scotland Coronary Prevention Study (WOSCOPS) placebo arm
KIF6 variant and traditional risk factors. In the placebo arm of the West of Scotland trial, the KIF6
variant was similar in magnitude to traditional risk factors and more informative than HDL-C
<40mg/dl, LDL-C >189 mg/dl, and hypertension. In this prospective nested case and control study,
patients were matched for age and smoking, and all were men. *Median level in placebo arm
The WTCCC study tested 377,857 SNPs for association with MI [50]. To adjust
for multiple testing of so many SNPs in a discovery study, a large significant value
of p <1.5108 was used. However, a significant association between the 719Arg
allele and CHD was observed in a sex-differentiated test. Among females, the
719Arg allele was not associated with increased risk of CHD (OR=0.91, 95% CI;
0.721.14). Among males, the 719Arg allele was associated with increased risk of
CHD (OR=1.18, 95% CI; 1.021.36). If 20% of the population from which the
WTCCC cases were drawn was receiving statin therapy prior to their CHD event,
the estimated risk of CHD among untreated cases for 719Arg carriers compared
with noncarriers, would have been modestly higher (OR=1.23, 95% CI; 1.021.42)
[52]. When KIF6 was tested as a prespecified hypothesis based on its previous
association with CHD risk in six prior studies, it was found to be associated with
risk in men [53].
The Ottawa Heart Study is a cross-sectional, case control design study investi-
gating the relationship of CAD as defined by arteriography in patients (onset
<55 years males, <65 years females) seen in the coronary catheterization laboratory
and lipid clinic (cases n = 1,540), with asymptomatic control subjects (controls
n=1,455) recruited from an elderly population without a cardiovascular disease
history (>65 years males, >70 years female) [51]. In this investigation, the KIF6
719Arg variant was not associated with risk of CAD by this definition. These
negative results could also be explained by the fact that 89% of the cases were on
statin therapy, which would have eliminated an association with risk in KIF6
carriers. The result of the Ottawa Heart Study supports the concept that the KIF6
variant is associated with clinical events such as MI rather than arteriographically
defined CAD.
11.6Pharmacogenomics
In addition to predicting risk, a genetic test can assist in clinical decisions, particu-
larly in regard to drug type or dosage. Examples include the dose of coumadin and
polymorphisms in the CYP450 2C9 and VKORC1 genes, and CYP450 2C19 and
the use of clopidogrel. The CYP2C9*3 allele and VKORC1 381 CC and TC geno-
types have been reported to explain 60.2% of the variability in daily coumadin dose
requirements [54]. Simple equations have been developed to assist the physician in
selecting the most effective individualized dose based on genotype differences [55].
Using such knowledge, it has been demonstrated that approximately 60% of adverse
outcomes attributed to coumadin could be avoided by utilizing this genetic knowl-
edge to determine the most appropriate dose for a given patient [56]. In 2007 the
Food and Drug Administration placed a boxed warning on coumadin that indicates
use of these genotypes is of clinical value (http://www.fda.gov/NewsEvents/
Newsroom/PressAnnouncements/2007/ucm108967.htm). The CYP 2C19 polymor-
phism is found in approximately 1/3 of the population and variants may lead to a
slow or rapid rate of clopidogrel metabolism. Carriers of the reduced function allele
have reduced platelet aggregation in response to clopidogrel and were reported to
have a relative 53% increase in endpoints in the Trial to Assess Improvement in
Therapeutic Outcomes by Optimizing Platelet Inhibition with Prasugrel-Thrombolysis
in MI (Triton-TIMI 38) [57]. The Hazard ratio for stent thrombosis increased
threefold in carriers of the reduced function allele. Similarly, the KIF6 719Arg poly-
morphism can assist physicians in lipid therapy selection that has the best chance of
improving clinical outcomes for individual patients.
In addition to independently predicting CHD risk, the KIF6 719Arg
polymorphism also predicts the response to statin therapy in regard to clinical
event reduction independent of LDL-C, HDL-C, triglyceride, and nonHDL-C
values. This has been demonstrated in several well-known randomized clinical
trials including CARE, WOSCOPS, PROSPER, and PROVE IT-TIMI 22. In these
trials, the presence of the KIF6 719Arg polymorphism identified a substantial and
statistically significant reduction of cardiac events in the statin-treated group that
was not related to changes in any of the lipid parameters including LDL. Implying
a mechanism of action unrelated to classic lipoprotein metabolism.
11.6.1KIF6 719Arg Polymorphism Predicts Response

to Statin Therapy for Coronary Events in the CARE
and WOSCOPS Trials
Multiple primary and secondary prevention trials have shown that statin therapy is
beneficial in preventing coronary death and serious coronary events with ARRs of
34% and RRRs of 2530% compared to placebo over a 5-year period. Statin
therapy also reduces CHD events in some individuals who received less benefit in
terms of LDL-C lowering. Identification of those individuals who may benefit from
statin therapy despite less than optimal LDL-C lowering may aid with treatment
compliance and reduce morbidity and mortality. Further, statins have an acceptable
safety profile, but they also have dose-related side effects. Thus, it may be advanta-
geous for physicians to use an individuals genetic information to help determine
the risk/benefit of aggressive vs. standard statin treatment.
Since the KIF6 719Arg risk variant was associated with both risk for MI in
CARE and CHD in WOSCOPS, Iakoubova etal. [47] asked whether carriers of
the KIF6 risk variant benefited from pravastatin treatment. In CARE, pravastatin
treatment reduced the relative risk of MI by 37% in carriers of the KIF6 719Arg
risk variant, and in WOSCOPS, pravastatin treatment reduced the relative risk of
CHD by 50% among KIF6 719Arg Carriers. In CARE, when genotype was not
considered, the ARR was 3.5% (Fig. 11.6). When KIF6 genotype was considered
in the genetic substudy, the ARR by pravastatin was 4.9% for KIF6 719Arg carri-
ers of the 719Arg risk variant and 1.4% for noncarriers. The NNT in CARE to
prevent one coronary event in the original study was 34 for all patients, yet the
NNT was only 20 for KIF6 719Arg carriers compared with 72 for noncarriers. In
WOSCOPS, when genotype was not considered, the ARR was 3.5%. When KIF6
genotype was considered, the ARR by pravastatin was 5.5% for carriers of the
CARE WOSCOPS
All Carriers Noncarriers All Carriers Noncarriers

0
2
Absolute Risk
Reduction 4
(%)
6
8
3.5% 4.9% 1.4% 3.5% 5.5% 0.1%
P = 0.005 P < 0.0001
Fig. 11.6 Differential CHD Reduction by Statin Rx KIF6 719Arg Carriers received the most
benefit. Differential reduction in CHD events, in response to statin treatment, based on KIF6
719Arg Carrier status. In WOSCOPS, risk reduction was significantly greater in carriers than in
noncarriers (p = 0.01)
KIF6 risk variant (p=0.0001), and 0.1% for noncarriers. The NNT in WOSCOPS
was 46 for all patients, yet only 18 for KIF6 719Arg Carriers compared to >100
for noncarriers.
11.6.2KIF6 Predicts Response to Statin Therapy in Elderly

Patients with Prior Vascular Disease in the PROSPER
Trial
The elderly are a special population with regards to risk for coronary events and may
require prevention and treatment strategies that differ from those appropriate for
middle-aged patients. For example, there is evidence that plasma cholesterol levels
do not predict risk of coronary events as well in the elderly as they do in middle-aged
populations [5861]. Nevertheless, observational and randomized clinical studies of
statins in elderly populations have shown significant reductions of mortality and
nonfatal coronary events, particularly in those with prior vascular disease [38, 5862].
Since risk for coronary events or response to therapy may differ between the elderly
and the general population, we asked whether the KIF6 719Arg risk variant could also
predict both risk for coronary events and clinical event response to statin therapy
in the PROSPER trial, as had previously been demonstrated in the CARE and
WOSCOPS studies.
KIF6 719Arg Carriers Noncarriers

Placebo
20 20
CHD Events (%)
15 HR=0.66 P=0.002 15 HR=0.94 P=0.64
10 10
5 Pravastatin 5
0 0
0 6 12 18 24 30 36 0 6 12 18 24 30 36
Months of follow up Months of follow up
Fig. 11.7 Differential CHD reduction by statin Rx KIF6 719Arg carriers received the most benefit
in PROspective Study of Pravastatin in the Elderly at Risk (PROSPER). Differential effect of statin
treatment on fatal and nonfatal CHD events in PROSPER subjects based on KIF6 719Arg carrier
status. Among PROSPER patients with prior vascular disease, carriers of KIF6 Arg risk allele
received significant (p = 0.002) reduction (34%) in coronary events in response to pravastatin
treatment vs. placebo. Among patients without prior vascular disease there was no significant event
reduction
The PROSPER prospective trial investigated the effects of pravastatin therapy on

reduction of fatal and nonfatal cardiovascular events in elderly patients (7082
years old) with and without prior cardiac and peripheral vascular disease [62]. The
study cohort consisted of 5,804 patients (48% men) randomized to treatment with
either 40 mg pravastatin per day or placebo and followed for an average of
3.2 years. In this study, a significant reduction of coronary events was observed
only among the 42% of patients with prior vascular disease treated with pravastatin,
but not among those patients without a history of prior vascular disease.
In the genetic substudy of PROSPER, carriers of the KIF6 719Arg risk variant
with prior vascular disease received substantial and significant reduction of
coronary events from statin therapy [63]. Among PROSPER patients with prior
vascular disease on pravastatin therapy, the ARR was 6.3% in 719Arg carriers vs.
1.2% in noncarriers. The NNT with pravastatin in KIF6 719Arg Carriers, was 16
compared to 83 in noncarriers. The RRR for coronary events for KIF6 719Arg
carriers was 33.6% (HR 0.66, 95% CI 0.520.86). However, noncarriers received
no significant risk reduction from pravastatin therapy (HR 0.94, 95% CI 0.69
1.28) (Fig. 11.7). Among those without prior vascular disease, neither the KIF6
719Arg carriers nor noncarriers benefited from statin therapy in terms of clinical
event reduction, a finding consistent with the original PROSPER trial where sig-
nificant event reduction with statin therapy was observed only in subjects with
prior vascular disease [62]. Interestingly, in patients with a history of prior vascu-
lar disease, KIF6 719Arg carriers and noncarriers received nearly identical reduc-
tions in LDL-C on pravastatintherapy suggesting that the risk reduction observed
in KIF6 719Arg carriers may be due to a mechanism that is independent of LDL-C
lowering.
11.6.3KIF6 Predicts Response to Statin Therapy After Acute

Coronary Syndromes in the PROVE IT-TIMI 22 Trial
The Pravastatin and Atorvastatin Evaluation and Infection Therapy Thrombosis

in Myocardial Infarction 22 (PROVE IT-TIMI 22) study compared high-dose
atorvastatin with standard-dose pravastatin treatment [64]. A genetic substudy of
this cohort explored whether high-dose atorvastatin therapy, compared to pravastatin
therapy, would significantly reduce coronary events in carriers of the KIF6 719Arg
risk variant and whether the clinical benefit would be greater in carriers than in
noncarriers [65].
The original PROVE IT-TIMI 22 study comprised 4,162 patients, who had been
hospitalized for an acute coronary syndrome within 10 days preceding enrollment
and randomized to high-dose atorvastatin (80mg/day) compared to standard-dose
pravastatin (40 mg/day) and to gatifloxacin vs. placebo using a double blind,
two-by-two factorial design. Genetic analysis was possible in 1,777 Caucasian
patients and the endpoint for the genetic study was the same as the original PROVE
IT-TIMI 22 study: death from any cause or major cardiovascular events, which
included myocardial infarction, documented unstable angina requiring hospitaliza-
tion, revascularization with either percutaneous coronary intervention or coronary-
artery bypass grafting, and stroke.
In the genetic study, high-dose atorvastatin therapy significantly reduced coro-
nary events in carriers of the KIF6 719Arg risk variant and this clinical benefit was
6.8-fold greater in KIF6 719Arg carriers than in noncarriers [65]. When genotype
was not considered in this genetic study, the RRR for coronary events was 27%,
favoring high-dose atorvastatin. When genotype was considered, in KIF6 719Arg
carriers, the RRR for death of any cause or major cardiovascular events was 41%
Fig. 11.8 Differential CHD reduction by statin Rx KIF6 719Arg carriers received the most benefit
in PROVE-IT. Differential reduction in CHD events in PROVE IT-TIMI 22 based on KIF6 719Arg
carrier status. KIF6 719Arg carriers received significantly greater benefit from 80 mg atorvastatin,
compared with 40 mg pravastatin, than did noncarriers. The number needed to treat (NNT) with
atorvastatin (vs. pravastatin) for 2 years to prevent one event was 10 for KIF6 719Arg carriers and
125 for noncarriers
719Arg Carriers Noncarriers
PS PER VE-IT PS PER E-IT

R E S CO OS O RE SCO OS OV
CA WO PR PR CA WO PR PR
0
0.1%
Reduction (%)
Absolute Risk
5
1.4% 1.2%
4.9% 5.5% 0.8%
10
6.3%
15
10.0%
Fig. 11.9 Differential response to statin therapy based on KIF6 719Arg polymorphism. Differential
reduction in CHD events in four clinical statin trials based on KIF6 719Arg carrier status. KIF6
719Arg carriers received substantial and significant reduction of absolute risk. KIF6 719Arg car-
riers received benefit from both a hydrophilic and lipophilic statin. No significant absolute risk
reduction (ARR) was observed in KIF6 noncarriers in these three placebo controlled, and one statin
dose comparison trials
favoring atorvastatin in a model adjusted for traditional risk factors, triglyceride and
CRP levels, and treatment with gatifloxacin. No significant RRR was observed in
noncarriers. The ARR was 10% favoring atorvastatin therapy for carriers of the
KIF6 719Arg variant compared to 0.8% for noncarriers. Thus, the number of
patients needed to treat with high-dose atorvastatin rather than standard-dose
pravastatin to prevent one acute coronary syndrome event was 10 for carriers of the
KIF6 719Arg variant compared to 125 for noncarriers. Further, among KIF6
719Arg carriers, this significant superiority of high-dose atorvastatin therapy was
evident as early as 30 days after the start of treatment, yet in noncarriers there was
no significant superior benefit of high-dose treatment at any point during the study
(Fig. 11.8). This early superiority of intensive therapy in KIF6 719Arg carriers
may be due to an early plaque-stabilizing effect, a pleiotropic effect that appears not
to be linked to LDL-lowering or to anti-inflammatory mechanisms related to the
reduction in CRP as KIF6 719Arg carriers and noncarriers did not differ in median
LDL-C, triglyceride, or CRP levels at baseline or during the trial [65].
11.6.4KIF6 Predicts Clinical Event Response to Statin Therapy

as Shown in Four Randomized Clinical Trials
Taken together, the genetic studies of CARE, WOSCOPS, PROSPER, and PROVE
IT-TIMI 22 suggest that the KIF6 719Arg variant influences both risk of coronary
events (Fig. 11.4) and response to statin therapy (Fig. 11.9) in middle-aged men and
women, and in the elderly with a prior history of vascular disease.
11.7Clinical Examples
The clinical utility of the KIF6 719Arg polymorphism can be illustrated in the
following two case studies.
11.7.1Case Study #1. Elderly Asymptomatic Individual

in Whom the Physician Has Discovered
a Moderately Elevated LDL-C
11.7.1.1Report
Patient CK is a 72-year-old, healthy Caucasian male with a prior history of myocar-

dial infarction at age 64 treated with a bare metal stent (BMS) and no further events.
His LDL-C was 116 at the time of the MI and he was not placed on statins.
11.7.1.2Current Evaluation
Trig 126mg/dl
Total cholesterol 180mg/dl
LDL-C 109mg/dl
HDL-C 46mg/dl
Creatinine 1.7mg/dl
11.7.1.3Clinical Decision Problem
Should a statin medication be started in a 72-year-old asymptomatic male with a

prior cardiovascular history and slightly elevated creatinine?
(a) If KIF6 719Arg=Negativeuse of statins would not significantly reduce
the risk of a future event and the slightly elevated creatinine suggests some
degree of renal impairment and thus increased side effect risk from statin
treatment.
(b) If KIF6 719Arg=Positiveuse of statins may significantly reduce the risk of a
future event and moves the statin risk/benefit decision towards the benefit side.
11.7.1.4Basis for Decision
Iakoubova, O. A., Robertson, M., Tong, C. H., Roswland, C. M., Catanese, J. J.,
Blauw, G. J., et al. (2010). KIF6 Trp719Arg polymorphism and the effect of statin
therapy in elderly patients: Results from the PROSPER study. European Journal of
Cardiovascular Prevention and Rehabilitation, 17(4), 455461.
11.7.2Case Study #2. 58-year-old Asymptomatic Executive

Treated with a Statin but With Poor compliance.
No Prior History of CVD
11.7.2.1Report
Patient PB is a 58-year-old busy male executive with moderate LDL-C elevation

in the past (LDL-C=140mg/dl) for which his physician had prescribed simvas-
tatin 40mg/day. The initial LDL-C response was encouraging with LDL-C low-
ered to 110mg/dl; but over the next 6 months, the physician noted that the LDL-C
increased to 142 mg/dl. Upon inquiry, the physician determined the increase in
LDL-C was due to poor compliance to the statin medication, in part, due to the
patients sense that he could feel no improvement and besides, he was too busy to
remember.
11.7.2.2Current Evaluation
Trig 144mg/dl
Total cholesterol 213mg/dl
LDL-C 142mg/dl
HDL-C 42mg/dl
Creatinine 1.0mg/dl
11.7.2.3Clinical Decision Problem
How to enhance compliance.

(a) If KIF6 719Arg=Positivepatient informed that he was born with a genetic
status that increased his heart attack risk but more importantly, allowed him to
benefit MORE from statin treatment, in regard to fewer heart attacks, than the
average person and thus he should improve his compliance since he is one of
the lucky ones.
(b) If KIF6 719Arg=Negativepatients informed that they were born with a
genetic status that adherence to heart disease prevention methods is not
only important but also indicates that combination therapy may benefit them
more than statin alone and thus they warrant further evaluation and perhaps
combination therapy.
11.7.2.4Basis for Decision
Iakoubova, O. A., Tong, C. H., Rowland, C. M., et al. (2008). Association of the Trp719Arg
polymorphism in kinesin-like protein 6 with myocardial infarction and coronary heart disease
in 2 prospective trials: The CARE and WOSCOPS trials. Journal of the American College of
Brown, G., Albers, J. J., Fisher, L. D., Schaefer, S. M., Lin, J. T., Kaplan, C., Zhao, X. Q., Bisson,
B. D., Fitzpatrick, V. F., & Dodge, H. T. (1990). Regression of coronary artery disease as a
result of intensive lipid-lowering therapy in men with high levels of apolipoprotein B. The New
Brown, G. B, Zhao, X. Q., Chait, A., Fisher, L. D., Cheung, M. C., Morse, J. S., Dowdy, A. A.,
Marino, E. K., Bolson, E. L., Alaupovic, P., Frohlich, J., & Albers, J. J. (2001). Simvastatin and
niacin, antioxidant vitamins, or the combination for the prevention of coronary disease. The New
11.8Conclusions
The KIF6 719Arg polymorphism has clinical relevance for two reasons, one for
risk determination, and one for statin clinical event benefit. The KIF6 719Arg
polymorphism identifies patients at increased risk for CV events in over 50,000
research subjects, and also divides patients into two groups in regard to the benefit
they may receive from statin therapy, specifically in regard to fewer clinical events.
This benefit appears to be independent of LDL-C change and hs-CRP change and
has been documented in multiple clinical trials.
These effects are seen in both men and women and are ethnically diverse. The
pharmacogenetic clinical event response findings have been shown in groups with
well controlled, moderate, and elevated LDL-C levels such as the PROVE IT-TIMI
22 study, the CARE, and the WOSCOPS study. Further, similar findings have been
documented in older individuals (>70 years) who have a prior history of vascular
disease but not in an older group without a prior history of vascular disease.
Individuals who are not carriers of the 719Arg allele may benefit from statin
therapy in regard to fewer clinical events but the likelihood of benefit is surprisingly
low, despite substantial LDL-C lowering. In this group, it seems reasonable to
investigate non-LDL causes of CHD and consider treatment with combination lipid
therapy such as nicotinic acid+statin based on the results of the FATS and HATS
investigations that achieved 8090% reductions in clinical events [66, 67].
Compliance to medication is a major problem faced by most physicians. KIF6
carriers have an even stronger reason to maintain compliance with their physician
prescribed statin medication.
By testing only once in a patients lifetime, the physician gains knowledge of
the presence or absence of the KIF6 719Arg polymorphism which allows the phy-
sician to: (1) utilize genetic SNP analysis to further refine cardiovascular risk
evaluation; and (2) allow the physician to utilize pharmacogenomic information to
personalize statin-induced lipid-lowering therapy with the goal of reducing future
cardiovascular events.
References
1. Tsunetomo, Y. (1983). Hagakure (1716) (W. S. Wilson, Trans., distributed in the USA by
Kodansha America Inc.). Tokyo, Japan: Kodansha America Inc.
2. Lipid Research Clinics Program. (1984). The lipid research clinics coronary primary preven-
tion trial results: I. Reduction of incidence of coronary heart disease, and II. The relationship
of reduction in incidence of coronary heart disease to cholesterol lowering. The Journal of the
American Medical Association, 251, 351374.
3. NCEP ATP-III. (2001). Expert panel on detection, evaluation, and treatment of high blood
cholesterol in adults. Executive summary of The Third Report of The National Cholesterol
Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High
Blood Cholesterol In Adults (Adult Treatment Panel III). The Journal of the American
Medical Association, 285, 24862497.
4. Superko, H. R., & King, S. III. (2008). Lipid management to reduce cardiovascular risk:
A new strategy is required. Circulation, 117, 560568.
5. Schwartz, R. G., Pearson, T. A., Kalaria, V. G., Mackin, M. L., Williford, D. J., Awasthi, A., etal.
(2003). Prospective serial evaluation of myocardial perfusion and lipids during the first six months
of pravastatin therapy: Coronary artery disease regression single photon emission computed
tomography monitoring trial. Journal of the American College of Cardiology, 42, 600610.
6. AHA Statistical Update. (2006). Heart disease and stroke statistics 2006 update. Circulation,
113, e85e151.
7. Dawber, T. R. (1980). The Framingham study. Cambridge, MA: Harvard University Press.
8. Brown, S. A., Hutchinson, R., Morrisett, J., Boerwinkle, E., Davis, C. E., Gotto, A. M., etal.;
ARIC Study Group. (1993). Plasma lipid, lipoprotein cholesterol, and apoprotein distributions
in selected US communities. Arteriosclerosis and Thrombosis, 13, 11391158.
9. Multiple Risk Factor Intervention Trial. (1982). Risk factor changes and mortality results.
Multiple Risk Factor Intervention Trial Research Group. The Journal of the American Medical
Association, 248, 14651477.
10. Chobanian, A. V., Bakris, G. L., Black, H. R., Cushman, W. C., Green, L. A., Izzo, J. L. Jr,
etal.; Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High
Blood Pressure. National Heart, Lung, and Blood Institute; National High Blood Pressure
Education Program Coordinating Committee. (2003). Seventh report of the Joint National
Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure.
Hypertension, 42, 12061252.
11. Coronary Drug Project Research Group. (1978). Natural history of myocardial infarction in
the coronary drug project: Long-term prognostic importance of serum lipid levels. The
American Journal of Cardiology, 42, 489498.
12. Committee of Principal Investigators. (1980). Report on a WHO cooperative trial on primary
prevention of ischaemic heart disease using clofibrate to lower serum cholesterol. Lancet, ii,
379384.
13. Illingworth, D. R., & Bacon, S. (1987). Hypolipidemic effects of HMG-CoA reductase inhibitors
in patients with hypercholesterolemia. The American Journal of Cardiology, 30, 33G42G.
14. NCEP The Expert Panel. (1988). Report of the National Cholesterol Education Program
Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults
(DeWitt S. Goodman Chairman). Archives of Internal Medicine, 148, 3669.
15. Singh, B. K., & Mehta, J. L. (2002). Management of dyslipidemia in the primary prevention
of coronary heart disease. Current Opinion in Cardiology, 17, 503511.
16. OKeefe, J. H., Cordain, L., Harris, W. H., Moe, R. M., & Vogel, R. (2004). Optimal low-
density lipoprotein is 50 to 70 mg/dl. Journal of the American College of Cardiology, 43,
21422146.
17. Little, W. C., Constantinescu, M., Applegate, R. J., Kutcher, M. A., Burrows, M. T., Kahl, F. R.,
etal. (1988). Can coronary angiography predict the site of a subsequent myocardial infarction
in patients with mild-to-moderate coronary artery disease? Circulation, 78, 11571166.
18. Fuster, V., Badimon, L., Badimon, J., & Chesebro, J. H. (1992). The pathogenesis of coronary
artery disease and the acute coronary syndromes. The New England Journal of Medicine, 326,
242250, 310318.
19. Fuster, V., Moreno, P. R., Fayad, Z. A., Corti, R., & Badimon, J. J. (2005). Atherothrombosis
and high-risk plaque: Part I: Evolving concepts. Journal of the American College of
20. Akosah, K. O., Schaper, A., & Cogbill, C. (2003). Schoenfeld preventing myocardial infarc-
tion in the young adult in the first place: How do the National Cholesterol Education Panel III
guidelines perform? Journal of the American College of Cardiology, 41, 14751479.
21. Shaar, J. A., Muller, J. E., Falk, E., etal. (2004). Terminology for high-risk and vulnerable
coronary artery plaques. European Heart Journal, 25, 10771082.
22. Falk, E., Shah, P. K., & Fuster, V. (1995). Coronary plaque disruption. Circulation, 92, 657671.
23. Hirokawa, N., & Noda, Y. (2008). Intracellular transport and kinesin superfamily proteins,
KIFs: Structure, function, and dynamics. Physiological Reviews, 88, 10891118.
24. Ikegami, K., & Setou, M. (2010). Unique post-translational modifications in specialized
microtubule architecture. Cell Structure and Function, 35(1), 1522.
25. Asbury, C. L. (2005). Kinesin: Worlds tiniest biped. Current Opinion in Cell Biology, 17,
8997.
26. Gennerich, A., & Vale, R. D. (2009). Walking the walk: How kinesin and dynein coordinate
their steps. Current Opinion in Cell Biology, 21, 5967.
27. Miki, H., Setou, M., Kaneshiro, K., & Hirokawa, N. (2001). All kinesin superfamily protein,
KIF, genes in mouse and human. Proceedings of the National Academy of Sciences of the
United States of America, 98, 70047011.
28. Wozniak, M. J., Milner, R., & Allan, V. (2004). N-terminal kinesins: Many and various.
Traffic, 5, 400410.
29. Hirokawa, N. (1998). Kinesin and dynein superfamily proteins and the mechanism of
organelle transport. Science, 279, 519526.
30. Seiler, S., Kirchner, J., Horn, C., Kallipolitou, A., Woehlke, G., & Schliwa, M. (2000). Cargo
binding and regulatory sites in the tail of fungal conventional kinesin. Nature Cell Biology, 2,
333338.
31. Barrett, T., Troup, D. B., Wilhite, S. E., etal. (2007). NCBI GEO: Mining tens of millions of
expression profiles database and tools update. Nucleic Acids Research, 35, D760D765.
32. King, J. Y., Ferrara, R., Tabibiazar, R., etal. (2005). Pathway analysis of coronary atheroscle-
rosis. Physiological Genomics, 23, 103118.
33. Seog, D. H., Lee, D. H., & Lee, S. K. (2004). Molecular motor proteins of the kinesin superfamily
proteins (KIFs): Structure, cargo and disease. Journal of Korean Medical Science, 19, 17.
34. Davani, S., Gozalo, C., Gambert, S., Chalmers, D., Gambert, P., Schiele, F., etal. (2010). The
polymorphism Trp719Arg in the kinesin-like protein 6 is associated with the presence of late
outgrowth endothelial progenitor cells in acute myocardial infarction. Atherosclerosis, 210(1),
4850.
35. Schunkert, H., Gotz, A., Braund, P., et al. (2008). Repeated replication and a prospective
meta-analysis of the association between chromosome 9p21.3 and coronary artery disease.
36. Shiffman, D., Ellis, S. G., Rowland, C. M., etal. (2005). Identification of four gene variants
associated with myocardial infarction. American Journal of Human Genetics, 77(4),
596605.
37. The ARIC Investigators. (1989). The Atherosclerosis Risk in Communities (ARIC) Study:
Design and objectives. American Journal of Epidemiology, 129, 687702.
38. White, A. D., Folsom, A. R., Chambless, L. E., etal. (1996). Community surveillance of coro-
nary heart disease in the Atherosclerosis Risk in Communities (ARIC) Study: Methods and
initial two years experience. Journal of Clinical Epidemiology, 49, 223233.
39. Morrison, A. C., Bare, L. A., Chambless, L. E., et al. (2007). Prediction of coronary heart
disease risk using a genetic risk score: The Atherosclerosis Risk in Communities Study.
American Journal of Epidemiology, 166, 2835.
40. Ridker, P. M., Cook, N. R., Lee, I. M., etal. (2005). A randomized trial of low-dose aspirin in
the primary prevention of cardiovascular disease in women. The New England Journal of
Medicine, 352, 12931304.
41. Shiffman, D., Chasman, D. I., Zee, R. Y., etal. (2008). A kinesin family member 6 variant is
associated with coronary heart disease in the Womens Health Study. Journal of the American
College of Cardiology, 51, 444448.
42. Cushman, M., Cornell, E. S., Howard, P. R., Bovill, E. G., & Tracy, R. P. (1995). Laboratory
methods and quality assurance in the Cardiovascular Health Study. Clinical Chemistry, 41,
264270.
43. Fried, L. P., Borhani, N. O., Enright, P., et al. (1991). The Cardiovascular Health Study:
Design and rationale. Annals of Epidemiology, 1, 263276.
44. Shiffman, D., OMeara, E. S., Bare, L. A., etal. (2008). Association of gene variants with
incident myocardial infarction in the Cardiovascular Health Study. Arteriosclerosis,
Thrombosis, and Vascular Biology, 28, 173179.
45. Sacks, F. M., Pfeffer, M. A., Moye, L. A., etal. (1996). The effect of pravastatin on coronary
events after myocardial infarction in patients with average cholesterol levels. Cholesterol and
Recurrent Events Trial investigators. The New England Journal of Medicine, 335, 10011009.
46. Shepherd, J., Cobbe, S. M., Ford, I., etal. (1995). Prevention of coronary heart disease with
pravastatin in men with hypercholesterolemia. West of Scotland Coronary Prevention Study
Group. The New England Journal of Medicine, 333, 13011307.
47. Iakoubova, O. A., Tong, C. H., Rowland, C. M., etal. (2008). Association of the Trp719Arg
polymorphism in kinesin-like protein 6 with myocardial infarction and coronary heart disease
in 2 prospective trials: The CARE and WOSCOPS trials. Journal of the American College of
48. Packard, C. J., OReilly, D. S., Caslake, M. J., et al. (2000). Lipoprotein-associated
phospholipase A2 as an independent predictor of coronary heart disease. West of Scotland
Coronary Prevention Study Group. The New England Journal of Medicine, 343,
11481155.
49. Iakoubova, O., Robertson, M., & Tong, C. H. (2009). KIF6 Trp719Arg polymorphism and the
effect of statin therapy in elderly patients: Results from the PROSPER study. (Manuscript
under review).
50. The Wellcome Trust Control Consortium. (2007). Genome wide association study of 14,000
cases of seven common diseases and 3,000 shared controls. Nature, 447, 661677.
51. Stewart, A. F., Dandona, S., Chen, L., etal. (2009). Kinesin family member 6 variant Trp719Arg
does not associate with angiographically defined coronary artery disease in the Ottawa Heart
Genomics Study. Journal of the American College of Cardiology, 53, 14711472.
52. Williams, P. D. (2010). Survival and other biases affecting case-control genetic studies of
coronary heart disease. Current Molecular Pharmacology (in submission).
53. Shiffman, D., Rowland, C. M., Thompson, J., Hall, A. S., Devlin, J. J., & Saman, N. J. (2009).
The KIF6719ARG allele is associated with increased risk of coronary heart disease among
males in the Wellcome Trust Case Control Consortium Study of Coronary Heart Disease.
Poster presentation at the British Atheroslcerosis Society Autumn Meeting: Genetics of
Complex Diseases, Queens College, Cambridge, UK.
54. Tham, L. S., Goh, B. C., Nafziger, A., Guo, J. Y., Wang, L. Z., Soong, R., et al. (2006).
A warfarin-dosing model in Asians that uses single-nucleotide polymorphisms in vitamin K
epoxide reductase complex and cytochrome P450 2C9. Clinical Pharmacology and
Therapeutics, 80, 346355
55. Sconce, E. A., Khan, T. I., Wynne, H. A., Avery, P., Monkhouse, L., King, B. P., etal. (2005).
The impact of CYP2C9 and VKORC1 genetic polymorphism and patient characteristics upon
warfarin dose requirements: Proposal for a new dosing regimen. Blood, 106(7), 23292333.
Epub 2005 Jun 9.
56. Voora, D., Eby, C., Linder, M. W., Milligan, P. E., Bukaveckas, B. L., McLeod, H. L., etal.
(2005). Prospective dosing of warfarin based on cytochrome P-450 2C9 genotype. Thrombosis
and Haemostasis, 93, 700705.
57. Antman, E. M., Wiviott, S. D., Murphy, S. A., Voitk, J., Hasin, Y., Widimsky, P., etal. (2008).
Early and late benefits of prasugrel in patients with acute coronary syndromes undergoing percu-
taneous coronary intervention: A TRITON-TIMI 38 (TRial to Assess Improvement in Therapeutic
Outcomes by Optimizing Platelet InhibitioN with Prasugrel-Thrombolysis in Myocardial
Infarction) analysis. Journal of the American College of Cardiology, 51, 20282033.
58. Schatz, I. J., Masaki, K., Yano, K., Chen, R., Rodriguez, B. L., & Curb, J. D. (2001).
Cholesterol and all-cause mortality in elderly people from the Honolulu Heart Program:
A cohort study. Lancet, 358, 351355.
59. Shipley, M. J., Pocock, S. J., & Marmot, M. G. (1991). Does plasma cholesterol concentration
predict mortality from coronary heart disease in elderly people? 18 year follow up in
Whitehall study. British Medical Journal (Clinical Research Ed.), 303, 8992.
60. Weverling-Rijnsburger, A. W., Blauw, G. J., Lagaay, A. M., Knook, D. L., Meinders, A. E., &
Westendorp, R. G. (1997). Total cholesterol and risk of mortality in the oldest old. Lancet,
350, 11191123.
61. Ray, K. K., Bach, R. G., Cannon, C. P., etal. (2006). Benefits of achieving the NCEP optional
LDL-C goal among elderly patients with ACS. European Heart Journal, 27(19),
23102316.
62. Shepherd, J., Blauw, G. J., Murphy, M. B., etal. (2002). Pravastatin in elderly individuals at
risk of vascular disease (PROSPER): A randomised controlled trial. Lancet, 360(9346),
16231630.
63. Iakoubova, O. A., Robertson, M., Tong, C. H., Roswland, C. M., Catanese, J. J., Blauw, G. J.,
et al. (2010). KIF6 Trp719Arg polymorphism and the effect of statin therapy in elderly
patients: Results from the PROSPER study. European Journal of Cardiovascular Prevention
and Rehabilitation, 17(4), 455461.
64. Cannon, C. P., Braunwald, E., McCabe, C. H., etal. (2004). Intensive versus moderate lipid
lowering with statins after acute coronary syndromes. The New England Journal of Medicine,
350, 14951504.
65. Iakoubova, O. A., Sabatine, M. S., Rowland, C. M., etal. (2008). Polymorphism in KIF6 gene
and benefit from statins after acute coronary syndromes: Results from the PROVE IT-TIMI
22 study. Journal of the American College of Cardiology, 51, 449455.
66. Brown, G., Albers, J. J., Fisher, L. D., Schaefer, S. M., Lin, J. T., Kaplan, C., etal. (1990).
Regression of coronary artery disease as a result of intensive lipid-lowering therapy in men
with high levels of apolipoprotein B. The New England Journal of Medicine, 323,
12891298.
67. Brown, G. B., Zhao, X. Q., Chait, A., Fisher, L. D., Cheung, M. C., Morse, J. S., etal. (2001).
Simvastatin and niacin, antioxidant vitamins, or the combination for the prevention of coronary
disease. The New England Journal of Medicine, 345, 15831592.
Part IV
Drugs that Cause Delayed
Hypersensitivity
Chapter 12
Abacavir
Elizabeth J. Phillips and Simon A. Mallal
Keywords Abacavir Abacavir pharmacokinetics Abacavir hypersensitivity

Pharmacovigilance Rechallenge HLA-B*5701 False positive clinical diagnosis
Abacavir patch testing Abacavir observational studies PREDICT-1 study
SHAPE study Negative predictive value of HLA-B*5701 Immunopathogenesis
of abacavir hypersensitivity HLA-B*5701 translation to clinic HLA-B*5701
quality assurance HLA-B*5701 laboratory testing HLA-B*5701 screening test
12.1Pharmacology and Clinical Use of Abacavir
Abacavir is a guanosine analog that competitively inhibits the reverse transcriptase of

the human immunodefiency virus (HIV) and is used in combination antiretroviral
therapy for the treatment of HIV. It was approved by the US Food and Drug
Administration in 1998 and has been in clinical use since that time. Abacavir is cur-
rently most commonly used as part of a fixed-dose combination, 600mg in combination
with lamivudine 300mg which is given once-daily as part of combination antiretroviral
therapy. It is also available as a liquid formulation and has been approved for use in
children. Abacavir is well absorbed with an absolute bioavailability of 83% and can be
taken without regard to food [1]. Although the parent drug has a short plasma half-life
E.J. Phillips(*)
Institute for Immunology & Infectious Diseases, Murdoch University, Murdoch, Western Australia
and
Sir Charles Gairdner Hospital, Nedlands, Perth, Western Australia
and
Royal Perth Hospital, Perth, Western Australia
and
School of Pathology & Laboratory Medicine, University of Western Australia, Perth, Western
Australia
e-mail: e.phillips@iiid.com.au
202 E.J. Phillips and S.A. Mallal
of only 2h, abacavir is metabolized to carbovir triphosphate which has an intracellular

half-life of 20h or longer [1, 2]. It is this long intracellular half-life of carbovir triphos-
phate that makes abacavir pharmacokinetically amenable to once-daily dosing and this
has been supported by clinical data [3, 4]. Unlike other nucleoside analog reverse tran-
scriptase inhibitors, abacavir is metabolized predominantly by the liver by two major
pathways: alcohol dehydrogenase and uridine diphosphate glucruonyltransferase. Less
than 2% is excreted unchanged in the urine [1]. In addition, unlike other antiretroviral
drugs that are metabolized by the liver such as nonnucleoside reverse transcriptase
inhibitors and protease inhibitors, abacavir is not significantly metabolized by cyto-
chrome (CYP) P450 enzymes. It also does not inhibit or induce CYP enzymes which
make drug interactions unlikely. Therapeutic drug monitoring has been infrequently
employed for abacavir as it is the intracellular concentrations of carbovir triphosphate
which would be most likely to be associated with drug effect and these are difficult and
costly to measure on a routine basis. There is also very little information on validated
intracellular target concentrations and their association with clinical efficacy.
12.2Abacavir Side Effects
The major treatment limiting side effect of abacavir, which is independent of the dose of
the drug, is a drug hypersensitivity syndrome. Abacavir hypersensitivity was first
described in the premarketing phase of drug development. In predominantly Caucasian
populations, it has been clinically diagnosed in approximately 8% of those initiating the
drug [5]. On first exposure to the drug symptoms of abacavir hypersensitivity usually
appears in the second week of treatment, a median of 9 days after its initiation. Initial
symptoms such as fever, malaise, diarrhea, nausea, and vomiting are common but can be
very nonspecific and hence challenging to separate from side effects of other antiretroviral
drugs, infections, or immune restoration/inflammatory disease. Rash tends to be a late
component of the abacavir hypersensitivity syndrome occurring in approximately 70% of
patients prior to drug discontinuation. The rash associated with abacavir hypersensitivity
is a mild to moderate maculopapular exanthema and not a blistering eruption with skin
separation characteristic of StevensJohnson syndrome or toxic epidermal necrolysis.
Symptoms of abacavir hypersensitivity should remit within 72h of stopping the drug and
this dechallenge used as one of the criteria that strengthens the likelihood of the clinical
diagnosis of abacavir hypersensitivity [5]. Rechallenge with abacavir following a reaction
clinically compatible with abacavir hypersensitivity can result in rapid and severe shock
and mortality and is therefore contraindicated [6]. The success of abacavir as an antiretro-
viral drug was predicated in the early postmarketing years by a robust clinical manage-
ment pharmacovigilence program that educated physicians and other healthcare workers
treating patients with HIV in the early recognition and management of abacavir hypersen-
sitivity. This included permanent discontinuation of abacavir in patients with signs and
symptoms compatible with abacavir. By necessity from a drug safety standpoint, the clini-
cal pharmacovigilence program led to many more patients being diagnosed with abacavir
hypersensitivity and discontinuing abacavir than truly had the disease. Patients themselves
were also engaged in this process of early recognition of abacavir hypersensitivity and
12 Abacavir 203
were given a warning card and a 24h contact line should they develop symptoms or
signs in the first 6 weeks of abacavir treatment.
12.3Pharmacogenetics of Abacavir Hypersensitivity
Early clues for a genetic basis of abacavir hypersensitivity included lower fre-
quency of the reactions described in those of African and African American
descent, and a case report describing the syndrome in a father and daughter [7, 8].
In March 2002, two independent research groups reported an association between
the histocompatibility class I allele, HLA-B*5701 and abacavir hypersensitivity [9,
10]. Both studies employed a candidate gene approach and had an overrepresenta-
tion of Caucasian men. The Western Australian population-based cohort study
provided complete case ascertainment of 200 patients exposed to abacavir [9]
whereas, the US GSK study was a retrospective case-control design [10]. In the
Western Australian study the HLA-B*5701 allele was present in 78% of patients
clinically diagnosed with abacavir hypersensitivity [9]. Despite the strong associa-
tion between HLA-B*5701 and abacavir hypersensitivity, 22% of patients in the
Australian study and 45% of patients in the US study with a clinical diagnosis of
abacavir hypersensitivity lacked HLA-B*5701 [9, 10]. This less than 100% nega-
tive predictive value of HLA-B*5701 for clinically diagnosed abacavir hypersensi-
tivity raised safety concerns about the potential clinical utility of HLA-B*5701 as
a screening test to prevent abacavir hypersensitivity. These concerns were further
heightened with a subsequent US case-control study that showed only 8% sensitiv-
ity of HLA-B*5701 for clinically diagnosed abacavir hypersensitivity in African
American populations [11]. It later became clear that these early studies showing
lower sensitivity of HLA-B*5701 for clinically diagnosed abacavir hypersensitivity
were hampered by false positive clinical diagnosis. This false positive clinical diag-
nosis was particularly apparent in Blacks and other non-White populations with a
lower carriage rate of HLA-B*5701 (Fig. 12.1). The problem with false positive
clinical diagnosis has been further highlighted in randomized double-blinded con-
trolled treatment trials where consistently 27% of patients found to have not even
received abacavir after unbinding of the study had been given a clinical diagnosis
of abacavir hypersensitivity [1215]. The problem of the high sensitivity but low
specificity of clinical diagnosis for abacavir hypersensitivity was largely overcome
by abacavir patch testing, a technique whereby increasing concentrations of aba-
cavir were applied in a petrolatum vehicle on the skin surface with standard com-
mercial patch tape. In 2002, it was noted that a high proportion of patients with
clinical syndromes compatible with abacavir hypersensitivity had clearly positive
skin patch tests 24h after application and that the histopathological picture of a skin
biopsy from a skin patch test was identical to that of the rash of acute abacavir
hypersensitivity reaction [16]. Three of four HLA-B*5701 negative clinical
abacavir hypersensitivity patients were accessible for follow-up in the original
Western Australian study returned and had negative abacavir skin patch tests and
subsequently went on to tolerate abacavir [17]. Subsequent patch test studies
W. EUROPE 5- 7%
MEDITERRANEAN UK MIDDLE EAST 1-2%

1-2% ~8%
(NB 5-7% Ashkenazi Jews)
US
Caucasian INDIA 5-20%
~8% CHINA 0%
US Asian (NB 2.5% N.E.
~1% provinces)
US JAPAN 0%
African-
American THAILAND
~2.5% 4 -10%*
US AUSTRALIA
Hispanic ~8%
~2%
S. AMERICAN *THAILAND B*57 carriage:

Caucasian Subsaharan Urban Bangkok 3.6%
5-7% AFRICA Thai Dai Lue (NE Thai) ~11%
<1% Southern Thai Muslim 3%
Fig. 12.1 Carriage rate of HLA-B*5701 in different populations (Adapted from Phillips, E.
(2006). Clinical Infectious Diseases: An Official Publication of the Infectious Diseases Society of
America, 43, 103105 by the University of Chicago Press)
showed 100% of patients with positive abacavir patch tests carried HLA-B*5701
[1821]. This suggested that abacavir patch testing identifies those with true immu-
nologically mediated abacavir hypersensitivity and could be used as a research tool
to circumvent the false positive clinical diagnosis.
Between 2006 to the present, several observational studies have been published
that have supported the utility of HLA-B*5701 testing in decreasing both true
immunologically mediated and the false-positive clinically diagnosed abacavir
hypersensitivity. The Western Australian cohort introduced routine HLA-B*5701
screening in early 2002 and subsequently published the prospective experience fol-
lowing the introduction of HLA-B*5701 screening between 2002 and 2005 [17, 22].
This study showed a significant reduction in abacavir hypersensitivity from 8%
before introduction of screening to 2% following postscreening. All three cases of
abacavir hypersensitivity occurred in HLA-B*5701 positive patients who had been
prescribed abacavir despite the positive screening test. No cases of abacavir hyper-
sensitivity were described among 138 HLA-B*5701-negative patients. This study also
showed a reduction in all cause discontinuation of abacavir potential highlighting the
positive downstream effects that HLA-B*5701 screening has on the confidence of
the patient and provider. Several other observational studies have also shown
significant reductions in patients diagnosed with clinical abacavir hypersensitivity
reaction following introduction of HLA-B*5701 screening [2328]. In the
Atazanavir/Ritonavir Induction/Simplification with Epzicom Study (ARIES), the
12 Abacavir 205
first randomized open-label clinical trial to utilize prospective HLA-B*5701

screening only 4/517 (0.8%) of patients received a clinical diagnosis of abacavir
hypersensitivity and all four had negative skin patch tests adding further support
that screening essentially eliminates true immunologically mediated abacavir
hypersensitivity syndrome [28].
The highest level of evidence to support the clinical utility of HLA-B*5701 for
the prevention of abacavir hypersensitivity syndrome has come from the Prospective
Randomized Evaluation of DNA Screening in a Clinical Trial (PREDICT-1) study
[20]. The PREDICT-1 study set the stage as being the first randomized pharmaco-
genetic study to examine the clinical effectiveness of a pharmacogenetic marker to
reduce the toxicity of a specific drug. In this study, a total of 1,956 predominantly
Caucasian (84%) patients, were enrolled across 265 sites in Europe and Australia,
and were randomized either to real-time HLA-B*5701 screening and exclusion
from abacavir based on the HLA status or a control arm that received retrospective
screening following exposure to abacavir [20]. The study was additionally unique
in utilizing coprimary endpoints which included clinical diagnosis of abacavir
hypersensitivity by the patients physician and clinically diagnosed abacavir hyper-
sensitivity which had been confirmed by positive patch testing. Abacavir patch
testing was conducted in all patients 610 weeks following a clinical diagnosis of
abacavir hypersensitivity as well as in 100 patients who were tolerant of abacavir
to test the specificity of patch testing. The results of the study were striking in that
patch test positive abacavir hypersensitivity was completely eliminated in the group
who underwent HLA-B*5701 screening. This indicates a 100% negative predictive
value of HLA-B*5701 for true immunologically mediated (patch test positive)
abacavir hypersensitivity. A significant reduction in clinically diagnosed abacavir
hypersensitivity from 7.8 to 3.4% also occurred. It is likely that most of these 3.4%
who were clinically diagnosed with abacavir hypersensitivity but patch test nega-
tive represent false-positive clinical diagnosis. This was supported by a multivariate
model which suggested that there were significantly more patients starting nonnu-
cleoside reverse transcriptase inhibitors and those with gastro-intestinal symptoms
using protease inhibitors in the HLA-B*5701 negative group that received a clini-
cal diagnosis of abacavir hypersensitivity. These finding suggest that symptom
overlap and confusion contributed to false-positive clinical diagnosis in these
patients. Indeed, a 3.6% rate of false-positive clinical diagnosis had been antici-
pated in the study design and was used for the power calculation for the clinical
outcome endpoint. The high anticipated false positive clinical diagnosis rate also
motivated the use of the coprimary endpoint of patch test confirming abacavir
hypersensitivity [20].
The generalizability of the PREDICT-1 study to all populations was limited by
its predominant Caucasian population. To look at the applicability of HLA-B*5701
testing to more diverse populations, the Study of Hypersensitivity to Abacavir and
Pharmacogenetic Evaluation (SHAPE) study was conducted to examine the sensi-
tivity and specificity of HLA-B*5701 for clinically diagnosed and patch test posi-
tive abacavir hypersensitivity in White and Black American populations [21]. This
study was a retrospective case-control study that applied patch testing to Black and
White cases with a clinical history consistent with abacavir hypersensitivity

reaction and HLA-B*5701 testing to both cases and controls tolerating abacavir. In
this study, 100% of both White and Black patch test positive cases carried HLA-
B*5701 supporting that HLA-B*5701 has 100% sensitivity and 100% negative
predictive value for immunologically mediated abacavir hypersensitivity that is
generalizable across race.
12.4Mechanism of Drug-Induced Delayed

Hypersensitivity Reaction
In parallel with the strong body of clinical evidence outlined above, insights into the
immunopathogenesis of abacavir hypersensitivity provides additional evidence to
support that abacavir hypersensitivity is restricted to the class I MHC allele HLA-
B*5701 and is associated with an abacavir specific CD8+-mediated immune response
[2931]. This process is exquisitely restricted by HLA-B*5701 as demonstrated by
elegant research showing that a single amino acid change within the peptide-binding
domain is enough to result in clinical and ex vivo tolerance. Any change in the pep-
tide biding groove appears to obliterate the ability of the HLA allele to accommodate
and bind the haptenated version of abacavir or its reactive metabolite [29]. This
explains why patients who carry other members of the B17 serogroup such as HLA-
B*58 and HLA-B*5703, that differ from HLA-B*5701 by only a few amino acids,
are tolerant to abacavir. Other experiments from healthy, abacavir naive HLA-B*5701
blood donors have shown that CD8+ T-cells proliferate in response to abacavir over
1114 days and can be restimulated by abacavir-exposed antigen-presenting cells
transfected with HLA-B*5701 that are devoid of other HLA antigens [29].
12.5Predictive Value of Genotype Testing

for Patients on Abacavir
Data from the PREDICT-1 study showed the positive predictive value of HLA-
B*5701 for abacavir hypersensitivity to be 55%, meaning that approximately 45%
of patients carrying HLA-B*5701 would not develop hypersensitivity on exposure
to abacavir. A follow up genetic study of the PREDICT-1, SHAPE and multina-
tional Australian-Canadian-Swiss Study attempted to look at factors driving aba-
cavir tolerance by looking at polymorphisms in several genes haplospecific to
HLA-B*5701 as well as KIR, CD14, and drug metabolizing genes such as alcohol
dehydrogenase comparing 95 abacavir patch test-positive patients and 43 HLA-
B*5701 patients tolerating abacavir. This study did not identify any definitive
abrogating factors in HLA-B*5701 patients tolerating abacavir but did suggest that
it is highly likely that these lie outside of the MHC [32].
12 Abacavir 207
12.6Application of HLA-B*5701 Screening to the Clinic
The process starting from discovery of the association between HLA-B*5701 to

the translation of the test into widespread clinical practice that has been endorsed
by international treatment guidelines illustrates the number of hurdles and steps
involved in getting a pharmacogenetic test to the clinic (Fig. 12.1). Although
HLA-B*5701 is most prevalent in White European populations (Fig. 12.2), current
evidence suggests that the implications of HLA-B*5701 positivity are similar
across race and the 100% negative predictive value of HLA-B*5701 for true
immunologically mediated abacavir hypersensitivity is broadly applicable to both
white and non-White race [20, 21]. This is extremely important from a practical
standpoint as in the globalized world where migration and population admixture
is frequent and it is not be possible to determine genetic racial background by
patient appearance or self-defined ethnicity [33]. Furthermore, screening is cost
effective even in races that have a very low prevalence of HLA-B*5701 because it
Fig. 12.2 A roadmap for the transplantation of pharmacogenetic research into clinical practice:
the abacavir example (Adapted from Phillips, E, Mallal, S. (2009). Personalized Medicine, 6,
393408, Future Medicine Ltd)
Table 12.1 The number needed to screen for HLA-B*5701 to prevent one case of abacavir hypersensitivity (ABC HSR) is dependent on the prevalence of the
208
abacavir hypersensitivity as well as the positive predictive value of HLA-B*5701 for ABC since there is such a high rate of false positive clinical diagnosis for
ABC HSR(27%) which is also significantly reduced with HLA-B*5701 screening. When this is taken into account, the number needed to treat to prevent once
case decreases from 32 to approximately 13. Explanatory tables are shown below. For each table a theoretical population of 100,000 has been assumed and the
number of individuals with the ADR is calculated based on its reported prevalence and the number of tolerant patients is calculated by subtracting this figure from
100,000. The sensitivity and specificity of the HLA test for the ADR have been taken from the literature and the numbers in each of the four cells (ad) calculated.
The positive and negative predictive values (PPV and NPV) and numbers needed to test (NNT) were then calculated from the numbers in each cell (ad).
HLA +ve HLA ve Total Prevalence=(a+b)/1000
ADR +ve a b a+b Prev=a+b/a+b+c+d
Sensitivity=a/a+b 100%
Drug tolerant c d c+d Specificity=d/c+d 100%
Total a+c b+d a+b+c+d= 100,000
PPV=a/a+ c100% NPV=d/b+ d100% Number need to prevent one
case=100,000/a
Abacavir true immunologically mediated+clinical false-positive

B*5701 +ve B*5701 ve Total Prevalence=8%
ABC HSR (including clinical 8,000 0 8,000 Sensitivity=100%
false-positive)
ABC tolerant 2,425 89,575 92,000 Specificity=97.4%
Total 10,425 89,575 1,00,000 NNT to prevent
PPV=76.7% NPV=100% one case=12.5
Abacavir true immunologically mediated

B*5701 +ve B*5701 ve Total Prevalence=3.1%
ABC HSR 3,150 0 3,150 Sensitivity=100%
ABC tolerant 2,550 94,300 96,850 Specificity=97.4%
Total 5,700 94,300 1,00,000 NNT to prevent one case=32
PPV=55%% NPV=100%
E.J. Phillips and S.A. Mallal
12 Abacavir 209
prevents both true immunologically mediated abacavir hypersensitivity and

false-positive clinical diagnosis, and in view of the consistent 27% rate of false-
positive clinical diagnosis [34]. Data from all studies to-date suggest that the
number needed to test to prevent one case of abacavir hypersensitivity would be
13 if taking into account both true and false-positive clinical diagnosis and 32 if
considering only prevention of true immunologically mediated abacavir hypersen-
sitivity (Table 12.1). Perhaps the most important role of HLA-B*5701 screening is
improving drug safety in that the major morbidity and mortality of abacavir hyper-
sensitivity occurs on repeat rather than initial exposure to abacavir, therefore,
primary HLA-B*5701 screening prevents the population at highest risk from ever
being exposed or sensitised to abacavir.
Most crucial in the successful translation of HLA-B*5701 screening into
routine clinical use has been the effective development of cost-effective and
quality assured laboratory testing [35]. The development of a internal quality
assurance program now driven by the Asian Pacific Histocompatibility and
Immunogenetics Association (APHIA) has been instrumental in the success and
safety of ongoing HLA-B*5701 testing globally. The development of molecular
and flow cytometric techniques was particularly important since the sequence
based typing used in HLA testing is expensive, can have a long turnaround time,
and has typically been performed in the specialized transplant laboratories.
PCR-based techniques such as sequence specific amplification and real-time
PCR melting curve analysis are sensitive and specific assays that have been
developed and have now been applied in many routine diagnostic laboratories
[36, 37]. Flow cytometric-based techniques have also been described and look
promising as cost-effective tests that can be added on to the CD4+ and CD8+
counts that are done at baseline and follow-up in all patients on antiretroviral
therapy [38]. Other PCR-based techniques have included a Taqman assay detect-
ing the single nucleotide polymorphism HCP5 rs2390529 within an endogenous
retrovirus [39]. Previous studies suggested that this marker was in complete
linkage disequilibrium with HLA-B*5701 and was therefore a haplospecific
marker. However, cases of patch test positive abacavir hypersensitivity have
occurred in patients positive for HLA-B*5701 but negative for HCP5 rs2395029
[32] which suggests that this marker cannot safely be used as a screening test for
the prevention of abacavir hypersensitivity. This underscores that any surrogate
that does not have 100% sensitivity for the presence of HLA-B*5701 cannot
safely be implemented as a screening test.
12.7Case Report
A 43-year-old Caucasian HIV-positive man presents asymptomatic with a

CD4+ count of 250/ml and an HIV viral load of 150,000 copies per ml.
Screening test comes back negative for HLA-B*5701 and he is commenced on
treatment with abacavir-lamivudine (600/300 mg) once daily and nevirapine
initially 200mg once daily with a plan to increase to 400mg once daily after
two weeks. Two weeks into treatment, he develops a maculopapular eruption
starting on his trunk and becoming generalized, but he is afebrile and otherwise
well. Laboratory tests reveal normal full blood count and liver function tests;
rash continues to be bothersome despite antihistamines and topical steroids; and
atazanavir/ritonavir is substituted for nevirapine. Within a few days the rash
has resolved.
12.8Summary
The patient being HLA-B*5701 negative in this case gives additional reassurance
that this is not abacavir hypersensitivity syndrome, particularly when drugs such
as abacavir and nevirapine that have overlapping toxicities are used together.
Rash does not occur as the initial symptom with abacavir hypersensitivity syn-
drome and isolated rash without fever and other symptoms does not meet criteria
for the clinical diagnosis of abacavir hypersensitivity. Nevirapine is a common
cause of isolated rash. Isolated rash can occur with abacavir, however this is not
known to be predicted by HLA-B*5701.
HLA-B*5701 is most useful as a screening test for patients who are naive to
abacavir. Although there is yet to be another HLA allele associated with aba-
cavir hypersensitivity, it is still possible that a rare HLA allele may yet be dis-
covered and HLA-B*5701 negativity should not be used as the basis for
rechallenge in a patient who has experienced a clinical syndrome compatible
with abacavir hypersensitivity.
References
1. Yuen, G. J., Weller, S., & Pakes, G. E. (2008). A review of the pharmacokinetics of abacavir.
Clinical Pharmacokinetics, 47(6), 351371.
2. Moyle, G., Boffito, M., Fletcher, C., Higgs, C., Hay, P. E., Song, I. H., et al. (2009). Steady-state
pharmacokinetics of abacavir in plasma and intracellular carbovir triphosphate following
administration of abacavir at 600 milligrams once daily and 300 milligrams twice daily in
human immunodeficiency virus-infected subjects. Antimicrobial Agents and Chemotherapy,
53(4), 15321538.
3. Lamarca, A., Clumeck, N., Plettenberg, A., Domingo, P., Fu, K., Craig, C., et al. (2006).
Efficacy and safety of a once-daily fixed-dose combination of abacavir/lamivudine compared
with abacavir twice daily and lamivudine once daily as separate entities in antiretroviral-
experienced hiv-1-infected patients (CAL30001 study). Journal of Acquired Immune
Deficiency Syndromes, 41(5), 598606.
4. Moyle, G. J., DeJesus, E., Cahn, P., Castillo, S. A., Zhao, H., Gordon, D. N., et al. (2005).
Abacavir once or twice daily combined with once-daily lamivudine and efavirenz for the treat-
ment of antiretroviral-naive HIV-infected adults: Results of the ziagen once daily in antiretro-
viral combination study. Journal of Acquired Immune Deficiency Syndromes, 38(4),
417425.
12 Abacavir 211
5. Hetherington, S., McGuirk, S., Powell, G., Cutrell, A., Naderer, O., Spreen, B., et al. (2001).
Hypersensitivity reactions during therapy with the nucleoside reverse transcriptase inhibitor
abacavir. Clinical Therapeutics, 23(10), 16031614.
6. Shapiro, M., Ward, K. M., & Stern, J. J. (2001). A near-fatal hypersensitivity reaction to aba-
cavir: Case report and literature review. AIDS Reader, 11(4), 222226.
7. Symonds, W., Cutrell, A., Edwards, M., Steel, H., Spreen, B., Powell, G., et al. (2002). Risk
factor analysis of hypersensitivity reactions to abacavir. Clinical Therapeutics, 24(4), 565573.
8. Peyriere, H., Nicolas, J., Siffert, M., Demoly, P., Hillaire-Buys, D., & Reynes, J. (2001).
Hypersensitivity related to abacavir in two members of a family. Annals of Pharmacotherapy,
35(10), 12911292.
9. Mallal, S., Nolan, D., Witt, C., Masel, G., Martin, A. M., Moore, C., et al. (2002). Association
between presence of HLA-B*5701, HLA-DR7, and HLA-DQ3 and hypersensitivity to HIV-1
reverse-transcriptase inhibitor abacavir. Lancet, 359(9308), 727732.
10. Hetherington, S., Hughes, A. R., Mosteller, M., Shortino, D., Baker, K. L., Spreen, W., et al.
(2002). Genetic variations in HLA-B region and hypersensitivity reactions to abacavir.
Lancet, 359(9312), 11211122.
11. Hughes, A. R., Mosteller, M., Bansal, A. T., Davies, K., Haneline, S. A., Lai, E. H., et al.
(2004). Association of genetic variations in HLA-B region with hypersensitivity to abacavir
in some, but not all, populations. Pharmacogenomics, 5(2), 203211.
12. Gulick, R. M., Ribaudo, H. J., Shikuma, C. M., Lalama, C., Schackman, B. R., Meyer Iii, W.
A., et al. (2006). Three- vs four-drug antiretroviral regimens for the initial treatment of HIV-1
infection: A randomized controlled trial. Journal of the American Medical Association,
296(7), 769781.
13. DeJesus, E., Herrera, G., Teofilo, E., Gerstoft, J., Buendia, C. B., Brand, J. D., et al. (2004).
Abacavir versus zidovudine combined with lamivudine and efavirenz, for the treatment of
antiretroviral-naive HIV-infected adults. Clinical Infectious Diseases, 39(7), 10381046.
14. Kityo, C., Walker, A. S., Gilks, C., Gibb, D. M., Ssali, F., Babiker, A. G., et al. (2008).
Twenty-four-week safety and tolerability of nevirapine vs. abacavir in combination with zido-
vudine/lamivudine as first-line antiretroviral therapy: a randomized double-blind trial
(NORA). Tropical Medicine & International Health, 13(1), 616.
15. Hernandez, J., Cutrell, A., Bonny, T., Castillo, S. A., Brothers, C., Hee, J., et al. (2003).
Diagnosis of abacavir hypersensitivity reactions among patients not receiving abacavir in two
blinded studies [Abstract 134]. Antiviral Therapy, 8, L88.
16. Phillips, E. J., Sullivan, J. R., Knowles, S. R., & Shear, N. H. (2002). Utility of patch testing
in patients with hypersensitivity syndromes associated with abacavir. AIDS, 16(16),
22232225.
17. Martin, A. M., Nolan, D., Gaudieri, S., Almeida, C. A., Nolan, R., James, I., et al. (2004).
Predisposition to abacavir hypersensitivity conferred by HLA-B*5701 and a haplotypic
Hsp70-Hom variant. Proceedings of the National Academy of Sciences of the United States of
America, 101(12), 41804185.
18. Phillips, E. J., Wong, G. A., Kaul, R., Shahabi, K., Nolan, D. A., Knowles, S. R., et al. (2005).
Clinical and immunogenetic correlates of abacavir hypersensitivity. AIDS, 19(9), 979981.
19. Shear, N. H., Milpied, B., Bruynzeel, D. P., & Phillips, E. J. (2008). A review of drug patch
testing and implications for HIV clinicians. AIDS, 22(9), 9991007.
20. Mallal, S., Phillips, E., Carosi, G., Molina, J. M., Workman, C., Tomazic, J., et al. (2008).
HLA-B*5701 screening for hypersensitivity to abacavir. The New England Journal of
Medicine, 358(6), 568579.
21. Saag, M. S., Balu, R., Phillips, E., Brachman, P., Martorell, C., Burman, W., et al. (2008).
High sensitivity of human leukocyte antigen-B*5701 as a marker for immunologically con-
firmed abacavir hypersensitivity in white and black patients. Clinical Infectious Diseases,
46(7), 11111118.
22. Rauch, A., Nolan, D., Martin, A., McKinnon, E., Almeida, C., & Mallal, S. (2006).
Prospective genetic screening decreases the incidence of abacavir hypersensitivity reactions
in the Western Australian HIV cohort study. Clinical Infectious Diseases, 43(1), 99102.
23. Zucman, D., Truchis, P. D., Majerholc, C., Stegman, S., & Caillat-Zucman, S. (2007).
Prospective screening for human leukocyte antigen-B*5701 avoids abacavir hypersensitivity
reaction in the ethnically mixed French HIV population. Journal of Acquired Immune
Deficiency Syndromes, 45(1), 13.
24. Reeves, I., Churchill, D., & Fisher, M. (2006). Screening for HLA-B*5701 reduces the frequency
of abacavir hypersensitivity reactions [Abstract]. Antiviral Therapy, 11:L11 (Abstract no. 14).
25. Trottier B., T. R., Nguyen V.K., & Machouf N. (2007). How effectively HLA screening can
reduce the early discontinuation of abacavir in real life? 4th International AIDS society con-
ference on HIV pathogenesis and treatment. Sydney, Australia, July 2225 (Abstract no.
MOPEB002).
26. Waters, L. J., Mandalia, S., Gazzard, B., & Nelson, M. (2007). Prospective HLA-B*5701
screening and abacavir hypersensitivity: A single centre experience. AIDS, 21(18),
25332534.
27. Lalonde, R. G., Thomas, R., Rachlis, A., Gill, M. J., Roger, M., Angel, J. B., et al. (2009).
Successful implementation of a national HLA-B*5701 genetic testing service in Canada.
Tissue Antigens, 75(1), 1218.
28. Young, B., Squires, K., Patel, P., Dejesus, E., Bellos, N., Berger, D., et al. (2008). First large,
multicenter, open-label study utilizing HLA-B*5701 screening for abacavir hypersensitivity
in North America. AIDS, 22(13), 16731675.
29. Almeida, C. A. M., Martin, A. M., Nolan, D., Lucas, A., Cameron, P. U., James, I., et al. (2008).
Cytokine profiling in abacavir hypersensitivity patients. Antiviral Therapy, 13(2), 281288.
30. Martin, A. M., Almeida, C. A., Cameron, P., Purcell, A. W., Nolan, D., James, I., et al. (2007).
Immune responses to abacavir in antigen-presenting cells from hypersensitive patients. AIDS,
21(10), 12331244.
31. Chessman, D., Kostenko, L., Lethborg, T., Purcell, A. W., Williamson, N. A., Chen, Z., et al.
(2008). Human leukocyte antigen class I-restricted activation of CD8+ T cells provides the
immunogenetic basis of a systemic drug hypersensitivity. Immunity, 29(1), 165. doi:10.1016/j.
immuni.2008.04.020.
32. Phillips, E., Nolan, D., Thorborn, D., Schaefer, M., Laird, R., Rauch, A., et al. (2008). Genetic
factors predicting abacavir hypersensitivity and tolerance in HLA-B*5701 positive individu-
als. European Journal of Dermatology, 18(2), 247.
33. Phillips, E., Nolan, D., & Mallal, S. (2008). The authors reply to comments on abacavir
hypersensitivity. New England Journal of Medicine, 358(23), 25152516.
34. Schackman, B. R., Scott, C. A., Walensky, R. P., Losina, E., Freedberg, K. A., & Sax, P. E.
(2008). The cost-effectiveness of HLA-B*5701 genetic screening to guide initial antiretroviral
therapy for HIV. AIDS, 22(15), 20252033.
35. Hammond, E., Almeida, C. A., Mamotte, C., Nolan, D., Phillips, E., Schollaardt, T. A., et al.
(2007). External quality assessment of HLA-B*5701 reporting: An international multicentre
survey. Antiviral Therapy, 12(7), 10271032.
36. Martin, A. M., Nolan, D., & Mallal, S. (2005). HLA-B*5701 typing by sequence-specific
amplification: Validation and comparison with sequence-based typing. Tissue Antigens, 65(6),
571574.
37. Hammond, E., Mamotte, C., Nolan, D., & Mallal, S. (2007). HLA-B*5701 typing: Evaluation
of an allele-specific polymerase chain reaction melting assay. Tissue Antigens, 70(1), 5861.
38. Martin, A. M., Krueger, R., Almeida, C. A., Nolan, D., Phillips, E., & Mallal, S. (2006).
A sensitive and rapid alternative to HLA typing as a genetic screening test for abacavir hyper-
sensitivity syndrome. Pharmacogenetics and Genomics, 16(5), 353357.
39. Colombo, S., Rauch, A., & Rotger, M. (2008). The HCP5 single-nucleotide polymorphism:
A simple screening tool for prediction of hypersensitivity reaction to abacavir. Journal of
Infectious Diseases, 198(9), 1415.
Chapter 13
Allopurinol
Pei Chen, Shuen-Iu Hung, Shih-Yang Chen, and Yuan-Tsong Chen
Keywords Gout Uric acid metabolism Hypersensitivity reaction Severe

c utaneous adverse reactions
13.1Pharmacology of Allopurinol
Allopurinol (ZYLOPRIM), chemically described as 1,5-dihydro-4H-pyrazolo (3,4-d)

pyrimidin-4-one, is a xanthine oxidase inhibitor used to treat patients with gout,
kidney stones, and hyperuricemia associated with cancer therapy.
Uric acid is a by-product of the breakdown of certain nucleotides/nucleosides
(purines) in the body. Hyperuricemia occurs when the body produces more uric
acid than can be eliminated, or when uric acid can not be adequately excreted from
the body. Allopurinol prevents the production of uric acid by blocking the activity
of the enzyme that converts purines to uric acid. Allopurinol is a structural analog
of the natural purine base, hypoxanthine (Fig. 13.1). The drug is therefore a com-
petitive inhibitor of xanthine oxidase, the enzyme responsible for the conversion of
hypoxanthine to xanthine and xanthine to uric acid. Allopurinol is metabolized to
the xanthine analog, oxypurinol (alloxanthine), that also inhibits xanthine
oxidase.
Increases of serum uric acid concentration, referred to as hyperuricemia, can
lead to the deposition of monosodium urate crystals in various tissues. This may
result in attacks of gout, urate nephropathy, and/or tophi.
Y.-T. Chen (*)

Institute of Biomedical Sciences, Academia Sinica, Taipei 11529, Taiwan
and
Department of Pediatrics, Duke University Medical Center,
Durham, North Carolina, USA
e-mail: chen0010@ibms.sinica.edu.tw
214 P. Chen et al.
Fig. 13.1 The mechanism by which allopurinol reduces production of uric acid
13.1.1Clinical Indications
Regardless of the cause of hyperuricemia, the first-line of pharmacologic therapy

to reduce serum urate concentrations in most gout patients is inhibition of xanthine
oxidase with allopurinol. When effective and well-tolerated, allopurinol is a cost-
effective option [1, 2]. Allopurinol is the most frequently used antihyperuricemic
agent, probably because of a convenient once-daily dosage regimen and the fact
that the drug can be used to treat both urate overproduction and underexcretion [1].
Other indications for allopurinol treatment are LeschNyhan syndrome, recurrent
uric acid kidney stones refractory to other treatments, certain enzyme/blood disorders,
and hyperuricemia associated with cancer chemotherapy [3, 4]. However, because
allopurinol is not innocuous, the drug is not recommended for treatment of asymp-
tomatic hyperuricemia.
13.1.2Target Patient Populations
The Third National Health and Nutrition Examination Survey (NHANES III) esti-
mated that 5.1 million patients in the USA were afflicted with gout between 1988 and
1994 [5] At present, at least three million in the US are thought to suffer from active
gout, and an additional 36 million individuals have a history of gout [6, 7]. Data from
a US managed care claims database revealed that the prevalence of gout has increased,
from 2.9 per 1,000 in 1990 to 5.2 per 1,000 in 1999 [8]. Gout is also becoming more
13 Allopurinol 215
prevalent in other countries, including New Zealand and Taiwan [9, 10]. The incidence
of gout has been found to correlate strongly with serum urate concentration, increasing
markedly when this exceeds 420mmol/L (7.0mg/dL) [11].
13.1.3Alternative Medications
Pharmacologic urate-lowering drugs include uricosurics (probenecid, sulfinpyrazone,

and benzbromarone), which increase the excretion of uric acid in the urine; xanthine
oxidase inhibitors (allopurinol and the recently approved febuxostat), which inhibit
uric acid production; and the experimental uricase (Rasburicase), which degrades
urate [1, 12]. Pharmacotherapy thus consists of a choice between a medication that
reduces urate production and a drug that increases urate excretion. Uricosuric drugs
such as probenecid and sulfinpyrazone increases renal urate clearance and are consid-
ered antihyperuricemic agents for patients with primary gout who present substan-
tially decreased renal urate excretion. Such drugs are contraindicated in patients with
high urinary concentrations of uric acid and are not recommended for patients
with chronic renal insufficiency. Allopurinol remains the most frequently used
antihyperuricemic agent because of a convenient once-daily dosage regimen and an
ability to treat both urate overproduction and underexcretion [1]. Febuxostat, a
recently developed xanthine oxidase inhibitor, is considered a reasonable alternative
for patients intolerant to allopurinol. Febuxostat is now approved in Europe and the
USA, but is not yet available in other countries.
13.1.4Dosing, Pharmacokinetics, and Pharmacodynamics
Allopurinol has been approved by the US Food and Drug Administration (FDA) at
doses up to 800mg/day. The expert consensus EULAR spelled out guidelines have
reinforced FDA dosing guidelines for allopurinol in patients with preserved renal
function [2, 13]. These guidelines recommend that allopurinol be initiated at a dose
of 100mg/day, and increased by 100mg/day every 14 weeks until the target serum
urate level (<6mg/dL) is achieved or the maximum appropriate allopurinol dose is
reached. FDA dosing guidelines have also advocated daily doses of 200300mg
allopurinol as adequate for most patients with mild gout, and an average daily dose
of 400600 mg allopurinol is expected to control hyperuricemia in patients with
moderately severe tophaceous gout [13]. However, the vast majority of allopurinol
prescriptions are for doses 300mg/day, which often fails to adequately treat the
hyperuricemia of gout. This situation has arisen from consideration of longstanding
guidelines for allopurinol use calibrated to renal function and serum concentrations
of oxypurinol, to avoid allopurinol hypersensitivity syndrome (HSS) [13]. These
guidelines are not evidence-based and fail to adequately treat hyperuricemia and
also fail to prevent allopurinol HSS [7, 13].
216 P. Chen et al.
A longer-term goal of allopurinol therapy is to prevent further attacks, eliminate

tophi, and prevent joint destruction, by consistently reducing the concentration of
urate [1, 2, 14]. As urate is insoluble in physiologic solutions at concentrations
exceeding 6.77.0 mg/dL, current guidelines for inhibiting ongoing urate crystal
deposition, reducing total body urate stores, and resolving macroscopic tophi rec-
ommend continuing (lifelong) reduction of serum urate concentrations to <6mg/dL
(approximately 360mmol/L), ideally to 56mg/dL [15]. Uric acid levels usually
begin to fall within 23 days of commencement of allopurinol treatment, but return
to original levels within 710 days if treatment is stopped. Thus, several months of
therapy may be required before gout can be controlled.
13.2Toxicities
Although allopurinol is well-tolerated by most patients, approximately 2% develop

a mild skin reaction (rash and itching) [3] and 0.4% suffer from allopurinol HSS,
a severe cutaneous adverse reaction that can be life-threatening [16]. This syndrome
may present with fever, eosinophilia, and rashes, ranging from maculopapular
through StevensJohnson syndrome (SJS) to toxic epidermal necrolysis (TEN) and
multiorgan involvement, with significant mortality and morbidity [17]. Potential
risk factors for allopurinol HSS include Han Chinese ethnicity, older age, and
underlying renal impairment [18]. The risk of allopurinol HSS may be reduced by
administering smaller doses of drug from 50 to 300mg/day, with the exact dose
being directly proportional to creatinine clearance [19]. However, such regimen is
not well adhered in clinical practice, and their effectiveness in reducing allopurinol
hypersensitivity reactions has been challenged [20].
13.3Pharmacogenetics and Pharmacogenomics of Allopurinol
Allopurinol is a frequent cause of severe cutaneous adverse reactions (SCAR),

including drug HSS, SJS, and TEN. Although rare, the mortality rate of patients
with allopurinol-induced SCAR can be as high as 26% [17, 21]. SJS is character-
ized by high fever; malaise; and a rapidly developing blistering exanthema of mac-
ules and target-like lesions, accompanied by mucosal involvement. TEN has similar
presentations with even more extensive skin detachment and a higher mortality rate
(3040%) [22]. HSS has systemic manifestations with multiorgan involvement, in
addition to exanthema [17].
A recent European multinational casecontrol study found that allopurinol has
become the drug most commonly associated with SJS/TEN in Europe and Israel,
accounting for 17.4% of 379 patients with SJS or TEN [23]. Allopurinol also plays
a major role in the induction of SJS or TEN in Asian populations [18, 24, 25]. Data
from 230 consecutive SJS/TEN patients assessed over a 5-year period (19972002)
13 Allopurinol 217
in the Chang Gung Memorial Hospital Health System of Taiwan showed that
allopurinol was the second leading cause of SJS/TEN, after carbamazepine [26].
However, because of recent awareness of carbamazepine-induced SJS, allopurinol
has now become the leading drug responsible for severe adverse reactions in
patients receiving compensation from the Taiwan Drug Relief Foundation (TDRF)
(http://www.tdrf.org.tw).
Drug hypersensitivity is typically dose-independent and unpredictable, resulting
from exaggerated immune-mediated reactions [27]. Susceptibility to such idiosyn-
cratic reactions is thought to be genetically determined, and familial predisposition
to allopurinol hypersensitivity has been reported [28]. To identify genetic factors
predisposing an individual to allopurinol-induced SCAR, we assessed polymor-
phisms in genes related to drug metabolism and the immune response in 51 well-
characterized patients with allopurinol-induced SCAR, including 30 with HSS, 13
with SJS, 5 with SJS/TEN and 3 with TEN, and compared allele frequencies in
such patients with those of 228 controls, including 135 allopurinol-tolerant sub-
jects, and 93 healthy individuals [29]. All participants were unrelated Han Chinese
residing in Taiwan. We found that the HLA-B*5801 allele was present in all 51
(100%) patients with allopurinol-induced SJS/TEN/HSS, but only in 15% (20/135)
of allopurinol-tolerant controls (corrected P = 4.7 1024, OR = 580.3, 95%
CI=34.49,780.9), and 20% (19/93) of healthy controls (corrected P=8.11021718,
OR = 393.5, 95% CI = 23.26,665.26) [29]. The occurrence of HLA-A*3303,
HLA-B*5801, HLA-Cw*0302, and HLA-DRB1*0301 alleles of the HLA-B*5801
ancestral haplotype was also significantly associated with allopurinol hypersensi-
tivity, as were several haplotype-specific polymorphisms near the HLA-B gene,
including SNPs located in BAT3, MSH5, and MICB. Recombinant mapping data
further concluded that the HLA-B*5801 allele itself was the major susceptibility
gene for allopurinol-induced SCAR in the Han Chinese population [29].
As the distribution of the HLA-B*5801 allele varies among different popula-
tions (e.g., 24% in Africans, 16% in Caucasians, 315% in Asian Indians, and
8.810.9% in Chinese (http://www.ebi.ac.uk/imgt/hla/stats.html, Accessed 10 May
2007), an association between this allele and allopurinol-induced SCAR may also
be present in other ethnic groups. Indeed, such an association has been observed in
patients from Europe, Japan, and Thailand [3033].
Data from a European study showed that 55% (15/27) of patients of European
ancestry with allopurinol-induced SJS/TEN carried the HLA-B*5801 allele [30],
compared with 1.5% (28/1,822) of controls, resulting in an odds ratio of 80 (95%
CI=34187; corrected P<106) (Table 13.1). Among 31 patients with allopurinol-
induced SJS/TEN enrolled in the European study, four were of non-European
ancestry (two from Asia, one from South America, and one from Africa) and all
four had the HLA-B*5801 allele [30]. In addition, all three Japanese patients with
allopurinol-associated SCAR carried the HLA-B*5801 allele [31], and a moderate
but statistically significant association was observed between HLA-B*5801 and
allopurinol-associated SJS/TEN in Japanese patients ([32]; Table 13.1). Recently,
HLA-B*5801 was found to be present in all of 27 patients with allopurinol-induced
SJS/TEN from Thailand [33]. Moreover, the risk of allopurinol-induced SJS/TEN
218
Table 13.1 HLA-B*5801 in allopurinol-induced severe cutaneous adverse reactions (SCAR)

Study number 1 2 (European study) 3 4
Study Non-European ancestry
population Han Chinesea Caucasianb (two Asians) Japanesec Thaid
Case 51/51 (100%) 15/27 (55%) 4/4 (100%) 7/13 (54%) 27/27 (100%)
Control 20/135 (15%) 28/1822 (1.5%) 6/493 (1.2%) 7/54 (13%)
Odds ratio 580.3 80 94.7 348.3
(95% CI) (34.49780.9) (34187) (24.4367.3) (19.26,336.9)
P value 4.71024e <106e 1.71109 1.611013
Reference Hung etal. [29] Lonjou etal. [30] Kaniwa etal. [32]; Wichittra etal. [33]
Dainichi etal. [31]
a
Case: Allopurinol-SCAR; control: Tolerant control
b
ase: Allopurinol-SJS/TEN; control: A mixed European population
c
Case: Allopurinol-SJS/TEN; control: Japanese population
d
Case: Allopurinol-SJS/TEN; control: Tolerant control
e
Adjusted using Bonferronis correction for multiple comparisons to account for observed alleles
P. Chen et al.
13 Allopurinol 219
was found to be significantly higher in patients with than without the HLA-B*5801
allele (OR 348.3; 95% CI=19.26336.9; P=1.611013) (Table 13.1).
Ethnically associated differences in the prevalence of HLA-B*5801 among
patients with allopurinol-associated SCAR (55% in European populations and
100% in Thailand and Southeast Asia, including Taiwan) may be linked to the very
different allelic frequencies in these populations. Indeed, the dbMHC data base
shows that the allele frequency of HLA-B*5801 is significantly higher in Thais
(8.6%) and Han Chinese (7.3%) than in Europeans (0.8%) or Japanese (0.61%).
In Taiwan, the allele frequency is 10% with carrier prevalence of 20% [29].
Whereas, all allopurinol-SCAR patients from Southeast Asia carried at least one
HLA-B*5801 allele, it was also found that additional patients tolerant to allopurinol
carried the allele. This suggests that the HLA-B*5801 allele is necessary but not
sufficient for development of allopurinol-SCAR. Other genes may also be involved
in the mechanism of pathogenesis, such as costimulatory molecules involved in the
interaction between antigen-presenting cells and T cell interaction [29].
13.4Clinical Utility of PGx Testing in Personalized Medicine
One important goal of pharmacogenomics is to prevent severe ADRs by using a

simple test to screen out patients at risk. Before any test can be used in clinical
practice, however, several important issues must be considered [34, 35]. These
include the incidence and severity of the adverse event; the sensitivity and specific-
ity of the predictive marker; and whether equally effective, alternative medications
are available for individuals who test positive.
Although the incidence of SCAR is relatively low compared with common dis-
eases, SCAR is serious and even life-threatening. Many surviving patients have
long-term complications, some of which result in permanent damage, such as
blindness and renal failure. The medical cost associated with ADRs is also very
high. Allopurinol is the most widely-prescribed urate-lowering agent in the world,
with about 2.8 million prescriptions per year written for the treatment of gout in the
US [6]. Recent extensive research has attempted to elucidate the roles of uric acid
and oxidative stress in the development of various diseases [36]. These studies have
indicated that the spectrum of conditions that can be treated with allopurinol may
expand to include metabolic syndrome and related disorders, chronic kidney
disease, and the adverse effects of cancer treatment [37]. As potential indications
increase, the number of prescriptions for allopurinol may also increase.
Data from a retrospective case-control study in Taiwan, using an allopurinol-
tolerant group as the control, found that the HLA-B*5801 allele would have 100%
sensitivity and 85% specificity in detecting the risk of allopurinol-associated SCAR
(Table 13.2). Assume the prevalence rate of 0.4% (four allopurinol-associated
SCAR in 1,000 allopurinol users), the presence of B*5801 allele would have a
2.6% positive predictive value (PPV) for detecting allopurinol-induced SCAR,
whereas absence thereof would have 100% negative predictive value (NPV)
220
Table 13.2 HLA-B*5801 as a test for Allopurinol-induced SCAR

Han Chinese European Thai
Affected Tolerant Affected Tolerant Affected Tolerant
Patients with the allele 51 (100%) 1,905 (15%) 15 (55.6%) 101 (1.5%) 29 (100%) 936 (12.96%)
Patients without the allele 0 (0%) 10,794 (85%) 12 (44.4%) 6,622 (98.5%) 0 (0%) 6,285 (87.04%)
Sensitivity (%) 100 55.6 100
Specificity (%) 85 98.5 87
Positive predictive value (%) 2.6 12.9 3.01
Negative predictive value (%) 100 99.8 100
Assumed prevalence rate (%) 0.4 0.4 0.4
ADRs SCAR SJS or TEN SJS or TEN
Reference Hung etal. [29] Lonjou [30] Wichittra etal. [33]
P. Chen et al.
13 Allopurinol 221
(Table 13.2). Similar sensitivity, specificity, PPV, and NPV results can be found in
European and Thai populations (Table 13.2).
Results from Han Chinese patients (Table 13.2) indicate that 15% of those who
test positive for the HLA-B*5801 allele may never develop SCAR and may there-
fore be unnecessarily denied the drug. However, the life-threatening consequences
of SJS/TEN, and the availability of alternative drugs, may justify the withholding
of allopurinol from such patients. Hyperuricemia in HLA-B*5801-positive gout
patients can be treated with uricosurics (e.g., probenecid, sulfinpyrazone, or benz-
bromazone), unless creatinine clearance is <50mL/min or urate overproduction has
been documented. Such patients may be treated with febuxostat, a newly approved
xanthine oxidase inhibitor.
In addition to the HLA-B*5801, other factors such as renal function or virus
infection may be risk factors for the development of allopurinol-induced SCAR
[17, 29, 38]. Current guidelines suggest that the allopurinol dose be calibrated to
renal function and serum concentration of oxypurinol to avoid allopurinol HSS.
However, lower doses (e.g., 300mg/day) may fail to adequately treat hyperurice-
mia. Moreover, these guidelines have not succeeded in prevention of allopurinol-
associated SCAR. Using the presence of the HLA-B*5801 allele as a pharmacogenetic
test, in conjunction with tests of renal function, the vast majority of patients at low
risk for development of allopurinol-induced SCAR (test-negative patients) can be
treated with higher doses of allopurinol sufficient to achieve normouricemia.
Prospective studies will be very helpful to determine the effectiveness of HLA-
B*5801 testing in prevention of allopurinol-associated SCAR. Such a study is
ongoing in Taiwan.
13.5Conclusion
Allopurinol is a frequent cause of SCAR, and the HLA B*5801 allele is a marker
of susceptibility. The clinical severity of SCAR, the availability of alternative medi-
cations, and the high sensitivity and specificity of the HLA B*5801 marker provide
a plausible basis for the development of tests to identify individuals at risk of this
potentially life-threatening condition caused by allopurinol.
References
1. Terkeltaub, R. A. (2003). Clinical practice. Gout. The New England Journal of medicine, 349,
16471655.
2. Zhang, W., Doherty, M., Bardin, T., Pascual, E., Barskova, V., Conaghan, P., et al. (2006).
EULAR Standing Committee for International Clinical Studies Including Therapeutics.
EULAR evidence based recommendations for gout. Part II: Management. Report of a task
force of the EULAR Standing Committee for International Clinical Studies Including
Therapeutics (ESCISIT). Annals of the Rheumatic Diseases, 65, 13121324.
222 P. Chen et al.
3. Wortmann, R. L. (2002). Gout and hyperuricemia. Current Opinion in Rheumatology, 14,

2816.
4. Chohan, S., Becker, M. A. (2009). Update on emerging urate-lowering therapies. Current
Opinion in Rheumatology, 21, 143149.
5. Kramer, H. M., Curhan, G. (2002). The association between gout and nephrolithiasis: the
National Health and Nutrition Examination Survey III, 19881994. American Journal of
Kidney Diseases: The Official Journal of the National Kidney Foundation, 40, 3742.
6. Lawrence, R. C., Felson, D. T., Helmick, C. G., Arnold, L. M., Choi, H., Deyo, R. A., etal.
(2008). National Arthritis Data Workgroup. Estimates of the prevalence of arthritis and other
rheumatic conditions in the United States. Part II. Arthritis and Rheumatism, 58, 2635.
7. Terkeltaub, R. (2009). Novel therapies for treatment of gout and hyperuricemia. Arthritis
Research & Therapy, 11, 236246.
8. Wallace, K. L., Riedel, A. A., Joseph-Ridge, N., Wortmann, R. (2004). Increasing prevalence
of gout and hyperuricemia over 10 years among older adults in a managed care population.
The Journal of Rheumatology, 31, 15821587.
9. Mikuls, T. R., Farrar, J. T., Bilker, W. B., Fernandes, S., Schumacher, H. R. Jr, Saag, K. G.
(2005). Gout epidemiology: results from the UK General PracticeResearch Database,
19901999. Annals of the Rheumatic Diseases, 64, 267272.
10. Roddy, E., Zhang, W., Doherty, M. (2007). The changing epidemiology of gout. Nature
Clinical Practice. Rheumatology, 3, 443449.
11. Campion, E. W., Glynn, R. J., DeLabry, L. O. (1987). Asymptomatic hyperuricemia. Risks and
consequences in the normative aging study. The American Journal of Medicine, 82, 421426.
12. Bieber, J. D., Terkeltaub, R. A. (2004). Gout: on the brink of novel therapeutic options for an
ancient disease. Arthritis and Rheumatism, 50, 24002414.
13. Chao, J., Terkeltaub, R. (2009). A critical reappraisal of allopurinol dosing, safety, and effi-
cacy for hyperuricemia in gout. Current Rheumatology Reports, 11, 135140.
14. Romeijnders, A. C., Gorter, K. J. (2002). Summary of the Dutch College of General
Practitioners Gout Standard. Nederlands Tijdschrift Voor Geneeskunde, 146, 309313.
15. Perez-Ruiz, F., Liot, F. (2007). Lowering serum uric acid levels: what is the optimal target
for improving clinical outcomes in gout? Arthritis and Rheumatism, 57, 13241328.
16. Pluim, H. J., van Deuren, M., Wetzels, J. F. (1998). The allopurinol hypersensitivity syndrome.
The Netherlands Journal of Medicine, 52, 10710.
17. Arellano, F., Sacristan, J. A. (1993). Allopurinol hypersensitivity syndrome: a review. The
Annals of Pharmacotherapy, 27, 337343.
18. Lee, H. Y., Ariyasinghe, J. T. N., Thirumoorthy, T. (2008). Allopurinol hypersensitivity
syndrome: a preventable severe cutaneous adverse reaction? Singapore Medical Journal, 49,
384387.
19. Hande, K. R., Noone, R. M., Stone, W. J. (1984). Severe allopurinol toxicity: description and
guidelines for prevention in patients with renal insufficiency. The American Journal of
Medicine, 76, 4756.
20. Vazquez-Mellado, J., Morales, E. M., Pacheco-Tena, C., Burgos-Vargas, R. (2001). Relation
between adverse events associated with allopurinol and renal function in patients with gout.
Annals of the Rheumatic Diseases, 60, 981983.
21. Aubock, J., Fritsch, P. (1985). Asymptomatic hyperuricaemia and allopurinol induced toxic
epidermal necrolysis. British Medical Journal, 290, 19691970.
22. Roujeau, J. C. (1994). The spectrum of StevensJohnson syndrome and toxic epidermal
necrolysis: a clinical classification. The Journal of Investigative Dermatology, 102, 28S30S.
23. Halevy, S., Ghislain, P. D., Mockenhaupt, M., Fagot, J. P., Bouwes Bavinck, J. N., Sidoroff,
A., etal. for the EuroSCAR Study Group. (2008). Allopurinol is the most common cause of
StevensJohnson syndrome and toxic epidermal necrolysis in Europe and Israel. Journal of
the American Academy of Dermatology, 58, 2532.
24. Lin, M. S., Dai, Y. S., Pwu, R. F., Chen, Y. H., Chang, N. C. (2005). Risk estimates for drugs
suspected of being associated with StevensJohnson syndrome and toxic epidermal necroly-
sis: a case-control study. Internal Medicine Journal, 35, 188190.
13 Allopurinol 223
25. Jantararoungtong, T., Tiamkao, S., Vannaprasath, S., Choonhakarn, C., Auvichayapat, N.,
Tassaneeyakul, W. (2009). StevensJohnson syndrome and toxic epidermal necrolysis in
Srinagarind hospital: a retrospective study of causative drugs and clinical outcome. Thai
Journal of Pharmacology, 31, 4143.
26. Hung, S. I., Chung, W. H., Chen, Y. T. (2005). HLA-B genotyping to detect carbamazepine-
induced StevensJohnson syndrome: implications for personalizing medicine. Personalazed
27. Roujeau, J. C. (2006). Immune mechanisms in drug allergy. Allergology International:
Official Journal of the Japanese Society of Allergology, 55, 2733.
28. Melsom, R. D. (1999). Familial hypersensitivity to allopurinol with subsequent desensitiza-
tion. Rheumatology, 38, 1301.
29. Hung, S. I., Chung, W. H., Liou, L. B., Chu, C. C., Lin, M., Huang, H. P., et al. (2005).
HLA*5801 allele as a genetic marker for severe cutaneous adverse reactions caused by
allopurinol. Proceedings of the National Academy of Sciences of the United States of America,
102, 41344139.
30. Lonjou, C., Borot, N., Sekula, P., Ledger, N., Thomas, L., Halevy, S., etal. 2008;RegiSCAR
study group. A European study of HLA-B in StevensJohnson syndrome and toxic epidermal
necrolysis related to five high-risk drugs. Pharmacogenetics and Genomics, 18, 99107.
31. Dainichi, T., Uchi, H., Moroi, Y., Furue, M. 2007;StevensJohnson syndrome, drug- induced
hypersensitivity syndrome and toxic epidermal necrolysis caused by allopurinol in patients
with a common HLA allele: what causes the diversity? Dermatology, 215, 8688.
32. Kaniwa, N., Saito, Y., Aihara, M., Matsunaga, K., Tohkin, M., Kurose, K., et al. JSAR
research group (2008) HLA-B locus in Japanese patients with anti-epileptics and allopurinol-
related StevensJohnson syndrome and toxic epidermal necrolysis. Pharmacogenomics, 9,
16171622.
33. Wichittra, T., Thawinee, J., Chen, P., Lin, P. Y., Somsak, T., Usanee, K., etal. (2009). Strong
association between HLA-B*5801 and allopurinol-induced StevensJohnson syndrome and toxic
epidermal necrolysis in a Thai population. Pharmacogenetics and Genomics, 19, 704709.
34. Nolan, D., Gaudieri, S., Mallal, S. (2003). Pharmacogenetics: a practical role in predicting
antiretroviral drug toxicity. Journal of HIV Therapy, 8, 3641.
35. Phillips, K. A., Veenstra, D., Van Bebber, S., Sakowski, J. (2003). An introduction to cost-
effectivness and cost-benefit analysis of pharmacogenomics. Pharmacogenomics, 4,
231239.
36. Pacher, P., Nivorozhkin, A., Szabo, C. (2006). Therapeutic effects of xanthine oxidase inhibi-
tors: Renaissance half a century after the discovery of allopurinol. Pharmacological Reviews,
58, 87114.
37. Suzuki, I., Yamauchi, T., Onuma, M., Nozaki, S. (2009). Allopurinol, an inhibitor of uric acid
synthesis CAN it be used for the treatment of metabolic syndrome and related disorders?
Drugs of Today, 45, 363368.
38. Suzuki, Y., Inagi, R., Aono, T., Yamanishi, K., Shiohara, T. (1998). Human herpesvirus 6
infection as a risk factor for the development of severe drug-induced hypersensitivity syn-
drome. Archives of Dermatology, 134, 11081112.
Chapter 14
Carbamazepine and Its Structurally-Related
Antiepileptics
Shuen-Iu Hung, Wen-Hung Chung, Jing-Jane Tsai,

and Yuan-Tsong Chen
Keywords Epilepsy Stevens Johnson syndrome Seizures
14.1Pharmacology of Carbamazepine
14.1.1Clinical Indications and Patient Populations

for Carbamazepine
Carbamazepine (CBZ) (Equetro, Carbatrol, Epitol, Tegretol) is one of a class

of antiepileptic drugs (AEDs) that aim to prevent or reduce the severity of abnormal
nerve activity in the brain, thereby blocking seizures [1, 2]. CBZ was initially
approved for the treatment of epilepsy in the US in 1974. Currently, the indications
for CBZ use include seizures, trigeminal neuralgia, and bipolar disorder [3]. CBZ
is now first-line therapy for the treatment of partial seizures and tonic-clonic
seizures. The mechanism of action of CBZ antiepileptic effects is thought to be
sodium channel blocking, limiting the repetitive firing of action potentials, slowing
the rate of recovery of voltage-activated Na+ channels from inactivation, decreasing
the activity of nerve cells, and preventing them from firing abnormally in the brain
[1, 3]. In addition, CBZ may also prevent abnormal brain signals from spreading to
other parts of the brain. Complex partial seizures most frequently arise from the
repetitive firing of action potentials in the temporal lobe of the brain and cause
impairment of consciousness with or without accompanying automatisms. Patients
who have complex partial seizures can be treated with CBZ with a greater response
and more significant improvement than patients with other types of seizures [1, 4].
Although CBZ is approved to treat various types of seizures (e.g., partial seizures,
Hung S.I. ()
Institute of Pharmacology, National Yang-Ming University,
Taipei, 11221, Taiwan
e-mail: sihung@ym.edu.tw
A.H.B. Wu and K.-T J. Yeo (eds.), Pharmacogenomic Testing in Current Clinical Practice, 225
Molecular and Translational Medicine,DOI 10.1007/978-1-60761-283-4_14,
226 S.-I. Hung et al.
generalized tonic-clonic seizures, or mixed seizure patterns), it is usually not

effective at treating absence seizures (petit mal seizures) [5].
In the past decade, in addition to its antiepileptic effects, CBZ has increasingly
become the medication of choice for the treatment of bipolar disorder, particularly
acute mania [6]. CBZ has been approved to treat manic or mixed episodes (which
include characteristics of both mania and depression) associated with bipolar
disorder, also known as manic depression [6]. CBZ also has good therapeutic
effects in trigeminal neuralgia (also known as tic douloureux), a nerve condition
that causes episodes of facial pain (typically cheek or jaw pain) [7]. It is thought
that CBZ works to treat trigeminal neuralgia by blocking the nerve signals that
cause pain and other unpleasant sensations. Moreover, CBZ also has antidiuretic
effects that may be associated with reduced levels of antidiuretic hormone.
14.1.2Alternative AED Medications for Carbamazepine
Alternative medication options for CBZ in the treatment of epilepsy include a

number of different AEDs. In addition to CBZ, the major drugs used in partial
seizures and generalized tonic-clonic seizures comprise other first-line AEDs such
as oxcarbazepine (OXC), phenytoin (PHT), pentobarbital (PHB), valproate (also
called valproic acid), and lamotrigine [LTG], among others; and also second-line
AEDs such as gabapentin, topiramate, vigabatrin, and levetiracetam, and the like
[8, 9]. As the mechanism of action of CBZ is mainly due to its prevention of repeti-
tive firing of action potentials by binding to the alpha-subunit of voltage-gated
sodium channels in neurons [10], the alternative AEDs (e.g., OXC, PHT, PHB,
valproate, LTG, topiramate, and vigabatrin) with similar mechanisms of action are
good choices for the treatment of partial seizures, generalized tonic-clonic seizures,
or mixed seizure patterns. In addition to the mechanism of action, the selection of
alternative medications for CBZ may be determined by the effectiveness, price, side
effects, cross-reactivity of skin reactions, and the practical experience of physi-
cians. Since side effects are commonly associated with the administration of AEDs,
monotherapy, and initiation with a single agent at a time are suggested. Furthermore,
many AEDs are hepatic enzyme inducers; for example, CBZ is a CYP450 enzyme
inducer. CBZ and its alternative AEDs can potentially interact with many other
medications [11]. Therefore, prescribing CBZ or one of its alternative AEDs in
combination with other medicines must be done with caution to observe for poten-
tial drug-drug interactions.
Among the alternative medications, OXC (Trileptal) is a 10-keto analog of
CBZ and acts as a prodrug [1]. The mechanism of action of OXC is similar to that
of CBZ, however, it is a less potent enzyme inducer than CBZ, and OXC is metabo-
lized by different enzymes [12]. As a prodrug, orally ingested OXC is rapidly
converted to its main active metabolite, a 10-monohydroxy derivative, which
is inactivated by glucuronide conjugation and eliminated by renal excretion. OXC
has been approved for monotherapy or adjunctive therapy for partial seizures in
adults and partial seizures in children aged 416.
14 Carbamazepine and Its Structurally-Related Antiepileptics 227
PHT (diphenylhydantoin, Dilantin) is a hydantoin and exerts antiseizure

activity without causing general depression of the CNS [1]. PHT is effective against
all types of partial and tonic-clonic seizures, but not absence seizure. The mecha-
nism of antiepileptic action of PHT is similar to CBZ, which is mediated by slowing
the rate of recovery of voltage-activated Na+ channels from inactivation, therefore
limiting the repetitive firing of action potentials [1]. The pharmacokinetic charac-
teristics of PHT are influenced by its binding to serum proteins, by the nonlinearity
of its elimination kinetics, and by its metabolism by CYPs. The side reactions/
toxicity of PHT includes cerebellar-vestibular effects, gastrointestinal symptoms,
gingival hyperplasia, osteomalacia, megaloblastic anemia, hirsutism, and hyper
sensitivity reactions, among others [1].
LTG (Lamictal) is a phenyltriazine derivative, and its antiepileptic action has
been proposed to involve the inhibition of voltage-sensitive sodium channels, a
mechanism similar to those of PHT and CBZ, and also involve the modulation of
presynaptic transmitter release of excitatory amino acids [1]. LTG is effective
against a broad spectrum of seizures, including partial and secondarily generalized
tonic-clonic seizures in adults, and Lennox-Gastaut syndrome in both children and
adults. In addition, LTG was also approved for maintenance treatment of bipolar I
disorder. LTG is completely absorbed from the gastrointestinal tract and is metabo-
lized primarily by glucuronidation. Common side effects of LTG include head-
aches, dizziness, and insomnia. In extremely rare cases, LTG has been known to
cause drug reaction with eosinophilia and systemic symptoms (DRESS), Stevens
Johnson syndrome (SJS), or toxic epidermal necrolysis (TEN) [1].
14.1.3Dosing, Pharmacokinetics and Pharmacodynamics

of Carbamazepine
CBZ (Equetro, Carbatrol, Epitol, Tegretol) comes in several different formulae,

including tablets, chewable tables, extended-release tablets, extended-release
capsules and suspension (liquid). All are administered orally. Several different
manufacturers make these drugs, and not all forms are approved for each indica-
tion. For example, Equetro is approved to treat bipolar disorder only, while
Carbatrol, Epitol, Tegretol, and generic CBZ are approved to treat epilepsy and
trigeminal neuralgia, but not bipolar disorder. The dosage of CBZ for treating
epilepsy depends upon a patients age and weight. For adults and children over
12years old, the recommended starting dose is 200mg twice daily. For children
aged 612 years, the recommended starting dose is 100mg twice daily. For children
under 6 years, the starting dose is calculated by weight: 1020mg/kg total per day,
divided into two or three doses per day (1020mg/kg/day b.i.d. or t.i.d.). When
combining CBZ with existing anticonvulsant therapy, it should be added gradually
while the other AEDs are maintained or gradually decreased [13].
The relationship between the dose of CBZ and concentrations of the drug in plasma
is not a simple one. Because of its limited aqueous solubility and high lipid solubility,
the absorption of CBZ is slow but complete following oral administration [1].
Approximately 75% of CBZ binds to plasma proteins. After oral administration of

CBZ tablets, its peak plasma level can be achieved after 45h [13]. CBZ is lipo-
philic and distributes rapidly into all tissues, including the liver, kidney, and brain.
The therapeutic concentrations of CBZ for adults are reported to be between 6 and
12 mg/ml, although considerable variation occurs. The concentrations of CBZ in
CSF appear to correspond to the levels of free drug in plasma. Side effects referable
to the CNS are frequent at concentrations above 9mg/ml. The elimination half-life
of CBZ averages 35h, ranging from 18 to 65h. Since CBZ induces hepatic meta-
bolic enzymes, autoinduction of CBZ decreases its half-life to 1020h with chronic
administration [1, 14, 15].
CBZ is metabolized in the liver where it exhibits autoinduction. As an inducer of
hepatic enzymes, CBZ can enhance the expression of the hepatic cytochrome (CYP)
P450 enzymes, including CYP2C, CYP3A, and UGT, thus enhancing the metabo-
lism of drugs degraded by these enzymes and altering clearance of these drugs [14,
15]. Hepatic P450 CYP3A4 is primarily responsible for biotransformation of CBZ.
It converts CBZ to CBZ-10, 11-epoxide (10, 11-epoxyCBZ), which also has anti-
epileptic activity since CBZ-10, 11-epoxide is able to limit sustained repetitive
firing at therapeutic concentration [16]. In the liver, CBZ-10, 11-epoxide is detoxi-
fied by microsomal epoxide hydratase to inactive compounds excreted in the urine
primarily as glucuronides.
14.1.4Side Effects and Toxicities of Carbamazepine
CBZ can cause two types of ADRs. Type A is dose-dependent and results from
acute overdose or during chronic therapy. CBZ may produce common type A
ADRs such as drowsiness, vertigo, fatigue, unsteadiness, dizziness, nausea, and
vomiting; these ADRs require monitoring and dose adjustment [17, 18]. Less com-
mon ADRs include dry mouth, ataxia, diplopia, and blurred vision [17, 18]. In
comparison, CBZ may produce idiosyncratic type B ADRs, which are dose-
independent, not related to its pharmacologic reactions, often serious and life-
threatening, and historically unpredictable. Type B ADRs experienced with CBZ
include hypersensitivity reactions, hepatitis, and blood dyscrasias (aplastic anemia,
agranulocytosis) [19, 20]. Among them, cutaneous adverse drug reactions (cADRs)
are the most common hypersensitivity reactions. cADRs can present in a variety of
ways from mild maculopapular eruption (MPE) with increasing severity to hyper-
sensitivity syndrome (HSS), SJS, and TEN [21].
MPE is characterized by cutaneous itchy, erythematous macules and papules,
which usually spontaneously resolve within 12 weeks after withdrawing the caus-
ative drugs. The rate of CBZ-induced MPE is ~3.7% [22]. HSS, also called drug
reaction with eosinophilia and systemic symptoms (DRESS), is characterized by
systemic manifestations with fever, hematologic abnormalities, and multiorgan
involvement (e.g., hepatitis and nephritis) in addition to skin rash [21]. MPE and
HSS are nonblistering cADRs. In comparison, SJS and TEN are two of the most
serious blistering cADRs with a 1050% mortality rate [21]. SJS and TEN are
characterized by a rapidly developing painful or burning blistering exanthema of
purpuric macules and target-like lesions accompanied by mucosal involvement and
skin detachment to a varying extent. SJS is defined as skin detachment of less than
10% and TEN as skin detachment greater than 30% [23]. Mucous membranes com-
monly affected by SJS and TEN include the eyes, oropharynx, and anogenital areas.
Survivors are at risk of permanent complications such as blindness due to corneal
damage [23].
14.2Pharmacogenetics for Carbamazepine Hypersensitivity
14.2.1HLA-B*1502 and CBZ-SJS/TEN in Han Chinese
As mentioned above, CBZ can cause a variety of hypersensitivity reactions, including

mild MPE to severe HSS and SJS/TEN. We previously studied the genetic polymor-
phisms of genes involved in immune-regulation and CYP-P450 single nucleotide
polymorphisms (SNPs) in 44 Han Chinese patients in Taiwan with CBZ-SJS/TEN
and compared their allele frequencies with those of 101 tolerant controls [24]. We
found that all 44 patients (100%) who developed CBZ-SJS/TEN carried the HLA-
B*1502 allele, while only 3% of the CBZ-tolerant group (odds ratio: 2,504, cor-
rected P=3.131027), and 8.6% of the 93 healthy subjects carried the allele [24].
Our extension study, which included 60 patients who experienced CBZ-SJS/TEN
and 144 tolerant controls, revealed that 59 of 60 CBZ-SJS/TEN patients tested posi-
tive for HLA-B*1502 allele, while only 6 (4.16%) of the 144 tolerant controls car-
ried the allele (odds ratio: 1,357, corrected P=1.61041) (Table 14.1) [25].
To determine whether the HLA-B gene itself or genes in the vicinity of the B
locus contributes to the main susceptibility of CBZ-SJS/TEN, we performed fine-
mapping in the MHC region using 220 SNPs, 20 short tandem repeat polymor-
phisms (STRPs), and HLA alleles [25]. We found that polymorphisms located
between HLA-DRA and HLA-C showed strong association. In particular, the TT
or GT genotypes of rs3130690, 36kb telemetric to the HLA-B locus, were present
in 98% CBZ-SJS/TEN patients, but only in 5% of tolerant controls. MICA*019,
47 kb centromeric to the HLA-B locus, was present in 95% of CBZ-SJS/TEN
patients, but only in 15% of tolerant controls. The recombinant map of Cw*0801-
HLABC-CA*119-rs3130690T-B*1502-MICA*019 defined the susceptible region
within 86kb (i.e., between the T allele of rs3130690 and MICA*019) flanking the
B*1502 gene in the 4Mb MHC region [25]. Within this 86kb region, HLA-B is
the only known gene. In addition, we have performed a whole genome scan using
Affymetrix GeneChip Human Mapping 100K set in a case-controlled association
study to identify additional markers/susceptibility genes other than HLA-B*1502
that might predispose individuals for CBZ-SJS/TEN. Using 56 cases/54 controls
for CBZ-SJS/TEN, we confirmed the previous observation that the most significant
association was observed in the SNPs of the HLA-B regions on chromosome 6,
Table 14.1 Genetic associations of carbamazepine hypersensitivity

Phenotype HLA association Population References
SJS/TEN HLA-B*1502: 98.3% (59/60) Han Chinese Hung etal.
[P=1.6 1041, OR 1357] in Taiwan [25]
SJS/TEN HLA-B*1502: 100% (4/4) Han Chinese Man etal. [27]
[P=1.48104, OR=71.9] in Hong Kong
SJS/TEN HLA-B*1502: 100% (6/6) Thai population Locharernkul
[P=0.0005, OR=25.5] etal. [28]
SJS/TEN No association with HLA-B*1502 Germany and France; Lonjou etal.
in Whites (0/8); HLA-B*1502: Vietnam, China, [32]
100% in Asian ancestry (4/4) Cambodia and
Reunion Island
SJS/TEN No association with HLA-B*1502 Japanese Kaniwa etal.
(0/7); weak association with [30] and
HLA-B*5901 (1/5) Ikeda etal.
[31]
HSS SNPs in motlin gene in HLA Han Chinese in Taiwan Hung etal. [25]
region [P=0.0046, OR 7.11];
no association with HLA-B*1502
MPE HLA-A*3101 [P=0.0022, OR Han Chinese in Taiwan Hung etal. [25]
17.5]; no association with HLA-
B*1502
MPE No association with HLA-B*1502 Han Chinese in Hong Man etal. [27]
Kong
MPE No association with HLA-B*1502 Thai population Locharernkul
etal. [28]
a
HSS/MPE Significant association with TNF- Caucasians in UK Alfirevic etal.
HLA-DR3-DQ2 haplotype [34]
[P=0.02, OR 3.2]; no
association with HLA-B*1502
SJS/TEN StevensJohnson syndrome/toxic epidermal necrolysis; HSS hypersensitivity syndrome;
MPE maculopapular eruption
a
CBZ-hypersensitivity: only two with blistering skin rashes
and we did not find additional SNPs that were significantly associated with CBZ-
SJS/TEN [26]. The above data suggested that HLA-B*1502 itself is the key genetic
susceptibility to CBZ-SJS/TEN.
14.2.2HLA-B*1502 and CBZ-SJS/TEN

in Asians and Caucasians
The same association between HLA-B*1502 and CBZ-SJS/TEN also have been
reported in independent studies from other Asian countries (Table 14.1). Man
etal. reported that 4 out of 4 (100%) Chinese patients with CBZ-SJS/TEN tested
positive for HLA-B*1502 compared to 14.5% in tolerant controls from Hong
Kong (odds ratio: 71.9, P=1.4810-4) [27]. Another independent study in a Thai
population showed that 6 out of 6 (100%) CBZ-SJS/TEN patients all tested positive
for HLA-B*1502, while only 8 of 50 (16%) of tolerant controls were positive for
the allele (odds ratio: 25.5, P=0.0005) [28]. Mehta etal. reported from India that
6 out of 8 CBZ-SJS patients had HLA-B*1502 while none of the ten controls were
found to be positive (odds ratio: 71.40, P=0.0014) [29]. However, a recent case
series from Japan showed no significant association between HLA-B*1502 and
CBZ-SJS/TEN in Japanese patients [30, 31]. None of the 12 Japanese patients
with CBZ-SJS/TEN carried the HLA-B*1502 allele [30, 31]. Instead, Ikeda etal.
suggested that HLA-B*5901 is one of the candidate markers for CBZ-induced SJS
in the Japanese population [31]. It should be noted that the frequency of the HLA-
B*1502 in the Japanese population is extremely low, almost nonexistent.
The genetic association between CBZ-SJS/TEN and HLA-B*1502 in a given
population seems to correlate with the allele frequency of HLA-B*1502. For
example, the allele frequency of HLA-B*1502 is ~0% in Caucasians, and a case
series including 12 patients with CBZ-SJS/TEN from Europe, including Germany
and France, showed no significant association between HLA-B*1502 and CBZ-
SJS/TEN in Caucasian patients [32, 33]. It is interesting to note that, among the 12
patients tested, 8 were Caucasian and none of these patients carried the allele; the
remaining 4 patients were from Vietnam, China, Cambodia, and Reunion Island
(also people of Chinese descent), and all 4 of these patients tested positive for
HLA-B*1502. This finding indicates that the genetic factor (HLA-B*1502 allele)
is the main risk factor for CBZ-induced SJS/TEN (Table 14.1).
14.2.3HLA, TNF-Alpha Alleles and CBZ-MPE/HSS
Genetic susceptibility of CBZ hypersensitivity seems phenotype-specific; HLA-

B*1502 association is specific for CBZ-SJS/TEN, but not for MPE or HSS.
We found that MPE was associated with HLA-A*3101 (odds ratio: 17.5, cor-
rected P=0.0022) and HSS with SNPs in the motlin gene (odds ratio 7.11, corrected
P = 0.0046) [25]. Case-series studies in Asian countries where strong HLA-
B*1502 association was found in CBZ-SJS/TEN also showed that CBZ-MPE was
not associated with the allele [27, 28]. In a large case-series study of 56 Caucasian
patients from the UK with CBZcADRs, primarily HSS, a significant association
was found between HLA-B*0801 and HLA-DR3, DQ2 and TNF-308 alleles
in patients with CBZ-hypersensitivity [34]. These data suggest that genetic suscepti-
bility to CBZ-hypersensitivity is not only ethnic-specific, but also phenotype-
specific (Table 14.1).
14.2.4HLA-B*1502 and Other AEDs-SJS/TEN
It is interesting to note, since OXC, the prodrug of CBZ, shares structural similarity
with CBZ, that three OXC-SJS cases have been reported to date, and two were
HLA-B*1502 positive [35, 37]. The OXC-SJS case who did not test positive for
HLA-B*1502 had no separation of epidermis and dermis as seen in typical cases of
SJS [36]. In addition, recent studies reported an increased frequency of HLA-
B*1502 in patients with PHT- and LTG-induced SJS/TEN in Hong Kong and
Thailand as compared to controls. Although the number of reported cases was
small, many tested positive for HLA-B*1502, including 1 of 2 LTG-TEN and 1 of
1 PHT-SJS patients with Han Chinese background, and 4 of 4 PHT-SJS Thai
patients (odds radio: 18.5, P=0.005) [27, 28]. Furthermore, we carried out a case-
control association study in Han Chinese residing in Taiwan and enrolled 26 PHT,
6 LTG-induced SJS/TEN patients, and 113 PHT-tolerant, 67 LTG-tolerant subjects,
respectively, who were on the drug for more than 3 months without experiencing
adverse reactions [37]. We found that HLA-B*1502 was present in 8 (30.8%) of 26
PHT-SJS/TEN subjects, while only in 9 (8%) of 113 PHT-tolerant subjects (OR=5.1
(95% CI 1.815.1), P = 0.0041). In comparison, HLA-B*1502 was present in 2
(33%) of 6 LTG-SJS patients and 6 (9%) of 67 LTG-tolerant subjects (OR=5.1
(95% CI, 0.833.8), P=0.1266) [37]. In contrast, LTG-induced SJS/TEN or HSS
in Caucasians was reported to associate with HLA-B*3801 or HLA-B*5801
weakly [33, 38].
14.3Clinical Utility of HLA-B*1502 Testing for Personalized

Medicine of CBZ
14.3.1Pharmacogenetic Test and FDA Recommendation
The value of a pharmacogenetic (PGx) test is related to its cost-effectiveness and

several other factors, including: (1) the incidence and severity of the adverse events,
(2) the sensitivity and specificity of the predictive markers, and (3) whether equally
effective, alternative medications are available for individuals who test positive.
Although the incidence of CBZ-SJS/TEN is low, and data on the pharmaco
economic analysis of cost-benefit for CBZ-SJS/TEN are not yet available because
SJS/TEN carries high mortality and morbidity and many surviving patients have
long-term complications (e.g., ocular damage, renal failure), it is valuable to perform
genotyping of HLA-B*1502 before prescribing CBZ for high-risk Southeast
Asians in whom HLA-B*1502 allele frequency is high in order to avoid these
life-threatening conditions.
Due to strong evidence of the association between HLA-B*1502 allele and
CBZ-SJS/TEN in some Asian countries as described above, the US Food and Drug
Administration (FDA) and similar regulatory agencies in Canada and Taiwan have
relabeled CBZ with genetic information, which notes that patients with ancestries
from areas in which HLA-B*1502 is present should be screened for HLA-B*1502
allele before starting treatment with CBZ [39, 40].
14.3.2Sensitivity, Specificity, and Predictive Values

of the PGx Test
The sensitivity, specificity, and predictive values of a pharmacogenetic test for a

specific genetic marker correlate with the allele frequency and disease prevalence in
the testing population. For example, considering CBZ-SJS/TEN in Han Chinese of
Taiwan, if using a CBZ-tolerant group as the control in a test for SJS/TEN, the HLA-
B*1502 allele would have 100% sensitivity and 97% specificity. Assuming a 0.25%
prevalence rate, the presence of HLA-B*1502 has a 7.7% positive predictive value
for detecting CBZ-SJS/TEN, whereas its absence has a 100% negative predictive
value [41]. The low positive predicted value is because there are HLA-B*1502
carriers who are tolerant to CBZ. Although the positive predictive value of the test
is low (~7.7%), in view of the severity of the diseases and availability of alternative
medicines, CBZ should not be prescribed in HLA-B*1502 carriers unless the
expected benefit clearly outweighs the increased risk of serious skin reactions [42].
14.3.3Alternative Medicines Following the PGx Test
Several alternative AEDs are available that are as equally effective as CBZ, there-
fore it is feasible to withhold CBZ from HLA-B*1502 carriers in favor of using
other AEDs. However, the choice of alternative AEDs should be made with caution
as some of the aromatic AEDs with similar chemical structures may have cross-
reactivity hypersensitivity reactions [43]. In the clinical setting, CBZ, OXC, PHT,
LTG, and PB have a 2030% chance of cross-reactivity of skin rashes [4446].
About 2533% of CBZ hypersensitive patients showed cross-sensitivity of skin
rashes to OXC, and 2570% of OXC-hypersensitivity patients showed cross-reactivity
to CBZ [4446]. Most of the cases showing cross-reactivity hypersensitivity are
primarily reported as mild skin rashes such as MPE. It is suggested that if patients
have ever suffered CBZ hypersensitivity, they are unfavorable for the administra-
tion of aromatic AEDs sharing a similar chemical structure, particularly prohibiting
OXC. As mentioned before, several case studies reported that SJS/TEN caused by
OXC, PHT, or LTG have been reported to associate with HLA-B*1502, therefore,
it is suggested to avoid OXC, PHT, or LTG as alternatives to CBZ therapy in
patients who test positive for the HLA-B*1502 allele.
14.4Case Report
Recently, we have carried out a prospective study in Taiwan aimed at determining

the value of screening HLA-B*1502 before prescribing CBZ to prevent CBZ-SJS/
TEN. In this study, we identified individuals at risk of CBZ-SJS/TEN by using
HLA-B*1502 genotyping and CBZ was not prescribed in those test positive for the
HLA-B*1502. We have enrolled more than 3,000 patients and the preliminary
results suggest that application of HLA-B*1502 genotyping as a screening tool
before patients taking CBZ can effectively reduce the incidence of these life-
threatening adverse drug reactions in our population [47].
14.5Conclusion
Screening for the HLA-B*1502 allele before starting treatment with CBZ in Asian
countries as well as for patients in non-Asian countries who are of Asian descent is
justified in view of the high frequency and seriousness of the consequences of SJS/
TEN, the high sensitivity and specificity of the marker, and the availability of alter-
native AEDs equally effective as CBZ. Similar chemicals, such as OXC, PHT, and
LTG should also be avoided in individuals who test positive for HLA-B*1502.
References
1. Hardman, J. G., Limbird, L. E., & Gilman, A. G. (2001). Goodman and Gilmans pharmaco-
logical basis of therapeutics (10th ed.). New York: McGraw-Hill.
2. Raines, A., & Standaert, F. G. (1969). Effects of anticonvulsant drugs on nerve terminals.
Epilepsia, 10, 211227.
3. Albani, F., Riva, R., & Baruzzi, A. (1995). Carbamazepine clinical pharmacology: A review.
Pharmacopsychiatry, 28, 235244.
4. Browne, T. R., & Holmes, G. L. (2001). Epilepsy. The New England Journal of Medicine, 344,
11451151.
5. Liporace, J., Sperling, M., & Dichter, M. (1994). Absence seizures and carbamazepine in
adults. Epilepsia, 35, 10261028.
6. Mann, J. J. (2005). The medical management of depression. The New England Journal of
Medicine, 353, 18191834.
7. OConnor, A. B., & Dworkin, R. H. (2009). Treatment of neuropathic pain: An overview of
recent guidelines. The American Journal of Medicine, 122(10 Suppl.), S22S32.
8. French, J. A., & Pedley, T. A. (2008). Clinical practice. Initial management of epilepsy. The
9. Brodie, M. J., & Dichter, M. A. (1996). Antiepileptic drugs. The New England Journal of
Medicine, 334, 168175.
10. Ambrosio, A. F., Soares-Da-Silva, P., Carvalho, C. M., & Carvalho, A. P. (2002). Mechanisms
of action of carbamazepine and its derivatives, oxcarbazepine, BIA 2-093, and BIA 2-024.
Neurochemical Research, 27, 121130.
11. Mintzer, S., & Mattson, R. T. (2009). Should enzyme-inducing antiepileptic drugs be consid-
ered first-line agents? Epilepsia, 50(Suppl. 8), 4250.
12. Ambrosio, A. F., Soares-Da-Silva, P., Carvalho, C. M., & Carvalho, A. P. (2002). Mechanisms
of action of carbamazepine and its derivatives, oxcarbazepine, BIA 2-093, and BIA 2-024.
Neurochemical Research, 27, 121130.
13. (2009). Tegretol prescribing information. East Hanover: Novartis Pharmaceuticals.
14. Bertilsson, L. (1978). Clinical pharmacokinetics of carbamazepine. Clinical Pharmacokinetics,
3, 128143.
1 5. Bauer, L. A. (2008). Applied clinical pharmacokinetics (2nd ed.). New York: McGraw-Hill.
16. Fagiolino, P., Vazquez, M., Olano, I., & Delfino, A. (2006). Systemic and presystemic conver-
sion of carbamazepine to carbamazepine-10,11-epoxide during long term treatment. Journal
of Epilepsy and Clinical Neurophysiology, 12, 1316.
17. Pellock, J. M. (1987). Carbamazepine side effects in children and adults. Epilepsia, 28(s3),
S64S70.
18. Porter, R. J. (1987). How to initiate and maintain carbamazepine therapy in children and
adults. Epilepsia, 28(s3), s59s63.
19. Gogtay, N. J., Bavdekar, S. B., & Kshirsagar, N. A. (2005). Anticonvulsant hypersensitivity
syndrome: A review. Expert Opinion on Drug Safety, 4, 571581.
20. Zaccara, G., Franciotta, D., & Perucca, E. (2007). Idiosyncratic adverse reactions to antiepi-
leptic drugs. Epilepsia, 48, 12231244.
21. Roujeau, J. C. (2005). Clinical heterogeneity of drug hypersensitivity. Toxicology, 209,
123129.
22. Arif, H., Buchsbaum, R., Weintraub, D., Koyfman, S., Salas-Humara, C., Bazil, C. W., et al.
(2007). Comparison and predictors of rash associated with 15 antiepileptic drugs. Neurology,
68, 17011709.
23. Roujeau, J. C. (1994). The spectrum of Stevens-Johnson syndrome and toxic epidermal
necrolysis: A clinical classification. The Journal of Investigative Dermatology, 102,
28S30S.
24. Chung, W. H., Hung, S. I., Hong, H. S., Hsih, M. S., Yang, L. C., Ho, H. C., et al. (2004).
Medical genetics: A marker for Stevens-Johnson syndrome. Nature, 428, 486.
25. Hung, S. I., Chung, W. H., Jee, S. H., Chen, W. C., Chang, Y. T., Lee, W. R., et al. (2006).
Genetic susceptibility to carbamazepine-induced cutaneous adverse drug reactions.
26. Hung, S. I., Chung, W. H., Fan, W. H., Ou Yang, C. W., Chen, C. H., Fann, C. S. J., et al.
(2006). Genome-wide scan for genetic markers associated with carbamazepine- and
allopurinol-induced severe cutaneous adverse reactions. In 2nd international drug hypersen-
sitivity conference, Liverpool, April 1821.
27. Man, C. B., Kwan, P., Baum, L., Yu, E., Lau, K. M., Cheng, A. S., et al. (2007). Association
between HLA-B*1502 allele and antiepileptic drug-induced cutaneous reactions in Han
Chinese. Epilepsia, 48, 10151018.
28. Locharernkul, C., Loplumlert, J., Limotai, C., Korkij, W., Desudchit, T., Tongkobpetch, S.,
etal. (2008). Carbamazepine and phenytoin induced Stevens-Johnson syndrome is associated
with HLA-B*1502 allele in Thai population. Epilepsia, 49, 20872091.
29. Mehta, T. Y., Prajapati, L. M., Mittal, B., Joshi, C. G., Sheth, J. J., Patel, D. B., et al. (2009).
Association of HLA-B*1502 allele and carbamazepine-induced Stevens-Johnson syndrome
among Indians. India Journal of Dermatology, Venereology, and Leprology, 75, 579582.
30. Kaniwa, N., Saito, Y., Aihara, M., Matsunaga, K., Tohkin, M., Kurose, K., et al. (2008). JSAR
research group. HLA-B locus in Japanese patients with anti-epileptics and allopurinol-related
Stevens-Johnson syndrome and toxic epidermal necrolysis. Pharmacogenomics, 9,
16171622.
31. Ikeda, H., Takahashi, Y., Yamazaki, E., Fujiwara, T., Kaniwa, N., Saito, Y., et al. (2010). HLA
Class I markers in Japanese patients with carbamazepine-induced cutaneous adverse reac-
tions. Epilepsia, 51, 297300.
32. Lonjou, C., Thomas, L., Borot, N., Ledger, N., de Toma, C., LeLouet, H., et al. (2006).
RegiSCAR Group. A marker for Stevens-Johnson syndrome: Ethnicity matters.
33. Lonjou, C., Borot, N., Sekula, P., Ledger, N., Thomas, L., Halevy, S., et al. (2008). RegiSCAR
study group. A European study of HLA-B in Stevens-Johnson syndrome and toxic epidermal
necrolysis related to five high-risk drugs. Pharmacogenetics and Genomics, 18, 99107.
34. Alfirevic, A., Jorgensen, A. L., Williamson, P. R., Chadwick, D. W., Park, B. K., &
Pirmohamed, M. (2006). HLA-B locus in Caucasian patients with carbamazepine hypersen-
sitivity. Pharmacogenomics, 7, 813818.
35. Chen, Y. C., Chu, C. Y., & Hsiao, C. H. (2009). Oxcarbazepine-induced Stevens-Johnson
syndrome in a patient with HLA-B*1502 genotype. Journal of the European Academy of
Dermatology and Venereology, 23, 702703.
36. Lin, L. C., Lai, P. C., Yang, S. F., & Yang, R. C. (2009). Oxcarbazepine-induced Stevens-Johnson
syndrome: A case report. The Kaohsiung Journal of Medical Sciences, 25, 8286.
37. Hung, S. I., Chung, W. H., Liu, Z. S., Chen, C. H., Hsih, M. S., Hui, R. C., et al. (2010).
Common risk allele in aromatic antiepileptic-drugs induced Stevens-Johnson syndrome and
toxic epidermal necrolysis in Han Chinese. Pharmacogenomics, 11, 349356.
38. Kazeem, G. R., Cox, C., Aponte, J., Messenheimer, J., Brazell, C., Nelsen, A. C., et al. (2009).
High-resolution HLA genotyping and severe cutaneous adverse reactions in lamotrigine-
treated patients. Pharmacogenetics and Genomics, 19, 661665.
39. FDA ALERT: Dangerous or even fatal skin reactions Carbamazepine (marketed as
Carbatrol, Equetro, Tegretol, and generics) Healthcare Professional Sheet text version.
Available at: http://www.fda.gov/Drugs/DrugSafety/Postmarket Drug Safety Information for
Patients and Providers/ucm124718.htm (accessed date: Jan. 25, 2010)
40. Ferrell, P. B. Jr., & McLeod, H. L. (2008). Carbamazepine, HLA-B*1502 and risk of Stevens-
Johnson syndrome and toxic epidermal necrolysis: US FDA recommendations.
41. Hung, S. I., Chung, W. H., & Chen, Y. T. (2005). HLA-B genotyping to detect carbamazepine-
induced Stevens-Johnson syndrome: Implications for personalizing medicine. Personalized
42. Grossman, I. (2007). Routine pharmacogenetic testing in clinical practice: Dream or reality?
43. Romano, A., Gueant-Rodriguez, R. M., Viola, M., Gaeta, F., Caruso, C., & Gueant, J. L.
(2005). Cross-reactivity among drugs: Clinical problems. Toxicology, 209, 169179.
44. Hirsch, L. J., Arif, H., Nahm, E. A., Buchsbaum, R., Resor, S. R., Jr., & Bazil, C. W. (2008).
Cross-sensitivity of skin rashes with antiepileptic drug use. Neurology, 71, 15271534.
45. Alvestad, S., Lydersen, S., & Brodtkorb, E. (2008). Cross-reactivity pattern of rash from cur-
rent aromatic antiepileptic drugs. Epilepsy Research, 80, 194200.
46. Seitz, C. S., Pfeuffer, P., Raith, P., Brocker, E. B., & Trautmann, A. (2006). Anticonvulsant
hypersensitivity syndrome: Cross-reactivity with tricyclic antidepressant agents. Annals of
Allergy, Asthma & Immunology, 97, 698702.
47. Chen, P., Shen, C. Y., Lin, J. J., Ong, C. T., Wu, S. L., Tsai, P. J., et al. (2009). Taiwan SJS
consortium. A prospective study of HLA-B*1502 genotyping in preventing carbamazepine-
induced Stevens-Johnson syndrome and toxic epidermal necrolysis. The 59th annual meeting
of the American society of human genetics, Honolulu, Hawaii.
Part V
Miscellaneous Drugs
Chapter 15
Pharmacogenetics of Flucloxacillin
and Amoxicillin-Clavulanate Associated
Hepatic Dysfunction/Injury
Hong-Kee Lee and Lionel D. Lewis
Keywords Drug-induced liver injury Anti-staphylococcal agent Beta-lactamase

inhibitor
15.1Introduction
Drug-induced liver injury (DILI) may be life-threatening and is one of the most
common reasons that prevent investigational drugs from reaching the market [1] or
newly approved drugs being withdrawn post marketing [2, 3]. DILI is a term that
describes various different types of hepatic damage, which includes cholestasis
(accumulation of bile in the biliary canaliculi and/or cessation of bile flow) and
hepatocellular damage [4]. In cholestatic hepatitis, the biochemical markers alka-
line phosphatase (ALP or Alk Phos) and gamma-glutamyltransferase (GGT) would
be markedly elevated. However, hepatocellular damage causes marked elevation of
aspartate aminotransferase (AST) and alanine aminotransferase (ALT). Jaundice
(hyperbilirubinemia) is usually present in both types of hepatic damage, however,
specific elevation of conjugated bilirubin indicates more of a cholestatic problem.
Despite extensive reporting of DILI in literature, the mechanisms causing the
hepatic injury remains complex and are often incompletely understood [5].
Hepatotoxicity can be predictable and dose-related, or unpredictable and idio-
syncratic. Dose-related hepatotoxicity is usually detected in the preclinical stages
of drug development and these investigational drugs often do not achieve regulatory
approval to be marketed. Idiosyncratic hepatotoxicity usually is not observed until
larger numbers of individuals are exposed to the drug, during the postmarketing
phase [6]. In this chapter, two antibiotics that can cause cholestatic hepatotoxicity
are discussed in detail.
L.D. Lewis(*)
Dartmouth Medical School & Dartmouth-Hitchcock Medical Center, Borwell 322W,
Lebanon, NH 03756, USA
e-mail: lionel.lewis@dartmouth.edu
240 H.-K. Lee and L.D. Lewis
15.2Flucloxacillin (Floxapen)
15.2.1General Pharmacology
Flucloxacillin was first marketed in the 1970s in Europe. It is the anti-staphylococcal

agent of choice in many countries, with the exception of the USA and Canada [7].
Flucloxacillin is a semisynthetic penicillin (Fig. 15.1, I) with a narrow-spectrum of
bactericidal activity. It has considerable activity against Gram-positive organisms
like penicillin sensitive and penicillinase producing Staphylococcus aureus,
Streptococcus pyogenes, and Streptococcus pneumoniae. It is not active against
Gram-negative bacilli, or Streptococcus faecalis and recently methicillin-resistant
Staphylococcus aureus have become resistant to it. Flucloxacillin is available in
parenteral and oral formulations and is well-absorbed from the gastrointestinal tract.
Presence of food in the gastrointestinal tract delays its absorption, resulting in lower
peak serum concentrations. Flucloxacillin is highly bound to serum proteins
(95% bound), and 10% of it is metabolized to flucloxacilloic acid (Fig. 15.1, II) [8].
The elimination half-life of flucloxacillin is 3060min, and about 50% of a dose is
excreted by the kidney within 6h of administration. Indications for flucloxacillin
include pneumonia, osteomyelitis, skin, soft tissue and wound infections, infected
burns, and cellulitis. Flucloxacillin can commonly cause gastrointestinal distur-
bances (nausea, vomiting, diarrhea) and less frequently causes hematological side
effects such as neutropenia and thrombocytopenia. It is contraindicated in patients
with history of hypersensitivity to b-lactam antibiotics and patients with previous
history of flucloxacillin-associated jaundice/hepatic dysfunction.
15.2.2Hepatotoxicity
Hepatotoxicity due to flucloxacillin was first described in a case report in 1982 [9].
Shortly after, a number of additional cases were reported [1012]. To date, 348 and
1,477 reports regarding suspected flucloxacillin-associated hepatotoxicity have
Fig. 15.1 Chemical structures of (I) flucloxacillin, and (II) flucloxacilloic acid
15 Pharmacogenetics of Flucloxacillin and Amoxicillin-Clavulanate 241
been registered in the Swedish Adverse Drug Reaction Database (SWEDIS) and
WHO Adverse Drug Reaction Database (VigiBase), respectively [13]. Classical
flucloxacillin-induced cholestasis usually present as painless jaundice with elevated
bilirubin (mainly conjugated direct bilirubin) and ALP [14]. A delay of several
days to weeks usually occurred between the start of treatment and onset of clinical
symptoms [11], which included nausea, abdominal pain, pruritus, and fever. The
course of the hepatotoxicity was typically protracted, averaging about 11 weeks,
ranging from weeks to months [7]. The risk of cholestasis due to flucloxacillin was
reported by Russman et al. to be 8.5 per 100,000 (1 in 12,000) first-time users
according to a cohort study using the data from the UK General Practice Research
Database [14]. Devereaux et al. reported an incidence rate of 1:15,000 [7].
Susceptibility to flucloxacillin-induced cholestasis was associated with the patients
age and duration of flucloxacillin treatment. Patients over 55 years of age who
received flucloxacillin treatment for over 14 days were at increased risk of developing
cholestasis [14, 15]. The incidence of flucloxacillin-induced cholestasis was also
found to be higher in females compared to males [14, 15].
Hepatotoxicity due to flucloxacillin was not discovered preclinically because
in vitro toxicity studies by Lakehal et al. showed that drug concentrations up to
1mM were not toxic to human hepatocytes or biliary epithelial cells [16]. However,
a minor metabolite of flucloxacillin, 5-hydroxymethylflucloxacillin, formed by
CYP3A4 and initially discovered by Thijssen [17, 18], was found to be toxic to
biliary epithelial cells invitro [16], but the concentration of this minor metabolite
invivo was only 1% of the concentration of the parent drug. Several hypotheses had
been proposed to explain the mechanism of DILI due to flucloxacillin, but no
consistent result had been observed to prove these hypotheses. Carey etal. proposed
that treatment with flucloxacillin could result in the formation of hepatic protein
adducts [19]. A serendipitous discovery that patients who were rechallenged with
flucloxacillin developed symptoms of eosinophilia, inflammatory infiltrates, and
fever suggested the involvement of the immune system in the pathophysiology of
this syndrome [9]. Maria etal. were unable to implicate an immunological basis for
flucloxacillin-associated DILI using lymphocyte proliferation assays [20].
15.2.3Pharmacogenetics
The potential genetic basis for DILI susceptibility was investigated and several
candidate genes such as those for drug transporters (e.g., ABCB4, ABCB11),
familial intrahepatic cholestasis type 1 gene (FIC1) [21], drug metabolizing
enzymes e.g., CYP3A4 and CYP3A5 [4], and human leukocyte antigen (HLA) [22]
were studied. Amongst these studies, the association between HLA and susceptibility
to flucloxacillin related DILI was found to be the strongest.
The UK DILIGEN study was intended to identify genetic determinants of DILI [23].
Genome-wide and candidate gene association studies (GWAS) were performed
on 51 cases with flucloxacillin-associated hepatic dysfunction and 282 controls
matched for sex and ancestry. Thirty-six out of the 51 cases were female and the
average age of onset of flucloxacillin related DILI was 63 years old, consistent with
the earlier studies of risk prediction where patients older than 55 years were more
susceptible to hepatic injury. Most of the cases (86%) were diagnosed as having
flucloxacillin-associated cholestasis. The Illumina Human1M BeadChip which
contained 1,072,820 markers was used to genotype cases and controls. 206,421
markers were discarded due to failure to meet standard quality control criteria. The
GWAS revealed one highly significant signal in the MHC region on chromosome 6
(Fig. 15.2) [23]. This single nucleotide polymorphism (SNP) associated with flu-
cloxacillin-associated DILI was rs2395029, with a P-value of 8.7 1033 (trend
test) and estimated odds ratio (OR) of 45 (95% CI=19.4105). Among the cases,
43 of them (84%) carried the risk allele (T), which has an allele frequency of
approximately 5% in the European population. In comparison, only 11% of the
controls subjects carried this allele. rs2395029 is a missense SNP in the HCP5 gene
which was reported by Colombo et al. to be in complete linkage disequilibrium
with HLA-B*5701 [24]. For the cases in the GWAS, HLA-B*5701 and rs2395029
genotypes correlated perfectly. This meant that HLA-B*5701 showed a highly
significant association with flucloxacillin-associated DILI. Patients possessing this
allele and treated with flucloxacillin were associated with an 80-fold increased risk
of developing DILI. Data from the same study also revealed that the HLA-
DRB1*0701-DQB1*0303 haplotype were significantly more common among
flucloxacillin-related hepatotoxicity compared to controls.
HLA-B*5701 is relatively common in Northern Europe but rarer in Africa and
Asia. This allele was also associated with the development of skin-hypersensitivity
reactions to abacavir (see Chap. 12). The pathophysiology of abacavir-related
skin reactions has been well detailed. An abacavir metabolite interacts with cyto-
toxic CD8+ T cells, with recognition dependent on the presence of the HLA-B*5701
antigen [25]. However, the mechanism associated with flucloxacillin-induced DILI
has not been so fully elucidated. There is no obvious structural similarity between
Fig. 15.2 Flucloxacillin DILI genome-wide association study data [23]. Each dot represents a
SNP. The x axis represents the position of the SNP on chromosomes. The y axis represents
the log10 of Cochran-Armitage trend P value of the SNP in the case-control association study.
The very strong signal in chromosome 6 lies in the regions coding for the MHC genes. Reproduced
with permission from publisher.
flucloxacillin and abacavir to explain why both adverse drug reactions are associated
with HLA-B*5701. Further studies are required to improve our understanding of
these observations.
15.3Amoxicillin-Clavulanate (e.g. Augmentin)
15.3.1General Pharmacology
Amoxicillin-clavulanate was first marketed in UK in 1981, and was FDA approved

in the US in 1984, and granted marketing approval in Australia in 1986 [26]. It is
widely used as an oral and parenteral antibacterial consisting of the combination of
a semisynthetic antibiotic amoxicillin, and the b-lactamase inhibitor, clavulanic acid
(Fig. 15.3) as its potassium salt. Amoxicillin is a 4-hydroxy analog of ampicillin,
derived from the basic penicillin nucleus, 6-aminopenicillanic acid. It has a broad-
spectrum of bactericidal activity against many gram-positive and gram-negative
microorganisms. However, amoxicillin was found to be susceptible to degradation
by b-lactamases. Clavulanic acid is produced by the fermentation of Streptomyces
clavuligerus. It is a b-lactam that is structurally related to the penicillins and
possesses the ability to inactivate a wide variety of b-lactamases by irreversibly
binding to the enzymes active site. Clavulanic acid is particularly active against the
clinically important plasmid-mediated b-lactamases responsible for transferred drug
resistance to penicillins and cephalosporins. The amoxicillin-clavulanic acid combi-
nation protects amoxicillin from degradation by b-lactamase enzymes and extends
the antibiotic spectrum of amoxicillin to include many normally resistant bacteria.
Amoxicillin-clavulanate is well absorbed from the gastrointestinal tract and can
be taken without regard to meals, but absorption of clavulanic acid is increased when
taken with food. Amoxicillin is 18% and clavulanic acid is 25% bound to serum
proteins. Amoxicillin diffuses readily into most body tissues and fluids with the
exception of the brain and spinal fluid. The half-life of amoxicillin is 1.3h and that
of clavulanic acid is 1h. Approximately, 5070% of the amoxicillin and 2540% of
the clavulanic acid are excreted unchanged in urine during the first 6h post dosing.
Concurrent administration of probenecid delays amoxicillin excretion but does not
Fig. 15.3 Chemical structures of (I) amoxicillin, and (II) clavulanic acid
delay renal excretion of clavulanic acid. Indications for amoxicillin-clavulanate

include lower respiratory tract infections, otitis media, sinusitis, skin infections, and
urinary tract infections. Amoxicillin-clavulanate can cause side effects of mild
gastrointestinal disturbances and rarely causes hemolytic anemia and thrombocy-
topenia. It is contraindicated in patients with a history of allergic reactions to any
penicillin and in patients with a previous history of cholestatic jaundice/hepatic
dysfunction associated with the antibiotic.
15.3.2Hepatotoxicity
The first case of DILI associated with amoxicillin-clavulanate was reported in

1988 [27], and several additional case reports soon followed [26, 28, 29]. More
recently, an Italian database of spontaneous reporting of suspected adverse drug
reactions defined that the amoxicillin-clavulanate combination caused a higher
incidence of DILI (4%) compared with amoxicillin alone (1%) [30]. The risk of
DILI associated with amoxicillin-clavulanate was reported to be 1 in 10,000
patients treated with the drug [31, 32]. Similarly to flucloxacillin, the risk of hepatic
injury increased with the patients age and duration of treatment with amoxicillin-
clavulanate [33]. Thomson et al. reported that patients over 55 years old had an
odds ratio of developing DILI due to amoxicillin-clavulanate of 16.1 (95%
CI=2.988.9) compared with patients less than 30 years old [34]. The combination
of advancing age and repeated intake of amoxicillin-clavulanate was reported to
increase the risk of DILI to 1 in 1,000 patients [31]. Men had an increased risk
compared to women in developing amoxicillin-clavulanate related DILI [34, 35].
15.3.3Pharmacogenetics
The cellular mechanism(s) underlying DILI associated with amoxicillin-clavulanate

is poorly understood. Several of the hypotheses concerning the pathophysiology of
this adverse drug reaction involves drug metabolism. Firstly, patients who metabo-
lize the drug differently compared to others or lack adequate protective mechanisms
to neutralize reactive metabolites may develop DILI. Secondly, patients with
immune systems that more readily recognize the neoantigens that were formed
when the drug combinations active metabolites interact with hepatocyte proteins
may also develop DILI [5]. Preliminary studies have shown that amoxicillin-
clavulanate associated hepatotoxicity, which is mainly cholestatic, is linked to
HLA-DRB1*1501 [36, 37], and the proposed immunological mechanism may be
mediated by the HLA class II system. The HLA Class II antigens present peptides
to the T-cell receptor of CD4+ helper T lymphocytes, and this leads to the subse-
quent development of an immune response and caused stimulation of B cells and
cytotoxic T cells [36]. Hautekeete etal. found that there was linkage disequilibrium
between HLA-DRB1*1501 and DRB5*0101, as well as DQB1*0602 [36].

ODonohue etal. independently found that HLA-DQA1*0102 was also in linkage
disequilibrium with the other three alleles [37]. These researchers also reported no
difference between patients who are homozygous, heterozygous, or negative for
HLA-DRB1*1501 and the severity/duration of jaundice, hepatic histology, or
biochemical pattern of hepatic injury. However, the HLA-DRB1*1501 haplotype
conferred susceptibility to primary sclerosing cholangitis [37].
15.4Conclusions
In summary, studies showed that there was an association between flucloxacillin-

associated cholestatic hepatitis and HLA-B*5701, and a similar association was
found between amoxicillin-clavulanate-associated cholestatic hepatitis and HLA-
DRB1*1501. However, prospective screening for the presence of these SNPs will
not be very useful due to the low prevalence of drug-associated cholestatic hepatitis.
The genotyping test may be helpful in establishing diagnosis in jaundice patients
with the suspected diagnosis of drug-induced cholestasis. However, this use of the
genotyping needs to be prospectively studied to confirm such utilization.
References
1. Welch, K. D., Wen, B., Goodlett, D. R., Yi, E. C., Lee, H., Reilly, T.P., etal. (2005). Proteomic
identification of potential susceptibility factors in drug-induced liver disease. Chemical
Research in Toxicology, 18, 924933.
2. Bohan, A. & Boyer, J. L. (2002). Mechanisms of hepatic transport of drugs: Implications for
cholestatic drug reactions. Seminars in Liver Disease, 22, 123136.
3. Larrey, D. & Pageaux, G. P. (2005). Drug-induced acute liver failure. European Journal of
4. Andrews, E. & Daly, A. K. (2008). Flucloxacillin-induced liver injury. Toxicology, 254,
158163.
5. Stricker, B. H. C. (1992). Drug-induced hepatic injury (2nd ed.). Amsterdam: Elsevier
Science Publishers B.V.
6. Kaplowitz, N. (2005). Idiosyncratic drug hepatotoxicity. Nature Reviews. Drug Discovery,
4, 489499.
7. Devereaux, B. M., Crawford, D. H., Purcell, P., Powell, L. W., & Roeser, H. P. (1995).
Flucloxacillin associated cholestatic hepatitis. An Australian and Swedish epidemic?
European Journal of Clinical Pharmacology, 49(12), 8185.
8. Cole, M., Kenig, M. D., & Hewitt, V. A. (1973). Metabolism of penicillins to penicilloic acids
and 6-aminopenicillanic acid in man and its significance in assessing penicillin absorption.
Antimicrobial Agents and Chemotherapy, 3, 463468.
9. Lobatto, S., Dijkmans, B. A., Mattie, H., & Van Hooff, J. P. (1982). Flucloxacillin-associated
liver damage. The Netherlands Journal of Medicine, 25, 4748.
10. Dobson, J. L., Angus, P. W., Jones, R., Crowley, P., & Gow, P. J. (2005). Flucloxacillin-
induced aplastic anaemia and liver failure. Transplant International: Official Journal of the
European Society for Organ Transplantation, 18, 487489.
11. Miros, M., Kerlin, P., Walker, N., & Harris, O. (1990). Flucloxacillin induced delayed
cholestatic hepatitis. Australian and New Zealand Journal of Medicine, 20, 251253.
12. Olsson, R., Wiholm, B. E., Sand, C., Zettergren, L., Hultcrantz, R., & Myrhed, M. (1992).
Liver damage from flucloxacillin, cloxacillin and dicloxacillin. Journal of Hepatology, 15,
154161.
13. Tornhage, C. J., Brunlof, G., & Wallerstedt, S. (2009). Severe hepatotoxic adverse reaction in
a healthy schoolgirl after treatment with flucloxacillin. Drug Healthcare Pat Safety, 1, 1719.
14. Russmann, S., Kaye, J. A., Jick, S. S., & Jick, H. (2005). Risk of cholestatic liver disease
associated with flucloxacillin and flucloxacillin prescribing habits in the UK: Cohort study
using data from the UK General Practice Research Database. British Journal of Clinical
15. Fairley, C. K., McNeil, J. J., Desmond, P., Smallwood, R., Young, H., Forbes, A., etal. (1993).
Risk factors for development of flucloxacillin associated jaundice. British Medical Journal
(Clinical Research Ed.), 306, 233235.
16. Lakehal, F., Dansette, P. M., Becquemont, L., Lasnier, E., Delelo, R., Balladur, P., et al.
(2001). Indirect cytotoxicity of flucloxacillin toward human biliary epithelium via metabolite
formation in hepatocytes. Chemical Research in Toxicology, 14, 694701.
17. Thijssen, H. H. (1979). Identification of the active metabolites of the isoxazolylpenicillins by
means of mass-spectrometry. The Journal of Antibiotics (Tokyo), 32, 10331037.
18 Thijssen, H. H. (1980). Analysis of isoxazolyl penicillins and their metabolites in body
fluids by high-performance liquid chromatography. Journal of Chromatography, 183(3),
339345.
19. Carey, M. A. & van Pelt, F. N. (2005). Immunochemical detection of flucloxacillin adduct
formation in livers of treated rats. Toxicology, 216, 4148.
20. Maria, V. A. & Victorino, R. M. (1997). Diagnostic value of specific T cell reactivity to drugs
in 95 cases of drug induced liver injury. Gut, 41, 534540.
21. Huwyler, J., Wright, M. B., Gutmann, H., & Drewe, J. (2006). Induction of cytochrome P450
3A4 and P-glycoprotein by the isoxazolyl-penicillin antibiotic flucloxacillin. Current Drug
Metabolism, 7, 119126.
22. Berson, A., Freneaux, E., Larrey, D., Lepage, V., Douay, C., Mallet, C., etal. (1994). Possible
role of HLA in hepatotoxicity. An exploratory study in 71 patients with drug-induced idiosyn-
cratic hepatitis. Journal of Hepatology, 20, 336342.
23. Daly, A. K., Donaldson, P. T., Bhatnagar, P., Shen, Y., Peer, I., Floratos, A., et al. (2009).
HLA-B*5701 genotype is a major determinant of drug-induced liver injury due to flucloxacil-
lin. Nature Genetics, 41, 816819.
24. Colombo, S., Rauch, A., Rotger, M., Fellay, J., Martinez, R., Fux, C., etal. (2008). The HCP5
single-nucleotide polymorphism: A simple screening tool for prediction of hypersensitiv-
ity reaction to abacavir. The Journal of Infectious Diseases, 198, 864867.
25. Chessman, D., Kostenko, L., Lethborg, T., Purcell, A. W., Williamson, N. A., Chen, Z., etal.
(2008). Human leukocyte antigen class I-restricted activation of CD8+ T cells provides the
immunogenetic basis of a systemic drug hypersensitivity. Immunity, 28, 822832.
26. Wong, F. S., Ryan, J., Dabkowski, P., Dudley, F. J., Sewell, R. B., & Smallwood, R. A. (1991).
Augmentin-induced jaundice. The Medical Journal of Australia, 154, 698701.
27. Van Den Broek, J. W. G., Buennemeyer, B. L. M., & Stricker, B. H. C. (1988). Cholestatic
hepatitis due to amoxicillin and clavulanic acid. Nederlands tijdschrift voor geneeskunde,
132, 14951497.
28. Hebbard, G. S., Smith, K. G., Gibson, P. R., & Bhathal, P. S. (1992). Augmentin-induced
jaundice with a fatal outcome. The Medical Journal of Australia, 156, 285286.
29. Limauro, D. L., Chan-Tompkins, N. H., Carter, R. W., Brodmerkel, G. J., Jr., & Agrawal, R. M.
(1999). Amoxicillin/clavulanate-associated hepatic failure with progression to Stevens-
Johnson syndrome. The Annals of Pharmacotherapy, 33, 560564.
30. Salvo, F., Polimeni, G., Moretti, U., Conforti, A., Leone, R., Leoni, O., etal. (2007). Adverse
drug reactions related to amoxicillin alone and in association with clavulanic acid: Data from
spontaneous reporting in Italy. The Journal of Antimicrobial Chemotherapy, 60, 121126.
31. Garcia Rodriguez, L. A., Stricker, B. H., & Zimmerman, H. J. (1996). Risk of acute liver
injury associated with the combination of amoxicillin and clavulanic acid. Archives of Internal
Medicine, 156, 13271332.
32. Hussaini, S. H., OBrien, C. S., Despott, E. J., & Dalton, H. R. (2007). Antibiotic therapy:
A major cause of drug-induced jaundice in southwest England. European Journal of
33. Larrey, D., Vial, T., Micaleff, A., Babany, G., Morichau-Beauchant, M., Michel, H.,
etal. (1992). Hepatitis associated with amoxycillin-clavulanic acid combination report of
15 cases. Gut, 33, 368371.
34. Thomson, J. A., Fairley, C. K., Ugoni, A. M., Forbes, A. B., Purcell, P. M., Desmond, P. V.,
etal. (1995). Risk factors for the development of amoxycillin-clavulanic acid associated jaun-
dice. The Medical Journal of Australia, 162, 638640.
35. Thomson, J. A., Fairley, C. K., McNeil, J. J., & Purcell, P. (1994). Augmentin-associated
jaundice. The Medical Journal of Australia, 160, 733734.
36. Hautekeete, M. L., Horsmans, Y., Van Waeyenberge, C., Demanet, C., Hearion, J., Verbist, L.,
etal. (1999). HLA association of amoxicillin-clavulanateinduced hepatitis. Gastroenterology,
117(5), 11811186.
37. ODonohue, J., Oien, K. A., Donaldson, P., Underhill, J., Clare, M., MacSween, R. N., etal.
(2000). Co-amoxiclav jaundice: Clinical and histological features and HLA class II associa-
tion. Gut, 47, 717720.
Chapter 16
Immunosuppressants Pharmacogenomics
Ping Wang
Keywords Immunosuppression Transplantation
This chapter will focus on three immunosuppressants used for posttransplantation

graft maintenance. These include the calcineurin inhibitors (CNIs), cyclosporine
and tacrolimus, and the cell proliferation inhibitor sirolimus. These drugs share
similar mechanism of action, metabolism pathway, and pharmacogenetics/pharma-
cogenomics. Other immunosuppressants will not be discussed here.
16.1Clinical Indication and Mechanism of Action
Cyclosporine, tacrolimus (also named FK506), and sirolimus (also named rapamycin)
are all used posttransplantation to suppress T-cell-mediated immune responses.
They are frequently used in combination with steroids and antiproliferative agents
such as azathioprine or mycophenolate mofetil. Tacrolimus has been shown to be
10100 times more potent than cyclosporine [1, 2]. Both tacrolimus and sirolimus
are used at much lower doses than cyclosporine. Two recent multisite randomized
trials suggest that regimens containing low-dose cyclosporine or low-dose tacroli-
mus are safe and effective, and in the case of tacrolimus, even more advantageous
for renal function, allograft survival, and acute rejection rates in renal transplant
recipients than standard-dose regimens [3, 4]. The target trough levels of 50100ng/mL
cyclosporine and 37ng/mL tacrolimus in these studies are much lower than the
traditionally suggested target trough ranges. See Wallemacq etal. [5] for proposed
target tacrolimus trough concentration guidelines for kidney, heart, and liver
transplantation by the 2007 European Consensus Conference on Tacrolimus
Optimization.
P. Wang(*)
The Methodist Hospital, Houston, TX, USA
e-mail: pwang@tmhs.org
A.H.B.Wu and K.-T.J. Yeo (eds.), Pharmacogenomic Testing in Current Clinical Practice, 249
250 P. Wang
Although chemically distinct, cyclosporine and tacrolimus both function at the

same step in the immune activation cascade by blocking the serine/threonine phos-
phatase activity of calcineurin [6]. The binding of foreign antigens to receptors on
the T cell surface triggers the activation of ras and the increase in the intracellular
calcium concentration. Through the facilitation of calcium/calmodulin, the phos-
phatase activity of calcineurin is activated. Calcineurin then dephosphorylate
NF-ATc, the cytosol subunit of the transcription factor NF-AT. Dephosphorylation
triggers nuclear translocation of NF-ATc, which binds to NF-ATn and leads to
transcription activation of cytokines and T-cell activation. Cyclosporine and tacroli-
mus form complexes with their respective immunophilins, cyclophilin and FKBP12,
which further engage calcium/calmodulin/calcineurin to form a pentamer and
inhibit the phosphatase activity of calcineurin, thereby blocking T-cell activation
and cytokine production.
Sirolimus is structurally related to tacrolimus. It also binds to FKBP12, but
inhibits mTOR (mammalian target of rapamycin), a key regulator of cell growth
and proliferation. The cell cycle of various cell types are arrested in G1 phase as a
result of mTOR inhibition [7]. The mechanism of action of the three drugs is
depicted in Fig. 16.1.
Fig. 16.1 Schematic drawing illustrating the mechanism of action for cyclosporine, tacrolimus,
and sirolimus. CnA and CnB are the catalytic and regulatory subunits of calcineurin, respectively
16 Immunosuppressants Pharmacogenomics 251
16.2Dosing Variability and Toxicities
Great interindividual and interethnic variability has been observed in the dose
requirement of the CNIs and sirolimus [8, 9]. Because of the variability, it is a challenge
in both clinical practice and clinical trials to achieve target concentration [10]. Oral
bioavailability of CNIs in African Americans is 2050% lower than in non-African
Americans. Studies have shown that absorption of orally-administered immunosup-
pressants, rather than hepatic metabolism and clearance, is the main source of vari-
ability [9, 11]. The CNIs are substrates of the cytochrome P450 3A (CYP3A)
enzymes, which are expressed in enterocytes and convert CNIs to their metabolites.
Absorption is also regulated by the drug efflux pump P-glycoprotein, which is
encoded by the MDR1/ABCB1 gene and expressed on the apical membrane of
enterocytes. After absorption, CNIs and metabolites are transported via the portal
vein to the liver, where CYP3As are also expressed and hydroxylation and dem-
ethylation metabolism takes place. Some metabolites are excreted in the bile. The
rest of the CNIs and metabolites enter systemic circulation. All three drugs are
highly lipophilic and distribute primarily into red blood cells (9095%), with minor
fractions appearing in plasma, lymphocytes, and granulocytes. The pharmacokinet-
ics of the CNIs is shown schematically in Fig. 16.2. The pathway for sirolimus is
not as clearly elucidated, but is thought to be similar to that of CNIs.
Fig. 16.2 Schematic drawing illustrating the pharmacokinetics of CNIs

252 P. Wang
The use of these immunosuppressants has resulted in dramatic improvements in

the short-term graft survival rate. However, long-term outcomes did not improve to
the same extent [12], especially in African Americans, who continue to have poor
graft survival rates, shorter graft half-lives, delayed graft function, and increased
risk for acute and chronic rejections. Part of this is due to the chronic graft dysfunc-
tion and immunosuppressants-associated side effects, attributable to the narrow
therapeutic indices of these drugs. The toxicities of cyclosporine and tacrolimus are
well-described, including nephrotoxicity, neurotoxicity, diabetes mellitus, hyper-
tension, hyperlipidemia, and gastrointestinal disturbances. Both cyclosporine and
tacrolimus cause renal damage, contributing to chronic graft dysfunction. Sirolimus
has a different toxicity profile. When used together with cyclosporine or tacrolimus,
sirolimus increases CNI exposure and synergizes with these drugs pharmacody-
namically, resulting in reduction in the dose requirement of CNIs. This benefits
renal function in the long term. However, sirolimus delays the clearance of circulat-
ing low-, intermediate-, and very low-density lipoproteins [13], which leads to
hyperlipidemia being the most concerning side effect. A high rate of death due to
cardiovascular diseases with well-functioning grafts is associated with sirolimus
use [14, 15]. Due to serious drug-related toxicities, especially CNI-related nephro-
toxicity, there has been great interest in minimizing or withdrawing CNI sometime
after transplantation, while maintaining adequate immunosupression and low rejec-
tion rates. Two recent trials, the CAESAR Study and the ELITE-Symphony Study
successfully demonstrated the feasibility and benefits of cyclosporine and tacroli-
mus reduction [3, 4]. However, it is also clear from the studies that complete with-
drawal of CNI is associated with significantly higher incidence of biopsy-proven
acute rejections [3]. For a review of CNI minimization, withdrawal, and avoidance
trials, refer to Srinivas and Meier-Kriesche [16].
16.3Therapeutic Drug Monitoring and Pharmacodynamic

Monitoring
In current clinical practice, dosing of cyclosporine, tacrolimus, and sirolimus is

guided by therapeutic drug monitoring (TDM). Although alternative strategies such
as C2 monitoring have been advocated for cyclosporine [17] and improved clinical
outcomes have been shown using such strategies [18], trough blood concentration (C0)
is still the parameter most commonly measured. The goal is to maintain the whole
blood trough concentration within a predefined therapeutic range, which is depen-
dent on the transplanted organ, the time after transplantation, and the analytical
method used to measure the drug. In reality, the target range also varies among dif-
ferent transplant centers.
The sample of choice for TDM is EDTA-anticoagulated whole blood, and the
drugs need to be extracted before measurement. Commercial immunoassays are
available for the monitoring of cyclosporine, tacrolimus, as well as sirolimus and
are widely used in clinical laboratories. These assays include the microparticle
enzyme immunoassay (MEIA, Abbott Diagnostics, Abbott Park, IL), the fluores-
cence polarization immunoassay (FPIA, Abbott Diagnostics), the enzyme-multiplied
immunoassay technique (EMIT, Dade Behring, Glasgow, DE and Abbott
Diagnostics), the antibody-conjugated magnetic immunoassay (ACMIA, Dade
Behring-Siemens, Deerfield, IL), the cloned enzyme donor immunoassay (CEDIA,
Microgenics, Fremont, CA), and the chemiluminescence microparticle immunoassay
(CMIA, Abbott Diagnostics,). An enzyme-linked immunosorbent assay (ELISA,
the PRO-Trac assay, Diasorin, Stillwater, MN) did not gain wide use. In recent
years, the use of laboratory-developed liquid chromatography-tandem mass spec-
trometry (LC-MS/MS) methods that can simultaneously quantify several immuno-
suppressants in one run gradually increased. The antibodies used in the MEIA
assays cross-react with some metabolites, some of which are biologically active.
Therefore, immunoassays show positive biases compared with methods using
LC-MS. The precision of the most widely used (MEIA) is poor at tacrolimus con-
centrations below 9ng/mL [19]. In view of recent clinical trial results demonstrating
benefits of decreased tacrolimus dose, the 2007 European Consensus Conference
on Tacrolimus Optimization recommended that analytical methods measuring tac-
rolimus should have a limit of quantitation (LOQ) lower than 1 ng/mL [5]. For
cyclosporine, the desired LOQ should be below 20 ng/mL [20]. Lab-developed
assays using LC-MS/MS measure the parent immunosuppressants with high speci-
ficity and can achieve LOQ below the recommended levels. The CMIA assay has
also been reported to conform to the recommendations [2123]. For a comparison
of the performance of current analytical methods for tacrolimus, refer to Wallemacq
etal. [5].
Although monitoring of whole blood trough concentration and adjusting dose
of CNIs and sirolimus according to the trough concentration has been standard
clinical practice for many years, drug level can only be measured after dose is
administered. Moreover, the correlation between drug dose and blood concentra-
tion is poor. It may take several months to achieve a stable dose using the trial-
and-error dose adjustment approach. Methods that can help improve dosing of
these drugs are being sought. These include research assays that measure the
concentration of CNIs in lymphocytes and transplanted tissues. Studies have
shown their correlation with acute rejections. However, it is unlikely that these
assays will make their way into clinical practice for technical and turnaround
time reasons. Pharmacodynamic monitoring has also been attempted by measur-
ing pentamer formation, calcineurin phosphatase activity, IL-2 production, and
T-cell cytometry and function. These assays are useful but are cumbersome to
perform in clinical settings. A method gauging the global immune cell function
has been approved by the Food and Drug Administration (FDA) for clinical use.
This assay uses the increase in ATP concentration in CD4+ T cells as a marker of
lymphocyte activation [24, 25]. However, the results do not correlate well with
tacrolimus concentrations [26], and no data have been published relating the
results to clinical outcomes.
254 P. Wang
16.4Pharmacogenetics and Pharmacogenomics
The pharmacogenetic studies of cyclosporine, tacrolimus, and sirolimus have generally

focused on CYP3A4, CYP3A5/CYP3AP1, and P-glycoprotein. Given that these
proteins are expressed in small intestine, liver, kidney as well as other tissues, the
genotypes of both the donor and the recipient need to be considered when relating
pharmacogenetic results to transplant outcomes. The findings are summarized in
Table 16.1.
The human CYP3A gene subfamily is located on chromosome 7 and consists of
CYP3A43, CYP3A4, CYP3A7, and CYP3A5. CYP3A43 is only expressed at very
low levels in some tissues and CYP3A7 is primarily a fetal enzyme.
The CYP3A4*1B allele is encoded by a single nucleotide polymorphism (SNP)
(A392G) located in the 5 promoter region of the gene. This polymorphism has
been linked to more aggressive forms and advanced stages of prostate cancer, espe-
cially in African Americans [27, 28]. However, its functional significance is still a
matter of debate. The *1B allele is rare in the Asian population [29]. The effect of
the *1B allele on tacrolimus pharmacokinetics was shown in one study in kidney
transplant recipients [30]. The association between CYP3A4*1B and cyclosporine
reported in healthy subjects in one study [31] was not repeated in other studies
involving healthy volunteers or renal transplant recipients [32, 33].
CYP3A5 contributes to at least half of the total hepatic CYP3A enzyme activity
and is expressed abundantly in the small intestine [34]. It is the major enzyme
responsible for the metabolism of CNIs and sirolimus.
The association between CYP3A5/CYP3AP1 and tacrolimus is more definite as
it has been reported in numerous studies. CYP3AP1 is a pseudogene that has a
polymorphism G-44A that is in linkage disequilibrium with the *1/*3 polymor-
phism (A6986G) of CYP3A5. The *3 allele of CYP3A5 causes cryptic splicing of
the mRNA, which generates a prematurely terminated protein product at amino
acid 109 [34]. Most studies to date have confirmed that the *3 allele results in
decreased tacrolimus clearance, decreased dose requirement, and shorter time
needed to achieve target trough concentration. Associations between the *3 allele
and tacrolimus-related toxicities have also been reported (Table 16.1) [30, 3551].
The CYP3A5*6 allele results in deletion of exon 7 and loss of functional CYP3A5.
The CYP3A5 *7 allele has a T insertion between codons 345/346 and also results
in a prematurely terminated protein product. Neither *6 nor *7 has been associated
with tacrolimus dose requirement. Of note, the prevalence of CYP3A5*3, *6, or *7
varies a lot with ethnicity. In Caucasians, 80% of individuals are *3 homozygotes,
and 1 in 500 carry the *6 allele. In contrast, only 30% of African Americans are *3
homozygotes, and 3 in 20 carry the *6 allele [34, 52]. CYP3A5*7 occurs with the
frequency of 1022% among African Americans, but is not found in Caucasians
and Asians [53, 54].
Although cyclosporine is also mainly metabolized by CYP3A5, the association
between CYP3A5 and cyclosporine pharmacokinetics or dose requirement is less
consistent throughout the literature (Table 16.1). One hypothesis is that there is a
Table 16.1 Effects of CYP3A4, CYP3A5, and MDR1 polymorphism on cyclosporine, tacrolimus, and sirolimus
CYP3A4 CYP3A5 MDR1
Cyclosporine Area under the curve (AUC)/dose No association [30, 7679]; dose-adjusted No association [31, 33, 39, 76, 80, 81]; 3435T carriers
ratio was affected by *1B [31]; trough concentrations were higher in *3 had higher oral clearance [82], higher AUC [83];
oral clearance was lower in *1B carriers[39] 3435C carriers had higher dose [84]; C-G-C
carriers [76]; no association haplotype was associated with lower AUCs [29];
[32, 33] carriers of 1236C had lower dose-adjusted peak
concentration and lower AUC [78]; 3435TT is a risk
16 Immunosuppressants Pharmacogenomics
of nephrotoxicity [73]
Tacrolimus *1B carriers required more *3 is associated with decreased tacrolimus No association between C3435T, G2677A/T, and dose
tacrolimus to reach target trough clearance, decreased dose requirement, [30, 38, 39, 4244, 47, 85, 86]; 2677A/T [87, 88]or
concentration [30] and shorter time to achieve target 3435C [35, 87, 88] required higher dose; 2677 SNP
concentration [30, 3551] and lower associated with tacrolimus-associated neurotoxicity
incidence of biopsy-proven [74]; less than three copies of T-129C, C3435T, and
nephrotoxicity in renal recipients [48], G2677A/T polymorphisms were associated with
but higher incidence of nephrotoxicity lower tacrolimus levels [70]
in liver recipients [49]
Sirolimus *1B carriers required higher dose No association [51]; *1 carriers had higher No association [51, 89]
[89] dose [89, 90] and higher clearance [91]
255
256 P. Wang
drug absorption barrier that is saturated by the much higher dose of cyclosporine
compared to tacrolimus or sirolimus [55].
The effect of MDR1 genotypes or haplotypes on immunosuppressant pharma-
cokinetics is also controversial. Three genetic variants, including C1236T in exon
12, G2677A/T in exon 21, and C3435T in exon 26, are studied most extensively in
the literature. These SNPs are linked with each other [56] and one study suggested
that MDR1 haplotype rather than genotypes of individual SNPs was more important
with regard to pharmacogenetics [29]. However, there is controversy regarding
whether these polymorphisms change either function or expression level of MDR1
[50, 5658]. Both positive and negative findings have been reported for association
between MDR1 variants and CNIs (Table 16.1). One confounding factor for the
MDR1 association studies is the prevalence of diarrhea in the transplant population.
Mycophenolate mofetil, which is often given to transplant recipients as an antipro-
liferative, as well as tacrolimus, can cause frequent diarrhea. Chronic diarrhea may
in turn damage the intestinal epithelium and disrupt mucosal gene expression [59].
It has been reported that diarrhea can increase trough tacrolimus levels due to
decreased intestinal P-glycoprotein activity [60, 61]. This confounds the associa-
tion between MDR1 genotype/haplotype and immunosuppressant dose.
The genes involved in the pharmacodynamics of immunosuppressants have not
been studied extensively. One study used a single-strand conformational polymor-
phism assay with limited sensitivity to screen for polymorphic variations around
the tacrolimus-FKBP12 binding sites in a relatively small population of Caucasians
and did not detect any polymorphisms [62]. More recent studies have reported
associations between genetic polymorphisms in the calcineurin gene and schizo-
phrenia [63, 64].
A recent candidate gene association study examined the correlation between
tacrolimus dose and polymorphisms in genes involved in pharmacokinetics and
pharmacodynamics of tacrolimus [65]. Seven hundred and sixty-eight polymor-
phism in fifteen candidate genes were screened, and five CYP3A4 and CYP3A5
genetic variants in linkage disequilibrium emerged as significantly associated with
stable tacrolimus dose. CYP3A5 *3 was identified as the one with the most signifi-
cant correlation, further pinning down the effect of this allele in tacrolimus
pharmacogenomics.
Genetic variants in non-CYP450 genes have also been reported to impact CNI
pharmacokinetics. For example, the C-25385T SNP of the pregnane X receptor
gene has been reported to significantly influence tacrolimus clearance [66]. The
pregnane X receptor is a nuclear receptor that is involved in the up-regulation of
drug-metabolizing enzymes such as CYP450s and drug transporters in response to
hormones and medications.
There are fewer studies focusing on the pharmacogenetics of sirolimus. Some
indicated that CYP3A4*1B or CYP3A5*1 carriers require higher sirolimus dose,
while others found no correlation (Table 16.1). The inconsistency is at least partly due
to the interaction between sirolimus and CNIs, which are sometimes coadministered.
There was no association between sirolimus and MDR1 genotype in these studies.
Large-scale SNP arrays and gene expression arrays have also been used to
p redict CNI-associated toxicities or to diagnose acute rejections and infections.
One study identified eight SNPs that could be used to predict CNI-associated
arrhythmia, ischemic heart disease, and heart failure in renal transplant recipients [67].
Using molecular signatures to diagnose rejections and infections (for example,
[68, 69]) has the potential to avoid invasive biopsies in the future and helps to
elucidate the biological mechanism underlying these processes.
16.5Clinical Utility of Pharmacogenetic Testing
Although CNIs are double-edged swords in transplantation, it is clear from recent

minimization/withdrawal studies that complete avoidance of CNIs is impossible at
this stage. It is therefore important to improve the dosing strategy to maximize
efficacy and minimize toxicities. Clinical pharmacogenetic testing may help to
achieve this goal.
Most studies, except for one, did not find difference in incidence of acute rejec-
tions between carriers and noncarriers of the CYP3A5*3 allele [30, 40, 7072].
Some attribute this to TDM, suggesting that blood concentration monitoring is able
to correct any genetic difference by reaching target trough concentration quickly.
However, significant delays have been observed in reaching target concentration in
CYP3A5 expressers in the first 2 weeks after kidney transplantation [40]. In the same
study, rejection episodes occurred earlier in CYP3A5 expressers. The absence of
any difference in acute rejection rates is likely due to the already low incidence of
acute rejections in the first years after transplantation and the relatively small
sample size in most studies. On the other hand, many reports indicate that donor or
recipient genotypes of CYP3As or MDR1 may predict the risk of drug-related side
effects, such as nephrotoxicity, neurotoxicity, hyperlipidemia, and hypertension
[48, 49, 73, 74]. Therefore, although pharmacogenetics may not be able to further
decrease the already-low incidence of acute rejection, it may help to identify
patients with increased susceptibility for drug-related toxicities. For these patients,
different immunosuppressive regimens or different target concentration may prove
beneficial.
Another potential utility of pharmacogenetic testing is the quick prediction of
the optimal initial and/or stable dose for an individual patient. One study developed
an algorithm to predict the stable dose of tacrolimus based on CYP3A5 genotype,
demographics, and interacting comedications [65]. The idea is to shorten the time
needed to achieve stable dose by having a prediction of what that dose should
be from the very beginning of therapy. At least two prospective randomized trials
are being carried out in Europe to investigate the benefits of dosing patients based
on CYP3A5 genotype. One study randomizes CYP3A5 expressers to either a stan-
dard initial tacrolimus dose or a twofold higher initial dose [55]. The other study
randomizes patients to either receive standard dose or to receive 75% of standard
dose for CYP3A5*1 carriers and 150% of standard dose for noncarriers [75].
258 P. Wang
Results of the latter study showed that a significantly higher proportion of patients
reached target trough concentration 3 days after initiation of tacrolimus in the
pharmacogenetics-guided group. In the same study, CYP3A*1/*1 homozygotes were
more likely to be underdosed, while CYP3A5*3/*3 carriers were more likely to be
overdosed in the nonpharmacogenetic group [75]. It is yet to be seen if clinical
outcomes of these prospective studies are impacted by pharmacogenetics.
It is unlikely that pharmacogenetics of immunosuppressants will replace con-
ventional TDM. The major advantage of pharmacogenetics is the knowledge of
patient genotype before the advent of immunosuppression. Used together with
TDM and possibly pharmacodynamic monitoring, pharmacogenetics may help to
improve long-term transplant outcomes.
16.6Clinical Pharmacogenomics Testing
No diagnostic assay has been approved by FDA for clinical genotyping of CYP3A4,
CYP3A5, or MDR1. Genetic variants, including CYP3A4 *1B, CYP3A5*3 and
MDR1 C1236T, C3435T, and G2677A/T, can be genotyped using methods such as
polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP)
or pyrosequencing. EDTA-anticoagulated whole blood is the preferred specimen,
but epithelial cells from buccal swabs can also be used. Partial automation of testing
is possible by using commercial kits (analyte specific reagents, ASRs) that can be
run on automated analyzers (for example, the INFINITI analyzer, AutoGenomics,
Carlsbad, CA). These assays need to be validated by clinical laboratories according
to Clinical Laboratory Improvement Amendments (CLIA) requirements before
they can be implemented clinically.
Since the major advantage of pharmacogenetics is the knowledge of genotype
before starting therapy, genotyping results should be available to the clinicians
before transplantation takes place. The turnaround time for the laboratory-developed
genotyping methods is usually 24h to several days. However, patients usually get
onto the transplantation waiting list long before surgery, which leaves enough time
for pharmacogenetic testing. An ideal time for such testing would be when the
patient is worked up for pretransplantation HLA typing.
16.7Cases
16.7.1Case 1. A 6-Month Long Trial-and-Error Tacrolimus

Dose Titration
A 34-year-old African American male received a living-donor kidney transplant with

zero mismatch on 8/8/2006 due to end-stage renal disease secondary to dia-betes
mellitus and hypertension. Daclizumab was given as induction therapy. Based on a

standard dosing protocol of 0.025mg/kg tacrolimus every 12h and a body weight of
105kg, he was started on 2.5mg tacrolimus every 12h, coadministered with myco-
phenolate mofetil and prednisone starting from postoperative day 2. He was dis-
charged on 8/12/2006, when his trough tacrolimus level was 12ng/mL. The trough
tacrolimus levels were then measured every time he came to a kidney clinic for
checkup. Tacrolimus dose was adjusted according to the trough levels with a target
concentration of 1015ng/mL in the first 6 months and 810ng/mL in the next 6
months. The trough level and tacrolimus dose are plotted in Fig. 16.3. Due to the
persistently low trough tacrolimus level, tacrolimus dose was elevated for several
times. After the dose was doubled to 5mg twice daily on 2/22/2007, the patients
trough level finally stayed relatively stable within the target range. The patient
was maintained on the dose for the following months and had consistently good
graftfunction. Genotyping for CYP3A5 revealed that the patient was homozygous for
the *1 allele.
Fig. 16.3 Graphs showing tacrolimus trough concentration and dose over time for the patient
in case 1
260 P. Wang
16.7.2Case 2. Tremors and Chronic Allograft Dysfunction

Due to Tacrolimus Overdose
A 20-year-old Caucasian male weighing 167kg received a kidney transplant due to

hypertension nephropathy. He was initiated on tacrolimus, mycophenolate mofetil,
and prednisone. Based on his body weight and age, tacrolimus was initiated at 5mg
twice per day. He was discharged with a trough tacrolimus level of 13 ng/mL.
At his checkup 2 weeks later, the tacrolimus trough level was above 20 ng/mL.
Tacrolimus dose was reduced to 3.5mg every 12h. The patient reported to have
tremors during the follow-up visits, and his serum creatinine concentration as well
as urine protein/creatinine ratio gradually went up. Tacrolimus dose was further
reduced to 2 mg every 12 h. The trough tacrolimus levels were mostly in the
1015ng/mL range, but occasionally went above 15ng/mL. His serum creatinine
levels fluctuated and he continued to have proteinuria and hypertension.
Approximately 1 year after transplantation, kidney biopsy revealed tubulointersti-
tial fibrosis and scarring, tubular atrophy, and vascular sclerosis, without evidence
of acute rejection. The histological changes were consistent with chronic allograft
nephropathy. Genotyping revealed that the CYP3A5 genotype of the patient was
*3/*3. Tacrolimus dose was further reduced to 1mg twice per day. The patients
allograft function gradually improved after the dose reduction.
References
1. Peters, D. H., Fitton, A., Plosker, G. L., & Faulds, D. (1993). Tacrolimus. A review of its
pharmacology, and therapeutic potential in hepatic and renal transplantation. Drugs, 46,
746794.
2. Spencer, C. M., Goa, K. L., & Gillis, J. C. (1997). Tacrolimus. An update of its pharmacology
and clinical efficacy in the management of organ transplantation. Drugs, 54, 925975.
3. Ekberg, H., Grinyo, J., Nashan, B., Vanrenterghem, Y., Vincenti, F., & Voulgari, A., et al.
(2007). Cyclosporine sparing with mycophenolate mofetil, daclizumab and corticosteroids in
renal allograft recipients: the CAESAR Study. American Journal of Transplantation,
7, 560570.
4. Ekberg, H., Tedesco-Silva, H., Demirbas, A., Vitko, S., Nashan, B., & Gurkan, A., et al.
(2007). Reduced exposure to calcineurin inhibitors in renal transplantation. The New England
Journal of Medicine, 20(357), 25622575.
5. Wallemacq, P., Armstrong, V. W., Brunet, M., Haufroid, V., Holt, D. W., & Johnston, A., etal.
(2009). Opportunities to optimize tacrolimus therapy in solid organ transplantation: Report of
the European consensus conference. Therapeutic Drug Monitoring, 31, 139152.
6. Ho, S., Clipstone, N., Timmermann, L., Northrop, J., Graef, I., & Fiorentino, D., etal. (1996).
The mechanism of action of cyclosporin A and FK506. Clinical Immunology and
Immunopathology, 80, S40S45.
7. Sehgal, S. N. (2003). Sirolimus: its discovery, biological properties, and mechanism of action.
Transplantation Proceedings, 35, 7S14S.
8. Dirks, N. L., Huth, B., Yates, C. R., & Meibohm, B. (2004). Pharmacokinetics of immunosup-
pressants: A perspective on ethnic differences. International Journal of Clinical Pharmacology
9. Mancinelli, L. M., Frassetto, L., Floren, L. C., Dressler, D., Carrier, S., & Bekersky, I., etal.
(2001). The pharmacokinetics and metabolic disposition of tacrolimus: A comparison across
ethnic groups. Clinical Pharmacology and Therapeutics, 69, 2431.
10. Ekberg, H., Mamelok, R. D., Pearson, T. C., Vincenti, F., Tedesco-Silva, H., & Daloze, P.
(2009). The challenge of achieving target drug concentrations in clinical trials: experience
from the Symphony study. Transplantation, 87, 13601366.
11. Kolars, J. C., Awni, W. M., Merion, R. M., & Watkins, P. B. (1991). First-pass metabolism of
cyclosporin by the gut. Lancet, 338, 14881490.
12. Meier-Kriesche, H. U., Schold, J. D., Srinivas, T. R., & Kaplan, B. (2004). Lack of improve-
ment in renal allograft survival despite a marked decrease in acute rejection rates over the
most recent era. American Journal of Transplantation, 4, 378383.
13. Hoogeveen, R. C., Ballantyne, C. M., Pownall, H. J., Opekun, A. R., Hachey, D. L., & Jaffe,
J. S., etal. (2001). Effect of sirolimus on the metabolism of apoB100-containing lipoproteins
in renal transplant patients. Transplantation, 72, 12441250.
14. Hricik, D. E. (1994). Posttransplant hyperlipidemia: the treatment dilemma. American
Journal of Kidney Diseases, 23, 766771.
15. Divakar, D., Bailey, R. R., Frampton, C. M., George, P. M., Walmsley, T. A., & Murphy, J.
(1991). Hyperlipidemia in stable renal transplant recipients. Nephron, 59, 423428.
16. Srinivas, T. R., & Meier-Kriesche, H. U. (2008). Minimizing immunosuppression, an alterna-
tive approach to reducing side effects: objectives and interim result. Clinical Journal of
American Society of Nephrology, 3, S101S116.
17. Cole, E., Midtvedt, K., Johnston, A., Pattison, J., & OGrady, C. (2002). Recommendations
for the implementation of Neoral C(2) monitoring in clinical practice. Transplantation,
15(73), S19S22.
18. Levy, G., Burra, P., Cavallari, A., Duvoux, C., Lake, J., & Mayer, A. D., et al. (2002).
Improved clinical outcomes for liver transplant recipients using cyclosporine monitoring
based on 2-hr post-dose levels (C2). Transplantation, 73, 953959.
19. Ghoshal, A. K., & Soldin, S. J. (2002). IMx tacrolimus II assay: Is it reliable at low blood
concentrations? A comparison with tandem MS/MS. Clinical Biochemistry, 35, 389392.
20. de Jonge, H., Naesens, M., & Kuypers, D. R. (2009). New insights into the pharmacokinetics
and pharmacodynamics of the calcineurin inhibitors and mycophenolic acid: Possible conse-
quences for therapeutic drug monitoring in solid organ transplantation. Therapeutic Drug
Monitoring, 31, 416435.
21. Wallemacq, P., Goffinet, J. S., OMorchoe, S., Rosiere, T., Maine, G. T., & Labalette, M.,
etal. (2009). Multi-site analytical evaluation of the Abbott ARCHITECT tacrolimus assay.
Therapeutic Drug Monitoring, 31, 198204.
22. De, B. K., Jimenez, E., De, S., Sawyer, J. C., & McMillin, G. A. (2009). Analytical perfor-
mance characteristics of the Abbott Architect i2000 Tacrolimus assay; comparisons with liq-
uid chromatography-tandem mass spectrometry (LC-MS/MS) and Abbott IMx methods.
Clinica Chimica Acta, 410, 2531.
23. Amann, S., Parker, T. S., & Levine, D. M. (2009). Evaluation of 2 immunoassays for monitor-
ing low blood levels of tacrolimus. Therapeutic Drug Monitoring, 31, 273276.
24. Kowalski, R. J., Zeevi, A., Mannon, R. B., Britz, J. A., & Carruth, L. M. (2007).
Immunodiagnostics: evaluation of functional T-cell immunocompetence in whole blood inde-
pendent of circulating cell numbers. Journal of Immunotoxicology, 4, 225232.
25. Bhorade, S. M., Janata, K., Vigneswaran, W. T., Alex, C. G., & Garrity, E. R. (2008). Cylex
ImmuKnow assay levels are lower in lung transplant recipients with infection. The Journal of
Heart and Lung Transplantation, 27, 990994.
26. Kowalski, R., Post, D., Schneider, M. C., Britz, J., Thomas, J., & Deierhoi, M., etal. (2003).
Immune cell function testing: an adjunct to therapeutic drug monitoring in transplant patient
management. Clinical Transplantation, 17, 7788.
27. Keshava, C., McCanlies, E. C., & Weston, A. (2004). CYP3A4 polymorphismspotential risk
factors for breast and prostate cancer: A HuGE review. American Journal of Epidemiology,
160, 825841.
262 P. Wang
28. Bangsi, D., Zhou, J., Sun, Y., Patel, N. P., Darga, L. L., & Heilbrun, L. K., etal. (2006). Impact
of a genetic variant in CYP3A4 on risk and clinical presentation of prostate cancer among
white and African-American men. Urologic Oncology, 24, 2127.
29. Chowbay, B., Cumaraswamy, S., Cheung, Y. B., Zhou, Q., & Lee, E. J. (2003). Genetic poly-
morphisms in MDR1 and CYP3A4 genes in Asians and the influence of MDR1 haplotypes
on cyclosporin disposition in heart transplant recipients. Pharmacogenetics, 13, 8995.
30. Hesselink, D. A., van Schaik, R. H., van der Heiden, I. P., van der Werf, M., Gregoor, P. J., &
Lindemans, J., etal. (2003). Genetic polymorphisms of the CYP3A4, CYP3A5, and MDR-1
genes and pharmacokinetics of the calcineurin inhibitors cyclosporine and tacrolimus.
31. Min, D. I., & Ellingrod, V. L. (2003). Association of the CYP3A4*1B 5-flanking region
polymorphism with cyclosporine pharmacokinetics in healthy subjects. Therapeutic Drug
Monitoring, 25, 305309.
32. Rivory, L. P., Qin, H., Clarke, S. J., Eris, J., Duggin, G., & Ray, E., etal. (2000). Frequency
of cytochrome P450 3A4 variant genotype in transplant population and lack of association
with cyclosporin clearance. European Journal of Clinical Pharmacology, 56, 395398.
33. von Ahsen, N., Richter, M., Grupp, C., Ringe, B., Oellerich, M., & Armstrong, V. W. (2001).
No influence of the MDR-1 C3435T polymorphism or a CYP3A4 promoter polymorphism
(CYP3A4-V allele) on dose-adjusted cyclosporin A trough concentrations or rejection inci-
dence in stable renal transplant recipients. Clinical Chemistry, 47, 10481052.
34. Kuehl, P., Zhang, J., Lin, Y., Lamba, J., Assem, M., & Schuetz, J., etal. (2001). Sequence
diversity in CYP3A promoters and characterization of the genetic basis of polymorphic
CYP3A5 expression. Nature Genetics, 27, 383391.
35. Macphee, I. A., Fredericks, S., Tai, T., Syrris, P., Carter, N. D., & Johnston, A., et al. (2002).
Tacrolimus pharmacogenetics: polymorphisms associated with expression of cytochrome p4503A5
and P-glycoprotein correlate with dose requirement. Transplantation, 74(11), 14861489.
36. Thervet, E., Anglicheau, D., King, B., Schlageter, M. H., Cassinat, B., & Beaune, P., etal.
(2003). Impact of cytochrome p450 3A5 genetic polymorphism on tacrolimus doses and
concentration-to-dose ratio in renal transplant recipients. Transplantation, 76, 12331235.
37. Macphee, I. A., Fredericks, S., Mohamed, M., Moreton, M., Carter, N. D., & Johnston, A., etal.
(2005). Tacrolimus pharmacogenetics: The CYP3A5*1 allele predicts low dose-normalized
tacrolimus blood concentrations in whites and South Asians. Transplantation, 79, 499502.
38. Tada, H., Tsuchiya, N., Satoh, S., Kagaya, H., Li, Z., & Sato, K., et al. (2005). Impact of
CYP3A5 and MDR1(ABCB1) C3435T polymorphisms on the pharmacokinetics of tacroli-
mus in renal transplant recipients. Transplantation Proceedings, 37, 17301732.
39. Haufroid, V., Mourad, M., Van Kerckhove, V., Wawrzyniak, J., De Meyer, M., & Eddour, D. C.,
etal. (2004). The effect of CYP3A5 and MDR1 (ABCB1) polymorphisms on cyclosporine
and tacrolimus dose requirements and trough blood levels in stable renal transplant patients.
40. MacPhee, I. A., Fredericks, S., Tai, T., Syrris, P., Carter, N. D., & Johnston, A., etal. (2004).
The influence of pharmacogenetics on the time to achieve target tacrolimus concentrations
after kidney transplantation. American Journal of Transplantation, 4, 914919.
41. Dai, Y., Hebert, M. F., Isoherranen, N., Davis, C. L., Marsh, C., & Shen, D. D., etal. (2006).
Effect of CYP3A5 polymorphism on tacrolimus metabolic clearance in vitro. Drug
Metabolism and Disposition, 34, 836847.
42. Tsuchiya, N., Satoh, S., Tada, H., Li, Z., Ohyama, C., & Sato, K., etal. (2004). Influence of
CYP3A5 and MDR1 (ABCB1) polymorphisms on the pharmacokinetics of tacrolimus in
renal transplant recipients. Transplantation, 78, 11821187.
43. Zhang, X., Liu, Z. H., Zheng, J. M., Chen, Z. H., Tang, Z., & Chen, J. S., et al. (2005).
Influence of CYP3A5 and MDR1 polymorphisms on tacrolimus concentration in the early
stage after renal transplantation. Clinical Transplantation, 19, 638643.
44. Goto, M., Masuda, S., Kiuchi, T., Ogura, Y., Oike, F., & Okuda, M., etal. (2004). CYP3A5*1-
carrying graft liver reduces the concentration/oral dose ratio of tacrolimus in recipients of
living-donor liver transplantation. Pharmacogenetics, 14, 471478.
45. Barrera-Pulido, L., Aguilera-Garcia, I., Docobo-Perez, F., Alamo-Martinez, J. M., Pareja-Ciuro, F.,
& Nunez-Roldan, A., etal. (2008). Clinical relevance and prevalence of polymorphisms in
CYP3A5 and MDR1 genes that encode tacrolimus biotransformation enzymes in liver trans-
plant recipients. Transplantation Proceedings, 40, 29492951.
46. Satoh, S., Kagaya, H., Saito, M., Inoue, T., Miura, M., & Inoue, K., et al. (2008). Lack of
tacrolimus circadian pharmacokinetics and CYP3A5 pharmacogenetics in the early and main-
tenance stages in Japanese renal transplant recipients. British Journal of Clinical Pharmacology,
66, 207214.
47. Op den Buijsch, R. A., Christiaans, M. H., Stolk, L. M., de Vries, J. E., Cheung, C. Y., &
Undre, N. A., etal. (2007). Tacrolimus pharmacokinetics and pharmacogenetics: Influence of
adenosine triphosphate-binding cassette B1 (ABCB1) and cytochrome (CYP) 3A polymor-
phisms. Fundamental & Clinical Pharmacology, 21, 427435.
48. Kuypers, D. R., de Jonge, H., Naesens, M., Lerut, E., Verbeke, K., & Vanrenterghem, Y.
(2007). CYP3A5 and CYP3A4 but not MDR1 single-nucleotide polymorphisms determine
long-term tacrolimus disposition and drug-related nephrotoxicity in renal recipients. Clinical
49. Fukudo, M., Yano, I., Yoshimura, A., Masuda, S., Uesugi, M., & Hosohata, K., etal. (2008).
Impact of MDR1 and CYP3A5 on the oral clearance of tacrolimus and tacrolimus-related
renal dysfunction in adult living-donor liver transplant patients. Pharmacogenetics &
Genomics, 18, 413423.
50. Kim, R. B., Leake, B. F., Choo, E. F., Dresser, G. K., Kubba, S. V., & Schwarz, U. I., etal.
(2001). Identification of functionally variant MDR1 alleles among European Americans and
African Americans. Clinical Pharmacology and Therapeutics, 70, 189199.
51. Mourad, M., Mourad, G., Wallemacq, P., Garrigue, V., Van Bellingen, C., & Van Kerckhove, V.,
et al. (2005). Sirolimus and tacrolimus trough concentrations and dose requirements after
kidney transplantation in relation to CYP3A5 and MDR1 polymorphisms and steroids.
Transplantation, 80, 977984.
52. Hesselink, D. A., van Gelder, T., & van Schaik, R. H. (2005). The pharmacogenetics of cal-
cineurin inhibitors: one step closer toward individualized immunosuppression?
53. Hustert, E., Haberl, M., Burk, O., Wolbold, R., He, Y. Q., & Klein, K., et al. (2001). The
genetic determinants of the CYP3A5 polymorphism. Pharmacogenetics, 11, 773779.
54. Roy, J. N., Lajoie, J., Zijenah, L. S., Barama, A., Poirier, C., & Ward, B. J., et al. (2005).
CYP3A5 genetic polymorphisms in different ethnic populations. Drug Metabolism and
Disposition, 33, 884887.
55. MacPhee, I. A., & Holt, D. W. (2008). A pharmacogenetic strategy for immunosuppression
based on the CYP3A5 genotype. Transplantation, 85, 163165.
56. Tanabe, M., Ieiri, I., Nagata, N., Inoue, K., Ito, S., & Kanamori, Y., etal. (2001). Expression of
P-glycoprotein in human placenta: relation to genetic polymorphism of the multidrug resistance
(MDR)-1 gene. The Journal of Pharmacology and Experimental Therapeutics, 297, 11371143.
57. Hoffmeyer, S., Burk, O., von Richter, O., Arnold, H. P., Brockmoller, J., & Johne, A., etal. (2000).
Functional polymorphisms of the human multidrug-resistance gene: multiple sequence variations
and correlation of one allele with P-glycoprotein expression and activity invivo. Proceedings of
the National Academy of Sciences of the United States of America, 28(97), 34733478.
58. Oselin, K., Gerloff, T., Mrozikiewicz, P. M., Pahkla, R., & Roots, I. (2003). MDR1 polymor-
phisms G2677T in exon 21 and C3435T in exon 26 fail to affect rhodamine 123 efflux in
peripheral blood lymphocytes. Fundamental & Clinical Pharmacology, 17, 463469.
59. Ziegler, T. R., Fernandez-Estivariz, C., Gu, L. H., Fried, M. W., & Leader, L. M. (2003).
Severe villus atrophy and chronic malabsorption induced by azathioprine. Gastroenterology,
124, 19501957.
60. Maes, B. D., Lemahieu, W., Kuypers, D., Evenepoel, P., Coosemans, W., & Pirenne, J., etal.
(2002). Differential effect of diarrhea on FK506 versus cyclosporine A trough levels and
resultant prevention of allograft rejection in renal transplant recipients. American Journal of
Transplantation, 2, 989992.
264 P. Wang
61. Lemahieu, W., Maes, B., Verbeke, K., Rutgeerts, P., Geboes, K., & Vanrenterghem, Y. (2005).
Cytochrome P450 3A4 and P-glycoprotein activity and assimilation of tacrolimus in trans-
plant patients with persistent diarrhea. American Journal of Transplantation, 5, 13831391.
62. Brogan, I. J., Pravica, V., & Hutchinson, I. V. (2000). Genetic conservation of the immunophilin-
binding domains of human calcineurin A1 and A2. Transplant Immunology, 8(2), 139141.
63. Gerber, D. J., Hall, D., Miyakawa, T., Demars, S., Gogos, J. A., & Karayiorgou, M., etal.
(2003). Evidence for association of schizophrenia with genetic variation in the 8p21.3 gene,
PPP3CC, encoding the calcineurin gamma subunit. Proceedings of the National Academy of
Sciences of the United States of America, 100, 89938998.
64. Liu, Y. L., Fann, C. S., Liu, C. M., Chang, C. C., Yang, W. C., & Hung, S. I., etal. (2007).
More evidence supports the association of PPP3CC with schizophrenia. Molecular Psychiatry,
12, 966974.
65. Wang, P., Mao, Y., Razo, J., Zhou, X., Wong, S. T. C., & Patel, S., et al. (2010). Using genetic
and clinical factors to predict tacrolimus dose in renal transplant recipients. Pharmaco
genomics, 11(10), 13891402.
66. Benkali, K., Marquet, P., Rerolle, J., Le Meur, Y., & Gastinel, L. (2008). A new strategy for
faster urinary biomarkers identification by Nano-LC-MALDI-TOF/TOF mass spectrometry.
BMC Genetics, 9, 541.
67. Mushiroda, T., Saito, S., Tanaka, Y., Takasaki, J., Kamatani, N., & Beck, Y., et al. (2005).
A model of prediction system for adverse cardiovascular reactions by calcineurin inhibitors
among patients with renal transplants using gene-based single-nucleotide polymorphisms.
Journal of Human Genetics, 50, 442447.
68. Saint-Mezard, P., Berthier, C. C., Zhang, H., Hertig, A., Kaiser, S., & Schumacher, M., etal.
(2009). Analysis of independent microarray datasets of renal biopsies identifies a robust
transcript signature of acute allograft rejection. Transplant International, 22, 293302.
69. Inkinen, K., Lahesmaa, R., Brandt, A., Katajamaa, M., Halme, L., & Hockerstedt, K., etal.
(2005). DNA microarray-based gene expression profiles of cytomegalovirus infection and
acute rejection in liver transplants. Transplantation Proceedings, 37, 12271229.
70. Roy, J. N., Barama, A., Poirier, C., Vinet, B., & Roger, M. (2006). Cyp3A4, Cyp3A5, and
MDR-1 genetic influences on tacrolimus pharmacokinetics in renal transplant recipients.
Pharmacogenetics & Genomics, 16, 659665.
71. Hesselink, D. A., van Schaik, R. H., van Agteren, M., de Fijter, J. W., Hartmann, A., & Zeier, M.,
et al. (2008). CYP3A5 genotype is not associated with a higher risk of acute rejection in
tacrolimus-treated renal transplant recipients. Pharmacogenetics & Genomics, 18, 339348.
72. Quteineh, L., Verstuyft, C., Furlan, V., Durrbach, A., Letierce, A., & Ferlicot, S., etal. (2008).
Influence of CYP3A5 genetic polymorphism on tacrolimus daily dose requirements and acute
rejection in renal graft recipients. Basic & Clinical Pharmacology & Toxicology, 103, 546552.
73. Hauser, I. A., Schaeffeler, E., Gauer, S., Scheuermann, E. H., Wegner, B., & Gossmann, J.,
etal. (2005). ABCB1 genotype of the donor but not of the recipient is a major risk factor for
cyclosporine-related nephrotoxicity after renal transplantation. Journal of the American
Society of Nephrology, 16, 15011511.
74. Yamauchi, A., Ieiri, I., Kataoka, Y., Tanabe, M., Nishizaki, T., & Oishi, R., et al. (2002).
Neurotoxicity induced by tacrolimus after liver transplantation: relation to genetic polymor-
phisms of the ABCB1 (MDR1) gene. Transplantation, 27(74), 571572.
75. Thervet, E., Loriot, M.A., Barbier, S., Buchler, M., Ficheux, M., & Choukroun, G., et al.
(2010). Optimization of initial tacrolimus dose using pharmacogenetic testing. Clinical
Pharmacology and Therapeutics, 87(6), 721726.
76. Hesselink, D. A., van Gelder, T., van Schaik, R. H., Balk, A. H., van der Heiden, I. P., &
van Dam, T., etal. (2004). Population pharmacokinetics of cyclosporine in kidney and heart
transplant recipients and the influence of ethnicity and genetic polymorphisms in the MDR-1,
CYP3A4, and CYP3A5 genes. Clinical Pharmacology and Therapeutics, 76, 545556.
77. Zhao, Y., Song, M., Guan, D., Bi, S., Meng, J., & Li, Q., etal. (2005). Genetic polymorphisms
of CYP3A5 genes and concentration of the cyclosporine and tacrolimus. Transplantation
Proceedings, 37, 178181.
78. Anglicheau, D., Thervet, E., Etienne, I., Hurault De Ligny, B., Le Meur, Y., & Touchard, G.,
etal. (2004). CYP3A5 and MDR1 genetic polymorphisms and cyclosporine pharmacokinetics
after renal transplantation. Clinical Pharmacology and Therapeutics, 75, 422433.
79. Kreutz, R., Zurcher, H., Kain, S., Martus, P., Offermann, G., & Beige, J. (2004). The effect of
variable CYP3A5 expression on cyclosporine dosing, blood pressure and long-term graft
survival in renal transplant patients. Pharmacogenetics, 14, 665671.
80. Balram, C., Sharma, A., Sivathasan, C., & Lee, E. J. (2003). Frequency of C3435T single
nucleotide MDR1 genetic polymorphism in an Asian population: phenotypic-genotypic cor-
relates. British Journal of Clinical Pharmacology, 56, 7883.
81. Singh, D., Alexander, J., Owen, A., Rustom, R., Bone, M., & Hammad, A., et al. (2004).
Whole-blood cultures from renal-transplant patients stimulated ex vivo show that the effects
of cyclosporine on lymphocyte proliferation are related to P-glycoprotein expression.
Transplantation, 27(77), 557561.
82. Yates, C. R., Zhang, W., Song, P., Li, S., Gaber, A. O., & Kotb, M., etal. (2003). The effect
of CYP3A5 and MDR1 polymorphic expression on cyclosporine oral disposition in renal
transplant patients. Journal of Clinical Pharmacology, 43, 555564.
83. Min, D. I., & Ellingrod, V. L. (2002). C3435T mutation in exon 26 of the human MDR1 gene
and cyclosporine pharmacokinetics in healthy subjects. Therapeutic Drug Monitoring, 24,
400404.
84. Bonhomme-Faivre, L., Devocelle, A., Saliba, F., Chatled, S., Maccario, J., & Farinotti, R.,
etal. (2004). MDR-1 C3435T polymorphism influences cyclosporine a dose requirement in
liver-transplant recipients. Transplantation, 78, 2125.
85. Goto, M., Masuda, S., Saito, H., Uemoto, S., Kiuchi, T., & Tanaka, K., etal. (2002). C3435T
polymorphism in the MDR1 gene affects the enterocyte expression level of CYP3A4 rather
than Pgp in recipients of living-donor liver transplantation. Pharmacogenetics, 12, 451457.
86. Mai, I., Perloff, E. S., Bauer, S., Goldammer, M., Johne, A., & Filler, G., etal. (2004). MDR1
haplotypes derived from exons 21 and 26 do not affect the steady-state pharmacokinetics of
tacrolimus in renal transplant patients. British Journal of Clinical Pharmacology, 58,
548553.
87. Zheng, H., Webber, S., Zeevi, A., Schuetz, E., Zhang, J., & Bowman, P., et al. (2003).
Tacrolimus dosing in pediatric heart transplant patients is related to CYP3A5 and MDR1 gene
polymorphisms. American Journal of Transplantation, 3, 477483.
88. Anglicheau, D., Verstuyft, C., Laurent-Puig, P., Becquemont, L., Schlageter, M. H., &
Cassinat, B., etal. (2003). Association of the multidrug resistance-1 gene single-nucleotide
polymorphisms with the tacrolimus dose requirements in renal transplant recipients. Journal
of the American Society of Nephrology, 14, 18891896.
89. Anglicheau, D., Le Corre, D., Lechaton, S., Laurent-Puig, P., Kreis, H., & Beaune, P., etal.
(2005). Consequences of genetic polymorphisms for sirolimus requirements after renal trans-
plant in patients on primary sirolimus therapy. American Journal of Transplantation, 5,
595603.
90. Le Meur, Y., Djebli, N., Szelag, J. C., Hoizey, G., Toupance, O., & Rerolle, J. P., etal. (2006).
CYP3A5*3 influences sirolimus oral clearance in denovo and stable renal transplant recipients.
91. Djebli, N., Rousseau, A., Hoizey, G., Rerolle, J. P., Toupance, O., & Le Meur, Y., etal. (2006).
Sirolimus population pharmacokinetic/pharmacogenetic analysis and Bayesian modelling in
kidney transplant recipients. Clinical Pharmacokinetics, 45, 11351148.
Index
A uric acid metabolism, 213

Abacavir uricosuric drugs, 215
false positive clinical diagnosis, 203 Amoxicillin-Clavulanate
HLA-B*5701 allele hepatotoxicity, 244
laboratory test, 209 pharmacology, 243244
negative predictive value, 203 Analyte specific reagent (ASR), 21
immunopathogenesis, hypersensitivity, 206 Angiotensin converting enzyme
nevirapine, 209 (ACE), 43
patch testing, 203 Antiepileptic drug (AED), 225
PCR-based technique, 209 Antiplatelet therapy, 140, 147
pharmacology and clinical use, 201202 Aspartate aminotransferase (AST), 239
PREDICT-1 study, 206 ASPE. See Allele-specific primer extension
rechallenge, 210 Aspirin. See Salicylates
side effects, 202203 ASR. See Analyte specific reagent
transplantation, 207 Atherosclerosis risk in communities
ACE. See Angiotensin converting enzyme (ARIC), 181
Acute coronary syndrome (ACS), Atorvastatin, 157
140, 190191 Augmentin. See Amoxicillin-Clavulanate
Acute lymphoblastic leukemia (ALL), 91 Azathioprine, 92
Adenosine diphosphate (ADP), 139
ADJUNCT biomarker, 53
Adverse drug reaction (ADR), 3, 48 B
AED. See Antiepileptic drug Bare metal stent (BMS), 192
Affymetrix technology, 18, 19 Beckman microarray, 33
Alanine aminotransferase (ALT), 239 b-lactamase inhibitor, 243
Alcoholism, 51
ALL. See Acute lymphoblastic leukemia
Allele-specific primer extension (ASPE) C
automated microarray, 20, 21 Calcineurin inhibitors (CNIs), 249
INFINITI assay, 21 Carbamazepine (CBZ)
Luminex assay, 21 AED medications, 226227
Allopurinol clinical
clinical testing, 219221 indications, 225226
gout, 214 utility, 232233
idiosyncratic reaction, 217 hypersensitivity
induced SCAR, 217219, 221 Asians and Caucasians trial, 230231
marker, 221 genetic susceptibility, 231
reduction mechanism, 213, 214 Han Chinese trial study, 229230
toxicity, 216 prodrug, 231
Molecular and Translational Medicine, DOI 10.1007/978-1-60761-283-4,
268 Index
Carbamazepine (CBZ) (cont.) Estrogen receptor (ER), 77

pharmacokinetics and pharmacodynamics, Evidence based PJ, 49
227228
screening value, 233
seizures, 249, 250 F
sensitivity and specificity, 234 FDA. See Food and Drug Administration
side effects and toxicity, 228229 Floxapen. See Flucloxacillin
Cardiovascular disease, 139 Flucloxacillin
Cardiovascular Health Study (CHS), 181 hepatotoxicity
Center for Medicaid and Medicare Services cholestasis, 241
(CMS), 6 hypothesis, 241
Cholestasis, 241 pharmacogenetics
Clinical Laboratory Improvement Act GWAS, 241
(CLIA), 8 HLA-B*5701, 242
Clopidogrel pharmacology
clinical efficacy, 140 antistaphylococcal agent, 240
mechanism, 139 chemical structure, 240
metabolism, 141 Fluorescence resonance energy transfer
pharmacogenetic testing, 144 (FRET), 23
polymorphism, 142 Food and Drug Administration (FDA), 253
therapeutic implications, 143144
therapy, dosage, 140141
CMS. See Center for Medicaid and Medicare G
Services Gamma-glutamyl carboxylase (GGC), 118
CNIs. See Calcineurin inhibitors Genome-Wide Association Studies (GWAS),
Coronary heart disease (CHD), 176 125126, 142
Creatine kinase (CK), 157 Gilberts syndrome, 63
CYP2D6 gene
activity, 53
genotyping, 83 H
inhibitor, 8182 Helicobacter pylori, 43
polymorphism, 8081 HLA-B*5701 allele
Cytochrome P4502C9, 120 quality assurance, 209
screening test, 203
translation, clinical use, 209
D Hydroxymethyl glutaryl coenzyme A reductase
Daubert standard, 50 (HMGCoA reductase), 156
Drug-eluting stent (DES), 140 Hypersensitivity syndrome (HSS), 215
Drug-induced liver injury (DILI) Hypoxanthine phosphoribosyl transferase
amoxicillin-clavulanate (HPRT), 92
hepatotoxicity, 244
pharmacology, 243244
flucloxacillin I
hepatotoxicity, 240241 ICER. See Incremental cost-effectiveness ratio
pharmacogenetics, 241243 IFCs. See Integrated fluidic circuits
pharmacology, 240 Immunosuppressant pharmacogenomics
genotyping test, 245 clinical
Drug reaction with eosinophilia and testing, 258
systemic symptoms utility, 257258
(DRESS), 228 cyclosporine and tacrolimus, 249
CYP3A4, CYP3A5, and MDR1 effects, 254
mechanism, 250
E polymorphism assay, 256
Epilepsy, 225 sirolimus, 250
eSensor technology, 1820 SNP array, 257
Index 269
therapeutic drug monitoring, 252253 Lipid Research Clinic-Coronary

variability and toxicity Primary Prevention Trial
CNI, 251 (LRC-CPPT), 175
transplantation, 252 Low-density lipoprotein cholesterol (LDLc)
Incremental cost-effectiveness ratio APOE, 157
(ICER), 38 CLMN, 159
INFINITITM and Luminex xTAG. See simvastatin, 160
Allele-specific primer extension Lymphocyte toxicity assay (LTA), 51
(ASPE)
Inflammatory bowel disease, 91
Inosine triphosphate pyrophophatase M
(ITPA), 101 Maculopapular eruption (MPE), 228
Integrated fluidic circuits (IFCs), 25 Major adverse cardiovascular event rate
International normalized ratios (INR), 119 (MACE)
International sensitivity index (ISI), 119 HMGCR, 161
International Warfarin Pharmacogenetics pleiotropic effect, 162
Consortium (IWPC), 122 Markov model, 38
Invader, 16 MassARRAY, 30
Irinotecan Matrix-assisted laser desorption
genetic markers, 68 ionization-time of light
germline variation, 67 (MALDI-TOF), 30
pharmacodynamics, 6162 Maximum tolerated dose (MTD), 64
pharmacoethnicity, 6566 Mercaptopurine, 91, 92
pharmacogenetics Molecular diagnosis, pharmacogenomics
drug transporter, 65 Affymetrix technology, 18
Gilberts syndrome, 63 allele-specific primer extension
pharmacokinetics automated microarray, 20, 21
biliary excretion, 61 INFINITI assay, 21
intravenous infusion, 60 Luminex assay, 21
structure, 59, 60 DNA extraction and quantitation
toxicity, 62 cell lysis, 17
UGT1A1 genotyping picogreen assay, 17
hematological toxicity, 67 eSensor technology, 1819
neutropenia, 66 genotyping methods, 30
Invader assay, 27, 28
mass spectrometry, 3032
K microarray-based genotyping platform,
KIF6 polymorphism 1718
719Arg trial, 187188 pyrosequencing
ARIC study, 181 DNA synthesis, 29
CHD, 181 enzymatic reaction, 30
CHS, 182 real-time PCR
clinical event, 191 hybridization probe, 2527
elderly patients, 188189 hydrolysis probe, 2325
marker, 184 sequencing
PROSPER study, 183 amplification, 28
variant, risk, 184 capillary electrophoresis, 28, 29
WHS, 181 SNP analysis, 16
WOSCOPS, 182183 Verigene technology
WTCCC study, 184 genotyping platform, 22, 23
RUO/ASR test, 22
MPE. See Maculopapular eruption
L MTD. See Maximum tolerated dose
LeschNyhan syndrome, 214 Myocardial perfusion imaging
Limit of quantitation (LOQ), 253 (MPI), 176
270 Index
N R
N-acetyl transferase 2 (NAT2), 5 Real-time PCR
National Comprehensive Cancer Network hybridization probe
(NCCN), 60 mechanism, 26
Negative predictive value (NPV), 219 URS, microfluidics cartridge, 26, 27
Neuromuscular side effects (NMSEs), 157 hydrolysis
Neutropenia, 66 multiplexing assay, 24
Nottingham score, 84, 85 TaqManTM probe, 23
Number needed to treat (NNT), 178 Relative risk reduction (RRR), 175
Restriction fragment length polymorphism
(RFLP), 258
O Rivaroxaban, 120, 132
Oral anticoagulation therapy, 132 Rosuvastatin, 157, 159
P S
Pain management, 53 Salicylates
Percutaneous coronary interventions (PCI), mechanism, 145
140 metabolism, clinical efficacy, 145
Personalized justice (PJ), 48 pharmacogenetics, 146147
Personalized medicine (PM), 11, 47, 48 therapeutic implications, 147
Personalized Medicine Coalition Selective estrogen receptor modulator
(PMC), 11 (SERM), 77
Pharmacogenetics, 244245 Selective serotonin reuptake inhibitor
Pharmacogenomic testing (SSRI), 81
abacavir, 42 Severe cutaneous adverse reaction (SCAR), 216
antipsychotic drugs, 42 Simvastatin, 160
aromatase inhibitor, 35 Single nucleotide polymorphisms (SNPs), 158
drug efficacy, 36 Single photon emission computed tomography
hypothetical analysis, 37 (SPECT), 176
QALY, 3738 SJS. See Stevens Johnson syndrome
reimbursement, 1011 Statin induced neuromyopathy (SINM), 157
thiopurine methyltransferase, 3941 Statin therapy
warfarin, 4142 acute coronary syndrome, 190191
Pharmacovigilence, 202 atherosclerosis
Physiogenomics (PG), 165166 plaque rupture, 179
Picogreen, 17 stenotic lesion, 179
PM. See Personalized medicine cholesterol lowering, 176177
PMC. See Personalized Medicine clinical status
Coalition genomic profile, 157
Polymerase chain reaction (PCR), 258 myalgia, 157
Positive predictive value (PPV), 219 clinical trials, 177178
Pravastatin, 157, 160, 161 efficacy and side effects
Prospective randomized evaluation of DNA myositis, 166
screening in a clinical trial physiogenomics, 165
(PREDICT-1), 205 elderly patients, 188, 189
Proton-pump inhibitors (PPIs), 141 genotype-guided
Pyrosequencing HMGCoA reductase, 156
DNA synthesis, 29 inhibition, 156
enzymatic reaction, 30 KIF6, 167
KIF6 vs. CHD, 180181
kinesin proteins, 179180
Q lipid-lowering, 194
Quality-adjusted life years (QALY), 3738 low-density lipoprotein cholesterol
Index 271
APOE, 157 hepatotoxicity, 95

CLMN, 159 intensification chemotherapy, 107
simvastatin, 160 pathway, 92, 94
mortality and prevention pharmacogenetic testing
HMGCR, 161 dosage, algorithm, 108
MACE, 161 myelosuppression, 106
pleiotropic effect, 162 plasma concentration, 104
pharacogenomics and KIF6, 186187 product labeling, 102
safety therapeutic index, 102
myalgia, 165 thioguanine, 107
myopathy, 164 TPMT genotype, 104
SLCO1B1, 162 TPMT polymorphism, 102
treatment protein activity, 98
asymptomatic executive, 193194 toxicity and treatment, 9698
asymptomatic individual, 192193 TPMT activity, 108
Stevens Johnson syndrome (SJS), 216, 227 TPMT SNPs and haplotypes, 98, 100, 101
Streptomyces clavuligerus, 243 Thiopurine methyl transferase (TPMT), 39
Study of hypersensitivity to abacavir and activity, 109
pharmacogenetic evaluation genotype, 105
(SHAPE), 205 polymorphism, 102
Surface-enhanced laser desorption and Toxic epidermal necrolysis (TEN), 216
ionization mass spectrometer Translational pharmacogenomics
(SELDI-TOF), 31 alcoholism, 52
algorithms, reports, and interpretations, 8
antidepressants, antipsychotics, 52
T barriers, 56
Tacrolimus clinical trials, 68
dysfunction, 260 cost-effectiveness, 8
trial, 258259 drug
Taiwan Drug Relief Foundation (TDRF), 217 sensitivity, 49
Tamoxifen toxicity, 4849
clinical practice education, training, 1112
ductal carcinoma, 84 evidence based PJ, 49
lobular carcinoma, 85 FDA initiatives, 6
competitive inhibitor, 78 forensic pathology, 5051
CYP2D6 in vitro, molecular diagnostics, 51
inhibitor, 8182 laboratory test, guidelines, 810
polymorphism, 8081 pain management, 53
hormonal therapy, 79 personalized medicine, 12
metabolism, 79 PM vs. PJ, 4748
treatment, 83 pros and cons, 54
TaqManTM probe, 24 reimbursement, 1011
TEN. See Toxic epidermal necrolysis terminology/definition, 5
Therapeutic drug monitoring (TDM), 78, therapeutic index, 3
252253 toxicity profile, 4
Thioguanine nucleotide (TGN), 92 warfarin, 5253
Thiopurine
allopurinol, 96
antimetabolite prodrug, 91 U
chemical structure, 92 Unitized reagent strip (URS), 27
clinical uses, 91, 93
conventional therapy, 95
gene associations and GWAS study, 101 V
genetic variation, 98 Verigene technology, 16
272 Index
Verigene technology (cont.) W

genotyping platform, 22, 23 Warfarin, 5253
RUO/ASR test, 22 distribution, 119
Vitamin K dosage
CYP 2C9, 120121 initial algorithm, 122
demographic data, 133 IWPC, 122124
dosage limitation, 124125
initial algorithm, 122 genetic variation, 120
IWPC, 122124 genotype testing, 127128
limitation, 124125 oral anticoagulation therapy, 132
epoxide reductase complex 1, pharmacodynamics, 117, 118
121122 pharmacogenomics
GWAS study, 125126 CYP 2C9, 120121
oral anticoagulation therapy, 132 epoxide reductase complex 1, 121122
pharmacogenomic testing Wellcome Trust Case Control Consortium
economic issue, 130 (WTCCC), 184
multicenter trial, 130132 West of Scotland Coronary Prevention Study
published trial, 128130 (WOSCOPS), 181
rivaroxaban, 132 White blood cell (WBC), 62
Vitamin K epoxide reductase (VKOR), 118 Womens Health Study (WHS), 181

(Molecular and Translational Medicine) Alan H.B. Wu, Kiang-Teck J. Yeo-Pharmacogenomic Testing in Current Clinical Practice_ Implementation in the Clinical Laboratory (Molecular and Translational Medi

Uploaded by

Copyright:

Available Formats

(Molecular and Translational Medicine) Alan H.B. Wu, Kiang-Teck J. Yeo-Pharmacogenomic Testing in Current Clinical Practice_ Implementation in the Clinical Laboratory (Molecular and Translational Medi

Uploaded by

Document Information

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

(Molecular and Translational Medicine) Alan H.B. Wu, Kiang-Teck J. Yeo-Pharmacogenomic Testing in Current Clinical Practice_ Implementation in the Clinical Laboratory (Molecular and Translational Medi

Uploaded by

Copyright:

Available Formats

What are some of the main drugs and conditions discussed in the book?

What are some of the main drugs and conditions discussed in the book?

What are some of the challenges of implementing pharmacogenomic testing?

What are some of the challenges of implementing pharmacogenomic testing?

Molecular and Translational Medicine

For other titles published in this series, go to

ISBN 978-1-60761-282-7 e-ISBN 978-1-60761-283-4

Springer Science+Business Media, LLC 2011

Printed on acid-free paper

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

The personalization of therapeutics promises to deliver a more efficacious and

therapeutic use. We have determined that companion diagnostic tests are

Part I Basic Concepts

1 Issues in Translation of Pharmacogenomics

2 Molecular Diagnostic Methods in Pharmacogenomics.......................... 15

3 Economics of Pharmacogenomic Testing in Clinical Practice............... 35

4 From Personalized Medicine to Personalized Justice:

Part III Specific Pharmacogenomic Targets: Cardiovascular Drugs

8 The Pharmacogenetics of Vitamin K Antagonist

9 Clopidogrel and Salicylates....................................................................... 139

10 Genotype-Guided Statin Therapy............................................................ 155

11 The Statin Response Gene: KIF6............................................................. 175

Part IV Drugs that Cause Delayed Hypersensitivity

14 Carbamazepine and Its Structurally-Related Antiepileptics................. 225

Part V Miscellaneous Drugs

15 Pharmacogenetics of Flucloxacillin and Amoxicillin-Clavulanate

16 Immunosuppressants Pharmacogenomics............................................... 249

Christine L.H. Snozek

Kiang-Teck J. Yeo, Nikolina Babic, and Alan H.B. Wu

Keywords Pharmacokinetics Pharmacodynamics Pharmacogenomics Barriers

1.1Clinical Laboratory Tests

The Institute of Medicine was instructed by the US Congress to perform a

Table 1.1 Examples of drugs undergoing pharmacogenomics rescue

1.2Terminology and Definitions

Pharmacogenetics is not a new discipline, and the importance of inheritance in the

understanding by prescribing physicians regarding available pharmacogenetictests;

Table 1.2 US Food and Drug Administration relabeling initiative

genetic information to guide warfarin therapy initiation and improve anticoagulation

1.4.4PGx Algorithms, Reports, and Interpretations

Complex genotyping results need to be readily translated into actionable decisions

1.4.5Laboratory Aspects and Guidelines

Patient Name Medical Record Age Sex

Ordering Physician DOB

Attending Physician Report Notes

430 C>T (*2) C/T

Normal warfarin _______________

Fig. 1.1 Sample report for warfarin sensitivity genotyping

Table 1.3 Current FDA-cleared pharmacogenomics assays

At present, there are no specific American Medical Association Current Procedural

Table 1.4 Genetic testing reference material coordination programa

1.4.7Education, Training, and Pharmacogenomics Resources

Nikolina Babic, Loren J. Joseph, and Kiang-Teck J. Yeo

Keywords Real-time polymerase chain reaction Pyrosequencing Mass

Rapid advances in pharmacogenomics research have facilitated the transfer of

Table 2.1 Different SNP analysis approaches

phate-glucyronosyltransferase 1A1 (UGT1A1), thiopurine S-methyltransferase

2.2DNA Extraction and Quantitation

DNA quality is critical for successful genotyping regardless of the platform or

2.3Microarray-Based Genotyping Platforms

Table 2.2 Genotyping platforms and respective pharmacogenomics tests

1.1Clinical Laboratory Tests

1.2Terminology and Definitions

1.4.4PGx Algorithms, Reports, and Interpretations

1.4.5Laboratory Aspects and Guidelines

1.4.7Education, Training, and Pharmacogenomics Resources

2.2DNA Extraction and Quantitation

2.3Microarray-Based Genotyping Platforms

2.3.3INFINITI and Luminex xTAG Allele-Specific

3.2Economic Outcome Measures

3.3Cost-Effectiveness Studies for Specific Pharmacogenomic

3.3.1Pharmacogenomic Testing for Thiopurine

3.3.2Pharmacogenomic Testing for Warfarin Dosing

3.3.3Pharmacogenomic Testing for Abacavir

3.3.4Pharmacogenomic Testing for Antipsychotic Drugs

3.3.5Miscellaneous Other Pharmacoeconomic Studies

In addition to warfarin, one should genotype HLA-B5701 [14] and HLA-B1502

4.1.3Evidence Base for Personalized Justice

4.3Forensic Pathology Perspectives

4.4.1Drug Hypersensitivity In Vitro Diagnostics

4.5Illustrative Cases and Scenarios

4.5.2Antidepressants and Antipsychotics