10 1 1 463 3921

Heparin-Induced
Thrombocytopenia
Third Edition, Revised and Expanded
edited by
Theodore E. Warkentin
McMaster University
Hamilton Regional Laboratory Medicine Program, and
Hamilton Health Sciences, General Site
Hamilton, Ontario, Canada
Andreas Greinacher
Ernst-Moritz-Arndt University Greifswald
Greifswald, Germany
MARCEL
MARCEL
DEKKER,
INC. NEWYORK BASEL
DEKKER
Although great care has been taken to provide accurate and current information,
neither the author(s) nor the publisher, nor anyone else associated with this publica-
tion, shall be liable for any loss, damage, or liability directly or indirectly caused or
alleged to be caused by this book. The material contained herein is not intended to
provide specific advice or recommendations for any specific situation.
Trademark notice: Product or corporate names may be trademarks or registered

trademarks and are used only for identification and explanation without intent to
infringe.
Library of Congress Cataloging-in-Publication Data

A catalog record for this book is available from the Library of Congress.
ISBN: 0-8247-5625-8
This book is printed on acid-free paper.
Headquarters
Marcel Dekker, Inc., 270 Madison Avenue, New York, NY 10016, U.S.A.
tel: 212-696-9000; fax: 212-685-4540
Distribution and Customer Service

Marcel Dekker, Inc., Cimarron Road, Monticello, New York 12701, U.S.A.
tel: 800-228-1160; fax: 845-796-1772
Eastern Hemisphere Distribution

Marcel Dekker AG, Hutgasse 4, Postfach 812, CH-4001 Basel, Switzerland
tel: 41-61-260-6300; fax: 41-61-260-6333
World Wide Web

http://www.dekker.com
The publisher offers discounts on this book when ordered in bulk quantities. For more
information, write to Special Sales/Professional Marketing at the headquarters
address above.
Copyright n 2004 by Marcel Dekker, Inc. All Rights Reserved.
Neither this book nor any part may be reproduced or transmitted in any form or by
any means, electronic or mechanical, including photocopying, microfilming, and re-
cording, or by any information storage and retrieval system, without permission in
writing from the publisher.
Current printing (last digit):
10 9 8 7 6 5 4 3 2 1
PRINTED IN THE UNITED STATES OF AMERICA

FUNDAMENTAL AND CLINICAL CARDIOLOGY
Editor-in-Chief
Samuel Z. Goldhaber, M.D.
Harvard Medical School
and Brigham and Women's Hospital
Boston, Massachusetts
Associate Editor, Europe

Henri Bounameaux, M.D.
University Hospital of Geneva
Geneva, Switzerland
1. Drug Treatment of Hyperlipidemia, edited by Basil M. Rifkind

2. Cardiotonic Drugs: A Clinical Review, Second Edition, Revised and
Eicpanded, edited by Carl V. Leier
3. Complications of Coronary Angioplasty, edited by Alexander J. R.
Black, H. Vernon Anderson, and Stephen G. Ellis
4. Unsfable Angina, edited by John D. Rutherford
5. Befa-Blockers and Cardiac Arrhythmias, edited by Prakash C. Deed
wania
6. Exercise and the Heart in Health and Disease, edited by Roy J. Shep-
hard and Henry S. Miller, Jr.
7. Cardiopulmonary' Physiology in Critical Care, edited by Steven M.
Scharf
8. Atherosclerotic Cardiovascular Disease, Hemostasis, and Endofhelial
Function, edited by Robert Boyer Francis, Jr.
9. Coronary Heart Disease Prevention, edited by Frank G. Yanowitz
10. Thrombolysis and Adjunctive Therapy for Acute Myocardial Infarction,
edited by Eric R. Bates
11. Stunned Myocardium: Properties, Mechanisms, and Clinical Mani-
festations, edited by Robert A. Kloner and Karin Przyklenk
12. Prevention of Venous Thromboembolism, edited by Samuel Z. Gold-
haber
13. Silent Myocardial Ischemia and Infarction: Third Edition, Peter F. Cohn
14. Congestive Cardiac Failure: Pathophysiology and Treatment, edited by
David B. Barnett, Hubert Pouleur, and Gary S. Francis
15. Heart Failure: Basic Science and Clinical Aspects, edited by Judith K.
Gwathmey, G. Maurice Briggs, and Paul D. Allen
16. Coronary Thrombolysis in Perspective: Principles Underlying Con-
junctive and Adjunctive Therapy, edited by Burton E. Sobel and Desire
Collen
17. Cardiovascular Disease in the Elderly Patient, edited by Donald D.
Tresch and Wilbert S. Aronow
18. Systemic Cardiac Embolism, edited by Michael D. Ezekowitz
19. Low-Molecular- Weight Heparins in Prophylaxis and Therapy of Throm-
boembolic Diseases, edited by Henri Bounameaux
20. Valvular Heart Disease, edited by Muayed Al Zaibag and Carlos M. G.
Duran
2 1. Implantable Cardioverter-Defibrillators: A Comprehensive Textbook,
edited by N. A. Mark Estes Ill, Antonis S. Manolis, and Paul J. Wang
22. Individualized Therapy of Hypertension, edited by Norman M. Kaplan
and C. Venkata S. Ram
23. Atlas of Coronary Balloon Angioplasty, Bernhard Meier and Vivek K.
Mehan
24. Lowering Cholesterol in High-Risk Individuals and Populations, edited
by Basil M. Rifkind
25. lnterventional Cardiology: New Techniques and Strategies for Diag-
nosis and Treatment, edited by Christopher J. White and Stephen R.
Ramee
26. Molecular Genetics and Gene Therapy of Cardiovascular Diseases,
edited by Stephen C. Mockrin
27. The Pericardium: A Comprehensive Textbook, David H. Spodick
28. Coronary Restenosis: From Genetics to Therapeutics, edited by Giora
Z. Feuerstein
29. The Endothelium in Clinical Practice: Source and Target of Novel
Therapies, edited by Gabor M. Rubanyi and Victor J. Dzau
30. Molecular Biology of Cardiovascular Disease, edited by Andrew R.
Marks and Mark B. Taubman
31. Practical Critical Care in Cardiology, edited by Zab Mohsenifar and P.
K. Shah
32. lntravascular Ultrasound Imaging in Coronary Artery Disease, edited
by Robert J. Siege1
33. Saphenous Vein Bypass GraR Disease, edited by Eric R. Bates and
David R. Holmes, Jr.
34. Exercise and the Heart in Health and Disease: Second Edition, Re-
vised and Expanded, edited by Roy J. Shephard and Henry S. Miller,
J r.
35. Cardiovascular Drug Development: Protocol Design and Methodology,
edited by Jeffrey S. Borer and John C. Somberg
36. Cardiovascular Disease in the Elderly Patient: Second Edition, Re-
vised and Expanded, edited by Donald D. Tresch and Wilbert S.
Aronow
37. Clinical Neurocardiology, Louis R. Caplan, J. Willis Hurst, and Mark I.
Chimowitz
38. Cardiac Rehabilitation: A Guide to Practice in the 27st Century, edited
by Nanette K. Wenger, L. Kent Smith, Erika Sivarajan Froelicher, and
Patricia McCall Comoss
39. Heparin-Induced Thrombocytopenia, edited by Theodore E. Warkentin
and Andreas Greinacher
40. Silent Myocardial lschemia and Infarction: Fourth Edition, by Peter F.
Cohn
41. Foundations of Cardiac Arrhythmias: Basic Concepts and Clinical
Approaches, edited by Peter M. Spooner and Michael R. Rosen
42. Interpreting Electrocardiograms: Using Basic Principles and Vector
Concepts, J. Willis Hurst
43. Heparin-Induced Thrombocytopenia: Second Edition, edited by The-
odore E. Warkentin and Andreas Greinacher
44. Thrombosis and Thromboembolism, edited by Samuel Z. Goldhaber
and Paul M. Ridker
45. Cardiovascular Plaque Rupture, edited by David L. Brown
46. Therapeutic Agents for Thrombosis and Thrombolysis: Second Edition,
Revised and Expanded, edited by Arthur A. Sasahara and Joseph
Loscalzo
47. Heparin-Induced Thrombocytopenia: Third Edition, Revised and Ex-
panded, edited by Theodore E. Warkentin and Andreas Greinacher
ADDITIONAL VOLUMES IN PREPARATION

To the late Professor Michael F. X. Glynn, for initiating my hemostasis in-
terests; to Dr. John G. Kelton, for amplifying these through boundless op-
portunities; and to Erica, Andrew, Erin, and Nathan, for downregulating my
passion, as a caring family must.
T.E.W.
To my co-workers and students for their contributions and efforts; to Sabine,

Sebastian, Anja, and Jan.
A.G.
Series Introduction
Although I used to think heparin-induced thrombocytopenia (HIT) was an

extraordinarily rare and elusive disease, nothing could be further from the
truth. If a high level of suspicion is maintained, HIT will be diagnosed fre-
quently. Suprisingly, it causes deep venous thrombosis and pulmonary em-
bolism, which tend to be massive and life-threatening, far more often than it
causes arterial thrombosis. These blood clots cannot be treated with standard
anticoagulation, and routine management may yield catastrophic results.
Therefore, it is imperative that clinical cardiologists, internists, hematologists,
and vascular medicine and surgery specialists familiarize themselves with
HIT. To achieve this objective, it is a pleasure to present this updated land-
mark book, Heparin-Induced Thrombocytopenia: Third Edition, Revised and
Expanded, edited by two of the most renowned clinical investigators in the
field.
I will keep my personal copy on my closest bookshelf to help me as I
consult on these perplexing patients. Nowhere else is the research from this
emerging field pulled together so comprehensively and lucidly as in this
scholarly and practical book.
Samuel Z. Goldhaber
v
Preface to the Third Edition
Since the appearance of the second edition of Heparin-Induced Thrombocy-

topenia over three years ago, new and important advances in the under-
standing and treatment of this paradoxical adverse reaction to heparin have
continued to emerge.
For example, mapping of the target epitopes recognized by HIT anti-
bodies to platelet factor 4 (a ‘‘self protein’’ found within platelets) rather than
on heparin itself helps to explain some of the ‘‘autoimmune’’ features of HIT,
such as its potential to present as thrombocytopenia and thrombosis several
days after stopping heparin (‘‘delayed-onset HIT’’), as well as the increased
risk of thrombosis that can persist for several days or weeks even after HIT is
recognized and the inciting agent, heparin, stopped. A new double-transgenic
animal model of HIT (with mice engineered to produce both human platelet
factor 4 and human platelet Fcg receptors provides direct evidence that plate-
let Fc receptor–mediated platelet activation by HIT antibodies indeed helps to
explain thrombocytopenia and the associated risk of thrombosis in HIT.
There is increasing recognition of the danger of microvascular throm-
bosis in HIT, particularly when warfarin or other coumarin anticoagulants
are given when HIT remains active. This can lead to a variant of coumarin-
induced necrosis, with a predisposition to involve the extremities, i.e., war-
farin-induced venous limb gangrene complicating HIT-associated deep-vein
thrombosis. The recognition of this syndrome underscores the importance of
continuing therapy with a direct thrombin inhibitor or danaparoid (where
available), and postponing warfarin therapy until substantial resolution of
the thrombocytopenia of HIT has occurred.
In 2002, danaparoid was withdrawn from the United States, leaving
the direct thrombin inhibitors, lepirudin and argatroban, as the two main
agents in that market for managing the difficult clinical situation of HIT
complicated by thrombosis. (As of October 2003, danaparoid remained avail-
able in Canada and the European Union.) Another recent clinical develop-
ment is the recognition that lepirudin bolus infusion can (rarely) trigger life-
vii
viii Preface to the Third Edition
threatening anaphylaxis, perhaps because of antibodies that have been

formed against this foreign protein (although lepirudin is manufactured using
recombinant biotechnology, its blueprint is drawn from the medicinal leech).
Perhaps most important, there is increasing agreement that patients
suspected as having HIT should not merely have their heparin stopped, but
should additionally have an alternative anticoagulant given, so as to reduce
the risk of subsequent thrombosis. Indeed, argatroban has FDA approval for
this novel indication of prophylaxis against thrombosis in HIT. Post-mar-
keting studies indicate that lepirudin is also effective in this situation. Thus,
physicians need to consider carefully the possibility of HIT in many hospi-
talized patients who develop thrombocytopenia during heparin therapy, or
even in patients who return to the emergency room with thrombosis and
thrombocytopenia following a recent hospitalization in which heparin was
given (delayed-onset HIT). If HIT is strongly suspected, alternative antico-
agulation is indicated.
All 20 chapters from the second edition have been revised, and two new
chapters added. One reviews pediatric HIT, and the other discusses use of a
new direct thrombin inhibitor, bivalirudin, in the context of preventing and,
possibly, treating HIT. Of course, for their generosity of time in updating and
adding to this book, we thank the contributors. And, for their invaluable ef-
forts, we thank (in Canada) Jo-Ann I. Sheppard, Aurelio Santos, Jr., James
W. Smith, and Maria Adamek, and (in Germany) Uta Alpen, Petra Eichler,
Norbert Lubenow, Lena Carlsson, and Theresia Lietz.
Despite the lower risk of HIT with low molecular weight heparin and
the novel factor Xa-inhibiting pentasaccharide (fondaparinux), compared
with unfractionated heparin, HIT does not seem to be going away. It remains
an issue particularly in cardiac surgery patients, where unfractionated heparin
remains the prevailing drug for providing anticoagulation during the cardiac
surgery itself, as well as in the postoperative period in many centers. Further,
the increasing awareness of HIT, and the increasing availability of laboratory
testing for HIT antibodies, means that more cases of HIT continue to be
recognized, even if, perhaps, the overall frequency of this reaction is in decline
(due to increasing use of low molecular weight heparin). Further, the common
concurrence of thrombocytopenia and heparin therapy in hospitalized pa-
tients, and the medicolegal consequences of a missed diagnosis of HIT, mean
that this diagnosis continues to be considered and discussed daily around the
world. We hope that this compilation of information and practical guidelines
on HIT diagnosis and treatment will assist the health care professional in
managing this challenging and fascinating disorder.
Andreas Greinacher
Preface to the Second Edition
Since the first edition of Heparin-Induced Thrombocytopenia appeared, there

is much new on this topic. In particular, there is growing awareness of the
intense ‘‘thrombin storm’’ characteristic of heparin-induced thrombocytope-
nia (HIT), especially after heparin has been stopped. Recently, another thera-
peutic option became available to manage this situation, namely, argatroban,
a synthetic, small-molecule, direct thrombin inhibitor. Significantly, the
indication approved by the Food and Drug Administration for this novel
anticoagulant was ‘‘for prophylaxis or treatment of thrombosis in patients
with HIT.’’
This approval of argatroban for prevention of thrombosis in HIT paral-
lels the growing recognition that even ‘‘isolated HIT’’ (i.e., HIT recognized on
the basis of thrombocytopenia alone, rather than because of a new throm-
botic complication) is associated with an unacceptably high risk of life- and
limb-threatening thrombosis, even when heparin has been promptly discon-
tinued because of a falling platelet count. Indeed, this view that isolated HIT
itself is an indication for prescribing an alternative anticoagulant in most pa-
tients has been accepted by the 2000 Consensus Conference on Antithrom-
botic Therapy of the American College of Chest Physicians.
Thus, there now exist three anticoagulant agents for which there is
consensus regarding efficacy treatment for HIT: danaparoid sodium, recom-
binant hirudin (lepirudin), and argatroban (listed in order of availability).
Approval status of the three drugs with respect to HIT varies in different
countries, but even if unapproved they may be available for ‘‘off-label’’ use, or
on a ‘‘compassionate-use’’ basis. Important differences in pharmacokinetics,
particularly in half-life, metabolism, and monitoring, mean that each of these
three agents will be appropriate in some of the complex clinical settings in
which HIT occurs.
Other developments in HIT include new understanding of the molecular
structure of the target antigen, a proposed role for activation of monocytes in
ix
x Preface to the Second Edition
thrombin generation, and new animal models for studying this fascinating
syndrome. Even the clinical syndrome itself is now better understood: The
influence of previous heparin exposure on the timing of onset of HIT has been
clarified, and the peculiar transience of HIT antibodies has been shown. These
clinical and laboratory insights provide a firmer basis for estimating pretest
probabilities of HIT in various clinical settings, and also give a scientific
rationale for considering re-exposure to heparin in a patient with previous
HIT but whose antibodies are no longer detectable.
Ironically, improvements in a laboratory testing also mean that not all
detectable HIT antibodies are truly pathogenic. Thus, physicians need to
interpret results of diagnostic assays in the clinical context, so as to estimate
accurately the posttest probability of HIT. A major aim of this book is to
provide the relevant information for understanding such a ‘‘clinipathologic’’
framework of HIT.
More and more, it seems, HIT is an issue in cases of alleged medical
malpractice. This is because HIT often occurs in patients who received heparin
for prophylaxis of thrombosis. So it can seem evident even to a nonmedical
person that something fundamentically went wrong if the patient ended up
with severe pulmonary embolism or even limb loss.
To address these new developments, and others, all chapters of the first
edition have been updated. In addition, two new chapters have been added,
one discussing the use of argatroban for management of HIT and the other
dealing with U.S. perspectives on medicolegal aspects of HIT.
Andreas Greinacher
Preface to the First Edition
An anticoagulant turns procoagulant; an antithrombotic causes thrombosis.

This is the fundamental paradox of heparin-induced thrombocytopenia
(HIT), an antibody-mediated prothrombotic drug reaction without parallel
in clinical medicine.
Heparin justifiably is listed as an ‘‘essential’’ drug by the World Health
Organization (1997): with a rapid onset of action, simple laboratory moni-
toring, and a low cost, heparin has benefited countless patients. And yet,
beginning some 40 years ago, a few physicians asserted that heparin caused
unusual and sometimes catastrophic thrombi in some of their patients who
received the drug for a week or more. Subsequently, two landmark studies led
by a vascular surgeon, Donald Silver, identified the key elements of the HIT
syndrome: thrombocytopenia, thrombosis, and heparin-dependent antibod-
ies in the patient’s blood (Rhodes et al., 1973, 1977; see Chap. 1). But key
questions remained: how can heparin cause thrombosis? What is the frequen-
cy of this event? How should these patients be treated? This book summarizes
a quarter-century of observation and study that has begun to provide answers
to these questions.
The reader will observe several ‘‘themes’’ in this book. One is that HIT
should be considered a clinicopathologic syndrome. This means that HIT
should be diagnosed only when 1) one or more unexpected clinical events
occur during heparin treatment (most commonly, thrombocytopenia with or
without thrombosis), and 2) heparin-dependent antibodies can be demon-
strated in the laboratory. A corollary is that the inability to demonstrate the
antibodies using reliable assays means that an alternative diagnosis must be
considered. Both editors view HIT through this ‘‘filter’’ of confirmatory
laboratory testing. For us, the laboratory has been crucial to defining the HIT
syndrome, by making it possible to distinguish patients who really have HIT
from those affected by the numerous other causes of thrombocytopenia
encountered in clinical medicine. Indeed, our own first studies on HIT,
presented at the XIIIth Congress of the International Society on Thrombosis
xi
xii Preface to the First Edition
and Hemostasis in Amsterdam, focused on improvements and innovations in

diagnostic testing using platelet activation assays (Greinacher et al., 1991;
Warkentin et al., 1991). Coincidentally, this was the same time scientific
meeting at which Jean Amiral and colleagues (1991) announced the identity of
the protein coantigen of HIT (platelet factor 4; PF4), providing another
diagnostic avenue (enzyme immunoassay). Thus, when we stepped onto the
patient wards, we increasingly relied on the laboratory to confirm or refute the
diagnosis of HIT. Through mutually reinforcing experiences of clinic and
laboratory, the nature of the HIT syndrome was unfolded. And, over time,
the wide spectrum of complications of HIT, and its high frequency in certain
clinical settings, became apparent.
Our focus on HIT as a clinicopathologic syndrome has implications for
the terminology we have used in this book. Because the causative role of
heparin can generally be established—in the appropriate clinical context—by
the demonstration of pathogenic, heparin-induced thrombocytopenia to de-
scribe this syndrome (i.e., heparin can be shown convincingly to have ‘‘in-
duced’’ the platelet count fall in a particular patient). In contrast, we use the
term nonimmune heparin-associated thrombocytopenia (nonimmune HAT) to
describe patients who have developed thrombocytopenia during heparin
treatment and in whom a pathogenic role for HIT antibodies cannot be
shown. In our view, this term unambiguously denotes that HIT antibodies are
not responsible for the thrombocytopenia, while leaving open the possibility
that heparin may have played a role in the causation of the platelet count fall
by nonimmune mechanisms (although coinciding thrombocytopenia from
another cause is probably the most frequent explanation for this event). We
have also introduced the term pseudo-HIT to indicate those patients with
nonimmune HAT that, by virtue of associated thrombosis or the timing of
onset of thrombocytopenia, closely mimics HIT (see Chap. 12).
A second theme of this book is the importance of in vivo thrombin gen-
eration in the pathogenesis of HIT. By virtue of antibody-mediated activation
of platelets and endothelium, and the neutralization of heparin by PF4
released from activated platelets, the HIT syndrome can be understood as a
prothrombotic disorder characterized by activation of the coagulation system.
This concept of HIT helps explain its association with venous as well as
arterial thrombosis (by analogy with other hypercoagulable states, such as
congenital deficiency of natural anticoagulant factors), and also the occa-
sional HIT patient with decompensated, disseminated intravascular coagu-
lation (DIC).
Marked thrombin generation in HIT also helps explain its association
with coumarin-induced venous limb gangrene, an unusual syndrome now
recognized as a potential complication of coumarin treatment of HIT-
associated deep venous thrombosis (see Chap. 3). This iatrogenic disorder
Preface to the First Edition xiii
represents perhaps the most striking of all the HIT treatment paradoxes (see
Chap. 13): two antithrombotic agents with distinct adverse event profiles that
interact to produce a profound disturbance in procoagulant–anticoagulant
balance, i.e., increased thrombin generation (secondary to HIT) together with
acquired, severe protein C deficiency (secondary to coumarin treatment).
Finally, the concept of HIT as a hypercoagulable state with in vivo thrombin
generation provides a rationale for understanding the efficacy of new thera-
pies that either reduce thrombin generation by inhibition of factor Xa (e.g.,
danaparoid) or directly inactive thrombin (e.g., lepirudin).
A third theme of this book is the peculiarly inconstant nature of HIT, in
particular, its variable frequency and clinical presentation among different
patient populations treated with heparin. Figure 1 depicts HIT as an iceberg
within which a variety of clinical and laboratory factors interact to influence
antibody formation, development of thrombocytopenia, and, finally, result-
ing clinical complications, such as thrombosis. A recent, novel concept is that
the size and buoyancy of the HIT icebergs can vary among different patient
populations who receive heparin. This concept of multiple icebergs of HIT is
shown in Chapter 4, Fig. 3. Unraveling the clinical and laboratory determi-
nants for these differences in the icebergs among patient populations is a
major challenge of current and future investigation.
Figure 1 The iceberg model of heparin-induced thrombocytopenia as an index for

this book.
xiv Preface to the First Edition
Why should HIT be the subject of a book? First and foremost, HIT is
common, and nonimmune HAT is very common. According to the Council
for International Organization of Medical Sciences (CIOMS III), adverse
drug reactions can be classified as common if they occur in 1–10% of patients,
and very common if they occur in 10% or more of patients. There is convincing
evidence that HIT occurs in as many as 5% of certain patient populations,
such as postoperative orthopedic patients receiving unfractionated heparin.
The clinical influence of HIT is substantial: about half of these patients
develop HIT-associated thrombosis. Nonimmune HAT occurs in as many as
30% of certain patient populations. Thus, physicians need to be able to re-
liably distinguish among the various thrombocytopenic disorders that occur
during heparin treatment. This will minimize the risk of inappropriate treat-
ments, such as failing to stop heparin administration in a patient with prob-
able HIT, or deciding to stop heparin in a patient with nonimmune HAT or
pseudo-HIT. Because HIT is a life- and limb-threatening iatrogenic illness
with many diagnostic and treatment pitfalls, medicolegal consequences of
caregiver’s action or inaction can be significant (see Chap. 18).
A second reason for the compilation of this book is that most of the
pieces of the HIT puzzle are now firmly in place. Consensus has emerged on
several key aspects of the syndrome, including the nature of its target antigen,
the participation of platelet and endothelial cell activation in the pathogenesis
of thrombosis, the frequency of HIT, and optimal laboratory testing. The
publication of this book reflects this coherence in our understanding of the
HIT syndrome. Yet there remain important unresolved issues: for example,
what is the fundamental nature of the ‘‘autoimune’’ response to the PF4–
heparin neoepitope? Why is the immune response to the HIT antigen so tran-
sient? Why do only a subset of patients who form HIT antibodies develop
clinical HIT?
Heparin has been, and will continue to be, one of the most important
agents for the prophylaxis and treatment of venous and arterial thromboem-
bolism. Consequently, HIT will continue to be an important management
problem for some time to come. Both of us have spent a decade of our sci-
entific careers providing, in the context of other investigators’ work, a rational
management approach aimed at minimizing morbidity and mortality among
the many patients who develop the most common immune-mediated adverse
drug reaction in clinical medicine. The importance of controlling thrombin
generation in HIT is now widely appreciated. The book should help guide
clinicians through the often paradoxical clinical and management problems
posed by patients with HIT (see Chaps. 13–17).
The book is a tribute to our scientific mentors, John G. Kelton and
Christian Mueller-Eckhardt, and the close cooperation of many of our scien-
tific colleagues and personal friends whose efforts made this project possible.
Preface to the First Edition xv
We would also like to acknowledge the help of many individuals in this

project. In North America, we thank Jo-Ann Sheppard for technical support
over the years, and also for preparing many of the figures in this book; Aurelio
Santos, Jr., for preparing the key Fig. 13.1; James W. Smith, for help with
vexing computer problems; Katherine Bean and Maria Adamek, for able
secretarial assistance; and Erica Warkentin, for checking references and man-
uscripts. In Germany, we thank Uta Alpen for excellent secretarial assistance
and Petra Eichler, Norbert Lubenow, Lena Carlsson, and Oliver Ranze for
valuable discussion and review of manuscripts.
Andreas Greinacher
REFERENCES
Amiral J, Bridey F, Dreyfus M, Vissac AM, Fressinaud E, Meyer D. Identification of

PF4 as a target for antibodies generated in heparin-induced thrombocytopenia:
development of a diagnostic test [abstr]. Thromb Haemost 1991; 65(suppl):865.
Greinacher A, Michels I, Mueller-Eckhardt C. Heparin induced platelet activation
(HIPA) test: a rapid and sensitive tool for diagnosing heparin associated
thrombocytopenia (HAT) and selecting a compatible heparin [abstr]. Thromb
Haemost 1991; 65(suppl):795.
Rhodes GR, Dixon RH, Silver D. Heparin induced thrombocytopenia with throm-
botic and hemorrhagic manifestations. Surg Gynecol Obstet 1973; 136:409–416.
Rhodes GR, Dixon RH, Silver D. Heparin induced thrombocytopenia: eight cases
with thrombotic-hemorrhagic complications. Ann Surg 1977; 186: 752–758.
Warkentin TE, Kelton JG. Determinants of platelet donor variability in diagnostic
testing for heparin-induced thrombocytopenia [abstr]. Thromb Haemost 1991;
65(suppl):1036.
World Health Organization. Seventh report of the WHO Expert Committee. The use
of essential drugs. WHO Technical Report Series. Geneva: World Health
Organization, 1997:37.
Contents
Series Introduction (Samuel Z. Goldhaber) v

Preface to the Third Edition vii
Preface to the Second Edition ix
Preface to the First Edition xi
Contributors xxi
1. History of Heparin-Induced Thrombocytopenia 1

2. Differential Diagnosis of Acute Thrombocytopenia 25

Volker Kiefel
3. Clinical Picture of Heparin-Induced Thrombocytopenia 53

4. Frequency of Heparin-Induced Thrombocytopenia 107

David H. Lee and Theodore E. Warkentin
5. Nonimmune Heparin–Platelet Interactions: Implications

for the Pathogenesis of Heparin-Induced Thrombocytopenia 149
McDonald K. Horne III
6. Heparin-Dependent Antigens in Heparin-Induced

Thrombocytopenia 165
Jean Amiral and Dominique Meyer
7. Molecular Immunopathogenesis of Heparin-Induced

Gian Paolo Visentin, Chao Yan Liu, and Richard H. Aster
xvii
xviii Contents
8. Role of Sulfated Polysaccharides in the Pathogenesis of

Heparin-Induced Thrombocytopenia 197
Susanne Alban and Andreas Greinacher
9. The Platelet Fc Receptor in Heparin-Induced

Gregory A. Denomme
10. Immune Vascular Injury in Heparin-Induced

Gowthami M. Arepally, Mortimer Poncz, and
Douglas B. Cines
11. Laboratory Testing for Heparin-Induced

Theodore E. Warkentin and Andreas Greinacher
12. Pseudo–Heparin-Induced Thrombocytopenia 313

13. Treatment of Heparin-Induced Thrombocytopenia: An

Overview 335
Andreas Greinacher and Theodore E. Warkentin
14. Danaparoid for the Treatment of Heparin-Induced

Beng Hock Chong and Harry N. Magnani
15. Lepirudin for the Treatment of Heparin-Induced

Andreas Greinacher
16. Argatroban Therapy in Heparin-Induced

Bruce E. Lewis and Marcie J. Hursting
17. Bivalirudin for the Treatment of Heparin-Induced

John R. Bartholomew
Contents xix
18. Hemodialysis in Heparin-Induced Thrombocytopenia 509

Karl-Georg Fischer
19. Management of Cardiopulmonary Bypass

Anticoagulation in Patients with Heparin-Induced
Bernd Poetzsch and Katharina Madlener
20. Heparin-Induced Thrombocytopenia in Children 553

Anne F. Klenner and Andreas Greinacher
21. Legal Aspects of Heparin-Induced Thrombocytopenia:

U.S. Perspectives 573
Kevin M. McIntyre and Theodore E. Warkentin
22. Legal Aspects of Heparin-Induced Thrombocytopenia:

European Perspectives 587
Klaus Ulsenheimer
Appendixes
1 Ten Clinical ‘‘Rules’’ for Diagnosing HIT 597
2 Estimating the Pretest Probability of HIT: The Four T’s 599
3 Platelet Count Monitoring for HIT 600
4 Treatment Recommendations 602
5 Danaparoid Dosing Schedules in HIT Patients 606
6 Dosing Schedules for Lepirudin Treatment of Patients with
HIT 608
7 Dosing Schedule for Lepirudin in Patients with HIT and
Renal Impairment 610
8 Dosing Schedules for Argatroban Treatment of Patients
with HIT (Approved Indications) 611
9 Timelines of an Episode of HIT 612
Index 613
Contributors
Susanne Alban, Ph.D. Senior Researcher and Lecturer, Institute of Phar-

macy, Christian-Albrechts University of Kiel, Germany
Jean Amiral, Ph.D. Scientific Director, HYPHEN BioMed, Neuville-sur-

Oise, France
Gowthami M. Arepally, M.D. Assistant Professor, Department of Medicine,

Duke University Medical Center, Durham, North Carolina, U.S.A.
Richard H. Aster, M.D. Professor, Department of Medicine, Medical

College of Wisconsin, and Senior Investigator, Blood Research Institute,
The Blood Center of Southeastern Wisconsin, Milwaukee, Wisconsin, U.S.A.
John R. Bartholomew, M.D. Program Director, Section of Vascular Med-

icine, Departments of Cardiovascular Medicine and Hematology/Oncology,
Cleveland Clinic Foundation, Cleveland, Ohio, U.S.A.
Beng Hock Chong, M.B.B.S., Ph.D., F.R.C.P.(Glasgow), F.R.A.C.P,

F.R.C.P.A. Professor and Head, Department of Medicine, St. George Clin-
ical School, The St. George Hospital, University of New South Wales, Koga-
rah, New South Wales, Australia
Douglas B. Cines, M.D. Professor, Department of Pathology and Labora-

tory Medicine, and Director, Coagulation Laboratories, University of Penn-
sylvania School of Medicine, Philadelphia, Pennsylvania, U.S.A.
Gregory A. Denomme, Ph.D. Assistant Professor, Department of Labora-

tory Medicine and Pathobiology, University of Toronto, and Scientist, Re-
search and Development, Canadian Blood Services, and Associate Staff
xxi
xxii Contributors
Scientist, Department of Pathology and Laboratory Medicine, Mount Sinai

Hospital, Toronto, Ontario, Canada
Karl-Georg Fischer, M.D. Division of Nephrology and General Medicine,

Department of Medicine, University Hospital Freiburg, Freiburg, Germany
Andreas Greinacher, M.D. Professor, Institute for Immunology and Trans-

fusion Medicine, Ernst-Moritz-Arndt University Greifswald, Greifswald,
Germany
McDonald K. Horne III, M.D. Senior Clinical Investigator, Hematology

Service, Department of Laboratory Medicine, Warren G. Magnuson Clinical
Center, National Institutes of Health, Bethesda, Maryland, U.S.A.
Marcie J. Hursting, Ph.D., D.A.B.C.C. Director, Clinical Science Consult-

ing, Potomac, Maryland, U.S.A.
Volker Kiefel, M.D. Professor, Department of Transfusion Medicine, Uni-

versity of Rostock, Rostock, Germany
Anne F. Klenner, M.D. Institute for Immunology and Transfusion Medi-

cine, Ernst-Moritz-Arndt University Greifswald, Greifswald, Germany
David H. Lee, M.D., F.R.C.P.(C) Associate Professor, Department of Med-

icine, Queen’s University, Kingston, Ontario, Canada
Bruce E. Lewis, M.D., F.A.C.C. Professor of Medicine, Loyola University

Medical Center, Maywood, Illinois, and Chief of Cardiology, Catholic
Health Partners, Chicago, Illinois, U.S.A.
Chao Yan Liu, M.D. Senior Research Scientist, Department of Pediatrics,

University at Buffalo, The State University of New York, Buffalo, New York,
U.S.A.
Katharina Madlener, M.D. Senior Physician, Department of Hemostasis

and Thrombosis, Kerckhoff-Klinik, Sprudelhof II, Bad Nauheim, Germany
Harry N. Magnani, M.D., B.S., M.Sc., Ph.D. Medical Consultant, Oss, The
Netherlands
Kevin M. McIntyre, M.D., J.D., F.A.C.C., F.A.C.P. Associate Clinical

Professor, Department of Medicine, Harvard Medical School, Department
Contributors xxiii
of Medicine, Boston VA Health Care System, and Brigham and Women’s

Hospital, Boston, Massachusetts, U.S.A.
Dominique Meyer, M.D. Professor of Hematology, INSERM U-143, Hôpi-

tal de Bicêtre, Bicêtre, France
Mortimer Poncz, M.D. Professor, Department of Pediatrics, and Chief,

Division of Pediatric Hematology, University of Pennsylvania School of
Medicine, and Children’s Hospital of Philadelphia, Philadelphia, Pennsylva-
nia, U.S.A.
Bernd Poetzsch, M.D. Professor, Department of Experimental Hematology

and Transfusion Medicine, Rheinische Friedrich-Wilhelms-University Bonn,
Bonn, Germany
Klaus Ulsenheimer, Dr.iur., Dr.rer.pol. Professor, Faculty of Law, Univer-

sity of Munich, Munich, Germany
Gian Paolo Visentin, M.D. Associate Professor, Department of Pediatrics,

and Director, Digestive Disease and Nutrition Laboratory, University at
Buffalo, The State University of New York, Buffalo, New York, U.S.A.
Theodore E. Warkentin, M.D., F.R.C.P.(C), F.A.C.P. Professor, Depart-

ment of Pathology and Molecular Medicine and Department of Medicine,
McMaster University, and Associate Head, Transfusion Medicine, Hamilton
Regional Laboratory Medicine Program, Hamilton, Ontario, Canada
1
History of Heparin-Induced
Thrombocytopenia
McMaster University and Hamilton Regional Laboratory Medicine Program,
I. THE DISCOVERY OF HEPARIN AND ITS FIRST

CLINICAL USE
The following account of the discovery and first clinical development of

heparin was recorded by the physiologist Charles H. Best (1959), a codiscov-
erer of insulin as well as a pioneer in studies of heparin. Incidentally, in 1916,
while working at Johns Hopkins University to characterize procoagulant
substances, Dr. Jay McLean (1916) identified a natural anticoagulant sub-
stance. Further studies of this material were performed by his supervisor, Dr.
Howell, who coined the term, ‘‘heparin’’ to indicate its first extraction from
˘
animal hepatic tissues (Gr. DN kaU [h epar], liver) (Howell and Holt, 1918).
Despite its in vitro anticoagulant action, the inability of heparin to prevent
platelet-mediated thrombosis (Shionoya, 1927) made it uncertain whether it
had antithrombotic potential. However, animal (Mason, 1924) and human
studies (Crafoord, 1937) showed that heparin could prevent thrombosis. By
the 1950s, heparin was established as an important therapeutic agent in the
treatment of venous and arterial thrombosis.
II. THE PARADOX OF HEPARIN AS A POSSIBLE CAUSE

OF THROMBOSIS
A. Weismann and Tobin
On June 1, 1957, at the Fifth Scientific Meeting of the International Society of
Angiology (North American Chapter) in New York, two physicians sug-
1
2 Warkentin
Figure 1 Photograph of Dr. Rodger Elmer Weismann, taken circa 1958.
gested that heparin might cause arterial embolism in some patients. Rodger E.
Weismann, a 43-year-old Assistant Professor of Clinical Surgery at the
Dartmouth Medical School (Fig. 1), and his Resident in Surgery, Dr. Richard
W. Tobin, presented their 3-year experience with 10 patients who developed
unexpected peripheral arterial embolism during systemic heparin therapy at
the Mary Hitchcock Memorial Hospital, in Hanover, New Hampshire. Their
first patient with this complication was reported in detail, and to this day
represents a classic description of the syndrome:
This 62-year-old white woman was admitted to the Hitchcock Hospital
Feb 8, 1955, with left retinal detachment, complicating longstanding
myopia. . . . Left scleral buckling was carried out on Feb. 10, and strict
bed rest was required during the ensuing three weeks. On her beginning
ambulation, on March 6, signs and symptoms of left iliofemoral throm-
bophlebitis were noted, for which systemic heparinization was begun
(. . . heparin sodium in divided subcutaneous doses, totaling 150–300 mg
per day . . .). On March 16 . . . . , after 10 days of anticoagulation therapy,
sudden signs of right common femoral arterial occlusion led to the diag-
nosis of common femoral arterial embolism. Successful femoral em-
bolectomy was carried out. She was kept on adequate heparinization
and made a satisfactory initial recovery until March 19, . . . when signs
of sudden occlusion of the distal aorta appeared.
History of HIT 3
. . . [P]rompt transperitoneal distal aortic and bilateral iliac embolecto-

mies were performed. In the ensuing 24 hours, because unsatisfactory
distal circulation persisted, the patient underwent left femoral explora-
tion, with negative findings, and right popliteal exploration, revealing an
embolus. She subsequently pursued a favorable course, . . . never showing
more serious ischemic changes than a small area of superficial gangrene of
the right great toe and several small areas of skin infarction of the right leg
(Weismann and Tobin, 1958).
The report included a photograph of the emboli removed from the distal
aorta and both iliac arteries, with the authors noting their ‘‘unusual length
and cylindrical shape, suggesting origin in [the] proximal aorta,’’ as well as a
corresponding photomicrograph of the embolus. The thromboemboli were
described by the authors as ‘‘pale, soft, salmon-colored clots’’ that ‘‘histo-
logically . . . were comprised mostly of fibrin, platelets and leukocytes; red cells
were rare.’’ This appearance was distinguishable from the typical appearance
of thrombi originating in the heart (i.e., mulberry-colored thrombi tending to
contain cellular elements of the blood in approximately normal proportions),
leading the authors to propose ‘‘the source for the emboli . . . to be aortic mu-
ral platelet-fibrin thrombi.’’
A summary of the 10 reported patients noted that the onset of arterial
embolism began between 7 and 15 days, inclusive, of commencing heparin
treatment (mean, day 10). Multiple thromboemboli occurred in nine patients;
six of the patients died as a direct result of these complications; two survived
with extensive amputations, and two were discharged with their extremities
intact. The temporal time frame was consistent with the later realization by
others that this syndrome represented an immune-mediated reaction initiated
by the heparin.
The authors noted that further embolization stopped when the hepa-
rin was discontinued, leading to their recommendation that ‘‘heparin should
be promptly reduced in dosage, and, if possible, discontinued if the presence
of fibrin-platelet thrombi adherent to the intima of the aorta is suspected.’’
Aggressive surgical management of emboli was also recommended, as some
limbs were salvageable in this way. The authors summarized well the clinical
dilemma: ‘‘In each instance there was a feeling of futility in the management
of the problem, due to anticipation of further emboli from the same or similar
sources. Heparin was badly needed to retard distal thrombosis; yet the agent
was probably seriously altering the integrity and attachment of the throm-
botic source’’ (Weismann and Tobin, 1958).
B. Roberts and Colleagues

The communication of Weismann and Tobin was met with considerable skep-
ticism. When a show of hands was elicited to indicate those surgeons who had
4 Warkentin
also observed similar events, none were raised (Weismann, personal commu-
nication, July 1998). However, a few years later, Brooke Roberts and col-
leagues from the University of Pennsylvania in Philadelphia described a series
of patients who were remarkably similar to those reported by Weismann and
Tobin (Roberts et al., 1964; Barker et al., 1966; Kaupp and Roberts, 1972).
The key features were summarized as follows:
To witness a series of apparently paradoxical events is disconcerting as
well as challenging. When such paradoxes involve totally unexpected
results following the use of a major therapeutic agent, it is at first
difficult to know whether the relationship is causal or merely
coincidental. When, however, the same series of events has been seen
repeatedly it is difficult to escape the conclusion that there is some causal
relationship, even though the mechanism by which it is accomplished
may be unknown. . . . During the last 9 years at the Hospital of the
University of Pennsylvania we have seen a group of 11 patients who
suffered unexplained arterial embolization for the first time while being
treated with heparin for some condition that could not of itself
reasonably be expected to cause arterial emboli. . . . All patients had
been receiving heparin for 10 days or more when the initial embolus
occurred. . . . All emboli removed were of a light color, seemingly made
up primarily of fibrin and platelets, and microscopically appeared to be
relatively free of red cells. . . . All patients in this group had multiple
emboli. . . . Of the 4 deaths, 3 were attributed to cerebral vascular
accidents presumably embolic in origin and 1 was thought to have
resulted from a perforation of the small bowel 2 weeks after the removal
of a mesenteric embolus (Roberts et al., 1964).
Roberts’ group also viewed the likely pathogenesis as that of emboli-
zation of platelet-fibrin–rich material originating within the aorta, rather than
the heart. Furthermore, they believed that the thrombi were initially formed
on aortic ulcerations that acted as a nidus for thrombus formation. This
pathogenesis was suggested by the observation that such adherent thrombi
could be removed from the proximal aorta in a few of the patients (Roberts et
al., 1964; Kaupp and Roberts, 1972).
C. An Immune Basis for Heparin-Induced Thrombosis?

The delay between initiation of heparin therapy and onset of embolization
caused Roberts and colleagues (1964) to speculate that the etiology could re-
present an ‘‘antiheparin factor,’’ resulting perhaps from ‘‘an antigen–anti-
body mechanism.’’ Furthermore, the observation that the first 21 patients
reported with this syndrome from both Hanover and Philadelphia had re-
ceived heparin exclusively by subcutaneous or intramuscular, rather than
intravenous, injection also was offered by Roberts’ group as support for im-
History of HIT 5
mune sensitization. Apparent heparin-induced thrombosis did not seem rare

to these investigators: at least 13 of 110 (12%) patients with peripheral arterial
emboli managed by the Philadelphia group over a decade were believed to
have been caused by preceding heparin treatment (Barker et al., 1966).
III. HEPARIN-INDUCED THROMBOCYTOPENIA

AND PARADOXICAL THROMBOSIS
A. Heparin-Induced Thrombocytopenia
Routine platelet count measurements were not a feature of hospital labora-
tory practice until the 1970s, and neither the Dartmouth nor Philadelphia
surgeons reported thrombocytopenia in their patients with heparin-induced
arterial thrombosis. Ironically, the first report of severe heparin-induced
thrombocytopenia involved a patient who did not develop paradoxical throm-
bosis. Natelson and coworkers (1969) reported on a 78-year-old man with
prostate carcinoma and pulmonary embolism, who on day 10 of treatment
with therapeutic-dose heparin developed severe thrombocytopenia. Three
days after discontinuing the heparin therapy, the patient’s fibrinogen fell to 1
g/L, attributed to carcinoma-associated disseminated intravascular coagula-
tion (DIC). Heparin treatment was restarted and, although fibrinogen levels
normalized, the platelet count fell to 5 109/L, rising to 115 109/L 6 days
after stopping heparin administration. Simultaneously, however, the fibrin-
ogen value fell to less than 0.5 g/L. When heparin was given for the third time,
the platelet count fell over 2 days to 10 109/L, although the fibrinogen values
again normalized. In vitro studies showed that heparin added to the patient’s
citrated platelet-rich plasma produced platelet count reductions. This early
report of severe heparin-induced thrombocytopenia is interesting, as it
illustrates the dichotomy of heparin reproducibly producing severe throm-
bocytopenia while at the same time maintaining anticoagulant activity
(correction of DIC). However, it remained for later workers to link throm-
bocytopenia and thrombosis to heparin therapy.
B. Rhodes, Dixon, and Silver: ‘‘

‘‘Heparin-Induced
Thrombocytopenia with Thrombotic and
Hemorrhagic Manifestations’’’’
Laboratory evidence implicating an immune basis for heparin-induced throm-
bocytopenia (HIT) was first provided by studies performed by a vascular
surgeon (Donald Silver; Fig. 2), in collaboration with two residents (Glen R.
Rhodes and R. H. Dixon). The first two patients described by Silver’s group
(Rhodes et al., 1973) developed severe thrombocytopenia (platelet count
6 Warkentin
Figure 2 Photograph of Dr. Donald Silver, taken circa 1975.
nadirs, 8 and 10 109/L), myocardial infarction, petechiae, and heparin re-

sistance, with complete platelet count recovery on discontinuing heparin treat-
ment. Both patients developed rapid recurrence of thrombocytopenia when
heparin rechallenges were given within 1 week of platelet count recovery.
The immune basis of this syndrome was suggested by several laboratory
observations. First, increased platelet consumption was suggested by in-
creased numbers of marrow megakaryocytes, as well as immediate recurrence
of thrombocytopenia on reexposure to heparin. Second, a circulating platelet-
activating substance was found in both patients’ blood: patient, but not con-
trol, serum resulted in aggregation of normal donor platelets in the presence
of heparin. Third, the possible identity of the aggregating agent as an
immunoglobulin G (IgG) was shown by fractionation of one patient’s serum
to show the presence of heparin-dependent, complement-fixing activities
within the IgG fraction.
A second report from this group (Rhodes et al., 1977) represented the
landmark study in establishing HIT as a distinct syndrome. Eight patients
History of HIT 7
were reported with thrombocytopenia that occurred during intravenous ther-

apeutic-dose or subcutaneous prophylactic-dose heparin. The mean platelet
count nadir was 25 (range, 5–54 109/L). The predominance of thrombotic,
rather than hemorrhagic, complications was demonstrated: seven patients
had new or recurrent thromboembolic events, and the remaining patient had a
stroke leading to evacuation of a temporal lobe hematoma. Complement-
fixing, heparin-dependent antibodies were identified in five of the patients.
The authors also cited the previous work by Weismann and Tobin (1958) and
Roberts and colleagues (1964) as likely representing the identical syndrome.
Thus, for the first time, the concept of an immune-mediated hypercoagulable
state, with a predisposition to arterial thromboembolism that occurred in
association with thrombocytopenia, was proposed.
C. Platelet-Activating Antibodies in the Pathogenesis of HIT

Although some limited studies of heparin-dependent platelet aggregation by
patient serum were performed in the classic study by Rhodes and colleagues
(1973), the next few years saw increasing emphasis on this characteristic
feature of HIT antibodies. In 1975, National Institutes of Health investigators
Fratantoni et al. described a patient who developed severe thrombocytopenia
(4 109/L) and pulmonary embolism while receiving therapeutic-dose
unfractionated heparin (UFH) to treat deep venous thrombosis. Recurrent
thrombocytopenia resulted following heparin rechallenge. The patient’s
serum produced both aggregation and serotonin release from normal platelets
in the presence of heparin. The platelet-activating factor was presumed, but
not proved, to be caused by an antibody.
During the next 5 years, at least eight groups of investigators reported
similar patients, confirming the presence of heparin-dependent, platelet-
activating antibodies (Babcock et al., 1976; Green et al., 1978; Nelson et al.,
1978; Trowbridge et al., 1978; Wahl et al., 1978; Cimo et al., 1979; Hussey et
al., 1979; Cines et al., 1980). Babcock and colleagues (1976) described five
patients who developed thrombocytopenia (mean platelet count nadir, 28
109/L) during heparin treatment; heparin-dependent antibodies were detected
that produced platelet factor 3 activity (i.e., patient globulin fractions in-
cubated with heparin, platelet-rich plasma, and celite-activated contact
product shortened the clotting time following recalcification). Three patients
developed thrombotic complications, and none developed hemorrhage. The
five patients were observed within a 6-week time span, leading the authors to
suggest that ‘‘this syndrome may occur more often than has previously been
suspected.’’
A consistent theme was evident from these various reports. Patients
developed arterial or venous thrombotic complications, in association with
8 Warkentin
thrombocytopenia that generally began after 5 or more days of heparin

treatment. A platelet-activating antibody that aggregated platelets suspended
in citrated plasma was usually detected. The platelet count nadirs seen in some
of the larger series (e.g., 33 and 48 109/L, respectively) observed by Cimo et
al. (1979) and Hussey et al. (1979), were higher than in previous reports,
indicating that as recognition of the syndrome grew, less severely thrombo-
cytopenic patients were recognized.
D. The ‘‘
‘‘White Clot Syndrome’’
’’
Jonathan Towne, a vascular surgeon in Milwaukee, reported with his
colleagues (1979) that the pale thrombi characteristic of this syndrome
consisted of fibrin–platelet aggregates (electron microscropy). These workers
coined the term ‘‘white clot syndrome’’ to describe the characteristic appear-
ance of these arterial thromboemboli. Ironically, their report is also the first to
note the occurrence of phlegmasia cerulea dolens (severe venous limb is-
chemia) that progressed to venous limb gangrene in two of their patients (i.e.,
a syndrome of limb loss due to extensive venous thrombosis without arterial
white clots). Nonetheless, the designation of white clot syndrome has become
virtually synonymous with HIT in both North America and Europe (Benha-
mou et al., 1985; Stanton et al., 1988), despite the lack of specificity of these
thrombi for HIT (see Chap. 12).
IV. NONIMMUNE HEPARIN-ASSOCIATED

THROMBOCYTOPENIA
A. Nonimmune Mechanisms in Heparin-Associated
Thrombocytopenia
Klein and Bell (1974) reported on two patients who developed severe throm-
bocytopenia, thrombotic complications, and DIC, with hypofibrinogenemia
and microangiopathic red cell abnormalities; i.e., these patients likely had
severe HIT. This experience prompted Bell to perform the first prospective
study investigating the frequency of thrombocytopenia complicating thera-
peutic-dose UFH (Bell et al., 1976). Sixteen of 52 patients (31%) developed a
platelet count fall to less than 100 109/L, and some of these patients
developed hypofibrinogenemia and elevated fibrin(ogen) degradation prod-
ucts. The authors speculated that a ‘‘thromboplastic contaminant’’ extracted
along with heparin from beef lung could explain the thrombocytopenia. A
subsequent randomized controlled trial by Bell and Royall (1980) found the
frequency of thrombocytopenia to be higher in patients who received bovine
heparin (26%) compared with heparin of porcine intestinal origin (8%).
History of HIT 9
These investigators found no platelet-activating antibodies in plasma

from the patients who developed thrombocytopenia (Alving et al., 1977),
leading Bell (1988) to challenge the view that an immune pathogenesis
explained HIT. However, as the Johns Hopkins group did not report
thrombotic complications in any of their 37 patients who developed throm-
bocytopenia in their prospective studies, and given the apparent early onset of
thrombocytopenia in many of their patients, it is likely that most of their
patients did not have immune-mediated HIT.
B. Nonimmune (Type I) Versus Immune (Type II)

Heparin-Induced Thrombocytopenia
A confusing situation arose. The terms ‘‘heparin-induced thrombocytope-
nia’’ or ‘‘heparin-associated thrombocytopenia’’ were often applied to any
patient who developed thrombocytopenia during heparin therapy, whether
presumed or proved to be caused by heparin-dependent antibodies or
otherwise. Investigators in Australia, led by Dr. Beng Chong (1981), also
observed patients with thrombocytopenia in whom heparin-dependent,
platelet-activating IgG antibodies could be identified. In a subsequent report
that appeared in the Lancet, two distinct syndromes of ‘‘heparin-induced
thrombocytopenia’’ were described by Chong and colleagues (1982). The
first, called ‘‘group 1,’’ developed severe, delayed-onset thrombocytopenia
with thrombotic complications in association with IgG antibodies that caused
platelet activation. In contrast, ‘‘group 2’’ patients had mild asymptomatic
thrombocytopenia of early onset.
In 1989, at a Platelet Immunobiology Workshop in Milwaukee, it was
suggested to Chong that terminology describing these two types of HIT be
formalized. Accordingly, Chong recommended the terms in a review article
that appeared in Blut (Chong and Berndt, 1989), although (in reverse of the
Lancet article nomenclature) the early, nonimmune disorder was named
‘‘HIT type I’’ and the later-onset, immune disorder referred to as ‘‘HIT type
II.’’ These terms subsequently became popular.
V. LABORATORY TESTING TO CHARACTERIZE THE HIT

SYNDROME
A. A Sensitive and Specific Platelet Activation Assay for HIT
Many clinical laboratories began to use platelet aggregation assays (Fratan-
toni et al., 1975; Babcock et al., 1976) to diagnose HIT. Problems with this
type of assay, however, included low sensitivity (Kelton et al., 1984) as well as
technical limitations in simultaneous evaluation of multiple patient and
control samples. In 1983–1984, while working as a research fellow in the
10 Warkentin
McMaster University laboratory of John Kelton, Dave Sheridan overcame

problems of low test sensitivity by showing that washed platelets, resuspended
in a buffer containing physiological concentrations of divalent cations, were
very sensitive to platelet activation by HIT sera (Sheridan et al., 1986). The
assay, known as the ‘‘platelet serotonin release assay (SRA),’’ was adapted
from a method of platelet washing developed at McMaster University by the
laboratory of Dr. Fraser Mustard. In particular, the emphasis on using phys-
iological calcium concentrations was based on observations that ‘‘artifacts’’
of agonist-induced platelet activation were caused by use of citrate anticoagu-
lation resulting in low plasma calcium concentrations. One example of an
artifact induced by citrate is that of two-phase aggregation triggered by ADP.
At physiological calcium concentrations, only weak single-phase aggregation
without thromboxane generation is triggered by ADP (Kinlough-Rathbone
et al., 1983). Fortuitously, the washed platelet technique previously developed
at McMaster University by Mustard and colleagues that Sheridan evaluated
for its HIT serum-sparing properties rendered platelets far more sensitive to
the platelet-activating properties of HIT antibodies than assays based on
citrated platelet-rich plasma. Modified washed platelet assays have subse-
quently been developed by other investigators (see Chap. 11).
Sheridan and colleagues also made the observation that heparin con-
centrations strongly influenced platelet activation by HIT sera: therapeutic
(0.05–1 U/mL), but not high (10–100 U/mL), heparin concentrations resulted
in platelet activation, i.e., the characteristic ‘‘two-point’’ serotonin release
activation profile of HIT. Later, Greinacher and colleagues (1994) showed
that high heparin concentrations in solution release platelet factor 4 (PF4)
from PF4–heparin complexes bound covalently to a solid phase, with a cor-
responding decrease in binding of HIT antibodies to the surface. Thus, the
inhibition of platelet activation by high heparin concentrations probably re-
sults from a similar disruption of the multimolecular antigen complex on the
platelet surface.
The high sensitivity of washed platelets to activation by HIT antibodies
led to new insights into the pathogenesis of platelet activation. For example, 2
years after describing their washed platelet assay for HIT, Kelton and
coworkers (1988) reported that the platelet activation process was critically
dependent on the platelet Fc receptor. This represented a fundamental new
pathobiological mechanism in a drug-induced thrombocytopenic disorder.
B. Prospective Studies of Serologically Defined HIT

Although several prospective studies of the frequency of HIT were performed
(see Chap. 4), until the 1990s none had systematically evaluated serum or
plasma from study participants for HIT antibodies. Often the distinction
between ‘‘early’’ and ‘‘late’’ thrombocytopenia was blurred. Thus, the relative
History of HIT 11
frequency and clinical importance of immune versus nonimmune HIT were

unclear. This is illustrated by a prospective study reported by Powers and
colleagues (1979) that found HIT to be ‘‘uncommon’’ during treatment with
porcine mucosal heparin, as ‘‘only’’ 4 of 120 (3%) patients developed
thrombocytopenia, in contrast with the 26–31% frequency of thrombocyto-
penia reported for bovine lung heparin. However, 2 of these 120 patients may
have died as a result of HIT-associated thrombosis (Warkentin and Kelton,
1990), underscoring the need for a specific laboratory marker for this
immune-mediated syndrome.
In a prospective study of HIT that performed systematic testing for HIT
antibodies (Warkentin et al., 1995), the authors showed the dramatic clinical
effects of HIT. Of 665 patients participating in a clinical trial of UFH versus
low molecular weight heparin (LMWH) after orthopedic surgery, 9 patients
developed ‘‘late’’ thrombocytopenia serologically confirmed to represent
HIT. These patients had a thrombotic event rate far greater than controls.
Moreover, the spectrum of thrombosis in HIT patients included venous
thromboembolism, rather than only the classic problem of arterial thrombo-
sis. This study also showed that early postoperative thrombocytopenia
occurred frequently, but was not explained by HIT antibodies (see Chap. 4).
However, even this study did not initially capture the complete clinical
profile of HIT. This is because it defined the platelet count fall indicating
possible HIT using the ‘‘standard’’ definition of thrombocytopenia, i.e., a
platelet count fall to less than 150 109/L (Warkentin et al., 1995).
Subsequent review of the database, together with correlative analysis of the
results of systematic serological testing for HIT antibodies (performed in
most study subjects), showed that this standard definition underestimated the
number of patients who had HIT (Warkentin et al., 2003a). Rather, a
proportional fall in platelet count (50% or greater)—in relation to the
postoperative peak platelet count—provided a more accurate definition of
thrombocytopenia (applicable at least to this postoperative patient popula-
tion). This improved definition identified twice as many patients as having
had HIT in this clinical trial, without compromising diagnostic specificity.
Indeed, the study suggested that the risk of immune HIT is about 5% (16/332
= 4.8%) in postoperative orthopedic surgery patients receiving UFH for a
week or more (see Chap. 4).
VI. THE TARGET ANTIGEN OF HIT: PLATELET FACTOR

4–HEPARIN
In 1992, Jean Amiral, working in the laboratory of Dominique Meyer, re-

ported that the antigen recognized by HIT antibodies was a complex between
heparin and platelet factor 4, an endogenous platelet a-granule protein (Ami-
12 Warkentin
ral et al., 1992). This important discovery led to an explosion of basic studies
in numerous laboratories that led to further characterization of the basic
pathogenesis of HIT (see Chaps. 6–8). Amiral’s discovery also led to the de-
velopment of new assays for HIT antibodies based on enzyme immunoassay
techniques (see Chap. 11).
The antigen site(s) recognized by HIT antibodies were identified as
being on PF4, rather than on heparin itself or a compound antigen (Li et al.,
2002) (see Chaps. 6–8). This observation highlights intriguing parallels
between HIT and the antiphospholipid syndrome. This latter disorder is also
characterized by pathogenic antibodies directed against one or more proteins
that express neoepitopes when bound to certain negatively charged phospho-
lipid surfaces (see Chap. 12). The presence of neoepitopes on the ‘‘self ’’
protein, PF4, suggests the HIT can be conceptualized as a transient, drug-
induced, platelet- and coagulation-activating autoimmune disorder. Indeed,
high-titer HIT antibodies that are able to activate platelets in vitro even in the
absence of pharmacologic heparin have been associated with the onset of
thrombocytopenia and thrombosis beginning several days after heparin has
been discontinued, so-called delayed-onset HIT (Warkentin and Kelton,
2001) (see Chap. 3).
VII. TREATMENT OF THROMBOSIS COMPLICATING HIT
The treatment of HIT is discussed in Chapters 13–17. Here we will discuss

only a few vignettes relating to the initial use of selected treatments for HIT.
A. Danaparoid Sodium
In 1982, a 48-year-old vacationing American developed deep venous throm-
bosis and pulmonary embolism following a transatlantic flight to Germany.
Heparin treatment was complicated by thrombocytopenia and progression of
venous thrombosis. Professor Job Harenberg of Heidelberg University, who
had performed phase I evaluations of the experimental glycosaminoglycan
anticoagulant danaparoid, requested this agent from the manufacturer (NV
Organon, The Netherlands). The platelet count recovered and the venous
thrombosis resolved (Harenberg et al., 1983, 1997). Over the next 6 years, this
patient developed recurrent thromboembolic events, each time successfully
treated with danaparoid. This favorable experience led to a named-patient,
compassionate-release program ending in March 1997, during which time
somewhat over 750 patients were treated with this agent. Additionally, Chong
and colleagues (2001) performed the first randomized, controlled clinical trial
evaluating danaparoid (see Chap. 14).
History of HIT 13
B. Recombinant Hirudin (Lepirudin)

The medicinal leech, Hirudo medicinalis, has been used for medical purposes
for many centuries. Given the observation that the medicinal leech can
prevent clotting of blood it has ingested, crude preparations derived from
this animal were given experimentally at the beginning of the twentieth cen-
tury. However, because this treatment’s daily cost (75 Reichsmark) in 1908
was equivalent to the monthly salary of a factory worker, it was judged to be
infeasible. After World War I, Haas, at Justus-Liebig University in Giessen,
began his experiments using crude extracts of leech heads for hemodialysis.
The major complication in these animal experiments was severe bleeding. The
first human hemodialysis patients were treated by him with hirudin during
dialysis when a more purified, but still crude protein extract of leech heads
became available (Haas, 1925).
In 1956, Dr. F. Markwardt began his work to extract the active com-
ponent of the leech at the Ernst-Moritz-Arndt University, in Greifswald. Still
today, elderly peasants in the small villages around Greifswald tell stories of
how they earned their pocket money by collecting leeches for the researchers
at the nearby medical school.
The production of large amounts of hirudin by recombinant technology
allowed assessment of this direct thrombin inhibitor in clinical trials. Dr. An-
dreas Greinacher, at that time working at the Justus-Liebig University in
Giessen, first used a recombinant hirudin (lepirudin) to anticoagulate a pa-
tient who developed acute HIT following heart transplantation. After Grein-
acher’s move to Greifswald, he further assessed the use of hirudin in patients
with HIT in two clinical studies that led to the first approval of a drug for
parenteral anticoagulation of patients with HIT in both the European
Community (March 1997) and the United States (March 1998) (Greinacher
et al., 1999).
C. Warfarin-Induced Venous Limb Gangrene

A theme of this book is the central importance of increased thrombin
generation in the pathogenesis of thrombosis complicating HIT. The recog-
nition that warfarin therapy can be deleterious in some patients with HIT
illustrates the importance of uncontrolled thrombin generation in HIT.
In December 1992, in Hamilton, Canada, while receiving ancrod and
warfarin treatment for deep vein thrombosis complicating HIT, a 35-year-old
woman developed progressive venous ischemia, culminating in venous limb
gangrene. This occurred despite a supratherapeutic international normalized
ratio (INR). The following day, Kelton observed an area of skin necrosis on
the abdomen of this patient, suggesting the diagnosis of warfarin-iduced skin
14 Warkentin
necrosis. The author questioned whether the warfarin had also contributed to
the pathogenesis of the venous limb gangrene. This hypothesis was directly
tested just 2 months later when a second young woman developed severe
phlegmasia cerulea dolens of an upper limb during treatment of deep vein
thrombosis complicating HIT with ancrod and warfarin. Treatment with
vitamin K and plasma given by pheresis reversed the phlegmasia. Further
laboratory studies supported this hypothesis of a disturbance in procoagu-
lant-anticoagulant balance during treatment of HIT with warfarin (Warken-
tin et al., 1997) (see Chaps. 3, 12, and 13).
Increasingly, HIT became viewed as a syndrome characterized by
multiple prothrombotic events, including not only platelet and endothelial
cell activation, but also profound activation of coagulation pathways. This
conceptual framework provides a rationale for antithrombotic therapy that
reduces thrombin generation in patients with HIT (Warkentin et al., 1998).
VIII. TREATMENT OF ISOLATED HIT
Isolated HIT refers to HIT diagnosed on the basis of thrombocytopenia

alone, rather than because of HIT-associated thrombosis. Often, the initial
reason for administering heparin includes routine postoperative prophylaxis
or a medical indication such as acute stroke or myocardial infarction. Until
the early 2000s, the standard approach upon suspecting HIT in such patients
was discontinuation of heparin, sometimes with substitution of oral anti-
coagulants.
A. Natural History of Isolated HIT

During the mid-1990s, new data indicated a high risk for venous thrombosis
in postoperative orthopedic patients who developed HIT, particularly for
pulmonary embolism (Warkentin et al., 1995) (see Chap. 3). Thus, HIT came
to be viewed as a dramatic, albeit transient, prothrombotic state, even when
the original indication for heparin was routine antithrombotic prophylaxis.
In July 1992, the author became aware of a 68-year-old patient whose
platelet count fell from 151 to 51 109/L between days 5 and 8 following
coronary artery bypass surgery, during routine postoperative heparin anti-
thrombotic prophylaxis. The heparin was stopped, and laboratory testing
confirmed HIT. The platelet count recovered, and the patient was discharged
to home on postoperative day 12. Three days later, the patient complained of
dyspnea, and then died suddenly. Postmortem examination showed massive
History of HIT 15
pulmonary embolism. This tragic outcome prompted the question: Is mere

cessation of heparin sufficient for a patient with isolated HIT?
To address this problem, the author performed a study of the natural
history of HIT (Warkentin and Kelton, 1996). From a database of patients
with serologically proven HIT, a 62-patient cohort with isolated HIT was
identified: the cumulative 30-day thrombotic event rate was 52.8% (see Fig. 2
in Chap. 4). The rate of thrombosis was similarly high whether heparin was
simply stopped or substituted with warfarin.
Similar findings were reported later by Wallis and colleagues (1999)
from Loyola University. These investigators also found a high frequency of
subsequent thrombosis (43 of 113, or 38%) in patients with isolated HIT
managed by discontinuation of heparin. Surprisingly, a trend was observed
for the highest risk of thrombosis in those patients in whom heparin was
stopped most promptly (see Table 5 in Chap. 4).
Further evidence supporting an unfavorable natural history of un-
treated HIT was provided by a prospective cohort study (Greinacher et al.,
2000). These investigators found that the thrombotic event rate was 6.1% per
day during the mean 1.7-day interval between diagnosis of HIT (and cessation
of heparin) and initiation of lepirudin therapy. This event rate corresponded
closely to the 10% rate of thrombosis observed in the Hamilton study in the
first 48 h following diagnosis of isolated HIT (Warkentin and Kelton, 1996).
B. Argatroban
A synthetic small-molecule thrombin inhibitor derived from L-arginine, now
known as argatroban, was used in Japan during the 1980s as a treatment for
chronic arterial occlusion (Tanabe, 1986). During this time, argatroban also
underwent investigation as treatment for HIT in Japan, particularly in the
setting of hemodialysis (Matsuo et al., 1988). In 1993, exclusive rights to the
compound for the United States and Canada were acquired from Mitsubishi-
Tokyo Pharmaceuticals, Inc. by Texas Biotechnology Corporation (TBC) of
Houston. In 1995, clinical evaluation of this agent for HIT began in the
United States, using a prospective, multicenter, open-label design with
historical controls, the ARG-911 study (Lewis et al., 2001) (see Chap. 16).
Two groups of patients were studied: HIT without thrombosis (i.e., isolated
HIT) and HIT complicated by thrombosis (heparin-induced thrombocyto-
penia/thrombosis syndrome, or HITTS). Eligibility was based on clinical
suspicion of HIT, and serological confirmation of the diagnosis, therefore,
was not required. Both patient groups received the identical therapeutic-dose
regimen of argatroban (initially, 2 Ag/kg/min, then adjusted by activated
partial thromboplastin time [aPTT]). The favorable results of the ARG-911
16 Warkentin
and subsequent studies (ARG-915, ARG-915X) led to the approval of

argatroban on June 30, 2000, by the U.S. Food and Drug Administration
(FDA) as ‘‘anticoagulant for prophylaxis or treatment of thrombosis in
patients with heparin-induced thrombocytopenia.’’ Thus, for the first time
in the United States, a drug was approved for the novel indication of
prevention of thrombosis in isolated HIT. A marketing partnership between
TBC and SmithKline Beecham (now, GlaxoSmithKline) commenced in
August 1997. Marketing of argatroban began on November 13, 2000.
C. Therapeutic-Dose Anticoagulation for Isolated HIT

The approval by the FDA of identical therapeutic-dose regimens of arga-
troban for both prophylaxis and treatment of HIT highlighted the emerging
view that HIT is a high-risk prothrombotic state. This contrasted with the
earlier concept that HIT was generally benign, provided that thrombocyto-
penia was promptly recognized and heparin discontinued. Further support
for the new view included studies showing HIT to be a profound hypercoag-
ulable state (markedly elevated molecular markers of in vivo thrombin
generation) (Warkentin et al., 1997; Greinacher et al., 2000) and recognition
that many patients already have subclinical deep vein thrombosis at the time
that isolated HIT is first recognized (Tardy et al., 1999).
Indeed, therapeutic doses of an alternative anticoagulant might be
generally applicable for treatment of most patients with isolated HIT (Farner
et al., 2001) (see also Chaps. 13–16). For example, although the prophylactic-
dose regimen of lepirudin for HIT is initially lower than the therapeutic-dose
regimen (0.10 mg/kg/h, rather than 0.15 mg/kg/h, and without an initial
lepirudin bolus), subsequent dose adjustments are made using the aPTT; thus,
the eventual infusion rate approaches the one given using the therapeutic
regimen. A high success rate (91.4%) was observed using such ‘‘prophylactic’’
doses of lepirudin for isolated HIT (Farner et al., 2001).
In contrast, the prophylactic-dose regimen using danaparoid (750 U bid
or tid) may be somewhat less effective than therapeutic-dose danaparoid
(usually, 150–200 U/h) for preventing new thromboembolic complications in
acute HIT: 81.4% versus 91.6% (Farner et al., 2001) (see Chap. 14). If this
difference is real, it could be explained by greater efficacy of the therapeutic-
dose regimen, in which at least twice as much danaparoid is usually given
(3600–4800 vs. 1500–2250 U/24 h). The implication of Farner’s study is that
the approved prophylactic-dose regimen of danaparoid may not be optimal,
either when used for its approved indication in Europe (i.e., prevention of
HIT-associated thrombosis) or for the corresponding ‘‘off-label’’ use for HIT
elsewhere (Warkentin, 2001) (see Chap. 13).
History of HIT 17
IX. REDUCING THE RISK OF HIT

A. Low Molecular Weight Heparin
For over 50 years, UFH has been used in numerous clinical situations. How-
ever, UFH has several limitations, and efforts to develop potentially superior
LMWH preparations began during the 1980s. Advantages of LMWH
included better pharmacokinetics (e.g., improved bioavailability, predictable
and stable dose response obviating the need for monitoring, lower risk of
resistance to anticoagulation, longer plasma half-life) and favorable benefit-
risk ratios in experimental animals (Hirsh, 1994; Hirsh et al., 2001). Advan-
tages of UFH include its low cost, widely available laboratory monitoring,
and potential for neutralization using protamine. But the question remained:
Was the risk of HIT lower with LMWH? This was an important and relevant
question, particularly as differences in risk of HIT exist even among UFH
preparations derived from different animal sources (see Chap. 4). As dis-
cussed earlier (Sec. V.B), there is indeed evidence that LMWH has both a
lower risk of HIT antibody formation and (more importantly) a lower risk of
HIT and HIT-associated thrombosis. Table 1 provides a historical timeline of
the introduction of the LMWH enoxaparin in the United States in various
clinical situations.
B. Fondaparinux
Fondaparinux (Arixtra) is a synthetic pentasaccharide modeled after the
antithrombin-binding site of heparin. It selectively binds to antithrombin,
causing rapid and specific inhibition of factor Xa. Fondaparinux does not
bind to PF4, and as a corollary, HIT antibodies fail to recognize PF4 mixed
with fondaparinux, both in platelet activation and PF4-dependent antigen
assays.
Preliminary evidence suggests that although HIT antibody formation
occasionally occurs in association with fondaparinux use, such antibodies fail
to react in HIT assays in which fondaparinux replaces UFH or LMWH in
vitro (Warkentin et al., 2003b). Thus, this pentasaccharide anticoagulant
seems unlikely to cause an adverse effect resembling HIT. Although no pa-
tients developed HIT with either LMWH (enoxaparin) or fondaparinux in the
two orthopedic surgery trials reported, the duration of anticoagulant therapy
may have been too brief to reveal a true difference in risk of immune throm-
bocytopenia between LMWH (frequency 0.1–1.0%) and fondaparinux (an-
ticipated negligible frequency). Fondaparinux is approved in the United
States, Canada, and the European Union for antithrombotic prophylaxis in
orthopedic surgery situations (Table 1).
18 Warkentin
Table 1 U.S. Approvals of Selected Anticoagulants During the Past Decade
Date of U.S. approval
Indication Enoxaparina Other (non-LMWH)
Prophylaxis after hip March 29, 1993 Fondaparinuxb

replacement surgery December 7, 2001
Prophylaxis after knee March 9, 1995 Fondaparinuxb
replacement surgery December 7, 2001
Prophylaxis after general May 6, 1997
(abdominal) surgery
Extended prophylaxis after January 30, 1998
hip replacement surgery
Prophylaxis for unstable angina March 27, 1998
and non–q wave myocardial
infarction (given together
with aspirin)
Acute deep-vein thrombosis, December 31, 1998
with or without pulmonary
embolism, together with warfarin
Prophylaxis in medical patients November 17, 2000
at risk for deep-vein thrombosis
or pulmonary embolism
Use as an anticoagulant in patients Bivalirudin
with unstable angina undergoing December 15, 2000
percutaneous transluminal
coronary angioplasty (PTCA)
Anticoagulation in patients Argatroban
with or at risk for HIT April 3, 2002
undergoing percutaneous
coronary interventions (PCI)
a
Other LMWH preparations have been approved (at later times) for various antithrombotic
indications (not shown).
b
Wording of fondaparinux approval in U.S.: ‘‘ARIXTRA is indicated for the prophylaxis of
deep vein thrombosis, which may lead to pulmonary embolism:
. in patients undergoing hip fracture surgery
. in patients undergoing hip replacement surgery
. in patients undergoing knee replacement surgery.’’
History of HIT 19
C. Bivalirudin
The 20-amino-acid hirudin analogue bivalirudin (Angiomax, formerly, Hir-
ulog) was used 10 years ago in the United States on a compassionate use basis
for the treatment of four patients with HIT (Nand, 1993; Reid and Alving,
1994; Chamberlin et al., 1995). Currently, this agent is approved in the United
States and Canada for anticoagulation of patients with unstable angina
undergoing percutaneous transluminal coronary angioplasty (Table 1) (see
Chap. 16). It is now seeing some off-label use for the treatment of HIT
(Francis et al., 2003). This agent is also being evaluated as an anticoagulant to
permit ‘‘on-pump’’ and ‘‘off-pump’’ cardiac surgery, in patients with or
without HIT (Warkentin and Greinacher, 2003). It is possible that increasing
use of bivalirudin or other nonheparin anticoagulants, e.g., argatroban
(Table 1) or ximelagatran (oral thrombin inhibitor currently under investi-
gation), in diverse clinical situations will result in reduced numbers of patients
developing HIT.
REFERENCES
Alving BM, Shulman NR, Bell WR, Evatt BL, Tack KM. In vitro studies of heparin-
associated thrombocytopenia. Thromb Res 1977; 11:827–834.
Amiral J, Bridey F, Dreyfus M, Vissac AM, Fressinaud E, Wolf M, Meyer D. Platelet
factor 4 complexed to heparin is the target for antibodies generated in heparin-
induced thrombocytopenia [letter]. Thromb Haemost 1992; 68:95–96.
Babcock RB, Dumper CW, Scharfman WB. Heparin-induced thrombocytopenia. N
Engl J Med 1976; 295:237–241.
Barker CF, Rosato FE, Roberts B. Peripheral arterial embolism. Surg Gynecol Obstet
1966; 123:22–26.
Bell WR. Heparin-associated thrombocytopenia and thrombosis. J Lab Clin Med
1988; 111:600–605.
Bell WR, Royall RM. Heparin-associated thrombocytopenia: a comparison of three
heparin preparations. N Engl J Med 1980; 303:902–907.
Bell WR, Tomasulo PA, Alving BM, Duffy TP. Thrombocytopenia occurring during
the administration of heparin. A prospective study in 52 patients. Ann Intern
Med 1976; 85:155–160.
Benhamou AC, Gruel Y, Barsotti J, Castellani L, Marchand M, Guerois C, Leclerc
MH, Delahousse B, Griguer P, Leroy J. The white clot syndrome or heparin
associated thrombocytopenia and thrombosis (WCS or HATT). Int Angiol
1985; 4:303–310.
Best CH. Preparation of heparin, and its use in the first clinical cases. Circulation 1959;
19:79–86.
Chamberlin JR, Lewis B, Wallis D, Messmore H, Hoppensteadt D, Walenga JM,
Moran S, Fareed J, McKiernan T. Successful treatment of heparin-associated
20 Warkentin
thrombocytopenia and thrombosis using Hirulog. Can J Cardiol 1995; 11:511–

514.
Chong BH, Berndt MC. Heparin-induced thrombocytopenia. Blut 1989; 58:53–57.
Chong BH, Grace CS, Rozenberg MC. Heparin-induced thrombocytopenia: effect of
heparin platelet antibody on platelets. Br J Haematol 1981; 49:531–540.
Chong BH, Pitney WR, Castaldi PA. Heparin-induced thrombocytopenia: association
of thrombotic complications with heparin-dependent IgG antibody that induces
thromboxane synthesis and platelet aggregation. Lancet 1982; 2:1246–1249.
Chong BH, Gallus AS, Cade JF, Magnani H, Manoharan A, Oldmeadow M, Arthur
C, Rickard K, Gallo J, Lloyd J, Seshadri P, Chesterman CN. Prospective
randomised open-label comparison of danaparoid with dextran 70 in the
treatment of heparin-induced thrombocytopaenia with thrombosis: a clinical
outcome study. Thromb Haemost 2001; 86:1170–1175.
Cimo PL, Moake JL, Weinger RS, Ben-Menachem Y, Khalil KG. Heparin-induced
thrombocytopenia: association with a platelet aggregating factor and arterial
thromboses. Am J Hematol 1979; 6:125–133.
Cines DB, Kaywin P, Bina M, Tomaski A, Schreiber AD. Heparin-associated throm-
bocytopenia. N Engl J Med 1980; 303:788–795.
Crafoord C. Preliminary report on post-operative treatment with heparin as a pre-
ventive of thrombosis. Acta Chir Scand 1937; 79:407–426.
Farner B, Eichler P, Kroll H, Greinacher A. A comparison of danaparoid and lepi-
rudin in heparin-induced thrombocytopenia. Thromb Haemost 2001; 85:950–
957.
Francis JL, Drexler A, Gwyn G, Moroose R. Bivalirudin, a direct thrombin inhibitor,
in the treatment of heparin-induced thrombocytopenia [abstr]. J Thromb Hae-
most 2003; 1(suppl 1):1909.
Fratantoni JC, Pollet R, Gralnick HR. Heparin-induced thrombocytopenia:
confirmation of diagnosis with in vitro methods. Blood 1975; 45:395–401.
Green D, Harris K, Reynolds N, Roberts M, Patterson R. Heparin immune throm-
bocytopenia: evidence for a heparin-platelet complex as the antigenic deter-
minant. J Lab Clin Med 1978; 91:167–175.
Greinacher A, Pötzsch B, Amiral J, Dummel V, Eichner A, Mueller-Eckhardt C.
Heparin-associated thrombocytopenia: isolation of the antibody and character-
ization of a multimolecular PF4-heparin complex as the major antigen. Thromb
Haemost 1994; 71:247–251.
Greinacher A, Völpel H, Janssens U, Hach-Wunderle V, Kemkes-Matthes B, Eichler P,
Mueller-Velten HG, Pötzsch B, for the HIT Investigators Group. Recom-
binant hirudin (lepirudin) provides safe and effective anticoagulation in
patients with heparin-induced thrombocytopenia. Circulation 1999; 99:73–80.
Greinacher A, Eichler P, Lubenow N, Kwasny H, Luz M. Heparin-induced throm-
bocytopenia with thromboembolic complications: meta-analysis of two pro-
spective trials to assess the value of parenteral treatment with lepirudin and its
therapeutic aPTT range. Blood 2000; 96:846–851.
Haas G. Versuche der Blutauswaschung am Lebenden mit Hilfe der Dialyse. Klin
Wochenschr 1925; 4:13–14.
History of HIT 21
Harenberg J, Zimmermann R, Schwarz F, Kubler W. Treatment of heparin-induced

thrombocytopenia with thrombosis by new heparinoid [letter]. Lancet 1983;
1:986–987.
Harenberg J, Huhle G, Piazolo L, Wang LU, Heene DL. Anticoagulation in patients
with heparin-induced thrombocytopenia type II. Semin Thromb Hemost 1997;
23:189–196.
Hirsh J. From bench to bedside: history of development of LMWHs. Orthop Rev
1994; 23(suppl 1):40–46.
Hirsh J, Warkentin TE, Shaughnessy SG, Anand SS, Halperin JL, Raschke R,
Granger C, Ohman EM, Dalen JE. Heparin and low-molecular-weight hapa-
rin: mechanisms of action, pharmacokinetics, dosing, monitoring, efficacy, and
safety. Chest 2001; 119(suppl):64S–94S.
Howell WH, Holt E. Two new factors in blood coagulation—heparin and pro-
antithrombin. Am J Physiol 1918; 47:328–341.
Hussey CV, Bernhard VM, McLean MR, Fobian JE. Heparin induced platelet
aggregation: in vitro confirmation of thrombotic complications associated with
heparin therapy. Ann Clin Lab Sci 1979; 9:487–493.
Kaupp HA, Roberts B. Arterial embolization during subcutaneous heparin therapy.
Case report. J Cardiovasc Surg 1972; 13:210–212.
Kelton JG, Sheridan D, Brain H, Powers PJ, Turpie AG, Carter CJ. Clinical useful-
ness of testing for a heparin-dependent platelet-aggregating factor in patients
with suspected heparin-associated thrombocytopenia. J Lab Clin Med 1984;
103:606–612.
Kelton JG, Sheridan D, Santos A, Smith J, Steeves K, Smith C, Brown C, Murphy
WG. Heparin-induced thrombocytopenia: laboratory studies. Blood 1988; 72:
925–930.
Kinlough-Rathbone RL, Packham MA, Mustard JF. Platelet aggregation. In: Har-
ker LA, Zimmerman TS, eds. Methods in Hematology. Measurements of
Platelet Function. Edinburgh: Churchill Livingstone, 1983:64–91.
Klein HG, Bell WR. Disseminated intravascular coagulation during heparin therapy.
Ann Intern Med 1974; 80:477–481.
Lewis BE, Wallis DE, Berkowitz SD, Matthai WH, Fareed J, Walenga JM,
Bartholomew J, Sham R, Lerner RG, Zeigler ZR, Rustagi PK, Jang IK, Rifkin
SD, Moran J, Hursting MJ, Kelton JG, for the ARG-911 Study Investigators.
Circulation 2001; 103:1838–1843.
Li ZQ, Liu W, Park KS, Sachais BS, Arepally GM, Cines AB, Poncz M. Defining a
second epitope for heparin-induced thrombocytopenia/thrombosis antibodies
using KKO, a murine HIT-like monoclonal antibody. Blood 2002; 99:1230–
1236.
Mason EC. Blood coagulation. The production and prevention of experimental
thrombosis and pulmonary embolism. Surg Gynecol Obstet 1924; 39:421–428.
Matsuo T, Chikahira Y, Yamada T, Nakao K, Ueshima S, Matsuo O. Effect of syn-
thetic thrombin inhibitor (MD805) as an alternative drug on heparin induced
thrombocytopenia during hemodialysis. Thromb Res 1988; 52:165–171.
McLean J. The thromboplastic action of cephalin. Am J Physiol 1916; 41:250–257.
22 Warkentin
Nand S. Hirudin therapy for heparin-associated thrombocytopenia and deep venous

thrombosis. Am J Hematol 1993; 43:310–311.
Natelson EA, Lynch EC, Alfrey CP Jr, Gross JB. Heparin-induced thrombocyto-
penia. An unexpected response to treatment of consumption coagulopathy. Ann
Intern Med 1969; 71:1121–1125.
Nelson JC, Lerner RG, Goldstein R, Cagin NA. Heparin-induced thrombocytopenia.
Arch Intern Med 1978; 138:548–552.
Powers PJ, Cuthbert D, Hirsh J. Thrombocytopenia found uncommonly during hep-
arin therapy. JAMA 1979; 241:2396–2397.
Reid T III, Alving BM. HirulogR therapy for heparin-associated thrombocytopenia
and deep venous thrombosis. Am J Hematol 1994; 43:352–353.
botic and hemorrhagic manifestations. Surg Gynecol Obstet 1973; 136:409–
416.
Rhodes GR, Dixon RH, Silver D. Heparin induced thrombocytopenia: eight cases
with thrombotic-hemorrhagic complications. Ann Surg 1977; 186:752–758.
Roberts B, Rosato FE, Rosato EF. Heparin—a cause of arterial emboli? Surgery 1964;
55:803–808.
Sheridan D, Carter C, Kelton JG. A diagnostic test for heparin-induced thrombo-
cytopenia. Blood 1986; 67:27–30.
Shionoya T. Studies on experimental extracorporeal thrombosis. III. Effects of certain
anticoagulants (heparin and hirudin) on extracorporeal thrombosis and on the
mechanism of thrombus formation. J Exp Med 1927; 46:19–26.
Stanton PE Jr, Evans JR, Lefemine AA, Vo RN, Rannick GA, Morgan CV Jr, Hinton
JP, Read M. White clot syndrome. South Med J 1988; 81:616–620.
Tanabe T. Clinical results of MD-805, antithrombin agent, on chronic arterial
occlusion. J Clin Ther Med 1986; 2:1645.
Tardy B, Tardy-Poncet B, Fournel P, Venet C, Jospe R, Dacosta A. Lower limb veins
should be systematically explored in patients with isolated heparin-induced
thrombocytopenia [letter]. Thromb Haemost 1999; 82:1199–1200.
Towne JB, Bernhard VM, Hussey C, Garancis JC. White clot syndrome. Peripheral
vascular complications of heparin therapy. Arch Surg 1979; 114:372–377.
Trowbridge AA, Caraveo J, Green JB III, Amaral B, Stone MJ. Heparin-related
immune thrombocytopenia. Studies of antibody-heparin specificity. Am J Med
1978; 65:277–283.
Wahl TO, Lipschitz DA, Stechschulte DJ. Thrombocytopenia associated with anti-
heparin antibody. JAMA 1978; 240:2560–2562.
Wallis DE, Workman DL, Lewis BE, Steen L, Pifarre R, Moran JF. Failure of early
heparin cessation as treatment for heparin-induced thrombocytopenia. Am J
Med 1999; 106:629–635.
Warkentin TE. Heparin-induced thrombocytopenia: yet another treatment paradox?
Thromb Haemost 2001; 85:947–949.
Warkentin TE, Greinacher A. Heparin-induced thrombocytopenia and cardiac sur-
gery. Ann Thorac Surg 2003; 76:638–648. [Corrected article to be reprinted in its
entirety in Ann Thorac Surg 2003, Dec. In press].
History of HIT 23
Warkentin TE, Kelton JG. Heparin and platelets. Hematol Oncol Clin North Am
1990; 4:243–264.
Warkentin TE, Kelton JG. A 14-year study of heparin-induced thrombocytopenia.
Am J Med 1996; 101:502–507.
Warkentin TE, Kelton JG. Delayed-onset heparin-induced thrombocytopenia and
thrombosis. Ann Intern Med 2001; 135:502–506.
Warkentin TE, Levine MN, Hirsh J, Horsewood P, Roberts RS, Gent M, Kelton JG.
Heparin-induced thrombocytopenia in patients treated with low-molecular-
weight heparin or unfractionated heparin. N Engl J Med 1995; 332:1330–1335.
Warkentin TE, Elavathil LJ, Hayward CPM, Johnston MA, Russett JI, Kelton JG.
The pathogenesis of venous limb gangrene associated with heparin-induced
thrombocytopenia. Ann Intern Med 1997; 127:804–812.
Warkentin TE, Chong BH, Greinacher A. Heparin-induced thrombocytopenia: to-
wards consensus. Thromb Haemost 1998; 79:1–7.
Warkentin TE, Roberts RS, Hirsh J, Kelton JG. An improved definition of immune
heparin-induced thrombocytopenia in postoperative orthopedic patients. Arch
Intern Med 2003a. In press.
Warkentin TE, Cook RJ, Marder VJ, Kelton JG. Comparison of heparin-induced
thrombocytopenia antibody (HIT-Ab) generation and in vitro cross-reactivity
after elective hip or knee replacement surgery in patients receiving antithrom-
botic prophylaxis with fondaparinux or enoxaparin [abstr]. Blood 2003b. In
press.
Weismann RE, Tobin RW. Arterial embolism occurring during systemic heparin
therapy. Arch Surg 1958; 76:219–227.
2
Differential Diagnosis of Acute
Thrombocytopenia
Volker Kiefel
University of Rostock, Rostock, Germany
I. INTRODUCTION
Platelet count is affected by the rate of platelet production, the platelet life
span, and the distribution of platelets among different compartments (Wint-
robe et al., 1981). These three aspects of platelet kinetics have been extensively
studied with platelets radiolabeled with 51Cr (Aster and Jandl, 1964) or 111In
(Heaton et al., 1979).
Platelet production is equivalent to platelet turnover in a ‘‘steady state’’
and can be estimated by determining platelet mean life span and count: ap-
proximately 44 109/L per day of platelets are produced by normal persons
(Branehög et al., 1974). Platelet turnover is decreased in certain marrow dis-
orders (e.g., aplastic anemia and hereditary thrombocytopenia) and often as
a result of cytotoxic drugs used for the therapy of malignant disease. Plate-
let distribution is mainly influenced by spleen size. Normally, approximately
30% of platelets are sequestered in the spleen. In patients with splenomegaly,
this fraction can increase to 90%, and mild to moderate thrombocytopenia
can result. Conversely, in splenectomized subjects, more than 90% of the total
platelet mass is circulating.
This chapter will focus on the differential diagnosis of thrombocy-
topenic states that are characterized by a shortened platelet survival that
is mediated by immune mechanisms. Pathological conditions of enhanced
platelet destruction caused by nonimmune mechanisms will also be briefly dis-
cussed. Autoantibodies responsible for autoimmune thrombocytopenic pur-
pura (AITP), acquired platelet dysfunction, cyclic thrombocytopenia, and
25
26 Kiefel
Table 1 Antigens on Platelet Glycoproteins, Determined Under Reducing

Conditions
Antigenic
Number of determinants
Glycoproteins subunits Molecular weight fora
GP IIb/IIIa, 3 IIba 125,000 allo, iso, auto,

CD41/CD61 IIh 22,000 drug
IIIa 105,000
GP Ib/IX/V, CD42 4 Iba 135,000 allo, auto, drug
Ibh 22,000
IX 17,000
V 85,000
GP Ia/IIa, VLA-2, 2 Ia 165,500 allo, auto
CD49b/CD29 IIa 130,000
GP IV, GP IIIb, CD36 1 88,000 iso, auto
CD109 1 175,000 allo
HLA Class I 2 heavy chain 45,000 allo
h2M 12,000
a
allo, alloantibody; auto, autoantibody; drug, drug-dependent antibody; iso, isoantibody; VLA,
very late activation antigen; h2M, h2-microglobulin.
some forms of drug-induced immune thrombocytopenia (e.g., gold-induced

thrombocytopenia), react with monomorphic epitopes on platelet glycopro-
teins present on platelets of all healthy individuals. Alloantibodies recognize
polymorphic, genetically determined epitopes on platelet glycoproteins. They
are a specific finding in sera of patients with post-transfusion purpura. Other
alloimmune thrombocytopenic disorders will not be discussed here. Drug-
dependent antibodies can mediate platelet destruction by recognizing mono-
morphic determinants on platelet glycoproteins in the presence of the
causative drug. Table 1 gives an overview of the platelet glycoproteins known
to carry antigenic determinants.
II. AUTOIMMUNE THROMBOCYTOPENIC PURPURA

A. Pathogenesis
Thrombocytopenia in AITP results from rapid clearance of platelets sensi-
tized with autoantibodies, usually of the IgG class, reacting with glycopro-
teins on the platelet surface (reviewed in Kiefel et al., 1992). Van Leeuwen and
Differential Diagnosis of Acute Thrombocytopenia 27
coworkers (1982a) were the first to identify the glycoprotein (GP) IIb/IIIa
complex as the major target antigen recognized by platelet autoantibodies in
AITP, a finding confirmed by others. Epitopes of some antibodies have been
assigned to GP IIIa (Beardsley et al., 1984) or GP IIb (Tomiyama et al., 1987).
The other major target antigen is the platelet GP Ib/IX complex (Woods et al.,
1984; Kiefel et al., 1991). Other ‘‘rare’’ autoantibodies have been shown to
react with GP Ia/IIa (Castaldi et al., 1989), GP V (Mayer and Beardsley,
1996), GMP-140 (CD62), and the thrombopoietin receptor (Malloy et al.,
1995).
B. Clinical Manifestations
In its various manifestations AITP is a relatively common disorder. Different
forms are usually distinguished by clinical criteria.
Acute Postinfectious Autoimmune Thrombocytopenia

Acute AITP usually affects children younger than 10 years of age, boys and
girls equally. The thrombocytopenia typically begins suddenly 10 days to 3
weeks after an acute viral infection (Waters, 1992). The incidence of AITP in
childhood is estimated at 1:25,000 children per year. Bleeding symptoms may
be severe in patients with platelet counts of fewer than 10 109/L. However,
most patients recover within 3 months, even if they have not been treated with
corticosteroids or intravenous IgG (ivIgG). By convention, AITP is referred
to as ‘‘chronic’’ if thrombocytopenia persists for more than 6 months. How-
ever, about 90% of children with AITP develop the acute self-limited form.
Chronic AITP
Chronic AITP is the most common form of immune thrombocytopenia in
adults, with women more frequently affected (3:1). The annual incidence has
been estimated at approximately 1:31,000 in adults (Frederiksen and
Schmidt, 1999). It may occur as ‘‘idiopathic’’ immune thrombocytopenia or
as ‘‘secondary AITP’’: that is, observed together with other immunological
diseases, such as systemic lupus erythematosus (Waters, 1992), rheumatoid
arthritis (Hegde et al., 1983), Crohn’s disease (Kosmo et al., 1986), primary
biliary cirrhosis (Panzer et al., 1990); malignant diseases, such as lymphopro-
liferative disorders (Hegde et al., 1983) and solid tumors (Mueller-Eckhardt et
al., 1983a); infectious diseases including viral hepatitis (Pawlotsky et al., 1995;
Ibarra et al., 1986), human immunodeficiency virus (HIV) infection (Van der
Lelie et al., 1987; Walsh et al., 1985); after bone marrow transplantation
(Benda et al., 1989); and in diseases of unknown origin, such as sarcoidosis
28 Kiefel
(Henke et al., 1986). AITP with concomitant warm-type autoimmune hemo-

lytic anemia is known as Evans’s syndrome (Waters, 1992).
Onset of hemorrhagic diathesis is often insidious in chronic AITP. Most
commonly, patients present with hemorrhagic manifestations of the skin, and
in more severe forms, mucosal bleeding (‘‘wet purpura’’) with bloody blisters
in the mouth, oozing from gums, epistaxis, melena, and menorrhagia
(Crosby, 1975). Concerns about life-threatening intracranial hemorrhage
are an important reason for therapy in AITP. The clinical course in AITP
is difficult to predict, and it is estimated that in 10% of patients in childhood
with acute AITP it will become chronic. The likelihood of chronic thrombo-
cytopenia is much higher in adults. Many patients experience remissions and
relapses that occur spontaneously or following infections. Chronic AITP may
be an early manifestation of systemic lupus erythematosus.
C. Diagnosis
Clinical diagnosis of idiopathic AITP is based on its typical clinical picture
and the exclusion of other causes for thrombocytopenia: isolated thrombo-
cytopenia without evidence of impaired thrombocytopoiesis (normal num-
bers of megakaryocytes in the bone marrow). Moreover, spleen size is normal
and no other conditions known to enhance platelet clearance are found, such
as disseminated intravascular coagulation (DIC), thrombotic thrombocyto-
penia purpura (TTP), or large hemangiomas characteristic of the Kasa-
bach-Merritt syndrome. Familial thrombocytopenia implies a nonimmune
pathogenesis (Greinacher and Mueller-Eckhardt, 1994), for most patients
with hereditary thrombocytopenia do not have enhanced platelet destruction
(Najean and Lecompte, 1990).
It may be difficult to diagnose AITP on clinical grounds in patients with
malignancy, because immune thrombocytopenia may coexist with spleno-
megaly or neoplastic marrow infiltration. Moreover, not all ‘‘platelet anti-
body tests’’ are diagnostically helpful (George et al., 1996). In particular,
‘‘platelet-associated immunoglobulin G’’ (PAIgG) assays developed as
‘‘platelet Coombs tests’’ focused on technical considerations, rather than
on clinical usefulness (Dixon et al., 1975; Mueller-Eckhardt et al., 1978;
Hegde et al., 1985; Follea et al., 1982; Morse et al., 1981; Kunicki et al., 1982;
Leporrier et al., 1979; McMillan et al., 1979; Court and LoBuglio, 1986;
Kiefel et al., 1987a). However, it appears that platelet-bound IgG in these
assays bears little or no direct relation to immune-mediated platelet destruc-
tion (Mueller-Eckhardt et al., 1982; Kiefel et al., 1986). Rather, PAIgG re-
flects IgG stored in the platelet a-granules (George, 1990). However, with the
advent of platelet glycoprotein-specific assays (McMillan et al., 1987; Kiefel et
al., 1987b; Kiefel, 1992), it is evident that IgG bound to GPs IIb/IIIa or Ib/IX
is specific for AITP.
D. Therapeutic Considerations
Therapy should be based on the degree of hemorrhagic diathesis observed.
Thus, patients with ‘‘wet purpura’’ are generally treated more aggressively
because they are considered to be at greatest risk for bleeding. In children with
severe acute AITP, it is important to reduce physical activity. Drugs that
inhibit platelet function, such as acetylsalicylic acid, should be avoided. If
therapy is necessary, prednisone at an initial dose of 1 mg/kg body weight
should be given for a limited period. The ivIgG preparations are very effective
in childhood AITP (Imbach et al., 1981, 1985), with doses of 2 g/kg body
weight (0.4 g/kg daily for 5 days or 1 g/kg for 2 days) usually effective for a
limited time in adult patients. Alternatively, blockade of immune phagocy-
tosis with sensitized autologous red blood cells has been proposed (Salama et
al., 1983; Becker et al., 1986): IgG anti-D (Rh0) may be given intravenously to
rhesus (D)-positive patients with AITP. Two doses of approximately 20 Ag/kg
body weight result in an increase in platelets in most patients. Although the
therapeutic effect of anti-D appears less rapidly than with high-dose ivIgG, it
is often more sustained. Immunosuppressive therapy with azathioprine
(Quiquandon et al., 1990) alone or together with corticosteroids may be at-
tempted in patients refractory to other forms of treatment. Refractory pa-
tients with dangerous bleeding complications have been successfully treated
with cyclophosphamide (Reiner et al., 1995). One of the most effective thera-
peutic measures in AITP is splenectomy. It should not be performed in
children younger than 6 years of age and not in the first 6 months of initially
diagnosed acute AITP. It results in partial or complete remissions in 50–80%
of patients (Shulman and Jordan, 1987). Possibly, patients with predomi-
nantly splenic sequestration of platelets have a higher chance of remission
after splenectomy (Najean et al., 1997). A good response to ivIgG may indi-
cate a higher remission rate after splenectomy (Law et al., 1997). Therapy of
AITP has been reviewed (Berchtold and McMillan, 1989; Waters, 1992; Eden
and Lilleyman, 1992; George et al., 1996).
E. Other Manifestations of Autoimmunity Against Platelets

Acquired Antibody-Mediated Platelet Dysfunction
In ‘‘typical’’ AITP, platelet autoantibodies induce thrombocytopenia with
relatively moderate bleeding tendency that often is less pronounced than that
30 Kiefel
observed with similar platelet counts caused by impaired thrombocytopoiesis

(Waters, 1992). Therefore, it has been concluded that platelet autoantibodies
normally do not, or only slightly, affect platelet function. In 1986, the first case
of a patient with normal platelet counts, but an IgG1 autoantibody-induced
platelet dysfunction resembling Glanzmann’s thrombasthenia, was described
(Niessner et al., 1986). Interestingly, patients with antibody-mediated platelet
dysfunction may develop immune thrombocytopenia (Kubota et al., 1989)
and vice versa (Meyer et al., 1991).
Cyclic Thrombocytopenia
Cyclic thrombocytopenia, which predominantly occurs in women, is charac-
terized by rhythmic fluctuations of platelet counts. These fluctuations are in
phase with the menstrual cycle, lowest platelets counts being observed during
menses. Normal to high platelet counts are observed at midcycle (Tomer et
al., 1989). In many patients with this condition, thrombocytopenia is the re-
sult of accelerated platelet clearance, as determined with 111In-labeled plate-
lets. In two of the three patients described by Tomer, autoantibodies with GP
Ib/IX specificity were identified during both the thrombocytopenic period
and the period with normal platelet counts. These authors correlated platelets
counts with changing densities of the Fcg-receptor on the patients’ autolo-
gous monocytes. In one patient studied by Menitove, IgG anti-GP IIb/IIIa
was found (Menitove et al., 1989). In another case, an IgM anti-GP IIb/IIIa
has been reported (Kosugi et al., 1994). Data from another group suggest that
the pathophysiology underlying the clinical picture of cyclic thrombocyto-
penia may be heterogeneous: an autoimmune form with cyclic changes in
platelet destruction and a distinct condition with cyclic changes in thrombo-
cytopoiesis (Nagasawa et al., 1995).
Onyalai
An exceptionally severe variant of AITP, onyalai, is observed in some black
populations in southern Africa. Whites living in the same regions do not
appear to suffer from this disease. In a series of 103 patients (Hesseling, 1987),
all patients presented with hemorrhagic bullae of the mucous membranes of
the oropharynx. Six died, four of cerebral hemorrhage and two of hemor-
rhagic shock. Clinical diagnosis is based on the criteria of AITP and, in
addition, to the presence of hemorrhagic bullae (Hesseling, 1992). Antibodies
against GP IIb/IIIa—often of the IgM class—have been implicated. The bone
marrow contains normal counts of megakaryocytes, but patients may be
anemic at presentation owing to blood loss. Therapeutic options are discussed
elsewhere (Hesseling, 1992).
III. POSTTRANSFUSION PURPURA

A. Pathogenesis
Posttransfusion purpura (PTP) is a rare, but severe, transfusion reaction.
Typically, 6–8 days after transfusion of whole blood, packed red blood cells,
or platelet concentrates, patients experience an abrupt platelet count drop.
The patient’s serum almost invariably contains a high-titered platelet-specific
alloantibody, typically reacting with a determinant on the platelet GP IIb/IIIa
complex. In most cases, it reacts with HPA-1a, but other specificities have
been observed (Table 2). The duration of this immune-mediated thrombocy-
topenia is normally limited to 5–60 days. Although the patient’s autologous
platelets do not carry the corresponding alloantigen, nevertheless, they
undergo enhanced destruction. The alloimmune response is anamnestic,
because nearly all patients with PTP have a documented prior exposure to
the platelet alloantigen during previous pregnancy or transfusion.
Although PTP was first described in 1961, its pathogenesis remains de-
bated even today. It has been suggested that circulating HPA-1a antigen from
the platelets of the immunizing blood product persists, and thus antigen–
antibody complexes are adsorbed onto the autologous platelets (Shulman et
al., 1961). Others have proposed that during the secondary, anamnestic im-
munization precipitating PTP, a second autoreactive antibody arises, causing
Table 2 Platelet Specific Alloantigens Implicated in Cases of PTP
Original Percent
Alloantigen designation positivea Localization Refs.
HPA-1a PlA1, Zwa 97.5 GP IIIa Shulman et al., 1961

HPA-1b PlA2, Zwb 30.8 GP IIIa Taaning et al., 1985;
Chapman et al., 1987
HPA-2b Koa, Siba 11.8 GP Iba Lucas et al., 1998
HPA-3a Baka, Leka 86.1 GP IIb Boizard and Wautier, 1984;
Keimowitz et al., 1986
HPA-3b Bakb 62.9 GP IIb Kickler et al., 1988;
Kiefel et al., 1989
HPA-4a Yukb, Pena >99.9 GP IIIa Simon et al., 1988
>99.7b
HPA-5b Bra, Zava 20.6 GP Ia Christie et al., 1991
HPA-15b Gova 60.2 CD109 Kelton et al., 1990;
Berry et al., 2000
a
Alloantigen frequency observed in Germany (Kiefel et al., 1993).
b
Alloantigen frequency observed in Japan.
32 Kiefel
the patient’s severe thrombocytopenia. It is our experience that alloantibodies

observed in the early anamnestic response in PTP are always high-titered.
Moreover, they can be eluted from the autologous (alloantigen-negative)
platelets in most cases studied (Kroll et al., 1993). Therefore, it can also be
hypothesized that this pseudospecific alloantibody for a limited time cross-
reacts with a structurally related epitope on the patient’s autologous platelets.
B. Clinical Picture
The clinical course of PTP has been summarized by the European PTP study
group (Mueller-Eckhardt et al., 1991). Women were predominantly affected
(99 out of 104). Mean age was 58.4 years. In 28 of 51 patients, marked febrile
transfusion reactions were observed following the transfusion that precipi-
tated PTP. The interval between transfusion and onset of severe thrombocy-
topenia was generally 6–10 days. In 68 out of 84 patients, the initial platelet
count was fewer than 10 109/L. Bleeding persisted for 3–37 days (mean, 10.2
days). Most patients required treatment. Fatal bleeding was not rare, occur-
ring in 7 of 75 patients (Shulman and Jordan, 1987) and in 2 of 38 patients
(Kroll et al., 1993) in two studies.
C. Diagnosis
The diagnosis of PTP should be considered in all patients with a sudden drop
in platelet count to fewer than 10 109/L. Thus, PTP and HIT may occur in
similar clinical situations, (i.e., about 1 week after surgery). In PTP, how-
ever, thrombocytopenia is more pronounced and associated with bleeding
(Lubenow et al., 2000), in contrast to the absence of petechiae and presence of
thromboembolic complications characteristic of HIT (see Chap. 3).
Moreover, PTP only occurs following recent transfusion. PTP has
occurred in association with delayed hemolytic transfusion reaction (Chap-
man et al., 1987; Maslanka and Zupanska, 1993). In a single case, PTP was
observed concurrently with drug-dependent immune hemolytic anemia
(Mueller-Eckhardt et al., 1987). Diagnosis is confirmed by detection of a
platelet-specific alloantibody against an epitope on platelet GP IIb/IIIa,
usually anti-HPA-1a. Platelets of the patient are always negative for the
corresponding antigen. Eluates prepared from patients’ autologous platelets
usually contain anti-HPA-1a (Kroll et al., 1993).
D. Therapeutic Considerations
The efficacy of corticosteroids is uncertain. In contrast, high-dose ivIgG
(Mueller-Eckhardt et al., 1983b) is clearly effective in most (Berney et al.,
1985; Chong et al., 1986; Mueller-Eckhardt and Kiefel, 1988) but not all cases
(Kroll et al., 1993). Thus, ivIgG is the treatment of choice for PTP. Platelet
transfusions are ineffective, even if platelets from HPA-1a-negative donors
are given (Gerstner et al., 1979). As the bleeding tendency is often pro-
nounced, immediate therapy following clinical diagnosis is mandatory.
IV. PASSIVE IMMUNE THROMBOCYTOPENIA
Acute thrombocytopenia in humans resulting from experimental transfer of

platelet autoantibodies has been observed in studies on the pathogenesis of
AITP (Harrington et al., 1951). However, inadvertent transfer of platelet
autoantibodies has not been recognized as a problem in clinical transfusion
practice. In contrast, platelet alloantibodies with HPA-1a (Moilan et al.,
1985; Ballem et al., 1987; Scott et al., 1988; Brunner-Bollinger et al., 1997) and
HPA-5b (Warkentin et al., 1992) specificities can cause abrupt-onset immune
thrombocytopenia. In three cases the antibody was transfused with plasma, in
one case by whole blood, and once by a red blood cell concentrate. The con-
dition may be accompanied by a febrile transfusion reaction. It is important
to investigate these cases to identify and exclude donors with ‘‘harmful’’ plate-
let alloantibodies.
V. DRUG-INDUCED IMMUNE THROMBOCYTOPENIA

A. Pathogenesis
Drugs can induce various immune-mediated cytopenias, such as immune
hemolytic anemia, neutropenia, and drug-induced immune thrombocytope-
nia (DIT) (Salama and Mueller-Eckhardt, 1992). Different mechanisms are
involved in causing DIT, with drug-dependent antibodies being most exten-
sively investigated. Drug-dependent antibodies typically react with mono-
morphic epitopes on virtually the same platelet glycoproteins recognized by
platelet-specific autoantibodies: GP Ib/IX (Kunicki et al., 1978; van Leeuwen
et al., 1982b; Berndt et al., 1985), GP V (Stricker and Shuman, 1986) or GP
IIb/IIIa (Christie et al., 1987, 1993; Pfueller et al., 1990; Visentin et al., 1991).
They bind to the platelet target antigen only in the presence of the causative
drug: thus, if the drug is removed from the buffer during in vitro studies, the
antibody detaches (see Fig. 2). The antibody specifically recognizes the
antigen on platelet glycoprotein by the Fab fragment (Christie et al., 1985).
The precise role of the drug in antigen–antibody binding is unclear. Experi-
mental studies suggest that in some individuals, the drug induces conforma-
34 Kiefel
tional change either in the antibody or in the platelet glycoprotein to induce

complementarity required for antibody binding (Shulman and Jordan, 1987).
The drugs most commonly implicated in DIT are quinidine and quinine.
Many substances have been suspected to induce this condition, but only a few
have been documented by appropriate laboratory analysis (Aster and
George, 1990; George et al., 1998). Table 3 lists drugs implicated in the
author’s laboratory. Sometimes a metabolite, rather than the drug itself,
mediates drug-dependent antibody binding to the platelet surface (Eisner and
Shahidi, 1972; Eisner and Kasper, 1972; Kiefel et al., 1987c; Meyer et al.,
1993). Similar observations have been made for drug induced immune
hemolysis (Salama and Mueller-Eckhardt, 1985).
B. Clinical Picture and Therapy

Thrombocytopenia begins at least 7 days after the first exposure to the drug
(Shulman and Jordan, 1987). Drug-dependent antibodies can persist; thus,
reexposure to the drug can cause a sudden drop in platelet count, and is not
recommended as a diagnostic maneuver. DIT often causes severe thrombo-
cytopenia and bleeding.
The most important measure is to discontinue the offending drug. This
is normally followed by a rise of platelet count. If bleeding symptoms are life-
threatening, transfusion of large doses of platelets together with ivIgG should
be considered.
Table 3 Drugs Inducing DIT
Drug n Comments
Quinidine 18
Quinine 3
Quinidine + quinine 5
Trimethoprim–sulfamethoxazole 6 1 metabolite—specific
ddAb (sulfamethoxazole)
Rifampicin (rifampin) 4
Nomifensine 2
Paracetamol (acetaminophen) 1 1 metabolite—specific ddAb
Carbamazepine 4
Diclofenac 5
Ibuprofen 3 1 metabolite—specific ddAb
Ranitidine 1
Vancomycin 1
Source: Institute for Clinical Immunology and Transfusion Medicine, University of Giessen.
C. Immune Thrombocytopenia Induced by GP IIb/IIIa

Inhibitors
Specific inhibitors of GP IIb/IIIa are used with increasing frequency during
cardiovascular interventions. Acute thrombocytopenia is a well-documented
side effect of abciximab, a chimeric antibody inhibiting GP IIb/IIIa receptor
function. In most cases, antibodies reacting with GP IIb/IIIa-abciximab
complexes seem to be involved, but many issues of the pathogenesis remain
unresolved. Typically, abciximab-induced thrombocytopenia occurs within
the first 24 hours, most frequently at 4 hours. About 0.7% of patients are
affected after first exposure to the drug (Jubelirer et al., 1999) and 4.6% of
patients after repeated administration of abciximab (Tcheng et al., 2001). The
rapid onset of thrombocytopenia even in those patients receiving the drug for
the first time suggests that preformed antibodies are involved. A complicating
issue is that pseudothrombocytopenia also can develop after abciximab
administration; this was reported by Sane et al. (2000) to occur in 2.1% of
patients (0.6% of placebo controls), whereas Schell et al. (2002) found
pseudothrombocytopenia in 27% of patients treated with abciximab. Immu-
nological tests for antibodies reacting with abciximab-coated platelets are of
little predictive value, as such antibodies are found in sera of more than 70%
of healthy normal individuals (Curtis et al., 2002; Kiefel, unpublished
observations). Whether these antibodies react with the abciximab molecule
itself or against structural changes of GP IIb/IIIa complex after binding of the
drug is not clear (Curtis et al., 2002).
Acute immune-mediated thrombocytopenia has also been observed
after treatment with the GP IIb/IIIa inhibitors tirofiban and eptifibatide
(Bougie et al., 2002). Sera from these patients contain naturally occurring
antibodies that react against GP IIb/IIIa in the presence of the drug.
However, they showed no reactivity with abciximab-treated platelets. Similar
antibodies have been identified in patients who developed thrombocytopenia
during treatment with the oral GP IIb/IIIa-receptor antagonists xemilofiban
and orbofiban (Brassard et al., 2002). However, in these patients the median
time of treatment before onset of thrombocytopenia was 11 days.
Therapy of thrombocytopenia induced by GP IIb/IIIa inhibitors is best
established for abciximab. In contrast with other forms of immune throm-
bocytopenia, abciximab-induced immune thrombocytopenia is most effec-
tively treated with platelet transfusions (Kereiakes et al., 1996). However,
only in few cases with severe thrombocytopenia does bleeding occur and
require therapy. In every suspected case, pseudothrombocytopenia should
be ruled out first by review of the blood smear to avoid inappropriate dis-
continuation of the GP IIb/IIIa inhibitor or giving an unnecessary platelet
transfusion.
36 Kiefel
D. Drug-Induced Autoimmune Thrombocytopenia

Some drugs cause autoimmune cytopenias by inducing autoantibodies indis-
tinguishable from those encountered in ‘‘idiopathic’’ autoimmune cytopenia.
A well-known example is a-methyldopa, which in 10–36% of patients induces
formation of red blood cell autoantibodies (Petz and Garratty, 1980).
However, only 1% of patients develop clinical hemolysis. Similarly, autoim-
mune thrombocytopenia has been observed during the course of gold therapy
(von dem Borne et al., 1986). The antibodies found in these patients do not
require the presence of the drug for binding to platelets in vitro (i.e., they
resemble autoantibodies).
VI. CONSUMPTIVE THROMBOHEMORRHAGIC DISORDERS

A. Pathogenesis
A heterogeneous group of events, including sepsis, malignancies, trauma,
obstetric complications, snake venoms, and hemolytic transfusion reactions,
can be complicated by a systemic syndrome characterized by dysregulated
thrombin formation, leading to activation and consumption of coagulation
factors and resulting in the formation of intravascular fibrin thrombi.
Secondary plasmin generation helps to lyse the fibrin formed. Additionally,
the vessel wall and platelets are usually involved in this pathological process of
‘‘disseminated intravascular coagulation’’ (DIC). Indeed, thrombocytopenia
is a common clinical manifestation of DIC (Mammen, 1998). Both bleeding
and widespread thrombotic microvascular occlusion, leading to organ failure,
can result from DIC.
B. Clinical Disorders
Septicemia
Disseminated intravascular coagulation can complicate infections, especially
with gram-negative bacteria (Marder et al., 1994; Mammen, 1998). It has been
suggested that thrombocytopenia in septic patients is the consequence of
immune-mediated platelet damage, based on the observation of elevated
PAIgG levels (Kelton et al., 1979). However, this does not prove an autoim-
mune basis for the thrombocytopenia (Shulman and Reis, 1994), for PAIgG is
often also elevated in thrombocytopenia of nonimmune origin.
The pathogenesis of DIC in sepsis is multifactorial and includes direct
endothelial damage and platelet activation by endotoxins, resulting in expo-
sure of procoagulant material. In addition, the cytokines interleukin-1 and
tumor necrosis factor increase tissue factor activity, thereby shifting the bal-
ance toward a prothrombotic tendency (Mammen, 1998). Plasminogen ac-

tivator inhibitor-1 (PAI-1) blocks plasmin generation during the course
of sepsis, thereby contributing to fibrin deposition in the microcirculation
(Müller-Berghaus, 1987).
Malignant Disease
About 9–15% of patients with cancer have DIC at some point during their
disease (Pasquini et al., 1995). Overt bleeding is uncommon; rather, recur-
rent thromboembolism is characteristic, an entity known as Trousseau’s syn-
drome. An exception: acute promyelocytic leukemia is often accompanied
by a severe DIC and bleeding, often induced or worsened by chemotherapy
(Marder et al., 1994).
Other Conditions
Disseminated intravascular coagulation occurs in obstetric situations char-
acterized by release of thrombogenic material [e.g., the retained dead fetus
syndrome (Marder et al., 1994; Baglin, 1996), amniotic fluid embolism, or
placental separation]. Bites of certain snakes may cause hypofibrinogenemia
induced by enzymes that clot fibrinogen or directly activate platelets. Severe
hemolytic transfusion reactions can cause DIC, especially in association with
red cell antibodies, causing intravascular complement-mediated hemolysis
(e.g., ABO-incompatible transfusion). DIC seems to be aggravated by
complement-mediated damage of endothelial cells. Whether red cell lysis
alone (not mediated by complement) is able to induce DIC in humans remains
unclear (Mollison et al., 1993; Baglin, 1996). Other conditions associated with
DIC are trauma and localized processes in which activation of coagulation
occurs within giant hemangiomas (Kasabach-Merritt syndrome) or aortic
aneurysms.
C. Diagnosis
‘‘Global’’ coagulation tests, such as prothrombin and activated partial
thromboplastin times, are usually prolonged; fibrinogen concentrations are
often reduced. However, these parameters can be normal in DIC. Fibrin
degradation products mirror the action of plasmin on fibrin clots, and there-
fore are elevated in most patients with DIC. The D-dimer test readily detects
cross-linked fibrin degradation products. Elevated prothrombin fragment
F1+2 levels reflect thrombin activation as one of the central mechanisms
underlying DIC. Examination of a blood smear sometimes will show red cell
fragmentation in DIC. On the other hand, a high percentage of red cell frag-
38 Kiefel
mentation suggests a microangiopathic hemolytic disorder (discussed subse-

quently). Laboratory diagnosis (Marder et al., 1994) and therapy (Marder
et al., 1994; Humphries, 1994; Baglin, 1996) of DIC are reviewed elsewhere.
VII. THROMBOTIC THROMBOCYTOPENIC PURPURA

AND THE HEMOLYTIC-UREMIC SYNDROME
Thrombotic thrombocytopenic purpura (TTP) is a severe disease character-

ized by intravascular platelet aggregation, nonimmune hemolytic anemia,
neurological symptoms and signs, and renal failure. Red cell fragmentation,
hemolysis with a negative direct antiglobulin test, and platelet-rich thrombi
occluding small blood vessels are characteristic.
Different hypotheses have been proposed to explain the peculiar
phenomenon of platelet deposition within the precapillary arterioles (Moake
and Eisenstaedt, 1994). Plasma of patients with TTP and the hemolytic-
uremic syndrome (HUS) contains unusually large von Willebrand factor
(vWF) forms (Moake et al., 1982) that may mediate platelet aggregation.
These can be explained by hereditary deficiency of vWF-cleaving protease
(ADAMTS13) (Bianchi et al., 2002) in patients with chronic, relapsing TTP.
An IgG inhibitor of the vWF-cleaving protease is responsible for the more
common acute, self-limited form of TTP (Furlan et al., 1998). Various disor-
ders can be associated with a TTP-like illness, including treatment with immu-
nosuppressive drugs, metastatic cancer or its therapy (Gordon and Kwaan,
1997), and infections.
Vascular damage in HUS is usually confined to the kidneys, and
neurological sequelae are less pronounced than in TTP (Moake and Eisen-
staedt, 1994). HUS developing after bloody diarrhea is most commonly
related to infections with verocytotoxin-producing bacteria [e.g., Escherichia
coli serotype O157 (Taylor and Monnens, 1998)]. Transfer of verocytotoxin to
the target organs and damage of endothelial cells has been implicated in the
pathogenesis of HUS following infectious diarrhea and enterocolitis. Other
forms of HUS may be caused by drugs (cyclosporine, tacrolimus, quinine) or
complement abnormalities, e.g., factor H (Ohali et al., 1998; Noris et al.,
1999). Moreover, HUS can be found associated with SLE, during pregnancy,
and following bone marrow transplantation.
The cornerstone of therapy for TTP is transfusion of homologous
plasma, usually given by plasma exchange. Corticosteroids and immunosup-
pressive drugs may be effective in cases of auto-anti-vWF-cleaving protease.
Details of clinical features and therapeutic options are discussed elsewhere
(Remuzzi, 1987; George and Aster, 1990; Moake and Eisenstaedt, 1994; All-
ford et al., 2003).
VIII. TECHNIQUES FOR CHARACTERIZATION

OF PLATELET ANTIBODIES
Laboratory testing for immune-mediated thrombocytopenia requires specific

knowledge of the underlying clinical problem. For example, an accurate drug
history is needed to evaluate DIT in vitro.
A. Analysis of Platelet Autoantibodies and Alloantibodies

in Serum Samples
Often, only serum samples from a thrombocytopenic patient are available
for study. Depending on the clinical problem, it may be useful to screen for
platelet-reactive serum antibodies. A reliable, standardized immunoglobulin-
binding assay for platelet antibodies is the platelet suspension immunofluo-
rescence test (von dem Borne et al., 1978). A simplified alternative is more
convenient for large-scale screening (Schneider and Schnaidt, 1981), but may
be less sensitive. Usually a ‘‘panel’’ of platelets with different alloantigens is
employed: platelet suspensions are incubated with the serum to be studied,
and immunoglobulin binding to platelets is determined by platelet immuno-
fluorescence. Reactivity with all platelets from the panel is often observed
with platelet autoantibodies, but may also occur with antibodies against
‘‘high-frequency’’ antigens (e.g., HPA-4a [Yukb]; Naka), or mixtures of
alloantibodies, if the panel does not include antigen-negative platelet suspen-
sions. The most compelling way to exclude that a broadly reactive alloanti-
body is an autoantibody is to test the serum against autologous platelets. With
the rare exception of PTP, alloantibodies do not react with autologous
platelets.
Whenever possible, the platelet glycoprotein target of the antibodies
should be identified. This is important because many sera of patients who
have previously been exposed to allogeneic blood cells via transfusion or
pregnancy contain ‘‘contaminating’’ alloantibodies reacting with HLA class I
antigens present in high density on platelets.
Laboratory diagnosis of AITP, PTP, DIT, and certain thrombocyto-
penic states in newborns is based on the characterization of platelet-specific
antibodies. This may be accomplished with assays that include electropho-
retic determination of molecular weight of glycoproteins, including immu-
noblot (Herman et al., 1986; Huisman, 1986) or (radio-) immunoprecipitation
(Mulder et al., 1984; Santoso et al., 1989; Smith et al., 1993). These techniques
are cumbersome and time-consuming. Therefore, assays that allow identifi-
cation of target antigens with well-characterized monoclonal antibodies are
now preferred. These include the monoclonal antibody immobilization of
platelet antigens (MAIPA) assay (Kiefel et al., 1987b; Kiefel, 1992) and the
40 Kiefel
immunobead assay (McMillan et al., 1987). These monoclonal antibody-

based immunoassays are suitable to detect, and discriminate among, anti-
bodies against GP IIb/IIIa, GP Ia/IIa, GP Ib/IX, GP V, HLA class I antigen,
and other structures of the platelet membrane.
Virtually all antibodies reacting with the HLA class I antigens, and most
antibodies against GP Ia/IIa, are alloantibodies. In contrast, GP IIb/IIIa and
GP Ib/IX/V carry determinants recognized by autoantibodies, drug-depen-
dent antibodies, and alloantibodies (see Tables 1 and 2). The serological
diagnosis of PTP is based on detection of platelet alloantibodies, mainly
against GP IIb/IIIa.
B. Characterization of Platelet-Bound Antibodies

Determination of specific autoantibodies against GPs IIb/IIIa and Ib/IX on a
patient’s autologous platelets performed with ‘‘direct’’ glycoprotein-specific
immunoassays (McMillan et al., 1987) is much more specific than quantita-
tion of PAIgG. As an alternative, testing of eluates prepared at pH 2.8 from
autologous patient platelets in an antibody-binding assay (platelet immuno-
fluorescence) is also specific for AITP (Kiefel et al., 1996). Direct immuno-
Figure 1 Summary of drug-dependent antibody detection in immunoglobulin-bind-

ing assays.
precipitation detected anti-HPA-5b on the patient’s platelets in a case of

passive alloimmune thrombocytopenia (Warkentin et al., 1992).
C. Determination of Drug-Dependent Antibodies

Drug-dependent antibodies against platelets can be characterized using var-
ious techniques. If antiglobulin-binding assays are used, they should include
the following steps (Fig. 1). Test platelets should be incubated in buffer
containing the drug, and the serum sample to be tested is added. Platelets are
washed with buffer containing the same concentration of the drug. Following
incubation in buffer with the conjugated antihuman IgG (also containing the
same drug concentration as the incubation mixture), platelets are again
washed, and binding of IgG is detected by enzyme-linked immunosorbent
assay (ELISA) or by radioimmunoassay. With each experiment, several
controls should be included (Table 4). Figure 2 shows a typical experiment
Table 4 Interpretation of Immunoglobulin-Binding Assays for Detection of

Drug-Dependent Antibodies
Drug or
Platelets metabolite Serum Reaction Interpretation
a
+ + Patient + Drug-dependent antibody
+ Patient b
+ + Normal donor serum b
+ Normal donor serum b
+ + Patient + Autoantibody
+ Patient + (alloantibody)
+ + Normal donor serum
+ Normal donor serum
+ + Patient + (Nonspecific) adsorption

+ Patient of immunoglobulins to
+ + Normal donor serum + platelets induced by
+ Normal donor serum the drug
+ + Patient Negative result

+ Patient
+ + Normal donor serum
+ Normal donor serum
a
The experiment for ddAb detection, as depicted in Figure 2.
b
Control experiments.
42 Kiefel
Figure 2 Quinidine-dependent antibodies binding to platelet in the presence (0.01

mM and 0.1 mM) and absence of the drug (0 mM drug, in normal saline). Antibody
binding to platelets was detected only with quinidine added to the washing buffer at the
same concentrations as the incubation mixture. ddab, drug-dependent antibodies;
ELISA, enzyme-linked immunosorbent assay; IgG, immunoglobulin G; plt, platelets.
with two quinidine-dependent platelet antibodies: evidently the drug-depen-

dent antibodies do not remain fixed to the platelet membrane if the drug is not
included in the washing buffer. As already discussed, drug-dependent anti-
bodies can be caused by a drug metabolite, rather than by the drug itself. In
this case, positive results are obtained only with metabolites. If metabolites
are renally excreted, urine from a person ingesting the drug may be sufficient
as a crude ‘‘metabolite preparation’’ (Kiefel et al., 1987c). This approach has
also been explored for the detection of drug-dependent antibodies against red
blood cells (Salama and Mueller-Eckhardt, 1985).
REFERENCES
Allford SL, Hunt BJ, Rose P, Machin SJ. Guidelines on the diagnosis and man-
agement of the thrombotic microangiopathic haemolytic anaemias. Br J Hae-
matol 2003; 120:556–573.
Aster RH, George JN. Thrombocytopenia due to enhanced platelet destruction by

immunologic mechanisms. In: Williams WJ, Beutler E, Erslev AJ, Lichtman
MA, eds. Hematology. New York: McGraw-Hill, 1990:1370–1398.
Aster RH, Jandl JH. Platelet sequestration in man. I. Methods. J Clin Invest 1964;
43:843–855.
Baglin T. Disseminated intravascular coagulation: diagnosis and treatment. Br Med J
1996; 312:683–687.
Ballem PJ, Buskard NA, Decary F, Doubroff P. Post-transfusion purpura secondary
to passive transfer of anti-P1Al by blood transfusion. Br J Haematol 1987; 66:
113–114.
Beardsley DS, Spiegel JE, Jacobs MM, Handin RI, Lux SE. Platelet membrane
glycoprotein IIIa contains target antigens that bind anti-platelet antibodies in
immune thrombocytopenias. J Clin Invest 1984; 74:1701–1707.
Becker T, Küenzlen E, Salama A, Mertens R, Kiefel V, Weiß H, Lampert F, Gaedicke
G, Mueller-Eckhardt C. Treatment of childhood idiopathic thrombocytopenic
purpura with rhesus antibodies (anti-D). Eur J Pediatr 1986; 145:166–169.
Benda H, Panzer S, Kiefel V, Mannhalter C, Hinterberger W, Lechner K, Mueller-
Eckhardt C. Identification of the target platelet glycoprotein in autoimmune
thrombocytopenia occurring after allogeneic bone marrow transplantation.
Blut 1989; 58:151–153.
Berchtold P, McMillan R. Therapy of chronic idiopathic thrombocytopenic purpura
in adults. Blood 1989; 74:2309–2317.
Berndt MC, Chong BH, Bull HA, Zola H, Castaldi PA. Molecular characterization of
quinine/quinidine drug-dependent antibody platelet interaction using mono-
clonal antibodies. Blood 1985; 66:1292–1301.
Berney SI, Metcalfe P, Wathen NC, Waters AH. Post-transfusion purpura responding
to high dose intravenous IgG: further observations on pathogenesis. Br J Hae-
matol 1985; 61:627–632.
Berry JE, Murphy CM, Smith GA, Ranasinghe E, Finberg R, Walton J, Brown J,
Navarrete C, Metcalfe P, Ouwehand WH. Detection of Gov system antibodies
by MAIPA reveals an immunogenicity similar to the HPA-5 alloantigens. Br J
Haematol 2000; 110:735–742.
Bianchi V, Robles R, Alberio L, Furlan M, Lämmle B. Von Willebrand factor-
cleaving protease (ADAMTS13) in thrombocytopenic disorders: a severely defi-
cient activity is specific for thrombotic thrombocytopenic purpura. Blood 2002;
100:710–713.
Boizard B, Wautier JL. Leka, a new platelet antigen absent in Glanzmann’s throm-
basthenia. Vox Sang 1984; 46:47–54.
Bougie DW, Wilker PR, Wuitschick ED, Curtis BR, Malik M, Levine S, Lind RN,
Periera J, Aster RH. Acute thrombocytopenia after treatment with tirofiban or
eptifibatide is associated with antibodies specific for ligand-occupied GPIIb/
IIIa. Blood 2002; 100:2071–2076.
Branehög I, Kutti J, Weinfeld A. Platelet survival and platelet production in idiopathic
thrombocytopenic purpura (ITP). Br J Haematol 1974; 27:127–143.
Brassard JA, Curtis BR, Cooper RA, Ferguson J, Komocsar W, Ehardt M, Kupfer S,
Maurath C, Swabb E, Cannon CP, Aster RH. Acute thrombocytopenia in pa-
44 Kiefel
tients treated with the oral glycoprotein IIb/IIIa inhibitors xemilofiban and
orbofiban: evidence for an immune etiology. Thromb Haemost 2002; 88:892–
897.
Brunner-Bollinger S, Kiefel V, Horber FF, Nydegger UE, Berchtold P. Antibody
studies in a patient with acute thrombocytopenia following infusion of plasma
containing anti-P1Al. Am J Hematol 1997; 56:119–121.
Castaldi PA, Mehrabani PA, Fournier DJ, Berndt MC. Autoimmune thrombocyto-
penia associated with B-CLL and an IgG autoantibody directed against the
human platelet GP Ia-IIa complex (VLA-2) [abstr]. Thromb Haemost 1989;
62:151.
Chapman JF, Murphy MF, Berney SI, Ord J, Metcalfe P, Amess JAL, Waters AH.
Post-transfusion purpura associated with anti-Bak1 and anti-PlA2 platelet anti-
bodies and delayed hemolytic transfusion reaction. Vox Sang 1987; 52:313–317.
Chong BH, Cade J, Smith JA, Tatoulis J. An unusual case of post-transfusion pur-
pura: good transient response to high-dose immunoglobulin. Vox Sang 1986;
51:182–184.
Christie DJ, Mullen PC, Aster RH. Fab-mediated binding of drug-dependent
antibodies to platelets in quinidine- and quinine-induced thrombocytopenia. J
Clin Invest 1985; 75:310–314.
Christie DJ, Mullen PC, Aster RH. Quinine- and quinidine platelet antibodies can
react with GP IIb/IIIa. Br J Haematol 1987; 67:213–219.
Christie DJ, Pulkrabek S, Putnam JL, Slatkoff ML, Pischel KD. Posttransfusion
purpura due to an alloantibody reactive with glycoprotein Ia/IIa (anti-HPA-
5b). Blood 1991; 77:2785–2789.
Christie DJ, Sauro SC, Cavanaugh AL, Kaplan ME. Severe thrombocytopenia in an
acquired immunodeficiency syndrome patient associated with pentamidine-de-
pendent antibodies specific for glycoprotein IIb/IIIa. Blood 1993; 82:3075–3080.
Court WS, LoBuglio AF. Measurement of platelet surface-bound IgG by a mono-
clonal 125I-anti-IgG assay. Vox Sang 1986; 50:154–159.
Crosby WH. Wet purpura, dry purpura. JAMA 1975; 232:744–745.
Curtis BR, Swyers J, Divgi A, McFarland JG, Aster RH. Thrombocytopenia after
second exposure to abciximab is caused by antibodies that recognize abciximab-
coated platelets. Blood 2002; 99:2054–2059.
Dixon R, Rosse WF, Ebbert L. Quantitative determination of antibody in idiopathic
thrombocytopenic purpura. Correlation of serum and platelet-bound antibody
with clinical response. N Engl J Med 1975; 292:230–236.
Eden OB, Lilleyman JS. Guidelines for management of idiopathic thrombocytopenic
purpura. Arch Dis Child 1992; 67:1056–1058.
Eisner EV, Kasper K. Immune thrombocytopenia due to a metabolite of para-
aminosalicylic acid. Am J Med 1972; 53:790–796.
Eisner EV, Shahidi NT. Immune thrombocytopenia due to a drug metabolite. N Engl J
Med 1972; 287:376–381.
Follea G, Mandrand B, Dechavanne M. Simultaneous enzymo-immunologic assays of
platelet associated IgG, IgM, and C3. A useful tool in assessment of immune
thrombocytopenias. Thromb Res 1982; 26:249–258.
Frederiksen H, Schmidt K. The incidence of idiopathic thrombocytopenic purpura in

adults increases with age. Blood 1999; 94:909–913.
Furlan M, Robles R, Galbusera M, Remuzzi G, Kyrle PA, Brenner B, Krause M,
Scharrer I, Aumann V, Mittler U, Solenthaler M, Lammle B. Von Willebrand
factor-associated protease in thrombotic thrombocytopenic purpura and the
hemolytic-uremic syndrome. N Engl J Med 1998; 339:1578–1584.
George JN. Platelet immunoglobulin G: its significance for the evaluation of throm-
bocytopenia and for understanding the origin of alpha-granule proteins. Blood
1990; 76:859–870.
George JN, Aster RH. Thrombocytopenia due to enhanced platelet destruction
by nonimmunologic mechanisms. In: Williams WJ, Beutler E, Erslev AJ,
Lichtman MA, eds. Hematology. 4th ed. New York: McGraw-Hill, 1990:1351–
1370.
George JN, Woolf SH, Raskob GE, Wasser JS, Aledort LM, Ballem PJ, Blanchette
VS, Bussel JB, Cines DB, Kelton JG, Lichtin AE, McMillan R, Okerbloom JA,
Regan DH, Warrier I. Idiopathic thrombocytopenic purpura: a practice guide-
line developed by explicit methods for the American Society of Hematology.
Blood 1996; 88:3–40.
George JN, Raskob GE, Shah SR, Rizvi MA, Hamilton SA, Osborne S, Vondracek T.
Drug-induced thrombocytopenia: a systematic review of published case reports.
Ann Intern Med 1998; 129:886–890.
Gerstner JB, Smith MJ, Davis KD, Cimo PL, Aster RH. Posttransfusion purpura:
therapeutic failure of PlAl-negative platelet transfusion. Am J Hematol 1979;
6:71–75.
Gordon LI, Kwaan HC. Cancer- and drug-associated thrombotic thrombocytopenic
purpura and hemolytic uremic syndrome. Semin Hematol 1997; 34:140–147.
Greinacher A, Mueller-Eckhardt C. Hereditary types of thrombocytopenia, an impor-
tant differential diagnosis in chronic thrombocytopenia. In: Sutor AH, Thomas
KB, eds. Thrombocytopenia in Childhood. Stuttgart: Schattauer, 1994:191–
198.
Harrington WJ, Minnich V, Hollingsworth JW, Moore CV. Demonstration of a
thrombocytopenic factor in the blood of patients with thrombocytopenic
purpura. J Lab Clin Med 1951; 38:1–10.
Heaton WA, Davis HH, Welch MJ, Mathias CJ, Joist JH, Sherman LA, Siegel BA.
Indium-111 a new radionuclide label for studying human platelet kinetics. Br J
Haematol 1979; 42:613–622.
Hegde UM, Williams K, Devereux S, Bowes A, Powell D, Fisher D. Platelet associated
IgG and immune thrombocytopenia in lymphoproliferative and autoimmune
disorders. Clin Lab Haematol 1983; 5:9–15.
Hegde UM, Ball S, Zuiable A, Roter BLT. Platelet associated immunoglobulins
(PAIgG and PAIgM) in autoimmune thrombocytopenia. Br J Haematol 1985;
59:221–226.
Henke M, Engler H, Engelhardt R, Löhr GW. Successful treatment of sarcoidosis-
associated thrombocytopenia refractory to corticosteroids by a single course of
human gammaglobulins. Klin Wochenschr 1986; 64:1209–1211.
46 Kiefel
Herman JH, Kickler TS, Ness PM. The resolution of platelet serologic problems using
Western blotting. Tissue Antigens 1986; 28:257–268.
Hesseling PB. Onyalai in Namibia. Clinical manifestations, haematological findings,
course and management of 103 patients in the Kavango territory. Trans R Soc
Trop Med Hyg 1987; 81:193–196.
Hesseling PB. Onyalai. Baillieres Clin Haematol 1992; 5:457–473.
Huisman JG. Immunoblotting: an emerging technique in immunohematology. Vox
Sang 1986; 50:129–136.
Humphries JE. Transfusion therapy in acquired coagulopathies. Hematol Oncol Clin
North Am 1994; 8:1181–1201.
Ibarra H, Zapata C, Inostroza J, Mezzano S, Riedemann S. Immune thrombocyto-
penic purpura associated with hepatitis A. Blut 1986; 52:371–375.
Imbach P, Barandun S, d’Apuzzo V, Baumgartner C, Hirt A, Morell A, Rossi E,
Schöni M, Vest M, Wagner HP. High-dose intravenous gammaglobulin for
idiopathic thrombocytopenic purpura in childhood. Lancet 1981; 1:1228–1231.
Imbach P, Berchtold W, Hirt A, Mueller-Eckhardt C, Rossi E, Wagner HP, Gaedicke
G, Joller P, Müller B, Barandun S. Intravenous immunoglobulin versus oral
corticosteroids in acute thrombocytopenic purpura in childhood. Lancet 1985;
2:464–468.
Jubelirer SJ, Koenig BA, Bates MC. Acute profound thrombocytopenia following
C7E3 Fab (abciximab) therapy: case reports, review of the literature and impli-
cations for therapy. Am J Hematol 1999; 61:205–208.
Keimowitz RM, Collins J, Davis K, Aster RH. Post-transfusion purpura associated
with alloimmunization against the platelet-specific antigen, Baka. Am J Hematol
1986; 21:79–88.
Kelton JG, Neame PB, Gauldie J, Hirsh J. Elevated platelet-associated IgG in the
thrombocytopenia of septicemia. N Engl J Med 1979; 300:760–764.
Kelton JG, Smith JW, Horsewood P, Humbert JR, Hayward CPM, Warkentin TE.
Gova/Govb alloantigen system on human platelets. Blood 1990; 75:2172–2176.
Kereiakes DJ, Essell JH, Abbottsmith CW, Broderick TM, Runyon JP. Abciximab-
associated profound thrombocytopenia: therapy with immunoglobulin and
platelet transfusion. Am J Cardiol 1996; 78:1161–1163.
Kickler TS, Herman JH, Furihata K, Kunicki TJ, Aster RH. Identification of Bakb, a
new platelet-specific antigen associated with posttransfusion purpura. Blood
1988; 71:894–898.
Kiefel V. The MAIPA assay and its applications in immunohematology. Transfusion
Med 1992; 2:181–188.
Kiefel V, Spaeth P, Mueller-Eckhardt C. Immune thrombocytopenic purpura: auto-
immune or immune complex disease? Br J Haematol 1986; 64:57–68.
Kiefel V, Jäger S, Mueller-Eckhardt C. Competitive enzyme-linked immunoassay for
the quantitation of platelet-associated immunoglobulins (IgG, IgM, IgA) and
complement (C3c, C3d). Vox Sang 1987a; 53:151–156.
Kiefel V, Santoso S, Weisheit M, Mueller-Eckhardt C. Monoclonal antibody-specific
immobilization of platelet antigens (MAIPA): a new tool for the identification
of platelet reactive antibodies. Blood 1987b; 70:1722–1726.
Kiefel V, Santoso S, Schmidt S, Salama A, Mueller-Eckhardt C. Metabolite-specific

(IgG) and drug-specific antibodies (IgG, IgM) in two cases of trimethoprim-
sulfamethoxazole-induced immune thrombocytopenia. Transfusion 1987c;
27:262–265.
Kiefel V, Santoso S, Glöckner WM, Katzmann B, Mayr WR, Mueller-Eckhardt C.
Posttransfusion purpura associated with an anti-Bakb. Vox Sang 1989; 56:93–
97.
Kiefel V, Santoso S, Kaufmann E, Mueller-Eckhardt C. Autoantibodies against
platelet glycoprotein Ib/IX: a frequent finding in autoimmune thrombocyto-
penic purpura. Br J Haematol 1991; 79:256–262.
Kiefel V, Santoso S, Mueller-Eckhardt C. Serological, biochemical and molecular
aspects of platelet autoantigens. Semin Hematol 1992; 29:26–33.
Kiefel V, Kroll H, Bonnert J, Unkelbach K, Katzmann B, Nebenführer Z, Santoso S,
Mueller-Eckhardt C. Platelet alloantigen frequencies in Caucasians: a serolog-
ical study. Transfus Med 1993; 3:237–242.
Kiefel V, Freitag E, Kroll H, Santoso S, Mueller-Eckhardt C. Platelet autoantibodies
(IgG, IgM, IgA) against glycoproteins IIb/IIIa and Ib/IX in patients with
thrombocytopenia. Ann Haematol 1996; 72:280–285.
Kosmo MA, Bordin G, Tani P, McMillan R. Immune thrombocytopenia and Crohn’s
disease [letter]. Ann Intern Med 1986; 104:136.
Kosugi S, Tomiyama Y, Shiraga M, Kashiwagi H, Nakao H, Kanayama Y, Kurata Y,
Matzuzawa Y. Cyclic thrombocytopenia associated with IgM anti-GPIIb-IIIa
autoantibodies. Br J Haematol 1994; 88:809–815.
Kroll H, Kiefel V, Mueller-Eckhardt C. Posttransfusionelle Purpura: Klinische und
immunologische Untersuchungen bei 38 Patientinnen. Infusionsthera Trans-
fusionsmed 1993; 20:198–204.
Kubota T, Tanoue K, Murohashi I, Nara N, Yamamoto N, Yamazaki H, Aoki N.
Autoantibody against platelet glycoprotein IIb/IIIa in a patient with non
Hodgkin’s lymphoma. Thromb Res 1989; 53:379–386.
Kunicki TJ, Johnson MM, Aster RH. Absence of the platelet receptor for drug-
dependent antibodies in the Bernard-Soulier syndrome. J Clin Invest 1978;
62:716–719.
Kunicki TJ, Koenig MB, Kristopeit SM, Aster RH. Direct quantitation of platelet-
associated IgG by electroimmunoassay. Blood 1982; 60:54–58.
Law C, Marcaccio M, Tam P, Heddle N, Kelton JG. High-dose intravenous immune
globulin and the response to splenectomy in patients with idiopathic
thrombocytopenic purpura. N Engl J Med 1997; 336:1494–1498.
Leporrier M, Dighiero G, Auzemery M, Binet JL. Detection and quantification of
platelet-bound antibodies with immunoperoxidase. Br J Haematol 1979; 42:
605–611.
Lubenow N, Eichler P, Albrecht D, Carlsson LE, Kothmann J, Rossocha WR, Hahn
M, Quitmann H, Greinacher A. Very low platelet counts in post-transfusion
purpura falsely diagnosed as heparin-induced thrombocytopenia. Report of
four cases and review of literature. Thromb Res 2000; 100:115–125.
Lucas GF, Pittman SJ, Davies S, Solanki T, Bruggemann K. Post-transfusion purpura
48 Kiefel
(PTP) associated with anti-HPA-1a, anti-HPA-2b and anti-HPA-3a antibodies.

Transfusion Med 1998; 7:295–299.
Malloy B, Noel P, Eaton D, Solberg L. Immune thrombocytopenia associated with an
antibody to c-Mpl, the thrombopoietin (TPO) receptor [abstr]. Blood 1995;
86(suppl 1):279a.
Mammen EF. The haematological manifestations of sepsis. J Antimicrob Chemother
1998; 41(suppl A):17–24.
Marder VJ, Feinstein DI, Francis CW, Colman RW. Consumptive thrombohemor-
rhagic disorders. In: Colman RW, Hirsh J, Marder VJ, Salzman EW, eds.
Hemostasis and Thrombosis. Basic Principles and Clinical Practice. 3rd ed.
Philadelphia: JB Lippincott, 1994:1023–1063.
Maslanka K, Zupanska B. Post-transfusion purpura and delayed haemolytic
transfusion reaction. Transfusion Med 1993; 3:281–284.
Mayer JL, Beardsley DS. Varicella-associated thrombocytopenia: autoantibodies
against platelet surface glycoprotein V. Pediatr Res 1996; 40:615–619.
McMillan R, Tani P, Mason D. A method that allows shipment of whole blood for the
assay of platelet-associated IgG. Blood 1979; 54:1201–1202.
McMillan R, Tani P, Millard F, Berchtold P, Renshaw L, Woods VL. Platelet-
associated and plasma anti-glycoprotein autoantibodies in chronic ITP. Blood
1987; 70:1040–1045.
Menitove JE, Pereira J, Hoffman R, Anderson T, Fried W, Aster RH. Cyclic throm-
bocytopenia of apparent autoimmune etiology. Blood 1989; 73:1561–1569.
Meyer M, Kirchmaier CM, Spangenberg P, Ströhl C, Breddin K. Acquired disorder of
platelet function associated with autoantibodies against membrane glycopro-
tein IIb-IIIa complex-1. Glycoprotein analysis. Thromb Haemost 1991; 65:491–
496.
Meyer T, Herrmann C, Wiegand V, Mathias B, Kiefel V, Mueller-Eckhardt C.
Immune thrombocytopenia associated with hemorrhagic diathesis due to ibu-
profen administration. Clin Invest 1993; 71:413–415.
Moake JL, Eisenstaedt RS. Thrombotic thrombocytopenic purpura and the hemolytic
uremic syndrome. In: Colman RW, Hirsh J, Marder VJ, Salzman EW, eds.
Hemostasis and Thrombosis. Basic Principles and Clinical Practice. 3rd ed.
Philadelphia: JB Lippincott, 1994:1064–1075.
Moake JL, Rudy CK, Troll JH, Weinstein MJ, Colannino NM, Azocar J, Seder RH,
Hong SL, Deykin D. Unusually large plasma factor VIII:von Willebrand factor
multimers in chronic thrombocytopenic purpura. N Engl J Med 1982; 307:1432–
1435.
Moilan J, Scott E, Dalmasso A. Transfusion reaction with severe thrombocytopenia
due to a passively acquired platelet specific antibody [abstr]. Transfusion 1985;
25:459.
Mollison PL, Engelfriet CP, Contreras M. Blood Transfusion in Clinical Medicine. 9th
ed. Oxford: Blackwell Scientific, 1993.
Morse BS, Giuliani D, Nussbaum M. Quantitation of platelet-associated IgG by
radial immunodiffusion. Blood 1981; 57:809–811.
Müller-Berghaus G. Septicemia and the vessel wall. In: Verstraete M, Vermylen J,
Lijnen R, Arnout J, eds. Thrombosis and Haemostasis 1987. Leuven: Leuven

University Press, 1987:619–671.
Mueller-Eckhardt C, Mahn I, Schulz G, Mueller-Eckhardt G. Detection of platelet
autoantibodies by a radioactive anti-immunoglobulin test. Vox Sang 1978; 35:
357–365.
Mueller-Eckhardt C, Mueller-Eckhardt G, Kayser W, Voss RM, Wegner J, Küenzlen
E. Platelet associated IgG, platelet survival and platelet sequestration in throm-
bocytopenic states. Br J Haematol 1982; 52:49–58.
Mueller-Eckhardt C, Küenzlen E, Kiefel V, Vahrson H, Graubner M. Cyclophospha-
mide-induced immune thrombocytopenia in a patient with ovarian carcinoma
successfully treated with intravenous gamma globulin. Blut 1983a; 46: 165–169.
Mueller-Eckhardt C, Küenzlen E, Thilo-Körner D, Pralle H. High-dose intravenous
immunoglobulin for post-transfusion purpura [letter]. N Engl J Med 1983b;
308:287.
Mueller-Eckhardt C, Allolio B, Salama A, Kiefel V, Deuss U. Nomifensine-dependent
immune hemolytic anemia and posttransfusion purpura in the same patient.
Transfusion 1987; 27:250–252.
Mueller-Eckhardt C, Kiefel V. High-dose IgG for post-transfusion purpura—revis-
ited. Blut 1988; 57:163–167.
Mueller-Eckhardt C, Kroll H, Kiefel V. The members of the European PTP Study
Group. Posttransfusion purpura. In: Kaplan-Gouet C, Schlegel N, Salmon C,
McGregor J, eds. Platelet Immunology: Fundamental and Clinical Aspects.
Paris: John Libbey Eurotext, 1991:249–255.
Mulder A, van Leeuwen EF, Veenboer GJM, Tetteroo PAT, von dem Borne AEGK.
Immunochemical characterization of platelet-specific alloantigens. Scand J
Haematol 1984; 33:267–274.
Nagasawa T, Hasegawa Y, Kamoshita M, Ohtani K, Komeno T, Itoh T, Shinagawa
A, Kojima H, Ninomiya H, Abe T. Megakaryopoiesis in patients with cyclic
thrombocytopenia. Br J Haematol 1995; 91:185–190.
Najean Y, Lecompte T. Genetic thrombocytopenia with autosomal dominant
transmission: a review of 54 cases. Br J Haematol 1990; 74:203–208.
Najean Y, Rain JD, Billotey C. The site of destruction of autologous 111In-labeled
platelets and the efficacy of splenectomy in children and adults with idiopathic
thrombocytopenic purpura: a study of 578 patients with 268 splenectomies. Br J
Haematol 1997; 97:547–550.
Niessner H, Clemetson KJ, Panzer S, Mueller-Eckhardt C, Santoso S, Bettelheim P.
Acquired thrombasthenia due to GP IIb/IIIa-specific autoantibodies. Blood
1986; 68:571–576.
Noris M, Ruggenenti P, Perna A, Orisio S, Caprioli J, Skerka C, Vasile B, Zipfel PF,
Remuzzi G. Hypocomplementemia discloses genetic predisposition to hemo-
lytic uremic syndrome and thrombotic thrombocytopenic purpura: role of
factor H abnormalities. J Am Soc Nephrol 1999; 10:281–293.
Ohali M, Shalev H, Schlesinger M, Katz Y, Kachko L, Carmi R, Sofer S, Landau D.
Hypocomplementemic autosomal recessive hemolytic uremic syndrome with
decreased factor H. Pediatr Nephrol 1998; 12:619–624.
50 Kiefel
Panzer S, Penner E, Nelson PJ, Prochatzka E, Benda H, Saurugger PN. Identification

of the platelet glycoprotein Ib/IIIa complex as a target antigen in primary biliary
cirrhosis-associated autoimmune thrombocytopenia. Evidence that platelet-
reactive autoantibodies can also bind to the mitochondrial antigen M2. J Auto-
immun 1990; 3:473–483.
Pasquini E, Gianni L, Aitini E, Nicolini M, Fattori PP, Cavazzini G, Desiderio F,
Monti F, Forghieri ME, Ravaioli A. Acute disseminated intravascular coagula-
tion syndrome in cancer patients. Oncology 1995; 52:505–508.
Pawlotsky JM, Bouvier M, Fromont P, Deforges L, Duval J, Dhumeaux D, Bierling
P. Hepatitis C virus infection and autoimmune thrombocytopenic purpura.
J Haematol 1995; 23:635–639.
Petz LD, Garratty G. Acquired Immune Hemolytic Anemias. New York: Churchill
Livingstone, 1980.
Pfueller SL, Bilston RA, Jane S, Gibson J. Expression of the drug-dependent antigen
for quinine-dependent antiplatelet antibodies on GP IIIa but not that on GP Ib,
IIb or IX on human endothelial cells. Thromb Haemost 1990; 63:279–281.
Quiquandon I, Fenaux P, Caulier MT, Pagniez D, Huart JJ, Bauters F. Re-evalua-
tion of the role of azathioprine in the treatment of adult chronic idiopathic
thrombocytopenic purpura: a report of 53 cases. Br J Haematol 1990; 74:223–
228.
Reiner A, Gernsheimer T, Slichter SJ. Pulse cyclophosphamide therapy for refractory
auto-immune thrombocytopenic purpura. Blood 1995; 85:351–358.
Remuzzi G. Thrombotic thrombocytopenic purpura and allied disorders. In: Ver-
straete M, Vermylen J, Lijnen R, Arnout J, eds. Thrombosis and Haemostasis
1987. Leuven: Leuven University Press, 1987:673–708.
Salama A, Mueller-Eckhardt C. The role of metabolite-specific antibodies in nomi-
fensine-dependent immune hemolytic anemia. N Engl J Med 1985; 313: 469–
474.
Salama A, Mueller-Eckhardt C. Immune-mediated blood cell dyscrasias related to
drugs. Semin Hematol 1992; 29:54–63.
Salama A, Mueller-Eckhardt C, Kiefel V. Effect of intravenous immunoglobulin in
immune thrombocytopenia. Competitive inhibition of reticuloendothelial
system function by sequestration of autologous red blood cells? Lancet 1983;
2:193–196.
Sane DC, Damaraju LV, Topol EJ, Cabot CF, Mascelli MA, Harrington RA,
Simoons ML, Califf RM. Occurrence and clinical significance of pseudothrom-
bocytopenia during abciximab therapy. J Am Coll Cardiol 2000; 36:75–83.
Santoso S, Kiefel V, Mueller-Eckhardt C. Immunochemical characterization of the
new platelet alloantigen system Bra/Brb. Br J Haematol 1989; 72:191–198.
Schell DA, Ganti AK, Levitt R, Potti A. Thrombocytopenia associated with c7E3 Fab
(abciximab). Ann Hematol 2002; 81:76–79.
Schneider W, Schnaidt M. The platelet adhesion immunofluorescence test: a modifi-
cation of the platelet suspension immunofluorescence test. Blut 1981; 43:389–
392.
Scott EP, Moilan-Bergeland J, Dalmasso AP. Posttransfusion thrombocytopenia
associated with passive transfusion of a platelet-specific antibody. Transfusion

1988; 28:73–76.
Shulman NR, Jordan JV. Platelet kinetics. In: Colman RW, Hirsh J, Marder VJ,
Salzman EW, eds. Hemostasis and Thrombosis. Basic Principles and Clinical
Practice. Philadelphia: KB Lippincott, 1987:431–451.
Shulman NR, Reis DM. Platelet immunology. In: Colman RW, Hirsh J, Marder VJ,
Salzman EW, eds. Hemostasis and Thrombosis: Principles and Clinical Prac-
tice. 3d ed. Philadelphia: JB Lippincott, 1994:414–468.
Shulman NR, Aster RH, Leithner A, Hiller MC. Immunoreactions involving platelets.
V. Post-transfusion purpura due to a complement-fixing antibody against a
genetically controlled platelet antigen. A proposed mechanism for thrombocy-
topenia and its relevance in ‘‘autoimmunity.’’ J Clin Invest 1961; 40:1597–1620.
Simon TL, Collins J, Kunicki TJ, Furihata K, Smith KJ, Aster RH. Posttransfusion
purpura associated with alloantibody specific for the platelet antigen, Pena. Am
J Hematol 1988; 29:38–40.
Smith JW, Hayward CPM, Warkentin TE, Horsewood P, Kelton JG. Investigation of
human platelet alloantigens and glycoproteins using non-radioactive immuno-
precipitation. J Immunol Methods 1993; 158:77–85.
Stricker RB, Shuman MA. Quinidine purpura: evidence that glycoprotein V is a target
platelet antigen. Blood 1986; 67:1377–1381.
Taaning E, Morling N, Ovesen H, Svejgaard A. Posttransfusion purpura and anti-Zwb
(-P1A2). Tissue Antigens 1985; 26:143–146.
Taylor CM, Monnens LAH. Advances in hemolytic uraemic syndrome. Arch Dis
Child 1998; 78:190–193.
Tcheng JE, Kereiakes DJ, Lincoff AM, George BS, Kleiman NS, Sane DC, Cines DB,
Jordan RE, Mascelli MA, Langrall MA, Damaraju L, Schantz A, Effron MB,
Braden GA. Abciximab readministration: results of the ReoPro Readministra-
tion Registry. Circulation 2001; 104:870–875.
Tomer A, Schreiber AD, McMillan R, Cines DB, Burstein SA, Thiessen AR, Harker
LA. Menstrual cyclic thrombocytopenia. Br J Haematol 1989; 71:519–524.
Tomiyama Y, Kurata Y, Mizutani H, Kanakura Y, Tsubakio T, Yonezawa T, Tarui S.
Platelet glycoprotein IIb as a target antigen in two patients with chronic idio-
pathic thrombocytopenic purpura. Br J Haematol 1987; 66:535–538.
Van der Lelie J, Lange JMA, Vos JJE, van Dalen CM, Danner SA, von dem Borne
AEGK. Autoimmunity against blood cells in human immunodeficiency virus
(HIV) infection. Br J Haematol 1987; 67:109–114.
van Leeuwen EF, van der Ven JTM, Engelfriet CC, von dem Borne AEGK. Specificity
of autoantibodies in autoimmune thrombocytopenia. Blood 1982a; 59:23–26.
van Leeuwen EF, Engelfriet CP, von dem Borne AEGK. Studies on quinine- and
quinidine-dependent antibodies against platelets and their reaction with platelet
in the Bernard-Soulier syndrome. Br J Haematol 1982b; 51:551–560.
Visentin GP, Newman PJ, Aster RH. Characteristics of quinine- and quinidine-in-
duced antibodies specific for platelet glycoproteins IIb and IIIa. Blood 1991;
77:2668–2676.
von dem Borne AEGK, Verheugt FWA, Oosterhof F, von Riesz E, Brutel de la Riviere
52 Kiefel
A, Engelfriet CP. A simple immunofluorescence test for the detection of platelet

antibodies. Br J Haematol 1978; 39:195–207.
von dem Borne AEGK, Pegels JG, van der Stadt RJ, van der Plas-van Dalen CM,
Helmerhorst FM. Thrombocytopenia associated with gold therapy: a drug-
induced autoimmune disease. Br J Haematol 1986; 63:509–516.
Walsh C, Krigel R, Lennette E, Karpatkin S. Thrombocytopenia in homosexual
patients. Prognosis, response to therapy, and prevalence of antibody to the
retrovirus associated with the acquired immunodeficiency syndrome. Ann
Intern Med 1985; 103:542–545.
Warkentin TE, Smith JW, Hayward CPM, Ali AM, Kelton JG. Thrombocytopenia
caused by passive transfusion of anti-glycoprotein Ia/IIa alloantibody (anti-
HPA-5b). Blood 1992; 79:2480–2484.
Waters AH. Autoimmune thrombocytopenia: clinical aspects. Semin Hematol 1992;
29:18–25.
Wintrobe MM, Lee GR, Boggs DR, Bithell TC, Foerster J, Athens JW, Lukens JN.
Clinical Hematology. 8th ed. Philadelphia: Lea & Febiger, 1981:1090–1127.
Woods VL, Kurata Y, Montgomery RR, Tani P, Mason D, Oh EH, McMillan R.
Autoantibodies against platelet glycoprotein Ib in patients with chronic immune
thrombocytopenic purpura. Blood 1984; 64:156–160.
3
Clinical Picture of Heparin-Induced
Thrombocytopenia
I. INTRODUCTION
Heparin-induced thrombocytopenia (HIT) is a distinct clinicopathologic syn-

drome caused by platelet-activating antibodies that recognize complexes of
platelet factor 4–heparin (PF4/H). Its strong association with venous and ar-
terial thrombosis represents a striking paradox. However, thrombocytopenia
itself is common in clinical medicine. Furthermore, heparin is usually given to
patients who either have thrombosis, or who are judged to be at high risk for
thrombosis. Thus, thrombocytopenia with or without thrombosis during hep-
arin treatment does not necessarily indicate a diagnosis of HIT. Indeed, sev-
eral disorders can closely resemble HIT (see Chap. 12).
On the other hand, HIT is associated with a wide spectrum of unusual
thrombotic and other complications (Table 1). Unrecognized HIT may have
been an important contributing factor in otherwise bizarre clinical events that
have occurred in certain heparin-treated patients (Anderson et al., 1981;
Solomon et al., 1988; Pfueller et al., 1990; Muntean et al., 1992). Laboratory
documentation of HIT antibodies has been crucial in determining the clinical
scope of the HIT syndrome. Accordingly, this chapter emphasizes clinical
data obtained from large prospective and retrospective studies that have used
diagnostic testing for HIT antibodies.
53
54 Warkentin
Table 1 Thrombotic and Other Sequelae of HIT

Venous thrombosis Arterial thrombosis Miscellaneous
Deep vein thrombosis Aortic or iliofemoral throm- Heparin-induced skin lesions

(DVT) (50%): new, bosis resulting in acute at heparin injection sites
progressive, recurrent; limb ischemia/infarction (10–20%):
lower limb (often (5–10%) or spinal cord Erythematous plaques
bilateral); upper limb infarction (rare) Skin necrosis
(at site of venous Acute thrombotic stroke Coumarin-induced skin
catheter); phlegmasia (3–5%) necrosis complicating HIT
cerulea dolens Myocardial infarction involving ‘‘central’’ sites
Coumarin-induced venous (3–5%) (breast, abdomen, thigh,
limb gangrene (f5–10% Cardiac intraventricular leg, etc.) (rare)
of DVT treated with or intra-atrial thrombosis, Acute systemic reactions
warfarin) in situ or via embolization postintravenous heparin
Pulmonary embolism of DVT (rare) bolus (f25% of sensitized
(25%): with or without Thrombosis involving patients who receive an
right-sided cardiac miscellaneous arteries intravenous heparin bolus):
intra-atrial or (rare): upper limb, renal, Inflammatory: e.g., fever,
intraventricular thrombi mesenteric, spinal, and chills, flushing
Cerebral dural sinus other arteries Cardiorespiratory: e.g.,
thrombosis (rare) Embolization of thrombus tachycardia, hypertension,
Adrenal hemorrhagic from heart or proximal dyspnea; cardiopulmonary
infarction (rare): aorta can also contribute arrest (rare)
bilateral (acute or to microvascular ischemic Gastrointestinal: nausea,
chronic adrenal syndromes vomiting, diarrhea
failure) or unilateral Neurological: transient
Disseminated intravascular coagulation (DIC), with hypofibri- global amnesia, headache
nogenemia and acquired natural anticoagulant deficiency,
causing multiple venous and arterial thromboses (rare)
Estimated frequencies of the various complications of HIT are taken from reports with serological con-
firmation of the diagnosis (Warkentin et al., 1995; Warkentin and Kelton, 1996; Warkentin et al., 1997).
‘‘Rare’’ indicates an estimated frequency <3% of HIT patients.
II. THROMBOCYTOPENIA
Thrombocytopenia, using the standard definition of a platelet count of less

than 150 109/L, is the most common clinical effect of HIT, occurring in 85–
90% of patients (Warkentin 1998a). An even higher proportion develop
‘‘thrombocytopenia’’ if a definition appropriate for the clinical situation is
used.
A. Timing
The characteristic delay of 5 or more days between initiation of heparin and
onset of thrombocytopenia was the major clue that led early investigators to
Clinical Picture of HIT 55
recognize the immune pathogenesis of HIT (Roberts et al., 1964; Rhodes et

al., 1973). King and Kelton (1984) noted that thrombocytopenia occurred
between days 6 and 15 for more than 90% of patients in whom HIT occurred
during their first exposure to heparin. In contrast, for patients who developed
HIT during a repeat course of heparin, the onset of thrombocytopenia was
often more rapid, occurring within 2 days. These data have been interpreted
as indicating an ‘‘anamnestic’’ (Gr., memory) or ‘‘secondary’’ immune
response in HIT, i.e., the immune system produces HIT antibodies more
quickly on reencountering an antigen ‘‘remembered’’ within its memory cell
repertoire. Recent data, however, suggest another explanation for these two
temporal profiles of HIT, typical and rapid (discussed subsequently).
Typical Onset of HIT

A prospective study of serologically confirmed HIT showed that the platelet
count typically begins to fall between days 5 and 10 (inclusive) of postoper-
ative subcutaneous heparin prophylaxis (Warkentin et al., 1995, 2003)
(Fig. 1). Note that the data refer to the day the platelet count begins to fall,
and not the later day on which an arbitrary threshold defining thrombocy-
topenia is crossed. This study also showed that most patients who developed
thrombocytopenia beginning after day 5 had HIT rather than another
explanation for the thrombocytopenia. The data suggest the following clinical
rule:
Rule 1
A thrombocytopenic patient whose platelet count fall began between
days 5 and 10 of heparin treatment (inclusive) should be considered to
have HIT unless proved otherwise (first day of heparin use is considered
‘‘day 0’’).
HIT-IgG antibodies generally are not detectable before day 5 of heparin
treatment, but are readily detectable using sensitive assays when the platelet
count first begins to fall due to HIT.
A recent study (Warkentin and Kelton, 2001a) that analyzed temporal
aspects of the platelet count fall in 243 patients with serological confirmed
HIT in relation to heparin use (both past and present) also found that the
onset of the platelet count fall typically occurs between days 5 and 10 (Fig. 2).
Interestingly, among these patients with typical onset of HIT, there was no
significant difference in the time to onset of HIT, irrespective of whether or
not the patients had been exposed to heparin in the past. For most patients
with typical onset of HIT, previous heparin exposure had occurred in the
‘‘remote’’ past, arbitrarily defined as more than 100 days previous (Fig. 2).
Gruel and colleagues (2003) have reported that the onset of the platelet
count fall may occur on average several days later in patients who develop
56 Warkentin
Figure 1 HIT in a clinical trial of postoperative orthopedic patients. (a) Serial

platelet counts of nine patients with HIT (platelet count nadir <150 109/L). The bold
line and shaded area indicate the mean (F2 SD) platelet count in the reference
population (367 patients who tested negative for HIT antibodies). The reference
population indicates the occurrence of postoperative thrombocytopenia (days 1–3),
followed by postoperative thrombocytosis (maximal, days 11–14). Nine patients
developed serologically confirmed HIT, with a platelet count fall to <150 109/L;
eight of the nine patients developed HIT-associated thrombosis (see insert for
description of the types of thrombi observed; all thrombi were venous, except for a
mesenteric artery thrombosis). (b) Serial platelet counts of nine patients with HIT
(platelet nadir >150 109/L, but platelet count fall >50%). Five patients developed
DVT (*). HIT developed in seven patients receiving unfractionated heparin (UFH)
and two receiving low molecular weight heparin (LMWH) (y). The platelet count fell
abruptly on postoperative day 13 when 5000 U of intravenous UFH was given. (c) Day
of onset of HIT for 18 patients observed in a clinical trial. HIT began between days 5
and 10, inclusive, in all 18 patients. Length of heparin treatment was variable; thus, the
remaining number of patients at risk for HIT for each day of follow-up is shown (n).
* For one of the patients, the platelet count began to fall on day 5 after UFH ‘‘flushes’’
were received through an intra-arterial catheter placed at the time of surgery. y The
platelet count fell abruptly on day 12 (postoperative day 13), together with symptoms
and signs of an acute systemic reaction, following administration of a 5000 U
intravenous UFH bolus (see Fig. 1a, Chap. 4). However, the first clinical manifestation
of HIT was on day 9 (erythematous skin lesions at heparin injection sites), and HIT
antibodies were first detected on day 5. z The platelet count fell abruptly on day 10 after
administration of a 5000 U intravenous UFH bolus, followed by therapeutic-dose
UFH infusion. However, positive HIT antibodies were first detected by PF4/H-EIA
on day 6 of treatment with subcutaneous UFH, 7500 U twice daily. (a, c, Warkentin
et al., 1995; Warkentin, 2000; b, Warkentin et al., 1995, 2003.)
Figure 1 Continued.
58 Warkentin
Figure 2 Temporal patterns of HIT in 243 patients in relation to previous treatment

with heparin. (A) Data are shown for the patients in whom the day of onset of HIT
could be determined to within a 3-day period. Among 170 patients with typical onset of
HIT, there was no significant difference in onset of HIT (median day), irrespective of
whether previous heparin exposure had been definite (6.5, n = 47), possible (7.0, n =
49), or unlikely (6.0, n = 74) ( p = 0.88, definite vs. unlikely). Among 120 patients who
had definite previous exposure to heparin, 73 had rapid onset of HIT. (B) For the
subgroup of patients with definite previous exposure to heparin, the 73 patients with
rapid onset of HIT invariably had been exposed to heparin within the past 100 days
(i.e., no patients shown at the asterisk [*]); in contrast, only 16/47 patients with typical
onset of HIT had been exposed to heparin within the past 100 days ( p < 0.001). (From
Warkentin and Kelton, 2001a).
HIT during low molecular weight heparin (LMWH) therapy. More time may
be required to generate clinically important levels of HIT-IgG so as to activate
platelets in the presence of PF4/LMWH, rather than PF4/H, complexes.
Diminishing Risk of HIT After Day 10

The risk of HIT decreases after the day 5–10 ‘‘window’’ passes (see Fig. 1c). In
my experience, a platelet count fall after day 10 is usually caused by another
pathological process, such as septicemia. In a notable exception, sometimes
an invasive procedure ‘‘resets the clock’’; that is, a platelet count fall that
begins on day 12 of a course of heparin that consists of two 6-day treatments
with heparin (before and after intervening surgery) is likely HIT. Perhaps the
surgery causes circumstances that favor seroconversion (e.g., release of PF4)
(see Chap. 6). Tholl and colleagues (1997) reported on a patient who for 9
years uneventfully received unfractionated heparin (UFH) for hemodialysis;
nevertheless, HIT complicating hemodialysis began shortly after the patient
underwent parathyroidectomy.
Rapid Onset of HIT

Sometimes patients develop rapid-onset HIT. This is defined as an unexpected
fall in the platelet count that begins soon after heparin is started. Indeed, it is
generally evident on the first postheparin platelet count, whether obtained
minutes, hours, or a day later. Patients who develop such a rapid fall in the
platelet count and who are confirmed serologically to have HIT antibodies
invariably have received heparin in the past (Warkentin and Kelton, 2001a;
Lubenow et al., 2002). A characteristic feature of this prior heparin exposure
has been recently identified: it invariably includes a recent exposure to
heparin, generally within the past 2–3 weeks, and almost invariably within
the past 100 days (Figs. 2 and 3).
This temporal profile of onset of HIT can be explained as follows: the
rapid fall in platelet count represents abrupt onset of platelet activation
caused by residual circulating HIT antibodies that resulted from the recent
heparin treatment, rather than antibodies newly generated by the subsequent
course of heparin.
This explanation is supported by other observations. First, for patients
with typical onset of HIT, there was no difference in its median day of onset,
irrespective of whether or not patients had previously been exposed to hepa-
rin. Second, patients did not generally develop thrombocytopenia that began
between days 2 and 4. Had there truly been an anamnestic immune response
more rapid than the usual 5- to 10-day period, one might have expected to
identify such a group of patients. Third, patients reexposed to heparin fol-
lowing disappearance of HIT antibodies do not necessarily form HIT anti-
60 Warkentin
Figure 3 A 49-year-old patient exhibiting both typical- and rapid-onset HIT: The
platelet count began to fall on day 6 of subcutaneous (sc) UFH injections given for
antithrombotic prophylaxis following neurosurgery (typical HIT). An abrupt fall in
platelet count occurred twice on day 18, each after a 5000 U intravenous (iv) UFH
bolus (rapid HIT). Symptoms and signs of acute systemic reaction occurred 10 minutes
after each bolus (dyspnea, tachypnea, hypertension, chest tightness, restlessness). Note
that the patient’s platelet count never fell below 150 109/L, even though her serum
tested strongly positive for HIT antibodies by serotonin release assay. She developed
proximal deep venous thrombosis (DVT) shortly after developing HIT.
bodies again; those who do appear to form antibodies after day 5 (Gruel et al.,
1990; Warkentin and Kelton, 2001a). Indeed, several patients with well-
documented previous HIT have received full treatment courses of heparin
several months or years later without incident (Warkentin and Kelton, 2001a;
Lindhoff-Last et al., 2002).
HIT Antibodies Are Transient

There is a plausible biological basis to explain why patients who develop
rapid-onset HIT have received heparin in the recent, rather than in the re-
mote, past: HIT antibodies are transient and become undetectable at a

median of 50 days (95% CI, 32–64 days) after first testing positive, using
the platelet serotonin release assay. The median time to a negative test is some-
what longer (85 days; 95% CI, 64–124 days) using a commercial antigen assay
(Fig. 4). At 100-day follow-up, the probability of the activation and antigen
assays being negative is approximately 90% and 60%, respectively (Warken-
tin and Kelton, 2001a).
Rule 2
A rapid fall in the platelet count soon after starting heparin therapy is
unlikely to represent HIT unless the patient has received heparin in the
recent past, usually within the past 100 days.
To summarize, the rapid fall in platelet count appears to be caused
by the repeat administration of heparin to a patient with residual circulating
HIT antibodies, rather than resulting from a rapid regeneration of HIT
antibodies.
Figure 4 Proportion of patients with HIT antibodies after an episode of HIT. The
time (in days) to a negative test by the activation assay (n = 144) or the antigen assay
(n = 93) is shown. The antigen test tended to become negative more slowly than did the
activation assay ( p = 0.007). (From Warkentin and Kelton, 2001a.)
62 Warkentin
A Hypothesis to Explain the Timing of HIT

There is a possible explanation for these unusual temporal features of HIT:
because the HIT antigen(s) is a ‘‘cryptic’’ autoantigen (or neoantigen)
comprised of two autologous substances (PF4 and heparin), HIT can be
regarded as an autoimmune disorder. Indeed, the target of the immune
response appears to be one of at least three dominant conformation-
dependent neoepitopes formed on PF4 when it binds to heparin (see Chaps.
6–8). There is evidence that transient IgG-mediated autoimmune responses
can occur, particularly when the responsible antibodies have relatively low
affinity for the neoepitope (thus having avoided prior clonal deletion as
occurs with lymphocytes that have high-affinity binding to autoantigens). In
this situation, the antibodies are generated only as long as the autoantigen is
present, thus explaining why there is a rapid fall in anti-PF4/H antibodies
soon after discontinuation of heparin. The affinity of the HIT antibodies may
be substantially enhanced when both Fab ‘‘arms’’ of the IgG molecule can
bind to linked epitopes, i.e., two PF4 molecules bound to a single heparin
molecule (Newman and Chong, 1999).
This hypothesis could explain several unusual aspects of the timing of
HIT, such as: (i) why HIT tends to occur fairly rapidly, beginning as soon as 5
days after starting heparin even in a patient who has never been exposed
previously to heparin (autoreactive T-cell or B-cell clones might already be
present in small numbers prior to starting heparin); (ii) why HIT occurs more
often in certain patient populations, such as postoperative patients (cytokine-
driven immune responses); and (iii) why HIT does not necessarily recur in
patients with a previous history of HIT who are subsequently treated with
heparin (there is a rapid loss of HIT antibodies following resolution of HIT,
and the specific circumstances that favored immune stimulation the first
time—e.g., large, stoichiometric concentrations of PF4 and heparin, occur-
ring in an inflammatory milieu—may not be recapitulated during the subse-
quent heparin exposure).
Implications for Repeat Use of Heparin in a Patient with a

History of HIT
The (1) transient nature of the HIT antibody, the (2) apparent minimum of 5
days to regenerate clinically significant HIT antibodies even in a patient who
once had HIT, and (3) the observation that HIT antibodies do not necessarily
recur, despite heparin rechallenge in a patient with definite prior HIT, all
suggest that it may be safe to readminister heparin to such patients. Fortu-
nately, this potentially risky situation is not frequently necessary, as there are
several alternative anticoagulants that can be substituted for heparin (see

Chaps. 13–19).
However, UFH is the unparalleled drug of choice in certain therapeutic
settings, particularly heart surgery when using cardiopulmonary bypass, or
vascular surgery. Furthermore, there are important disadvantages of newer
anticoagulants for these procedures (see Chap. 19). In my opinion, therefore,
for patients with a remote history of HIT (>100 days) who require cardiac or
vascular surgery, a rational approach is to prove serologically that HIT anti-
bodies are no longer present, and then to give heparin for a brief time to
permit the surgery (Olinger et al., 1984; Pötzsch et al., 2000; Warkentin and
Kelton, 2001a) (see also Chap. 19). We have even used this approach suc-
cessfully in a patient who required heparin for major vascular surgery 1
month following an episode of HIT, when the HIT antibodies had just
become undetectable. After surgery, it seems prudent to avoid postoperative
heparin completely and to administer an alternative anticoagulant, such as
danaparoid, lepirudin, or argatroban, as indicated. The actual risk of recur-
rent HIT beginning 5–10 days later, either following a transient intraoperative
heparin exposure or even during prolonged postoperative heparin use, is
unknown, but may be low.
For planning a brief reexposure to heparin in a patient who had HIT in
the past few weeks or months, a dilemma would arise if the follow-up patient
serum now tested negative using a sensitive activation assay (e.g., platelet
serotonin release assay), but positive by antigen assay. There is evidence that
activation assays are better at detecting clinically significant levels of HIT
antibodies (Warkentin et al., 2000) (see Chap. 11). Thus, use of heparin in this
situation might be a reasonable option, provided that one had confidence in
the activation assay performed, the antigen assay result was ‘‘weak’’ (e.g.,
0.400–0.750 OD units), there was a strong indication for surgery requiring
heparin, and there was limited experience with an alternative anticoagulant.
Continued watchful waiting is another option, given the transience of HIT
antibodies.
Sensitization by Incidental Heparin Exposure

Sensitizing exposures to heparin can be relatively obscure. For example, in-
cidental use of intraoperative line ‘‘flushes’’ that were not even documented
in the medical records has led to HIT antibody formation or acute onset of
HIT, with tragic consequences (Brushwood, 1992; Ling and Warkentin,
1998). Greinacher and colleagues (1992) reported a patient who developed
recurrent HIT when reexposed to heparin present in prothrombin complex
concentrates. Physicians should suspect possible heparin exposure in a patient
64 Warkentin
whose clinical course suggests HIT, especially if the patient was recently hos-
pitalized or has undergone procedures in which heparin exposure may have
occurred.
Delayed Onset of HIT

Rarely, HIT begins several days after discontinuing heparin therapy or
persists for several weeks even though heparin administration has been
stopped (Castaman et al., 1992; Tahata et al., 1992; Warkentin and Kelton,
2001b; Rice et al., 2002; Warkentin and Bernstein, 2003; Shah and Spencer,
2003) (see Fig. 5). A dramatic case encountered by the author was a female
outpatient who presented with transient global amnesia and a platelet count
of 40 109/L 7 days after receiving two doses of UFH; despite the diagnosis
and serologic confirmation of HIT and avoidance of all heparin, this patient’s
Figure 5 Delayed onset of HIT: a 68-year-old woman who received UFH for heart
surgery was noted to have a platelet count of 40 109/L on postoperative day 19 and a
‘‘rash’’ of her lower extremities. She presented on day 38 with symptomatic DVT and
developed rapid-onset recurrent thrombocytopenia after receiving iv UFH. The pa-
tient was successfully treated with danaparoid sodium (D.S.) and warfarin. In retro-
spect, the thrombocytopenia first observed on postoperative day 19 almost certainly
was caused by delayed onset of HIT.
platelet count fell over the next 4 days to 14 109/L, along with laboratory
evidence for DIC (low fibrinogen and elevated fibrin D-dimer levels). This
patient’s platelet counts gradually recovered to normal over several months,
during which time recurrent thrombotic events were managed successfully
with an alternative anticoagulant.
The unusual clinical course of these patients could be related to very
high titers of platelet-activating IgG antibodies (Warkentin and Kelton,
2001b). Moreover, substantial platelet activation in vitro can be caused by
some of these patients’ sera even in the absence of added heparin. This find-
ing of substantial heparin-independent platelet activation resembles that
described in other patients with drug-induced immune thrombocytopenia,
in which prolonged thrombocytopenia has been reported in association with
drug-independent binding of IgG to platelets (Kelton et al., 1981). Given the
apparent rarity of these cases, it is perhaps surprising that this syndrome does
not occur more frequently, given that HIT—once initiated—resembles some-
what an autoimmune disorder, with IgG recognizing an autologous protein,
PF4. On the other hand, earlier discharge from the hospital and a higher
index of suspicion for this syndrome might mean that delayed onset of HIT
will become a relatively more common presentation of HIT in the future.
Delayed onset of HIT, however, should not be confused with delayed
clinical manifestation of HIT-associated thrombosis. For example, Fig. 3
shows a patient who developed typical onset of HIT while receiving postop-
erative heparin prophylaxis. However, isolated HIT was not clinically
recognized, and the patient presented subsequently with a DVT and a normal
platelet count; when heparin boluses were given, rapid onset of thrombocy-
topenia occurred. Presumably, subclinical HIT-associated DVT that began
during the episode of isolated HIT progressed to symptomatic thrombosis in
the absence of anticoagulation. In contrast, patients with delayed onset of
HIT develop thrombocytopenia several days after the use of heparin and
are thrombocytopenic when they present with thrombosis. Exacerbation of
thrombocytopenia occurs if further heparin is given.
The existence of delayed onset of HIT presents a diagnostic dilemma in
patients who are no longer receiving heparin but who develop thrombocyto-
penia 5 or more days after placement of a heparin-coated device, e.g., certain
intravascular grafts or stents (Cruz et al., 2003). Such a puzzling situation of
delayed onset of thrombocytopenia post–vascular surgery prompted inves-
tigators to postulate heparin contamination of a graft (the manufacturer
insisted otherwise) (Bürger et al., 2001). In my view, either delayed onset or a
protracted course of thrombocytopenia could reflect the generation and
persistence of unusual ‘‘autoimmune’’ HIT antibodies without the need to
invoke continuing exposure to heparin.
66 Warkentin
B. Severity of Thrombocytopenia
Figure 6 shows the platelet count nadirs of 142 patients with laboratory-
proved HIT in one medical community: the median platelet count nadir was
approximately 60 109/L (Warkentin, 1998a). This contrasts with ‘‘typical’’
drug-induced immune thrombocytopenic purpura (DITP; e.g., caused by
quinine/quinidine, sulfa antibiotics, or rifampin [see Chap. 2]), for which the
median platelet count nadir is 15 109/L or less, and patients usually develop
bleeding (Pedersen-Bjergaard et al., 1997). The platelet count is 15 109/L or
Figure 6 Platelet count nadirs in 142 patients with serologically confirmed HIT: the
data are taken from a study of 127 patients with serologically confirmed HIT that used
a definition of <150 109/L (Warkentin and Kelton, 1996), together with a group of
15 patients diagnosed with serologically confirmed HIT over the same time period
whose platelet count nadir was >150 109/L. There is a lognormal distribution of the
platelet count nadirs, with a median platelet count of 59 109/L. HIT is only occa-
sionally complicated by very severe thrombocytopenia. HIT-associated thrombosis
occurred in most patients irrespective of the severity of the platelet count nadir. For
comparison, schematic platelet count nadir distributions are shown for typical drug-
induced immune thrombocytopenic purpura (DITP) and atypical DITP caused by
abciximab. (Modified from Warkentin, 1998a.)
fewer in only about 5% of patients with HIT (Warkentin, 2003). But even in
this minority of HIT patients with very severe thrombocytopenia, thrombo-
sis, rather than bleeding, predominates. Patients with atypical drug-induced
thrombocytopenic purpura caused by anti-GPIIb/IIIa-blocking drug (e.g.,
abciximab [ReoPro]) appear to develop severity of thrombocytopenia inter-
mediate between that of typical DITP and HIT (Fig. 6).
Definition of Thrombocytopenia
Figure 6 illustrates that HIT is associated with thrombosis even when the
platelet count nadir is more than 150 109/L. This suggests that the standard
definition of thrombocytopenia (<150 109/L) may be inadequate for many
patients with HIT. Particularly in postoperative patients, a major fall in the
platelet count can occur without the nadir falling to less than 150 109/L (see
Figs. 1b and 3). Indeed, studies indicate that a 50% or greater fall in the plate-
let count from the postoperative peak is strongly associated with HIT anti-
bodies, even when the platelet count nadir remains higher than 150 109/L
(Ganzer et al., 1997; Warkentin et al., 2003). Moreover, this patient subgroup
is at increased risk for thrombosis.
Rule 3
A platelet count fall of more than 50% from the postoperative peak
between days 5 and 14 after surgery associated with heparin treatment
can indicate HIT even if the platelet count remains higher than 150
109/L.
It is possible that a greater than 50% platelet count fall definition is also
appropriate for medical patients (Girolami et al., 2003). Regardless of the
patient population, a clinician should have a high index of suspicion when
unexpected large-percentage declines in the platelet count occur during hepa-
rin treatment, irrespective of whether an arbitrary absolute threshold for
‘‘thrombocytopenia’’ is crossed.
Platelet Count Monitoring in Patients Receiving Heparin

In postoperative patients, the onset of HIT coincides with rising platelet
counts (postoperative thrombocytosis); thus, the platelet count profile of
HIT resembles an ‘‘inverted V’’ (}; see Fig. 1a, b). The postoperative peak
platelet count preceding HIT is often higher than the preoperative platelet
count. Therefore, the postoperative peak platelet count is the appropriate
baseline for calculating the magnitude of a subsequent platelet count fall
(Table 2).
Table 2 Determining the Day of Onset of Thrombocytopenia: A 35-Year-Old Woman Who Developed HIT After Heart Surgery
Day 0
Preoperative (surgery) Day 1 Day 2 Day 3 Day 4 Day 5 Day 6 Day 7 Day 8 Day 9 Day 10 Day 11
Heparin UFH Line Nil Nil UFH 5000 UFH 5000 UFH 5000 UFH 5000 D.S D.S D.S D.S
used during flushes b.i.d. sc b.i.d. sc b.i.d. sc b.i.d. sc
CPB
Platelet 227 98 137 209 255 300 374 378 310 224 (PEa) 166 171 161 (nadir)
count
Percent Platelet fall during day 0–4 is unlikely to be HIT unless there was Rising Peak 18% 41% 56% No 57%
platelet recent heparin use (past 100 days) and the magnitude of the platelet platelet (378 ! (378 ! (378 ! further (378 !
count platelet fall is greater than expected. count count 310) 224) 166) fall 161)
fall
a
Pulmonary embolism (PE) occurred on postoperative day 8, in association with a platelet count fall of 41%, from 378 (postoperative peak) to 224 109/L.
The platelet count began to fall on day 7. The case illustrates why it is wrong to use the preoperative platelet count value as the ‘‘baseline,’’ as the fall in platelet
count from 227 (preoperative) to 224 (day 7) would be considered trivial, even though HIT-associated pulmonary embolism occurred. The preoperative and
first three postoperative days are in shaded boxes to indicate that these data should be censored in the interpretation of platelet counts in HIT. In this patient,
the abrupt fall in platelet count from 227 to 98 (day 0) is expected (heart surgery). This patient was treated successfully with danaparoid sodium (D.S.), with
longer-term anticoagulation with warfarin. The patient’s clinical course is also shown in Fig. 4B in Chap. 12.
Figure 1 Pathogenesis of HIT: a central role for thrombin generation. (See Fig.
13.1, p. 340 for full legend.)
Figure 2 Model of the interaction between argatroban and thrombin. (See Fig.
16.2, p. 440.)
Figure 3 (a) Model of the human platelet factor 4 tetramer. (b) AC dimer view of
the amino acids (‘‘ring of charge’’) crucial for heparin binding. [C-terminal a-helix
residues encompassing lysines 61–62 and 65–66 (cyan), arginines 20, 22, and 49
(green), and histidine 23, threonine 25, lysine 46 (yellow).] (See Fig. 7.2, p. 185, for
full legend. From a, Zhang et al., 1994; b, Loscalzo et al., 1985; Mayo et al., 1995a.)
Figure 4 (a) Heparin-induced erythematous plaques. (b) Heparin-induced skin

necrosis. (See Fig. 3.13a,b, p. 86, for full legend.)
Figure 5 Primary and secondary structure of platelet factor 4 (PF4) in relation to
HIT neoepitopes. (Top) 3-D representation of the PF4 tetramer, indicating two
neoepitope sites (per monomer). The ‘‘ring of positive charge’’ is formed by lysine
residues in the C-terminus (light blue) and other lysine and arginine residues (dark
blue). (Bottom) The linear sequence of the 70-amino acid polypeptide of a single PF4
molecule is shown. (From Li et al., 2002.) (See Fig. 7.4, p. 187.)
Figure 6 Ischemic limb syndromes in HIT. (a) Warfarin-induced venous limb
gangrene complicating HIT-associated deep vein thrombosis. (b) Warfarin-induced
digital necrosis complicating HIT and Raynaud’s phenomenon (from Warkentin et
al., 2004). (c) Livedo reticularis (thigh) and (d) patchy necrosis (foot) complication
HIT with hypofibrinogenemic DIC (no warfarin, no DVT). (See Figs. 3.9, p. 77,
Fig. 3.11, p. 80, and Fig. 3.12a,b, p. 82, for full legends.)
HIT-Associated Thrombosis Without Trombocytopenia

Anecdotal reports indicate that HIT-associated thrombosis can occur in the
absence of thrombocytopenia, as conventionally defined (Phelan, 1983;
Hach-Wunderle et al., 1994; Warkentin, 1996a, 1997; Houston, 2000). How-
ever, most of these patients do have an associated fall in the platelet count,
although the nadir remains higher than 150 109/L. Perhaps the most dra-
matic example of this phenomenon was a patient with essential thrombocy-
themia who developed serologically confirmed HIT: the platelet count fell by
49% from 1235 to 633, i.e., concomitant ‘‘thrombocytopenia’’ and thrombo-
cytosis (Risch et al., 2000).
A study suggested that HIT antibody formation without thrombocy-
topenia is not associated with a thrombosis rate greater than control patients
(Warkentin et al., 1995, 2003). However, the subset of patients who formed
HIT antibodies and whose platelet count fell by 50% or more—but remained
above 150 109/L—did have an increased risk for thrombosis (odds ratio,
6.0). Figure 7 illustrates this concept of the central importance of thrombo-
cytopenia (defined broadly as a large relative fall in the platelet count) in
determining risk for thrombosis. These observations provide indirect evi-
Figure 7 ‘‘Iceberg’’ model of HIT: Model A indicates that thrombosis occurs in

patients who develop HIT antibody formation and thrombocytopenia. This model is
supported by clinical data. In contrast, model B indicates the possibility of HIT
antibody formation contributing to thrombosis without the intermediary process of
thrombocytopenia. Although anecdotal experience suggests occasional patients con-
sistent with model B, controlled studies indicate that HIT antibody formation without
thrombocytopenia does not have an increased frequency of thrombosis, compared
with controls (Warkentin et al., 1995, 2003). Note that thrombocytopenia is broadly
defined and includes patients with large relative falls in the platelet count, even if the
platelet nadir is >150 109/L. (From Warkentin, 1999.)
70 Warkentin
dence suggesting that in vivo platelet activation by HIT antibodies probably

contributes to the pathogenesis of HIT-associated thrombosis.
Platelet Count Recovery Following Discontinuation of Heparin

The median time to platelet count recovery to more than 150 109/L after
stopping heparin administration is about 4 days, although several more days
may be required for the platelet count to reach a stable plateau. In patients
with very severe HIT, the platelet count may take 2 weeks or more to recover
(Warkentin, 1998a). Unlike nonimmune heparin-associated thrombocytope-
nia, the platelet count will generally not recover in patients with HIT unless
the heparin is discontinued.
III. THROMBOSIS
A. The HIT Paradox: Thrombosis but Not Hemorrhage
Table 1 summarizes the clinical spectrum and approximate frequency of
clinical sequelae associated with HIT. Spontaneous hemorrhage is not
characteristic of HIT, and petechiae are not typically observed, even in those
occasional patients whose platelet count is less than 10 109/L. Bleeding
complications were not increased over controls in two prospective studies of
HIT (Cipolle et al., 1983; Warkentin et al., 1995).
Rule 4
Petechiae and other signs of spontaneous bleeding are not clinical fea-
tures of HIT, even in patients with very severe thrombocytopenia.
The explanation for this clinical feature is unknown, but could be
related to unique pathophysiological aspects of HIT, such as in vivo platelet
activation, generation of procoagulant, platelet-derived microparticles,
and procoagulant alterations of endothelium and monocytes (see Chaps. 9
and 10).
B. HIT Is a Hypercoagulable State

A large controlled study (Warkentin et al., 1995, 2003) concluded that HIT is
independently associated with thrombosis, even in a patient population at
high baseline risk for thrombosis (postoperative orthopedic patients). More-
over, both venous and arterial thrombosis was seen. Thus, HIT can be
considered a hypercoagulable state (Table 3), a designation consistent with
increased in vivo thrombin generation seen in almost all patients with HIT
(Warkentin et al., 1997; Greinacher et al., 2000).
Table 3 The Prothrombotic Nature of HIT: Comparison with Other

Hypercoagulable States
Odds ratio
Hypercoagulable state for thrombosis
Heparin-induced thrombocytopenia:
Platelet count <150 109/L 36.9
Platelet fall >50% beginning z5 days of heparin 12.4
Platelet fall >50%, but platelet count remains >150 109/L 6.0
Factor V Leiden 6.6
Congenital protein C deficiency 14.4
Congenital protein S deficiency 10.9
Congenital antithrombin deficiency 24.1
Dysfibrinogenemia 11.3
Lupus anticoagulant 5.4
Source: Warkentin, 1995; Warkentin et al. 1995, 2003.
C. Timing of Thrombotic Complications

Thrombosis occurs in association with HIT in at least four ways. Only the
last three situations are conventionally considered as HIT-associated throm-
bosis. First, thrombosis can precede heparin treatment, for which it usually
represents the initial indication for heparin therapy. Second, HIT can be the
presenting clinical manifestation of HIT, often occurring early during the
platelet count fall. Indeed, new thrombosis is the initial clinical manifestation
in about half of all HIT patients (Warkentin and Kelton, 1996; Greinacher
et al., 1999) (see Fig. 1a).
Third, thrombosis can occur during the period of thrombocytopenia
or early platelet count recovery despite discontinuation of the heparin (dis-
cussed subsequently). Finally, thrombosis can occur following platelet count
recovery (Gallus et al., 1987; Warkentin and Kelton, 1996). In these patients,
it is possible that subclinical thrombosis occurred during the thrombocyto-
penia, but became clinically evident only later. The term heparin-induced
thrombocytopenia–thrombosis (syndrome), also known as HITT or HITTS, is
sometimes used to describe patients with HIT-associated thrombosis.
Natural History of ‘‘Isolated HIT’’

There is a high probability of subsequent thrombosis even when heparin
administration is stopped because of thrombocytopenia caused by HIT. A
retrospective cohort study (Warkentin and Kelton, 1996) identified 62
patients with serologically confirmed HIT in whom the diagnosis was clin-
72 Warkentin
ically suspected because of thrombocytopenia alone, and not because of

signs and symptoms indicative of possible new thrombosis. Thus, this co-
hort was identified without an apparent recognition bias caused by symp-
tomatic thrombosis. Nevertheless, the 30-day thrombosis event rate was
about 50% (see Fig. 2 in Chap. 4). This high frequency of thrombosis oc-
curred whether the heparin administration was simply stopped or substituted
by warfarin.
In 1999, Wallis and colleagues provided further support for this con-
cept that isolated HIT had an unfavorable natural history. In their retro-
spective cohort study of 113 patients with serologically confirmed HIT, these
workers also found a relatively high risk of thrombosis (23–38% at 30-day
follow-up, depending on whether patients who developed thrombosis at
the time heparin was stopped are included) in patients with isolated HIT
managed by cessation of heparin. Further, early cessation of heparin (within
48 h after a 50% or greater fall in platelet count) did not appear to reduce risk
of thrombosis, compared with patients in whom heparin was discontinued
later.
Meta-analysis of two prospective cohort studies also found a high
initial thrombotic event rate (6.1% per day after stopping heparin therapy
and before beginning alternative anticoagulant therapy with lepirudin
(Greinacher et al., 1999, 2000) (see Fig. 4 in Chap. 15). Taken together,
these large retrospective and prospective cohort studies suggest the follow-
ing rule:
Rule 5
HIT is associated with a high frequency of thrombosis despite discon-
tinuation of heparin therapy with or without substitution by coumarin:
the initial rate of thrombosis is about 5–10% per day over the first 1–2
days; the 30-day cumulative risk is about 50%.
About 5% of patients (3 of 62) in the largest study died suddenly, two with
proved or probable pulmonary embolism (Warkentin and Kelton, 1996).
This experience supports the recommendation that further anticoagulation
be considered for patients in whom isolated HIT has been diagnosed (Hirsh
et al., 1998, 2001; Warkentin et al., 1998) (see Chaps. 1 and 13–16).
D. Clinical Factors in the Pathogenesis of HIT-Associated

Thrombosis
Clinical factors help determine the location of thrombosis in HIT. For
example, Makhoul and colleagues (1986) observed prior vessel injury (e.g.,
recent angiography) in 19 of 25 patients with lower limb HIT-associated

thrombosis. Similarly, central venous catheters are crucial for the occurrence
of an upper limb DVT in patients with HIT (Hong et al., 2003).
Prospective studies of HIT in medical patients show that venous and
arterial thrombotic events occur in approximately equal numbers; in contrast,
there is a marked predominance of venous thrombosis when HIT occurs in
surgical patients (see Table 4 in Chap. 4). In a retrospective study, Boshkov
and colleagues (1993) found that HIT patients with cardiovascular disease
were more likely to develop arterial thrombosis, whereas venous thrombosis
was strongly associated with the postoperative state.
Rule 6
Localization of thrombosis in patients with HIT is strongly influenced
by independent acute and chronic clinical factors, such as the post-
operative state, atherosclerosis, or the location of intravascular catheters
in central veins or arteries.
E. Venous Thrombosis
Large case series suggest that venous thrombotic complications predominate
in HIT (Warkentin and Kelton, 1996; Nand et al., 1997) (see Table 4 in Chap.
4). Indeed, pulmonary embolism occurs more often than all arterial throm-
botic events combined. Furthermore, the strength of association between
HIT and venous thromboembolism increases in relation to the severity of
thrombosis (Table 4). Other unusual venous thrombotic events complicating
HIT include cerebral vein (dural sinus) thrombosis (v.i.), hepatic vein throm-
bosis (Theuerkauf et al., 2000), and perhaps retinal vein thrombosis (Nguyen
et al., 2003). Thus:
Rule 7
In patients receiving heparin, the more unusual or severe a subsequent
thrombotic event, the more likely the thrombosis is caused by HIT.
Regardless of the severity of thrombosis, in any patient who develops a
symptomatic venous or arterial thrombosis while receiving heparin, the plate-
let count should be measured to evaluate whether HIT could be present.
Lower Limb DVT

Lower limb DVT is the most frequent thrombotic manifestation of HIT.
Many venous thrombi are extensive and are often bilateral (see Table 4).
74
Table 4 Association of HIT and Thrombosis
Thrombosis rate in:

Odds ratio
Patient population (Ref.) Thrombosis HIT Controls (95% CI) p-Value
Post–orthopedic Proximal DVT 8/18 (44.4%) 26/647 (4.0%) 19.1 (5.9–58.3) <0.001
surgerya (Warkentin Bilateral proximal DVT 2/18 (11.1%) 4/647 (0.6%) 20.1 (1.7–150) 0.01
et al., 1995, 2003) Pulmonary embolism 2/18 (11.1%) 2/647 (0.3%) 40.3 (2.7–572) 0.004
Any thrombosis 13/18 (72.2%) 112/647 (17.3%) 12.4 (4.0–45.2) <0.001
Patients with central lineb Upper-limb DVT 14/145 (9.7%) 3/484 (0.6%) 17.1 (4.9–60.5) <0.001
(Hong et al., 2003)
Medicala (Girolami et al., Any thrombosis 3/5 (60%) 21/593 (3.5%) 40.8 (5.2–163) <0.001
2003)
a
HIT defined as >50% platelet count fall.
b
HIT defined as any abnormal platelet count fall with positive HIT serology (platelet fall was >50% in 93% of study patients).
Warkentin
Sometimes the DVT is sufficiently severe on clinical grounds as to merit use of

the term ‘‘phlegmasia cerulea dolens’’ (i.e., an inflamed, blue, painful limb).
However, progression of phlegmasia to venous limb gangrene is rare in the
absence of coumarin anticoagulation (discussed subsequently).
There is slight left-sided predominance involving lower limb DVT: we
found that 76/137 (56%) of lower limb DVT complicating HIT involved the
left lower limb (Hong et al., 2003), a similar proportion as in control patients
(57%). A slight left-sided predominance (f55 vs. f45%) for lower limb DVT
has also been noted in non-HIT populations (Kerr et al., 1990; Markel et al.,
1992). This is attributed to the left iliac vein crossing the left iliac artery,
causing an increase in left-sided lower limb venous pressures. Pregnancy
amplifies further this phenomenon, thus explaining the marked predomi-
nance (>95%) of left lower limb DVT in pregnancy (Ginsberg et al., 1992).
Upper Limb DVT

Upper limb DVT is relatively common in HIT, occurring in about 5% of
patients with HIT (Hong et al., 2003). Notably, in these patients the upper
limb DVT occurred at the site of a current or recent central venous catheter.
Most (86%) of the patients therefore had right upper limb DVT complicating
HIT, reflecting strong physician preference to using the right neck veins for
insertion of central lines. This study suggests that a systemic hypercoagulable
state (HIT) interacts with a local factor (location of central lines) to result in
clinical events (upper limb DVT).
Recurrence of Venous Thromboembolism

Gallus and colleagues (1987) identified HIT as a significant risk factor for
recurrence of venous thromboembolism in a prospective treatment study: 3 of
the 9 patients with HIT developed recurrent venous thromboembolism,
compared with 12 of the 223 patients in whom HIT was not diagnosed (odds
ratio, 8.8; p < 0.01).
Warfarin-Induced Venous Limb Gangrene

Venous limb gangrene is one of two clinical syndromes associated with HIT
in which coumarin anticoagulation paradoxically plays an important path-
ogenic role (Fig. 8). Venous limb gangrene is defined as acral (extremity)
necrosis that occurs in a limb affected by DVT. Additional features include
(1) absence of large artery occlusion (i.e., there are palpable or doppler-
76 Warkentin
Figure 8 Coumarin-induced skin necrosis (CISN): HIT is associated with two forms
of CISN: (1) venous limb gangrene, affecting extremities with active deep vein throm-
bosis, and (2) ‘‘classic’’ CISN, which involves central (nonacral) tissues, such as breast,
abdomen, thigh, flank, and leg, among other tissue sites. CISN complicating HIT
typically manifests as venous limb gangrene (f90%) (Warkentin et al., 1997, 1999),
whereas CISN in other settings most commonly affects central tissues (f90%) (Cole
et al., 1988). (From Warkentin, 1996b.)
identifiable pulses); (2) extensive thrombotic occlusion of large and small

veins, as well as venules; and (3) the characteristic hallmark of a supra-
therapeutic international normalized ratio (INR), generally >4.0. (Fig. 9).
Anticoagulation with warfarin, phenprocoumon, or other coumarins is
a crucial factor to explain the progression of DVT to venous limb gangrene
(Warkentin, 1996b; Warkentin et al., 1997). A case–control study of 8 patients
with HIT-associated venous limb gangrene found a higher median INR, com-
pared with 58 control HIT patients treated with warfarin for DVT who did
not develop venous gangrene (5.8 vs. 3.1; p < 0.001). Laboratory studies
Figure 9 Warfarin-associated venous limb gangrene. Progression of DVT to acral

necrosis (leading to below-the-knee amputation) occurred despite the presence of pal-
pable arterial foot pulses in this 49-year-old woman with HIT treated with warfarin
(INR = 7.2 at the onset of limb gangrene). (From Warkentin et al., 1997.) (See color
insert, Fig. 6a.)
showed a characteristic hemostatic profile for patients with venous gangrene:

persisting in vivo thrombin generation (elevated thrombin–antithrombin
complex levels), together with reduced protein C activity (Fig. 10). The high
INR is a surrogate marker for severely reduced protein C (through parallel
coumarin-induced reduction in factor VII). Thus, venous limb gangrene
appears to result from a profound disturbance in procoagulant-anticoagulant
balance.
The association between venous limb gangrene and HIT was first re-
ported by Towne and colleagues (1979). They noted a prodrome of phleg-
masia cerulea dolens before progression to distal gangrene (information on
possible coumarin treatment was not given). Other reports of venous limb
gangrene complicating HIT, however, do suggest that warfarin had been used
during the evolution to necrosis (Thomas and Block, 1992; Hunter et al.,
1993; Kaufman et al., 1998).
Patients have also developed venous limb gangrene during combined
treatment with both ancrod and warfarin (Warkentin et al., 1997; Gupta
et al., 1998); because thrombin generation increases during treatment of
HIT with ancrod (Warkentin, 1998b; see Fig. 2 in Chap. 13), ancrod could
predispose to a greater risk for venous gangrene during warfarin treatment.
Recently, several patients have been reported who developed venous
limb gangrene during the transition to coumarin from parenteral anticoagu-
78 Warkentin
Figure 10 Thrombin-antithrombin (TAT) complexes compared with protein C

activity in patients with HIT: Each data point represents TAT complexes and protein
C activity per single treatment day per patient. In both panels the open symbols
represent three patients with warfarin-induced venous limb gangrene and one patient
with phlegmasia cerulea dolens (open squares). The diagonal line represents an
arbitrary ratio of TAT complex to protein C of 400. (Left) Results when HIT was first
diagnosed and before warfarin therapy. Control samples included 8 patients (closed
circles) who subsequently received warfarin for DVT without developing venous limb
gangrene and 14 patients without DVT who did not later receive warfarin. (Right)
Results in 16 patients who were receiving warfarin for HIT, including 4 patients (open
symbols) who developed venous limb gangrene/phlegmasia and 12 patients (closed
circles) who received warfarin without developing venous limb gangrene. The data
suggest that patients who develop venous limb gangrene or phlegmasia have a higher
ratio of TAT to protein C, consistent with a disturbance in procoagulant–anti-
coagulant balance during warfarin treatment of HIT. (From Warkentin et al., 1997.)
lation with a direct thrombin inhibitor (lepirudin or argatroban) (Smythe

et al., 2002; Srinivasan et al., 2003). Typically, patients had symptomatic DVT
in the affected limb and had their direct thrombin inhibitor started and
stopped while they remained thrombocytopenic. Additionally, the INR was
supratherapeutic at the time that limb ischemia or gangrene occurred after
stopping the direct thrombin inhibitor. This experience indicates that the
transition from parenteral anticoagulation to coumarin therapy should pro-
ceed cautiously, as suggested by the following ‘‘rule’’ (see also Chap. 13):
Rule 8
Venous limb gangrene is characterized by (1) in vivo thrombin gen-
eration associated with acute HIT; (2) active DVT in the limb(s) affected
by venous gangrene; and (3) a supratherapeutic INR during coumarin
anticoagulation. This syndrome can be prevented by (1) delaying

initiation of coumarin anticoagulation during acute HIT until there has
been substantial recovery of the platelet count (to at least 100–150
109/L) while receiving an alternative parenteral anticoagulant (e.g.,
lepirudin, argatroban, danaparoid), and only if the thrombosis has
clinically improved; (2) initiating coumarin in low, maintenance doses
(e.g., 2–5 mg warfarin); (3) ensuring that both parenteral and oral
anticoagulant overlap for at least 5 days, with at least the last 2 days in
the target therapeutic range; and (4) if applicable, physicians should
reverse coumarin anticoagulation with vitamin K in a patient recognized
with acute HIT after coumarin therapy has been commenced.
The frequency of venous limb gangrene in HIT patients with DVT who
receive warfarin is unknown. This complication happened in 8 of 66 (12.1%;
95% CI 5.4–22.5%) patients with HIT-associated DVT treated with warfarin
(with or without ancrod) in Hamilton; venous limb gangrene was a more
frequent cause of limb loss in HIT patients than was arterial occlusion in this
medical community. Venous gangrene also occurred in 1 of 21 (4.8%; 95% CI
0.12–23.8%) patients treated with phenprocoumon in Germany (Greinacher
et al., 2000). In contrast, this complication was not observed by Wallis and
colleagues in any of 51 patients who received warfarin with a diagnosis of
HIT, although only 16 patients received warfarin to manage HIT-associated
thrombosis (95% CI for 0/16 0–20.6%). Besides cotherapy with ancrod,
factors that could influence the risk for venous gangrene include the dosing of
coumarin, the rate of coagulation factor turnover/consumption related to
DIC, and vitamin K deficiency.
Rarely, coumarin therapy contributes to microvascular thrombosis and
acral limb ischemia in the absence of DVT. Figure 11 (also see color insert,
Fig. 6b) shows multiple digital necrosis of the right hand complicating the
initiation of warfarin therapy (maximal INR = 4.3) in a patient with
Raynaud’s phenomenon who developed HIT following aortic valve replace-
ment for adenocarcinoma-associated noninfective thrombotic endocarditis
(Warkentin et al., 2004). Although digital necrosis occurred in all four limbs
in this patient, only the right foot (which exhibited the greatest amount of
ischemic necrosis) was found to have DVT by duplex ultrasonography. It was
hypothesized that microcirculatory disturbances secondary to paraneoplastic
Raynaud’s phenomenon interacted with altered procoagulant–anticoagulant
balance (secondary to HIT and warfarin therapy) to cause this dramatic
clinical syndrome.
Cerebral Venous (Dural Sinus) Thrombosis

Thrombosis of the dural venous sinuses is an unusual cause of stroke in
HIT patients that was first reported by Stevenson (1976). Often, there is a
80 Warkentin
Figure 11 Warfarin-associated multiple digital necrosis of the right hand in a 61-

year-old woman with paraneoplastic Raynaud’s phenomenon and adenocarcinoma-
associated thrombotic endocarditis who developed HIT following aortic valve re-
placement surgery (see text for additional clinical details). (From Warkentin et al.,
2004.) (See color insert, Fig. 6b.)
second hypercoagulable state, such as pregnancy (Van der Weyden et al.,

1983; Calhoun and Hesser, 1987) or myeloproliferative disease (Kyritsis et
al., 1990), that may have interacted with HIT to cause this complication.
Platelet-rich ‘‘white clots’’ were identified in the superior sagittal venous
sinus in one necropsy study (Meyer-Linderberg et al., 1997). Clinicians
should have a high index of suspicion for dural sinus thrombosis when a
patient develops progressive focal neurological signs, decreased level of
consciousness, seizures, or headache during or soon after stopping heparin
treatment (Beland et al., 1997; Pohl et al., 1999, 2000; Warkentin and
Bernstein, 2003). Treatment includes immediate discontinuation of heparin,
use of an alternative anticoagulant, and possibly, intravenous gammaglo-
bulin (see Chap. 13).
Adrenal Hemorrhagic Infarction

Clinicians should suspect bilateral adrenal hemorrhagic infarction when
thrombocytopenic patients develop abdominal pain and hypotension in
association with heparin treatment (Arthur et al., 1985; Dahlberg et al.,
1990; Ernest and Fisher, 1991; Delhumeau and Granry, 1992; Bleasel et al.,
1992; Kovacs et al., 2001). Fever and hyponatremia occur in some patients.
These patients require corticosteroid replacement to prevent death from acute
or chronic adrenal failure (Rowland et al., 1999). Unilateral adrenal hemor-
rhagic infarction typically presents with ipsilateral flank pain without signs of
adrenal failure (Warkentin, 1996a). HIT explained at least 5% of patients
with adrenal hemorrhage at one institution (Vella et al., 2001).
This hemorrhagic manifestation of HIT is caused by thrombosis of
adrenal veins leading to hemorrhagic necrosis of the glands. Other hy-
percoagulable states associated with adrenal necrosis include dissem-
inated intravascular coagulation (DIC) complicating meningococcemia
(Waterhouse-Friderichsen syndrome) and the antiphospholipid antibody
syndrome (McKay, 1965; Carette and Jobin, 1989).
DIC and Acquired Anticoagulant Deficiency

Although increased thrombin generation occurs in virtually all patients with
HIT, decompensated DIC, defined as reduced fibrinogen levels or an otherwise
unexplained increase in the INR, is relatively uncommon, occurring in about
5–10% of patients (Natelson et al., 1969; Klein and Bell, 1974; Zalcberg et al.,
1983; Castaman et al., 1992; Betrosian et al., 2003). Protein C consumption is
also well compensated, as protein C levels are usually within the normal range
when HIT is diagnosed (Warkentin et al., 1997).
Nevertheless, acquired natural anticoagulant failure from DIC could
contribute to thrombosis in some patients with HIT. Markedly reduced anti-
thrombin levels were found in a young woman with three-limb DVT and
bilateral adrenal infarction complicating HIT; following recovery, antithrom-
bin levels were normal (unpublished observations of the author). This hypo-
thesis implies that plasmapheresis could benefit patients by correcting
acquired anticoagulant deficiency; if so, the replacement fluid must be plasma,
rather than albumin, to correct antithrombin and other natural anticoagulant
deficiencies.
Other patients with HIT-associated DIC evince clinical signs of micro-
vascular thrombosis. For example, Figure 12 shows livedo reticularis and
patchy foot necrosis (despite palpable foot pulses) in a postoperative
cardiac surgery patient with HIT (platelet count nadir, 39 109/L)
complicated by hypofibrinogenemic DIC. Evidence for acquired natural
anticoagulant failure included mildly reduced antithrombin levels (0.76 U/
mL; normal, 0.77–1.30 U/mL) and moderately reduced protein C activity
(0.50 U/mL; normal, 0.70–1.80 U/mL) that subsequently resolved. Free
protein S levels were normal (1.12 U/mL; normal, 0.62–1.38 U/mL).
Evidence for DIC included a fibrinogen of 1.2 g/L (normal, 1.5–4.0 g/L)
that rose to 4.7 g/L one week later during therapeutic-dose danaparoid
therapy, a strongly positive protamine sulfate paracoagulation assay (4+
reactivity at 15 min; normal, no reactivity), a fibrin D-dimer level that was
greater than 2000 Ag/L (normal,<500 Ag/L), and the presence of red cell
82 Warkentin
Figure 12 Clinical manifestations of DIC. (a) Livedo reticularis. (b) Patchy ischemic
necrosis of right foot. This 70-year-old woman developed HIT-associated DIC with
hypofibrinogenemia, elevated INR, and reduced antithrombin and protein C activity
levels 9 days after emergency cardiac surgery for cardiac catheterization-associated
dissection of the left main coronary artery (see text for additional clinical information).
(See color insert, Figs. 6c and 6d.)
fragments. Additionally, the INR was elevated at 1.6 (normal, 0.9–1.2),

even though coagulation factors VII, V, X, and II all measured between
0.73 to 0.83 U/mL (normal, 0.50–1.50 U/mL). The anticoagulant treatment
was successful in avoiding limb amputation. In my experience, limb
ischemia and necrosis associated with DIC that occurs in the absence of
large artery thrombotic occlusion or warfarin therapy is the least common
explanation for limb loss in HIT.
Livedo reticularis is also discussed on pp. 88–89.
Congenital Hypercoagulability and HIT-Associated Thrombosis

Gardyn and associates (1995) reported a patient with fatal HIT and wide-
spread microvascular thrombosis. The investigators identified heterozygous
factor V Leiden (G1691A mutation) in this patient, and they speculated that
this contributed to the severe clinical course. However, the complications may
also have been related to the treatment with low molecular weight heparin
(LMWH) and warfarin.
The interaction between factor V Leiden and thrombotic sequelae of
HIT was formally investigated in a study of 165 patients with HIT, 16 (9.7%)
of whom had factor V Leiden (Lee et al., 1998). No increase in the number or
severity of venous or arterial thrombosis was seen. This result is not sur-
prising, as thrombosis occurs in about 50–75% of patients with HIT (War-
kentin and Kelton, 1996). Thus, even if the most common congenital
hypercoagulable disorders, factor V Leiden and the prothrombin G20210A
mutation (each occurring in about 5% of the population), were strongly as-
sociated with increased risk for thrombosis in HIT, only a few HIT-associated
thromboses could thereby be explained.
Carlsson and colleagues (2003) studied 142 patients with HIT (79 with
thrombosis) to determine whether any of 10 established or putative platelet re-
ceptor or clotting factor polymorphisms (including factor V Leiden and pro-
thrombin G20210A mutation) was associated with thrombosis. None was
found.
Lindhoff-Last et al. (2002) also found no association between factor V
Leiden or prothrombin G20210A mutation and thrombosis in a smaller study
of 21 patients. However, they found that more HIT patients had elevated
factor VIII levels (at mean 29-month follow-up) than matched normal
controls (16/21 vs. 4/21). The significance of this finding is unclear.
F. Arterial Thrombosis
Lower limb artery thrombosis was the first recognized complication of HIT
(Weismann and Tobin, 1958; Roberts et al., 1964; Rhodes et al., 1973, 1977).
84 Warkentin
Arterial thrombosis most commonly involves the distal aorta (e.g., saddle
embolism) or the large arteries of the lower limbs, leading to acute limb
ischemia with absent pulses. Sometimes, platelet-rich thromboemboli from
the left heart or proximal aorta explain acute lower limb arterial ischemia
(Vignon et al., 1996). Other arterial thrombotic complications that are
relatively common in HIT include acute thrombotic stroke and myocardial
infarction. The relative frequency of arterial thrombosis in HIT by location,
namely, lower limb artery occlusion >> stroke syndrome > myocardial in-
farction (Benhamou et al., 1985; Kappa et al., 1987; Warkentin and Kelton,
1996; Nand et al., 1997), is reversed from that observed in the non-HIT
population (myocardial infarction > stroke syndrome >> lower limb artery
occlusion).
Uncommon but well-described arterial thrombotic events in HIT in-
clude mesenteric artery thrombosis (bowel infarction), brachial artery throm-
bosis (upper limb gangrene), and renal artery thrombosis (renal infarction).
Multiple arterial thrombotic events are quite common, as are recurrences
following surgical thromboembolectomy, especially if further heparin is given
during or after surgery. Occasionally, microembolization of thrombus orig-
inating from the heart or aorta causes foot or toe necrosis with palpable arte-
rial pulses.
Angiographic Appearance
Lindsey and colleagues (1979) reported a distinct angiographic appearance of
heparin-induced thromboembolic lesions, described as ‘‘broad-based, isolat-
ed, gently lobulated excrescences which produced 30–95% narrowing of the
arterial lumen. The abrupt appearance of such prominent luminal contour
deformities in arterial segments that were otherwise smooth and undistorted
was unexpected and impressive. . . . In each case, the lesions were located
proximal to sites of arterial occlusion.’’ The radiologic and surgical experi-
ence described suggests that distal embolization of ‘‘white’’ clots composed of
‘‘platelet-fibrin aggregates’’ accounted for the limb ischemia.
G. Graft, Prosthetic Device, and Extracorporeal Circuit

Thrombosis
Heparin-induced thrombocytopenia predisposes to thrombosis of blood in
contact with native or prosthetic grafts or vascular fistulae, valve or other
intravascular prostheses, as well as extracorporeal circuits (Towne et al., 1979;
Silver et al., 1983; Bernasconi et al., 1988; AbuRahma et al., 1991; Lipton and
Gould, 1992; Hall et al., 1992). This presents serious management problems
in certain situations, such as renal hemodialysis (see Chap. 18). Clinicians
should check for unexpected platelet count declines, and test for HIT anti-
bodies, in patients who develop thrombosis of grafts, prostheses, or other
devices during heparin treatment.
IV. MISCELLANEOUS COMPLICATIONS OF HIT

A. Heparin-Induced Skin Lesions at Subcutaneous
Injection Sites
Clinical Picture
Skin lesions that occur at the site(s) of subcutaneous heparin injection are a
manifestation of the HIT syndrome. For unknown reasons, only 10–20% of
patients who form HIT antibodies during subcutaneous UFH or LMWH
treatment develop these lesions. Furthermore, about 75% of patients who
develop heparin-induced skin lesions do not develop thrombocytopenia, even
though heparin-dependent, platelet-activating HIT antibodies are readily de-
tectable (Warkentin, 1996a, 1997).
The skin abnormalities range in appearance from indurated, erythem-
atous nodules or plaques (Fig. 13a) to frank necrotizing lesions (Fig. 13b) (see
color insert, Fig. 4) that start 5 or more days (median, day 8) after beginning
heparin injections (Hasegawa 1984; MacLean et al., 1990; Wütschert et al.,
1999). The lesions can occur earlier if there was recent treatment with heparin
given by another route that resulted in formation of HIT antibodies. Some
erythematous plaques have an eczematous appearance. Necrotic lesions ty-
pically consist of a central black eschar surrounded by a cuff of induration and
erythema (Fig. 13b). Complex skin lesions can result—for example, several
discrete areas of necrosis (each lesion corresponding to a different heparin
injection site), each with a surrounding violaceous halo, with all circum-
scribed by a diffuse erythema. Even the least severe forms of heparin-induced
skin lesions usually cause pain or pruritus.
Both UFH and LMWH can cause these reactions. Patients who develop
UFH-induced skin lesions generally will develop further lesions if LMWH is
substituted for the UFH (Bircher et al., 1990). In contrast, it is uncommon for
danaparoid to cause skin lesions in these patients.
Histopathology
Lymphocyte infiltration of the upper and middermis that can extend into the
epidermis characterizes the erythematous plaque (Bircher et al., 1990).
Dermal and epidermal edema (spongiosis) is observed in lesions that appear
eczematous. The T lymphocytes of helper–suppressor (CD4+) phenotype
86 Warkentin
Figure 13 Heparin-induced skin lesions. (a) Heparin-induced erythematous

plaques: UFH injections into the lower abdomen resulted in painful erythematous
plaques beginning on day 7 of sc UFH treatment; at this time, the platelet count fell
only by 9% from 340 to 311 109/L. HIT antibody seroconversion from a negative
baseline was shown using the serotonin release assay (from 0% to 84% serotonin
release). (From Warkentin, 1996a.) (b) Heparin-induced skin necrosis: UFH injections
into the right anterior thigh led to skin necrosis: a large black eschar with irregular
borders is surrounded by a narrow band of erythema. The platelet count fell to 32
109/L; despite stopping heparin, the patient developed symptomatic proximal DVT 10
days later. (See color insert, Fig. 4.)
predominate, together with CD1+/DR+ dendritic (Langerhans) cells, con-

sistent with a type IV delayed hypersensitivity immune response. Cytokine
synthesis by activated CD4 cells could explain the peripheral blood eosino-
philia that has been reported in a few patients (Bircher et al., 1994). In con-
trast, histopathology of lesions associated with cutaneous necrosis usually
shows intravascular thrombosis of dermal vessels, with or without perivas-
cular inflammation and red cell extravasation of variable degree (Hall et al.,
1980; Kearsley et al., 1982; Cohen et al., 1988; MacLean et al., 1990; Balestra
et al., 1994).
Management
Heparin-induced skin lesions should be considered a marker for the HIT
syndrome. Platelet count monitoring, if not already being performed, should
be initiated and continued for several days, even after stopping heparin
administration. The reason is that some patients develop a fall in platelet
count, together with thrombosis (often affecting limb arteries), that begins
several days after stopping the heparin (Warkentin, 1996a, 1997). An alter-
native anticoagulant, such as danaparoid, lepirudin, or argatroban, should be
given, particularly in patients whose original indication for anticoagulation
still exists or who develop progressive thrombocytopenia. The skin lesions
themselves should be managed conservatively whenever possible, although
some patients require debridement of necrotic tissues followed by skin graft-
ing (Hall et al., 1980).
Rule 9
Erythematous or necrotizing skin lesions at heparin injection sites should
be considered dermal manifestations of the HIT syndrome, irrespective
of the platelet count, unless proved otherwise. Patients who develop
thrombocytopenia in association with heparin-induced skin lesions are
at increased risk for venous and, especially, arterial thrombosis.
B. Classic Coumarin-Induced Skin Necrosis

Classic coumarin-induced skin necrosis (CISN) is a very rare complication
of oral anticoagulant therapy (Cole et al., 1988). In its classic form, it is
characterized by dermal necrosis, usually in a central (nonacral) location,
such as breast, abdomen, thigh, or leg, that begins 3–6 days after starting
therapy with warfarin or other coumarin anticoagulants (see Fig. 8). Initially,
there is localized pain, induration, and erythema that progresses over hours to
central purplish-black skin discoloration and blistering, ultimately evolving
to well-demarcated, full-thickness necrosis involving skin and subdermal tis-
sues. Some patients require surgical debridement. Case reports suggest that
88 Warkentin
congenital deficiency of natural anticoagulant proteins, especially protein C,

is sometimes a pathogenic factor (Broekmans et al., 1983; Comp, 1993).
There is evidence that HIT also predisposes to classic CISN (Celoria
et al., 1988; Cohen et al., 1989; Warkentin et al., 1999; Srinivasan et al., 2003).
Theoretically, this could result from increased consumption of anticoagulant
factors, thereby leading to greater reduction in protein C in the setting of
increased thrombin generation in HIT (Tans et al., 1991; Warkentin et al.,
1997). However, central lesions of CISN seem less likely to complicate HIT
than the related syndrome of coumarin-induced venous limb gangrene (War-
kentin et al., 1997, 1999). Perhaps active DVT in HIT localizes the progressive
microvascular thrombosis to acral tissues already affected by extensive
venous thrombosis.
C. Other Heparin-Associated Skin Lesions

Skin Necrosis in the Absence of Coumarin Therapy
Other patients have developed skin lesions during intravenous heparin ther-
apy, or at locations otherwise distant from subcutaneous injection sites, in
the absence of coumarin therapy. Hartman and colleagues (1988) reported
a man who received intravenous heparin for saphenous vein thrombosis:
the platelet count fell from 864 to 44 109/L (day 10). On day 7, when the
platelet count had fallen by 33% to 575 109/L, progressive necrosis of
skin in the thigh at the region of the thrombosed vein occurred, necessi-
tating surgical excision. Thrombosis of veins and capillaries, with arterial
sparing, was noted. Balestra et al. (1994) reported a patient who developed
thrombocytopenia (75 109/L) and skin necrosis of the thigh on day 9 of
subcutaneous injections of LMWH given into the lower abdominal wall.
A skin biopsy showed small vessel thrombosis with a mild inflammatory
reaction.
Other Skin Lesions Associated with Heparin Treatment

Livedo Reticularis. The bluish, reticulated (network-like), mottled
appearance of livedo reticularis was reported in a patient with HIT
complicating intravenous UFH given for atrial fibrillation after heart surgery
(Gross et al., 1993). This patient also had DIC, microangiopathic peripheral
blood abnormalities, and fibrin thrombi noted within small dermal vessels.
The livedo appearance results from microvascular thrombosis, with slowing of
blood flow and dilation of the horizontally oriented dermal venous drainage
channels (Copeman, 1975). Figure 12a (see p. 82) shows livedo reticularis
associated with HIT and DIC.
Urticaria and Other Miscellaneous Lesions. Other dermatological con-

sequences of heparin treatment do not appear to be related to HIT. These
range from common lesions (ecchymosis) to rare effects of intravenous hepa-
rin, such as vasculitis (Jones and Epstein, 1987) and cutaneous necrosis with
hemorrhagic bullae (Kelly et al., 1981). Some patients have developed wide-
spread urticarial lesions, sometimes accompanied by angioedema, during
treatment with subcutaneous or intravenous heparin (Odeh and Oliven, 1992;
Patriarca et al., 1994). In one patient skin testing suggested a generalized
reaction against the preservative chlorbutol (Dux et al., 1981). Although
LMWH injections were claimed to have caused distal extremity dermal
lesions in a patient with HIT (Payne and Kovacs, 2003), it is possible these
were related to concomitant warfarin therapy.
D. Acute Systemic Reactions Following an Intravenous

Heparin Bolus
Acute systemic reaction (ASR) refers to a variety of symptoms and signs that
characteristically begin 5–30 min after an intravenous heparin bolus is given
to a patient with circulating HIT antibodies (Nelson et al., 1978; Warkentin
et al., 1992, 1994; Popov et al., 1997; Ling and Warkentin, 1998; Warkentin,
2002) (Table 5; see Fig. 3). Only about one quarter of at-risk patients who
receive a heparin bolus develop such a reaction. The most common signs and
symptoms are fever and chills, hypertension, and tachycardia. Less common
are flushing, headache, chest pain, dyspnea, tachypnea, and large-volume
diarrhea. In some patients, severe dyspnea is the predominant sign, termed
‘‘pseudo-pulmonary embolism’’ by Popov and colleagues (1997); multiple
Table 5 Clinical Features of Acute Systemic Reactions Following Intravenous

Heparin Bolus
Timing: onset 5–30 min after intravenous heparin bolus

Clinical context: recent use of heparin (past 5–100 days)
Laboratory features: abrupt, reversible fall in the platelet count
Signs and symptoms:
Inflammatory: chills, rigors, fever, flushing
Cardiorespiratory: tachycardia, hypertension, tachypnea, dyspnea, chest pain or
tightness, cardiopulmonary arrest (rare)
Gastrointestinal: nausea, vomiting, diarrhea
Neurological: headache, transient global amnesia (rare)
90 Warkentin
small perfusion defects on radionuclide lung scans can be shown (Nelson

et al., 1978; Ling and Warkentin, 1998). Fatal cardiac and respiratory arrest
has been reported (Ansell et al., 1986; Platell and Tan, 1986; Hewitt et al.,
1998).
An abrupt fall in the platelet count invariably accompanies these re-
actions. However, the platelet count drop is often transient (see Fig. 1A in
Chap. 4). Thus, physicians should determine the platelet count immediately
on suspecting the diagnosis and test for HIT antibodies. Heparin must be
discontinued, as further use can lead to fatal complications (Ling and War-
kentin, 1998).
Rule 10
Any inflammatory, cardiopulmonary, or other unexpected acute event
that begins 5–30 minutes after an intravenous heparin bolus should be
considered acute HIT unless proved otherwise. The postbolus platelet
count should be measured promptly and compared with prebolus levels,
because the platelet count fall is abrupt and often transient.
The clinical features of postheparin bolus ASR are not typical of IgE-
mediated anaphylaxis (i.e., urticaria, angioedema, and hypotension are not
seen). Rather, the syndrome resembles febrile transfusion reactions com-
monly observed after platelet transfusions, suggesting a common pathogen-
esis of proinflammatory cytokines associated with cellular activation (Heddle
et al., 1994). Moreover, there are similarities between ASR and the admin-
istration of ADP in humans, including acute dyspnea, tachycardia, and tran-
sient thrombocytopenia (Davey and Lander, 1964).
A few patients have developed acute, transient impairment of antero-
grade memory (i.e., the ability to form new memories) following an intrave-
nous heparin bolus in association with acute HIT (Warkentin et al., 1994;
Pohl et al., 2000). This syndrome resembles that of transient global amnesia,
a well-characterized neurological syndrome of uncertain pathogenesis.
E. Heparin Resistance
Difficulty in maintaining therapeutic anticoagulation despite increasing hep-
arin dosage, or heparin resistance, is a common finding in patients with HIT-
associated thrombosis (Rhodes et al., 1977; Silver et al., 1983). Possible
explanations include neutralization of heparin by PF4 released from activated
platelets (Padilla et al., 1992) or pathophysiological consequences of platelet-
derived microparticles (Bode et al., 1991). Heparin resistance is not specific
for HIT, however, and occurs in many patients with extensive thrombosis of
various etiologies (e.g., cancer).
V. SPECIAL CLINICAL SITUATIONS

A. Cardiac and Neurological Complications of HIT
Although HIT can affect almost any organ system, some clinical specialties
observe a wider spectrum of thrombotic and other sequelae of HIT. Table 6
lists complications encountered in cardiology and neurology.
B. HIT in Pregnancy
Heparin-induced thrombocytopenia has complicated UFH treatment given
for venous thromboembolism complicating pregnancy (Van der Weyden
et al., 1983; Meytes et al., 1986; Copplestone and Oscier, 1987; Greinacher
et al., 1992) or the postpartum period (Calhoun and Hesser, 1987). HIT seems
to be rare in this patient population; no pregnant patients have been
diagnosed with HIT over a 20-year period in Hamilton. Plasma glycosami-
noglycans are increased during pregnancy (Andrew et al., 1992), which could
contribute to lower frequency or pathogenicity of HIT antibodies. HIT
antibodies cross the placenta (Greinacher et al., 1993), so it is at least
theoretically possible that a heparin-treated newborn delivered from a mother
with acute HIT could develop this drug reaction.
Pregnant patients with HIT have developed unusual events, such as
cerebral dural sinus thrombosis (Van der Weyden et al., 1983; Calhoun and
Hesser, 1987). Treatment options for pregnant patients with life-threatening
thrombosis include danaparoid or fondaparinux as these drugs do not cross
the placenta (see Chaps. 13 and 14). The more benign syndrome of heparin-
induced skin lesions without thrombocytopenia has also been reported in
pregnant patients (Drouet et al., 1992). Danaparoid was reported to be
effective in a patient who developed LMWH-induced skin lesions (de Saint-
Blanquat et al., 2000).
C. HIT in Children and Neonates

There are anecdotal reports of HIT occurring in children, some as young as 3
months of age (Laster et al., 1987; Oriot et al., 1990; Potter et al., 1992;
Murdoch et al., 1993; Klement et al., 1996; Butler et al., 1997; Ranze et al.,
1999, 2001) (see Chap. 20). However, not all of these patients underwent
confirmatory testing with specific diagnostic assays. HIT in children has a
similar, often dramatic clinical course, as is seen in adults. The frequency of
HIT in the pediatric population is unknown.
The frequency and clinical import of HIT in neonates receiving heparin
in intensive care settings is controversial. Spadone and colleagues (1992) in-
vestigated 34 newborn infants (average gestational age, 29 weeks) who devel-
92 Warkentin
Table 6 Cardiological and Neurological Complications of HIT
Cardiological complications
Myocardial infarction (Rhodes et al., 1973; Van der Weyden et al., 1983)
Occlusion of saphenous vein grafts post–coronary artery bypass surgerya
Intra-atrial thrombus (left and rightb heart chambers) (Scheffold et al., 1995;
Olbrich et al., 1998)
Intraventricular thrombus (left and rightb heart chambers) (Commeau et al., 1986;
Dion et al., 1989; Vignon et al., 1996)
Prosthetic valve thrombosis (Bernasconi et al., 1988; Vazquez-Jimenez et al., 1999)
Right heart failure secondary to massive pulmonary embolism
Cardiac arrest postintravenous heparin bolus (Ansell et al., 1986; Platell and Tan,
1986; Hewitt et al., 1998)
Neurological complications
Stroke syndrome
In situ thrombosis
Progressive stroke in patients receiving heparin for treatment of stroke
(Ramirez-Lassepas et al., 1984)
Cardiac embolization (Scheffold et al., 1995)
Cerebral vein (dural venous sinus) thrombosis (Van der Weyden et al., 1983;
Kyritsis et al., 1990; Meyer-Lindenberg et al., 1997; Warkentin and Bernstein,
2003); complicating pregnancy (Calhoun and Hesser, 1987)
Amaurosis fugax (Theuerkauf et al., 2000)
Ischemic lumbosacral plexopathy (Jain, 1986)
Paraplegia, transient (Maurin et al., 1991) or permanent (Feng et al., 1993),
associated with distal aortic thrombosis
Transient global amnesia (Warkentin et al., 1994)
Headachec
a
Thrombosis preferentially affects saphenous vein grafts rather than internal mammary artery
grafts (Liu et al., 2002; Ayala et al., 2002).
b
Although adherent thrombi that likely developed in situ have been reported (Dion et al.,
1989), emboli originating from limb veins can explain right-sided intra-atrial or intra-
ventricular clots.
c
Headache as a feature of HIT is suggested by (1) its occurrence in patients with acute systemic
reactions post–heparin bolus (see Fig. 1a, Chap. 4) and (2) its concurrence with onset of
thrombocytopenia in several patients who developed HIT in a clinical trial (unpublished ob-
servations of the author).
oped thrombocytopenia or thrombosis, beginning an average of 22 days after

starting heparin therapy. Platelet aggregation studies suggested the presence
of HIT antibodies in 41% of these neonates. Aortic thrombosis complicating
umbilical artery catheter use was the most common complication. Another
group (Butler et al., 1997), also using platelet aggregation studies, reported a
neonate who may have developed fatal HIT shortly after birth. More specific
activation or antigen assays were not performed in either study, however. A
recent study of 108 neonates who received UFH flushes found no HIT anti-
bodies using a sensitive antigen assay (Klenner et al., 2003).
D. HIT in Bone Marrow Transplantation

Given the widespread use of heparin to maintain patency of indwelling cath-
eters, it is surprising that there are few reports of HIT in patients undergoing
intensive anticancer chemotherapy. Two reports describe patients with
apparent HIT complicating allogeneic or autologous marrow or stem cell
transplantation (Tezcan et al., 1994; Sauer et al., 1999). Subclavian vein
thrombosis occurred in one patient. It is possible that the combination of
intensive chemotherapy and treatment-induced thrombocytopenia reduces
the likelihood of HIT antibody formation or clinical expression of HIT.
There is an intriguing report of a man recently recovered from HIT
who was about to receive autologous marrow transplantation. When his
marrow was collected into heparin anticoagulant, substantial ex vivo throm-
bus formation occurred, preventing adequate cell collection (Bowers and
Jones, 2002).
VI. ESTIMATING THE PRETEST PROBABILITY OF HIT

A. Scoring Systems for HIT
Various scoring systems to estimate the probability of HIT based upon
clinical information have been published, usually for the purpose of evaluat-
ing new laboratory tests for HIT (Greinacher et al., 1994; Pouplard et al.,
1999; Alberio et al., 2003). These systems have included the platelet count
recovery following heparin cessation, which limits their applicability for
judging the clinical likelihood of HIT in ‘‘real time’’ when a thrombocytope-
nic patient receiving heparin is first evaluated. Further, these scoring systems
were developed before various features of the timing and severity of platelet
count fall in HIT were understood.
B. The ‘‘
‘‘Four Ts’’
’’
A new scoring system, the ‘‘4 Ts,’’ has been developed that takes advantage of
new information regarding the clinical features of HIT (Warkentin, 2003;
Warkentin and Heddle, 2003). Platelet count recovery is not a criterion,
because this information often is not available at initial evaluation, or heparin
may not have been stopped. For simplicity, four clinical features are assessed,
given scores of 0, 1, or 2 (Table 7). Thus, the maximal total score is 8.
94
Table 7 Estimating the Pretest Probability of HIT: The ‘‘Four Ts’’
Points (0, 1, or 2 for each of 4 categories: maximum possible score = 8)a
2 1 0
9
Thrombocytopenia (acute) Nadir, 20–100 (at least Nadir, 10–19 10 /L or Nadir, <10 109/L
30% fall); or any >50% any 30–50% fall (or or any <30% fall
fall (nadir z20) >50% fall associated
with heart surgery)
Timingb of platelet count Clear onset between days Consistent with day 5–10 Platelet count fall V4
fall, thrombosis, or other 5–10 or V1 day (if fall, but not clear (e.g., days without recent
sequelae (first day of heparin heparin exposure within missing platelet counts) or heparin exposure
course = day zero) past 30 days) V1 day (heparin exposure
within past 31–100 days)
or platelet fall after day 10
Thrombosis or other sequelae New thrombosis; skin Progressive or recurrent None
(e.g., skin lesions, ASR) necrosis; ASR after iv thrombosis; erythematous
heparin bolus skin lesions; suspected
thrombosis (not yet proven);
asymptomatic upper-limb
DVT
Other cause of No explanation (besides HIT) Possible other cause Definite other cause is
thrombocytopenia for platelet count fall is evident is evident present
not evident
Abbreviations: ASR, acute systemic reaction (see Table 5); DVT, deep venous thrombosis.
a
Pretest probability score: 6–8 = high; 4–5 = intermediate; 0–3 = low.
b
First day of immunizing heparin exposure considered day zero; the day the platelet count begins to fall is considered the day of onset of
thrombocytopenia (it generally takes 1–3 more days until an arbitrary threshold that defines thrombocytopenia is passed). The scoring system
shown here has undergone minor modifications from previously published scoring systems (Warkentin, 2003; Warkentin and Heddle, 2003).
Warkentin
Estimated pretest probabilities of HIT thereby range from low (0–3) to high
(6–8), with an intermediate score (4–5) indicating moderate risk.
Maximal scores for each category are given when the clinical features
are highly consistent with HIT. Thus, a patient will score 8 if there is a
substantial fall in the platelet count that begins 5–10 days after commencing
heparin, together with thrombosis, and where no other plausible cause is
apparent during clinical assessment. Even a patient with no clinical evidence
of thrombosis can be assigned a high pretest probability (score 6 of 8) if the
clinical features are otherwise consistent with HIT. Another feature of this
system is that very low platelet count values (i.e., 10–19 and<10 109/L)
score only 1 and 0 points, respectively, thus reducing the chance that a patient
with posttransfusion purpura (PTP) or DITP will be misclassified as HIT and
inappropriately given anticoagulant therapy.
C. Clinical Use of a Scoring System

A practical use of the scoring system is to help make initial clinical decisions
regarding therapy. Based on preliminary evaluation, we believe it is likely that
a low score (0–3) is associated with a very low risk (<5%) of clinically sig-
nificant HIT antibodies (defined arbitrarily as >50% serotonin release in a
washed platelet activation assay) (see Chap. 11). In contrast, a high score
(6–8) appears to be associated with a high risk (>80%) of such strong HIT
antibodies. Further, 50–75% of patients evaluated for clinical HIT will have
low or high scores. This leaves a smaller number of patients in whom the
clinical suspicion of HIT is more uncertain (score 4–5) and in whom the results
of diagnostic testing will be especially important for supporting (or refuting)
the diagnosis of HIT.
REFERENCES
AbuRahma AF, Boland JP, Witsberger T. Diagnostic and therapeutic strategies of

white clot syndrome. Am J Surg 1991; 162:175–179.
Alberio L, Kimmerle S, Baumann A, Taleghani BM, Biasiutti FD, Lämmle B. Rapid
determination of anti-heparin/platelet factor 4 antibody titers in the diagnosis of
heparin-induced thrombocytopenia. Am J Med 2003; 114:528–536.
Anderson KC, Kihajda FP, Bell WR. Diagnosis and treatment of anticoagulant-
related adrenal hemorrhage. Am J Hematol 1981; 11:379–385.
Andrew M, Mitchell L, Berry L, Paes B, Delorme M, Ofosu F, Burrows R, Khambalia
B. An anticoagulant dermatan sulfate proteoglycan circulates in the pregnant
woman and her fetus. J Clin Invest 1992; 89:321–326.
96 Warkentin
Ansell JE, Clark WP Jr, Compton CC. Fatal reactions associated with intravenous
heparin [letter]. Drug Intell Clin Pharm 1986; 20:74–75.
Arthur CK, Grant SJB, Murray WK, Isbister JP, Stiel JN, Lauer CS. Heparin-
associated acute adrenal insufficiency. Aust NZ J Med 1985; 15:454–455.
Ayala E, McDonough RF, Morgensztern D, Kharfan-Dabaja MA, Byrnes JJ. Heparin-
induced thrombocytopenia presenting with thrombosis of multiple saphenous
vein grafts (abstr). Blood 2002; (suppl):3894.
Balestra B, Quadri P, Demarmels Biasiutti F, Furlan M, Lämmle B. Low molecular
weight heparin-induced thrombocytopenia and skin necrosis distant from in-
jection sites. Eur J Haematol 1994; 53:61–63.
Beland B, Busse H, Loick HM, Ostermann H, van Aken H. Phlegmasia cerulea dolens,
cerebral venous thrombosis, and fatal pulmonary embolism due to heparin-
induced thrombocytopenic thrombosis syndrome. Anesth Analg 1997; 85:1272–
1274.
Benhamou AC, Gruel Y, Barsotti J, Castellani L, Marchand M, Guerois C, Leclerc
MH, Delahousse B, Griguer P, Leroy J. The white clot syndrome or heparin-
associated thrombocytopenia and thrombosis (WCS or HATT). Int Angiol
1985; 4:303–310.
Bernasconi F, Metivet F, Estrade G, Garnier D, Donatien Y. Thrombose d’une pros-
thèse valvulaire mitrale au cours d’une thrombopénie induite par l’héparine.
Traitement fibrinolytique Presse Méd 1988; 17:1366.
Betrosian AP, Theodossiades G, Lambroulis G, Kostantonis D, Balla M, Papani-
kolaou M, Georgiades G. Heparin-induced thrombocytopenia with pulmonary
embolism and disseminated intravascular coagulation associated with low-
molecular-weight heparin. Am J Med Sci 2003; 325:45–47.
Bircher AJ, Flückiger R, Buchner SA. Eczematous infiltrated plaques to subcutaneous
heparin: a type IV allergic reaction. Br J Dermatol 1990; 123:507–514.
Bircher AJ, Itin PH, Büchner SA. Skin lesions, hypereosinophilia, and subcutaneous
heparin [letter]. Lancet 1994; 343:861.
Bleasel JF, Rasko JEJ, Rickard KA, Richards G. Acute adrenal insufficiency sec-
ondary to heparin-induced thrombocytopenia–thrombosis syndrome. Med J
Aust 1992; 157:192–193.
Bode AP, Castellani WJ, Hodges ED, Yelverton S. The effect of lysed platelets on
neutralization of heparin in vitro with protamine as measured by the activated
coagulation time (ACT). Thromb Haemost 1991; 66:213–217.
Boshkov LK, Warkentin TE, Hayward CPM, Andrew M, Kelton JG. Heparin-
induced thrombocytopenia and thrombosis: clinical and laboratory studies. Br J
Haematol 1993; 84:322–328.
Bowers MJ, Jones FGC. Thrombus in harvested marrow from a patient with recent
heparin-induced thrombocytopenia. Br J Haematol 2002; 119:294.
Broekmans AW, Bertina RM, Loeliger EA, Hofmann V, Klingemann HG. Protein C
and the development of skin necrosis during anticoagulant therapy [letter].
Thromb Haemost 1983; 49:251.
Brushwood DB. Hospital liable for allergic reaction to heparin used in injection flush.
Am J Hosp Pharm 1992; 49:1491–1492.
Bürger T, Tautenhahn J, Böck M, Fahlke J, Halloul Z, Lippert H. Can a coated
Dacron vascular graft maintain a heparin-induced thrombocytopenia type II?

Langenbecks Arch Surg 2001; 386:267–271.
Butler TJ, Sodoma LJ, Doski JJ, Cheu HW, Berg ST, Stokes GN, Lancaster KJ.
Heparin-associated thrombocytopenia and thrombosis as the cause of a fatal
thrombus on extracorporeal membrane oxygenation. J Pediatr Surg 1997; 32:
768–771.
Calhoun BC, Hesser JW. Heparin-associated antibody with pregnancy: discussion of
two cases. Am J Obstet Gynecol 1987; 156:964–966.
Carette S, Jobin F. Acute adrenal insufficiency as a manifestation of the anticardiolipin
syndrome? Ann Rheum Dis 1989; 48:430–431.
Carlsson LE, Lubenow N, Blumentritt C, Kempf R, Papenberg S, Schröder W,
Eichler P, Herrmann FH, Santoso S, Greinacher A. Platelet receptor and clot-
ting factor polymorphisms as genetic risk factors for thromboembolic compli-
cations in heparin-induced thrombocytopenia. Pharmacogenetics 2003; 13:253–
258.
Castaman G, Ruggeri M, Girardello R, Rodeghiero F. An unusually prolonged case
of heparin-induced thrombocytopenia and disseminated intravascular coagu-
lation. Haematologica 1992; 77:174–176.
Celoria GM, Steingart RH, Banson B, Friedmann P, Rhee SW, Berman JA. Coumarin
skin necrosis in a patient with heparin-induced thrombocytopenia—a case re-
port. Angiology 1988; 39:915–920.
Cipolle RJ, Rodvoid KA, Seifert R, Clarens R, Ramirez-Lassepas M. Heparin-
associated thrombocytopenia: a prospective evaluation of 211 patients. Ther
Drug Monit 1983; 5:205–211.
Cohen GR, Hall JC, Yeast JD, Field-Kriese D. Heparin-induced cutaneous necrosis
in a postpartum patient. Obstet Gynecol 1988; 72:498–499.
Cohen DJ, Briggs R, Head HD, Acher CW. Phlegmasia cerulea dolens and its as-
sociation with hypercoagulable states: case reports. Angiology 1989; 40:498–
500.
Cole MS, Minifee PK, Wolma FJ. Coumarin necrosis—a review of the literature.
Surgery 1988; 103:271–277.
Commeau P, Grollier G, Charbonneau P, Troussard X, Lequerrec A, Bazin C, Potier
JC. Thrombopénie immuno-allergique induite par l’héparine responsable d’une
thrombose intraventriculaire gauche. Therapie 1986; 41:345–347.
Comp PC. Coumarin-induced skin necrosis. Incidence, mechanisms, management and
avoidance. Drug Safety 1993; 8:128–135.
Copeman PWM. Livedo reticularis. Signs in the skin of disturbance of blood viscosity
and of blood flow. Br J Dermatol 1975; 93:519–522.
Copplestone A, Oscier DG. Heparin-induced thrombocytopenia in pregnancy [letter].
Br J Haematol 1987; 65:248.
Cruz D, Karlsberg R, Takano Y, Vora D, Tobis J. Subacute stent thrombosis asso-
ciated with a heparin-coated stent and heparin-induced thrombocytopenia.
Cathet Cardiovasc Intervent 2003; 58:80–83.
Dahlberg PJ, Goellner MH, Pehling GB. Adrenal insufficiency secondary to adrenal
hemorrhage. Two case reports and a review of cases confirmed by computed
tomography. Arch Intern Med 1990; 150:905–909.
98 Warkentin
Davey MG, Lander H. Effect of adenosine diphosphate on circulating platelets in

man. Nature 1964; 201:1037–1039.
Delhumeau A, Granry JC. Heparin-associated thrombocytopenia [letter]. Crit Care
Med 1992; 20:1192.
de Saint-Blanquat L, Simon L, Toubas MF, Hamza J. Traitement par le danaparoı̈de
de sodium au cours de la grossesse chez une patiente présentant une allergie
cutanée aux héparines de bas poids moléculaire. Ann Fr Anesth Réanim 2000;
19:751–754.
Dion D, Dumesnil JG, LeBlanc P. In situ right ventricular thrombus secondary to
heparin induced thrombocytopenia. Can J Cardiol 1989; 5:308–310.
Drouet M, Le Pabic F, Le Sellin J, Bonneau JC, Sabbah A. Allergy to heparin. Spe-
cial problems set by pregnant women. Allergol Immunopathol 1992; 20:225–
229.
Dux S, Pitlik S, Perry G, Rosenfeld JB. Hypersensitivity reaction to chlorbutol-
preserved heparin [letter]. Lancet 1981; 1:149.
Ernest D, Fisher MM. Heparin-induced thrombocytopaenia complicated by bilateral
adrenal haemorrhage. Intensive Care Med 1991; 17:238–240.
Feng WC, Singh AK, Bert AA, Sanofsky SJ, Crowley JP. Perioperative paraplegia and
multiorgan failure from heparin-induced thrombocytopenia. Ann Thorac Surg
1993; 55:1555–1557.
Gallus AS, Goodall KT, Tillett J, Jackaman J, Wycherley A. The relative contri-
butions of antithrombin III during heparin treatment, and of clinically recog-
nisable risk factors, to early recurrence of venous thromboembolism. Thromb
Res 1987; 46:539–553.
Ganzer D, Gutezeit A, Mayer G, Greinacher A, Eichler P. Thromboembolieprophy-
laxe als Auslöser thrombembolischer Komplikationen. Eine Untersuchung zur
Inzidenz der Heparin-induzierten Thrombozytopenie (HIT) Typ II. Z Orthop
Ihre Grenzgeb 1997; 135:543–549.
Gardyn J, Sorkin P, Kluger Y, Kabili S, Klausner JM, Zivelin A, Eldor A. Heparin-
induced thrombocytopenia and fatal thrombosis in a patient with activated
protein C resistance. Am J Hematol 1995; 50:292–295.
Ginsberg JS, Brill-Edwards P, Burrows RF, Bona R, Prandoni P, Büller HR, Lensing
A. Venous thrombosis during pregnancy: leg and trimester of presentation.
Girolami B, Prandoni P, Stefani PM, Tanduo C, Sabbion P, Eichler P, Ramon R,
Baggio G, Fabris F, Girolami A. The incidence of heparin-induced thrombo-
cytopenia in hospitalized medical patients treated with subcutaneous unfrac-
tionated heparin: a prospective cohort study. Blood 2003; 101:2955–2959.
Greinacher A, Michels I, Mueller-Eckhardt C. Heparin-associated thrombocytope-
nia: the antibody is not heparin specific. Thromb Haemost 1992; 67:545–549.
Greinacher A, Eckhardt T, Mussmann J, Mueller-Eckhardt C. Pregnancy complicated
by heparin associated thrombocytopenia: management by a prospectively in
vitro selected heparinoid (Org 10172). Thromb Res 1993; 71:123–126.
Greinacher A, Amiral J, Dummel V, Vissac A, Kiefel V, Mueller-Eckhardt C. Lab-
oratory diagnosis of heparin-associated thrombocytopenia and comparison of
platelet aggregation test, heparin-induced platelet activation test, and platelet
factor 4/heparin enzyme-linked immunosorbent assay. Transfusion 1994; 34:

381–385.
Greinacher A, Völpel H, Janssens U, Hach-Wunderle V, Kemkes-Matthes B, Eichler
P, Mueller-Velten HG, Pötzsch B, for the HIT Investigators Group.
Recombinant hirudin (lepirudin) provides safe and effective anticoagulation
in patients with heparin-induced thrombocytopenia: a prospective study.
Circulation 1999; 99:73–80.
bocytopenia with thromboembolic complications: meta-analysis of 2 prospec-
tive trials to assess the value of parenteral treatment with lepirudin and its
Gross AS, Thompson FL, Arzubiaga MC, Graber SE, Hammer RD, Schulman G,
Ellis DL, King LE Jr. Heparin-associated thrombocytopenia and thrombosis
(HATT) presenting with livedo reticularis. Int J Dermatol 1993; 32:276–279.
Gruel Y, Lang M, Darnige L, Pacouret G, Dreyfus X, Leroy J, Charbonnier B. Fatal
effect of reexposure to heparin after previous heparin-associated thrombocy-
topenia and thrombosis [letter]. Lancet 1990; 336:1077–1078.
Gruel Y, Pouplard C, Nguyen P, Borg JY, Derlon A, Juhan-Vague I, Regnault V,
Samama M. Biological and clinical features of low-molecular-weight heparin-
induced thrombocytopenia. Br J Haematol 2003; 121:786–792.
Gupta AK, Kovacs MJ, Sauder DN. Heparin-induced thrombocytopenia. Ann Phar-
macother 1998; 32:55–59.
Hach-Wunderle V, Kainer K, Krug B, Müller-Berghaus G, Pötzsch B. Heparin-
associated thrombosis despite normal platelet counts [letter]. Lancet 1994;
344:469–470.
Hall AV, Clark WF, Parbtani A. Heparin-induced thrombocytopenia in renal failure.
Clin Nephrol 1992; 38:86–89.
Hall JC, McConahay D, Gibson D. Heparin necrosis. An anticoagulation syndrome.
JAMA 1980; 244:1831–1832.
Hartman AR, Hood RM, Anagnostopoulos CE. Phenomenon of heparin-induced
thrombocytopenia associated with skin necrosis. J Vasc Surg 1988; 7:781–
784.
Hasegawa GR. Heparin-induced skin lesions. Drug Intell Clin Pharm 1984; 18:313–
314.
Heddle NM, Klama L, Singer J, Richards C, Fedak P, Walker I, Kelton JG. The role
of the plasma from platelet concentrates in transfusion reactions. N Engl J Med
1994; 331:625–628.
Hewitt RL, Akers DL, Leissinger CA, Gill JI, Aster RH. Concurrence of anaphylaxis
and acute heparin-induced thrombocytopenia in a patients with heparin-
induced antibodies. J Vasc Surg 1998; 28:561–565.
Hirsh J, Warkentin TE, Raschke R, Granger C, Ohman EM, Dalen JE. Heparin
and low-molecular-weight heparin. Mechanisms of action, pharmacokinetics,
dosing considerations, monitoring, efficacy, and safety. Chest 1998; 114(suppl):
489S–510S.
Granger C, Ohman EM, Dalen JE. Heparin and low-molecular-weight heparin:
100 Warkentin
mechanisms of action, pharmacokinetics, dosing, monitoring, efficacy, and

Hong AP, Cook DJ, Sigouin CS, Warkentin TE. Central venous catheters and upper-
extremity deep-vein thrombosis complicating immune heparin-induced throm-
bocytopenia. Blood 2003; 101:3049–3051.
Houston DS. Heparin-induced thrombocytopenia without thrombocytopenia in a pa-
tient with essential thrombocythemia [letter]. Am J Hematol 2000; 65:331–
332.
Hunter JB, Lonsdale RJ, Wenham PW, Frostick SP. Heparin-induced thrombosis: an
important complication of heparin prophylaxis for thromboembolic disease in
surgery. Br Med J 1993; 307:53–55.
Jain A. Ischemic lumbosacral plexus neuropathy secondary to possible heparin-
induced thrombosis following aortoiliac bypass [abstr]. Arch Phys Med Rehabil
1986; 67:680.
Jones BF, Epstein MT. Cutaneous heparin necrosis associated with glomeruloneph-
ritis. Australas J Dermatol 1987; 28:117–118.
Kappa JR, Fisher CA, Berkowitz HD, Cottrell ED, Addonizio VP Jr. Heparin-
induced platelet activation in sixteen surgical patients: diagnosis and manage-
ment. J Vasc Surg 1987; 5:101–109.
Kaufman BR, Zoldos J, Bentz M, Nyström NÅ. Venous gangrene of the upper
extremity. Ann Plast Surg 1998; 40:370–377.
Kearsley JH, Jeremy RW, Coates AS. Leukocytoclastic vasculitis and skin necrosis
following subcutaneous heparin calcium. Aust NZ J Med 1982; 12:288–289.
Kelly RA, Gelfand JA, Pincus SH. Cutaneous necrosis caused by systemically ad-
ministered heparin. JAMA 1981; 246:1582–1583.
Kelton JG, Meltzer D, Moore J, Giles AR, Wilson WE, Barr R, Hirsh J, Neame PB,
Powers JP, Walker I, Bianchi F, Carter CJ. Drug-induced thrombocytopenia is
associated with increased binding of IgG to platelets both in vivo and in vitro.
Blood 1981; 58:524–529.
Kerr TM, Cranley JJ, Johnson JR, Lutter KS, Riechmann GC, Cranley RD, True
MA, Sampson M. Analysis of 1084 consecutive lower extremities involved with
acute venous thrombosis diagnosed by duplex scanning. Surgery 1990; 108:520–
527.
King DJ, Kelton JG. Heparin-associated thrombocytopenia. Ann Intern Med 1984;
100:535–540.
Klein HG, Bell WR. Disseminated intravascular coagulation during heparin therapy.
Ann Intern Med 1974; 80:477–481.
Klement D, Rammos S, von Kries R, Kirschke W, Kniemeyer HW, Greinacher A.
Heparin as a cause of thrombus progression. Heparin-associated thrombocy-
topenia is an important differential diagnosis in paediatric patients even with
normal platelet counts. Eur J Pediatr 1996; 155:11–14.
Klenner A, Fusch C, Rakow A, Beyersdorff E, Wander K, Lietz T, Eichler P,
Greinacher A. Benefit and risk of heparin addition in maintaining peripheral
venous catheters in neonates–a prospective double-blind, placebo-controlled
trial (abstr). J Thromb Haemost 2003; 1(suppl 1):OC423.
Kovacs KA, Lam YM, Pater JL. Bilateral massive adrenal hemorrhage. Assessment
of putative risk factors by the case-control method. Medicine (Balt) 2001;
80:45–53.
Kyritsis AP, Williams EC, Schutta HS. Cerebral venous thrombosis due to heparin-
induced thrombocytopenia. Stroke 1990; 21:1503–1505.
Laster J, Cikrit D, Waler N, Silver D. The heparin-induced thrombocytopenia syn-
drome: an update. Surgery 1987; 102:763–770.
Lee DH, Warkentin TE, Denomme GA, Lagrotteria DD, Kelton JG. Factor V Leiden
and thrombotic complications in heparin-induced thrombocytopenia. Thromb
Haemost 1998; 79:50–53.
Lindhoff-LastE,WenningB,SteinM, GerdsenF,BauersachsR,WagnerR.Riskfactors
and long-term follow-up of patients with the immune type of heparin-induced
thrombocytopenia. Clin Appl Thrombosis/Hemostasis 2002; 8:347–352.
Lindsey SM, Maddison FE, Towne JB. Heparin-induced thromboembolism: angio-
graphic features. Radiology 1979; 131:771–774.
Ling E, Warkentin TE. Intraoperative heparin flushes and acute heparin-induced
thrombocytopenia. Anesthesiology 1998; 89:1567–1569.
Lipton ME, Gould D. Case report: heparin-induced thrombocytopenia—a compli-
cation presenting to the vascular radiologist. Clin Radiol 1992; 45:137–138.
Liu JC, Lewis BE, Steen LH, Grassman ED, Bakhos M, Blakeman B, Wrona L, Leya
F. Patency of coronary artery bypass grafts in patients with heparin-induced
thrombocytopenia. Am J Cardiol 2002; 89:979–981.
Lubenow N, Kempf R, Eichner A, Eichler P, Carlsson LE, Greinacher A. Heparin-
induced thrombocytopenia: temporal pattern of thrombocytopenia in relation
to initial use or reexposure to heparin. Chest 2002; 122:37–42.
MacLean JA, Moscicki R, Bloch KJ. Adverse reactions to heparin. Ann Allerg 1990;
65:254–259.
Makhoul RG, Greenberg CS, McCann RL. Heparin-associated thrombocytopenia
and thrombosis: a serious clinical problem and potential solution. J Vasc Surg
1986; 4:522–528.
Markel A, Manzo RA, Bergelin RO, Strandness DE Jr. Pattern and distribution of
thrombi in acute venous thrombosis. Arch Surg 1992; 127:305–309.
Maurin N, Biniek R, Heintz B, Kierdorf H. Heparin-induced thrombocytopenia and
thrombosis with spinal ischaemia—recovery of platelet count following a
change to a low molecular weight heparin [letter]. Intensive Care Med 1991;
17:185–186.
McKay DG. Late manifestations of intravascular coagulation—tissue necrosis. In:
McKay DG, ed. Disseminated Intravascular Coagulation. An Intermediary
Mechanism of Disease. New York: Harper & Row, 1965:392–471.
Meyer-Lindenberg A, Quenzel E-M, Bierhoff E, Wolff H, Schindler E, Biniek R. Fata
cerebral venous sinus thrombosis in heparin-induced thrombotic thrombocy-
topenia. Eur Neurol 1997; 37:191–192.
Meytes D, Ayalon H, Virag I, Weisbort Y, Zakut H. Heparin-induced thrombocy-
topenia and recurrent thrombosis in pregnancy. A case report. J Reprod Med
1986; 31:993–996.
102 Warkentin
Muntean W, Finding K, Gamillscheg A, Zenz W. Multiple thromboses and couma-

rin-induced skin necrosis in a young child with antiphospholipid antibodies.
Thromb Haemorrh Disord 1992; 5:43–45.
Murdoch IA, Beattie RM, Silver DM. Heparin-induced thrombocytopenia in
children. Acta Paediatr 1993; 82:495–497.
Nand S, Wong W, Yuen B, Yetter A, Schmulbach E, Gross Fisher S. Heparin-induced
thrombocytopenia with thrombosis: incidence, analysis of risk factors, and
clinical outcomes in 108 consecutive patients treated at a single institution. Am J
Hematol 1997; 56:12–16.
Natelson EA, Lynch EC, Alfrey CP Jr, Gross JB. Heparin-induced thrombocytope-
nia. An unexpected response to treatment of consumption coagulopathy. Ann
Intern Med 1969; 71:1121–1125.
Nelson JC, Lerner RG, Goldstein R, Cagin NA. Heparin-induced thrombocytopenia.
Arch Intern Med 1978; 138:548–552.
Newman PM, Chong BH. Further characterization of antibody and antigen in
heparin-induced thrombocytopenia. Br J Haematol 1999; 107:303–309.
Nguyen QD, Do DV, Feke GT, Demirjian ZN, Lashkari K. Heparin-induced anti-
heparin-platelet antibody associated with retinal venous thrombosis. Ophthal-
mology 2003; 110:600–603.
Odeh M, Oliven A. Urticaria and angioedema induced by low-molecular-weight hep-
arin [letter]. Lancet 1992; 340:972–973.
Olbricht K, Wiersbitzky M, Wacke W, Eichler P, Zinke H, Schwock M, Möx B,
Kraatz G, Motz W, Greinacher A. Atypical heparin-induced thrombocytopenia
complicated by intracardiac thrombus, effectively treated with ultra-low-dose
rt-PA lysis and recombinant hirudin (Lepirudin). Blood Coagul Fibrinolysis
1998; 9:273–277.
Olinger GN, Hussey CV, Olive JA, Malik MI. Cardiopulmonary bypass for patients
with previously documented heparin-induced platelet aggregation. J Thorac
Cardiovasc Surg 1984; 87:673–677.
Oriot D, Wolf M, Wood C, Brun P, Sidi D, Devictor D, Tchernia G, Huault G.
Thrombopénie sévère induite par l’héparine chez un nourrisson porteur d’une
myocardite aiguë. Arch Fr Pediatr 1990; 47:357–359.
Padilla A, Gray E, Pepper DS, Barrowcliffe TW. Inhibition of thrombin generation by
heparin and low molecular weight (LMW) heparins in the absence and presence
of platelet factor 4 (PF4). Br J Haematol 1992; 82:406–413.
Patriarca G, Rossi M, Schiavino D, Schinco G, Fais G, Varano C, Schiavello R. Rush
desensitization in heparin hypersensitivity: a case report. Allergy 1994; 49:292–
294.
Payne SM, Kovacs MJ. Cutaneous dalteparin reactions associated with antibodies of
heparin-induced thrombocytopenia. Ann Pharmacother 2003; 37:655–658.
Pedersen-Bjergaard U, Andersen M, Hansen PB. Drug-induced thrombocytopenia:
clinical data on 309 cases and the effect of corticosteroid therapy. Eur J Clin
Pharmacol 1997; 52:183–189.
Pfueller SL, David R, Firkin BG, Bilston RA, Cortizo F. Platelet aggregating IgG
antibody to platelet surface glycoproteins associated with thrombosis and
thrombocytopenia. Br J Haematol 1990; 74:336–341.
Phelan BK. Heparin-associated thrombosis without thrombocytopenia. Ann Intern

Med 1983; 99:637–638.
Platell CFE, Tan EGC. Hypersensitivity reactions to heparin: delayed onset throm-
bocytopenia and necrotizing skin lesions. Aust NZ J Surg 1986; 56:621–623.
Pohl C, Klockgether T, Greinacher A, Hanfland P, Harbrecht U. Neurological com-
plications in heparin-induced thrombocytopenia. Lancet 1999; 353:1678–1679.
Pohl C, Harbrecht U, Greinacher A, Theuerkauf I, Biniek R, Hanfland P, Klockgether
T. Neurologic complications in immune-mediated heparin-induced thrombo-
cytopenia. Neurology 2000; 54:1240–1245.
Popov D, Zarrabi MH, Foda H, Graber M. Pseudopulmonary embolism: acute
respiratory distress in the syndrome of heparin-induced thrombocytopenia. Am
J Kidney Dis 1997; 29:449–452.
Potter C, Gill JC, Scott JP, McFarland JG. Heparin-induced thrombocytopenia in a
child. J Pediatr 1992; 121:135–138.
Pötzsch B, Klövekorn WP, Madlener K. Use of heparin during cardiopulmonary
bypass in patients with a history of heparin-induced thrombocytopenia [letter].
N Engl J Med 2000; 343:515.
Pouplard C, Amiral J, Borg JY, Laporte-Simitsidis S, Delahousse B, Gruel Y.
Decision analysis for use of platelet aggregation test, carbon 14-serotonin
release assay, and heparin-platelet factor 4 enzyme-linked immunosorbent assay
for diagnosis of heparin-induced thrombocytopenia. Am J Clin Pathol 1999;
111:700–706.
Ramirez-Lassepas M, Cipolle RJ, Rodvold KA, Seifert RD, Strand L, Taddeini L,
Cusulos M. Heparin-induced thrombocytopenia in patients with cerebrovas-
cular disease. Neurology 1984; 34:736–740.
Ranze O, Ranze P, Magnani HN, Greinacher A. Heparin-induced thrombocytopenia
in paediatric patients—a review of the literature and a new case treated with
danaparoid sodium. Eur J Pediatr 1999; 158(suppl 3):S130–S133.
Ranze O, Rakow A, Ranze P, Eichler P, Greinacher A, Fusch C. Low-dose danaparoid
sodium catheter flushes in an intensive care infant suffering from heparin-
induced thrombocytopenia. Pediatr Crit Care Med 2001; 2:175–177.
Rhodes GR, Dixon RH, Silver D. Heparin-induced thrombocytopenia with throm-
Rhodes GR, Dixon RH, Silver D. Heparin-induced thrombocytopenia: eight cases
with thrombotic-hemorrhagic complications. Ann Surg 1977; 186:752–758.
Rice L, Attisha WK, Drexler A, Francis JL. Delayed-onset heparin-induced throm-
bocytopenia. Ann Intern Med 2002; 136:210–215.
Risch L, Pihan H, Zeller C, Huber AR. ET gets HIT-thrombocytotic heparin-induced
thrombocytopenia (HIT) in a patient with essential thrombocythemia (ET).
Blood Coagul Fibrinolysis 2000; 11:663–667.
Roberts B, Rosato FE, Rosato EF. Heparin—a cause of arterial emboli? Surgery 1964;
55:803–808.
Rowland CH, Woodford PA, de Lisle-Hammond J, Nair B. Heparin-induced throm-
bocytopenia: thrombosis syndrome and bilateral adrenal hemorrhage after pro-
phylactic heparin use. Aust NZ J Med 1999; 29:741–742.
Sauer M, Gruhn B, Fuchs D, Altermann WW, Greinacher A, Völpel H, Zintl F.
104 Warkentin
Anticoagulation with recombinant hirudin following bone marrow transplan-

tation in a patient with activated protein C resistance and heparin-induced
antibodies showing cross-reactivity to the heparinoid danaparoid. Med Pediatr
Oncol 1999; 32:457–458.
Scheffold N, Greinacher A, Cyran J. Intrakardiale Thrombenbildung bei Heparin-
assoziierter Thrombozytopenie Typ II. Dtsch Med Wochenschr 1995; 120:519–
522.
Shah MR, Spencer JP. Heparin-induced thrombocytopenia occurring after discontin-
uation of heparin. J Am Board Fam Pract 2003; 16:148–150.
Silver D, Kapsch DN, Tsoi EKM. Heparin-induced thrombocytopenia, thrombosis,
and hemorrhage. Ann Surg 1983; 198:301–306.
Smythe MA, Warkentin TE, Stephens JL, Zakalik D, Mattson JC. Venous limb
gangrene during overlapping therapy with warfarin and a direct thrombin in-
hibitor for immune heparin-induced thrombocytopenia. Am J Hematol 2002;
71:50–52.
Solomon SA, Cotton DWK, Preston FE, Ramsay LE. Severe disseminated intra-
vascular coagulation associated with massive ventricular mural thrombus fol-
lowing acute myocardial infarction. Postgrad Med J 1988; 64:791–795.
Spadone D, Clark F, James E, Laster J, Hoch J, Silver D. Heparin-induced throm-
bocytopenia in the newborn. J Vasc Surg 1992; 15:306–312.
Srinivasan AF, Rice L, Bartholomew JR, Rangaswamy C, La Perna L, Thompson JE,
Murphy S, Baker KR. Warfarin-induced skin necrosis and venous limb gang-
rene in the setting of heparin-induced thrombocytopenia. Arch Intern Med. In
press.
Stevenson MM. Trombocytopenia during heparin therapy [letter]. N Engl J Med 1976;
295:1200–1201.
Tahata T, Miki S, Kusuhara K, Ueda Y, Okita Y, Matsuo S. [Delayed onset of heparin
induced thrombocytopenia: a case report]. Nippon Kyobu Geka Gakkai Zasshi
1992; 40:456–458.
Tans G, Rosing J, Thomassen MC, Heeb MJ, Zwaal RF, Griffin JH. Comparison of
anticoagulant and procoagulant activities of stimulated platelets and platelet-
derived microparticles. Blood 1991; 77:2641–2648.
Tezcan AZ, Tezcan H, Gastineau DA, Armitage JO, Haire WD. Heparin-induced
thrombocytopenia after bone marrow transplantation: report of two cases.
Bone Marrow Transplant 1994; 14:487–490.
Theuerkauf I, Lickfett L, Harbrecht U, Pohl C, Fischer HP, Pfeifer U. Segmental
hepatic vein thrombosis associated with heparin-induced thrombocytopenia II.
Virchows Arch 2000; 436:88–91.
Tholl U, Greinacher A, Overdick K, Anlauf M. Life-threatening anaphylactic reaction
following parathyroidectomy in a dialysis patient with heparin-induced throm-
bocytopenia. Nephrol Dial Transplant 1997; 12:2750–2755.
Thomas D, Block AJ. Thrombocytopenia, cutaneous necrosis, and gangrene of the
upper and lower extremities in a 35-year-old man. Chest 1992; 102:1578–1580.
Towne JB, Bernhard VM, Hussey C, Garancis JC. White clot syndrome. Peripheral
vascular complications of heparin therapy. Arch Surg 1979; 114:372–377.
Van der Weyden MB, Hunt H, McGrath K, Fawcett T, Fitzmaurice A, Sawers RJ,
Rosengarten DS. Delayed-onset heparin-induced thrombocytopenia. A poten-

tially malignant syndrome. Med J Aust 1983; 2:132–135.
Vazquez-Jimenez JF, Janssens U, Sellhaus B, Hermanns B, Huegel W, Hanrath P,
Messmer BJ. Thrombosis of a mitral valve prosthesis in a patient with heparin-
induced thrombocytopenia type II. J Thorac Cardiovasc Surg 1999; 118:751–
753.
Vella A, Nippoldt TB, Morris JC III. Adrenal hemorrhage: a 25-year experience at the
Mayo Clinic. Mayo Clin Proc 2001; 76:161–168.
Vignon P, Guéret P, Francßois B, Serhal C, Fermeaux V, Bensaid J. Acute limb ischemia
and heparin-induced thrombocytopenia: the value of echocardiography in
eliminating a cardiac source of arterial emboli. J Am Soc Echocardiogr 1996;
9:344–347.
Med 1999; 106:629–635.
Warkentin TE. Hemostasis and atherosclerosis. Can J Cardiol 1995; 11(suppl C):29C–
34C.
Warkentin TE. Heparin-induced skin lesions. Br J Haematol 1996a; 92:494–497.
Warkentin TE. Heparin-induced thrombocytopenia: IgG-mediated platelet activa-
tion, platelet microparticle generation, and altered procoagulant/anticoagulant
balance in the pathogenesis of thrombosis and venous limb gangrene compli-
cating heparin-induced thrombocytopenia. Transfusion Med Rev 1996b; 10:
249–258.
Warkentin TE. Heparin-induced thrombocytopenia, heparin-induced skin lesions,
and arterial thrombosis. Thromb Haemost 1997; 77(suppl):562.
Warkentin TE. Clinical presentation of heparin-induced thrombocytopenia. Semin
Hematol 1998a; 35(suppl 5):9–16.
Warkentin TE. Limitations of conventional treatment options for heparin-induced
thrombocytopenia. Semin Hematol 1998b; 35(suppl 5):17–25.
Warkentin TE. Heparin-induced thrombocytopenia: a clinicopathologic syndrome.
Thromb Haemost 1999; 82(suppl):439–447.
Warkentin TE. Venous thromboembolism in heparin-induced thrombocytopenia.
Curr Opin Pulm Med 2000; 6:343–351.
Warkentin TE. Heparin-induced thrombocytopenia and the anesthesiologist. Can J
Anesth 2002; 49(suppl):S36–S49.
Warkentin TE. Heparin-induced thrombocytopenia: pathogenesis and management.
Br J Haematol 2003; 121:535–555.
Warkentin TE, Bernstein RA. Delayed-onset heparin-induced thrombocytopenia and
cerebral thrombosis after a single administration of unfractionated heparin. N
Engl J Med 2003; 348:1067–1069.
Warkentin TE, Heddle NM. Laboratory diagnosis of immune heparin-induced throm-
bocytopenia. Curr Hematol Rep 2003; 2:148–157.
Warkentin TE, Kelton JG. Interaction of heparin with platelets, including heparin-
induced thrombocytopenia. In: Bounameaux H, ed. Low-Molecular-Weight
Heparins in Prophylaxis and Therapy of Thromboembolic Diseases. New York:
Marcel Dekker, 1994:75–127.
106 Warkentin

Am J Med 1996; 101:502–507.
Warkentin TE, Kelton JG. Temporal aspects of heparin-induced thrombocytopenia.
N Engl J Med 2001a; 344:1286–1292.
thrombosis. Ann Intern Med 2001b; 135:502–506.
Warkentin TE, Soutar RL, Panju A, Ginsberg JS. Acute systemic reactions to intra-
venous bolus heparin therapy: characterization and relationship to heparin-
induced thrombocytopenia [abstr]. Blood 1992; 80(suppl 1):160a.
Warkentin TE, Hirte HW, Anderson DR, Wilson WEC, O’Connell GJ, Lo RC.
Transient global amnesia associated with acute heparin-induced thrombocyto-
penia. Am J Med 1994; 97:489–491.
Warkentin TE, Sikov WM, Lillicrap DP. Multicentric warfarin-induced skin necrosis
complicating heparin-induced thrombocytopenia. Am J Med 1999; 62:44–48.
Warkentin TE, Sheppard JA, Horsewood P, Simpson PJ, Moore JC, Kelton JG.
Impact of the patient population on the risk for heparin-induced thrombocy-
topenia. Blood 2000; 96:1703–1708.
Intern Med 2003. In press.
Warkentin TE, Whitlock RP, Teoh KHT. Warfarin-associated multiple digital ne-
crosis complicating heparin-induced thrombocytopenia and Raynaud’s phe-
nomenon after aortic valve replacement for adenocarcinoma-associated
thrombotic endocarditis. Am J Hematol 2004. In press.
Wütschert R, Piletta P, Bounameaux H. Adverse skin reactions to low molecular
weight heparins: frequency, management and prevention. Drug Safety 1999; 20:
515–525.
Zalcberg JR, McGrath K, Dauer R, Wiley SJ. Heparin-induced thrombocytopenia
with associated disseminated intravascular coagulation. Br J Haematol 1983;
54:655–660.
4
Frequency of Heparin-Induced
Thrombocytopenia
David H. Lee
Queen’s University, Kingston, Ontario, Canada
I. INTRODUCTION
Thrombocytopenia is a common problem encountered in hospitalized pa-

tients. For patients receiving heparin, there are three general explanations for
thrombocytopenia: (1) heparin-induced thrombocytopenia (HIT), (2) non-
idiosyncratic heparin-induced platelet activation (see Chap. 5), and—perhaps
most often—(3) an unrelated clinical problem, either common (e.g., hemodi-
lution, septicemia) or rare (e.g., posttransfusion purpura, drug-induced
immune thrombocytopenic purpura) (see Chaps. 2 and 12). The availability
of sensitive and specific laboratory assays (e.g., enzyme immunoassay [EIA]
and serotonin release assay [SRA]) for pathogenic HIT antibodies means that
patients with HIT can usually be readily distinguished from the other
conditions (see Chap. 11). However, the role of heparin in causing thrombo-
cytopenia because of nonimmune platelet activation cannot readily be
separated from other common medical problems encountered in hospitalized
patients, either on clinical or laboratory criteria. Furthermore, these two
conditions can coexist (Chong and Castaldi, 1986). Thus, the term ‘‘nonim-
mune heparin-associated thrombocytopenia’’ (nonimmune HAT) has been
recommended to describe patients who develop thrombocytopenia during
107
108 Lee and Warkentin
heparin treatment in which a role for HIT antibodies cannot be implicated

(Warkentin et al., 1998a).
Unfortunately, many early studies of HIT frequency either did not
perform laboratory testing or used relatively insensitive or nonspecific assays
to diagnose HIT. In contrast, more recent studies have used one, or even two,
sensitive and complementary assays. Perhaps for this reason, the under-
standing of the frequency and clinical import of HIT has shifted over the
years. Formerly, the range of views on immune-mediated HIT were diver-
gent: it was considered both nonexistent (Bell, 1988) and common (Kelton,
1986). Nevertheless, both viewpoints acknowledged that thrombosis re-
sulting from HIT was very uncommon. Today’s perspective on HIT is
very different. The frequency of HIT is now shown to be variable, partly
depending on patient population and type of heparin used. For example,
the frequency ranges from less than 1% (cardiac medical patients) to 5%
(orthopedic surgical patients) receiving unfractionated heparin (UFH); HIT
antibody formation following UFH use ranges from 2% (cardiac medical
patients) to 15% (orthopedic surgical patients) to 50% (cardiac surgical
patients). Most importantly, however, it is now becoming clear that the risk
for thrombosis in patients who develop HIT is at least 33–50%, a frequency
that is far greater than in control patients who do not develop HIT (see
Chap. 3).
The biological basis for this variability in frequency of HIT and HIT
antibody formation is now apparent. The HIT antigen is a cryptic auto-
antigen, or neoantigen, on platelet factor 4 (PF4) that is formed when PF4
binds to heparin (see Chaps. 6–8). Only stoichiometric concentrations of
heparin and PF4 will form the antigen. Thus, it can be hypothesized that the
frequency of HIT antibody formation will be influenced not only by heparin
dose and composition, but also by circulating PF4 levels. Conditions associ-
ated with fluctuating, but at times high, circulating PF4 and heparin levels
(e.g., cardiac surgery) might be ideal for immunization to the HIT antigen.
Thus, real differences in HIT frequency observed among prospective studies
can be understood in a biologically plausible context.
The key role of the pathogenic HIT antibodies and the availability of
sensitive and specific assays for their detection suggest that HIT should be
considered a clinicopathologic syndrome. Consequently, this chapter will
focus on studies that have used in vitro testing to evaluate HIT antibodies.
However, other features known to be useful to diagnose HIT, such as the
timing of the onset of thrombocytopenia in typical HIT, and the rapid platelet
count fall on heparin rechallenge, will also be used (see Chap. 3). The
importance of confirmatory laboratory testing should not be underestimated:
prospective (Greinacher et al., 1994; Lee et al., 1996) and retrospective (Look
et al., 1997) studies suggest that only 25–55% of sera referred for evaluation
Frequency of HIT 109
Figure 1 Initially unrecognized HIT during prospective studies. (a) A 57-year-old

man developed skin lesions at the sites of LMWH (enoxaparin) injections on day 9.
An acute systemic reaction (ASR; see Chap. 3) developed after a 5000 U intravenous
UFH bolus. This patient was recognized as having had the HIT syndrome only
following systematic testing for HIT antibodies performed later (Warkentin et al.,
1995, 2003a). (b) A 73-year-old woman developed bowel infarction necessitating
resection while receiving UFH prophylaxis after hip replacement surgery. The
thrombocytopenia was initially attributed to ‘‘sepsis.’’ However, the patient was later
recognized as having the HIT syndrome following systematic testing for HIT anti-
bodies performed later (Warkentin et al., 1995, 2003a).
Table 1 Explanations for Variable Frequency of HIT Among Prospective

Studies
Biological explanations
1. Patient population studied (frequency of HIT antibody formation differs
among patient populations, possibly because of differences in platelet
activation and PF4 release)
2. Type of heparin used (UFH more immunogenic than LMWH; bovine lung
heparin more immunogenic than porcine mucosal heparin; possibly, lot-to-lot
variability in immunogenicity of heparin)
3. Variable duration of heparin treatment (HIT typically begins between days 5
and 10)
4. Dose of heparin used (dose-dependent thrombocytopenia)
Technical explanations
1. Variable definition of thrombocytopenia used
2. Differing baseline platelet counts permitted for study entry
3. Requirement to repeat platelet count testing to confirm thrombocytopenia
4. Variable intensity of platelet count surveillance
5. Variable intensity of surveillance for thrombotic events
6. Failure to exclude nonimmune heparin-associated thrombocytopenia
a. Lack of use of in vitro test for HIT antibodies
b. Use of insensitive or nonspecific HIT antibody assays
c. Inclusion of patients with ‘‘early’’ thrombocytopenia
d. Failure to exclude patients whose platelet count recovered during continued
heparin treatment
e. Failure to exclude patients with other explanations for thrombocytopenia
test positive for HIT antibodies. Furthermore, systematic analysis of a large

clinical trial of heparin treatment (Warkentin et al., 1995, 2003a) revealed
several patients in whom unusual clinical events subsequently linked to HIT
were initially attributed to other problems (Fig. 1).
Table 1 lists various biological and technical explanations that underlie
the reported variability in frequency of HIT among the prospective studies.
We will begin our discussion by summarizing an important technical problem
in many studies, namely, the failure to exclude patients with early, nonim-
mune HAT.
II. EARLY- VERSUS LATE-ONSET THROMBOCYTOPENIA
The distinction between thrombocytopenia that begins early (within 4 days)

or late (5 or more days after beginning heparin treatment) is a simple clinical
feature that is useful to distinguish nonimmune HAT, which begins early,

from (immune) HIT, which begins late. For this assessment, the first day of
heparin use is considered day 0 (see Table 2 in Chap. 3). There is an important
exception to this rule of timing for HIT: a rapid fall in platelet count on
starting heparin therapy can represent acute HIT, but only if a patient already
has circulating HIT antibodies, usually the result of a recent heparin
exposure. HIT antibodies are transient, which could explain why the risk
for rapid-onset HIT is restricted to a period of about 100 days following
exposure to heparin (Warkentin and Kelton, 2001) (see Chap. 3).
Typically, nonimmune HAT begins 1–2 days after starting heparin
administration and resolves during continued heparin therapy (Johnson et al.,
1984; Chong and Berndt, 1989; Warkentin and Kelton, 1994; Warkentin et
al., 1995; Greinacher, 1995). The platelet count fall is usually mild, with a
nadir between 75 and 150 109/L. This early platelet count fall may be the
result of a direct activating effect of heparin on platelets (Chong and Ismail,
1989; Chong and Castaldi, 1986) or of comorbid clinical factors.
Early nonimmune HAT occurs in up to 30% of patients receiving
heparin (Bell et al., 1976; Nelson et al., 1978; Warkentin et al., 1995). Sys-
tematic serological investigation of patients with early thrombocytopenia was
performed in one study comparing UFH with low molecular weight heparin
(LMWH) for post-operative antithrombotic prophylaxis in patients who
underwent hip replacement surgery (Warkentin et al., 1995). With 150
109/L as a platelet count threshold, early thrombocytopenia was observed in
189/665 (28%) patients; however, HIT antibodies were not detected in any of
the 98 patients tested, and platelet count recovery to more than 150 109/L
within 3 days occurred despite continuing the heparin (Warkentin et al.,
1995). No difference in the frequency of early thrombocytopenia was ob-
served between patients who received UFH (28%) and those who received
LMWH (29%). This suggests that unrelated clinical factors, such as peri-
operative hemodilution with fluid and blood products, were primarily
responsible. In contrast, the onset of late thrombocytopenia (i.e., between
days 5 and 10 of heparin treatment) was strongly associated with the
formation of HIT antibodies and occurred significantly more frequently in
the patients who received UFH (discussed subsequently).
III. FREQUENCY OF IMMUNE HIT
Tables 2 and 3 list prospective studies of the frequency of HIT that

employed in vitro testing for HIT antibodies or were studies in which the
likelihood of HIT could be judged based on information provided, espe-
cially the timing of the onset of thrombocytopenia. Relevant variables
112
Table 2 The Frequency of HIT: Prospective Studies of HIT in Medical Patients Using In Vitro Testing of Patient Serum/Plasma
for HIT Antibodies or Indicating a High Likelihood of HIT Based on Timing of Platelet Count Fall
Frequency of (immune) HIT (%)
Major Timing of Definition of
indication In vitro Route, Bovine Porcine platelet fall thrombocytopenia
Study for heparin test dose UFH UFH LMWH reported? (109/L)
Comparisons between bovine UFH and porcine UFH [studies and data in bold are randomized, controlled trials (RCTs)]
Ansell et al., 1980 VTE PRP(SR) iv ther 4/21 (19.0) 0/22 (0) Yes <150
(RCT)
Green et al., 1984 VTE HIPA iv ther 2/45 (4.4) 0/44 (0) Yes <150
(RCT)
Powers et al., 1984 VTE SRAa iv ther 2/65a (3.1) 0/66 (0) Yes <150
(RCT)
Bailey et al., 1986 VTE, ATE No test iv ther 1/21 (4.8) 0/22 (0) Yes <100
(RCT)
Cipolle et al., 1983; VTE, ATE PRP iv ther 6/100b (6.0) 1/111 (0.9) Yes <100
Ramirez-Lassepas [stroke] [3/54] (5.6) (1/83) (1.2)
et al., 1984 [stroke]
Predominant treatment for venous (VTE) or arterial (ATE) thromboembolism
Bell et al., 1976
Alving et al., 1977 VTE, ATE PRP/SRAc iv ther 3/52c (5.8) Yes <100
Powers et al., 1979 VTE No test iv ther 2/120d (1.7) Yesd <150
Gallus et al., 1980 VTE PRP iv ther 3/166e (1.8) Yes <100
sc proph 0/5 (0)
Holm et al., 1980 VTE No test iv ther 0/90f (0) Yes <100
Monreal et al., 1989 VTE No test iv ther 2/89 (2.2) Low <100
sc proph 0/49 (0) 0/43 (0) platelets
Lee and Warkentin
day 8
Kakkasseril et al., VTE, ATE PRP iv ther 4/142g (2.8) No <100
1985
Malcolm et al., 1979 Multiple PRP iv ther 1/66h (1.5) Yes <100
sc proph 0/38 (0)
Rao et al., 1989 Multiple PRP(SR) iv ther 0/94 (0) NA <100
sc proph 0/99 (0)
Girolami et al., 2003 VTE, ATE HIPA, EIA sc proph 5/360 (1.4) Yes >50% fall
Frequency of HIT
sc ther 0/238 (0)

Lindhoff-Last et al., VTE EIA iv ther 1/356 (0.3)i Yes <100 or
2002 sc ther 0/720(0)i >50% fall
Predominant treatment for myocardial infarction or acute coronary syndromes (MI/ACS)
Kappers-Klunne MI/ACS HIPA, EIA iv ther 1/358 (0.3) Yes <60 and
et al., 1997 >50% fall
2/358 (0.6) (<120)
Romeril et al., 1982 MI/ACS PRP sc proph 0/45 (0) NA <150
Weitberg et al., 1982 MI/ACS No test sc proph 0/50 (0) NA <150
Johnson et al., 1984 MI/ACS No test sc proph 0/66 (0) Yes <150
Intensive care unit
Verma et al., 2003 Multiple EIA, SRA Multiple 1/267 (0.4) Yes >33% fall
Hemodialysis
Yamamoto et al., New-onset EIA, PRP iv ther 6/154 (3.9) Yes Clotting and
1996 hemodialysis >20% fall
3/154 (1.9) (>50% fall)
Healthy volunteers
Saffle et al., 1980 — PRP sc proph 0/25 (0) 0/14 (0) NA <150
113
Table 2 Continued
Frequency of (immune) HIT (%)
114

Schwartz et al., 1985 — No test Bolus iv 3/20 (15.0) 0/10 (0) Yes <150
ther
HIT was excluded if the platelet count rose during continued heparin after an early fall (e.g., Johnson et al., 1984). Also, where uncertainty existed
as to the number of patients with probable HIT, the lower number was indicated in the table, to avoid overestimating the number of patients with
HIT (contrast the analysis shown in Table 4). Some data relating to Cipolle et al. (1983) were obtained by personal communication, as reported
(Warkentin and Kelton, 1991). The study by Gallus et al. (1980) was excluded because the source of heparin was not specified. Some reports (e.g.,
Nelson et al., 1978) were excluded because timing of thrombocytopenia was not reported.
Abbreviations: ATE, arterial thromboembolism; EIA, PF4-heparin enzyme-immunoassay; HIPA, heparin-induced platelet activation test; iv ther,
intravenous therapeutic-dose heparin; LMWH, low molecular weight heparin; MI/ACS, myocardial infarction/acute coronary syndromes; PRP,
HIT assay using citrated platelet-rich plasma (PRP/SR, with serotonin release); sc proph, subcutaneous prophylactic-dose heparin; sc ther,
subcutaneous therapeutic; RCT, randomized controlled trial; SRA, serotonin release assay using washed platelets; UFH, unfractionated heparin;
VTE, venous thromboembolism.
a
Powers et al. (1984) described five patients who developed thrombocytopenia during bovine UFH use (none during porcine UFH use); two
patients whose platelet counts fell beginning on day 7 (to nadir of 41 109/L) and on day 8 (to nadir of 53 109/L) had positive testing for HIT
antibodies by SRA; test results are not available for one patient whose platelet count fell on day 5 (excluded from Table 2, but included in Table 4);
one patient with proved HIT developed progression of deep venous thrombosis, as indicated in Table 4 (laboratory records of Dr. J. G. Kelton).
b
Cipolle et al. (1983) described ten patients who received bovine UFH who may have had HIT; only six are included here, all of whom developed a
platelet count fall between days 7 and 10. Ramirez-Lassepas et al. (1984) examined the subgroup with underlying cerebrovascular disease and
reported six patients who developed thrombocytopenia after receiving bovine heparin. Only the three who developed late thrombocytopenia are
reported here.
c
Only 8 of 16 thrombocytopenia patients underwent PRP testing (all negative) and only 5 underwent SRA testing. The 3 who developed late
thrombocytopenia (HIT) had negative PRP testing, but did not have SRA testing.
d
Information on timing of platelet count fall to determine the likelihood of HIT was obtained by personal communication, as described (Warkentin
and Kelton, 1991).
e
Origin of heparin is uncertain and may have used mucosal heparin from sheep; at least 3, and as many as 5, patients appeared to have HIT based
on in vitro testing and timing of platelet count fall.
f
One case of early thrombocytopenia due to DIC reported in original paper was excluded.
Lee and Warkentin
g
At least four, and as many as nine, patients appeared to have HIT; the four that tested positive for HIT antibodies are included here.
h
One patient appeared to have HIT based on positive heparin rechallenge, despite negative in vitro HIT test.
i
Thrombocytopenic patients with negative EIA testing were excluded.
Table 3 The Frequency of HIT: Prospective Studies of Surgical Patients Using Confirmatory In Vitro Laboratory Testing
of Patient Serum or Plasma or Indicating a High Likelihood of HIT Based on Timing of Platelet Count Fall
Frequency of (immune) HIT

Frequency of HIT
Comparisons between porcine UFH and LMWH

Leyvraz et al., Orthopedic PRP sc proph 2/204 (1.0) 0/205 (0) Yes <100, >40% fall
1991
Warkentin et al., Orthopedic SRA sc proph 9/332 (2.7) 0/333 (0) Yes <150
1995, 2003
16/332 (4.8) 2/333 (0.6) >50% fall
Other studies
Louridas, 1991 Vascular PRPa iv ther, 5/114b (4.4) Yes <100
sc proph
Warkentin et al., Orthopedic SRA/EIA sc proph 2/246 (0.8) Yes <150 or
1998b >50% fall
Ganzer et al., Orthopedic HIPA sc proph 15/307 (4.9) Yes >50% fall
1997
Trossaert et al., Cardiac PRP/EIA sc proph 0/51 (0) Yes Not stated
1998
Warkentin et al., Cardiac SRA sc proph 1/100 (1.0) Yes >50% fall
2000
Pouplard et al., Cardiac SRA, EIA sc proph 9/263 (3.4) 1/370 (0.3) No <100 or
1999, 2002 >40% fall
Abbreviations: EIA, PF4-heparin enzyme-linked immunosorbent assay; HIPA, heparin-induced platelet activation test (aggregation of washed
platelets): LMWH, low molecular weight heparin; PRP, HIT assay using citrated platelet-rich plasma; sc proph, subcutaneous prophylactic-dose
heparin; SRA, serotonin release assay using washed platelets; UFH, unfractionated heparin.
a
Ineffective testing was used (platelet aggregation without heparin).
115
b
Of seven patients with thrombocytopenia reported, two were excluded because of early onset of thrombocytopenia.
influencing the frequency of HIT include the type of heparin used and the
patient population.
A. Frequency of HIT in Medical Patients and Normal

Volunteers: Comparison of UFH of Bovine Versus
Porcine Origin
Five randomized trials (Bell and Royall, 1980; Green et al., 1984; Powers et
al., 1984; Ansell et al., 1985; Bailey et al., 1986) and one nonrandomized study
(Cipolle et al., 1983; Ramirez-Lassepas et al., 1984) compared the frequency
of HIT during treatment with UFH that was derived either from bovine lung
or porcine intestinal mucosa. In addition, the frequency of HIT was evaluated
in normal volunteers in one randomized (Schwartz et al., 1985) and one
nonrandomized prospective study (Saffle et al., 1980) involving porcine and
bovine heparins. The study of Bell and Royall (1980) has been excluded from
primary analysis because neither laboratory testing for HIT antibodies nor
data on the timing of onset of thrombocytopenia were provided.
Taken together, the four randomized controlled trials in medical
patients strongly suggest that bovine UFH is more likely to cause HIT than
porcine UFH, as all nine patients with HIT had received UFH of bovine
origin ( p = 0.0059 by Mantel-Haenszel) (see Table 2). Similarly, an increased
frequency of HIT in patients receiving bovine lung heparin was suggested in
the nonrandomized comparison by Cipolle and colleagues (1983) (6/100, 6%
vs. 1/111, 0.9%; p = 0.055), as well as in the study of Bell and Royall (1980)
(26% vs. 8%; p < 0.005), although this latter study probably included
patients with nonimmune HAT.
A significantly higher frequency of HIT antibody formation with bovine
UFH was observed in a study of cardiac surgical patients randomized
between bovine and porcine UFH (Francis et al., 2003). Although a smaller
cardiac surgery study (Konkle et al., 2001) failed to detect a difference in
antibody formation, blood samples were obtained only until postoperative
day 5, which may have been too early to detect the majority of HIT antibodies
(Warkentin, 2003).
A higher frequency of immune HIT with bovine lung heparin is
biologically plausible. Bovine heparin has a higher sulfate:disaccharide ratio
than does porcine heparin (Casu et al., 1983), and it is better able to activate
platelets in vitro (Barradas et al., 1987). These properties could lead to greater
platelet activation in vivo and, consequently, greater potential for PF4 re-
lease. Moreover, the bovine heparin chains would be expected to better form
the large multimolecular complexes that compose the target antigen for HIT
antibodies.
Lot-to-lot variability within heparin of a particular animal source could

also contribute to variable frequency of HIT. Stead and coworkers (1984)
reported a striking cluster of six patients with pulmonary embolism compli-
cating HIT identified within a few weeks at one institution. A particular lot of
bovine lung heparin in use in the operating room was linked to these events:
patient serum-induced platelet aggregation occurred in the presence of this
particular lot of heparin, but not when other lots of bovine lung heparin from
the same manufacturer were used.
B. Frequency of HIT in Medical Patients Treated

with Porcine Mucosal UFH
Table 2 also lists the frequency of HIT observed in several prospective studies
that have evaluated medical patients receiving intravenous, therapeutic-dose
porcine UFH, usually for venous thromboembolism (VTE). Excluding a
study of hemodialysis, an overall frequency of HIT of slightly less than 1% is
suggested (0.7% [11/1604]; 95% CI 0.3–1.2%). This is a relatively low
number, particularly when one considers that, paradoxically, the frequency
appears to be much higher in postoperative surgical patients who received
lower (prophylactic) doses of porcine heparin (discussed subsequently).
In contrast, HIT seemed to occur more often in a prospective study of
acute hemodialysis patients receiving porcine UFH (Yamamoto et al., 1996).
Whether this is a real difference that reflects increased platelet activation (and
PF4 release) during hemodialysis or reflects a more sensitive definition of
thrombocytopenia (any platelet count fall associated with line clotting) re-
quires further investigation.
There is an anecdotal report of a patient developing HIT when she re-
ceived UFH to permit extracorporeal exposure of her leukocytes to ultravi-
olet irradiation (photopheresis) for treatment of lupus (Dittberner et al.,
2002).
C. Frequency of HIT in Surgical Patients Treated

with Porcine Mucosal UFH
Two large prospective studies suggest that HIT is an important problem in
orthopedic patients receiving UFH (Warkentin et al., 1995, 2003a; Ganzer
et al., 1997). When using a definition of a 50% fall in platelet count that began
on or after day 5 of heparin treatment, and that was confirmed by serologic
testing for HIT antibodies, both studies observed a frequency of HIT of about
5% (see Table 3). Each study used porcine mucosal heparin, derived from a
different manufacturer, that was given by the subcutaneous route at a dosage
of 15,000 U/day. Thrombocytopenia that was likely attributable to HIT has
also been observed in several other clinical trials of patients receiving UFH
following orthopedic surgery (see Warkentin et al., 1995, for discussion), but
only one study (Leyvraz et al., 1991) reported confirmatory in vitro testing.
There is little prospective information of the frequency of HIT in other
postoperative surgical populations treated with UFH (see Table 3). Three
studies have been performed on postoperative cardiac surgical patients who
also received postoperative UFH in addition to high doses of heparin during
preceding cardiopulmonary bypass (Trossaert et al., 1998; Warkentin et al.,
2000; Pouplard et al., 1999) (see Table 3). Pooling the three studies, about
2.4% of the patients developed serologically confirmed HIT. Interestingly the
frequency of HIT in this population appears to be lower than in orthopedic
patients receiving UFH, even though the cardiac surgical patients appear to
have a higher frequency of formation of HIT antibodies (Warkentin et al.,
2000).
Isolated limb perfusion (ILP) with melphalan employs extracorporeal
circulation (and thus high-dose UFH) to treat melanoma or unresectable
sarcoma limited to an extremity. In one study, HIT occurred in 3 of 108
patients (2.8%), who also received subcutaneous UFH prophylaxis following
ILP (Masucci et al., 1999). The occurrence of arterial thrombosis and partial
limb amputation in two of these patients with HIT led the investigators to
discontinue routine UFH prophylaxis post-ILP. The hypothesis that ILP is a
high-risk situation for HIT was supported by a prospective study showing
HIT antibody seroconversion in nine of nine patients who underwent this
procedure (despite not receiving postoperative UFH prophylaxis), with eight
patients having HIT antibodies ‘‘strong’’ enough to cause serotonin release.
D. HIT is Less Frequent in Orthopedic Patients Receiving

LMWH Compared with UFH Prophylaxis
Anecdotal reports indicate that HIT can occur during treatment with LMWH
(Ball et al., 1989; Tardy et al., 1990; de Raucourt et al., 1996; Plath et al., 1997;
Elalamy et al., 1996; Warkentin, 1998; Gruel et al., 2003; Ng and Lee, 2003).
Clinical trial data suggest that the frequency is low, however. Using a sensitive
definition for HIT (>50% fall on or after day 5 and confirmed by positive
HIT antibodies), two studies in Hamilton found an overall frequency of only
4/439 (0.9%; 95% CI 0.25–2.32%) for HIT complicating use of LMWH given
for postoperative orthopedic patients (Warkentin et al., 1995, 1998b, 2003a).
In contrast, using the same definition of HIT, patients who received UFH had
a much higher frequency of HIT, 16/332 (4.8%; 95% CI 2.78–7.71%)
(Warkentin et al., 1995, 2003a). A similar high frequency of UFH-induced
HIT was observed in a German study, 15/307 (4.9%; 95% CI, 2.76–7.93)
(Ganzer et al., 1997).
The strongest evidence that LMWH is indeed associated with a lower

frequency of HIT was provided by a randomized trial that directly compared
the frequency of HIT between the two types of heparin (Warkentin et al.,
1995, 2003a). The frequency of HIT in patients treated with the LMWH
preparation, enoxaparin (itself derived from porcine mucosal heparin), was
lower than that seen in patients treated with porcine UFH, irrespective of
whether a standard definition (platelet fall to <150 109/L on or after day 5
of heparin treatment) or a more sensitive definition (>50% platelet count fall
from the postoperative peak) of thrombocytopenia was used. The frequency
of HIT antibody formation also differed between the two patient groups,
using either the serotonin release assay (Warkentin et al., 1995) or the PF4-
heparin EIA (Warkentin et al., 2000). Thrombocytopenia also appeared to be
infrequent in other trials of LMWH (Leyvraz et al., 1991; Simonneau et al.,
1997; ENOXACAN Study Group, 1997).
UFH appeared also to lead to greater frequency of HIT antibody for-
mation than the LMWH reviparin (Clexane) in a randomized trial of post–hip
and knee surgery patients (Ahmad et al., 2003a). HIT antibodies occurred
somewhat more often in knee surgery patients. These same investigators also
examined HIT antibody formation in orthopedic patients immobilized in a
plaster cast who were randomized to receive reviparin or placebo (Ahmad et
al., 2003b). A surprising finding was that the number of patients who
apparently formed HIT antibodies (by EIA) was higher in the placebo group
(10 cases vs. 6). No patient in either study developed clinical HIT.
Gruel and colleagues (2003) performed a systematic study that identi-
fied 11 patients with HIT (3 with HIT-associated thrombosis) that had been
exclusively treated with a LMWH preparation (dalteparin, nadroparin, or
enoxaparin). Clinical and serologic features were similar to patients with HIT
developing during UFH, except that there was evidence that thrombocyto-
penia may begin somewhat later during LMWH therapy. Based upon
estimated relative use of UFH and LMWH in France, the authors estimated
the frequency of HIT to be 40-fold less with LMWH, compared with UFH.
E. Frequency of HIT During Pregnancy

HIT appears to be uncommon during pregnancy even with UFH treatment.
Fausett and colleagues (2001) reported that none of 244 pregnant women
developed HIT during UFH use, although HIT occurred in 10 of 244 (4%)
nonpregnant patients who received UFH ( p = 0.0014). In a literature review,
Sanson and coworkers (1999) identified no cases of HIT among 486 women
who received LMWH during pregnancy. Ellison et al. (2000) studied 57
pregnancies in 50 patients and also found no episodes of HIT in pregnant
women who received enoxaparin. More recently, Lepercq and colleagues
(2001) found no cases of HIT in 624 pregnancies among 604 women treated
with LMWH.
F. Role of Incidental UFH Flushes in the Frequency of HIT

There are two ways that incidental exposure to heparin by ‘‘flushing’’ of
intravascular catheters can affect the frequency or clinical effect of HIT. First,
such minor heparin exposures can trigger formation of HIT antibodies (Ling
and Warkentin, 1998; Warkentin et al., 1998b). And second, in patients who
have already formed potent HIT antibodies for any reason, any ongoing or
recurrent heparin exposure—including small-dose exposure—could lead to
recurrence or exacerbation of thrombocytopenia or thrombosis (Rice and
Jackson, 1981). Indeed, several patients have been reported in whom severe
HIT occurred while only small amounts of heparin were being given as flushes
to maintain the patency of intravascular catheters (Doty et al., 1986; Heeger
and Backstrom, 1986; Kappa et al., 1987; Rama et al., 1991; Brushwood,
1992; Parney and Steinke, 2000).
In most of the reports of patients developing HIT during LMWH
treatment, recent prior exposure to UFH was not excluded. Indeed, incidental
exposure to UFH by intraoperative invasive catheters could lead to formation
of HIT antibodies that are inappropriately attributed to later postoperative
LMWH prophylaxis (Shumate, 1995). However, if true, it would suggest that
the apparent difference in immunogenicity between UFH and LMWH could
be even greater than initially reported.
To address this issue, a randomized, double-blind clinical trial was per-
formed to test the hypothesis that incidental exposure to UFH by intra-
operative invasive lines, rather than postoperative LMWH antithrombotic
prophylaxis, was the predominant explanation for postoperative HIT anti-
body formation (Warkentin et al., 1998b). Patients were randomized to
receive either UFH or normal saline flushes during surgery. However, the
data obtained essentially ruled out the hypothesis: the frequency of HIT
antibodies was not higher in the patients who were randomized to receive
UFH flushes (2.2 vs. 2.7%; p = 0.73). Rather, the results suggested that
postoperative LMWH prophylaxis administered to both groups was the pre-
dominant factor in causing HIT antibody formation. However, HIT antibody
formation occurred in two patients who received UFH flushes, but who
subsequently were given warfarin anticoagulation. Because intraoperative
UFH flushes occasionally result in formation of high levels of HIT antibodies
that can lead to life-threatening, acute HIT if therapeutic-dose UFH is ad-
ministered a few weeks later (Ling and Warkentin, 1998), and because there is
no clinical benefit to flushing intravascular catheters with UFH (Warkentin et
al., 1998b), it seems reasonable to recommend that normal saline flushes be

considered for routine flushing of intravascular catheters used during surgery.
It is possible that heparin flushes for venous access devices in cancer pa-
tients can cause HIT antibody formation. In a serosurveillance study, Mayo
and colleagues (1999) found that about one third of 49 such patients tested
formed low levels of HIT antibodies (detected by EIA) at least once. However,
only one patient developed a positive serotonin release assay, and no patient
developed thrombocytopenia. These data are in keeping with our own expe-
rience that HIT is very uncommon in this patient population.
In recent years, many centers have substituted saline for heparin to
intermittently ‘‘flush’’ peripheral venous catheters. This is because saline
flushing of such devices ‘‘locked’’ between use have similar patency rates as
when heparin flushes are used (Randolph et al., 1998a). In contrast, heparin
may help prolong the patency of intra-arterial central venous, and pulmonary
artery catheters (Randolph et al., 1998b), and consequently exposure to hep-
arin by these routes remains common.
G. HIT and Heparin-Coated Devices

Heparin can be bonded to artificial surfaces (Larsson et al., 1987), either
through ionic attachment, as used for pulmonary artery catheters (Eldh and
Jacobsson, 1974), or by end-linked covalent bonding (e.g., Carmeda BioAc-
tive Surface, or CBAS) (Larm et al., 1983). CBAS has been used for car-
diopulmonary bypass circuits and filters (Borowiec et al., 1992a,b, 1993),
extracorporeal membrane oxygenation (ECMO) devices (Koul et al., 1992),
and coronary stents (Serruys et al., 1996). During use in patients, ionically
attached heparin is displaced by albumin from the catheter surface, where it
could contribute to HIT (discussed subsequently). End-linked heparin is an
effective and longer-lasting anticoagulant, as the immobilized, but flexi-
ble, heparin chains are able to interact with fluid-phase antithrombin and
thrombin (Elgue et al., 1993). Nevertheless, the end-linked, but relatively
unconstrained, heparin is capable of interacting with PF4 (Suh et al., 1998).
Therefore, it is theoretically possible that covalent heparin-bonded devices
could result in formation of HIT antibodies or could cause HIT in a patient
who has formed antibodies. Alternatively, even covalently bonded heparin
might ‘‘leach’’ into blood by proteolytic mechanisms, thereby contributing in
a more conventional way to the pathogenesis of HIT (Almeida et al., 1998a).
Use of heparin-coated pulmonary catheters in contributing to HIT has
been implicated by Laster and Silver (1988). These workers reported 10 pa-
tients with HIT whose platelet counts did not rise until the removal of their
heparin-coated pulmonary catheters, despite discontinuing all other sources
of heparin. Incubation of the heparin-coated catheters with platelets in the
presence of patient sera resulted in ‘‘catheter-induced’’ platelet aggregation.

Based on the identification of four such cases, during which time 1112 hepa-
rin-coated catheters had been used, they estimated the frequency of catheter-
associated HIT to be 0.4%.
H. HIT Caused by Other Sulfated Polysaccharides

The cryptic HIT autoantigen comprises conformationally altered PF4 when it
forms a multimolecular complex with heparin. Other negatively charged
polysaccharides can interact with PF4 to produce the HIT antigen (Wolf et
al., 1983; Greinacher et al., 1992a,b,c; Anderson 1992) (see Chap. 8). These
considerations explain why a number of high-sulfated polysaccharides, 10 or
more subunits in length, have been reported to cause a syndrome of throm-
bocytopenia and thrombosis that essentially mimics HIT. These drugs include
the semisynthetic 5-carbon subunit–based ‘‘heparinoid’’ pentosan polysulfate
(Gouault-Heilman et al., 1985; Vitoux et al., 1985; Follea et al., 1986; Goad
et al., 1994; Tardy-Poncet et al., 1994; Rice et al., 1998), as well as polysul-
fated chondroitin sulfate (Bouvier, 1980; Wolf et al., 1983; Greinacher et al.,
1992a). The frequency of immune-mediated thrombocytopenia, with or with-
out thrombosis, after exposure to these compounds is unknown, but may be
high.
I. Variable Duration of Heparin Treatment

As HIT typically begins 5–10 days after starting therapy with heparin, it
follows that the length of heparin treatment can influence the risk for HIT
(e.g., a 10-day course is far more likely to produce clinical HIT than a 3-day
treatment period). On the other hand, there is evidence that the risk of HIT
begins to decrease after 10 days of uninterrupted heparin use (see Fig. 1C,
Chap. 3). In a large study of postoperative orthopedic surgical patients
receiving postoperative heparin prophylaxis, no patient developed HIT anti-
bodies after day 10, even though many patients received heparin for up to
14 days (Warkentin et al., 1995). These data are consistent with a ‘‘point
exposure’’ model for risk of HIT in this patient population, such as a brief
time shortly after surgery, when high circulating PF4 levels coincide with the
first few subcutaneous heparin injections. However, even if HIT antibody
formation occurs during the day 5–10 window period, thrombocytopenia
itself can occur somewhat later, particularly if a larger dose of heparin is
given, or UFH is substituted for LMWH (see Fig. 1a). The characteristic
timing of HIT should assist clinicians in focusing their platelet count
monitoring for HIT during the critical time period, so that the early diagnosis
of HIT is improved (see Sec. VI).
J. Heparin Dose-Dependence in HIT

Analysis of individual patients with HIT often shows dose-dependence; that
is, mild thrombocytopenia during subcutaneous heparin prophylaxis is fol-
lowed by a marked drop in platelet count if the patient then receives thera-
peutic-dose heparin (Fig. 1a).
However, dose-dependence of HIT is not readily apparent when
reviewing prospective studies of HIT (see Tables 2 and 3). However, this
could be explained by differences in frequency of HIT among different patient
populations that confounds the influence of heparin dose. For example,
among medical patients, porcine UFH is more likely to result in HIT when
given in therapeutic, rather than prophylactic, doses. This difference, if real,
could reflect dose-dependence of heparin in HIT. On the other hand, the
relatively high frequency of HIT in surgical patients (up to 5%) receiving
‘‘only’’ prophylactic-dose porcine UFH more likely reflects differences in risk
related to this patient population.
IV. FREQUENCY OF THROMBOSIS COMPLICATING HIT
Ironically, although thrombosis was the first manifestation of the HIT

syndrome, first recognized over 40 years ago (Weismann and Tobin, 1958),
widespread recognition that thrombosis was a common complication of HIT
did not occur until recently. Indeed, until 1995 no study of HIT had compared
the frequency of thrombosis with a matched control population (Warkentin
et al., 1995). This study quantitated the strength of the association between
HIT and thrombosis and noted that the more unusual the thrombotic event
(e.g., bilateral deep venous thrombosis, pulmonary embolism), the stronger
the association with HIT (see Chap. 3).
Table 4 summarizes the thrombotic events that have been observed
during prospective studies of HIT. The major observation is that thrombosis
is relatively common in HIT patients, occurring in approximately one third of
medical patients and about one half of postoperative surgical patients. The
data also support findings from a prior retrospective study (Boshkov et al.,
1993) that found the type of thrombotic event complicating HIT was
influenced by the patient population. Table 4 suggests that the ratio of arterial
to venous thrombosis is about 1:1 in medical patients, many of whom might
have had arterial disease as their basis for hospitalization. Additionally, the
therapeutic-dose heparin used in many of these studies may have partially
protected against venous thromboembolic complications, although it may
not have prevented platelet-mediated arterial occlusion. In contrast, there
appears to be a strong predisposition to venous thromboembolism in post-
124
Table 4 Proportion of Patients with HIT Developing HIT-Associated Thrombosis in Prospective Studies Employing In Vitro
Laboratory Testing or in Which Data on Timing of Platelet Count Fall Were Reported
Patients with
HIT (using Thrombotic
Major Definition of more sensitive complication of HIT
indication In vitro Number thrombocytopenia definition
Study for heparin test treated (109/L) of HIT) Venous Arterial
Medical patients (only intravenous therapeutic dose or hemodialysis included in table)

Ansell et al., 1980 VTE PRP(SR) 43 <150 4 0 0
Green et al., 1984; VTE HIPA 89 <150 2 1 1
Green, 1986
Powers et al., 1984 VTE SRA 65 <150 3 1 0
Bailey et al., 1986 VTE, ATE No test 43 <100 1 0 1
Cipolle et al., 1983; Cerebrovascular PRP 137a <100 5 0 1
Ramirez-Lassepas ischemia (>40% fall) (16) Not stated (6)
et al., 1984
Powers et al., 1979 VTE No test 120 <150 2 1 1
Gallus et al., 1980 VTE PRP 166 <100 5 1 0
Monreal et al., 1989 VTE No test 89 <100 2 1 0
Kakkasseril et al., 1985 VTE, ATE PRP 142 <100 9 2 2
Malcolm et al., 1979 Multiple PRP 66 <100 1 1 0
Kappers-Klunne Acute coronary HIPA, EIA 358 <120, 2 1 0
et al., 1997 syndromes >30% fall
(<60 or (1) (1) (0)
Lee and Warkentin
>50% fall)
Yamamoto et al., 1996 Hemodialysis PRP, EIA 154 Clotting, 5 0 1
platelet fall
Total medical: Venous/arterial thrombosis ratio = 9/7 = 1.3 1472 41 9 7
Orthopedic surgical patients (total joint arthroplasty)
Warkentin et al., 1995, Hip SRA 332 <150 9 7 1
2003a
(>50% fall) (18) (12) (1)
Frequency of HIT
Warkentin et al., 1998b Hip, knee SRA 246 <150, >50% 2 0 0

Ganzer et al., 1997 Hip, knee HIPA 307 >50% fall 15 5c 0
Leyvraz et al., 1991 Hip PRP 175 <100, 2 2 0
>40% fall
Total orthopedic: Venous/arterial thrombosis ratio = 14/1 = 14.0 1060 28 14 1
(37) (19) (1)
Surgical, other
Louridas, 1991 Vascular PRPb 114 <100 5 0 0
Where there was uncertainty over the numbers of patients with HIT, the higher estimated value was indicated in the table, to minimize the bias
toward a high frequency of HIT-associated thrombosis (contrast analysis shown in Table 2).
Abbreviations: ATE, arterial thromboembolism; EIA, PF4-heparin enzyme-linked immunosorbent assay; HIPA, heparin-induced platelet ac-
tivation test (aggregation of washed platelets); LMWH, low molecular weight heparin; MI/ACS, myocardial infarction or acute coronary
syndromes; PRP, HIT assay using citrated platelet-rich plasma (PRP/SR, with serotonin release); SRA, serotonin release assay using washed
platelets; UFH, unfractionated heparin; VTE, venous thromboembolism.
a
Detailed clinical data on thrombosis were available only on the subset of patients with cerebrovascular disease (n = 137).
b
Ineffective testing was used (platelet aggregation without heparin).
c
Another five patients developed venous thrombosis in association with a positive HIPA assay, but the platelet count did not fall by >50%.
125
operative orthopedic patients who have developed HIT (venous:arterial ratio

at least 14:1) (see Table 4).
The retrospective identification of patients with serologically confirmed
HIT permits analysis of large groups of HIT patients (Table 5). This provides
an alternative assessment of the spectrum of thrombotic complications
in HIT. Three large studies (Warkentin and Kelton, 1996; Nand et al. 1997;
Wallis et al., 1999) showed a predominance of venous thrombosis complicat-
ing HIT. Indeed, pulmonary embolism was even more frequent than all types
of arterial thromboses combined.
In contrast, a different spectrum of thrombotic complications was
reported by investigators at the University of Missouri–Columbia Health
Sciences Center (Silver et al., 1983; Laster et al., 1987; Almeida et al., 1998b).
Arterial, rather than venous, thromboembolism predominates in these pa-
tient series. Because this work is from the perspective of a vascular surgery
service, it is possible that patients with arterial thrombosis are either more
likely to be recognized as having HIT, or greater numbers of patients with
preexisting arteriopathy are treated with heparin and thus at higher risk for
developing arterial thrombosis if HIT develops.
Another pattern that emerges from the Missouri series is a progressively
decreasing frequency of reported thrombotic or hemorrhagic complications,
from 61% in 1983, to 23% in 1987, then to 7.4% in 1998. The authors believe
this to be the result of earlier recognition of HIT. However, an alternative
explanation could be greater awareness of HIT over time, and thus a higher
likelihood of identifying patients with less severe HIT. Indeed, a study by
Wallis and colleagues (1999) suggests that earlier recognition of HIT may not
reduce the risk of thrombosis (see Table 5).
A progressive reduction in HIT-associated mortality over time was also
observed by the Missouri group (see Table 5). However, early discontinuation
of heparin was not associated with significantly lower mortality in another
study (Wallis et al., 1999). This issue is complicated by the observation that
deaths apparently unrelated to thrombosis are relatively common in patients
with HIT (Warkentin and Kelton, 1996; Greinacher et al., 1999).
It is possible that nonthrombotic mortality may be higher than expected
by chance in patients with HIT. This speculation is based on the observation
that only a minority of patients who form HIT antibodies develop HIT (dis-
cussed subsequently); a corollary to this statement is that comorbid factors
that tend to result in increased pathogenicity of HIT antibodies may also
independently contribute to increased patient morbidity and mortality (i.e.,
patients with septicemia or multisystem organ failure may be more likely to
have platelet activation in the presence of HIT antibodies than ‘‘well’’
patients).
Table 5 Frequency of Thrombosis Complicating HIT in Retrospective Cohort Studies
Mean platelet Patients with Ratio of

Patients count nadir thrombosis venous/arterial Number of
Study with HIT (109/L) (%) thrombosis deaths (%)
Frequency of HIT
Warkentin and Kelton, 1996 127 59a 97 (76%) 4.3 26 (20%)

Subgroup with ‘‘isolated’’ 62 57 32 (52%)b 4 13 (21%)
thrombocytopenia
Nand et al., 1997 108 58 32 (29%) 2.5 5 (5%)c
Wallis et al., 1999 113 54 43 (38%) 1.4 31 (27%)
Subgroup with heparin 40 56 18 (45%) 1.4 10 (25%)
cessation <48 h
Subgroup with heparin 73 54 25 (34%) 1.4 21 (29%)
cessation >48 h
Silver et al., 1983 62 Range: 5–83 38 (61%) 0.6 20 (32%)d
Laster et al., 1987 169 57 30 (18%) 0.5 20 (12%)
Almeida et al., 1998b 94e >108 7 (7%) 0.6f 0 (0%)
a
The mean platelet count nadir for 127 patients with HIT and platelet count <150 109/L, and the median platelet count nadir for all 142
patients diagnosed with HIT (including those whose platelet count nadir was z150 109/L), were both 59 109/L (Warkentin, 1998a) (see
Fig. 6 in Chap. 3).
b
The cumulative 30-day frequency of new thrombosis in patients with isolated thrombocytopenia following recognition of HIT was 52.8% by
Kaplan-Meier analysis.
c
Only deaths in patients who developed thrombosis were reported. Total number of deaths in the HIT cohort was not reported.
d
Fourteen of the 20 deaths were judged to be caused by HIT-associated thrombosis.
e
Of 100 consecutive patients with positive in vitro testing. 6 were previously known to have heparin-dependent antibodies and were not
subsequently reexposed to heparin.
f
Two thromboses of arteriovenous grafts were excluded from classification into arterial or venous thrombosis.
127
A. Natural History of Isolated HIT

Isolated HIT is defined as the initial recognition of HIT because of throm-
bocytopenia alone, rather than because symptoms or signs of thrombosis
draw attention to the possibility of underlying HIT. A large retrospective
cohort study (Warkentin and Kelton, 1996) suggests that the subsequent
frequency of new, progressive, or recurrent thrombosis is relatively high in
such a patient population with serologically confirmed HIT (see Fig. 2).
Although these data are retrospective, the investigators attempted to mini-
mize bias. First, the date that the HIT assay was ordered was used as an
objective marker of first suspicion of the diagnosis of HIT. Second, patients
were excluded from analysis if there was any evidence in the medical records
to suggest the possibility of new signs or symptoms of thrombosis that may
have caused the physician to suspect HIT. In other words, efforts were made
to identify patients in whom HIT was suggested because of thrombocytopenia
alone. Finally, only objectively documented new, progressive, or recurrent
thrombotic events were analyzed.
Figure 2 Cumulative frequency of thrombosis in HIT patients presenting with

isolated thrombocytopenia. Approximately 50% of HIT patients initially recognized
with isolated thrombocytopenia developed objective evidence for thrombosis during
the subsequent 30-day period. The 1- and 2-day thrombotic event rates were ap-
proximately 10 and 18%, respectively. (From Warkentin and Kelton, 1996.)
The study identified 62 patients who met the definition of isolated

thrombocytopenia. The 30-day cumulative risk for thrombosis in this study
was 52.8% (see Fig. 2). This risk did not differ whether the heparin had been
discontinued, or whether warfarin had been substituted for the heparin.
Similar findings were reported from a much smaller earlier study performed
in Europe (Boon et al., 1994). This high risk for thrombosis in HIT is also
supported by a prospective study (Warkentin et al., 1995), in which five of six
HIT patients developed thrombosis either on the first day that their platelet
count fell below 150 109/L or within the next few days despite the dis-
continuation of heparin.
Subsequent to the Hamilton study, a report by Wallis and colleagues
(1999) from Loyola University confirmed the high risk for thrombosis among
patients in whom HIT is identified by platelet count monitoring, even with
discontinuation of heparin (see Table 5). Overall, the 30-day thrombotic event
rate was 43/113 (38%), with a ratio of venous to arterial thrombosis of just
1.4. The relatively low predominance of venous thrombosis could be ex-
plained by the large number of patients (59%) in this study who developed
HIT following cardiac surgery (i.e., a patient population had relatively high
risk for arterial thrombosis).
An intriguing finding of the Wallis report is that early cessation of
heparin did not appear to improve clinical outcomes. For 40 of the 113
patients with HIT (35%), heparin was discontinued within 48 h of onset of
thrombocytopenia (defined as a platelet count fall to less than 100 109/L, or
a greater than 50% fall from the peak platelet count after initiating heparin).
Indeed, there was a trend to a higher rate of thrombosis in the patients with
early heparin cessation, compared with the remaining 65% of patients in
whom heparin was stopped later (45 vs. 34%; p = 0.26) (see Table 5).
Further evidence supporting an unfavorable natural history of untreat-
ed HIT was provided by a large prospective cohort study (Greinacher et al.,
2000). These investigators found that the thrombotic event rate was 6.1% per
day during the mean 1.7-day interval between diagnosis of HIT (and cessation
of heparin) and initiation of lepirudin therapy. This event rate (6.1 1.7 =
10.4%) corresponds closely to the 10% rate of thrombosis observed in the
Hamilton study in the first 48 h following diagnosis of isolated HIT
(Warkentin and Kelton, 1996) (see Fig. 2).
Taken together, these studies of the natural history of isolated HIT
provide the basis for the recent recommendation that prophylactic anticoag-
ulant therapy is appropriate for most patients with isolated HIT (Hirsh et al.,
2001) (see Chaps. 1 and 13–16). Other data to support this concept include
the high probability of detecting subclinical deep vein thrombosis by duplex
ultrasonography in patients with isolated HIT (Tardy et al., 1999), as well as
the persistence of marked in vivo thrombin generation for several days in
Table 6 Studies Describing Systematic Screening for HIT Antibodies Using Sensitive Assays in Patients Receiving Heparin
Heparin (porcine Patients
130
UFH used unless HIT assay Number of with HIT Patients with
Study Trial design otherwise indicated) used patients antibodies (%) HIT (%)
Medical patients
Amiral et al., 1996 Retrospective iv ther UFH EIA-IgM/A/G 19 (17.4)
109 1 (0.9)a
EIA-Ig 3 (2.8)
Kappers-Klunne Prospective iv ther UFH EIA-IgG 9 (2.5) 2 (0.6)
358
et al., 1997 HIPA 30 (8.4) 0 (0)
Hemodialysis patients
Greinacher et al., 1996 Prevalence study iv ther UFH HIPA 165 7 (4.2%) 0 (0)
de Sancho et al., 1996 Prevalence study iv ther UFH EIA-IgM/A/G 45 0 (0) 0 (0)
EIA-IgM/G 4 (3.1)
iv ther UFH 128 0 (0)b
f EIA-IgG 3 (2.3)
Boon et al., 1996 Prevalence study EIA-IgM/G 1 (0.8)
LMWH 133 0 (0)b
f EIA-IgG 1 (0.8)
Luzzatto et al., 1998 Prevalence study iv ther UFH EIA-IgG 50 6 (12.0) 0 (0)
Orthopedic postoperative surgical patients
Warkentin et al.,
sc proph UFH EIA-IgG 29 (14.1)
205 10 (4.9)
1995, 2000 Substudy of RCT SRA 19 (9.3)
sc proph LMWH EIA-IgG 11 (6.0)
182 2 (1.1)
SRA 5 (2.7)c
Warkentin et al., 2000 Prospective sc proph LMWH EIA-IgG 22 (8.6)
257 2 (0.8)
SRA 9 (3.5)c
Amiral et al., 1996 Retrospective sc proph LMWH EIA-IgM/A/G 8 (8.0)
100 0 (0)
EIA-IgG 2 (2.0)
Cardiac postoperative surgical patients (all received porcine UFH at cardiopulmonary bypass except where otherwise stated)
Visentin et al., 1996 Retrospective CPB: UFH EIA-IgM/G 44 27 (61.4)d
0 (0)
(NPH) EIA-IgG 23 (52.3)
Bauer et al., 1997 Prospective CPB: bovine UFH EIA-IgM/A/G 111 57 (51.4)e
0 (0)
(NPH) SRA 23 (52.3)
Trossaert et al., 1998 Retrospective CPB: UFH EIA-IgM/A/G 51 14 (27.5)f
Lee and Warkentin
0 (0)
sc proph UFH EIA-IgG 9 (17.6)
PRP 2 (3.9)

EIA-IgM/A/G 46 (29.3)
Pouplard et al., 1999 Prospective sc proph UFH EIA-IgG 157 24 (15.3) 6 (3.8)
6 (3.8)
SRA
EIA-IgM/A/G 37 (21.6)
sc proph LMWH EIA-IgG 171 24 (14.0) 0 (0)
SRA 2 (1.2)
Warkentin et al., 2000 Prospective CPB: UFH EIA-IgG 50 (50.0)
100 1 (1.0)
Frequency of HIT
sc proph UFH SRA 20 (20.0)

Francis et al., 2003 Prospective RCT CPB: bovine UFHg EIA-IgM/A/G 99 44 (44.4)h 0 (0)
CPB: porcine UFHg EIA-IgM/A/G 108 33 (30.6)h 0 (0)
Vascular surgical patients
Jackson et al., 1998 Prospective iv ther UFH EIA-IgM/G 54 3 (5.6) j
Lindhoff-Last Prospective iv ther UFH EIA-IgM/A/G 50 17 (34.0)
et al., 2000 7 days, EIA-IgG 48 6 (12.5)
0 (0)
then sc proph HIPA 48 3 (6.3)i
LMWH
Abbreviations: CPB, cardiopulmonary bypass; EIA, PF4-heparin enzyme-linked immunosorbent assay (-IgM/A/G, one or more of IgM, IgA,
and IgG antibodies present; -IgG, IgG antibodies only present); HIPA, heparin-induced platelet activation assay; iv ther, intravenous therapeutic-
dose heparin; LMWH, low molecular weight heparin; MI, myocardial infarction; RCT, randomized controlled trial; sc proph, subcutaneous
prophylactic dose heparin: SRA, serotonin release assay using washed platelets; UFH, unfractionated heparin; VTE, venous thromboembolism,
NPH, no postoperative heparin.
a
Thrombocytopenia defined as platelet count fall >50% from baseline.
b
Two patients with HIT antibodies had mild thrombocytopenia, but a causal relation to heparin was not stated.
c
20% serotonin release used in Warkentin et al. (2000) rather than 50% serotonin release cutoff used in Warkentin et al. (1995).
d
Ten patients had HIT-IgG preoperatively: incidence of new seroconversion was 17/44 (38.6%).
e
Incidence of new seroconversion was 43% for EIA and 9% for SRA.
f
Two patients had HIT-IgG preoperatively: incidence of new seroconversion was 12/51 (23.5%) for EIA-IgM/A/G and 7/51 (13.7%) for EIA-IgG.
None had a positive aggregation assay preoperatively.
g
80.2% underwent CPB, remainder ‘‘off-pump’’ surgery; 18.8% received postoperative UFH, LMWH, or both.
h
Excludes patients testing positive for HIT antibodies at baseline.
i
Includes only HIPA patients who also tested positive in EIA.
j
131
Two patients had HIT-IgG preoperatively: incidence of new seroconversion was 1/54 (1.9%).
patients with acute HIT even following discontinuation of heparin (Warken-

tin et al., 1997; Greinacher et al., 2000).
B. Summary of Observations from Prospective and

Retrospective Studies
Observations emerging from these studies include the following:
1. The risk of thrombosis in patients with HIT is higher than pre-
viously recognized (up to 50%), and remains high despite the
discontinuation of heparin. Mortality in patients with HIT is sig-
nificant, although it remains uncertain what proportion is related to
HIT-associated thrombosis, and whether these can be prevented by
effective treatment.
2. Most thrombotic events are venous, rather than arterial, although
this predominance may not be observed in patient populations at
high risk for arterial disease. Pulmonary embolism may be the most
frequent life-threatening consequence of HIT.
V. POPULATION-BASED STUDIES OF HIT ANTIBODY

SEROCONVERSION
Usually, serological investigation for HIT antibodies is performed on pa-

tients who develop thrombocytopenia during heparin treatment. Since 1995,
however, many studies have performed systematic studies of HIT antibody
seroconversion using sensitive assays (EIA, SRA, or both), irrespective of
whether or not thrombocytopenia occurred. Some interesting insights into the
pathogenesis of HIT have emerged from these reports.
As shown in Table 6, three general types of patient population have
been investigated: medical patients receiving therapeutic-dose UFH; ortho-
pedic patients receiving UFH or LMWH; and cardiac surgical patients
receiving UFH or LMWH. There appear to be distinct frequencies of HIT
antibody formation, as well as varying risks of ‘‘breakthrough’’ of HIT,
among these different populations (Fig. 3). Several observations emerge from
these studies:
1. The prevalence of seroconversion depends on the diagnostic assay
used. PF4–heparin EIA is more sensitive than the SRA for the
detection of HIT antibodies (Bauer et al., 1997; Pouplard et al.,
1999; Warkentin et al., 2000); however, this increase in sensitivity
does not necessarily translate into greater predictive value for
clinical HIT (see Chap. 11).
Figure 3 Variable frequency of HIT antibody formation and clinical HIT among
different patient populations treated with UFH or LMWH. A schematic ‘‘iceberg,’’
shown on lower line, illustrates the relation among HIT-associated thrombosis,
thrombocytopenia, HIT antibodies detected by serotonin release assay (SRA), and
HIT antibodies detected by enzyme-immunoassay (EIA). The size of the iceberg re-
flects the relative frequency of HIT antibody formation by EIA (i.e., the cardiac-UFH
iceberg is about six times larger than the orthopedic-LMWH iceberg [50 vs. 8%
frequency of HIT antibody formation]). Noteworthy aspects include the observation
that HIT-associated thrombosis is most common in orthopedic-UFH patients, even
though HIT antibody formation is most common in cardiac-UFH patients, and the
observation that orthopedic-LMWH has a higher frequency of thrombosis than does
medical-UFH.
2. With use of PF4–heparin EIA, the frequency of seroconversion

following cardiac surgery approaches 50% (Visentin et al., 1996;
Bauer et al., 1997; Warkentin et al., 2000). A high frequency of
seroconversion (13–20%) was also observed using the SRA. Despite
the highest frequency of HIT seroconversion reported in this patient
population, the likelihood of developing HIT appears to be less than
in other orthopedic patients also treated with postoperative UFH.
3. Seroconversion occurs frequently without thrombocytopenia or
thrombosis. Indeed, most patients who form HIT antibodies do
not develop HIT. The proportion who develop HIT, however, is
highest among the patients who have a positive SRA. This sug-
gests that HIT antibodies ‘‘strong’’ enough to activate platelets are
more likely to be clinically significant. Patient-dependent factors

also must be important, however, because the probability of a
positive SRA indicating clinical HIT ranges from about <10%
(cardiac surgery) to approximately 50% (orthopedic surgery).
4. Regardless of which diagnostic assay is used, new seroconversion
occurs more frequently after exposure to UFH than LMWH (War-
kentin et al., 1995, 2000, 2003a; Amiral et al., 1996; Lindhoff-Last
et al., 2002; Ahmad et al., 2003a).
A. HIT in Patients Undergoing Cardiac Surgery

Three prospective studies have evaluated the frequency of HIT in postoper-
ative cardiac surgical patients who also have received postoperative anti-
thrombotic prophylaxis with UFH (Trossaert et al., 1998; Pouplard et al.,
1999, 2002; Warkentin et al., 2000). Pooling the data, the frequency of HIT
appears to be about 2%. This frequency is consistent with a number of
retrospective studies (Glock et al., 1988; Walls et al., 1992a,b; Singer et al.,
1993) that reported a frequency of HIT of up to 5%, but overall, also noted a
frequency of about 2% (Table 7). Furthermore, HIT was associated with a
Table 7 Frequency of HIT and Thrombosis in Retrospective Studies of HIT

in Cardiovascular Surgery Patients
Patients Ratio of Total deaths

Patients Patients with HIT and venous: in patients
at risk, with HIT, thrombosis, arterial with HIT,
Study n n (%) n (%) thrombosis n (%)
Walls et al., 4261 82 (1.9) 31 (38) 0.3:1 23 (28)

1992a
Walls et al., 764 35 (4.5) 17 (49) 0.3:1 15 (43)
1992b
Visentin et al., 51 0 (0) — — —
1996
Glock et al., — 21 17 (81) 0.7:1 8 (38)
1988
Singer et al., 1500 11 (0.75) 7 (64)a 0.3:1b 2 (18)
1993
a
In seven patients, 17 thrombotic events occurred.
b
Precise number of arterial and venous events is unclear from the published data. For this
analysis, of six limb amputations associated with intravascular catheters or devices, five were
assumed to be arterial and one venous, based on the type of intravascular catheter or device
associated with the amputated limb.
risk of thrombosis of 38–81%, and with an overall mortality of 18–43% in

these studies. In contrast to the orthopedic patient population, the predom-
inant thrombotic event appears to be arterial.
Only one study has examined the influence of postoperative antithrom-
botic prophylaxis with UFH or LMWH on the frequency of HIT antibody
formation and HIT following heart surgery (Pouplard et al., 1999, 2002). In
this prospective, but nonrandomized comparative trial, one group of patients
(n = 263) judged to be at highest risk for thrombosis received postoperative
therapeutic-dose UFH (200 U/kg once daily by subcutaneous injection,
adjusted by activated partial thromboplastin time, for at least 10 days). The
second group of patients (n = 370) received LMWH.
Overall, formation of HIT antibodies was similar in the two patient
groups, using a commercial EIA that detects IgM, IgA, and IgG antibodies.
However, among 19 patients who tested positive in the platelet activation
assay (serotonin release), 10 evinced a platelet count fall consistent with HIT.
Further, there appeared to be an association with HIT and postoperative
treatment with UFH (9/263=3.4%) compared with LMWH (1/370 = 0.3%)
(Pouplard et al., 2002). Thus,
1. The frequency of HIT antibody formation following heart surgery is
influenced primarily by UFH given at cardiopulmonary bypass,
rather than the type of heparin preparation given postoperatively.
2. Among patients who form HIT antibodies following heart surgery,
the risk of HIT may be higher in those receiving UFH (however,
this cannot be definitively concluded because the study was non-
randomized and UFH and LMWH dosing differed).
3. The study indicates a greater clinical usefulness of the platelet sero-
tonin release assay (SRA), compared with the PF4–heparin EIA (see
Chap. 11).
The potential to reduce risk of ‘‘breakthrough’’ of HIT among postcardiac
surgery patients who form HIT antibodies is a major reason why antithrom-
botic agents with low (danaparoid) or possibly absent (antithrombin-binding
pentasaccharide, fondaparinux) cross-reactivity against PF4–heparin might
be ideal anticoagulants for this clinical situation (Warkentin et al., 2003) (see
Chap. 8).
B. Anti-Xa–Inhibiting Pentasaccharide
Fondaparinux (Arixtra) is a novel antithrombin-binding pentasaccharide
that inhibits factor Xa without inhibiting thrombin (factor IIa). It has been
shown to be safe and effective for antithrombotic prophylaxis following
orthopedic surgery. Fondaparinux does not bind to PF4, and therefore its use
might avoid HIT. In systematic studies of HIT antibody formation associated
with fondaparinux or enoxaparin prophylaxis after elective hip or knee re-
placement therapy, low frequencies of HIT antibody formation were seen
with both anticoagulants (Warkentin et al., 2003b). However, whereas HIT
antibodies invariably ‘‘cross-reacted’’ with enoxaparin, they failed to recog-
nize PF4 mixed with fondaparinux (by fluid-phase EIA) (see Chap. 11).
Thus, fondaparinux offers the possibility of absent or negligible risk of
causing HIT.
C. Implications of Subclinical HIT Antibody Seroconversion

Two retrospective studies found that patients with acute coronary syndrome
who either had HIT antibodies at presentation (Williams et al., 2003) or
developed HIT antibodies following heparin treatment (Mattioli et al., 2000)
had a higher frequency of vascular events during long-term follow-up ranging
from 1 month to 1 year, despite absence of clinical HIT in any patient. Un-
recognized confounders could explain these findings; e.g., a short interval
from recent hospitalization/heparin exposure (former study) or a longer du-
ration time of hospitalization (latter study) could be associated independently
with adverse outcomes and greater likelihood of forming HIT antibodies.
VI. VARIABLE FREQUENCY OF HIT: IMPLICATIONS

FOR PLATELET COUNT MONITORING
Until recently, studies of HIT frequency have yielded seemingly confusing

and inconsistent results. However, as argued in this chapter, by taking into
consideration (1) type of heparin used, (2) patient population treated, and (3)
laboratory and clinical evidence to distinguish (immune) HIT from nonim-
mune HAT, distinct profiles for HIT antibody seroconversion, HIT itself,
and HIT-associated thrombosis can be discerned (see Fig. 3). New research
questions will be generated in the search for the biological basis for these
intriguing differences in HIT risk. But perhaps the most important insight to
emerge from these collective studies is the simple and clinically relevant ob-
servation that new, progressive, or recurrent thrombosis occurs in at least 33–
50% of patients who develop proven HIT. This underscores the need for
prompt recognition and urgent therapy in all patients suspected of having this
adverse drug reaction.
Practically, these findings suggest strategies for platelet count monitor-
ing in patients receiving heparin. Some physicians are hesitant to institute
Table 8 Recommendations for Platelet Count Monitoring for Heparin-Induced

Thrombocytopenia During Heparin Treatment
1. Monitoring for typical-onset HIT: stratifying the intensity of platelet count

monitoring for HIT based upon its risk
A. Patients at highest risk for HIT (1–5%) (e.g., postoperative patients
receiving prophylactic-dose UFH after major surgery): monitoring during
heparin therapy, at least every second day from day 4 to day 14a
Patients receiving therapeutic-dose UFH: platelet count monitoring once
dailyb from day 4 to day 14a
B. Patients at intermediate risk for HIT (0.1–1%) (e.g., medical/obstetrical
patients receiving prophylactic-dose UFH, or postoperative patients
receiving prophylactic-dose LMWH, or postoperative patients receiving
intravascular catheter ‘‘flushes’’ with UFH): monitoring during heparin
therapy, at least every 2 or 3 days from day 4 to day 14a, when practicalc
C. Patients at low risk for HIT (<0.1%) (e.g., medical/obstetrical patients
receiving prophylactic- or therapeutic-dose LMWH, or medical patients
receiving only intravascular catheter ‘‘flushes’’ with UFH): routine platelet
count monitoring is not recommendedd
2. Monitoring for rapid-onset HIT: for a patient recently exposed to heparin
(within the past 100 days), a repeat platelet count within 24 hours following
reinitiation of heparin
3. When to suspect HIT:
A relative (proportional) platelet count fall of 50% or greater that is otherwise
clinically unexplained should be considered suspicious for HIT, even if the
platelet count nadir remains above 150 109/L.
For any patient who develops thrombosis during or within several days after
heparin therapy, or who develops an unusual clinical event in association with
heparin therapy (e.g., inflammatory or necrotic skin lesions at heparin injec-
tion sites, acute systemic reaction post–intravenous heparin therapy), a repeat
platelet count should be measured promptly and compared with recent values.
These are draft recommendation (Seventh American College of Chest Physicians Consensus
Conference on Antithrombotic Therapy, September 2003). Readers should consult the pub-
lication (Warkentin and Greinacher, 2004) to obtain the final recommendations.
a
The crucial time period for monitoring ‘‘typical-onset’’ HIT is between days 4 to 14 (first day
of heparin = day 0), where the highest platelet count from day 4 (inclusive) onwards
represents the ‘‘baseline.’’ Platelet count monitoring can cease before day 14 when heparin is
stopped.
b
Once-daily platelet count monitoring recommended as daily blood draws required for aPTT
monitoring.
c
Frequent platelet count monitoring may not be practical when UFH or LMWH is given to
outpatients.
d
Monitoring as per ‘‘intermediate’’ risk is appropriate if UFH was given before initiating
LMWH.
regular platelet count monitoring for HIT. One explanation is the almost
ubiquitous use of heparin in hospitalized patients. Thus, a requirement that
regular, perhaps even daily, platelet count monitoring be performed seems
excessive. Additionally, there is no convincing evidence that regular platelet
count monitoring can prevent the thrombotic complications of HIT if the
physician response is merely to stop the heparin (Wallis et al., 1999).
However, a worthy consideration is that instituting alternative, parenteral
anticoagulation may prevent thrombosis in patients recognized as having
isolated HIT.
These comments notwithstanding, marked differences in risk for HIT
are apparent among different patient populations. Thus, it seems prudent to
recommend that patients at the highest risk of HIT, and for HIT-associated
thrombosis (e.g., postoperative patients receiving UFH, or any patient
receiving bovine lung UFH) should have platelet counts monitored regularly,
perhaps at least every other day. For patients whose risk for HIT appears to
be 0.1–1% (e.g., medical patients receiving UFH, surgical patients receiving
LMWH), less frequent monitoring may be appropriate. Since HIT is unlikely
to occur before day 5, or after day 14, the monitoring could be performed 2 or
3 times per week from days 4 to 14. Most patients have frequent complete
blood counts performed during the first few days of hospitalization, so
comparative platelet count results for days 0–3 are usually available.
Two recent consensus conferences have examined the issue of platelet
count monitoring for HIT (Warkentin, 2002; Warkentin and Greinacher,
2004). Although the recommendations were not identical, they had in common
the concept of stratifying the intensity of platelet count monitoring based upon
the risk of developing HIT and focusing the monitoring during the time when
HIT usually occurs. Table 8 summarizes draft recommendations (Warkentin
and Greinacher, 2004).
Regardless of the intensity of surveillance, all physicians who monitor
platelet counts need to understand how to distinguish HIT from nonimmune
HAT, because diagnostic confusion may lead to inappropriate decisions to
discontinue heparin therapy in patients with nonimmune HAT who otherwise
require anticoagulation because of high risk for thrombosis. Irrespective of
whether platelet count monitoring is being performed, HIT should be
considered promptly in the differential diagnosis of any patient who develops
symptoms or signs of new, progressive, or recurrent thrombosis during or
within a few days of discontinuing heparin treatment.
ACKNOWLEDGMENTS
Studies described in this chapter were supported by grants from the Heart
and Stroke Foundation of Ontario. Dr. Warkentin was a Research Scholar of
the Heart and Stroke Foundation of Canada. Dr. Lee was supported by a
Research Fellowship from the Heart and Stroke Foundation of Canada.
REFERENCES
Ahmad S, Haas S, Hoppensteadt DA, Lietz H, Reid U, Bender N, Messmore HL,

Misselwitz F, Bacher P, Gaikwad BS, Jeske WP, Walenga JM, Fareed J.
Differential effects of clivarin and heparin in patients undergoing hip and
knee surgery for the generation of anti-heparin–platelet factor 4 antibodies.
Thromb Res 2003a; 108:49–55.
Ahmad S, Bacher P, Lassen MR, Hoppensteadt DA, Leitz H, Misselwitz F, Walenga
JM, Fareed J. Investigations of the immunoglobulin subtype transformation
of anti-heparin–platelet factor 4 antibodies during treatment with a low-
molecular-weight heparin (Clivarin) in orthopedic patients. Arch Pathol Lab
Med 2003b; 127:584–588.
Almeida JI, Liem TK, Silver D. Heparin-bonded grafts induce platelet aggregation
in the presence of heparin-associated antiplatelet antibodies. J Vasc Surg
1998a; 27:896–901.
Almeida JI, Coats R, Liem TK, Silver D. Reduced morbidity and mortality rates of
the heparin-induced thrombocytopenia syndrome. J Vasc Surg 1998b; 27:309–
316.
Alving BM, Shulman NR, Bell WR, Evatt BL, Tack KM. In vitro studies of heparin-
induced thrombocytopenia. Thromb Res 1977; 11:827–834.
Amiral J, Peynaud-Debayle E, Wolf M, Bridey F, Vissac A-M, Meyer D. Generation
of antibodies to heparin-PF4 complexes without thrombocytopenia in patients
treated with unfractionated or low-molecular-weight heparin. Am J Hematol
1996; 52:90–95.
Anderson GP. Insights into heparin-induced thrombocytopenia. Br J Haematol
1992; 80:504–508.
Ansell J, Slepchuk N Jr, Kumar R, Lopez A, Southard L, Deykin D. Heparin-induced
thrombocytopenia: a prospective study. Thromb Haemost 1980; 43:61–65.
Ansell JE, Price JM, Shah S, Beckner RR. Heparin-induced thrombocytopenia:
What is its real frequency? Chest 1985; 88:878–882.
Bailey RT Jr, Ursick JA, Heim KL, Hilleman DE, Reich JW. Heparin-associated
thrombocytopenia: a prospective comparison of bovine lung heparin, manu-
factured by new process, and porcine intestinal heparin. Drug Intell Clin
Pharm 1986; 20:374–378.
Ball A, L’Huillier AM, Dreyfuss L, Porte JL, Barthel JC. Thrombopénie à la
Fraxiparine. Une observation. Presse Méd 1989; 18:1254–1255.
Barradas MA, Mikhailidis DP, Epemolu O, Jeremy JY, Fonseca V, Dandona P.
Comparison of the platelet proaggregatory effect of conventional unfraction-
ated heparins and a low molecular weight heparin fraction (CY 222). Br J
Haematol 1987; 67:451–457.
Bauer TL, Arepally G, Konkle BA, Mestichelli B, Shapiro SS, Cines DB, Poncz M,
McNulty S, Amiral J, Hauck WW, Edie RN, Mannion JC. Prevalence of
heparin-associated antibodies without thrombosis in patients undergoing car-

diopulmonary bypass surgery. Circulation 1997; 95:1242–1246.
Bell WR. Heparin-associated thrombocytopenia and thrombosis. J Lab Clin Med
1988; 111:600–605.
Bell WR, Royall RM. Heparin-associated thrombocytopenia: a comparison of three
heparin preparations. N Engl J Med 1980; 303:902–907.
Bell WR, Tomasulo PA, Alving FM, Duffy TP. Thrombocytopenia occurring during
the administration of heparin. A prospective study in 52 patients. Ann Intern
Med 1976; 85:155–160.
Boon DMS, Michiels JJ, Stibbe J, van Vliet HHDM, Kappers-Klunne MC. Heparin-
induced thrombocytopenia and antithrombotic therapy [letter]. Lancet 1994;
344:1296.
Boon DMS, van Vliet HHDM, Zietse R, Kappers-Klunne MC. The presence of
antibodies against a PF4-heparin complex in patients on haemodialysis [letter].
Borowiec J, Thelin S, Bagge L, Hultman J, Hansson H-E. Decreased blood loss after
cardiopulmonary bypass using heparin-coated circuit and 50% reduction of
heparin dose. Scand J Thorac Cardiovasc Surg 1992a; 26:177–185.
Borowiec J, Thelin S, Bagge L, Nilsson L, Venge P, Hansson HE. Heparin-coated
circuits reduce activation of granulocytes during cardiopulmonary bypass. A
clinical study. J Thorac Cardiovasc Surg 1992b; 104:642–667.
Borowiec JW, Bylock A, van der Linden J, Thelin S. Heparin coating reduces blood
cell adhesion to arterial filters during coronary bypass: a clinical study. Ann
Thorac Surg 1993; 55:1540–1545.
induced thrombocytopenia and thrombosis: clinical and laboratory studies. Br
J Haematol 1993; 84:322–328.
Bouvier C. In: Van Aken WG, ed. Thrombocytopenia (and consumption coagu-
lopathy) induced by heparin. A case report [discussion]. Scand J Haematol 1980;
25 (suppl 36):85–90.
Brushwood DB. Hospital liable for allergic reaction to heparin used in injection
flush. Am J Hosp Pharm 1992; 49:1491–1492.
Casu B, Johnson EA, Mantovani M, Mulloy B, Oreste P, Pescador R, Prino G, Torri
G, Zoppetti G. Correlation between structure, fat-clearing and anticoagulant
properties of heparins and heparan sulphates. Arzneimittal, forschung Drug
Res 1983; 33:135–142.
Chong BH, Castaldi PA. Platelet proaggregating effect of heparin: possible mecha-
nism for nonimmune heparin-associated thrombocytopenia. Aust NZ J Med
1986; 16:715–716.
Chong BH, Ismail F. The mechanism of heparin-induced platelet aggregation. Eur J
Haematol 1989; 43:245–251.
Cipolle RJ, Rodvoid KA, Seifert R, Clarens R, Ramirez-Lassepas M. Heparin-
associated thrombocytopenia: a prospective evaluation of 211 patients. Ther
Drug Monit 1983; 5:205–211.
de Raucourt E, Vinsonneau C, Juvin K, Fischer AM, Meyer G. Heparin-induced

thrombocytopenia with thrombotic complications during prophylactic treat-
ment with low-molecular-weight heparin. Blood Coagul Fibrinolysis 1996; 7:
786–788.
de Sancho M, Lema MG, Amiral J, Rand J. Frequency of antibodies directed
against heparin-platelet factor 4 in patients exposed to heparin through
chronic hemodialysis [letter]. Thromb Haemost 1996; 75:695–696.
Dittberner T, Schöttler E, Ranze O, Greinacher A, Knobler R. Heparin-induced
thrombocytopenia: a complication in extracorporeal photochemotherapy (pho-
topheresis) [letter]. J Am Acad Dermatol 2002; 47:452–453.
Doty JR, Alving BM, McDonnell DE, Ondra SL. Heparin-associated thrombocy-
topenia in the neurosurgical patient. Neurosurgery 1986; 19:69–72.
Elalamy I, Potevin F, Lecrubier C, Bara L, Marie JP, Samama MM. A fatal low-
molecular-weight heparin-associated thrombocytopenia after hip surgery: pos-
sible usefulness of PF4-heparin ELISA test. Blood Coagul Fibrinolysis 1996;
7:665–671.
Eldh P, Jacobsson B. Heparinized vascular catheters: a clinical trial. Radiology 1974;
111:289–292.
Elgue G, Blombäck M, Olsson P, Riesenfeld J. On the mechanism of coagulation
inhibition on surfaces with end point immobilized heparin. Thromb Haemost
1993; 70:289–293.
Ellison J, Walker ID, Greer IA. Antenatal use of enoxaparin for prevention and
treatment of thromboembolism in pregnancy. BJOG 2000; 107:1116–1121.
ENOXACAN Study Group. Efficacy and safety of enoxaparin versus unfractionated
heparin for prevention of deep vein thrombosis in elective cancer surgery: a
double-blind randomized multicentre trial with venographic assessment. Br J
Surg 1997; 84:1099–1103.
Fausett MB, Vogtlander M, Lee RM, Esplin MS, Branch DW, Rodgers GM, Silver
RM. Heparin-induced thrombocytopenia is rare in pregnancy. Am J Obstet
Gynecol 2001; 185:148–152.
Follea G, Hamandijan I, Trzeciak MC, Nedey C, Streichenberger R, Dechavanne
M. Pentosane polysulfate associated thrombocytopenia. Thromb Res 1986;
42:413–418.
Francis JL, Palmer GJ, Moroose R, Drexler A. Comparison of bovine and porcine
heparin in heparin antibody formation after cardiac surgery. Ann Thorac Surg
2003; 75:17–22.
Gallus AS, Goodall KT, Beswick W, Chesterman CN. Heparin-associated throm-
bocytopenia: case report and prospective study. Aust NZ J Med 1980; 10:25–
31.
Ganzer D, Gutezeit A, Mayer G, Greinacher A, Eichler P. Thromboemboliepro-
phylaxe als auslöser thrombembolischer Komplicationen. Eine Untersuchung
zur inzidenz der Heparin-induzierten Thrombozytopenie (HIT) Typ II. Z
Orthop 1997; 135:543–549.
Baggio G, Fabris F, Girolami A. The incidence of heparin-induced throm-
bocytopenia in hospitalized medical patients treated with subcutaneous un-

fractionated heparin: a prospective cohort study. Blood 2003; 101:2955–2959.
Glock Y, Szmil E, Boudjema B, Boccalon H, Fournial G, Cerene AL, Puel P.
Cardiovascular surgery and heparin-induced thrombocytopenia. Int Angiol
1988; 7:238–245.
Goad KE, Home MK III, Gralnick HR. Pentosan-induced thrombocytopenia:
support for an immune complex mechanism. Br J Haematol 1994; 88:803–808.
Gouault-Heilman M, Payen D, Contant G, Intrator L, Huet Y, Schaeffer A. Throm-
bocytopenia related to synthetic heparin analogue therapy [letter]. Thromb
Haemost 1985; 54:557.
Green D. Heparin-induced thrombocytopenia. Med J Aust 1986; 144(suppl):HS37–
HS39.
Green D, Martin GJ, Shoichet SH, DeBacker N, Bomalaski JS, Lind RN. Throm-
bocytopenia in a prospective, randomized, double-blind trial of bovine and
porcine heparin. Am J Med Sci 1984; 288:60–64.
Greinacher A. Antigen generation in heparin-associated thrombocytopenia: the
nonimmunologic type and the immunologic type are closely linked in their
pathogenesis. Semin Thromb Hemostas 1995; 21:106–116.
Greinacher A, Michels I, Schäfer M, Kiefel V, Muller-Eckhardt C. Heparin-
associated thrombocytopenia in a patient treated with polysulphated chon-
droitin sulphate: evidence for immunological crossreactivity between heparin
and polysulphated glycosaminoglycan. Br J Haematol 1992a; 81:252–254.
Greinacher A, Drost W, Michels I, Leitl J, Gottsmann M, Kohl HG, Glaser M,
Mueller-Eckhardt C. Heparin-associated thrombocytopenia successfully
treated with the heparinoid Org 10172 in a patient showing cross-reaction to
LMW heparins. Ann Haematol 1992b; 64:40–42.
Greinacher A, Michels I, Muller-Eckardt C. Heparin-associated thrombocytopenia:
the antibody is not heparin-specific. Thromb Haemost 1992c; 67:545–549.
Greinacher A, Amiral J, Dummel V, Vissac A, Keifel V, Mueller-Eckhardt C. Labo-
ratory diagnosis of heparin-associated thrombocytopenia and comparison of
factor 4/heparin enzyme-linked immunosorbent assay. Transfusion 1994;
34:381–385.
Greinacher A, Zinn S, Wizemann, Birk UW. Heparin-induced antibodies as a risk
factor for thromboembolism and haemorrhage in patients undergoing chronic
haemodialysis [letter]. Lancet 1996; 348:764.
P, Mueller-Velten HG, Pötzsch B. Recombinant hirudin (lepirudin) provides
effective and safe anticoagulation in patients with the immunologic type of
heparin-induced thrombocytopenia. Circulation 1999; 99:73–80.
Greinacher A, Eichler P, Lubenow N, Kwasny H, Luz M. Heparin-induced
thrombocytopenia with thromboembolic complications: meta-analysis of two
prospective trials to assess the value of parenteral treatment with lepirudin and
its therapeutic aPTT range. Blood 2000; 96:846–851.
Gruel Y, Pouplard C, Nguyen P, Borg JY, Derlon A, Juhan-Vague I, Regnault V,
Samama M, and the French Heparin-Induced Thrombocytopenia Study

Group. Br J Haematol 2003; 121:786–792.
Heeger PS, Backstrom JT. Heparin flushes and thrombocytopenia [letter]. Ann In-
tern Med 1986; 105:143.
Granger C, Ohman Em, Dalen JE. Heparin and low-molecular-weight heparin.
Mechanisms of action, pharmacokinetics, dosing, monitoring, efficacy, and
Holm HA, Eika C, Laake K. Thrombocytes and treatment with heparin from
porcine mucosa. Scand J Haematol 1980; 36(suppl):81–84.
Jackson MR, Gillespie DL, Chang AS, Longenecker EG, Peat RA, Alving B. The
incidence of heparin-induced antibodies in patients undergoing vascular sur-
gery: a prospective study. J Vasc Surg 1998; 28:439–445.
Johnson RA, Lazarus KH, Henry DH. Heparin-induced thrombocytopenia: a pros-
pective study. Am J Hematol 1984; 17:349–353.
Kakkasseril JS, Cranley JJ, Panke T, Grannan K. Heparin-induced thrombocyto-
penia: a prospective study of 142 patients. J Vasc Surg 1985; 2:382–384.
induced platelet activation in sixteen surgical patients: diagnosis and ma-
nagement. J Vasc Surg 1987; 5:101–109.
Kappers-Klunne MC, Boon DMS, Hop WCJ, Michiels JJ, Stibbe J, van der Zwaan
C, Koudstaal PJ, van Vliet HHDM. Heparin-induced thrombocytopenia and
thrombosis: a prospective analysis of the incidence in patients with heart and
cerebrovascular diseases. Br J Haematol 1997; 96:442–446.
Kelton JG. Heparin-induced thrombocytopenia. Haemostasis 1986; 16:173–186.
Konkle BA, Bauer TL, Arepally G, Cines DB, Poncz M, McNulty S, Edie RN,
Mannion JD. Heparin-induced thrombocytopenia: bovine versus porcine hepa-
rin in cardiopulmonary bypass surgery. Ann Thorac Surg 2001; 71:1920–1924.
Koul B, Vesterqvist O, Egberg N, Steen S. Twenty-four-hour heparin-free veno–
right ventricular ECMO: an experimental study. Ann Thorac Surg 1992;
53:1046–1051.
Larm O, Larsson R, Olsson P. A new non-thrombogenic surface prepared by
selective covalent binding of heparin via a modified reducing terminal residue.
Biomater Med Devices Artif Organs 1983; 11:161–173.
Larsson R, Larm O, Olsson P. The search for thromboresistance using immobilized
heparin. Ann NY Acad Sci 1987; 516:102–115.
Laster J, Silver D. Heparin-coated catheters and heparin-induced thrombocytopenia.
J Vasc Surg 1988; 7:667–672.
Laster J, Cikrit D, Walker N, Silver D. The heparin-induced thrombocytopenia
syndrome: an update. Surgery 1987; 102:763–770.
Lee DH, Warkentin TE, Denomme GA, Hayward CPM, Kelton JG. A diagnostic
test for heparin-induced thrombocytopenia: detection of platelet micropar-
ticles using flow cytometry. Br J Haematol 1996; 95:724–731.
Lepercq J, Conard J, Borel-Derlon A, Darmon JY, Boudignat O, Francoual C,
Priollet P, Cohen C, Yvelin N, Schved JF, Tournaire M, Borg JY. Venous
thromboembolism during pregnancy: a retrospective study of enoxaparin

safety in 624 pregnancies. BJOG 2001; 108:1134–1140.
Leyvraz PF, Bachmann F, Hoek J, Büller HR, Postel M, Samama M, Vandenbroek
MD. Prevention of deep vein thrombosis after hip replacement: randomised
comparison between unfractionated heparin and low molecular weight hepa-
rin. Br Med J 1991; 303:543–548.
Lindhoff-Last E, Eichler P, Stein M, Plagemann J, Gerdsen F, Wagner R, Ehrly
AM, Bauersachs R. A prospective study on the incidence and clinical relevance
of heparin-induced antibodies in patients after vascular surgery. Thromb Res
2000; 97:387–393.
Lindhoff-Last E, Nakov R, Misselwitz F, Breddin HK, Bauersachs R. Incidence and
clinical relevance of heparin-induced antibodies in patients with deep vein
thrombosis treated with unfractionated or low-molecular-weight heparin. Br J
Haematol 2002; 118:1137–1142.
Ling E, Warkentin TE. Intraoperative heparin flushes and subsequent acute heparin-
induced thrombocytopenia. Anesthesiology 1998; 89:1567–1569.
Look KA, Sahud M, Flaherty S, Zehnder JL. Heparin-induced platelet aggrega-
tion vs platelet factor 4 enzyme-linked immunosorbent assay in the diagnosis
of heparin-induced thrombocytopenia-thrombosis. Am J Clin Pathol 1997;
108:78–82.
Louridas G. Heparin-induced thrombocytopenia. S Afr J Surg 1991; 29:50–52.
Luzzatto G, Bertoli M, Cella G, Fabris F, Zaia B, Girolami A. Platelet count, anti-
heparin/platelet factor 4 antibodies and tissue factor pathway inhibitor plasma
antigen level in chronic dialysis. Thromb Res 1998; 89:115–122.
Malcolm ID, Wigmore TA, Steinbrecher UP. Heparin-associated thrombocytope-
nia: low frequency in 104 patients treated with heparin of intestinal mucosal
origin. Can Med Assoc J 1979; 120:1086–1088.
Masucci IP, Calis KA, Bartlett DL, Alexander HR, Horne MK III. Thrombocy-
topenia after isolated limb or hepatic perfusions with melphalan: the risk of
heparin-induced thrombocytopenia. Ann Surg Oncol 1999; 6:476–480.
Mattioli AV, Bonetti L, Sternieri S, Mattioli G. Heparin-induced thrombocytopenia
in patients treated with unfractionated heparin: prevalence of thrombosis in a
1 year follow-up. Ital Heart J 2000; 1:39–42.
Mayo DJ, Cullinane AM, Merryman PK, Horne MK III. Serologic evidence of
heparin sensitization in cancer patients receiving heparin flushes of venous
access devices. Support Care Cancer 1999; 7:425–427.
Monreal M, Lafoz E, Salvador R, Roncales J, Navarro A. Adverse effects of three
different forms of heparin therapy: thrombocytopenia, increased transami-
nases, and hyperkalemia. Eur J Clin Pharmacol 1989; 37:415–418.
Nand S, Wong W, Yuen B, Yetter A, Schmulbach E, Gross Fisher S. Heparin-
induced thrombocytopenia with thrombosis: incidence, analysis of risk factors,
and clinical outcomes in 108 consecutive patients treated at a single institution.
Am J Hematol 1997; 56:12–16.
Nelson JC, Lerner RG, Goldstein R, Cagin NA. Heparin-induced thrombocytope-
nia. Arch Intern Med 1978; 138:548–552.
Ng HJ, Lee LH. Heparin-induced thrombocytopenia: acknowledging its presence in

low-molecular-weight heparin therapy. Int J Hematol 2003; 77:185–187.
Parney IF, Steinke DE. Heparin-induced thrombocytopenia and thrombosis fol-
lowing sub-arachnoid hemorrhage. Case report. J Neurosurg 2000; 93:136–
139.
Plath J, Schulze R, Barz D, Krammer B, Steiner M, Anders O, Mach J. Necrotizing
skin lesions induced by low-molecular-weight heparin after total knee arthro-
plasty. Arch Orthop Trauma Surg 1997; 116:443–445.
Pouplard C, May MA, Iochmann S, Amiral J, Vissac AM, Marchand M, Gruel Y.
Antibodies to platelet factor 4-heparin after cardiopulmonary bypass in pa-
tients anticoagulated with unfractionated heparin or a low-molecular-weight
heparin: clinical implications for heparin-induced thrombocytopenia. Circu-
lation 1999; 99:2530–2536.
Pouplard C, May MA, Regina S, Maakaroun A, Fusciardi J, Gruel Y. Changes in
the platelet count after cardiopulmonary bypass can efficiently predict the
development of pathogenic heparin-dependent antibodies [abstr]. Blood 2002;
100:16a–17a.
Powers PJ, Cuthbert D, Hirsh J. Thrombocytopenia found uncommonly during
heparin therapy. JAMA 1979; 241:2396–2397.
Powers PJ, Kelton JG, Carter CJ. Studies on the frequency of heparin-associated
thrombocytopenia. Thromb Res 1984; 33:439–443.
Rao AK, White GC, Sherman L, Colman R, Lan G, Ball AP. Low incidence of
thrombocytopenia with porcine mucosal heparin. A prospective multicentre
study. Arch Intern Med 1989; 149:1285–1288.
Rama BN, Haake RE, Bander SJ, Ghasem-Zadeh A, Gorla C. Heparin-flush
associated thrombocytopenia-induced hemorrhage: a case report. Nebr Med J
1991; 76:392–394.
Ramirez-Lassepas M, Cipolle RJ, Rodvold KA, Seifert RD, Strand L, Taddeini L,
Cusulos M. Heparin-induced thrombocytopenia in patients with cerebrovas-
cular ischemic disease. Neurology 1984; 34:736–740.
Randolph AG, Cook DJ, Gonzales CA, Andrew M. Benefit of heparin in peripheral
venous and arterial catheters: systematic review and meta-analysis of ran-
domised controlled trials. BMJ 1998a; 316:969–975.
Randolph AG, Cook DJ, Gonzales CA, Andrew M. Benefit of heparin in central
venous and pulmonary artery catheters. A meta-analysis of randomized con-
trolled trials. Chest 1998b; 113:165–171.
Rice L, Jackson D. Can heparin cause clotting? Heart Lung 1981; 10:331–335.
Rice L, Kennedy D, Veach A. Pentosan induced cerebral sagittal sinus thrombosis: a
variant of heparin induced thrombocytopenia. J Urol 1998; 160:2148.
Romeril KR, Hickton CM, Hamer JW, Heaton DC. Heparin-induced thrombocy-
topenia: case reports and a prospective study. NZ Med J 1982; 95:267–269.
Saffle JR, Russon J Jr, Dukes GE, Warden GD. The effect of low-dose heparin
therapy on serum platelet and transaminase levels. J Surg Res 1980; 28:297–
305.
Sanson BJ, Lensing AWA, Prins MH, Ginsberg JS, Barkagan ZS, Lavenne-
Pardonge E, Brenner B, Dulitzky M, Nielsen JD, Boda Z, Turi S, MacGillavry

MR, Hamulyák K, Theunissen IM, Hunt BJ, Büller HR. Safety of low-
molecular-weight heparin in pregnancy: a systematic review. Thromb Haemost
1999; 81:668–672.
Schwartz KA, Royer G, Kaufman DB, Penner JA. Complications of heparin
administration in normal individuals. Am J Hematol 1985; 19:355–363.
Serruys PW, Emanuelsson H, van der Giessen W, Lunn AC, Kiemeney F, Macaya
C, Rutsch W, Heyndrickx G, Suryapranata H, Legrand V, Goy JJ, Materne P,
Bonnier H, Morice M-C, Fajadet J, Belardi J, Colombo A, Garcia E, Ruygrok
P, de Jaegere P, Morel M-A on behalf of the Benestent-II Study Group.
Heparin-coated Palmaz-Schatz stents in human coronary arteries. Early out-
come of the Benestent-II pilot study. Circulation 1996; 93:412–422.
Shumate MJ. Heparin-induced thrombocytopenia [letter]. N Engl J Med 1995;
333:1006–1007.
Silver D, Kapsch DN, Tsoi EK. Heparin-induced thrombocytopenia, thrombosis,
and hemorrhage. Ann Surg 1983; 198:301–306.
Simonneau G, Sors H, Charbonnier B, Page Y, Laaban JP, Azarian R, Lauent M,
Hirsch JL, Ferrari E, Bosson JL, Mottier D, Beau B. A comparison of low-
molecular-weight heparin with unfractionated heparin for acute pulmonary
embolism. N Engl J Med 1997; 337:663–669.
Singer RL, Mannion JD, Bauer TL, Armenti FR, Edie RN. Complications from
heparin-induced thrombocytopenia in patients undergoing cardiopulmonary
bypass. Chest 1993; 104:1436–1440.
Stead RB, Schafer AI, Rosenberg RD, Handin RI, Josa M, Khuri SF. Heterogeneity
of heparin lots associated with thrombocytopenia and thromboembolism. Am
J Med 1984; 77:185–188.
Suh JS, Aster RH, Visentin GP. Antibodies from patients with heparin-induced
thrombocytopenia/thrombosis recognize different epitopes on heparin: platelet
factor 4. Blood 1998; 91:916–922.
Tardy-Poncet B, Tardy B, Grelac F, Reynaud J, Mismetti P, Bertrand JC, Guyotat
D. Pentosan polysulfate-induced thrombocytopenia and thrombosis. Am J
Hematol 1994; 45:252–257.
Tardy B, Page Y, Tardy-Poncet B, Comtet C, Zeni F, Bertrand JC. Thrombopénie
induite par une héparine de bas poids moléculaire [letter]. Therapie 1990;
45:453.
Tardy B, Tardy-Poncet B, Fournel P, Venet C, Jospe R, Dacosta A. Lower limb
veins should be systematically explored in patients with isolated heparin-
Trossaert M, Gaillard A, Commin PL, Amiral J, Vissac AM, Fressinaud E. High
incidence of anti-heparin/platelet factor 4 antibodies after cardiopulmonary
bypass. Br J Haematol 1998; 101:653–655.
Verma AK, Levine M, Shalansky SJ, Carter CJ, Kelton JG. Frequency of heparin-
induced thrombocytopenia in critical care patients. Pharmacotherapy 2003;
23:745–753.
Visentin GP, Malik M, Cyganiak KA, Aster RH. Patients treated with unfrac-
tionated heparin during open heart surgery are at high risk to form antibodies
reactive with heparin: platelet factor 4 complexes. J Lab Clin Med 1996;
128:376–383.
Vitoux JF, Roncato M, Hourdbhaigt P, Aiach M, Fiessinger J-N. Heparin-induced
thrombocytopenia and pentosan polysulfate: treatment with a low molecular
weight heparin despite in vitro platelet aggregation [letter]. Thromb Hemost
1985; 55:294–295.
Med 1999; 106:629–635.
Walls JT, Curtis JJ, Silver D, Boley TM, Schmaltz RA, Nawarawong W. Heparin-
induced thrombocytopenia in open heart surgical patients: sequelae of late
recognition. Ann Thorac Surg 1992a; 53:787–791.
Walls JT, Boley TM, Curtis JJ, Silver D. Heparin-induced thrombocytopenia in
patients undergoing intra-aortic balloon pumping after open heart surgery.
ASAIO J 1992b; 38:M574–M576.
thrombocytopenia. Semin Hematol 1998; 35(suppl 4):17–25.
Warkentin TE. Platelet count monitoring and laboratory testing for heparin-induced
thrombocytopenia: recommendations of the College of American Patholo-
gists. Arch Pathol Lab Med 2002; 126:1415–1423.
Warkentin TE. Pork or beef ? Ann Thorac Surg 2003; 75:15–16.
Warkentin TE, Greinacher A. Heparin-induced thrombocytopenia: recognition,
treatment, and prevention. Chest 2004. In press.
Warkentin TE, Kelton JG. Heparin-induced thrombocytopenia. Prog Hemost
Thromb 1991; 10:1–34.
Heparins in Prophylaxis and Therapy of Thromboembolic Diseases. New
York: Marcel Dekker, 1994:75–127.
Am J Med 1996; 101:502–507.
N Engl J Med 2001; 344:1286–1292.
Heparin-induced thrombocytopenia in patients treated with low-molecular
Warkentin TE, Chong BH, Greinacher A. Heparin-induced thrombocytopenia:
towards consensus. Thromb Haemost 1998a; 79:1–7.
Warkentin TE, Ling E, Ho A, Sheppard JI. ‘‘Incidental’’ unfractionated heparin
(UFH) vs normal saline (NS) flushes for intraoperative invasive catheters
and the frequency of formation of heparin-induced thrombocytopenia IgG
antibodies (HIT-IgG): a randomized, controlled trial [abstr]. Blood 1998b;

92(suppl 1):91b.
Warkentin TE, Sheppard JI, Horsewood P, Simpson PJ, Moore JC, Kelton JG.
Impact of the patient population on the risk of heparin-induced thrombocy-
topenia. Blood 2000; 96:1703–1708.
after elective hip or knee replacement surgery in patients receiving anti-
thrombotic prophylaxis with fondaparinux or enoxaparin [abstr]. Blood 2003b.
In press.
Weitberg AB, Spremulli E, Cummings FJ. Effect of low-dose heparin on the platelet
count. South Med J 1982; 75:190–192.
Williams RT, Damaraju LV, Mascelli MA, Barnathan ES, Califf RM, Simoons ML,
Deliargyris EN, Sane DC. Anti-platelet factor 4/heparin antibodies. An in-
dependent predictor of 30-day myocardial infarction after acute coronary
ischemic syndromes. Circulation 2003; 107:2307–2312.
Wolf H, Nowak H, Wick G. Detection of antibodies interacting with glycosami-
noglycans polysulfate in patients treated with heparin or other polysulfated
glycosaminoglycans. Int Arch Allergy Appl Immunol 1983; 70:157–163.
Yamamoto S, Koide M, Matsuo M, Suzuki S, Ohtaka M, Saika S, Matsuo T.
Heparin-induced thrombocytopenia in hemodialysis patients. Am J Kidney
Dis 1996; 28:82–85.
5
Nonimmune Heparin–Platelet
Interactions: Implications for the
Pathogenesis of Heparin-Induced
Thrombocytopenia
McDonald K. Horne III
Warren G. Magnuson Clinical Center, National Institutes of Health,
Bethesda, Maryland, U.S.A.
I. INTRODUCTION
Almost as soon as heparin was introduced into clinical medicine, the new drug
was reported to cause immediate small, but consistent, reductions in platelet
count (Sappington, 1939). Later it was also found to produce platelet dys-
function (Heiden et al., 1977), accounting for at least some of its hemorrhagic
risk (Hirsh, 1984; John et al., 1993). These effects, which most likely result
from direct contact between the sulfated glycosaminoglycans and platelets,
are distinct from the role heparin plays in immune-mediated heparin-induced
thrombocytopenia (HIT). However, direct heparin–platelet binding is critical
in the pathogenesis of HIT as well (Horne and Hutchison, 1998). Therefore,
the various ‘‘nonimmune’’ heparin–platelet interactions will be reviewed.
II. HEPARIN BINDING TO PLATELETS
Appreciation of the functional effects of heparin on platelets led to studies of

heparin binding to these cells, which is specific and saturable (Sobel and
Adelman, 1988; Horne, 1988; Horne and Chao, 1989). The negative charge
density of heparin is largely responsible for its binding specificity (Horne,
149
150 Horne
1988). Polysaccharides with various primary structures can displace heparin

from platelets if the molecules are sufficiently charged (Horne, 1988; Grei-
nacher et al., 1993) (Table 1). The identity of the platelet binding site(s), which
provides a complementary positive charge, is uncertain. One report indicates
that glycoprotein IIb/IIIa (integrin aIIbh3) contains a heparin-binding site
(Sobel et al., 2001), but this is inconsistent with other studies (Horne, 1988,
1991).
Next to negative charge density, molecular size has the greatest effect on
polysaccharide binding to platelets. Heparin molecular weight, for example,
affects both its platelet-binding affinity and capacity (Horne and Chao, 1990).
As medicinal heparin is polydisperse (i.e., comprises a mixture of molecules
varying in mass from about 4000 to 30,000 Da), the mass of a mole of heparin
(approximately 6 1023 molecules) depends on the mean size of the molecules
in the sample. The maximum number of molecules bound per platelet is
approximately the same for heparin species with molecular weights between
about 5000 and 15,000 Da (see Table 1). However, larger molecules bring
more glycosaminoglycan mass to the platelet surface than smaller molecules
(Fig. 1). Therefore, when heparin-binding capacity is expressed in terms of
mass, rather than moles or molecules, the capacity of larger heparins is greater
than that of smaller heparins.
Similar distinctions apply to the parameters of binding affinity. Longer
heparin molecules contain more potential platelet-binding domains than
shorter molecules. Therefore, a large heparin species can half-saturate
platelets at a lower molar concentration (Kd) than a smaller heparin species,
although the concentration of platelet-binding domains in the suspension is
the same for both species at half-saturation (Horne and Chao, 1990).
Table 1 Platelet Binding Parameters for Heparin Fractions of Different Molecular Mass
Heparin Sulfate/ Dissociation constant Binding capacity

Mr range carboxylate
(Da) (mol/mol) (mg/L) (nM) (mg/1015 cells) (molecules/cell)
14,000–16,000 2.0 F 0.29 a

4.6 F 1.1 310. F 73 66. F 2.5 2600. F 100
9,500–10,500 1.8 F 0.26 3.9 F 2.1 390. F 210 56. F 8.4 3400. F 500
4,500–5,500 1.9 F 0.15 3.2 F 1.0 640. F 200 23. F 5.7 2800. F 680
2,700–3,300 1.7 F 0.25 4.0 F 2.0 1300. F 650 10. F 5.4 2000. F 1100
a
Values are means F1 standard deviation.
Source: Horne and Chao, 1990.
Nonimmune Heparin–Platelet Interactions 151
Figure 1 Schematic binding of heparin to platelets comparing heparin of Mr 5000

with heparin of Mr 15,000. Each ‘‘platelet-binding domain’’ of heparin is hypothesized
to have Mr > 3000, whereas heparin-binding sites on the platelets (indicated as
++++) can bind 7000 Da heparin. Therefore, each binding site is not quite filled
with Mr 5000 heparin, but is too occupied to allow binding of a second heparin
molecule. In contrast, Mr 15,000 heparin has adequate length to occupy two binding
sites, but physical constraints, such as limited heparin flexibility and spacial
distribution of binding sites, allow it to occupy only one site at a time. The scheme
is consistent with the binding parameters shown in Table 1.
152 Horne
Because of its high charge density, as well as the high linear flexibility
confered by its constituent L-iduronic acid residues, heparin binds readily
to a variety of basic plasma proteins, which theoretically could compete with
platelets for heparin (Casu et al., 1988; Young et al., 1994). However, heparin
binding to only two plasma proteins, antithrombin and fibronectin, interferes
with heparin-induced platelet activation (Salzman et al., 1980; Chong and
Ismail, 1989) or with binding of heparin to platelets (Horne and Chao, 1990).
III. NONIDIOSYNCRATIC HEPARIN-INDUCED PLATELET

ACTIVATION
The functional consequence of heparin binding to platelets is subtle cell

stimulation. Antibody-independent activation of platelets by heparin in vitro
has been reported from many laboratories. However, the results of these
studies have varied, presumably because of differences in experimental
conditions. In plasma, for example, heparin alone causes slight platelet
aggregation, whereas platelets suspended in laboratory buffers are reported
to aggregate either briskly or not at all in response to heparin (Eika, 1972;
Salzman et al., 1980; Westwick et al., 1986; Chong and Ismail, 1989). In
citrate-anticoagulated plasma, heparin also potentiates platelet activation by
agonists, such as ADP and collagen (Holmer et al., 1980; Chen and Sylvén,
1992; Xiao and Théroux, 1998; Aggarwal et al., 2002; Klein et al., 2002), and
this effect is more pronounced in patients with acute illness, arterial disease,
and anorexia nervosa (Mikhailidis et al., 1985; Reininger et al., 1996; Burgess
and Chong, 1997).
The platelet proaggregatory effect of heparin does not appear to be an
artifact of low ionized calcium concentration in the presence of citrate
anticoagulant: Chen and colleagues (1992) observed that heparin enhanced
collagen-induced platelet aggregation in a dose-dependent fashion even in
whole blood anticoagulated with hirudin (i.e., physiological calcium concen-
trations). On the other hand, the responsiveness of washed platelets to
agonists, when resuspended in buffers containing physiological calcium, has
been reported to be both increased and decreased by heparin (Saba et al.,
1984; Westwick et al., 1986). Although the data are not always consistent, this
much seems clear: direct heparin-induced platelet aggregation requires met-
abolic energy and is mediated by fibrinogen; therefore, it depends on platelet
fibrinogen receptors (platelet glycoprotein IIb/IIIa complexes) and divalent
cations (Chong and Ismail, 1989). There is also evidence that heparin can
antagonize platelet inhibition by prostacyclin (Saba et al., 1979; Eldor and
Weksler, 1979; Fortini et al., 1985; Berglund and Wallentin, 1991).
The properties of heparin that influence its platelet binding also

influence its stimulatory effects on platelets: heparin of a high molecular
weight is more active than low molecular weight heparin, and heparin with
low affinity for antithrombin and fibronectin is more active (because it is more
available) than heparin with high affinity for these plasma proteins (Salzman
et al., 1980; Holmer et al., 1980; Westwick et al., 1986; Chong and Ismail,
1989; Brace and Fareed, 1990; Xiao and Théroux, 1998; Aggarwal et al., 2002;
Klein et al., 2002). The latter observation implies that the anticoagulant
(antithrombin-dependent) activities of heparin are distinct from its platelet
stimulatory effects. Furthermore, nonheparin polysaccharides can mimic the
proaggregatory effects of heparin on platelets if they are sufficiently large and
charged (Tiffany and Penner, 1981). In contrast, heparan sulfate (the pre-
dominant anticoagulant glycosaminoglycan found in danaparoid) has negli-
gible platelet-activating properties, as it has a relatively low degree of
sulfation, despite sharing a carbohydrate backbone similar to heparin (Lin-
dahl and Kjellén, 1991; Burgess and Chong, 1997).
IV. PLATELET-RELATED PROHEMORRHAGIC EFFECTS

OF HEPARIN
Paradoxically, despite the in vitro evidence that heparin stimulates platelets,

there is evidence that heparin causes bleeding partly because of its effects on
platelet function (Hirsh, 1984; John et al., 1993). Heparin, for example, causes
prolongation of the skin-bleeding time unrelated to any effects on platelet
counts (Hjort et al., 1960; Heiden et al., 1977; Kelton, 1986; Ljungberg et al.,
1988). Also, the structural characteristics of heparin associated with platelet
stimulation in vitro (e.g., increased heparin size and sulfation; decreased
affinity for antithrombin) are associated with enhanced bleeding in animal
models (Hjort et al., 1960; Carter et al., 1982; Ockelford et al., 1982; Fer-
nandez et al., 1986; Borowska et al., 1988; Van Ryn-McKenna et al., 1989).
The apparent inhibition of platelet function in vivo may be related to
two specific actions of heparin: inhibition of thrombin-induced platelet
activation and reduction of von Willebrand factor (vWF)-dependent platelet
function. Thrombin is a ‘‘strong’’ platelet activator [i.e., it stimulates platelet
secretion without intermediate platelet aggregation (Ware and Coller, 1995)].
However, in the presence of antithrombin, heparin essentially eliminates
stimulation of platelets by thrombin (Westwick et al., 1986; Cofrancesco et
al., 1988). This effect is likely responsible for the marked prolongation of
bleeding time seen in patients receiving high doses of heparin during heart
surgery (Kestin et al., 1993). Heparin also binds to vWF, preventing binding
of vWF to platelets (Sobel et al., 1991, 1992). This reduces vWF-mediated
154 Horne
subendothelial adhesion of platelets flowing at high shear rates, perhaps also

contributing to heparin-related prolongation of the bleeding time.
V. NONIMMUNE HEPARIN-ASSOCIATED
THROMBOCYTOPENIA (NONIMMUNE HAT)
Nonimmune HAT describes the common clinical situation in which a patient

develops a fall in the platelet count within the first few days of receiving
heparin. Often, there are concomitant clinical factors to explain the throm-
bocytopenia [e.g., hemodilution, bacteremia, or disseminated intravascular
coagulation (DIC)]. In some patients, however, it is possible that a direct
proaggregatory effect of the heparin is responsible for the platelet count fall
(Salzman et al., 1980). The designation associated helps to convey the un-
certain role of heparin in causing thrombocytopenia in an individual patient,
and the term nonimmune distinguishes this syndrome from (immune-mediat-
ed) HIT (Warkentin et al., 1998).
Nonimmune HAT is typically mild, transient, and clinically inconse-
quential (Gollub and Ulin, 1962; Johnson et al., 1984; Chong, 1988; War-
kentin and Kelton, 1994). There is debate whether this represents a real in vivo
phenomenon, or whether the apparent thrombocytopenia is instead related to
ex vivo platelet aggregation (Davey and Lander, 1968). Indeed, some inves-
tigators were unable to show this event at all (Heinrich et al., 1988; Xiao and
Théroux, 1998). Sometimes, however, apparent nonimmune HAT is a dra-
matic clinical syndrome that can be confused with HIT (Chong et al., 1982)
(see Chap. 12).
Balduini et al. (1993) observed that an early fall in platelet count was
more frequent and of greater magnitude in patients receiving heparin
following streptokinase therapy for acute myocardial infarction, compared
with control patients who received streptokinase alone. The heparin-treated
patients also showed greater ex vivo spontaneous platelet aggregation,
suggesting that heparin may have had a direct proaggregatory effect.
VI. HEPARIN–PLATELET INTERACTIONS IN THE

PATHOGENESIS OF HIT
Heparin also binds to several proteins that are secreted from stimulated
platelets. One of these, platelet factor 4 (PF4), binds heparin with unusually
high affinity (Capitanio et al., 1985; Loscalzo et al., 1985; Horne, 1993).
Unlike antithrombin or fibronectin, however, PF4 does not prevent heparin
from binding to platelets. In fact, it appears that heparin can bind to platelets
in complexes with PF4 (Horne and Hutchison, 1998).
Heparin-bound PF4 has been identified as the primary target of the
antibodies characteristic of HIT (Amiral et al., 1992; Visentin et al., 1994;
Kelton et al., 1994). Because platelets have specific-binding sites for PF4, this
protein was originally assumed to mediate the attachment of HIT immune
complexes to the platelet surface (Kelton et al., 1994). However, it was
subsequently demonstrated that heparin, rather than PF4, serves this pur-
pose; that is, the binding to platelets of heparin-PF4 complexes occurs at
heparin-binding sites, rather than PF4-binding sites (Greinacher et al., 1993;
Horne and Hutchison, 1998).
The sequence of interactions leading to platelet activation has recently
been described (Newman and Chong, 2000). It begins with the nonimmune
binding of heparin to the cells, which stimulates them to secrete a relatively
small amount of PF4. The released PF4 binds to heparin in solution to reveal
the target for HIT IgG antibodies (HIT-IgG). Immune complexes of heparin,
PF4, and HIT-IgG then attach to the platelets at heparin-binding sites. This
facilitates contact between the Fc termini of the IgG molecules and the
platelet Fc receptors, thereby activating the cells further and causing more
PF4 secretion. As this sequence accelerates, platelet aggregation occurs.
Because immune complexes comprised of heparin, PF4, and IgG from
HIT patients bind to platelets at heparin-binding sites, they must compete
with free heparin for binding to the platelet surface (Horne and Alkins, 1996;
Horne and Hutchison, 1998). When free heparin is in molar excess, heparin–
PF4–IgG complexes are displaced from the platelet surface (Fig. 2). When
PF4 is in excess, there is no free heparin, and binding of the complexes is the
only option.
Variable heparin–PF4 stoichiometry may explain why some patients
develop HIT-IgG without developing thrombocytopenia, and why HIT is
more common in certain clinical settings than in others, such as following
surgery (Boshkov et al., 1993; Amiral et al., 1995, 1996; Warkentin et al.,
1995, 2000; Visentin et al., 1996; Kappers-Klunne et al., 1997; Bauer et al.,
1997). When a patient is given heparin, the plasma concentration of PF4 rises
because PF4 is displaced from the endothelial surface, where it is normally
bound to heparan sulfate (Dawes et al., 1982; Rao et al., 1983; O’Brien et al.,
1985). PF4 neoantigens (or cryptic antigens) are formed, leading to the HIT
immune response (Chong and Newman, 1997). However, complexes of
heparin, PF4, and IgG are harmless unless they become bound to platelets,
and this cannot happen as long as there is sufficient free heparin to compete
effectively for the limited number of platelet-binding sites (Horne and
Hutchison, 1998). Therefore, as long as heparin remains in molar excess over
156 Horne
Figure 2 Drawing illustrating the importance of the molar ratio of heparin to PF4 in
determining binding of HIT immune complexes to platelets. Maximal binding of HIT-
IgG occurs when heparin and PF4 are present in equimolar concentrations. Note that
IgG Fc interactions with platelet Fc receptors are not shown in this figure. (From
Horne and Hutchison, 1998.)
PF4, heparin–PF4–IgG binding to platelets is minimized, and platelet

activation and thrombocytopenia do not develop.
The importance of heparin–PF4 stoichiometry in HIT was demonstrat-
ed in a serotonin-release assay using platelets from a patient with the gray
platelet syndrome, which lack PF4 (Horne and Alkins, 1995). Platelet release
was stimulated only as PF4 was added and approached or exceeded molar
excess. In the absence of PF4 or with 5 nM PF4 (heparin:PF4 f10:1), 0.1 unit/
mL heparin (f50 nM) and HIT antibody caused 4–5% release. However, 25
nM PF4 (heparin:PF4 f2:1) caused 20% release, 50 nM PF4 (heparin:PF4 f
1:1) caused 34% release, and 100 nM PF4 (heparin:PF4 f1:2) caused 60%
release.
In most clinical settings, free heparin is in considerable molar excess
over PF4. For example, therapeutic concentrations of heparin (0.2–0.4 U/
mL) correspond to about 100–200 nmol/L of heparin. When heparin is given
to normal individuals, plasma concentrations of PF4 from endothelial

reservoirs reach only about 8 nM (Dawes et al., 1982). For PF4 concen-
trations to approach 100–200 nmol/L, marked activation of circulating
platelets is necessary. Complete activation of platelets in a concentration of
250 109/L will generate a plasma PF4 concentration of about 200 nM
(Horne, 1993). Therefore, a molar excess of PF4 over therapeutic concen-
trations of heparin would be highly unlikely outside extreme clinical circum-
stances. On the other hand, prophylactic doses of heparin (e.g., 5000 U every
8–12 h by subcutaneous injection) administered in a setting associated with
some platelet activation (e.g., postoperative orthopedic patients) might well
produce molar ratios of heparin and PF4 that would favor platelet binding of
heparin–PF4 complexes, and—if an immune response had occurred—platelet
binding of heparin–PF4–IgG complexes. Indeed, such scenarios are the ones
in which HIT is reported most frequently (Warkentin et al., 1995, 2000;
Ganzer et al., 1997).
VII. IMPLICATIONS OF NONIMMUNE HEPARIN BINDING TO

PLATELETS FOR THE PREVENTION OR TREATMENT
OF HIT
Heparin’s molecular size influences its platelet-binding affinity and capacity

and its stimulating effect on platelets. Similarly, heparin’s affinity and
capacity for PF4 are related to its size: large heparin molecules have higher
affinity for PF4 and can bind several PF4 molecules, whereas smaller heparin
molecules have less affinity for PF4 and can bind fewer PF4 molecules. (Bock
et al., 1980; Lane et al., 1984; Marshall et al., 1984). Therefore, it should be no
surprise that low molecular weight heparin is associated with a lower
frequency of HIT than standard heparin (Warkentin et al., 1995; Lindhoff-
Last et al., 2002) and that in some instances low molecular weight heparin
has been given to patients with HIT without adverse consequences (Slocum
et al., 1996). Smaller heparins complex less efficiently with PF4, thereby
reducing immunogenicity, and bind less avidly to platelets, thereby reducing
immune complex binding to the cells. Indeed, the very smallest heparin, the
antithrombin-binding pentasaccharide (Mr 1714), theoretically is an ideal
anticoagulant agent for patients with HIT (see Chap. 8). The pentasaccharide
binds to neither platelets nor PF4 and therefore cannot support platelet
activation by HIT antibodies (Elalamy et al., 1995; Walenga et al., 1997;
Ahmad et al., 1999).
Similarly, the safety and efficacy of treating HIT patients with danapa-
roid, a so-called heparinoid, can be explained by the fact that its major
component (approximately 84% heparin sulfate) does not bind to platelets
158 Horne
(Horne, 1988; Magnani, 1993). On the other hand, danaparoid sometimes

cross-reacts with HIT antibodies in laboratory tests for HIT. This is perhaps
mediated by a minor component of danaparoid (about 12% dermatan
sulfate), which does have weak affinity for both platelets and PF4 (Barber
et al., 1972; Horne, 1988).
REFERENCES
Aggarwal A, Sobel BE, Schneider DJ. Decreased platelet reactivity in blood anti-
coagulated with bivalirudin or enoxaparin compared with unfractionated hepa-
rin: implications for coronary intervention. J Thromb Thrombolysis 2002; 13:
161–165.
Ahmad S, Jeske WP, Walenga JM, Hoppensteadt DA, Wood JJ, Herbert JM, Mess-
more HL, Fareed J. Synthetic pentasaccharides do not cause platelet activation
by anti-heparin-platelet factor 4 antibodies. Clin Appl Thromb Hemost 1999;
5:259–266.
Amiral J, Bridey F, Wolf M, Boyer-Neumann C, Fressinaud E, Vissac AM, Peynaud-
Debayle E, Dreyfus M, Meyer D. Antibodies to macromolecular platelet factor
4–heparin complexes in heparin-induced thrombocytopenia: a study of 44 cases.
Amiral J, Peynaud-Debayle E, Wolf M, Bridey F, Vissac AM, Meyer D. Generation of
antibodies to heparin–PF4 complexes without thrombocytopenia in patients
1996; 52:90–95.
Balduini CL, Noris P, Bertolino G, Previtali M. Heparin modifies platelet count and
function in patients who have undergone thrombolytic therapy for acute
myocardial infarction [letter]. Thromb Haemost 1993; 69:522–523.
Barber AF, Kaser-Glanzmann R, Jakabova M, Luscher EF. Characterization of a
chondroitin 4-sulfate proteoglycan carrier for heparin neutralizing activity
(platelet factor 4) released from human blood platelets. Biochim Biophys Acta
1972; 286:312–329.
McNulty S, Amiral J, Hauck WW, Edie RN, Mannion JD. Prevalence of
Berglund U, Wallentin L. Influence on platelet function by heparin in men with
unstable coronary artery disease. Thromb Haemost 1991; 66:648–651.
Bock PE, Luscombe M, Marshall SE, Pepper DS, Holbrook JJ. The multiple
complexes formed by the interaction of platelet factor 4 with heparin. Biochem J
1980; 191:769–776.
Borowska A, Lauri D, Maggi A, Dejana E, De Gaetano G, Donati MB, Pangrazzi J.

Impairment of primary haemostasis by low molecular weight heparins in rats.
Br J Haematol 1988; 68:339–344.
induced thrombocytopenia and thrombosis: clinical and laboratory studies.
Br J Haematol 1993; 84:322–328.
Brace LD, Fareed J. Heparin-induced platelet aggregation. II. Dose/response
relationships for two low molecular weight heparin fractions (CY 216 and CY
222). Thromb Res 1990; 59:1–14.
Burgess JK, Chong BH. The platelet proaggregating and potentiating effects of un-
fractionated heparin, low molecular weight heparin and heparinoid in intensive
care patients and healthy controls. Eur J Haematol 1997; 58:279–285.
Capitanio AM, Niewiarowski S, Rucinski B, Tuszynski GP, Cierniewski CS, Her-
shock D, Kornecki E. Interaction of platelet factor 4 with human platelets.
Biochim Biophys Acta 1985; 839:161–173.
Carter CJ, Kelton JG, Hirsh J, Cerskus A, Santos AV, Gent M. The relationship
between the hemorrhagic and antithrombotic properties of low molecular
weight heparin in rabbits. Blood 1982; 59:1239–1245.
Casu B, Petitou M, Provasoli M, Sinaÿ P. Conformational flexibility: a new concept
for explaining binding and biological properties of iduronic acid-containing
glycosaminoglycans. Trends Biochem Sci 1988; 13:221–225.
Chen J, Sylvén C. Heparin potentiation of collagen-induced platelet aggregation is
related to the GPIIb/GPIIIa receptor and not to the GPIb receptor, as tested by
whole blood aggregometry. Thromb Res 1992; 66:111–120.
Chen J, Karlberg KE, Sylvén C. Heparin enhances platelet aggregation irrespective of
anticoagulation with citrate or with hirudin. Thromb Res 1992; 67:253–262.
Chong BH. Heparin-induced thrombocytopenia. Blood Rev 1988; 2:108–114.
Chong BH, Ismail F. The mechanism of heparin-induced platelet aggregation. Eur J
Haematol 1989; 43:245–251.
Chong BH, Newman PM. The antigenic epitope in heparin-induced thrombocytope-
nia resides solely on platelet factor 4 [abstr]. Blood 1997; 90(suppl 1):461a.
Chong BH, Pitney WR, Castaldi PA. Heparin-induced thrombocytopenia: association
of thrombotic complications with heparin-dependent IgG antibody that induces
thromboxane synthesis and platelet aggregation. Lancet 1982; 2:1246–1249.
Cofrancesco E, Colombi M, Manfreda M, Pogliani EM. Effect of heparin and related
glycosaminoglycans (GAGs) on thrombin-induced platelet aggregation and re-
lease. Haematologica 1988; 73:471–475.
Davey MG, Lander H. Effect of injected heparin on platelet levels in man. J Clin
Pathol 1968; 21:55–59.
Dawes J, Pumphrey CW, McLaren KM, Prowse CV, Pepper DS. The in vivo release of
human platelet factor 4 by heparin. Thromb Res 1982; 27:65–76.
Eika C. The platelet aggregating effect of eight commercial heparins. Scand J
Haematol 1972; 9:480–482.
Elalamy I, Lecrubier C, Potevin F, Abdelouahed M, Bara L, Marie JP, Samama MM.
Absence of in vitro cross-reaction of pentasaccharide with the plasma heparin-
160 Horne
dependent factor of twenty-five patients with heparin-associated thrombocyto-

penia [letter]. Thromb Haemost 1995; 74:1384–1385.
Eldor A, Weksler BB. Heparin and dextran sulfate antagonize PGI2 inhibition of
platelet aggregation. Thromb Res 1979; 16:617–628.
Fernandez F, N’guyen P, Van Ryn J, Ofosu FA, Hirsh J, Buchanan MR. Hemorrhagic
doses of heparin and other glycosaminoglycans induce a platelet defect. Thromb
Res 1986; 43:491–495.
Fortini A, Modesti PA, Abbate R, Gensini GF, Neri Serneri GG. Heparin does not
interfere with prostacyclin and prostaglandin D2 binding to platelets. Thromb
Res 1985; 40:319–328.
laxe als Auslöser thrombembolischer Komplikationen. Eine Untersuchung zur
Inzidenz der Heparin-induzierten Thrombozytopenie (HIT) Typ II. Z Orthop
1997; 135:543–549.
Gollub S, Ulin AW. Heparin-induced thrombocytopenia in man. J Lab Clin Med
1962; 59:430–435.
Greinacher A. Antigen generation in heparin-associated thrombocytopenia: the non-
immunologic type and the immunologic type are closely linked in their patho-
genesis. Semin Thromb Hemost 1995; 21:106–116.
Greinacher A, Michels I, Liebenhoff U, Presek P, Mueller-Eckhardt C. Heparin-
associated thrombocytopenia: immune complexes are attached to the platelet
membrane by the negative charge of highly sulphated oligosaccharides. Br J
Haematol 1993; 84:711–716.
Heiden D, Mielke CH Jr, Rodvien R. Impairment by heparin of primary haemostasis
and platelet [14C]-5-hydroxytryptamine release. Br J Haematol 1977; 36:427–
436.
Heinrich D, Görg T, Schulz M. Effects of unfractionated and fractionated heparin on
platelet function. Haemostasis 1988; 18(suppl 3):48–54.
Hirsh J. Heparin-induced bleeding. Nouv Rev Fr Hematol 1984; 26:261–266.
Hjort PF, Borchgrevink CF, Iversen OH, Stormorken H. The effect of heparin on the
bleeding time. Thromb Diath Haemorrh 1960; 4:389–399.
Holmer E, Lindahl U, Bäckström G, Thunberg L, Sandberg H, Söderström G, An-
dersson L-O. Anticoagulant activities and effects on platelets of a heparin frag-
ment with high affinity for antithrombin. Thromb Res 1980; 18:861–869.
Horne MK III. Heparin binding to normal and abnormal platelets. Thromb Res 1988;
51:135–144.
Horne MK III. Heparin binds normally to platelets digested with Streptomyces griseus
protease. Thromb Res 1991; 61:155–158.
Horne MK III. The effect of secreted heparin-binding proteins on heparin binding to
platelets. Thromb Res 1993; 70:91–98.
Horne MK III, Alkins BR. Importance of PF4 in heparin-induced thrombocytopenia:
confirmation with gray platelets [letter]. Blood 1995; 85:1408–1409.
Horne MK III, Alkins BR. Platelet binding of IgG from patients with heparin-induced
thrombocytopenia. J Lab Clin Med 1996; 127:435–442.
Horne MK III, Chao ES. Heparin binding to resting and activated platelets. Blood
1989; 74:238–243.
Horne MK III, Chao ES. The effect of molecular weight on heparin binding to
platelets. Br J Haematol 1990; 74:306–312.
Horne MK III, Hutchison KJ. Simultaneous binding of heparin and platelet factor-4
to platelets: further insights into the mechanism of heparin-induced thrombo-
cytopenia. Am J Hematol 1998; 58:24–30.
John LCH, Rees GM, Kovacs IB. Inhibition of platelet function by heparin. An
etiologic factor in postbypass hemorrhage. Thorac Cardiovasc Surg 1993; 105:
816–822.
Johnson RA, Lazarus KH, Henry DH. Heparin-induced thrombocytopenia: a pro-
spective study. Am J Hematol 1984; 17:349–353.
Kappers-Klunne MC, Boon DMS, Hop WCJ, Michiels JJ, Stibbe J, van der Zwaan C,
Koudstaal PJ, van Vliet HHDM. Heparin-induced thrombocytopenia and
thrombosis: a prospective analysis of the incidence in patients with heart and
cerebrovascular diseases. Br J Haematol 1997; 96:442–446.
Kelton JG. Heparin-induced thrombocytopenia. Haemostasis 1986; 16:173–186.
Kelton JG, Smith JW, Warkentin TE, Hayward CPM, Denomme GA, Horsewood
P. Immunoglobin G from patients with heparin-induced thrombocytopenia
binds to a complex of heparin and platelet factor 4. Blood 1994; 83:3232–
3239.
Kestin AS, Valeri CR, Khuri SF, Loscalzo J, Ellis PA, MacGregor H, Birjiniuk V,
Ouimet H, Pasche B, Nelson MJ, Benoit SE, Rodino LJ, Barnard MR, Mi-
chelson AD. The platelet function defect of cardiopulmonary bypass. Blood
1993; 82:107–117.
Klein B, Faridi A, von Tempelhoff GF, Heilmann L, Mittermayer C, Rath W. A whole
blood flow cytometric determination of platelet activation by unfractionated
and low molecular weight heparin in vitro. Thromb Res 2002; 108:291–296.
Lane DA, Denton J, Flynn AM, Thunberg L, Lindahl U. Anticoagulant activities of
heparin oligosaccharides and their neutralization by platelet factor 4. Biochem J
1984; 218:725–732.
Lindahl U, Kjellén L. Heparin or heparan sulfate—what is the difference? Thromb
Haemost 1991; 66:44–48.
Lindhoff-Last E, Nakov R, Misselwitz F, Breddin HK, Bauersachs R. Incidence and
clinical relevance of heparin-induced antibodies in patients with deep vein
thrombosis treated with unfractionated or low-molecular-weight heparin. Br J
Haematol 2002; 118:1137–1142.
Ljungberg B, Beving H, Egberg N, Johnsson H, Vesterqvist O. Immediate effects of
heparin and LMW heparin on some platelet and endothelial derived factors.
Thromb Res 1988; 51:209–217.
Loscalzo J, Melnick B, Handin RI. The interaction of platelet factor four and
glycosaminoglycans. Arch Biochem Biophys 1985; 240:446–455.
Magnani HN. Heparin-induced thrombocytopenia (HIT): an overview of 230 patients
treated with Orgaran (Org 10172). Thromb Haemost 1993; 70:554–561.
Marshall SE, Luscombe M, Pepper DS, Holbrook JJ. The interaction of platelet factor
4 with heparins of different chain length. Biochim Biophys Acta 1984; 797:34–
39.
Mikhailidis DP, Barradas MA, Jeremy JY, Gracey L, Wakeling A, Dandona P.
162 Horne
Heparin-induced platelet aggregation in anorexia nervosa and in severe pe-

ripheral vascular disease. Eur J Clin Invest 1985; 15:313–319.
Newman PM, Chong BH. Heparin-induced thrombocytopenia: new evidence for the
dynamic binding of purified anti-PF4-heparin antibodies to platelets and the
resultant platelet activation. Blood 2000; 96:182–187.
O’Brien JR, Etherington MD, Pashley MA. The heparin-mobilisable pool of platelet
factor 4: a comparison of intravenous and subcutaneous heparin and Kabi
heparin fragment 2165. Thromb Haemost 1985; 54:735–738.
Ockelford PA, Carter CJ, Cerskus A, Smith CA, Hirsh J. Comparison of the in vivo
hemorrhagic and antithrombotic effects of a low antithrombin-III affinity
heparin fraction. Thromb Res 1982; 27:679–690.
Rao AK, Niewiarowski S, James P, Holt JC, Harris M, Elfenbein B, Bastl C. Effect of
heparin on the in vivo release and clearance of human platelet factor 4. Blood
1983; 61:1208–1214.
Reininger CB, Greinacher A, Graf J, Lasser R, Steckmeier B, Schweiberer L. Platelets
of patients with peripheral arterial disease are hypersensitive to heparin.
Thromb Res 1996; 81:641–649.
Saba HI, Saba SR, Blackburn CA, Hartmann RC, Mason RG. Heparin neutralization
of PGI2: effects upon platelets. Science 1979; 205:499–501.
Saba HI, Saba SR, Morelli GA. Effect of heparin on platelet aggregation. Am J
Hematol 1984; 17:295–306.
Salzman EW, Rosenberg RD, Smith MH, Lindon JN, Favreau L. Effect of heparin
and heparin fractions on platelet aggregation. J Clin Invest 1980; 65:64–73.
Sappington SW. The use of heparin in blood transfusions. JAMA 1939; 113:22–25.
Slocum MM, Adams JG Jr, Teel R, Spadone DP, Silver D. Use of enoxaparin in
patients with heparin-induced thrombocytopenia syndrome. J Vasc Surg 1996;
23:839–843.
Sobel M, Adelman B. Characterization of platelet binding of heparins and other
glycosaminoglycans. Thromb Res 1988; 50:815–826.
Sobel M, McNeill PM, Carlson PL, Kermode JC, Adelman B, Conroy R, Marques D.
Heparin inhibition of von Willebrand factor-dependent platelet function in
vitro and in vivo. J Clin Invest 1991; 87:1787–1793.
Sobel M, Soler DF, Kermode JC, Harris RB. Localization and characterization of a
heparin binding domain peptide of human von Willebrand factor. J Biol Chem
1992; 267:8857–8862.
Sobel M, Fish WR, Toma N, Luo S, Bird K, Mori K, Kusumoto S, Blystone SD,
Suda Y. Heparin modulates integrin function in human platelets. J Vasc Surg
2001; 33:587–594.
Tiffany ML, Penner JA. Heparin and other sulfated polyanions: their interaction with
the blood platelet. Ann NY Acad Sci 1981; 370:662–667.
Van Ryn-McKenna J, Ofosu FA, Hirsh J, Buchanan MR. Antithrombotic and
bleeding effects of glycosaminoglycans with different degrees of sulphation. Br J
Haematol 1989; 71:265–269.
Visentin GP, Ford SE, Scott JP, Aster RH. Antibodies from patients with heparin-
induced thrombocytopenia/thrombosis are specific for platelet factor 4 com-
plexed with heparin or bound to endothelial cells. J Clin Invest 1994; 93:81–88.
Visentin GP, Malik M, Cyganiak KA, Aster RH. Patients treated with unfractionated
heparin during open heart surgery are at high risk to form antibodies reactive
with heparin:platelet factor 4 complexes. J Lab Clin Med 1996; 128:376–383.
Walenga JM, Jeske WP, Bara L, Samama MM, Fareed J. Biochemical and phar-
macologic rationale for the development of a synthetic heparin pentasaccharide.
Thromb Res 1997; 86:1–36.
Ware AJ, Coller BS. Platelet morphology, biochemistry, and function. In: Beutler E,
Lichtman MA, Coller BS, Kipps TJ, eds. Williams Hematology. 5th ed. New
York: McGraw-Hill, 1995:1161–1201.
Warkentin TE, Sheppard JA, Horsewood P, Simpson PJ, Moore JC, Kelton JG.
topenia. Blood 2000; 96:1703–1708.
Westwick J, Scully MF, Poll C, Kakkar VV. Comparison of the effects of low
molecular weight heparin and unfractionated heparin on activation of human
platelets in vitro. Thromb Res 1986; 42:435–447.
Xiao Z, Théroux P. Platelet activation with unfractionated heparin at therapeutic
concentrations and comparisons with a low-molecular-weight heparin and with
a direct thrombin inhibitor. Circulation 1998; 97:251–256.
Young E, Wells P, Holloway S, Weitz J, Hirsh J. Ex-vivo and in-vitro evidence that low
molecular weight heparins exhibit less binding to plasma proteins than un-
fractionated heparin. Thromb Haemost 1994; 71:300–304.
6
Heparin-Dependent Antigens in
Jean Amiral
HYPHEN BioMed, Neuville-sur-Oise, France
Dominique Meyer
INSERM U-143, Hôpital de Bicêtre, Bicêtre, France
I. INTRODUCTION
About 10 years ago, platelet factor 4 (PF4) complexed to heparin (PF4-H)

was identified as the major target antigen for heparin-dependent antibodies
involved in the pathogenesis of immune heparin-induced thrombocytopenia
(HIT) (Amiral et al., 1992, 1995; Gruel et al., 1993; Greinacher et al., 1994,
1995; Kelton et al., 1994; Visentin et al., 1994). Occasionally, other antigens
can be involved, such as interleukin-8 (IL-8) or neutrophil-activating peptide-
2 (NAP-2), two CXC chemokines of the PF4 superfamily (Amiral et al.,
1996a). There is now increasing evidence that the risk of HIT depends on the
type of heparin used, its sulfation grade, the therapy duration, and the
patient’s clinical context (Kelton, 1992; Warkentin and Kelton, 1996) (see
Chap. 4). However, many questions remain unresolved: How are antibodies
generated? Why are they observed in only a subgroup of patients receiving
heparin? How do they become pathogenic in only a few of the latter? Indeed,
antibodies to PF4–H develop surprisingly often in many heparin-treated
patients, especially in the context of platelet activation (e.g., heart surgery
using cardiopulmonary bypass) (Amiral et al., 1996b; Visentin et al., 1996).
Clinical complications of HIT are especially associated with high-titer
PF4–H antibodies of the IgG isotype, usually in patients with comorbid
disease who are receiving unfractionated heparin (UFH). The frequency of
165
166 Amiral and Meyer
HIT is lower with the use of low molecular weight heparin (LMWH)
(Warkentin et al., 1995). However, recent studies suggest that this complica-
tion can also develop in the absence of IgG isotypes (Amiral et al., 1996c). In
some patients with apparent HIT, only IgA or IgM isotypes are present,
usually in high concentrations.
In this chapter, the current understanding of PF4–H antibody genera-
tion and its contribution to the complications of HIT will be discussed.
Formation of the PF4–H antigen complexes and their binding to blood and
endothelial cells, thus targeting the immune response onto these cells (Cines
et al., 1987; Visentin et al., 1994; Visentin and Aster, 1995; Horne and Hut-
chison, 1998), will be analyzed. Finally, the possibility that HIT can be caused
in the absence of detectable antibodies to PF4–H will be discussed, including
the hypothesis that preexisting antibodies to other chemokines could become
pathogenic during heparin treatment.
II. ANTIGENICITY OF PF4 IN THE PRESENCE OF HEPARIN
Antibodies to self-antigens, including certain autologous plasma proteins,

can develop as result of immune dysfunction, resulting in chronic autoim-
mune disease. Sometimes, however, formation of complexes between an
autologous protein and a foreign substance leads to new antigens, potentially
even on the self-protein. These can be described as cryptic autoantigens, or
neoantigens. The immune stimulation resulting from such an altered self-
epitope quickly abates when the inducing foreign substance is no longer
present. Such a model appears relevant to explain some of the clinical
complications observed in HIT (see Chap. 3). In HIT, PF4 constitutes the
self-antigen, with one or more cryptic auto-epitopes formed when complexes
are formed under optimal conditions with the foreign substance, heparin
(discussed subsequently). Thus, the antibodies to PF4–H complexes effec-
tively behave as autoantibodies in the presence of heparin (Shoenfeld, 1997).
PF4 is a positively charged tetrameric glycoprotein member of the CXC
chemokine family (Brandt and Flad, 1992). The tetramer forms by sequential
noncovalent association of PF4 monomers: two dimers are formed that self-
associate into the fundamental tetrameric structure. As found within platelet
a-granules, PF4 is released into blood only after platelet activation, such as
seen with trauma, surgery, atherosclerosis (Dunlop et al., 1987), diabetes,
cardiopulmonary bypass, inflammation, cancer, infections, and so on. In
vivo, PF4 has many different biological functions, including immunoregula-
tion, inhibition of megakaryocytopoiesis and angiogenesis, and mediation of
cell response. When released from platelets, PF4 is in a complex of eight
tetramers linked to a chondroitin-containing proteoglycan dimer: the entire
Heparin-Dependent Antigens in HIT 167
complex has a molecular weight (MW) of 350 kDa. The PF4 complexes can
also bind to endothelial cell proteoglycans (heparan sulfate). Heparin, when
present, having a greater affinity for PF4, displaces PF4 from the endothelial
cell glycosaminoglycans, thereby forming PF4–H complexes that are released
into the circulation.
The interaction between heparin and PF4 has been intensively studied
(Bock et al., 1980; Cowan et al., 1986; Stuckey et al., 1992; Maccarana and
Lindahl, 1993). In the presence of a stoichiometric concentration of heparin
and PF4 (which corresponds to 27 international units [IU] of heparin per
milligram of PF4), multimolecular PF4–H complexes (Greinacher et al.,
1994; Amiral et al., 1995) are generated. With stoichiometric concentrations,
heparin wraps around the PF4 molecule, altering its structure and rendering it
antigenic. Figure 1 shows the different complexes that can be formed between
heparin and PF4, depending on the respective concentrations of both sub-
stances. Only multimolecular complexes are believed to be antigenic in
heparin-treated patients. Thus, the immunogenicity of complexes is strictly
dependent on the respective concentrations of heparin and PF4. If we
consider the usual therapeutic range for heparin (0.1–1 IU/mL), the amount
of PF4 required for the generation of multimolecular PF4–H complexes is
from 3 to 40 Ag/mL. In patients undergoing cardiopulmonary bypass who
receive higher heparin concentrations (up to 3 IU/mL), the corresponding
Figure 1 Formation of heparin and PF4 complexes at different concentrations of

heparin and PF4: In the presence of stoichiometric concentrations of both substances,
multimolecular complexes are formed. Heparin then wraps around the PF4 tetramer,
altering its structure and rendering it antigenic.
higher PF4 concentrations required for the formation of the immunogenic

PF4–H complexes may result from intense platelet activation, resulting from
exposure of blood to the extracorporeal circuit. In general, the existence of
favorable conditions allowing the formation of multimolecular PF4–H com-
plexes may depend as much on the underlying disease that is promoting
platelet activation as on the dose of heparin given (see Chap. 4).
The intensity of the heparin-dependent immune response thus depends
on the presence and, presumably, persistence of the multimolecular PF4–H
complexes. In particular, high concentrations of PF4–H complexes may be
important in triggering an immune response. However, heparin concentra-
tions vary considerably in treated patients, and the concentrations allowing
PF4–H complex formation may occur frequently. But, if low PF4 concen-
trations are present, formation of immunogenic complexes can occur only at
corresponding very low levels of heparin (e.g., 0.027 IU/mL of heparin for 100
ng/mL of PF4, which is the approximate PF4 concentration in normal
subjects receiving heparin). The chances of developing a significant immune
response in this setting would be low. The potentially important role of
individual responsiveness to a given PF4 antigenic stimulus is unknown.
Although antibodies to PF4–H complexes are present in most patients
who develop HIT, they are absent in some patients with apparent HIT,
including patients with positive activation assays for HIT antibodies. Anti-
bodies to IL-8 or to NAP-2 have been observed in some of these patients
(Amiral et al., 1996a; Regnault et al., 2003), but in others, no specific heparin-
dependent antibodies have been identified. As discussed later, antibodies to
IL-8 or to NAP-2 are generated by mechanisms different from those involved
in PF4–H antibody formation, and may be true autoantibodies (Bendtzen et
al., 1995).
III. PATHOGENICITY OF HEPARIN-DEPENDENT

ANTIBODIES
Anti-PF4–H antibodies of the IgG isotype are present in at least 80% of

patients with clinical HIT. In the remaining cases, only IgA, IgM, or both,
isotypes, in high concentrations, are observed. These intriguing observations
require explanation for how these antibodies trigger thrombocytopenia, with
or without thrombosis.
Antibodies to PF4–H are believed to become pathogenic when they
interact with platelets or other blood and endothelial cells. This can occur
only if the PF4–H complexes bind to the cell surfaces, predominantly through
their heparin-binding sites (Van Rijn et al., 1987; Horne and Alkins, 1996;
Horne and Hutchison, 1998), but possibly also through PF4-binding sites
(Capitanio et al., 1985; Rybak et al., 1989). Although the HIT antibodies
recognize PF4–H complexes in the fluid phase (Newman et al., 1998), it is
uncertain whether this typically occurs in vivo before interaction of PF4–H–
IgG complexes with the platelet surface, or whether HIT antibodies only bind
after PF4–H complexes are attached to the platelet surface.
Regardless, the clinical state of patients—determining the extent of
platelet and endothelial cell activation—seems to be a key factor for deter-
mining whether clinical HIT results (Boshkov et al., 1993; Reininger et al.,
1996). This contribution occurs in several ways: activated platelet generate
high PF4 concentrations that can complex with heparin, and activated cells
also expose a higher density of heparin-binding sites (Horne and Chao, 1989).
Furthermore, these platelets may be more readily activated by heparin-
dependent antibodies. This situation occurs in patients with acute or chronic
blood activation associated with cardiopulmonary bypass, atherosclerosis,
inflammation, infections, cancer, diabetes, orthopedic surgery, among others.
Another factor determining HIT antibody formation is the type of
heparin used for heparin binding to PF4, which depends on its oligosaccha-
ride composition, polysaccharide length, and grade of sulfation (Lindahl et
al., 1994; Greinacher et al., 1995). Formation of PF4–H complexes requires a
heparin molecule with at least 12–14 oligosaccharide units and a high
sulfation grade (more than three sulfate groups per disaccharide) (Amiral et
al., 1995). Furthermore, binding of heparin to blood and endothelial cells also
increases with heparin molecule length and sulfation grade (Sobel and Adel-
man, 1988; Horne and Chao, 1990; Harenberg et al., 1994). Heparin structure
thus has a dual effect in HIT: it is required to form PF4–H complexes and also
to target these complexes onto cells. These factors could explain the higher
frequency of PF4–H antibody development and of HIT in patients receiving
UFH, compared with LMWH. With UFH, PF4–H complexes are more easily
formed and require a lower heparin concentration than with LMWH. For
the latter drug, only the subset of molecules containing at least 12–14 oligo-
saccharide units (MW > 3600 Da) can generate immunoreactive PF4–H
complexes. Thus, because LMWH has a lower propensity to form PF4–H
complexes and binds less readily to platelets and endothelial cells, LMWH
therapy may be less likely to result in thrombocytopenia even in the presence
of pathogenic HIT antibodies.
PF4–H–reactive antibodies targeted to platelets induce platelet activa-
tion, resulting in thrombocytopenia and, often, thrombosis. Occasionally,
heparin-induced thrombosis occurs in the absence of thrombocytopenia
(Hach-Wunderle et al., 1994; Bux-Gewehr et al., 1996). Platelet activation
by the IgG isotype antibodies is mediated by interaction with the platelet
FcgRIIA receptors (Kelton et al., 1994; Denomme et al., 1997). Some studies
suggest an important role for FcgRIIA polymorphism (Brandt et al., 1995;
Burgess et al., 1995). However, the role of the FcgRIIA receptor polymor-
phism is controversial (Arepally et al., 1997; Denomme et al., 1997; Suh et al.,
1997; Bachelot-Loza et al., 1998) (Chap. 9).
Platelet activation might also occur through other mechanisms, such as
direct antibody binding to exposed cell antigens (Rubinstein et al., 1995), a
phenomenon that is dependent on the antigen electric charge (Schattner et al.,
1993). Heparin is highly electronegative. Evidence for direct activation
through antigen binding is supported by the positive platelet aggregation
produced by some patient plasma samples containing only antibodies of the
IgM or IgA isotypes. Furthermore, in vivo, platelets are in their blood and
endothelial environment. Formation of heparin-containing immune com-
plexes on cell surfaces can initiate blood and endothelial cell interactions, and
this can enhance the activating effect. Cell–cell interactions may occur and be
amplified through release products that chemoattract and activate cells, or
through transcellular metabolism (Nash, 1994; Marcus et al., 1995). Platelet
products (e.g., PF4) and platelet-derived microparticles (Warkentin et al.,
1994) can induce activation of leukocytes (Aziz et al., 1995; Jy et al., 1995;
Petersen et al., 1996). Leukocyte-release products, such as cathepsin G, can
directly activate platelets and cleave h-thromboglobulin to the active chemo-
kine NAP-2, thus establishing an amplification loop. Platelet–leukocyte
aggregates can form in vivo contributing to vascular occlusion, especially in
limb vessels (Fig. 2). In a recent study, antibodies to PF4–H from patients
with HIT were shown to induce synthesis of tissue factor by monocytes in the
presence of PF4 and heparin (Pouplard et al., 2001). This could be a com-
plementary pathway for inducing thrombosis.
Various characteristics of PF4–H antibodies are another key factor for
induction of HIT. Platelet activation induced by PF4–H antibodies is usually
weak and is only pathogenic when amplified. This is demonstrated by the
variable lag phase observed in platelet aggregation studies with different plas-
mas or sera from HIT patients. Antibody concentration is an important factor
for determining the extent of platelet activation. Antibody affinity is also very
important: the higher the affinity, the lower the concentration of antibodies
required for activating platelets. Recently, a subset of antibodies to PF4–H
complexes that had platelet-activating properties was isolated in three pa-
tients with HIT. These antibodies had the highest avidity for PF4–H. In con-
trast, the bulk of antibodies to PF4–H in these patients had no effect on plate-
let activation (Amiral et al., 2000). When IgM or IgA isotypes are present,
affinity for PF4–H complexes is usually lower than that of IgG isotypes and,
consequently, high concentrations are necessary for pathogenicity. Lastly,
HIT antibodies do not all bind to the same epitope on PF4–H complexes, and
this specificity could be an important factor in their action (Horsewood et al.,
1996; Pouplard et al., 1997; Suh et al., 1998). At least two neoepitopes have
Figure 2 Occurrence of cell–cell interactions at the neighborhood of blood

activation or inflammation sites: Presence of heparin-dependent antibodies increases
the amount of cells available at these sites, amplifies cell–cell interactions and cellular
activation, and can lead to blood clotting or release of circulating cell aggregates.
PF4-H, heparin–platelet factor 4; IL-8, interleukin-8; NAP-2, neutrophil-activating
peptide 2; hTG, h-thromboglobulin.
been identified on PF4 that are distinct from the ‘‘region of positive charge’’ to
which heparin binds (Ziporen et al., 1998; Li et al., 2002) (see also color insert
and Chap. 7). Thus, anti-PF4–H antibodies are not equivalent, and those with
the strongest affinity are most pathogenic.
Recent data show that primary platelet activation in HIT involves ADP
receptors (Polgár et al., 1998), and that platelet aggregation involves GPIIb/
IIIa (Hérault et al., 1997; Jeske et al., 1997). These findings further emphasize
the importance of platelet activation amplification loops for producing the
clinical manifestations of HIT.
IV. PREEXISTING ANTICHEMOKINE ANTIBODIES
Preexisting antibodies to chemokines, such as IL-8 or NAP-2, or possibly to

PF4 itself, may be present in some patients even before heparin therapy
(Sylvester et al., 1992; Bendtzen et al., 1995). These antibodies may occur
naturally and have a regulatory role in inflammation (Reitamo et al., 1993). In
some disease states, they are present at high concentration. Antibodies to IL-
8 are most common (Reitamo et al., 1993). However, in some patients true
autoantibodies to PF4 alone can also be observed. In the absence of heparin,
these antibodies do not demonstrate clear pathogenicity. During heparin
therapy, PF4 and other chemokines are released into the circulation from
their storage pools. Heparin localizes these chemokines to blood and endo-
thelial cells. Thus, naturally occurring heparin-dependent antibodies could
then be targeted to these cells, initiating immune injury. The amount of
chemokine–heparin complexes bound to blood and endothelial cells depends
on different factors: the amount of releasable chemokines (i.e., the patient’s
clinical state); the type and dose of heparin used; and the presence of activated
cells with an increased capacity to bind heparin-like antibodies against
PF4–H complexes. As with antibodies against PF4–H complexes, these natu-
ral antichemokine antibodies could initiate cell activation and cell–cell inter-
actions, as well as generate circulating cell aggregates that could lead to vessel
occlusion.
V. CONCLUSIONS
The conditions that permit formation of the molecular PF4–H target antigen
for HIT antibodies involve the properties of the heparin used, dose and
duration of therapy, and the clinical context of the treated patient. Immuno-
reactive complexes between PF4 and heparin are formed only under certain
conditions. Their formation in high concentrations is facilitated if underlying
disease favors platelet activation and release. Similar conditions enhance the
pathogenicity of the HIT-generated antibodies. These considerations help
unravel the apparent random generation of HIT antibodies in heparin-treated
patients, as well as the seemingly random occurrence of thrombotic events.
REFERENCES

factor 4 complexed to heparin is the target for antibodies generated in heparin

4 heparin complexes in heparin-induced thrombocytopenia: a study of 44 cases.
Amiral J, Marfaing-Koka A, Wolf M, Alessi MC, Tardy B, Boyer-Neumann C, Vissac
AM, Fressinaud E, Poncz M, Meyer D. Presence of auto-antibodies to inter-
leukin-8 or neutrophil-activating peptide-2 in patients with heparin-associated-
thrombocytopenia. Blood 1996a; 88:410–416.
Amiral J, Peynaud-Debayle E, Wolf M, Bridey F, Vissac AM, Meyer D. Generation of
antibodies to heparin–PF4 complexes without thrombocytopenia in patients
treated with unfractionated or low molecular weight heparin. Am J Hematol
1996b; 52:90–95.
Amiral J, Wolf M, Fischer AM, Boyer-Neumann C, Vissac AM, Meyer D. Patho-
genicity of IgA and/or IgM antibodies to heparin–PF4 complexes in patients
with heparin-induced thrombocytopenia. Br J Haematol 1996c; 92:954–959.
Amiral J, Pouplard C, Vissac AM, Walenga JM, Jeske W, Gruel Y. Affinity puri-
fication of heparin-dependent antibodies to platelet factor 4 developed in
heparin-induced thrombocytopenia: biological characteristics and effects on
platelet activation. Br J Haematol 2000; 109:336–341.
Arepally G, McKenzie SE, Jiang X-M, Poncz M, Cines DB. FcgRIIA H/R131 poly-
morphism, subclass-specific IgG anti-heparin/platelet factor 4 antibodies and
clinical course in patients with heparin-induced thrombocytopenia and throm-
bosis. Blood 1997; 89:370–375.
Aziz KA, Cawley JC, Zuzel M. Platetets prime PMN via released PF4: mechanism of
priming and synergy with GM–CSF. Br J Haematol 1995; 91:846–853.
Bachelot-Loza C, Saffroy R, Lasne D, Chatellier G, Aiach M, Rendu F. Importance of
the FcgRIIA-Arg/His-131 polymorphism in heparin-induced thrombocytope-
nia diagnosis. Thromb Haemost 1998; 79:523–528.
Bendtzen K, Hansen MB, Ross C, Poulsen LK, Svenson M. Cytokines and
autoantibodies to cytokines. Stem Cells 1995; 13:206–222.
Bock PE, Luscombe M, Marshall SE, Pepper DS, Holbrook JJ. The multiple com-
plexes formed by the interaction of platelet factor 4 with heparin. Biochem J
1980; 191:769–776.
induced thrombocytopenia and thrombosis: clinical and laboratory studies. Br J
Haematol 1993; 84:322–328.
Brandt E, Flad HD. Structure and function of platelet-derived cytokines of the h-
thromboglobulin/interleukin 8 family. Platelets 1992; 3:295–305.
Brandt J, Isenhart CE, Osborne JM, Ahmed A, Anderson CL. On the role of platelet
FcgRIIa phenotype in heparin-induced thrombocytopenia. Thromb Haemost
1995; 74:1564–1572.
Burgess JK, Lindeman R, Chesterman CN, Chong BH. Single amino acid mutation of
Fcg receptor is associated with the development of heparin-induced thrombo-
cytopenia. Br J Haematol 1995; 91:761–766.
Bux-Gewehr I, Helmling E, Sefert UT. HAT type II and platelets within a normal
range. Kardiologia 1996; 85:656–660.

Cines DB, Tomaski A, Tannenbaum S. Immune endothelial-cell injury in heparin-
associated thrombocytopenia. N Engl J Med 1987; 316:581–589.
Cowan SW, Bakshi EN, Machin KJ, Isaacs NW. Binding of heparin to human platelet
factor 4. Biochem J 1986; 234:485–488.
Denomme GA, Warkentin TE, Horsewood P, Sheppard JI, Warner MN, Kelton JG.
Activation of platelets by sera containing IgG1 heparin-dependent antibodies:
an explanation for the predominance of the FcgRIIa ‘‘low responder’’ (his131)
gene in patients with heparin-induced thrombocytopenia. J Lab Clin Med 1997;
130:278–284.
Dunlop MG, Prowse CV, Dawes J. Heparin-induced platelet factor 4 release in pa-
tients with atherosclerotic peripheral vascular disease. Thromb Res 1987; 46:
409–410.
ization of a multimolecular PF4–heparin complex as the major antigen. Thromb
Haemost 1994; 71:247–251.
Greinacher A, Alban S, Dummel V, Franz G, Mueller-Eckhardt C. Characterization
of the structural requirements for a carbohydrate based anticoagulant with a
reduced risk of inducing the immunological type of heparin-associated
thrombocytopenia. Thromb Haemost 1995; 74:886–892.
Gruel Y, Boizard-Boval B, Wautier JL. Further evidence that alpha-granule com-
ponents such as platelet factor 4 are involved in platelet–IgG–heparin inter-
actions during heparin-associated thrombocytopenia. Thromb Haemost 1993;
70:374–375.
Hach-Wunderle V, Kainer K, Krug B, Müller-Berghaus G, Pötzsch B. Heparin-
associated thrombosis despite normal platelet counts. Lancet 1994; 344:469–
470.
Harenberg J, Malsch R, Piazolo L, Heene DL. Binding of heparin to human leu-
kocytes. Haemostaseologie 1994; 14:16–24.
Hérault JP, Lalé A, Savi P, Pflieger AM, Herbert JM. In vitro inhibition of heparin-
induced platelet aggregation in plasma from patients with HIT by SR 121566, a
newly developed Gp IIb/IIIa antagonist. Blood Coagul Fibrinolysis 1997;
8:206–207.
Horne MK III, Alkins BR. Platelets binding of IgG from patients with heparin-
induced thrombocytopenia. J Lab Clin Med 1996; 127:435–442.
1989; 74:238–243.
Horne MK III, Chao ES. The effect of molecular weight on heparin binding to plate-
lets. Br J Haematol 1990; 74:306–312.
Horsewood P, Warkentin TE, Hayward CPM, Kelton JG. The epitope specificity of
Jeske WP, Walenga JM, Szatkowski E, Ero M, Herbert JM, Haas S, Bakhos M. Effect
of glycoprotein IIb-IIIa antagonists on the HIT serum induced activation of
Jy W, Mao WW, Horstman LL, Tao J, Ahn YS. Platelet microparticles bind, activate
and aggregate neutrophils in vitro. Blood Cells Mol Dis 1995; 21:217–231.
Kelton JG, Smith JW, Warkentin TE, Hayward CPM, Denomme GA, Horsewood P.
Immunoglobulin G from patients with heparin-induced thrombocytopenia
binds to a complex of heparin and platelet factor 4. Blood 1994; 83:3232–3239.
Kelton J. Pathophysiology of heparin-induced thrombocytopenia [letter]. Br J
Haematol 1992; 82:778–784.
Li ZQ, Liu W, Park KS, Sachais BS, Arepally GM, Cines DB, Poncz M. Defining a
1236.
Lindahl U, Lidholt K, Spillmann D, Kjellen L. More to ‘‘heparin’’ than anticoagulant.
Throm Res 1994; 75:1–32.
Maccarana M, Lindahl U. Mode of interaction between platelet factor 4 and heparin.
Glycobiology 1993; 3:271–277.
Marcus AJ, Safier LB, Broekman MJ, Islam N, Fliessbach JH, Hajjar KA, Kaminski
WE, Jendraschak E, Silverstein RL, von Schacky C. Thrombosis and inflam-
mation as multicellular processes: significance of cell–cell interactions. Thromb
Haemost 1995; 74:213–217.
Nash GB. Adhesion between neutrophils and platelets: a modulator of thrombotic and
inflammatory events? Thromb Res 1994; 74:S3–S11.
Newman PM, Swanson RL, Chong BH. Heparin-induced thrombocytopenia: IgG
binding to PF4-heparin complexes in the fluid phase and cross-reactivity with
low molecular weight heparin and heparinoid. Thromb Haemost 1998; 80:292–
297.
Petersen F, Lidwig A, Flad HD, Brandt E. TNF-a renders human neutrophils
responsive to platelet factor 4. J Immunol 1996; 156:1954–1962.
Polgár J, Eichler P, Greinacher A, Clemetson KJ. Adenosine diphosphate (ADP) and
ADP receptor play a major role in platelet activation/aggregation induced by
sera from heparin-induced thrombocytopenia patients. Blood 1998; 91:549–554.
Pouplard C, Amiral J, Borg JY, Vissac AM, Delahousse B, Gruel Y. Differences in
specificity of heparin-dependent antibodies developed under low-molecular-
weight-heparin therapy and higher cross-reactivity with Orgaran. Br J
Haematol 1997; 99:273–280.
Pouplard C, Iochmann I, Renard O, Hérault O, Colombat P, Amiral J, Gruel Y.
Induction of monocyte tissue factor expression by antibodies to platelet factor 4
developed in heparin-induced thrombocytopenia. Blood 2001; 97:3300–3302.
Regnault V, de Maistre E, Carteaux JP, Gruel Y, Nguyen P, Tardy B, Lecompte T.
Platelet activation induced by human antibodies to interleukin-8. Blood 2003;
101:1419–1421.

Thromb Res 1996; 81:641–649.
Reitamo S, Remitz A, Varga J, Ceska M, Effenberger F, Jimenez S, Uitto J. Dem-
onstration of interleukin 8 and autoantibodies to interleukin 8 in the serum of
patients with sytemic sclerosis and related disorders. Arch Dermatol 1993; 129:
189–193.
Rubinstein E, Boucheix C, Worthington RE, Carroll RC. Anti-platelet antibody in-
teractions with Fcg receptor. Semin Thromb Haemost 1995; 21:10–22.
Rybak ME, Gimbrone MA, Davies PF, Handin RI. Interaction of platelet factor four
with cultured vascular endothelial cells. Blood 1989; 73:1534–1539.
Schattner M, Lazzari M, Trevani AS, Malchiodi E, Kempfer AC, Isturiz MA, Geffner
JR. Activation of human platelets by immune complexes prepared with
cationized human IgG. Blood 1993; 82:3045–3051.
Shoenfeld Y. Heparin-induced thrombocytopenia as an autoimmune disease: idio-
typic evidence for the role of anti-heparin–PF4 autoantibodies. Isr J Med Sci
1997; 33:243–245.
Sobel M, Adelman B. Characterization of platelet binding of heparins and other
glycosaminoglycans. Thromb Res 1988; 50:815–826.
Stuckey JA, St Charles R, Edwards B. A model of the platelet factor 4 complex with
heparin. Proteins 1992; 14:277–287.
Suh JS, Malik MI, Aster RH, Visentin GP. Characterization of the humoral immune
response in heparin-induced thrombocytopenia. Am J Hematol 1997; 54:196–
201.
factor 4. Blood 1998; 91:916–922.
Sylvester L, Yoshimura T, Sticherling M, Schröder JM, Ceska M, Peichi P, Leonard
EJ. Neutrophil attractant protein-1–immunoglobulin G immune complexes and
free anti-NAP-1 antibody in normal human serum. J Clin Invest 1992; 90:471–
481.
Van Rijn JLML, Trillou M, Mardiguian J, Tobelem G, Caen J. Selective binding of
heparins to human endothelial cells. Implications for pharmacokinetics.
Thromb Res 1987; 45:211–222.
Visentin GP, Aster RH. Heparin induced thrombocytopenia and thrombosis. Curr
Opin Hematol 1995; 2:351–357.
plexed with heparin or bound to endothelial cells. J Clin Invest 1994; 93:81–
88.
Visentin GP, Malik M, Cyganiak KA, Aster RH. Patients with unfractionated heparin
during open heart surgery are at high risk to form antibodies reactive with
heparin: platelet factor 4 complexes. J Lab Clin Med 1996; 128:376–383.
Am J Med 1996; 101:502–507.
Warkentin TE, Hayward CPM, Boshkov LK, Santos AV, Sheppard JI, Bode AP,
Kelton JG. Sera from patients with heparin-induced thrombocytopenia
generate platelet-derived microparticles with procoagulant activity: an expla-
nation for the thrombotic complications of heparin-induced thrombocytopenia.
Blood 1994; 84:3691–3699.
Ziporen L, Li ZQ, Park KS, Sabnekar P, Liu WY, Arepally G, Shoenfeld Y, Kieber-
Emmons T, Cines DB, Poncz M. Defining an antigenic epitope on platelet factor
4 associated with heparin-induced thrombocytopenia. Blood 1998; 92:3250–
3259.
7
Molecular Immunopathogenesis of
Gian Paolo Visentin and Chao Yan Liu

University at Buffalo, The State University of New York,
Buffalo, New York
Richard H. Aster
Medical College of Wisconsin and Blood Research Institute,
The Blood Center of Southeastern Wisconsin,
Milwaukee, Wisconsin
I. INTRODUCTION
Thrombocytopenia occurs commonly during heparin therapy, usually as a

transient fall in platelet count 1–3 days after initiation of treatment. In most
patients this is of no clinical significance, and platelet levels return to normal
within 3 days, with or without discontinuing the heparin administration. In
contrast, a relatively small group of patients develop thrombocytopenia, with
a characteristic delay that is usually 5–10 days after starting heparin therapy,
although some patients recently exposed to heparin develop an abrupt onset
of thrombocytopenia. Paradoxically, many of these patients experience ve-
nous or arterial thromboembolism (see Chap. 3). To early investigators, this
profile of a delay in onset of thrombocytopenia, as well as abrupt recurrence
on rechallenge, suggested an immune pathogenesis (see Chap. 1).
Today, there is an emerging consensus that this immune-mediated
syndrome, designated heparin-induced thrombocytopenia (HIT), is an im-
portant life- and limb-threatening complication of heparin therapy. HIT is
more common in patients receiving certain types of heparin, such as unfrac-
tionated heparin (UFH) of beef lung versus porcine mucosal origin, or un-
fractionated versus low molecular weight forms of porcine-derived heparin
179
180 Visentin et al.
(see Chap. 4). HIT can be triggered by standard therapeutic-dose heparin,

low-dose (prophylactic) treatment (Hrushesky, 1978), low molecular weight
heparin(LMWH) (Lecompte, 1991; Tardy, 1991), and even by minute quan-
tities given to ‘‘flush’’ intravascular catheters (Heeger and Backstrom, 1986;
Ling and Warkentin, 1998; Kelton and Warkentin, 1998). Various aspects of
the pathogenesis of this disorder are also summarized in Chapters 5, 6, 8, 9,
and 10.
Early investigations showed that IgG antibodies associated with HIT
could induce platelet activation in the presence of pharmacological (0.2–1 U/
mL) or even lower doses of heparin. By taking advantage of this property, two
‘‘activation’’ diagnostic tests were developed for HIT: the platelet aggregation
test (PAT) and the serotonin release assay (SRA) (see Chap. 11). Because
activation of platelets by IgG from patients with HIT in the presence of hep-
arin could be inhibited by a monoclonal antibody that blocks the platelet
FcgRIIA receptor (Kelton et al., 1988), it was assumed that the antibodies
react with heparin to form immune complexes, which, in turn, activate plate-
lets. However, early studies failed to demonstrate binding of heparin-induced
antibodies to platelets in the presence of heparin, in contrast to the behavior
of platelet-reactive antibodies induced by other drugs, such as quinidine and
quinine. Moreover, the putative heparin–IgG complexes could not be iden-
tified in most studies (Green et al., 1978; Warkentin and Kelton, 1991; Grei-
nacher et al., 1992). Thus, it remained unclear how heparin induces platelet
activation and thrombocytopenia in patients with HIT.
A new understanding of the pathogenesis of HIT and associated throm-
bosis emerged when Amiral and coworkers (1992) suggested that antibodies
in HIT might be specific for complexes of heparin and platelet factor 4 (PF4),
rather than for heparin alone. We (Visentin et al., 1994) and others (Grei-
nacher et al., 1994; Kelton et al., 1994) confirmed these findings. We added the
observation that HIT antibodies recognize PF4 bound to heparan sulfate,
normally found on the surface of endothelial cells in the form of proteoglycan,
and speculated that binding of antibodies to PF4 on endothelial cells might
promote endothelial cell damage predisposing to thrombosis (Visentin et al.,
1994). These advances enabled the development of hypotheses to explain
thrombocytopenia and thrombosis in HIT, but our understanding of the
pathogenesis of this disorder is still incomplete.
II. HEPARIN AND PLATELET FACTOR 4
Heparin and heparan sulfate constitute a distinct class of glycosaminogly-

cans (GAG). GAGs are long, linear polymers composed of repeating di-
Molecular Immunopathogenesis of HIT 181
saccharide subunits. Heparin and heparan sulfate belong to a family of

polysaccharide species, the chains of which are made up of alternating 1-4-
linked and variously sulfated residues of hexuronic (D-glucuronic or L-
iduronic) acid and D-glucosamine. The two substances differ in their
hexuronic acid composition and pattern of substitution, with heparin
having a higher content of sulfates and, consequently, a greater linear
charge density. Commercially available heparin preparations are hetero-
geneous and polydisperse, consisting of polysaccharide fragments rang-
ing in length from 3,000 to 30,000 Da (10–100 saccharide residues) (see
Chap. 8).
Heparan sulfate, together with chondroitin sulfate and dermatan sul-
fate, is widely distributed in all tissues, whereas heparin is found only in lung,
ileum, skin, lymph nodes, thymus, and appendix, where mast cells are
concentrated (Gomes and Dietrich, 1982). Metachromatic granules of mast
cells are the major reservoir of heparin (Metcalfe et al., 1979). Heparan sulfate
and other GAGs are also found in mast cell granules, but are expressed mainly
on the surface of nearly all adherent mammalian cells in the form of
proteoglycans, consisting of oligosaccharides covalently linked to a core pro-
tein (syndecan) (Höök et al., 1984).
PF4, a heparin-binding protein normally found in platelet a-granules,
is secreted when platelets are activated by various stimuli. Human PF4 is a
member of a large family of homologous proteins, encoded by genes located
on chromosomes 4 and 17, which have been designated ‘‘chemokines,’’ and
are involved in chemotaxis, coagulation, inflammation, and cell growth
(Oppenheim et al., 1991; Rollins, 1997; Luster, 1998). This family has been
separated into four branches, designated CX3C, CXC, CC, and C, based on
the relative position of the first two conserved cysteines. PF4 belongs to the
CXC family, which includes, among others, interleukin-8 (IL-8), interferon-
g–inducible protein (IP-10), platelet basic protein (PBP), and two proteins
derived from PBP by proteolytic cleavage: h-thromboglobulin (h-TG) and
neutrophil-activating protein-2 (NAP-2). Human PF4 is a symmetrical,
tetrameric molecule made up of identical subunits, each containing 70
amino acid residues of known sequence (Poncz et al., 1987), including
two disulfide bonds, a single tyrosine, but no tryptophan. The molecule is
positively charged at physiological pH (Handin and Cohen, 1976). The
crystal structure of human PF4 has been resolved (Zhang et al., 1994).
Lysine residues on the exterior faces of a-helices at the COOH-terminus of
each monomer are critical for heparin binding (Loscalzo et al., 1985).
However, residues located elsewhere on the tetramer are probably also
important for this interaction (Maccarana and Lindahl, 1993; Mayo et al.,
1995b).
182 Visentin et al.
III. NATURE OF THE EPITOPES RECOGNIZED

BY HIT ANTIBODIES
A. The Role of Polyanion
The HIT antibodies fail to recognize PF4 or heparin alone, but bind avidly to
the PF4–heparin complex (Visentin et al., 1994). Antibody epitopes, there-
fore, could be composed of either combinatorial epitopes, consisting partly
of heparin and partly of PF4, or conformational epitopes on the PF4 mole-
cule induced by heparin binding. Alternatively, a conformational change
elsewhere on the PF4 molecule, created when the complex forms, could be
targeted.
Heparin is a linear polyanion, and Maccarana and Lindahl (1993) have
suggested that it binds to positively charged PF4 by nonspecific, electrostatic
interactions, rather than by specific oligosaccharide sequence recognition.
However, Stringer and Gallagher (1997) have described a sequence on hepa-
ran sulfate consisting of a 9 kDa fragment, with sulfated domains at each end
separated by a central, N-acetylated region, that may confer some specificity
for PF4 binding. Regardless of whether PF4–heparin interaction is to some
extent specific, non-GAG molecules can be substituted for heparin in detect-
ing HIT antibodies. Kelton et al. (1994) found that highly sulfated poly-
saccharides, including heparan sulfate, pentosan polysulfate, and dextran
sulfate, could be used, provided that they contained 1.0–1.5 sulfate groups per
saccharide residue. Chondroitin sulfates A, B, and C, containing an average
of only 0.5 sulfates per saccharide residue, were inactive. Highly sulfated but
low molecular weight substances such as glucose-1,3,6-trisulfate, 1,2-cyclo-
hexanediol disulfate, and heparin disaccharide, were likewise inactive. Grei-
nacher and colleagues (1992, 1995) also characterized the structural
requirements of polysaccharides active in generating HIT antibody epitopes.
They showed that the h 1,4-linkage between disaccharides, characteristic of
heparin and other GAGs, was not essential, that heparin fractions containing
fewer than 10 residues were unable to promote platelet serotonin release by
HIT antibodies, and that branched glucan sulfates were more effective than
linear glucan sulfates of the same molecular weight. Similarly, Amiral and
coworkers (1995) found that the extent of polysaccharide sulfation is posi-
tively correlated with the ability to interact with PF4 in facilitating the binding
of HIT antibodies.
Studies conducted in our laboratory (Visentin et al., 2001) showed that
UFH of bovine and porcine origin, as well as LMWH, formed complexes with
PF4 that were recognized equally well by a panel of HIT antibodies. In studies
with heparin fragments of known size, a length of at least 10 saccharide
residues was required to form complexes with PF4 that reacted (weakly) with
this antibody panel. For optimal antibody recognition, fragments containing

at least 12 saccharide residues were required. Also, sulfated GAGs other than
heparin (e.g., heparan sulfate) as well as non-GAG sulfated polysaccharides
(e.g., fucoidan and dextran sulfate) behaved similarly to heparin in their abil-
ity to form antibody-binding complexes with PF4. However, the heparinoid
anticoagulant danaparoid (Orgaran), a mixture of nonheparin low molecular
weight GAGs having a low degree of sulfation, formed complexes that reacted
with only about one third of patient samples tested (Visentin et al., 2001). Our
findings, together with those of Kelton, Greinacher, and Amiral already cited,
indicate that the ability of GAGs and other sulfated polysaccharides to
substitute for heparin in promoting platelet activation by HIT antibodies
and to form complexes with PF4 to which the antibodies bind is directly re-
lated to the size and degree of sulfation of the polysaccharide.
To determine whether or not a polysaccharide structure is necessary
for the formation of HIT antibody epitopes, we evaluated a series of linear,
nonsaccharide, polyanionic compounds and, unexpectedly, polyvinyl sul-
fate, polyvinyl sulfonate, polyvinyl phosphate, polyvinyl phosphonate, and
polyanethole sulfonate all react with PF4 to form complexes recognized by
HIT-associated antibodies (Visentin et al., 2001) (Fig. 1). Thus, neither a
saccharide chain nor sulfate side groups is essential for a polyanion to react
with PF4 and to create sites for antibody binding, arguing strongly against the
possibility that HIT antibodies are specific for ‘‘compound epitopes’’ con-
sisting partially of GAG and partially of peptide sequence at sites where the
molecules making up the complex come into close contact. This observation,
together with the finding that HIT antibodies fail to recognize heparin com-
plexed with protamine (unpublished observation), excludes the possibility
that they recognize a configuration of the sulfated saccharide that is stabilized
on binding to a small, positively charged, spherical protein.
It appears likely, therefore, that sites for antibody binding are created
when linear polyanionic compounds bind to PF4 and alter its three-dimen-
sional configuration. Heparin-induced antibodies associated with HIT bind
avidly to complexes formed between PF4 and heparin fragments attached by
end-linkage to agarose beads, but fail to recognize PF4 complexed with
heparin molecules immobilized by multiple cross-linkages (Suh et al., 1998).
Thus, another requirement for the formation of heparin–PF4 complexes for
which HIT antibodies are specific is that the saccharide chain making up the
heparin molecule must be in a flexible, relatively unconstrained state.
Although heparin–PF4 complexes have not yet yielded to structural
analysis, some informative data about the nature of heparin–PF4 interac-
tion and its effect on PF4 structure are available. Both bovine (St. Charles et
al., 1989) and human (Zhang et al., 1994) PF4 tetramers have been crystallized
184 Visentin et al.
Figure 1 Nonspecific role of heparin and other polyanions that lead to neoepitope
formation on PF4. (From Warkentin, 2003.)
and have similar structure. Each PF4 monomer consists of a COOH-terminal

amphiphilic a-helix overlying a three-stranded antiparallel h-sheet, a struc-
ture typical of CXC chemokine family members (Luster, 1998). Two PF4
monomers associate side by side to produce a six-stranded antiparallel h-
sheet, with overlying antiparallel a-helices (AB dimer). Each AB dimer as-
sociates with an identical CD dimer through surface interaction between the
h-sheets. The elements of PF4 structure are shown schematically in Figure 2a.
Crystallographic studies have shown that both bovine (St. Charles et al.,
1989) and human (Zhang et al., 1994) PF4 contain a ring of positively charged
lysine, arginine, and histidine residues that encircle the tetramer along a line
perpendicular to the a-helices and are available for interaction with solvent.
Modeling studies (Stuckey et al., 1992) support the possibility that a nega-
tively charged heparin molecule, containing 18 saccharide residues (MWf
5.4 kDa), interacts with these positively charged residues spanning about half
the tetramer. Mayo et al. (1995a,b) created a PF4 mutant (PF4-M2) in which
the NH2-terminal 11 residues were replaced by eight residues from the ho-
mologous CXC chemokine interleukin-8, to create a tetramer that binds
heparin with the same avidity as native PF4, but is more nearly symmetrical
around all three axes, facilitating NMR structural analysis. Their data,
Figure 2 (a) Computer-generated model (WebLab ViewerPro; Molecular Simu-

lation Inc., San Diego, CA) of the human PF4 tetramer, based on the crystallographic
coordinates. (b) AC dimer view of human PF4: the amino acid residues crucial for
heparin binding are displayed. (From: a, Zhang et al., 1994; b, Loscalzo et al., 1985;
Mayo et al., 1995a). (See color insert, Fig. 3.)
contrary to PF4–heparin–binding models that center around COOH-termi-

nal a-helix lysines, indicate that arginines 20, 22, and 49, and to a lesser extent
histidine 23, threonine 25, and lysine 46, are also important for heparin bind-
ing (see Fig. 2b [see color insert, Fig. 3b] and Fig. 3). On the basis of these
findings, it was speculated that heparin does not bind perpendicularly to the
a-helices of the AB dimer, as had been suggested (Stuckey et al., 1992), but
instead reacts with the a-helix at an angle, interacting preferentially with
PF4 along the AD dimer, where it would encounter arginine and other
positively charged residues. In either model, it is plausible that binding of a
linear polyanion of sufficient length and linear charge density to positively
charged residues on the surface of PF4 could cause the structural rearrange-
ment throughout the entire tetramer necessary for generation of HIT anti-
body epitopes.
On the basis of these reports and our own observations, it is possible to
propose a model of how heparin and other linear polyanions react with PF4 to
produce configurational changes in the tetramer and create sites for HIT
antibody binding. We suggest that linear polyanions, such as heparin, that
carry appropriately spaced, strong negative charges interact with PF4 by
binding to the ring of positive charges extending between the A and D or B
and C subunits, or both. The minimum length for a fully active polyanion is
about 50 Å, equivalent to six disaccharide subunits (12-mer), with each di-
186 Visentin et al.
Figure 3 Amino acid composition of human PF4 monomers: Residues crucial for
heparin binding [COOH-terminal a-helix residues encompassing lysines 61–62 and
65–66 (Loscalzo et al., 1985), arginines 20, 22, and 49, histidine 23, threonine 25, and
lysine 46 (Mayo, 1995b)] are boxed. Filled circles identify the six residues (37, 38, 39,
49, 55, and 57) in the 47 COOH-terminal region of PF4 at which human and rat PF4
differ.
saccharide measuring about 8.4 Å in length (Visentin et al., 2001). Reconfig-

uration of the tetramer, resulting from binding of the polyanion, creates the
neoepitope(s) for which HIT antibodies are specific.
B. The Role of Protein

Only a few investigators have attempted to map the actual epitopes on hepa-
rin–PF4 complexes recognized by HIT antibodies. Horsewood et al. (1996)
studied a total of 29 antibodies from patients with HIT that were positive in
the PF4–heparin enzyme-linked immunosorbent assay (ELISA) and in the
serotonin release assay (SRA). Five of these antibodies also reacted with
reduced alkylated PF4 in the presence of heparin. The same five antibodies
also recognized a peptide containing the 19 COOH-terminal amino acid
residues of the PF4 monomer, a region that encompasses a positively charged
a-helical domain thought to be critical for heparin binding (Loscalzo et al.,
1985). However, neither reduced PF4 nor the COOH-terminal peptide could
inhibit binding of HIT antibodies to heparin–PF4 complexes, even at high
concentrations. Therefore, the clinical significance of the five antibodies is
uncertain.
Amiral and coworkers (1996a) studied a subgroup of 15 patients
thought to have HIT whose antibodies were positive in a platelet aggregation
test, but negative in heparin–PF4 ELISA. Nine of these patients had anti-
bodies that recognized NAP-2 or IL-8, or both, two members of the CXC
chemokine family that are homologous with PF4. These findings are of inter-
est because five of the nine patients had thrombotic episodes. However, re-
actions of these antibodies against NAP-2 or IL-8 in their normal config-

urations (not immobilized on plastic) were not described, and their relation to
antibodies that recognize heparin–PF4 complexes is uncertain.
Ziporen et al. (1998) studied the binding of antibodies from 50 HIT
patients to different constructs of PF4, which contained a single amino acid
substitution, and chimeric proteins, which contained various portions of hu-
man PF4 and NAP-2. Mutation to alanine of three (K62, K65, K66) of the
four lysine residues in the COOH-terminal a-helix, had only minimal effect
on the binding of HIT antibodies, and the K61 ! A mutation reduced
antibody binding by only about 50%, suggesting that the COOH-terminal
lysines of PF4 do not constitute the major antigenic site for HIT antibodies.
NH2-terminal PF4–NAP-2 chimeras exhibited only slightly reduced antibody
binding. In contrast, the PF4–NAP-2 chimera, in which the portion of PF4
lying between the third and fourth cysteine residue (amino acids 37–47) was
Figure 4 Primary and secondary structure of platelet factor 4 (PF4) in relation to

HIT neoepitopes. (Top) 3-D representation of the PF4 tetramer, indicating two
neoepitope sites (per monomer). The ‘‘ring of positive charge’’ is formed by lysine
residues in the C-terminus (light blue) and other lysine and arginine residues (dark
blue). (Bottom) The linear sequence of the 70-amino acid polypeptide of a single PF4
molecule is shown. (From Li et al., 2002.) (See color insert, Fig. 5.)
188 Visentin et al.
substituted by the corresponding NAP-2 sequence, was almost totally non-

reactive.
With a different approach we found that, although human PF4 has
74% protein sequence identity to bovine and rat PF4, neither bovine nor rat
PF4 complexed to heparin are recognized by HIT antibodies (Visentin, 1999).
Yet, rat PF4 differs from its human counterpart at only 6 of its 47 COOH-
terminal amino acids (Doit et al., 1987; Poncz et al., 1987) (see Fig. 3). To
characterize the binding sites for HIT antibodies on PF4–heparin, we con-
structed seven PF4 mutants in which the human sequence (reactive) was con-
verted to the corresponding residues of rat PF4 (nonreactive) and determined
the effect of each change on HIT antibody binding to the construct complexed
with heparin. The PF4 constructs tested were comparable with wild-type PF4
in their avidity for heparin. Each of 15 antibodies from HIT patients recog-
nized PF4–heparin complexes containing PF4 constructs bearing mutations:
E4 ! S, L11 ! V, and T16 ! S at the NH2-terminus, or A57 ! V at the
COOH-terminus just as well as wild-type human PF4-heparin complexes. In
contrast, complexes containing other COOH-terminal mutants: P37 ! A/
T38 ! V/A39 ! P, R49 ! S, and L55 ! R exhibited varying degrees of
reduced binding. The HIT antibodies tested recognized PF4 mutated at
Figure 5 Spectratype analysis of the hV 6.1 family; PBMC from a patient with
HIT were cultured in the presence or in the absence of the antigen (heparin–PF4).
‘‘Oligoclonal’’ expansion of T cells is observed only under stimulation with heparin–
PF4.
positions 49 and 55 only at a higher ratio of heparin to PF4 (0.8 U/mL vs. 0.5
U/mL). None of the 15 antibodies recognized peptides comprising the 26 or
15 COOH-terminal amino acid residues of the PF4 monomer, or reduced
alkylated human PF4 either in presence or absence of heparin.
These results, together with the observations by Ziporen and associates
(1998), point to the region of PF4 between the third and fourth cysteine
residues as the major antigenic site for HIT antibody binding (Fig. 4 [see color
insert, Fig. 5]). Li et al. (2002) using a series of mouse/human PF4 chimeras
identified another antigenic site, on PF4–heparin that requires both P34 and
an intact N-terminus (Fig. 4 [see color insert, Fig. 5]). The latter results,
together with our studies utilizing biotin-labeled affinity-purified HIT anti-
bodies in a competitive inhibition assay (Suh et al., 1998), indicate that at
least three dominant HIT antibody recognition sites can be distinguished
and further support the idea that HIT antibodies recognize conformation-
dependent ‘‘neoepitopes’’ formed on PF4 when it binds to heparin.
IV. THE CELLULAR IMMUNE RESPONSE
The finding that HIT antibodies can be of the IgM, IgG, or IgA isotype
(Visentin et al., 1994; Greinacher et al., 1994; Kelton et al., 1994; Amiral et al.,
1995, 1996b; Arepally et al., 1997; Suh et al., 1997) indicates that class switch-
ing, likely requiring helper T cells, takes place in patients mounting a humoral
immune response to heparin–PF4. Although HIT is a drug-induced disorder,
parallels for the role of T cells in HIT may be drawn from studies of auto-
immune conditions, such as systemic lupus erythematosus, systemic sclerosis,
and insulin autoimmune syndrome (Ito et al., 1993; Crow et al., 1994; Ku-
wana et al., 1995b). In both lupus and scleroderma, T-helper cells mediate
antigen-specific autoantibody production by B cells (Adams et al., 1991;
Mohan et al., 1993; Kuwana et al., 1995a).
We hypothesize that the heparin–PF4 complex not only is the target for
antibody, but also is the stimulus for T-cell activation, and have used T-cell
receptor (TCR) spectratyping (Maslanka et al., 1995), also called immuno-
scope (Cochet et al., 1992; Pannetier et al., 1993), and clonotyping (Maslanka
et al., 1996) to characterize the T-cell response to PF4–heparin complexes in
HIT. The TCRs of more than 95% of peripheral blood T cells are composed
of two highly variable a- and h-chain glycoproteins, which function together
in a complex with five other invariant molecules (CD3 complex) on the surface
of the cell. The genes encoding the TCR h-chain subunit undergo sequential
rearrangments analogous to that of the immunoglobulin superfamily of
genes, during which D and J segments first, and then V segments, are com-
bined to form various VDJ sequences (LaRoque and Robinson, 1996). Diver-
190 Visentin et al.
sity is further increased by the random removal and insertion of nucleotides

between the V, D, and J segments. The resulting VDJ sequence encodes the so-
called third complementarity region or CDR3 loop, the primary region in-
volved in peptide recognition by the TCR (LaRoque and Robinson, 1996).
Through allelic exclusion, only a single h-chain is expressed on the surface of a
T cell, thus the h-chain CDR3 sequence provides a clonal marker for T-cell
lineages and has been used to assess T-cell repertoires. TCR spectratyping is
a polymerase chain reaction (PCR)–based technique that provides a readout
of TCR h-chain diversity in a given T-cell population. TCR clonotying is a
refinement of spectratyping whereby oligonucleotide probes specific to a
given CDR3 loop region from a particular TCR h-chain (thus a ‘‘clonotype’’)
are used to detect the presence or absence of the clonotype in a given T-cell
population.
Culture of peripheral blood mononuclear cels (PBMC) from patients
experiencing HIT incubated with heparin–PF4 complexes, but not heparin
or PF4 alone, leads to selective expansion of T-cell subsets (Liu et al., 2000;
Bacsi et al., 2001). On in vitro culturing of PBMC from two HIT patients,
the PF4–heparin complexes preferentially stimulated CD4 T cells expressing
TCR with h-chains of the V 5.1 family, with a shared core CDR3 region
amino acid motif (PGTG) (Bacsi et al., 1999). In a study of a third HIT pa-
tient, we found heparin–PF4–specific expansion of several hV 17 TCR clono-
types with yet another shared core CDR3 region amino acid motif (TSG)
(Bacsi et al., 2001). However, T-cell lines derived from this third patient and
maintained in culture in the presence of heparin–PF4 demonstrated selective
expansion of the hV 6.1 (Fig. 5) and 17 families sharing the conserved core
GTG motif previously identified in the hV5.1 family of the first two HIT pa-
tients (Liu et al., 2001).
Our findings provide evidence for the existence of T-cell subpopula-
tions specific for heparin–PF4 complexes in the peripheral blood of patients
experiencing HIT and suggest that a common CDR3 TCR motif may be im-
portant for recognition of a peptide derived from PF4 processed by antigen-
presenting cells in the presence of heparin. The observations are consistent
with the possibility that only a limited number of helper T cells are used in
mounting an antibody response to heparin–PF4. Further studies of the cel-
lular immune response in HIT may lead to insights concerning drug-induced
breakdown of immune tolerance to a self-protein.
It is presently unclear why patients with HIT mount a brisk humoral
immune response to an autologous protein (PF4). PF4 is undoubtedly pro-
cessed, under normal circumstances, by antigen-presenting cells (APC) with-
out triggering immunity. One possible explanation for the induction of a
response to PF4 after injection of heparin is that heparin may perturb the
processing of PF4 by APC in a way such that peptides not ordinarily pro-
duced (cryptic peptides) are generated and presented to T cells in the context
of class II MHC molecules. Studies in murine systems provide examples of
‘‘autoimmune’’ states triggered by exogenous agents that perturb protein
processing (Hess et al., 1991; Griem et al., 1996), but this phenomenon is not
well characterized in the context of human disease. Our studies support a
model in which pharmacological doses of heparin cause aberrant processing
of PF4 by APCs, leading to the presentation of peptides not ordinarily seen by
the immune system. This hypothesis can be tested directly if T-cell clones can
be developed from mononuclear cells responding to heparin–PF4 cultures.
It appears that multiple factors influence the formation of antibodies
specific for heparin–PF4 complexes in patients receiving heparin. Currently,
there is no evidence to support genetic predisposition as a basis for antibody
formation in patients receiving heparin. Unlike the situation in alloimmune
thrombocytopenia (de Waal et al., 1986; Mueller-Eckhardt et al., 1989), no
connection between HIT and human leukocyte antigens (HLA) has been
found (Greinacher and Mueller-Eckhardt, 1993). IgM antibodies specific for
heparin–PF4 complexes are a common finding in HIT (Visentin et al., 1994,
1999), indicating a primary immune response, and it could be speculated that
patients who received UFH previously may be at greater risk to produce
heparin–PF4–specific antibodies and develop HIT, if rechallenged with hep-
arin. However, Cadroy et al. (1994) described a patient with a history of HIT
who mounted a brisk IgM response when challenged again with UFH 3 years
later. A report by Warkentin and Kelton (2001) suggests that there is no
anamnestic immune response in HIT (i.e., patients either have ‘‘typical’’ HIT
[onset at days 5–10] or ‘‘rapid’’ HIT, the latter apparently caused by residual
circulating HIT antibodies, rather than a secondary immune response). Fur-
thermore, HIT did not necessarily recur in patients who were exposed to
heparin a second time.
V. IMPLICATIONS
The identification of mutations of human PF4 that lead to loss of HIT anti-
body binding will not necessarily localize the epitopes at which antibodies
attach because the actual binding site(s) could be elsewhere in the PF4 tet-
ramer. Moreover, HIT antibodies appear to recognize multiple sites on PF4–
heparin (Suh et al., 1998). Because the PF4 molecule is a nearly symmetrical
tetramer (Ibel et al., 1986), the HIT epitope could be expressed four times
on each heparin–PF4 heterodimer, creating the potential for even a single
antibody clone to react with four sites on a PF4 tetramer complexed with hep-
arin. Studies from our group (Visentin et al., 1994, 1996) and others (Amiral
et al., 1995; Arepally et al., 1995) have shown that, although antibodies
192 Visentin et al.
reactive with heparin–PF4 complexes are nearly always present in patients

with HIT, not all patients who form such antibodies experience thrombosis,
or even thrombocytopenia. Factors that could predispose antibody formers
to develop the HIT syndrome include the formation of unusually potent
(high-titer) antibodies (Suh et al., 1997) and the presence of underlying
conditions, congenital or acquired, that predispose to thrombosis. It can be
speculated that antibodies recognizing certain sites on heparin–PF4 form
immune complexes that are particularly effective in activating platelets. The
same antibodies might be more likely to promote vessel injury when they bind
to PF4 complexed with GAG on endothelial cells. Alternatively, patients who
make antibodies that recognize multiple sites on heparin–PF4 may be more
likely to produce pathogenic immune complexes, leading to more severe
symptomatology.
ACKNOWLEDGMENTS
Supported in part by Grants HL-64704, HL-13629, and HL-44612 from the

National Heart, Lung, and Blood Institute.
REFERENCES
Adams S, Leblanc P, Datta SK. Junctional region sequences of T-cell receptor beta-
chain genes expressed by pathogenic anti-DNA autoantibody-inducing helper T
cells from lupus mice: possible selection by cationic autoantigens. Proc Natl
Acad Sci USA 1991; 88:11271–11275.
Amiral J, Bridey F, Wolf M. Antibodies to macromolecular platelet factor-4 heparin
complexes in heparin-induced thrombocytopenia: a study of 44 cases. Thromb
Haemost 1995; 73:21–28.
Amiral J, Marfaing-Koka A, Wolf M, Alessi MC, Tardy B, Boyer-Neumann C, Vissac
AM, Fressinaud E, Poncz M, Meyer D. Presence of autoantibodies to inter-
leukin-8 or neutrophil-activating peptide-2 in patients with heparin-associated
thrombocytopenia. Blood 1996a; 88:410–416.
Amiral J, Wolf M, Fischer AM, Boyer-Neumann C, Vissac AM, Meyer D. Patho-
with heparin-induced thrombocytopenia. Br J Haematol 1996b; 92:954–959.
Arepally G, Reynolds C, Tomaski A, Amiral J, Jawad A, Poncz M, Cines DB.
Comparison of PF4–heparin ELISA assay with the 14C-serotonin release assay
in the diagnosis of heparin-induced thrombocytopenia. Am J Clin Pathol 1995;

104:648–654.
Arepally G, McKenzie SE, Jiang XM, Poncz M, Cines DB. FcgRIIA H/R131
polymorphism, subclass-specific IgG anti-heparin/platelet factor 4 antibodies
and clinical course in patients with heparin-induced thrombocytopenia and
thrombosis. Blood 1997; 89:370–375.
Bacsi S, De Palma R, Visentin GP, Gorski J, Aster RH. Complexes of heparin and
platelet factor 4 specifically stimulate T cells from patients with heparin-induced
thrombocytopenia/thrombosis. Blood 1999; 94:208–215.
Bacsi S, Geoffrey R, Visentin GP, De Palma R, Aster RH, Gorski J. Identification of T
cells responding to a self-protein modified by an external agent. Hum Immunol
2001; 62:113–124.
Cadroy Y, Amiral J, Raynaud H, Brunel P, Mazaleyrac A, Sauer M, Sie P. Evolution
of antibodies anti-PF4/heparin in a patient with a history of heparin-induced
thrombocytopenia reexposed to heparin [letter]. Thromb Haermost 1994; 71:
247–251.
Cochet M, Pannetier C, Regnault A, Darche S, Leclerc C, Kourilsky P. Molecular
detection and in vivo analysis of the specific T cell response to a protein antigen.
Eur J Immunol 1992; 22:2639–2647.
Crow MK, DelGiudice-Asch G, Zehetbauer JB, Lawson JL, Brot N, Weissach H,
Elkon KB. Autoantigen-specific T cell proliferation induced by the ribosomal
P2 protein in patients with systemic lupus erythematosus. J Clin Invest 1994;
94:345–352.
de Waal LP, van Dalen CM, Engelfriet CP, von dem Borne AEGK. Alloimmu-
nization against the platelet specific Zwa antigen, resulting in neonatal allo-
immune thrombocytopenia or posttransfusion purpura, is associated with
the supertypic DRw52 antigen including DR3 and DRw6. Hum Immunol
1986; 17:45–53.
Doi T, Greenberg SM, Rosenberg RD. Structure of the rat platelet factor 4 gene: a
marker for megakaryocyte differentiation. Mol Cell Biol 1987; 7:898–904.
Gomes PB, Dietrich CP. Distribution of heparin and other sulfated glycosaminogly-
cans in vertebrates. Comp Biochem Physiol 1982; 73B:857–863.
Green D, Harris K, Reynolds N, Roberts M, Patterson R. Heparin immune thrombo-
cytopenia: evidence for heparin–platelet complex as the antigenic determinate.
J Lab Clin Med 1978; 91:167–175.
Greinacher A, Mueller-Eckhardt C. Heparin-associated thrombocytopenia: no asso-
ciation of immune response with HLA. Vox Sang 1993; 65:151–153.
nia: the antibody is not heparin specific. Thromb Haemost 1992; 67:545–549.
Haemostas 1994; 71:247–251.
of the structural requirements for a carbohydrate-based anticoagulant with a
194 Visentin et al.
reduced risk of inducing the immunological type of heparin-associated throm-

bocytopenia. Thromb Haemost 1995; 74:886–892.
Griem P, Panthel K, Kalbacher H, Gleichmann E. Alteration of a model antigen by
Au(III) leads to T cell sensitization to cryptic peptides. Eur J Immunol 1996;
26:279–287.
Handin RJ, Cohen HJ. Purification and binding properties of human platelet factor
four. J Biol Chem 1976; 251:4273–4282.
Heeger PS, Backstrom JT. Heparin flushes and thrombocytopenia. Ann Intern Med
1986; 105:143.
Hess EV. Drug-related lupus [review]. Curr Opin Rheumatol 1991; 3:809–814.
Höök M, Kjellén L, Johansson S, Robinson J. Cell-surface glycosaminoglycans. Annu
Rev Biochem 1984; 53:847–869.
Hrushesky WJ. Subcutaneous heparin-induced thrombocytopenia. Arch Intern Med
1978; 138:1489–1491.
Ibel K, Polland GA, Baldwin JP, Pepper DS, Luscombe M, Holbrook JJ. Low-
resolution structure of the complex of human platelet factor 4 with heparin
determined by small-angle neutron scattering. Biochim Biophys Acta 1986;
870:58–63.
Ito Y, Nieda M, Uchigata Y, Nishimura M, Tokunaga K, Kuwata S, Obata F,
Tadokoro K, Hirata Y, Omori Y, Juji T. Recognition of human insulin in the
context of HLA DRB1*0406 products by T cells of insulin autoimmune
syndrome patients and healthy donors. J Immunol 1993; 151:5770–5776.
925–930.
Kelton JG, Warkentin TE. Heparin-induced thrombocytopenia. Diagnosis, natural
history, and treatment options [review]. Postgrad Med 1998; 103:169–171, 175–
178.
Kuwana M, Medsger TA Jr, Wright TM. T cell proliferative response induced by
DNA topoisomerase I in patients with systemic sclerosis and healthy donors.
J Clin Invest 1995a; 96:586–596.
Kuwana M, Medsger TA Jr, Wright TM. T and B cell collaboration is essential for the
autoantibody response to DNA topoisomerase I in systemic sclerosis. J Im-
munol 1995b; 155:2703–2714.
LaRoque R, Robinson MA. Diversity in the human T cell receptor beta chain [review].
Hum Immunol 1996; 48:3–11.
Lecompte T. Thrombocytopenia associated with low-molecular-weight heparin [let-
ter]. Lancet 1991; 338:1217.

1236.
Ling E, Warkentin TE. Intraoperative heparin flushes and subsequent acute heparin-
induced thrombocytopenia. Anesthesiology 1998; 89:1567–1569.
Liu CY, Gorski J, Aster RH, Visentin GP. Identification of a T-cell receptor CDR3
motif associated with the cellular immune response in heparin-induced throm-
bocytopenia/thrombosis [abstr]. Blood 2000; 96(suppl):817a.
Liu CY, Gorski J, Aster RH, Visentin GP. T cells from a patient experiencing heparin-
induced thrombocytopenia/thrombosis (HIT/T) share a common T cell recep-
tor CDR3 motif (GTG) when cultured in the presence of the putative antigen:
PF4–heparin [abstr]. Thromb Haemost 2001; July(suppl):OC217.
Loscalzo J, Melnick B, Handin RI. The interaction of platelet factor 4 and glyco-
saminoglycans. Arch Biochem Biophys 1985; 240:446–455.
Luster AD. Chemokins—chemotactic cytokines that mediate inflammation [review].
N Engl J Med 1998; 338:436–445.
Maccarana M, Lindahl U. Mode of interaction between platelet factor 4 and heparin.
Glycobiology 1993; 3:271–277.
Máslanka K, Piatek T, Gorski J, Yassai M, Gorski J. Molecular analysis of T cell
repertoires. Spectratypes generated by multiplex polymerase chain reaction and
evaluated by radioactivity or fluorescence. Hum Immunol 1995; 44:28–34.
Maslanka K, Yassai M, Gorski J. Molecular identification of T cells that respond in a
primary bulk culture to a peptide derived from a platelet glycoprotein
implicated in neonatal alloimmune thrombocytopenia. J Clin Invest 1996; 98:
1802–1808.
Mayo KH, Roongta V, Ilyina E, Milius R, Barker S, Quinlan C, La Rosa G, Daly TJ.
NMR solution structure of the 32-kDa platelet factor 4 ELR-motif N-terminal
chimera: a symmetric tetramer. Biochemistry 1995a; 34:11399–11409.
Mayo KH, Ilyina E, Roongta V, Dundas M, Joseph J, Lai CK, Maione T, Daly TJ.
Heparin binding to platelet factor-4. An NMR and site-directed mutagenesis
study: arginine residues are crucial for binding. Biochem J 1995b; 312:357–
365.
Metcalfe DD, Lewis RA, Silbert JE, Rosenberg RD, Wasserman SI, Austen KF.
Isolation and characterization of heparin from human lung. J Clin Invest 1979;
64:1537–1543.
Mohan C, Adams S, Stanik V, Datta SK. Nucleosome: a major immunogen for
pathogenic autoantibody-inducing T cells of lupus. J Exp Med 1993; 177:1367–
1381.
Mueller-Eckhardt C, Kiefel V, Kroll H, Mueller-Eckhardt G. HLA-DRw6, a new
immune response marker for immunization against the platelet alloantigen Bra.
Vox Sang 1989; 57:90–91.
Oppenheim JJ, Zachariae COC, Mukaida N, Matsushima K. Properties of the novel
proinflammatory supergene ‘‘intercrine’’ cytokine family. Annu Rev Immunol
1991; 9:6–648.
Pannetier C, Cochet M, Darche S, Casrouge A, Zöller M, Kourilsky P. The sizes of
CDR3 hypervariable regions of the murine T-cell receptor h chains vary as a
196 Visentin et al.
function of the recombined germ-line segments. Proc Natl Acad Sci USA 1993;
90:4319–4323.
Poncz M, Surrey S, LaRocco P, Weiss MJ, Rappaport EF, Conway TM, Schwartz E.
Cloning and characterization of platelet factor 4 cDNA derived from a human
erythroleukemic cell line. Blood 1987; 69:219–223.
Rollins BJ. Chemokines [review]. Blood 1997; 90:909–928.
St. Charles R, Walz DA, Edwards BFP. The three-dimensional structure of bovine
platelet factor 4 at 3.0-A resolution. J Biol Chem 1989; 264:2092–2099.
Stringer SE, Gallagher JT. Specific binding of the chemokine platelet factor 4 to
heparan sulfate. J Biol Chem 1997; 272:20508–20514.
Stuckey JA, St. Charles R, Edwards DFP. A model of the platelet factor 4 complex
with heparin. Proteins 1992; 14:277–287.
201.
factor 4 complexes. Blood 1998; 91:916–922.
Tardy B. Thrombocytopenia associated with low-molecular-weight heparin [letter].
Lancet 1991; 338:1217.
Visentin GP. Heparin-induced thrombocytopenia: molecular pathogenesis. Thromb
Haemost 1999; 82:448–456.
Visentin GP, Ford SE, Scott PJ, Aster RH. Antibodies from patients with heparin-
with heparin: platelet factor 4 complexes. J Lab Clin Med 1996; 128:376–383.
Visentin GP, Moghaddam M, Beery SE, McFarland JG, Aster RH. Heparin is not
required for detection of antibodies associated with heparin-induced thrombo-
cytopenia thrombosis. J Lab Clin Med 2001; 138:22–31.
Warkentin TE. Heparin-induced thrombocytopenia: pathogenesis and management
[review]. Br J Haematol 2003; 12:535–555.
Warkentin TE, Kelton JG. Heparin-induced thrombocytopenia. Prog Hemost
Thromb 1991; 10:1–34.
Warkentin TE, Kelton JG. Temporal aspects of heparin-induced-thrombocytopenia.
N Engl J Med 2001; 344:1286–1292.
Zhang X, Chen L, Bancroft DP, Lai CK, Maione TE. Crystal structure of recom-
binant human platelet factor 4. Biochemistry 1994; 33:8361–8366.
Ziporen L, Li ZQ, Park KS, Sabnekar P, Liu WY, Arepally G, Shoenfeld Y, Kieber-
Emmons T, Cines DB, Poncz M. Defining an antigenic epitope on platelet factor
4 associated with heparin-induced thrombocytopenia. Blood 1998; 92:3250–
3259.
8
Role of Sulfated Polysaccharides
in the Pathogenesis of Heparin-
Induced Thrombocytopenia
Susanne Alban
Christian-Albrechts University of Kiel, Kiel, Germany
Andreas Greinacher
Ernst-Moritz-Arndt University Greifswald, Greifswald, Germany
I. INTRODUCTION
Unfractionated heparin (UFH) and low molecular weight heparin (LMWH)

are the anticoagulants of choice when parenteral anticoagulation is required.
Both can be given subcutaneously or intravenously, and both are effective in a
variety of clinical settings (Hirsh et al., 2001). UFH in particular has several
limitations. These include its poor bioavailability after subcutaneous injec-
tion as well as the marked variability in the anticoagulant response to UFH
treatment in patients with acute thromboembolism (Hirsh, 1991; Young et al.,
1992). Another problem is the risk of inducing heparin-induced thrombocy-
topenia (HIT). These limitations are closely linked (Greinacher, 1995): the
underlying cause is the high density of negative charges of the heparin
molecule, leading to nonspecific binding of heparin to plasma proteins other
than antithrombin (AT). This results in inhibition of the anticoagulant effects
of heparin, as well as changes in the conformational structure of the proteins
following binding to heparin, with the potential for exposure of neoepitopes,
or cryptic epitopes, toward which an immune response can be induced.
In this chapter, the mechanism and structural requirements for complex
formation between sulfated carbohydrates, especially heparin, and proteins,
197
198 Alban and Greinacher
such as platelet factor 4 (PF4), are reviewed. The pathophysiological con-

sequences of these interactions in causing HIT are summarized. From these
considerations, the prospects for development of carbohydrate-based heparin
alternatives that would not cause immune thrombocytopenia are discussed.
II. INTERACTIONS OF PF4 WITH SULFATED

CARBOHYDRATES
A. Structure of PF4
Heparin activity is neutralized by PF4, a protein released from the a-granules
of activated platelets (Sear and Poller, 1973; Klener and Kubisz, 1978; Luscher
and Kaser-Glanzman, 1975; Niewiarowski, 1976; Walsh, 1976), which at-
taches to the endothelial surface by binding to glycosaminoglycans (GAGs)
(Novotny et al., 1993). PF4 is a compact homotetrameric globular protein with
a subunit molecular weight (MW) of 7780 Da (70 amino acid residues per
subunit) (Kaplan and Niewiarowski, 1985; Mayo et al., 1995) containing 6.0%
arginine, 3.2% histidine, and 12.3% lysine basic amino acids (Moore et al.,
1975). The NH2-terminal residues form antiparallel h-sheet-like structures
that induce non-covalent associations between dimers and also contribute to
the cohesion of the tetrameric unit. Furthermore, electrostatic interactions of
multiply charged amino acid side chains and hydrogen-bonding interactions
at the AB/CD dimer interface serve to stabilize the tetrameric structure. The
COOH-terminal a-helices, which contain four lysine residues, are arranged as
antiparallel pairs on the surface of each extended h-sheet (St. Charles et al.,
1989). The lysine residues are predominantly on one side, resulting in a ‘‘ring of
positive charge’’ that runs perpendicularly across the helices (Stuckey et al.,
1992; Zhang et al., 1994) (see Figs. 2 and 4 [color inserts Figs. 3 and 5 ], Chap. 7).
B. Structure of Heparin
Heparin is a polydisperse mixture of GAGs with MWs ranging from 5 to 40
kDa, with an average MW of 13 kDa (Linhardt and Toida, 1997). It is com-
posed of alternating D-glucosamine residues linked 1!4 to either L-iduronic
acid or D-glucuronic acid (Casu, 1985). The principal repeating unit in
heparin is the trisulfated disaccharide [! 4)-O-a-L-iduronic acid-2-sulfate
(1 ! 4)-O-a-D-glucosamine-2,6-disulfate (1 !] (Fig. 1), which represents 75–
90% of the heparin chain (Linhardt et al., 1992). The remaining 10–25% of
disaccharide units differ in their degree and positions of sulfation (Linhardt
et al., 1988). Besides, there are disaccharides consisting of unsulfated glucu-
ronic acid and/or N-acetylglucosamine. With a SO
3 :COO ratio of 2.0–2.5,
heparin is the GAG with the highest charge density. By binding to domains
containing positively charged amino acids, especially arginine and lysine, it
Role of Sulfated Polysaccharides in HIT 199
Figure 1 Main disaccharide unit of heparin composed of 75–90% heparin.
interacts with many proteins, resulting in manifold biological activities. The

most prominent example is a well-defined pentasaccharide sequence with a
central a-D-glucosamine-2,3,6-trisulfate unit, which binds specifically to AT
(Choay, 1989). About 30% (range 10–50%) of the heparin chains contain this
pentasaccharide (Fig. 2). These molecules are called high-affinity heparin in
contrast to the low-affinity heparin without this AT-binding site (Casu, 1990).
AT is a natural serine protease inhibitor that controls blood coagulation by
forming equimolar covalent complexes with certain coagulation enzymes. The
anticoagulant action of heparin is based mainly on accelerating the slow rate
of factor Xa (FXa) and thrombin (FIIa) inhibition by AT (Björk et al., 1989).
Whereas the heparin pentasaccharide is sufficient for FXa inhibition, throm-
bin inhibition requires a minimum heparin chain length of 18 monosacchar-
ides (5400 Da) to permit simultaneous binding of heparin to both AT and
thrombin.
C. PF4-Sulfated Polysaccharide Complexes

Platelet factor 4 has the highest affinity to heparin among proteins stored
within the platelet a-granules. Heparin molecules bind to PF4 by interactions
with the positively charged residues on the surface of PF4 (see also Chap. 7).
Stuckey and coworkers (1992) suggested that heparin is bound to PF4 by be-
ing wrapped around the tetramer along the ring of positive charge. A heparin
molecule with 16–18 monosaccharides interacts with PF4 by spanning about
half of the tetramer. As a consequence, only very long molecules are able to
wrap around the complete tetramer. Mayo and coworkers (1995) identified a
loop containing Arg-20, Arg-22, His-23, and Thr-25, as well as Lys-46 and
Arg-49, which are more relevant for heparin binding than the terminal
COOH-lysines. For optimal interaction with PF4, a heparin molecule should
200
Figure 2 Pentasaccharide sequence of the AT-binding site of heparin: sulfate groups essential for the AT-binding
are encircled.
Alban and Greinacher
consist of at least 12 monosaccharides (Visentin, 1999; Mikhailov et al., 1999).

At low concentrations (0.1–1.0 IU/mL) of heparin and high concentrations of
PF4, several PF4 tetramers compete for heparin binding. This permits
binding of a heparin chain to more than one PF4 tetramer. Particularly if a
heparin molecule is longer than 16 monosaccharides, it is able to bind to, and
thereby bridge, two PF4 tetramers. Thus, at certain concentrations of heparin
and PF4, large, multimolecular PF4–heparin complexes are formed that can
become dissociated in the presence of high heparin concentrations (Bock et
al., 1980; Greinacher et al., 1995, 1994c) (see Fig. 2, Chap. 5).
Only heparin molecules containing 16 or more monosaccharides com-
pletely bind to immobilized PF4, resulting in total neutralization of their
antifactor Xa (anti-Xa) and anti-thrombin (anti-IIa) activities, whereas pro-
gressively smaller oligosaccharides (without anti-IIa activity) become increas-
ingly resistant to neutralization of their anti-Xa activity by PF4 (Denton et al.,
1983; Lane et al., 1984). Because of their reduced sensitivity to inactivation by
PF4, LMWHs are more active than UFH in platelet-rich plasma (Beguin et
al., 1989), despite their lower activity in platelet-poor plasma (Samama et al.,
1994). However, in contrast to their anti-Xa activity, the anti-IIa activity of
LMWHs, which is mediated by molecules with a MW of more than 5400 Da,
can be completely neutralized by higher PF4 concentrations (Padilla et al.,
1992; Bendetowicz et al., 1994).
Formation of the PF4–heparin complex is independent of the AT-
binding site, because heparin of either low or high affinity to AT binds to PF4
with a similar apparent Kd (Loscalzo et al., 1985). The interaction appears to
be mediated by electrostatic interactions, as shown by studies of heparin oli-
gosaccharides with different charge densities (Maccarana and Lindahl, 1993).
Therefore, the complexes are dissociable. Indeed, heparin can be displaced
from PF4 by sulfated polysaccharides, such as other GÁGs (Handin and
Cohen, 1976), dextran sulfate (Loscalzo et al., 1985), or xylan sulfate (Camp-
bell et al., 1987). The molar ratios required for complex formation increase in
the order: UFH < LMWH < heparan sulfate < dermatan sulfate < chon-
droitin-6-sulfate < chondroitin-4-sulfate (Handin and Cohen, 1976). Besides
the degree of sulfation (DS) and MW, other structural parameters, such as the
type of the uronic acid and the location of the sulfate groups on the amino
sugar in the case of GAGs, influence the affinity of a polysaccharide to PF4
(Table 1).
D. Interactions of PF4 with Sulfated Polysaccharides In Vivo

Intravenous injection of heparin causes an increase in plasma PF4 level,
whereas subcutaneous injection does not (O’Brien et al., 1985). The maximum
amount of PF4 released corresponds to only about 5% of total platelet PF4
Table 1 Main Disaccharide Units of Mammalian Glycosaminoglycans, Arranged by

Increasing Affinity to PF4
Glycosaminoglycan Main disaccharide unit DSa
Hyaluronic acid [4)-h-D-GlcpA-(1 ! 3)-h-D-GlcpNAc-(1 !] 0

Keratan sulfate [3)-h-D-Gal-(1 ! 4)-h-D-GlcpNAc6S-(1 !] <1.0
Chondroitin sulfate A [4)-h-D-GlcpA-(1 ! 3)-h-D-GalpNAc4S-(1 !] <1.0
Chondroitin sulfate C [4)-h-D-GlcpA-(1 ! 3)-h-D-GalpNAc6S-(1 !] <1.0
Dermatan sulfate = ChS B [4)-a-L-IdopA-(1 ! 3)-h-D-GalpNAc4S-(1 !] 1.0
Heparan sulfate [4)-h-D-GlcpA-(1 ! 4)-a-D-GlcpNAc6S-(1 !] 1.0–1.5
Low molecular weight heparinb [4)-a-L-IdopA2S-(1 ! 4)-a-D-GlcpN2S,6S-(1 !] 2.0–2.5
Unfractionated heparin [4)-a-L-IdopA2S-(1 ! 4)-a-D-GlcpN2S,6S-(1 !] 2.0–2.5
a
DS, degree of sulfation (sulfate groups per disaccharide unit).
b
Produced by degradation of unfractionated heparin.
(Dawes et al., 1982). Platelets do not release PF4 when incubated with heparin
in vitro (see Chap. 5). However, in vivo, heparin and some other GAGs are
able to increase plasma PF4 levels (Cella et al., 1986). Thus, endothelial-
bound, rather than platelet-stored, PF4 seems to be the predominant source
of the PF4 released by heparin. Most likely, heparin and other high-sulfated
polysaccharides are able to displace PF4 from endothelial heparan sulfate in
relation to their affinity for PF4 (O’Brien et al., 1985).
III. PF4–HEPARIN COMPLEXES AS THE MAJOR ANTIGEN

RECOGNIZED BY HIT ANTIBODIES
A. Formation of Immune Complexes by HIT Antibodies
Two types of platelet count reduction associated with heparin treatment must
be distinguished (Greinacher, 1995; Warkentin et al., 1995, 1998). Mild
thrombocytopenia that occurs at the beginning of heparin treatment is most
common, usually with high doses of heparin or under certain clinical circum-
stances (e.g., following thrombolytic therapy or during the perioperative
period). Known as nonimmune heparin-associated thrombocytopenia, the
platelet count fall is typically unaccompanied by clinically adverse events, and
the platelet count recovers despite continued use of heparin. In contrast, HIT
typically occurs between the 5th and 20th days after starting heparin therapy.
HIT is often associated with thromboembolic complications.
Whereas nonimmune heparin-associated thrombocytopenia may be
caused by direct platelet-activating effects of heparin (see Chap. 5), HIT re-
sults from an immune mechanism (Amiral et al., 1992). However, both
thrombocytopenic syndromes are closely linked in their pathogenesis (Grei-
nacher, 1995), as the strong anionic character of heparin plays a pathogenic

role for each. Heparin adheres to both endothelial-bound and platelet-derived
PF4, with a potential further PF4 increase resulting from the platelet-acti-
vating effects of heparin (Horne and Hutchison, 1998; Newman and Chong,
2000). Heparin binding to PF4 exposes at least two neoepitopes, or cryptic
autoantigens, on PF4 (Li et al., 2002). Some patients develop antibodies,
predominantly IgG, but also IgM or IgA isotypes (Amiral, 1997) against the
multimolecular PF4–heparin complexes, which thus represent the major
antigen of HIT (Visentin et al., 1994; Greinacher et al., 1994c). Most HIT
antibodies recognize noncontiguous conformational epitopes on the PF4
molecule that are produced when at least four to eight PF4 molecules are
bound together by heparin (Horsewood et al., 1996; Newman and Chong,
1999). At PF4:heparin ratios equivalent to those prepared for use in PF4–
heparin immunoassays, the PF4 protein exhibits a shift from a globular state
to a more flexible, partially folded state (Mikhailov et al., 1999), thereby po-
tentially exposing cryptic antigens. The antibodies recognize two, and prob-
ably three, distinct sites on the PF4–heparin complexes (Suh et al., 1998; Li et
al., 2002). Furthermore, the heparin molecules must be in a flexible, relatively
unconstrained state to react with PF4 in such a way that they create sites for
HIT antibody binding.
In a few cases, PF4 alone can be recognized by the HIT antibodies as
shown by the reaction of purified HIT antibodies with both PF4–heparin
complexes and PF4 in the absence of heparin (Greinacher et al., 1994c; New-
man and Chong, 1999; Amiral et al., 2000). Here, endogenous GAGs may
take the role of heparin. However, patients have been reported with ‘‘delayed-
onset HIT’’ in which thrombocytopenia and thrombosis began several days
after heparin was stopped (Warkentin and Kelton, 2001). Sera from some of
these patients can activate platelets strongly in vitro even in the absence of
added heparin. One patient developed high levels of antibodies against PF4–
heparin complexes, together with thrombocytopenia and multiple thrombo-
ses, beginning about one week after a single injection of only 5000 U of
heparin (Warkentin and Bernstein, 2003).
In some patients with acute myocardial infarction, HIT antibodies were
apparently detected at baseline, even though the patients had never previously
been exposed to heparin (Suzuki et al., 1997). This might be due to platelet
activation connected with release of PF4 binding to endogenous GAGs.
Antibodies with cross-reactivity to PF4–heparin complexes may have been
generated against such endogenous GAG-PF4 complexes, even before the
first heparin treatment.
A retrospective analysis of patients with acute coronary syndrome
found that the presence of anti–PF4–heparin antibodies at onset was associ-
ated with a higher risk of acute myocardial infarction and death (Williams
et al., 2003). This observation requires prospective confirmation.
B. Effects of HIT Antibody-Containing Immune Complexes

Heparin-induced thrombocytopenia antibodies bind to PF4–heparin com-
plexes by their F(abV)2 domains (Horne and Alkins, 1996; Newman and
Chong, 2000), with the predominant immunoglobulin isotype being IgG
(Amiral et al., 1996b). Thus, divalent IgG binding to multimolecular PF4–
heparin complexes leads to the formation of large immune complexes
containing HIT–IgG on the platelet surface. The interaction of the HIT–
IgG Fc with the platelet FcgIIa receptors leads to cross-linking of these
receptors on the same or adjacent platelets, which triggers platelet activation
(Kelton et al, 1988, 1994; Chong et al., 1989a) (see Chap. 9). The HIT
antibody-mediated platelet activation can be inhibited by a monoclonal anti-
body specific for the FcgIIa receptor, by high concentrations of Fc fragments
derived from normal IgG, and by excess heparin saturating all binding sites
on PF4, and thus preventing the formation of multimolecular complexes
(Greinacher et al., 1994b; Visentin et al., 1994).
Besides these effects on platelets, polyclonal HIT antibodies bind to
endothelial cells (Cines et al., 1987; Visentin et al., 1994). The most convincing
evidence demonstrating that these antibodies are the same ones that cause
platelet activation was provided by classic adsorption–elution experiments
(Greinacher et al., 1994c). Purified IgG obtained from sera of HIT patients
gave positive reactions in both activation (serotonin release) and antigen
(anti-PF4–heparin) assays. This IgG fraction was then adsorbed using
cultured endothelial cells and, after extensive washing, the cells were eluted.
The eluate again tested positive in both activation and antigen assays. Thus,
these experiments showed that the antibodies recognize the same epitope on
platelets, endothelial cells, and PF4–heparin complexes coated onto a micro-
titer plate. It appears most likely that the epitope on endothelial cells com-
prises surface GAGs (Cines et al., 1987; Greinacher et al., 1994c; Visentin et
al., 1994). Endothelial cell activation by HIT antibodies can be inhibited by
excess heparin, but not by anti-FcgIIa receptor monoclonal antibodies.
In addition to platelet and endothelial cell activation, there is concom-
itant activation of coagulation, as shown by marked elevations in thrombin–
AT complex levels (Warkentin et al., 1997; Warkentin, 1998; Greinacher et
al., 2000). The simultaneous activation of platelets, endothelium, and coag-
ulation factors could explain the development of thrombocytopenia com-
bined with thrombosis or disseminated intravascular coagulation in patients
with HIT.
C. Importance of HIT Antibodies in Clinical HIT

HIT antibodies occur commonly in heparin-treated patients. However, as
many patients develop neither thrombocytopenia nor thrombosis (Amiral
et al., 1996a; Kappers-Klunne et al., 1997; Arepally et al., 1997;), it is evident

that pathogenicity requires additional factors. Two possible factors are high
titers of HIT antibodies (Suh et al., 1997), as well as optimal (equimolar) con-
centrations of heparin and PF4 in the blood circulation, such that formation
of macromolecular PF4–heparin antigen complexes is permitted (Horne and
Alkins, 1996; Horne and Hutchison, 1998). Thus, during low-dose heparin
prophylaxis in a setting of minimal platelet activation, clinical HIT may occur
less often than in a patient receiving high heparin doses together with ac-
tivated platelets (Fondu, 1995). In accordance with this working hypothesis,
HIT antibodies are most frequently induced by UFH in patients following
cardiopulmonary bypass surgery (f50%), followed by patients undergoing
major orthopedic surgery (f15%), and least frequently in medical patients
(f3%) (see Chap. 4).
Further factors favoring the development of clinical HIT are prethrom-
botic or inflammatory situations (e.g., open heart surgery) (Visentin et al.,
1996), greater susceptibility of the platelets to activation by HIT antibodies
(Salem and van der Weyden, 1983), perhaps mediated by differences in PF4
binding to platelets (Capitanio et al., 1985), and increased expression of
FcgIIa receptors (Chong et al., 1993). Also, the polymorphism of the FcgIIa
receptor at position Arg–His131 seems to be associated with a predisposition
to HIT (Carlsson et al., 1998). Consequently, although HIT antibodies play
an important role in the pathogenesis of clinical HIT, they are not the only
factor responsible for its clinical manifestation. Thus, although monitoring
for HIT antibodies should identify patients at risk for HIT (Elalamy et al.,
1996), it appears unlikely that this would lead to improved clinical outcomes
versus simply monitoring the platelet count to make an early diagnosis of
HIT.
IV. CROSS-REACTIVITY OF HIT ANTIBODIES WITH OTHER

SULFATED CARBOHYDRATES
A. Interactions with Low Molecular Weight Heparins
Generation of the HIT antigen depends not only on the concentration, but
also on the chain length of heparin. LMWH preparations (Table 2) have
reduced affinity for platelets, endothelial cells, and plasma proteins, such as
PF4 (Horne, 1993; O’Brien et al., 1985; Turpie, 1996). Accordingly, LMWH is
less likely to form multimolecular complexes with PF4 (Greinacher et al.,
1993); hence, they may induce an immune response less often than UFH. This
is corroborated by a prospective study in which patients receiving LMWH
after hip replacement surgery had a lower frequency of HIT antibody
formation than patients receiving UFH (Warkentin et al., 1995).
206
Table 2 Characteristics of Commercial LMWHs
INN (Brand name) Degradation method Mean MW (kDa) Anti-Xa (U/mg) Anti-Xa: Anti-IIa ratioa
Ardeparin sodium Peroxidation at elevated temperature 4.0–6.0 120 F 25 1.7–2.4

(Normiflo)
Bemiparin sodiumb Basic degradation in a nonaqueous 3.6 80–90 8.1
(Hibor) media and fractionation
Certoparin sodium Hydrolysis with isoamylnitrite 4.2–6.2 80–120 1.5–2.5
(Mono-Embolex NM)
Dalteparin sodiumc Hydrolysis with HNO2 5.6–6.4 110–120 1.9–3.2
(Fragmin)
Enoxaparin sodiumc Benzylation and alkaline 3.5–5.5 95–125 3.3–5.3
(Clexane, Lovenox) h-elimination
Nadroparin sodiumc Hydrolysis with HNO2 3.6–5.0 95–135 2.5–4.0
(Fraxiparin) and fractionation
Parnaparin sodiumc Radical-catalyzed degradation 4.0–6.0 75–110 1.5–3.0
(Fluxum) with H2O2 and Cu salts
Reviparin sodium Hydrolysis with HNO2 3.5–4.5 105 3.6–6.3
(Clivarin)
Tinzaparin sodiumc Enzymatic (heparinylase) 5.6–7.5 70–120 1.5–2.5
(Innohep) h-elimination
INN, International Nonproprietary Name.
a
Ratio of anti-FXa activity (U/mg) to antithrombin activity (U/mg).
b
Bemiparin is the first example of the second-generation LMWHs, which are defined to have a mean MW <4.0 kDa, a proportion of fragments
>6.0 kDa<15%, and an aXa: aIIa ratio >4:1.
c
From Monograph in European Pharmacopoeia, 4th ed.
Alban and Greinacher
Despite its lower immunogenicity, LMWH exhibits nearly 100% in vi-

tro cross-reactivity to HIT antibodies using sensitive assays (Greinacher et al.,
1994a,b; Warkentin et al., 1995; Amiral et al., 1996b; Amiral, 1997). The small
variations found with different LMWH preparations are probably based on
their individual structural parameters, such as DS and MW (Fareed et al.,
1988). Homogeneous heparin fragments containing 20, 18, 16, 14, and 12
carbohydrate residues form multimolecular complexes recognized by the
antibodies; fragments containing 10 residues induce antigen formation only
weakly; fragments containing 8 and 6 residues are less reactive (Amiral et al.,
1995; Greinacher et al., 1995) or nonreactive (Visentin et al., 2001). Small
heparin molecules may bind to PF4 (Bock et al., 1980; Denton et al., 1983;
Greinacher et al., 1995), but only large molecules are able to bridge four to
eight PF4 molecules (one to two PF4 tetramers), thus producing the non-
contiguous conformational epitopes recognized by most HIT antibodies
(Horsewood et al., 1996).
B. Interactions with Other Sulfated Carbohydrates

The formation of platelet-activating immune complexes is not limited to
heparin (Greinacher et al., 1992, 1993). Various other sulfated polysaccha-
rides, and even polyvinylsulfonate, bind PF4 to form antigen complexes
recognized by HIT antibodies. This cross-reaction depends on their structure,
especially on their DS and MW (Greinacher et al., 1992, 1995; Kelton et al.,
1994; Amiral et al., 1995). In vitro assays demonstrate that pentosan poly-
sulfate, dextran sulfate, as well as a highly sulfated chondroitin sulfate can
substitute for heparin. In contrast, neither dextran, dermatan sulfate, de-
N-sulfated heparin, sulfated glucosamine (Weimann et al., 2001), nor the
AT-binding pentasaccharide react in these assays. Accordingly, pentosan
polysulfate and highly sulfated chondroitin sulfate have induced thrombo-
cytopenia and thrombosis in vivo (Greinacher et al., 1993; Tardy et al., 1994)
(see Chap. 4). The corresponding antibodies can be detected by conventional
PF4-heparin enzyme-linked immunosorbent assay (PF4-H ELISA), demon-
strating the cross-reactivity with heparin (Gironell et al., 1996).
C. Relation Between the Anticoagulant Activity of

h-1,3-Glucan Sulfates and Their Cross-Reaction
with HIT-Associated Antibodies
To establish the structural requirements for the anticoagulant activity of
sulfated carbohydrates, as well as for the development of platelet-activating
immune complexes in the presence of HIT antibodies, we synthesized struc-
turally well-defined sulfated polysaccharides (Greinacher et al., 1995). The

resulting h-1,3-glucan sulfates (GluS) varied in their DS, MW, sulfation pat-
tern, and chemically introduced glycosidic side chains (Fig. 3). Although these
heparinoids differ structurally from heparin, they exhibit structure-dependent
anticoagulant as well as antithrombotic activities (Alban et al., 1995; Franz
and Alban, 1995). They also induce platelet activation in the presence of HIT
antibodies (Greinacher et al., 1995). Therefore, neither uronic acids, amino
groups, nor the a-1,4- or h-1,4-glycosidic linkages found in heparin are es-
sential for these biological properties.
An increase in the DS results in improved anticoagulant activity and,
after binding to PF4, an increased formation of HIT antibody-binding sites.
The MW is a second important structural parameter for anticoagulant po-
tency of a sulfated polysaccharide, as well as its capacity to cause platelet
activation in the presence of HIT antibodies. Fractions with hydrodynamic
volumes between 38 and 60 kDa showed the most prominent effects (Alban
and Franz, 1994a; Greinacher et al., 1995) (the hydrodynamic volumes were
determined by gel permeation chromatography using neutral pullulans as
MW standards; because these have lower hydrodynamic volumes owing to
the missing sulfate groups, the measured hydrodynamic volumes are higher
than the real MW; e.g., UFH had a mean hydrodynamic volume of 30 kDa).
Therefore, this MW range seems to represent the optimal chain length both
for the interaction with proteins involved in the coagulation cascade as well as
with PF4 to form HIT antigens. Beyond the optimal chain length, higher
Figure 3 Repeating unit of h-1,3-glucan sulfates: The primary OH group in po-

sition 6(**) is preferentially sulfated. Glycosidic-branched h-1,3-glucan sulfates are
substituted by a glucose, rhamnose, or arabinose unit, respectively, in position 6.
concentrations are required to form multimolecular PF4–GluS complexes

(Greinacher et al., 1995).
Compared with linear GluS having similar DS and MW, glycosidic-
branched products generally exhibit higher anticoagulant activity than the
respective linear derivatives (Alban, 1993, 1997). Glycosidic substitution
changes the three-dimensional structure of the polysaccharide chain, enhanc-
ing its flexibility and improving the interaction with proteins (Kindness et al.,
1980). As the side chains are more accessible to sulfation, they represent
clusters of negative charges (Alban and Franz, 1994b), facilitating binding
to PF4, which results in an increased cross-reactivity with HIT antibodies.
V. IMPLICATIONS FOR THE DEVELOPMENT OF

CARBOHYDRATE-BASED HEPARIN ALTERNATIVES
A. Structural Requirements of Carbohydrate-Based Heparin
Alternatives
A carbohydrate-based antithrombotic drug with a reduced risk of inducing
HIT antigen(s) should meet the following criteria (Greinacher et al., 1995):
1. The molecule should not be branched to reduce its flexibility and
to minimize charge clusters.
2. Its DS should be lower than 1.0 per monosaccharide, if its chain
length exceeds 10 monosaccharides.
3. Its MW should be lower than 2.4 kDa (about seven monosac-
charides), if its DS is between 1.0 and 1.8.
4. If the MW is higher than 2.4 kDa and the DS higher than 1.0, then
at least the therapeutic concentration must be lower than that
exhibiting cross-reactivity with HIT antibodies.
B. Danaparoid
Danaparoid sodium (Orgaran) is an alternative anticoagulant that is effective
for treating patients with HIT (see Chap. 14). This heparinoid consists of a
depolymerized mixture of GAGs extracted from porcine intestinal mucosa,
with a mean MW of 6 kDa. Its components are approximately 80% low
molecular weight heparan sulfate, 10% dermatan sulfate, 5% chondroitin
sulfate, and a small proportion of heparan sulfate (4%) with high affinity for
AT (Meuleman, 1992). Apart from the minor AT-binding heparan sulfate
component, the constituents of danaparoid have a DS per monosaccharide
between 0.5 and 0.7, as well as a low MW. Thus, the two important require-
ments to form multimolecular complexes with PF4 are not met. This is con-
sistent with the low cross-reactivity rate of danaparoid (about 10%) (Wilde
and Markham, 1997) (see Chaps. 11 and 14). As danaparoid inhibits platelet
activation by HIT antibodies even in the presence of heparin (Chong et al.,
1989b), it is possible that the GAG mixture binds to PF4 without producing
the antigen. Consequently, less PF4 is available for the small amount of
higher-sulfated heparan sulfate molecules responsible for AT binding and,
presumably, PF4 binding resulting in cross-reactivity with HIT antibodies
C. Pentasaccharides
Within the scope of developing new carbohydrate-based antithrombotics,
fondaparinux, a fully synthetic, chemically defined pentasaccharide (formerly
named Org31540/SR90107A, MW = 1728 Da; DS = 1.6; 700 anti-Xa U/
mg), has been developed, which corresponds to the AT-binding site of heparin
(Petitou et al., 1997) (Fig. 4). By its highly specific binding to AT, fondapa-
rinux selectively inhibits factor Xa and thus prevents thrombin generation
(Bauer et al., 2002). In four phase III clinical trials evaluating the prevention
of venous thromboembolism after major orthopedic surgery (>7300 patients),
fondaparinux showed superiority over the LMWH enoxaparin without
increasing clinically important bleeding (Turpie et al., 2001; Eriksson et al.,
2001; Bauer et al., 2001, Lassen et al., 2002; Turpie et al., 2002a,b) and has
recently been approved for this indication. At present, fondaparinux is under
further investigation for antithrombotic prophylaxis in other clinical settings,
as well as for treatment of deep vein thrombosis, pulmonary embolism, and
Figure 4 Chemical structure of the synthetically produced pentasaccharide Org

31540/SR90107A (MW = 1728 kDa; DS = 1.6; 864 anti-Xa U/mg), with eight sulfate
groups corresponding to the natural antithrombin-binding site.
acute coronary syndromes (The Rembrandt Investigators, 2000; Coussement

et al., 2001; Eriksson et al., 2003).
As expected, owing to the structure–activity relations previously dis-
cussed, fondaparinux did not cross-react with HIT antibodies in any concen-
tration tested, either in the PF4–H-ELISA or in the serotonin release assay
(Amiral et al., 1997; Greinacher et al., 1995; Ahmad et al., 1999). Immune
thrombocytopenia attributable to fondaparinux has not been observed in any
of the clinical studies. Of interest, a few sera from patients treated with
fondaparinux have tested positive in a PF4-dependent ELISA and have
caused platelet activation in vitro in the presence of added heparin, although
no cross-reactivity with fondaparinux itself could be shown (Warkentin et al.,
2003). This paradox is not understood.
In contrast to fondaparinux, we have observed in our laboratory that a
more highly sulfated pentasaccharide, Org 32701 (MW = 1991 Da; DS =
2.0) (Fig. 5) (Herbert et al., 1996), induces platelet activation in the presence of
HIT antibodies. This demonstrates that certain highly sulfated oligosacchar-
ides are indeed able to bind to PF4 and thus form the HIT neoantigen. But
whether such a highly sulfated pentasaccharide itself could induce clinical
HIT cannot yet be answered.
D. Specifically Designed Oligosaccharides

Pentasaccharides such as fondaparinux or the long-acting idraparinux (Her-
bert et al., 1998) have minimal, if any, undesirable interactions with blood and
vessel components, but their anticoagulant activity is limited to AT-mediated
Figure 5 Chemical structure of the synthetically produced pentasaccharide, Org

32701 (MW = 1991 kDa; DS = 2; 1150 anti-Xa U/mg), with a higher degree of
sulfation (ten sulfate groups) than the natural antithrombin-binding site.
FXa inhibition. Additional thrombin inhibitory properties might further

improve the anticoagulant efficacy of heparin-related oligosaccharides. Un-
fortunately, as with heparin, lengthening the sulfated oligosaccharide chain
increases nonspecific binding that could have undesirable effects, such as
binding to PF4 and associated risk of HIT. Thus, Petitou and coworkers
(1999) synthesized ‘‘heparin mimetics’’ that inhibited thrombin, but failed
to bind other proteins, particularly PF4. The most promising structure is
the hexadecasaccharide SR123781A, which is undergoing phase I evaluation
(Herbert et al., 2001). It is obtained from glucose through a convergent syn-
thesis and consists of an AT-binding pentasaccharide sequence linked to a
thrombin-binding domain via a neutral methylated hexasaccharide ‘‘spacer.’’
It specifically catalyzes the AT-mediated inhibition of FXa (IC50 = 77 F 5
ng/mL, 297 F 13 U/mg) and thrombin (IC50 = 4.0 F 0.5 ng/mL, 150 F 30 U/
mg), without effect on heparin cofactor II and without binding to PF4. Com-
pared with UFH and LMWH in animal studies, SR123781A exhibited a
highly favorable antithrombotic:bleeding ratio. This compound did not ac-
tivate platelets in the presence of plasma from HIT patients, suggesting that it
will not induce the HIT antigen. As the methylated spacer substitutes for the
sulfated carbohydrates, the minimally required chain length of eight sulfated
monosaccharides required for binding to PF4 (Maccarana and Lindahl, 1993)
is not present.
E. Conclusions
From experiments with well-defined GluS, the various structural require-
ments for a sulfated carbohydrate to form the HIT antigen have become clear.
Given this detailed knowledge, at least three carbohydrate-based anticoagu-
lant options can be proposed that should have a negligible risk for inducing
clinical HIT:
1. Mixtures of GAGs consisting predominantly of low-sulfated car-
bohydrates with correspondingly limited capacity to form anti-
genic complexes with PF4: A prototype of such an anticoagulant is
danaparoid.
2. Oligosaccharides with antithrombotic activity similar to the AT-
binding pentasaccharide: One such agent appears promising: fon-
daparinux did not cause HIT in more than 4000 patients treated
after orthopedic surgery.
3. GAGs with highly sulfated, but short, regions that are connected
by nonsulfated ‘‘spacers’’: Hereby, the thrombin-binding site and
the AT-specific pentasaccharide can be expressed in a single mole-
cule without reaching the critical length of a sulfated chain critical
for HIT antigen formation (Petitou et al., 1999).
The increasing use of LMWH already seems to have reduced the incidence of
HIT. We propose that the problem of HIT can be avoided further by using
anticoagulants meeting the foregoing outlined criteria in our treatment
arsenal.
ACKNOWLEDGMENTS
Part of this work was supported by Bayerischer Habilitations-Förderpreis

1996/Hans Zehetmair-Preis (S.A.) and Deutsche Forschungsgemeinschaft,
DFG Gr 1096/2-1 and Gr 1096/2-2 (A.G.).
REFERENCES
Ahmad S, Jeske WP, Walenga JM, Hoppensteadt DA, Wood JJ, Herbert JM,
Messmore HL, Fareed J. Synthetic pentasaccharides do not cause platelet ac-
tivation by antiheparin-platelet factor 4 antibodies. Clin Appl Thromb Hemost
1999; 5:259–266.
Alban S. Synthese und physiologische Testung neuartiger Heparinoide. Ph.D dis-
sertation, University of Regensburg, Germany, 1993.
Alban S. Carbohydrates with anticoagulant and antithrombotic properties. In: Wit-
czak ZJ, Nieforth KA, eds. Carbohydrates in Drug Design. New York: Marcel
Dekker, 1997:209–276.
Alban S, Franz G. Anticoagulant activity of curdlan sulfates in dependence on their
molecular weight. Pure Appl Chem 1994a; 66:2403–2406.
Alban S, Franz G. Gas liquid chromatography–mass spectrometry analysis of anti-
coagulant active curdlan sulfates. Semin Thromb Hemost 1994b; 20:152–158.
Alban S, Jeske W, Welzel D, Franz G, Fareed J. Anticoagulant and antithrombotic
actions of a semisynthetic h-1,3-glucan sulfate. Thromb Res 1995; 78:201–210.
Amiral J. Le facteur 4 plaquettaire, cible des anticorps anti-héparine: application au
diagnostic biologique de la thrombopénie induite par l’éparine (TIH). Ann
Med Intern 1997; 148:142–149.
Amiral J, Bridey F, Dreyfus M, Vissac AM, Fressinaud E, Wolf M, Meyer D.
Platelet factor 4 complexed to heparin is the target for antibodies generated in
heparin-induced thrombocytopenia [letter]. Thromb Haemost 1992; 68:95–96.
Amiral J, Bridey F, Wolf M, Boyer-Neumann C, Fressinaud E, Vissac AM, Pey-
naud-Debayle E, Dreyfus M, Meyer D. Antibodies to macromolecular platelet
factor 4-heparin complexes in heparin-induced thrombocytopenia: a study of
44 cases. Thromb Haemost 1995; 73:21–28.
Amiral J, Peynaud-Debayle E, Wolf M, Bridey F, Vissac AM, Meyer D. Generation
of antibodies to heparin–PF4 complexes without thrombocytopenia in
patients treated with unfractionated or low-molecular-weight heparin. Am J
Hematol 1996a; 52:90–95.
Amiral J, Wolf M, Fischer A, Boyer-Neumann C, Vissac A, Meyer D. Pathogenicity
of IgA and/or IgM antibodies to heparin–PF4 complexes in patients with hep-

arin-induced thrombocytopenia. Br J Haematol 1996b; 92:954–959.
Amiral J, Lormeau JC, Marfaing-Koka A, Vissac AM, Wolf M, Boyer-Neumann C,
Tardy B, Herbert JM, Meyer D. Absence of cross-reactivity of SR90107A/
ORG31540 pentasaccharide with antibodies to heparin–PF4 complexes devel-
oped in heparin-induced thrombocytopenia. Blood Coagul Fibrinolysis 1997;
8:114–117.
Amiral J, Pouplard C, Vissac AM, Walenga JM, Jeske W, Gruel Y. Affinity pu-
rification of heparin-dependent antibodies to platelet factor 4 developed in
heparin-induced thrombocytopenia: biological characteristics and effects on
platelet activation. Br J Haematol 2000; 109:336–341.
Arepally G, McKenzie SE, Jiang XM, Poncz M, Cines DB. FcgRIIA H/R131 poly-
morphism, subclass-specific IgG anti-heparin/platelet factor 4 antibodies and
clinical course in patients with heparin-induced thrombocytopenia and throm-
bosis. Blood 1997; 89:370–375.
Bauer KA, Eriksson BI, Lassen MR, Turpie AG. Fondaparinux compared with
enoxaparin for the prevention of venous thromboembolism after elective ma-
jor knee surgery. N Engl J Med 2001; 345:1305–1310.
Bauer KA, Hawkins DW, Peters PC, Petitou M, Herbert JM, Van Boeckel CA,
Meuleman DG. Fondaparinux, a synthetic pentasaccharide: the first in a new
class of antithrombotic agents—the selective factor Xa inhibitors. Cardiovasc
Drug Rev 2002; 20:37–52.
Beguin S, Mardiguian J, Lindhout T, Hemker HC. The mode of action of low mo-
lecular weight heparin preparation (PK10169) and two of its major components
on thrombin generation in plasma. Thromb Haemost 1989; 61:30–34.
Bendetowicz AV, Kai H, Knebel R, Caplain H, Hemker HC, Lindhout T, Beguin S.
The effect of subcutaneous injection of unfractionated and low molecular
weight heparin on thrombin generation in platelet rich plasma a study in
human volunteers. Thromb Haemost 1994; 72:705–712.
Björk I, Olson ST, Shore JD. Molecular mechanisms of the accelerating effect of hep-
arin on the reaction between antithrombin and clotting proteinases. In: Lane
DA, Lindahl U, eds. Heparin, Chemical and Biological Properties, Clinical
Applications. London: Edward Arnold, 1989:229–255.
Bock PE, Luscombe M, Marshall SE, Pepper DS, Holbrook JJ. The multiple com-
plexes formed by the interaction of platelet factor 4 with heparin. Biochem J
1980; 191:769–776.
Campbell A, Nesheim ME, Doctor VM. Mechanism of potentiation of antithrombin
III [AT-III] inhibition by sulfated xylans. Thromb Res 1987; 47:341–352.
Carlsson LE, Santoso S, Baurichter G, Kroll H, Papenberg S, Eichler P, Westerdaal
NAC, Kiefel V, van de Winkel JGJ, Greinacher A. Heparin-induced throm-
bocytopenia: new insights into the impact of the FcgRIIa-R-H131 polymor-
phism. Blood 1998; 92:1526–1531.
Casu B. Structure and biological activity of heparin. Adv Carbohydr Chem Biochem
1985; 43:51–134.
Casu B. Heparin structure. Haemostasis 1990; 20(suppl 1):62–73.
Cella G, Scattolo N, Luzzatto G, Stevanato F, Vio C, Girolami ASO. Effects on
platelets and on the clotting system of four glycosaminoglycans extracted from
hog mucosa and one extracted from aortic intima of the calf. J Med 1986;
17:331–346.
Choay J. Structure and activity of heparin and its fragments: an overview. Semin
Thromb Hemost 1989; 15:359–364.
Chong BH, Fawaz I, Chestermann CN, Berndt MC. Heparin-induced thrombocy-
topenia: mechanism of interaction of the heparin-dependent antibody with
platelets. Br J Haematol 1989a; 73:235–240.
Chong BH, Ismail F, Cade J, Gallus AS, Gordon S, Chesterman CN. Heparin-
induced thrombocytopenia: studies with a new molecular weight heparinoid,
Org 10172. Blood 1989b; 73:1592–1596.
Chong BH, Pilgrim RL, Cooley MA, Chesterman CN. Increased expression of plate-
let IgG Fc receptors in immune heparin-induced thrombocytopenia. Blood
1993; 81:988–993.
Coussement PK, Bassand JP, Convens C, Vrolix M, Boland J, Grollier G, Michels R,
Vahanian A, Vanderheyden M, Rupprecht HJ, Van de Werf F; PENTA-
LYSE investigators. A synthetic factor-Xa inhibitor (ORG31540/SR9017A) as
an adjunct to fibrinolysis in acute myocardial infarction. The PENTALYSE
study. Eur Heart J 2001; 22:1716–1724.
Dawes J, Pumphrey CW, McLaren KM, Prowse CV, Pepper DS. The in vivo release
of human platelet factor 4 by heparin. Thromb Res 1982; 27:65–76.
Denton J, Lane DA, Thunberg L, Slater AM, Lindahl U. Binding of platelet factor 4
to heparin oligosaccharides. Biochem J 1983; 209:455–460.
Elalamy I, Potevin F, Lecrubier C, Bara L, Marie JP, Samama MM. A fatal low-
molecular-weight heparin-associated thrombocytopenia after hip surgery: pos-
sible usefulness of PF4–heparin ELISA test. Blood Coagul Fibrinolysis 1996;
7:665–671.
Eriksson BI, Bauer KA, Lassen MR, Turpie AG. Fondaparinux compared with
enoxaparin for the prevention of venous thromboembolism after hip-fracture
surgery. N Engl J Med 2001; 345:1298–1304.
Eriksson BI, Lassen MR, PENTasaccharide in HIp-FRActure Surgery Plus In-
vestigators. Duration of prophylaxis against venous thromboembolism with
fondaparinux after hip fracture surgery: a multicenter, randomized, placebo-
controlled, double-blind study. Arch Intern Med 2003; 163:1337–1342.
Fareed J, Walenga JM, Hoppensteadt D, Haun X, Racanelli A. Comparative study
on the in vitro and in vivo activities of seven low-molecular-weight heparins.
Haemostasis 1988; 18(suppl 3):3–15.
Fondu P. Heparin-associated thrombocytopenia: an update. Acta Clin Belg 1995;
50:343–357.
Franz G, Alban S. Structure–activity relationship of antithrombotic polysaccharide

derivatives. Int J Biol Macromol 1995; 17:311–314.
Gironell A, Altes A, Arboix A, Fontcuberta J, Munoz Z, Marti-Vilalta JL. Pentosan
polysulfate-induced thrombocytopenia: a case diagnosed with an ELISA test
used for heparin-induced thrombocytopenia. Ann Hematol 1996; 73:51–62.
Greinacher A, Michels I, Mueller-Eckhardt C. Heparin-associated thrombocyto-
penia: the antibody is not heparin specific. Thromb Haemost 1992; 67:545–
549.
membrane by the negative charge of highly sulphated oligosaccharides. Br J
Haematol 1993; 84:711–716.
Greinacher A, Feigl M, Mueller-Eckhardt C. Crossreactivity studies between sera of
patients with heparin-associated thrombocytopenia and a new low molecular
weight heparin, reviparin. Thromb Haemost 1994a; 72:644–645.
Greinacher A, Amiral J, Dummel V, Vissac AM, Kiefel V, Mueller-Eckhardt C.
Laboratory diagnosis of heparin-associated thrombocytopenia, comparison of
platelet aggregation test, heparin-induced platelet activation (HIPA) test, and
PF4/heparin ELISA. Transfusion 1994b; 34:381–385.
Heparin-associated thrombocytopenia: isolation of the antibody and charac-
terization of a multimolecular PF4–heparin complex as the major antigen.
Thromb Haemost 1994c; 71:247–251.
Greinacher A, Eichler P, Lubenow N, Luz M. Heparin-induced thrombocytopenia
with thromboembolic complications: meta-analysis of two prospective trials to
assess the value of parenteral treatment with lepirudin and its therapeutic
aPTT range. Blood 2000; 96:846–851.
Handin RI, Cohen HJ. Purification and binding properties of human platelet factor
four. J Biol Chem 1976; 251:4273–4282.
Herbert JM, Heráult JP, Bernat A, van Amsterdam RG, Vogel GM, Lormeau JC,
Petitou M, Meuleman DG. Biochemical and pharmacological properties of
SANORG 32701. Comparison with the ‘‘synthetic pentasaccharide’’ (SR
90107/ORG 31540) and standard heparin. Circ Res 1996; 79:590–600.
Herbert JM, Herault JP, Bernat A, van Amsterdam RG, Lormeau JC, Petitou M,
van Boeckel C, Hoffman P, Meuleman DG. Biochemical and pharmacological
properties of SANORG 34006, a potent and long-acting synthetic pentasac-
charide. Blood 1998; 91:4197–4205.
Herbert JM, Herault JP, Bernat A, Savi P, Schaeffer P, Driguez PA, Duchaussoy P,
Petitou M. SR123781A, a synthetic heparin mimetic. Thromb Haemost 2001;

85:852–860.
Hirsh J. Heparin. N Engl J Med 1991; 324:1565–1574.
Granger C, Ohman EM, Dalen JE. Heparin and low-molecular-weight hepa-
rin: mechanism of action, pharmacokinetics, dosing, monitoring, efficacy, and
Horne MK III. The effect of secreted heparin-binding proteins on heparin binding to
Horne MK III, Alkins BR. Platelet binding of IgG from patients with heparin-
induced thrombocytopenia. J Lab Clin Med 1996; 127:435–442.
Horsewood P, Warkentin TE, Hayward CP, Kelton JG. The epitope specificity of
Kaplan KL, Niewiarowski S. Nomenclature of secreted platelet proteins—report of
the Working Party on Secreted Platelet Proteins of the Subcommittee on Plate-
lets. Thromb Haemost 1985; 53:282–284.
Kappers-Klunne MC, Boon DM, Hop WC, Michiels JJ, Stibbe J, van der Zwaan C,
Koudstaal PJ, van Vliet HH. Heparin-induced thrombocytopenia and throm-
bosis: a prospective analysis of the incidence in patients with heart and cere-
brovascular diseases. Br J Haematol 1997; 96:442–446.
925–930.
Kelton JG, Smith JW, Warkentin TE, Hayward CP, Denomme GA, Horsewood P.
Kindness G, Long WF, Williamson FB. Anticoagulant effects of sulphated poly-
saccharides in normal and antithrombin III-deficient plasmas. Br J Pharmacol
1980; 69:675–677.
Klener P, Kubisz PSO. Platelet heparin-neutralizing activity (platelet factor 4). Acta
Univ Carol Med (Praha) 1978; 24:79–86.
Lane DA, Denton J, Flynn AM, Thunberg L, Lindahl U. Anticoagulant activities of
heparin oligosaccharides and their neutralization by platelet factor 4. Biochem
J 1984; 218:725–732.
Lassen MR, Bauer KA, Eriksson BI, Turpie AG. Postoperative fondaparinux versus
preoperative enoxaparin for prevention of venous thromboembolism in elec-
tive hip-replacement surgery: a randomised double-blind comparison. Lancet
2002; 359:1715–1720.
1236.
Linhardt RJ, Toida T. Heparin oligosaccharides: new analogues development and

applications. In: Witczak ZJ, Nieforth KA, eds. Carbohydrates in Drug De-
sign. New York: Marcel Dekker, 1997:277–341.
Linhardt RJ, Rice KM, Kim YS, Lohse DL, Wang HM, Loganathan D. Mapping
and quantification of the major oligosaccharide components of heparin. Bio-
chem J 1988; 254:781–787.
Linhardt RJ, Ampofo SA, Fareed J, Hoppensteadt D, Mulliken JB, Folkman J.
Isolation and characterization of human heparin. Biochemistry 1992; 31:
12441–12445.
Loscalzo J, Melnick B, Handin RI. The interaction of platelet factor four and gly-
cosaminoglycans. Arch Biochem Biophys 1985; 240:446–455.
Luscher EF, Kaser-Glanzman R. Platelet heparin-neutralizing factor (platelet factor
4). Thromb Diath Haemorrh 1975; 33:66–72.
Maccarana M, Lindahl U. Mode of interaction between platelet factor 4 and hepa-
rin. Glycobiology 1993; 3:271–277.
Mayo KH, Roongta V, Ilyina E, Milius R, Barker S, Quinlan C, La Rosa G, Daly
TJ. NMR solution structure of the 32-kDa platelet factor 4 ELR-motif N-
terminal chimera: a symmetric tetramer. Biochemistry 1995; 34:11399–11409.
Meuleman DG. Orgaran (Org 10172): its pharmacological profile in experimental
models. Haemostasis 1992; 22:58–65.
Mikhailov D, Young HC, Linhardt RJ, Mayo KH. Heparin dodecasaccharide bind-
ing to platelet factor-4 and growth-related protein-alpha: induction of a par-
tially folded state and implications for heparin-induced thrombocytopenia. J
Biol Chem 1999; 274:25317–25329.
Moore S, Pepper DS, Cash JD. Platelet antiheparin activity. The isolation and char-
acterisation of platelet factor 4 released from thrombin-aggregated washed
human platelets and its dissociation into subunits and the isolation of mem-
brane-bound antiheparin activity. Biochim Biophys Acta 1975; 379:370–384.
dynamic binding of purified anti-PF4–heparin antibodies to platelets and the
Niewiarowski S. Report of the Working Party on Platelets. Platelet factor 4 (PF4),
platelet protein with heparin neutralizing activity. Thromb Haemost 1976; 36:
273–276.
Novotny WF, Maffi T, Mehta RL, Milner PG. Identification of novel heparin-
releasable proteins, as well as the cytokines midkine and pleiotrophin, in hu-
man postheparin plasma. Arterioscler Thromb 1993; 13:1798–1805.
O’Brien JR, Etherington MD, Pashley MA. The heparin-mobilisable pool of platelet
factor 4: a comparison of intravenous and subcutaneous heparin and Kabi
heparin fragment 2165. Thromb Haemost 1985; 54:735–738.
Padilla A, Gray E, Pepper DS, Barrowcliffe TW. Inhibition of thrombin generation
by heparin and low molecular weight (LMW) heparins in the absence and
presence of platelet factor 4 (PF4). Br J Haematol 1992; 82:406–413.
Petitou M, Duchaussoy P, Jaurand G, Gourvenec F, Lederman I, Strassel JM, Barzu

T, Crepon B, Herault JP, Lormeau JC, Bernat A, Herbert JM. Synthesis and
pharmacological properties of a close analogue of an antithrombotic penta-
saccharide (SR 90107A/ORG 31540). J Med Chem 1997; 40:1600–1607.
Petitou M, Herault JP, Bernat A, Driguez PA, Duchaussoy P, Lormeau JC, Herbert
JM. Synthesis of thrombin-inhibiting heparin mimetics without side effects.
Nature 1999; 398:417–422.
Rembrandt Investigators. Treatment of proximal deep vein thrombosis with a novel
synthetic compound (SR90107A/ORG31540) with pure anti-factor Xa activ-
ity: a phase II evaluation. Circulation 2000; 102:2726–2731.
St. Charles R, Walz DA, Edwards BF. The three-dimensional structure of bovine
platelet factor 4 at 3.0-A resolution. J Biol Chem 1989; 264:2092–2099.
Salem HH, van der Weyden MB. Heparin-induced thrombocytopenia. Variable plate-
let-rich plasma reactivity to heparin-dependent platelet aggregating factor. Pa-
thology 1983; 15:297–299.
Samama MM, Bara L, Gerotziafas GT. Mechanisms for the antithrombotic activity
in man of low molecular weight heparins (LMWHs). Haemostasis 1994; 24:
105–117.
Sear CH, Poller L. Antiheparin activity of human serum and platelet factor 4.
Thromb Diath Haemorrh 1973; 30:93–105.
Stuckey JA, St. Charles R, Edwards BF. A model of the platelet factor 4 complex
with heparin. Proteins 1992; 14:277–287.
201.
factor 4. Blood 1998; 91:916–922.
Suzuki S, Koide M, Sakamoto S, Yamamoto S, Matsuo M, Fujii E, Matsuo T. Early
onset of immunological heparin-induced thrombocytopenia in acute myocar-
dial infarction. Blood Coagul Fibrinolysis 1997; 8:13–15.
Tardy PB, Tardy B, Grelac F, Reynaud J, Mismetti P, Bertrand JC, Guyotat D.
Pentosan polysulfate-induced thrombocytopenia and thrombosis. Am J Hem-
atol 1994; 88:803–808.
Turpie AGG. New therapeutic opportunities for heparins: what does low molecular
weight heparin offer? J Thromb Thrombolysis 1996; 3:145–149.
Turpie AGG, Gallus AS, Hoek JA, for the Pentasaccharide Investigators. A syn-
thetic pentasaccharide for the prevention of deep-vein thrombosis after total
hip replacement. N Engl J Med 2001; 344:619–625.
Turpie AGG, Bauer KA, Eriksson BI, Lassen MR. Postoperative fondaparinux
versus postoperative enoxaparin for prevention of venous thromboembolism
after elective hip-replacement surgery: a randomised double-blind trial. Lancet
2002a; 359:1721–1726.
Turpie AGG, Bauer KA, Eriksson BI, Lassen MR. Fondaparinux vs enoxpaparin
for the prevention of venous thromboembolism in major orthopedic surgery: a
meta-analysis of 4 randomized double-blind studies. Arch Intern Med 2002b;

162:1833–1840.
Visentin GP. Heparin-induced thrombocytopenia: molecular pathogenesis. Thromb
Haemost 1999; 82:448–456.
Visentin GP, Malik M, Cyganiak KA, Aster RH. Patients treated with unfrac-
tionated heparin during open heart surgery are at high risk to form antibodies
reactive with heparin: platelet factor 4 complexes. J Lab Clin Med 1996; 128:
376–383.
required for detection of antibodies associated with heparin-reduced thrombo-
cytopenia/thrombosis. J Lab Clin Med Zool 138:22–31.
Walsh PN. Platelets, heparin, and blood coagulation. In: Kakkar VV, Thomas DP,
eds. Heparin: Chemistry and Clinical Usage. London: Academic Press, 1976:
125–143.
Warkentin TE, Bernstein RA. Delayed-onset heparin-induced thrombocytopenia
and cerebral thrombosis after a single administration of unfractionated
heparin. N Engl J Med 2003; 348:1067–1069.
Warkentin TE, Levine MN, Hirsh J, Horsewood P, Roberts RS, Gent M, Kelton
JG. Heparin-induced thrombocytopenia in patients treated with low molecular
after elective hip or knee replacement surgery in patients receiving anti-
thrombotic prophylaxis with fondaparinux or enoxaparin [abstr]. Blood, 2003.
In press.
William RT, Damaraju LV, Mascelli MA, Barnathan ES, Califf RM, Simoons ML,
Deliargyris EN, Sane DC. Anti-platelet factor 4/heparin antibodies: an in-
dependent predictor of 30-day myocardial infraction after acute coronary
ischemic syndromes. Circulation 2003; 107:2307–2312.
Weimann G, Lubenow N, Selleng K, Eichler P, Albrecht D, Greinacher A. Glu-
cosamine sulfate does not crossreact with the antibodies of patients with he-
parin-induces thrombocytopenia. Eur J Haematol 2001; 66:195–199.
Wilde MI, Markham A. Danaparoid. A review of its pharmacology and clinical use
in the management of heparin-induced thrombocytopenia. Drugs 1997; 54:

903–924.
Young E, Prins MH, Levine MN, Hirsh J. Heparin binding to plasma proteins, an
important mechanism for heparin resistance. Thromb Haemost 1992; 67:639–
643.
Zhang X, Chen L, Bancroft DP, Lai CK, Maione TE. Crystal structure of recom-
binant human platelet factor 4. Biochemistry 1994; 33:8361–8366.
9
The Platelet Fc Receptor in
Gregory A. Denomme
University of Toronto, Canadian Blood Services,
and Mount Sinai Hospital, Toronto, Ontario, Canada
I. INTRODUCTION
Heparin-induced thrombocytopenia (HIT) is a unique immune-mediated

disorder. HIT is common, occurring in as many as 5% of certain patient
populations. Affected patients often develop the paradox of thrombosis, but
not bleeding, despite having thrombocytopenia. One possible reason for this
unique clinical profile is the central role of the platelet Fcg receptor (FcgR) IIa
in mediating platelet activation in HIT. Indirect evidence suggesting a cru-
cial role for platelet activation in the pathogenesis of HIT is the observation
that thrombocytopenia caused by HIT antibodies is strongly associated with
thrombosis, whereas formation of HIT antibodies without thrombocytope-
nia is not (Warkentin et al., 1995).
It has been known for several years that HIT results from a predom-
inant IgG immune response to antigenic determinants involving heparin
bound to the surface of the platelet membrane (Green et al., 1978). Thus,
the pathogenesis of HIT resembles a type II immune reaction, i.e., a cytotoxic
antibody response (Roitt et al., 1985). However, typical features of a type II
immune response, such as phagocytosis, killer cell activity, or complement-
mediated lysis, do not seem to predominate in HIT. Instead, thrombocytope-
nia results primarily from IgG binding to platelet factor 4–heparin (PF4–H)
complexes on the platelet surface. The HIT–IgG within these large multimo-
lecular immune complexes interacts with the platelet FcgRIIa; cross-linking
of the receptors causes platelet activation, aggregation, and granule release
223
224 Denomme
(Chong et al., 1981). Furthermore, HIT antibodies activate endothelium in

vitro by interaction with PF4–heparan sulfate complexes (Cines et al., 1987;
Greinacher et al., 1994a; Visentin et al., 1994). However, unlike platelets,
human endothelium (with the exception of placental villous endothelial cells
and a subset of endothelial cells found in the superficial dermal vascular
plexus) do not express any Fcg receptors, either constitutively or in the setting
of immune complex diseases (Sedmak et al., 1991; Groger et al., 1996). Thus,
platelet activation and endothelial activation in HIT probably arise from
fundamentally distinct processes. Other effects of HIT include the formation
of platelet-leukocyte aggregates and the release of tissue factor from mono-
cytes (Khairy et al., 2001; Pouplard et al., 2001).
One of the most important unanswered questions in the pathophysiol-
ogy of this disorder is an explanation for why only a few patients who develop
HIT antibodies become thrombocytopenic. This problem has led investiga-
tors to study the role of FcgRIIa in explaining, at least partly, the heterog-
enous clinical sequelae among patients with HIT. This chapter will (a) review
the structure and function of the platelet FcgRIIa; (b) describe the mechanism
of HIT antibody-induced platelet activation by FcgRIIa; and (c) summarize
the studies that have attempted to identify a role for the FcgRIIa in modifying
clinical manifestations of HIT.
gRIIa STRUCTURE, DISTRIBUTION,

II. PLATELET Fcg
AND FUNCTION
The platelet FcgRIIa is a member of a family of structurally related glyco-

proteins, many of which are expressed on hematopoietic cells (Table 1).
Twelve different transcripts have been reported, derived from eight distinct
genes and grouped into three different classes: I, II, and III (for review see van
de Winkel and Capel, 1993; Rascu et al., 1997; Gessner et al., 1998). Allelic
polymorphic variants add yet another level of diversity for FcgRIIa,
FcgRIIIa, and FcgRIIIb. The genomic organization of the FcgR genes on
chromosome 1q23 was resolved by Su and coworkers (2002). The multigenic
region is approximately 1 mb in size and is in the following gene order and
orientation: cen—FCGR2A (5V-3V)—FCGR3A (3V-5V)—FCGR2C (3V-5V)—
FCGR3B (3V-5V)—FCGR2B (5V-3V)—tel.
The affinity for IgG varies among these isoforms and polymorphic
variants. Most notably, the FcgRIIa–His131 allele has a significantly higher
affinity for human IgG2 than FcgRIIa–Arg131 (Warmerdam et al., 1991). Fur-
thermore, FcgRIIIa–Val158 and FcgRIIIb–NA1 bind nearly twice as much
IgG as FcgRIIIa–Phe158 and FcgRIIIb–NA2, respectively (Salmon et al.,
1990; Koene et al., 1997). One allelic variant of FcgRIIc has a nonsense
codon, which results in the absence of expression of the glycoprotein if
Table 1 The Family of Fcg Receptors: Molecular and Structural Characteristics and Tissue Distribution
FcgRI (CD64) FcgRII (CD32) FcgRIII (CD16)
Genes IA, IB, IC IIA, IIB, IIC IIIA, IIIB

Functional variantsa None IIa: Gln/Lys127 IIIa: Leu/Arg/His48
IIa: Arg/His131 IIIa: Phe/Val158
IIc: Gln/Stop codon13 IIIb: NA1/NA2
IIIb: Ala/Asp60 (SH)
RNA transcriptsb IA1 IIA1, IIA2 IIIA
IB1, IB2 IIB1, IIB2, IIB3 IIIB
IC IIC
Glycoprotein expressedc Ia IIa1, sIIa2 IIIa
IIb1, sIIb2, IIb3 IIIb (GPI linked)
Platelet Fc Receptor in HIT
IIc
Molecular weight (kDa) 72 40 50–80
Extracellular Ig-like domains 3 2 2
Intracellular tyrosine motif d None ITAM(IIa, IIc) None
ITIM (IIb)
Noncovalent-associated subunits g-Chain None h-Chain, g-chain, ~-chain
Affinity constant 108M1 <107M1 IIIa: 3 107M1
IIIb: <107M1
IgG subclass aviditye 3=1 > 4{2 IIa-Arg131 3 > 1 { 2 > 4 IIIa-Val158 1 > 3 { 2,4
IIa-His131 3 > 1 = 2 > 4 IIIa-Val158 > IIIa-Phe158 for 1 and 3
IIb1 3 > 1 > 4 { 2 IIIb-NA1 > IIIb-NA2 for 1 and 3
Hematopoietic cell distribution CD34 progenitor cells, monocytes, IIa: platelets, endothelial cells, IIIa: monocytes, macrophages, NK
macrophages, dendritic cells monocytes, macrophages, cells, T cells
eosino-/baso-/neutrophils, IIIb: neutrophils
Langerhans/dendritic cells
IIb: B cells, monocytes
IIc: NK cells
a
Allelic polymorphisms that show differences in IgG binding; NA1/NA2 variants have multiple amino acid differences (Ory et al., 1989); SH+ individuals
(Bux et al., 1997) carry three copies of FcgRIIIB (Koene et al., 1998); lack of expression of IIc is due to a nonsense mutation (Metes et al., 1998).
b
Multiple mRNA transcripts from FcgRIB, IIA, and IIB, are the result of alternative splicing of primary transcripts.
c
Soluble forms of FcgRIIa and IIb (sIIa, sIIb) are devoid of the hydrophobic transmembrane exon; GPI, glycosylphosphatidylinositol.
d
ITAM, immunoreceptor tyrosine-based activation motif; ITIM, immunoreceptor tyrosine-based inhibition motif.
225
e
Numbers represent the relative order of IgG subclass binding to variants of FcgRIIa (Warmerdam et al., 1990), FcgRIIIa (Koene et al., 1997; Wu
et al., 1997), and FcgRIIIb (Salmon et al., 1990; Bredius et al., 1994b).
226 Denomme
inherited in the homozygous state (Metes et al., 1998). When cross-linked,

each isoform participates in biological activities through distinct signal trans-
duction pathways that affect cell functions, including antigen presentation,
immune complex clearance, phagocytosis and the oxidative burst, release of
various cytokines and intracellular granular mediators, antibody-dependent
cellular cytotoxicity (ADCC), and negative down-regulation of antibody pro-
duction or phagocytosis.
Only FcgRIIa is expressed on platelets (Rosenfeld et al., 1985; Kelton
et al., 1987). The receptor is a single a-chain, 40-kDa glycoprotein, with an
extracellular region consisting of two immunoglobulin-like, disulfide-linked
domains responsible for ligand binding, a transmembrane region, and an in-
tracellular domain that incorporates an immunoreceptor tyrosine-based acti-
vation motif (ITAM) essential for intracellular signal transduction (Qiu et al.,
1990). The gene comprises seven exons. A soluble form of the receptor is pro-
duced by alternative splicing of primary RNA transcripts to exclude exon five
containing the transmembrane region (Rappaport et al., 1993). The nucleo-
tide region encoding the ITAM is unique in that the two tyrosine motifs are
separated by 12 amino acids, rather than the usual 7 (Brooks et al., 1989).
At the amino acid level, the extracellular domain of FcgRIIa share 96%
identity with FcgRIIb and FcgRIIc (Brooks et al., 1989). A G ! A poly-
morphism at nucleotide position 519 of the cDNA (position 512 in Genbank
Accession M90724) is responsible for the Arg/His131 allelic functional vari-
ants (Clark et al., 1989). An additional polymorphism, an A ! G at nu-
cleotide 207 of the cDNA, results in a Gln–Trp27 substitution in the mature
polypeptide (Warmerdam et al., 1991). However, the Arg–His131 position is
near or within the binding region for IgG Fc (Hulett et al., 1995), and it is this
polymorphism that is associated with the affinity differences for human IgG2
(Warmerdam et al., 1991). More recently, another polymorphism proximal to
Arg131 that affects IgG2 binding has been found in a single healthy individual
(Norris et al., 1998): a lysine substitution for glutamine at position 127 dem-
onstrated a significant increase in FcgR-mediated phagocytosis in this homo-
zygous FcgRIIa–Arg131 individual.
FcgRIIa has low affinity for IgG (<107 M1) and interacts only with
multivalent antigen–antibody complexes (Warmerdam et al., 1991; Parren
et al., 1992). The copy number of FcgRIIa expressed on normal resting
platelets varies among healthy individuals, but is stable for a given individual
(Rosenfeld et al., 1987). Studies using Scatchard analysis report roughly a
threefold variation among individuals, with the number of binding sites rang-
ing from 600 to 1500 molecules per platelet when assayed using intact murine
monoclonal antibody IV.3 (McCrae et al., 1990), and 1500 to more than 4500
for the monovalent (Fab) preparation of IV.3 (Tomiyama et al., 1992; Brandt
et al., 1995). There are no major differences in FcgRIIa number between
Platelet Fc Receptor in HIT 227
males and female, among platelets from persons of different ages, or among
platelets representing the three possible genotypic classes of the FcgRIIa-Arg/
His131 allelic variants (Brandt et al., 1995).
III. IMMUNOGLOBULIN G AGONISTS AND PLATELET

ACTIVATION
gIIa Receptor-Mediated Platelet Activation
A. Fcg
Murine monoclonal antibodies, usually of the IgGl subclass, against CD9
were among the first studied for their platelet-activating properties. Subse-
quently, it was determined that monoclonal antibodies to glycoprotein (GP)
IIb/IIIa, h2-microglobulin, GP IV (CD36), and other selected antigens can
activate platelets (for review, see Rubinstein et al., 1995). In each instance,
platelet activation occurs by a consistent mechanism: first, there is Fab-
mediated binding to surface-expressed platelet antigens, then the Fc portion
of the antibodies interacts with platelet FcgRIIa. Evidence for FcgRIIa-
dependency includes the inhibition of platelet activation by a murine mono-
clonal anti-FcgRIIa antibody (IV.3). However, some platelet GPs (e.g., GP
Ib) do not support activation by monoclonal antibodies; others support ac-
tivation despite their usual sequestered location within platelets (e.g., mul-
timerin), and still other GPs (e.g., GP IIb/IIIa) support activation by only
certain monoclonal antibodies (Horsewood et al., 1991). This suggests that
specific factors, such as target protein membrane mobility and localization of
the target epitope, that permit formation of multimolecular GP antigen–IgG–
FcgRIIa complexes, are crucial for platelet activation mediated by FcgRIIa
clustering. These murine IgG Fc moieties within the platelet surface immune
complexes can interact with either FcgRIIa on the same platelet (intraplate-
let activation) or with FcgIIa receptors located on other platelets in close
proximity (interplatelet activation) (Anderson et al., 1991; Horsewood et al.,
1991).
Complexed human IgG is also a potent stimulator of platelet activation.
Karas and coworkers (1982) showed that trimeric human IgG and larger
immune complexes had significant affinity for platelet FcgRIIa. King et al.
(1990) showed that trimeric IgG molecules are necessary for platelet activa-
tion. Furthermore, heat-aggregated IgG also is a potent agonist for platelet
activation (Warkentin et al., 1994; Warkentin and Sheppard, 1999), as are
streptokinase–antistreptokinase antibodies (Lebrazi et al., 1995) and PF4–H–
containing immune complexes (Greinacher et al., 1994a).
HIT-IgG causes the generation of thromboxane A2 and associated
platelet granule release (Chong et al., 1981). Indeed, several different ‘‘acti-
vation assays’’ have been developed that detect HIT antibodies by their ability
228 Denomme
to cause resting platelets to aggregate (Greinacher et al., 1991), effect granule

release (Sheridan et al., 1986), or generate platelet-derived microparticles
(Warkentin et al., 1994) (see Chap. 11).
B. Procoagulant, Platelet-Derived Microparticles

Platelet activation by various agonists leads to procoagulant alterations of the
platelet membrane. This includes loss of the usual membrane asymmetry (i.e.,
with platelet activation there is increased transbilayer movement of phospha-
tidylserine from the inner to the outer platelet membrane). This membrane
‘‘flip-flop’’ is a consequence of a calcium-dependent enzyme (‘‘scramblase’’)
that serves to undo the membrane asymmetry actively maintained in resting
platelets by other enzymes (aminophospholipid translocase and ‘‘floppase’’)
(Bevers et al., 1999). Additionally, platelet activation also leads to profound
morphological changes that correlate with procoagulant activity, including
the generation of procoagulant, platelet-derived ‘‘microparticles’’ (Sims et al.,
1989).
Serum and purified IgG from patients with HIT, as well as other ‘‘IgG
agonists’’ (immune complexes and murine platelet-activating monoclonal
IgG), also generate platelet-derived microparticles via the platelet FcgIIa
receptors; in contrast, quinine- and quinidine-dependent sera do not produce
microparticles, even though they lead to far greater drug-dependent binding
of IgG to platelets, compared with HIT samples (Warkentin et al., 1994).
Indeed, HIT serum is superior in generating platelet-derived microparticles,
and in producing platelet procoagulant activity (as shown using a phospho-
lipid-dependent assay, the Russell’s viper venom time), than physiological
agonists such as thrombin, collagen, and ADP. Only the nonphysiological
agonist calcium ionophore produced greater numbers of microparticles and
procoagulant activity than did HIT sera (Warkentin and Sheppard, 1999).
Flow cytometry can be used to detect platelet-derived microparticles
generated by HIT antibodies or other platelet agonists (Fig. 1). This technique
has been used to develop diagnostic assays for HIT. For example, relative
quantitation of platelets and microparticles can be achieved using particle size
(‘‘forward scatter’’) and fluorescein-labeled platelet GP-specific monoclonal
antibodies (Lee et al., 1996). Tomer (1997) used labeled annexin V, a protein
that binds to phosphatidylserine expressed on activated platelets and micro-
particles, to test for HIT (see Chap. 11).
There is some uncertainty as to whether the ‘‘microparticles’’ detected
by flow cytometry represent true microparticles, or rather platelets that have
undergone considerable morphological changes during activation. Use of
high-sensitivity settings for signal thresholding on orthogonal light scatter,
Figure 1 Flow cytometric analysis of platelets activated with HIT sera, thrombin,
and calcium ionophore. Platelets (p) and microparticles (mp) were identified using
fluorescence (FL1, FITC anti-GPIba, y-axis) and size (forward scatter, x-axis). Nega-
tive controls included: (A) platelets incubated in buffer alone; (B) platelets incubated
with patients serum, which tested negative for HIT, in the presence of 0.1 U/mL
heparin; and (C) platelets incubated with HIT serum with no heparin added. Micro-
particles were generated by (D) platelets incubated with HIT serum in the presence of
0.1 U/mL heparin, as well as (E) thrombin-treated platelets (1 U/mL) and (F) calcium
ionophore-treated platelets (10 AM). The percentage of microparticles for each experi-
ment is shown in the lower left quadrant. (From Hughes et al., 2000.)
combined with fluorescence gating on platelet antigens, allows detection of

significant increases in total particle count, suggesting that at least some true
microparticles can be detected by flow cytometry (Bode and Hickerson, 2000).
However, these techniques remain to be tested with HIT sera. Nevertheless,
that true microparticles can be generated by HIT antibodies is suggested by a
recent study that used confocal microscopy and scanning/transmission
electron microscopy for their detection (Hughes et al., 2000) (Fig. 2).
230 Denomme
Figure 2 Electron microscopy of negatively stained platelets activated in situ with

HIT serum. Platelets were allowed to settle on bovine serum albumin-coated Formvar
grids and then incubated with (A) serum testing negative for HIT antibodies or heparin
(not shown), or (B) HIT serum in the presence of heparin, 0.1 U/mL. Platelets were
then fixed with 2% glutaraldehyde and negatively stained with 2% phosphotungstic
acid. Whereas unactivated platelets demonstrated round or discoid shapes (see A),
platelets activated by HIT serum demonstrated numerous surrounding microparticles
ranging in size from <0.1 to 1.0 Am in diameter. (Original magnification 13,000.)
(From Hughes et al., 2000.)
C. ADP Potentiation of Platelet Activation

Adenosine diphosphate (ADP) is an important autocrine stimulator of
platelet activation by HIT-IgG (Chong et al., 1981). This observation was
confirmed by Anderson and Anderson (1990), who showed that FcgRIIa-
mediated activation was augmented by ADP. Although ADP potentiates
platelet activation by many agonists, Polgár and coworkers (1998) found that
pretreatment of platelets with a potent ADP receptor antagonist completely
blocked the activity of HIT sera. This observation indicates that ADP and a
functional ADP receptor are crucial to FcgRIIa activation by HIT-IgG.
However, it should be pointed out that patients receiving ADP receptor an-
tagonists (e.g., clopidogrel) can still develop HIT and HIT-associated throm-
bosis (Warkentin, unpublished observations).
g Receptor–Mediated Signal Transduction

D. Fcg
FcgRIIa activation, as a result of ligand (IgG Fc) binding and by action of
phosphatidylinositol-3 kinase and phospholipase Cg2 (PLCg2), leads to
release of diacylglycerol (DAG) and inositol triphosphate (IP3) and mobili-
zation of internal calcium stores, culminating in platelet aggregation (Ander-
son and Anderson, 1990). Ligand binding leads to clustering of the FcgRIIa,
causing Src family protein tyrosine kinases to phosphorylate the ITAM re-
gion of FcgRIIa (Chacko et al., 1994; Huang et al., 1992). Following FcgRIIa
phosphorylation, additional tyrosine kinase activity (e.g., p72syk) increases
through the noncovalent interaction of their SH2 domains with the phos-
phorylated FcgRIIa ITAMs (Greinacher et al., 1994b; Yanaga et al., 1995;
Chacko et al., 1996). Subsequently, PLCg2 is phosphorylated by p72syk
(Blake et al., 1994); this phosphorylation is dependent on phosphatidylino-
sitol-trisphosphate [PtdIns(3,4,5)P3] (Gratacap et al., 1998). PLCg2 activa-
tion is crucial for the generation of DAG and IP3.
More recently, Gratacap and coworkers (2000) showed that FcgRIIa
activation alone does not produce sufficient levels of PtdIns(3,4,5)P3 to cause
PLCg2 activation, platelet release, and aggregation. Additionally, ADP re-
ceptor activation by Gi-protein signaling leads to the generation of PtdIns
(3,4,5)P3, which combined with activation of FcgRIIa generates optimal lev-
els of PtdIns(3,4,5)P3, leading to PLCg2 phosphorylation. Activated PLCg2
then generates DAG and IP3 from PtdIns(4,5)P2, mobilizing calcium, and
effecting platelet aggregation (Fig. 3). Moreover, lipid rafts appear to play an
Figure 3 Fcg receptor-mediated signal transduction. PF4–heparin–IgG complexes

(1) bind to FcgRIIa, causing receptor clustering (2). The immunoreceptor tyrosine
activation motifs (ITAMs) on FcgRIIa are phosphorylated by src protein tyrosine
kinases (PTKsrc) (3). The phosphorylated ITAMs interact with SH domains on p72syk
to phosphorylate phospholipase Cg2 (PLCg2) (4). ADP receptors, activated via ADP
and Gi proteins, generate phosphatidylinositol-trisphosphate (PIP3) (5), which helps
phosphorylate PLCg2. Activated PLCg2 acts on PIP2 to generate inositol triphos-
phate (IP3), and diacylglycerol (DAG) from phosphatidylinositol-bisphosphate (6).
IP3 mobilizes Ca++ to the intracellular space via Ca++ channels (7) and together with
DAG activates downstream protein kinase C (PKC) signaling pathways (8).
232 Denomme
important role in the organization of the FcgRIIa/ADP receptor/PLCg2

signaling pathway (Bodin et al., 2003). These findings help to explain the
previous observations that ADP scavengers (e.g., apyrase) fully inhibit plate-
let aggregation by HIT-IgG (Polgár et al., 1998).
gRIIa ACTIVATION IN HIT

IV. Fcg
Although an association between heparin treatment and paradoxical throm-

bosis was first suspected about 40 years ago (Weismann and Tobin, 1958;
Roberts et al., 1964), it was Rhodes and colleagues (1973) who first
provided evidence that serum from HIT patients contained a substance,
most likely IgG, that aggregated normal platelets in the presence of heparin.
This observation was confirmed by Fratantoni et al. (1975), who reported a
simple indirect aggregation method for detecting HIT antibodies. In 1986,
Sheridan and coworkers (1986) reported a washed platelet activation assay,
employing radiolabeled serotonin, as an activation endpoint that was sensi-
tive and specific for detecting clinically significant HIT antibodies. This same
group later reported that platelet activation by HIT antibodies was platelet
FcgRIIa-dependent, as it could be completely abrogated by a murine
monoclonal anti-FcgRIIa antibody, IV.3 (Kelton et al., 1988). Other workers
confirmed the central importance of the platelet Fc receptor in mediating
platelet activation in HIT (Adelman et al., 1989; Chong et al., 1989a,b).
Subsequently, Amiral and colleagues (1992) reported that the major target
antigen for HIT-IgG was PF4 complexed to heparin, a finding quickly
confirmed by other workers (Greinacher et al., 1994a; Kelton et al., 1994;
Visentin et al., 1994).
A. Dynamic Model of Platelet Activation in HIT

The initial event in HIT is the binding of HIT-IgG Fab to PF4–heparin com-
plexes on the platelet surface. Thus, HIT-IgG binds to platelets even if the Fc
receptors are blocked (Newman and Chong, 2000). These investigators fur-
ther showed that platelet activation by HIT-IgG is a dynamic process: ini-
tially, tiny amounts of PF4–heparin complexes form on the platelet surface.
HIT-IgG binds to these complexes (through IgG Fab), then engaging and
cross-linking the FcgRIIa by the Fc moieties of the HIT-IgG. This triggers
platelet activation and degranulation (including release of the crucial poten-
tiator, ADP). The released PF4 binds heparin and forms more complexes
containing antigen on the platelet surface. Thus, positive feedback accelerates
platelet activation. HIT-IgG also causes the release of tissue factor and
interleukin-8 (IL-8) from monocytes (Arepally and Mayer, 2001). In addition,
antibodies to IL-8 (a chemokine structurally related to PF4) have been

reported in some HIT patients. It appears that these antibodies can activate
platelets (Regnault et al., 2003).
gRIIa Numbers
B. Platelet Fcg
Variable expression of FcgRIIa numbers among individuals could affect
susceptibility to immune complex diseases (Rosenfeld et al., 1987), or even to
HIT. The number of platelet surface–expressed FcgRIIa molecules is in-
creased dramatically in patients with HIT (Chong et al., 1993b). However,
increased FcgRIIa expression is also seen after in vitro activation of platelets
by HIT antibodies. Thus, elevated FcgRIIa numbers may be a consequence of
platelet activation in HIT, rather than a proximate cause. This notion is sup-
ported by the fact that increased platelet FcgRIIa levels are seen in patients
with atherothrombosis and diabetes mellitus (Calverley et al., 2002). It re-
mains uncertain whether high baseline (pre-HIT) FcgRIIa numbers repre-
sents an important risk factor for HIT.
gRIIa
C. Plasma-Soluble Fcg
Soluble FcgRIIa, which is released from a-granules on platelet activation by
thrombin, has been demonstrated in plasma (Gachet et al., 1995). The soluble
form of the receptor lacks the amino acids for the transmembrane domain, a
result of alternative splicing that removes exon 5 from primary transcripts
(Rappaport et al., 1993). However, the relative amount of membrane versus
soluble FcgRIIa is fixed (Keller et al., 1993). Gachet and colleagues (1995)
reported that approximately 2 ng of soluble FcgRIIa is produced from 109
platelets. This value equals 2 ag, or 300 molecules, per platelet compared with
roughly 10 times as many molecules on the platelet surface. A much larger
amount of plasma-soluble FcgRIIa would be needed to inhibit significantly
PF4–H immune complexes from binding to platelet FcgRIIa. Therefore,
there is likely no effect of soluble FcgRIIa on membrane FcgRIIa-dependent
platelet activation, especially considering that immune complexes formed on
the platelet surface would sterically hinder soluble receptor interaction.
Moreover, plasma levels of soluble FcgRIIa are higher in patients with
HIT than in heparin-treated or other nonthrombocytopenic controls, pre-
sumably as a marker of in vivo platelet activation in HIT (Saffroy et al., 1997).
D. Plasma IgG Concentrations

Plasma IgG levels appear to influence platelet activation and aggregation by
HIT sera. With a platelet-rich plasma (PRP) aggregation test to detect HIT
234 Denomme
antibodies, Chong et al. (1993a) showed variable platelet sensitivity to

aggregation that was stable over time among different platelet donors. Chong
and coworkers showed that the addition of purified human IgG to the PRP
inhibited platelet aggregation by HIT sera, with complete inhibition at 40 mg/
mL. It is possible that the effect of purified IgG is due to the presence of small
IgG oligomers, because Karas et al. (1982) demonstrated that monomeric
IgG does not bind to the platelet FcgRIIa. Furthermore, Greinacher et al.
(1994b) showed that different preparations of intravenous IgG (ivIgG) for
therapeutic use varied in their ability to inhibit HIT antibody-induced platelet
serotonin release. Only Cohn alcohol-fractionated ivIgG preparations re-
tained the ability to inhibit the reaction at concentrations that can be achieved
in vivo (20 mg/mL). Preparations that were treated to reduce IgG oligomers
did not inhibit heparin-dependent platelet serotonin release. Although the use
of ivIgG to treat HIT does not appear to be common, it has some rationale in
certain clinical settings (see Chap. 13).
gRIIa-Arg/His131 Polymorphism
E. Fcg
There is an arginine-histidine (Arg/His) polymorphism at amino acid 131 of
the human FcgRIIa (Clark et al., 1989, Warmerdam et al., 1990). This allelic
variation affects the ability of human platelets to be activated by murine mono-
clonal IgG1 as well as by human IgG2 (Horsewood et al., 1991; Tomiyama et
al., 1992; Parren et al., 1992; Bachelot et al., 1995). This prompted Burgess et
al. (1995) to suggest that inherited FcgRIIa receptor variants could be a risk
factor for developing HIT. In a small cohort of patients, they found an over-
representation of the FcgRIIa–His131 variant. They hypothesized that IgG2
might be an important IgG subclass among HIT-IgG, as this could explain an
apparent association between HIT and the FcgRIIa–His131 variant.
However, subsequent reports argued against this hypothesis: IgG1
rather than IgG2 was the predominant subclass among HIT-IgG (Arepally
et al., 1997; Denomme et al., 1997; Suh et al., 1997). Nevertheless, in support
of a biological basis for a possible increased frequency of FcgRIIa–His131,
two groups found that HIT antibodies, including those that were predomi-
nantly IgG1, preferentially activated washed platelets of the His131 variant in
vitro (Denomme et al., 1997; Bachelot-Loza et al., 1998). However, Brandt et
al. (1995) found the opposite activation profile in platelet aggregation studies
using citrated platelet-rich plasma (i.e., the Arg131 variant was preferentially
activated by HIT plasmas). No consensus has emerged either among the six
studies that investigated whether one of the FcgRIIa–Arg/His131 phenotypes
predominated among patients with HIT: three studies show an overrepresen-
tation of FcgRIIa–His131 (Burgess et al., 1995; Brandt et al., 1995; Denomme
et al., 1997); two studies found no correlation with either variant (Arepally
et al., 1997; Bachelot-Loza et al., 1998); and one study (the largest) showed the
reverse correlation (Carlsson et al., 1998). This topic is considered in detail in
Section V.
F. Animal Models of HIT

One of the earliest animal models of HIT used the natural immune process of
anti-idiotypic antibody production to invoke expression of HIT-IgG in mice
(Blank et al., 1997, 1999). Mice immunized with HIT-IgG developed anti-
idiotypic IgG that now recognized PF4–heparin. Such mice developed throm-
bocytopenia within 4 days of the administration of unfractionated heparin.
Unfortunately, this model has limited use, as the mice did not develop throm-
bosis (perhaps because murine platelets lack FcIIa receptors).
Other investigators (Arepally et al., 2000) developed a murine mono-
clonal antibody, termed KKO, by immunizing mice with PF4–heparin. This
murine IgG2bn monoclonal antibody mimics HIT-IgG, as it requires both
PF4 and heparin to activate human platelets through their FcgRIIa. How-
ever, besides lacking Fc receptors, another problem of the murine system is
that mouse PF4 is not recognized by HIT-IgG or KKO. To overcome these
problems, Reilly and colleagues (2001) produced transgenic mice that express
both human FcgRIIa and human PF4. In these animals, addition of KKO
caused thrombocytopenia and death, including thrombosis of the lung vas-
culature. Thus, this murine model may prove useful to address questions of
antibody titer, FcgRIIa numbers, and so on, in influencing the clinical ex-
pression of HIT.
Murine transgenic models are useful for studying other aspects of anti-
body-mediated thrombocytopenia. For example, McKenzie and coworkers
(1999) demonstrated that transgenic mice expressing human FcgRIIa on
platelets and monocytes became more thrombocytopenic than their matched
wild-type littermates on being administered antimouse platelet IgG. Further,
when FcR g-chain knockout mice (which do not develop thrombocytopenia
in these experiments as all FcgRI and IIIa receptors are lacking) were cross-
bred with FcgRIIa transgenic mice, severe immune thrombocytopenia was
observed in these FcgRIIa transgenic FcR g-chain knockout mice.
However, when platelet-activating (anti-CD9) IgG was administered to
FcgRIIa transgenic mice, even more severe thrombocytopenia resulted, com-
pared with the previously studied antimouse platelet (nonactivating) IgG
(Taylor et al., 2000). Further, severe thrombosis, shock, and death developed
in the FcgRIIa transgenic mice crossed with FcR g-chain knockout mice.
Moreover, splenectomy facilitated anti-CD9-mediated shock in FcgRIIa
transgenic mice. Thus, the authors concluded that phagocytosis by mono-
cytes–macrophages of IgG-sensitized platelets may have a protective role in
236 Denomme
preventing thrombosis. These data have implications for HIT, as there may be
a balance between platelet activation by HIT-IgG (predisposing to thrombo-
sis) and clearance of platelets by monocytes–macrophages (protecting some-
what against thrombosis).
Unlike mice, primate platelets possess FcgRIIa. Thus, a primate model
for HIT may be feasible, as suggested by a recent report (Ahmad et al., 2000).
The animals (Macaca mulatta) used do not express the human Arg–His poly-
morphism, perhaps explaining why less variability in platelet activation re-
sponse to HIT-IgG was observed in these in vitro studies. The primate model
may have value in evaluating therapeutic agents for HIT (Untch et al., 2002).
g Receptors in HIT
G. Monocyte Fcg
Monocytes and macrophages possess several different classes of FcgR (see
Table 1), and thus may play a part in influencing the frequency and severity of
both thrombocytopenia and thrombosis in HIT. One role, discussed in the
previous section, involves their potential to influence the balance between
platelet activation and reticuloendothelial-mediated platelet clearance in
HIT. Another function recently proposed for monocytes is that of contrib-
uting to the procoagulant state in HIT, (i.e., a role posited previously for
endothelial cells) (see Chap. 10). Pouplard and colleagues (2001) found that
by adding HIT-IgG and PF4 (or PF4–heparin) directly to isolated monocytes
or to whole blood, the monocytes produced tissue factor (TF), an effect that
could be inhibited by high concentrations of heparin. Arepally and Mayer
(2001) found that monocytes expressed surface TF when incubated with PF4
either in the presence of HIT-IgG or the HIT-mimicking murine monoclonal
antibody, KKO. Because monocytes express sulfated proteoglycans on their
surface, PF4 binding to monocytes can occur in the absence of added heparin.
These studies raise the possibility that monocytes play an important role in
the pathogenesis of the procoagulant state characteristic of HIT.
gRIIa POLYMORPHISMS IN DISEASE

V. Fcg
gRIIa Polymorphism
A. Determining the Fcg
The FcgRIIa–Arg/His131 polymorphism was first identified on the basis of
functional differences effected by anti-CD3 monoclonal antibodies of the
murine IgG1 subclass (Tax et al., 1983, 1984). Proliferation assays distin-
guished ‘‘high’’ and ‘‘low’’ responders relative to the effects of these anti-CD3
murine monoclonal antibodies on T-cell–dependent mitogenesis. Subsequent-
ly, individuals bearing the FcgRIIa–Arg131 phenotype were identified as the
‘‘high responders’’ and the functional differences between the two polymor-
phic variants were later confirmed using other FcgRIIa-dependent assays,
such as erythrocyte antigen-rosetting, phagocytosis, and platelet activation
(Clark et al., 1989; Warmerdam et al., 1991; Parren et al., 1992; Salmon et al.,
1992). Murine monoclonal IgG1 antibodies activate platelets of all three Arg/
His131 phenotypes, but the homozygous FcgRIIa–Arg131 variant requires less
murine monoclonal antibody for platelet activation to occur.
The high-affinity binding of human IgG2 to FcgRIIa results when
histidine is substituted at amino acid 131 of the mature protein (Warmerdam
et al., 1991). FcgRIIa–His131 has a greater affinity for human IgG2, but a
lower affinity for murine IgG1. Therefore, the terms high and low responder,
used historically for the effects of murine monoclonal antibodies on Arg131
and His131 FcgRIIa phenotypes, respectively, is confusing, as the opposite
reaction profile is observed with human IgG2. The high/low responder ter-
minology has been largely replaced in favor of referring simply to the amino
acid polymorphism.
The FcgRIIa–Arg/His131 variant polymorphism can be determined in
three ways: (a) by functional assay, such as T-cell–dependent proliferation or
murine monoclonal antibody activation; (b) by specific binding using 41H16,
a monoclonal antibody the Fab of which binds exclusively to the FcgRIIa–
Arg131 variant; and (c) by molecular genotyping. Four DNA-based methods
have been developed to genotype for the FcgRIIa–Arg/His131 nucleotide
substitution (Clark et al., 1991; Osborne et al., 1994; Bachelot et al., 1995;
Jiang et al., 1996; Denomme et al., 1997). In one technique, the presence of the
FcgRIIa–Arg/His131 variant gene is determined using genomic DNA and a
sequence-specific primer–polymerase chain reaction (PCR) assay. Two PCR
reactions are necessary, each containing a common primer paired with a
unique primer having different 3V-ends to detect the presence of the G or A
variant nucleotide (Clark et al., 1991). This method has been modified using
different sequence-specific primers (Flesch et al., 1998) or using a nested
sequence-specific PCR (Carlsson et al., 1998). In a second technique, flank-
ing primers are used to amplify a region containing the nucleotide poly-
morphism, followed by dot-blotting and hybridization with allele-specific,
single-stranded oligonucleotide probes (Osborne et al., 1994; Burgess et al.,
1995; Denomme et al., 1997). In a third technique, Bachelot and coworkers
(1995) developed a denaturing gradient gel electrophoresis assay that distin-
guishes between the FcgRIIa–Arg/His131 variants also using flanking primers
that amplify a region containing the polymorphism. Last, restriction endo-
nuclease digestion of PCR-amplified genomic DNA has been developed using
one primer immediately proximal to the polymorphic site and containing a
mutation such that the polymorphism creates a restriction enzyme site for
only one of the alleles (Jiang et al., 1996).
238 Denomme
gRIIa Polymorphism in Infectious

B. Influence of Fcg
or Autoimmune Disease
A few studies have examined whether expression of the FcgRIIa–Arg/His131
polymorphism influences susceptibility to infectious or autoimmune disease.
In theory, the weaker binding of human IgG2 to the FcgRIIa–Arg131 variant
suggests that this gene might be overrepresented among patients with
recurrent infections characterized by certain microbes with polysaccharide
coats (i.e., involving an IgG2 antibody response) and overrepresented in
disease characterized by circulating immune complexes (because phagocytic
cells bearing the FcgRIIa–His131 variant would clear these complexes more
readily). Certainly, a skewed genotypic distribution favoring the FcgRIIa-
Arg131 variant has been noted in patients with Haemophilus influenzae
infections (Sanders et al., 1994) and meningococcal septic shock (Bredius
et al., 1994a). Furthermore, there is also predominance of FcgRIIa–Arg131 in
patients with elevated levels of immune complexes and glomerulonephritis
complicating systemic lupus erythematosus (Duits et al., 1995) (Table 2).
gRIIa Polymorphism in HIT

C. Fcg
It was logical to hypothesize that the platelet FcgRIIa–Arg/His131 polymor-
phism would influence the clinical expression of HIT. First, platelets from
normal individuals exhibit considerable variability in their activation by HIT
sera (Salem and Van der Weyden, 1983; Pfueller and David, 1986; Warkentin
et al., 1992). Second, many patients who form HIT antibodies during heparin
treatment do not develop thrombocytopenia (Warkentin et al., 1995; Amiral
et al., 1996; Suh et al., 1997). Third, the inciting role of heparin, a sulfated
carbohydrate, suggested that there could be an important role for HIT
antibodies of IgG2 subclass—that is, the subclass with higher affinity for
FcgRIIa–His131 that is predominantly formed in response to carbohydrate
antigens (Herrmann et al., 1992). (However, HIT epitopes form on the protein
PF4 when it undergoes conformation change bound to heparin) (see Chaps.
6–8). Consequently, it was speculated that the FcgRIIa variant distribution in
HIT would differ significantly from a random control population, and
especially differ from patients who did not develop thrombocytopenia during
heparin treatment (Denomme et al., 1997; Bachelot-Loza et al., 1998).
The six studies investigating the role of the FcgRIIa–Arg/His131 poly-
morphism have not yielded uniform results (Fig. 4). Three studies showed a
predominance of the His131 variant in patients with HIT that was significant,
compared with control patients. Together with evidence that HIT antibodies
preferentially activate platelets in vitro from individuals bearing the FcgRIIa–
His131 polymorphism (Denomme et al., 1997; Bachelot-Loza et al., 1998), it
was suggested that FcgRIIa–His131 predominance could reflect a greater
Table 2 Role of FcgRIIa–Arg/His131 Polymorphism in Disease
Predominant
FcgRIIa
Disease variant Comment
Infections by Arg131 Reduced binding of IgG2 by FcgRIIa–Arg131

encapsulated -MPS cells reduces phagocytosis,
bacteria conferring susceptibility to infections
with bacteria bearing polysaccharide
capsules (Haemophilus, meningococcus).
Immune complex Arg131 Reduced clearance of IgG-containing
nephritis (SLE) immune complexes by FcgRIIa–Arg131
-MPS cells leads to greater glomerular
deposition of immune complexes.
HIT with Arg131 Reduced clearance of IgG-containing
thrombosis immune complexes by FcgRIIa–Arg131
-MPS cells leads to greater immune
complex-dependent activation of platelets
and endothelial cells (Carlsson et al., 1998).
HIT with or without His131 Increased activation by HIT-IgG1 and
thrombosis HIT-IgG2 of FcgRIIa–His131 platelets,
without significant role for MPS cells
(Denomme et al., 1997).
Abbreviations: Arg, arginine; FcgRIIa, FcgIIa receptor; His, histidine; MPS, mononuclear phagocytic
system (reticuloendothelial system); SLEF, systemic lupus erythematosus.
potential for these platelets to be activated in vivo by HIT antibodies (see

Table 2). Two relatively small studies did not show any significant differences
in the Arg/His131 phenotypes between HIT patients and controls.
However, the largest of the six studies, involving 389 patients (i.e., more
than the 260 HIT patients reported in the previous five studies combined),
showed an increase in the frequency of the Arg131, rather than the His131,
variant in patients with HIT (Carlsson et al., 1998). Moreover, these workers
made the intriguing observation that the increase in Arg131 phenotype
occurred only in the subset of patients whose HIT was complicated by
thrombosis. These investigators proposed that reduced clearance of IgG-
containing immune complexes by phagocytic cells bearing FcgRIIa–Arg131
leads to greater immune complex-dependent activation of platelets and
endothelial cells, thus predisposing to thrombosis (see Table 2). Although
Arepally et al. (1997) did not observe a significant increase in the Arg131
phenotype among HIT patients with thrombosis, their subset of HIT patients
with thrombosis was much smaller than that reported by Carlsson (23 vs. 68
240 Denomme
Figure 4 FcgRIIa–His131 gene frequencies in six studies of HIT are shown: The
first four studies were from North American centers, the last two from Europe.
Although the first three studies showed predominance of His131 in patients with HIT,
the last study showed predominance off Arg131 in patients with HIT complicated by
thrombosis. A complicating feature is the difference in gene frequencies between
certain control populations [e.g., between Denomme et al. (1997) and Carlsson et al.
(1998)]. Not shown in the figure is the significant difference between control patients
in the studies by Carlsson and Brandt ( p = 0.013).
patients). On the other hand, when Pouplard and colleagues (1999) examined
the Arg/His131 gene frequency among patients who formed antibodies against
PF4–heparin following cardiac surgery, they noted that platelet levels were
significantly lower only in the Arg/Arg131 group, when compared with
patients who did not form antibodies.
The explanation for the differences among these various studies is not
readily apparent. However, a complicating aspect is noted in Fig. 4: the fre-
quency of the His131 phenotype is higher in the European controls (Bachelot-

Loza et al., 1998; Carlsson et al., 1998), compared with the North American
and Australian control populations (Burgess et al., 1995; Brandt et al., 1995;
Denomme et al., 1997; Arepally et al., 1997), an observation consistent with
population allele frequencies reported by Rascu et al. (1997). Indeed, pairwise
m2 analysis for the frequency of the His131 genotype among the various
control groups shows that the control population of Carlsson’s study differs
from that reported by Denomme and Brandt (see Fig. 4). The FcgRIIa–Arg/
His131 polymorphism varies among populations: in whites and African
Americans, the gene frequencies have roughly a 50:50 balance (Osborne
et al., 1994; Lehrnbecher et al., 1999). In contrast, in the Japanese and Chinese
populations, the His131 gene frequency is approximately 75% (Rascu et al.,
1997; Osborne et al., 1994). It is possible that unrecognized differences in
population between HIT patients and controls could be important. For ex-
ample, whereas samples from HIT patients could be referred from a wider
geographic area, control patients might have been obtained from a localized
area. None of the six studies reported on the Arg/His131 phenotype distribu-
tion among heparin-treated patients who formed HIT antibodies, but who
did not develop thrombocytopenia (i.e., the ideal control group for assessing
the influence of the FcgRIIa polymorphism).
In summary, the role of the FcgRIIa–Arg/His131 polymorphism in
contributing to the pathogenesis of HIT remains controversial. Regardless
of its ultimate resolution, the elucidation of the biological basis for differences
in frequency of FcgRIIa phenotype between HIT patients, with or without
thrombosis, and control subjects will provide new insights into the patho-
genesis of immune-mediated disease.
ACKNOWLEDGMENTS
The author wishes to thank Dr. Lena E. Carlsson for her helpful review of
the manuscript. Some of the studies described were supported by a Career
Development Fellowship Award of the Canadian Blood Services. The
author was a Bayer/Canadian Blood Services/Medical Research Council
Scholar.
REFERENCES
Adelman B, Sobel M, Fujimura Y, Ruggeri ZM, Zimmerman TS. Heparin-associated

thrombocytopenia: observations on the mechanism of platelet aggregation. J
Lab Clin Med 1989; 113:204–210.
242 Denomme
Ahmad S, Jeske WP, Walenga JM, Aldabbagh A, Iqbal O, Fareed J. Human anti-
heparin-platelet factor 4 antibodies are capable of activating primate platelets:
towards the development of a HIT model in primates. Thromb Res 2000;
100:47–54.
Amiral J, Bridey F, Dreyfus M, Vissaco AM, Fressinaud E, Wolf M, Meyer D. Platelet
induced thrombocytopenia. Thromb Haemost 1992; 68:95–96.
Amiral J, Peynaud-Debayle E, Wolf M, Bridey F, Vissac A-M, Meyer D. Generation
of antibodies to heparin-PF4 complexes without thrombocytopenia in patients
1996; 52:90–95.
Anderson GP, Anderson CL. Signal transduction by the platelet Fc receptor. Blood
1990; 76:1165–1172.
Anderson GP, van de Winkel JGJ, Anderson CL. Anti-GPIIb/IIIa (CD41)
monoclonal antibody-induced platelet activation requires Fc receptor-depend-
ent cell-cell interaction. Br J Haematol 1991; 79:75–83.
Arepally G, McKenzie SE, Jiang X-M, Poncz M, Cines DB. FcgRIIA H/R131
Arepally GM, Mayer IM. Antibodies from patients with heparin-induced thrombo-
cytopenia stimulate monocytic cells to express tissue factor and secrete inter-
leukin-8. Blood 2001; 98:1252–1254.
Arepally GM, Kamei S, Park KS, Kamei K, Li ZQ, Liu W, Siegel DL, Kisiel W,
Cines DB, Poncz M. Characterization of a murine monoclonal antibody that
mimics heparin-induced thrombocytopenia antibodies. Blood 2000; 95:1533–
1540.
Bachelot C, Saffroy R, Gandrille S, Aiach M, Rendu F. Role of FcgRIIA gene
polymorphism in human platelet activation by monoclonal antibodies. Thromb
Haemost 1995; 74:1557–1563.
the FcgRIIa-Arg/His-131 polymorphism in heparin-induced thrombocytope-
Bevers EM, Comfurius P, Dekkers DW, Zwaal RF. Lipid translocation across the
plasma membrane of mammalian cells. Biochim Biophys Acta 1999; 1439:317–
330.
Blake RA, Asselin J, Walker T, Watson SP. Fc gamma receptor II stimulated for-
mation of inositol phosphates in human platelets is blocked by tyrosine kinase
inhibitors and associated with tyrosine phosphorylation of the receptor. FEBS
Lett 1994; 342:15–18.
Blank M, Cines DB, Arepally G, Eldor A, Afek A, Shoenfeld Y. Pathogenicity of
human antiplatelet factor 4/heparin in vivo: generation of mouse anti-PF4/
heparin and induction of thrombocytopenia by heparin. Clin Exp Immunol
1997; 108:333–339.
Blank M, Eldor A, Tavor S, Ziporen L, Cines DB, Arepally G, Afek A, Shoenfeld Y. A
mouse model for heparin-induced thrombocytopenia. Semin Hematol 1999;

36(suppl 1):12–16.
Bode AP, Hickerson DHM. Characterization and quantitation by flow cytometry of
membranous microparticles formed during activation of platelet suspensions
with ionophore or thrombin. Platelets 2000; 11:259–271.
Bodin S, Viala C, Ragab A, Payrastre B. A critical role of lipid rafts in the organization
of a key FcgRIIa-mediated signaling pathway in human platelets. Thromb
Haemost 2003; 89:318–330.
Brandt JT, Isenhart CE, Osborne JM, Ahmed A, Anderson CL. On the role of platelet
FcgRIIa phenotype in heparin-induced thrombocytopenia. Thromb Haemost
1995; 74:1564–1572.
Bredius RGM, Derkz BHF, Fijen CAP, de Wit TPM, de Haas M, Weening RS, van de
Winkel JGJ, Out TA. Fc gamma IIa (CD32) polymorphism in fulminant
meningococcal septic shock in children. J Infect Dis 1994a; 170:848–853.
Bredius RGM, Fijen CAP, de Haas M, Kuijper EJ, Weening RS, van de Winkel JGJ,
Out TA. Role of neutrophil FcgRIII (CD32) and FcgRIII (CD16) polymorphic
forms in phagocytosis of human IgG1- and IgG3-opsonized bacteria and
erythrocytes. Immunology 1994b; 83:624–630.
Brooks DG, Qiu WQ, Luster AD, Ravetch JV. Structure and expression of human
IgG FcRII (CD32). Functional heterogeneity is encoded by the alternatively
spliced products of multiple genes. J Exp Med 1989; 170:1369–1385.
Burgess JK, Lindeman R, Chesterman CN, Chong BH. Single amino acid mutation of
Fcg receptor is associated with the development of heparin-induced thrombo-
cytopenia. Br J Haematol 1995; 91:761–766.
Bux J, Stein EL, Bierling P, Fromont P, Clay ME, Stoncek DF, Santoso S. Char-
acterization of a new alloantigen (SH) on the human neutrophil Fcg receptor
IIIb. Blood 1997; 89:1027–1034.
Calverley DC, Brass E, Hacker MR, Tsao-Wei DD, Espina BM, Pullarkat VA, Hodis
HN, Groshen S. Potential role of platelet FcgRIIA in collagen-mediated platelet
activation associated with atherothrombosis. Atherosclerosis 2002; 164:261–
267.
NAC, Kiefel V, van de Winkel JGJ, Greinacher A. Heparin-induced thrombo-
cytopenia: new insights into the impact of the FcgRIIa-R-H131 polymorphism.
Blood 1998; 92:1526–1531.
Chacko GW, Duchemin A-M, Coggeshall KM, Osborne JM, Brandt JT, Anderson
CL. Clustering of the platelet Fcg receptor induces noncovalent association with
the tyrosine kinase p72syk. J Biol Chem 1994; 269:32435–32440.
Chacko GW, Brandt JT, Coggeshall KM, Anderson CL. Phosphoinositide 3-kinase
and p72syk noncovalently associate with the low affinity Fcg receptor on human
platelets through an immunoreceptor tyrosine-based activation motif. J Biol
Chem 1996; 271:10775–10781.
Chong BH, Grace CS, Rozenberg MC. Heparin-induced thrombocytopenia: effect of
heparin platelet antibody on platelets. Br J Haematol 1981; 49:531–540.
Chong BH, Fawaz I, Chesterman CN, Berndt MC. Heparin-induced thrombocyto-
244 Denomme
penia: mechanism of interaction of the heparin-dependent antibody with

platelets. Br J Haematol 1989a; 73:235–240.
Chong BH, Castaldi PA, Berndt MC. Heparin-induced thrombocytopenia: effect of
rabbit IgG, and its Fab and Fc fragments on antibody-heparin-platelet inter-
action. Thromb Res 1989b; 55:291–295.
Chong BH, Burgess J, Ismail F. The clinical usefulness of the platelet aggregation test
for the diagnosis of heparin-induced thrombocytopenia. Thromb Haemost
1993a; 69:344–350.
Chong BH, Pilgrim RL, Cooley MA, Chesterman CN. Increased expression of platelet
IgG Fc receptors in immune heparin-induced thrombocytopenia. Blood 1993b;
81:988–993.
Clark MR, Clarkson SB, Ory PA, Stollman N, Goldstein IM. Molecular basis for a
polymorphism involving Fc receptor II on human monocytes. J Immunol 1989;
143:1731–1734.
Clark MR, Stuart SG, Kimberly RP, Ory PA, Goldstein IM. A single amino acid
distinguishes the high-responder from the low-responder form of Fc receptor II
on human monocytes. Eur J Immunol 1991; 21:1911–1916.
Denomme GA, Warkentin TE, Horsewood P, Sheppard J-AI, Warner MN, Kelton
JG. Activation of platelets by sera containing IgG1 heparin-dependent
antibodies: an explanation for the predominance of the FcgRIIa ‘‘low
responder’’ (his131) gene in patients with heparin-induced thrombocytopenia.
J Lab Clin Med 1997; 130:278–284.
Duits AJ, Bootsma H, Derksen RHWM, Spronk PE, Kater L, Kallenberg CGM,
Capel PJA, Westerdaal NAC, Spierenburg GT, Gmelig-Meyling FHJ, van de
Winkel JGJ. Skewed distribution of IgG Fc receptor IIa (CD32) polymorphism
is associated with renal disease in systemic lupus erythematosus patients.
Arthritis Rheum 1995; 39:1832–1836.
Flesch BK, Bauer F, Neppert J. Rapid typing of the human Fcg receptor IIA
polymorphism by polymerase chain reaction amplification with allele-specific
primers. Transfusion 1998; 38:174–176.
Fratantoni JC, Pollet R, Gralnick HR. Heparin-induced thrombocytopenia:
confirmation of diagnosis with in vitro methods. Blood 1975; 45:395–401.
Gachet C, Astier A, de la Salle H, de la Salle C, Fridman WH, Cazenave J-P, Hanau D,
Teillaud J-L. Release of FcgRIIa2 by activated platelets and inhibition of anti-
CD9-mediated platelet aggregation by recombinant FcgRIIa2. Blood 1995;
85:698–704.
Gessner JE, Heiken H, Tamm A, Schmidt RE. The IgG Fc receptor family. Ann
Hematol 1998; 76:231–248.
Green D, Harris K, Reynolds N, Roberts M, Patterson R. Heparin immune throm-
bocytopenia: evidence for a heparin-platelet complex as the antigenic deter-
minant. J Lab Clin Med 1978; 91:167–175.
Gratacap MP, Payrastre B, Viala C, Mauco G, Plantavid M, Chap H. Phosphati-
dylinositol 3,4,5-triphosphate-dependent stimulation of phospholipase C-
gamma2 is an early key event in FcgammaRIIA-mediated activation of human

platelets. J Biol Chem 1998; 273:24314–24321.
Gratacap MP, Hérault JP, Viala C, Ragab A, Savi P, Herbert JM, Chap H, Plantavid
M, Payrastre B. FcgRIIA requires a Gi-dependent pathway for an efficient
stimulation of phosphoinositide 3-kinase, calcium mobilization, and platelet
aggregation. Blood 2000; 96:3439–3446.
Greinacher A, Michels I, Kiefel V, Mueller-Eckhardt C. A rapid and sensitive test for
diagnosing heparin-associated thrombocytopenia. Thromb Haemost 1991; 66:
734–736.
Haemost 1994a; 71:247–251.
Greinacher A, Liebenhoff U, Kiefel V, Presek P, Mueller-Eckhardt C. Heparin-
associated thrombocytopenia: the effects of various intravenous IgG prepara-
tions on antibody mediated platelet activation—a possible new indication of
high dose i.v. IgG. Thromb Haemost 1994b; 71:641–645.
Gröger M, Sarmay G, Fiebiger E, Wolff K, Petzelbauer P. Dermal microvascular
endothelial cells express CD32 receptors in vivo and in vitro. J Immunol 1996;
156:1549–1556.
Herrmann DJ, Hamilton RG, Barington T, Frasch CE, Arakere G, Mäkelä O,
Mitchell LA, Nagel J, Rijkers GT, Zegers B, Danve B, Ward JI, Brown CS.
Quantitation of human IgG subclass antibodies to Haemophilus influenzae type
b capsular polysaccharide. J Immunol Methods 1992; 148:101–114.
Horsewood P, Hayward CPM, Warkentin TE, Kelton JG. Investigation of the
mechanisms of monoclonal antibody-induced platelet activation. Blood 1991;
78:1019–1026.
Huang MM, Indik Z, Brass LF, Hoxie JA, Schreiber AD, Brugge JS. Activation of Fc
gamma RII induces tyrosine phosphorylation of multiple proteins including Fc
gamma RII. J Biol Chem 1992; 267:5467–5473.
Hughes M, Hayward CPM, Warkentin TE, Horsewood P, Chorneyko KA, Kelton
JG. Morphological analysis of microparticle generation in heparin-induced
thrombocytopenia. Blood 2000; 96:188–194.
Hulett MD, Witort E, Brinkworth RI, McKenzie IFC, Hogarth PM. Multiple regions
of human FcgRII (CD32) contribute to the binding of IgG. J Biol Chem 1995;
270:21188–21194.
Jiang X-M, Arepally G, Poncz M, McKenzie SE. Rapid detection of the FcgRIIA-H/
R131 ligand-binding polymorphism using an allele-specific restriction enzyme
digestion (ASRED). J Immunol Methods 1996; 199:55–59.
Karas SP, Rosse WF, Kurlander RJ. Characterization of the IgG-Fc receptor on
human platelets. Blood 1982; 60:1277–1282.
Keller MA, Cassel DL, Rappaport EF, McKenzie SE, Schwartz E, Surrey S.
fluorescence-based RT PCR analysis: determination of the ratio of soluble to
membrane-bound forms of FcgRIIA transcripts in hematopoietic cell lines.
PCR Methods Appl 1993; 3:32–38.
246 Denomme
Kelton JG, Smith JW, Santos AV, Murphy WG, Horsewood P. Platelet IgG Fc
receptor. Am J Hematol 1987; 25:299–310.
925–930.
Khairy M, Lasne D, Brohard-Bohn B, Aich M, Rendu F, Bachelot-Loza C. A
new approach in the study of the molecular and cellular events implicated
in heparin-induced thrombocytopenia. Formation of leukocyte-platelet aggre-
gates. Thromb Haemost 2001; 85:1090–1096.
King M, McDermott P, Schreiber AD. Characterization of the Fc-gamma receptor on
human platelets. Cell Immunol 1990; 128:462–479.
Koene HR, Kleijer M, Algra J, Roos D, von dem Borne AEGK, de Haas M.
FcgRIIIa-158V/F polymorphism influences the binding of IgG by natural killer
cell FcgRIIIa, independently of the FcgRIIIa-48L/R/H phenotype. Blood 1997;
90:1109–1114.
Koene HR, Kleijer M, Roos D, de Haas M, von dem Borne AEGK. FcgRIIIB gene
duplication: evidence for presence and expression of three distinct FcgRIIIB
genes in NA(1+,2+)SH(+) individuals. Blood 1998; 91:673–679.
Lebrazi J, Helft G, Abdelouahed M, Elalamy I, Mirshahi M, Samama MM, Lecompte
T. Human anti-streptokinase antibodies induce platelet aggregation in an Fc
receptor (CD32) dependent manner. Thromb Haemost 1995; 74:938–942.
Lee DP, Warkentin TE, Denomme GA, Hayward CPM, Kelton JG. A diagnostic test
for heparin-induced thrombocytopenia: detection of platelet microparticles
using flow cytometry. Br J Haematol 1996; 95:724–731.
Lehrnbecher T, Foster CH, Zhu S, Leitman SF, Goldin LR, Huppi K, Chanock SJ.
Variant genotypes of the low-affinity Fcgamma receptors in two control popu-
lations and a review of low-affinity Fcgamma receptor polymorphisms in con-
trol and disease populations. Blood 1999; 94:4220–4232.
McCrae KR, Shattil SJ, Cines DB. Platelet activation induces increased Fcg receptor
expression. J Immunol 1990; 144:3920–3927.
McKenzie SE, Taylor SM, Malladi P, Yuhan H, Cassel DL, Chien P, Schwartz E,
Schreiber AD, Surrey S, Reilly MP. The role of the human Fc receptor FcgRIIA
in the immune clearance of platelets: a transgenic mouse model. J Immunol
1999; 162:4311–4318.
Metes D, Ernst LK, Chambers WH, Sulica A, Herberman RB, Morel PA. Expression
of functional CD32 molecules on human NK cells is determined by an allelic
polymorphism of the FcgammaRIIC gene. Blood 1998; 91:2369–2380.
dynamic binding of purified anti-PF4–heparin antibodies to platelets and the
Norris CF, Pricop L, Millard S, Taylor SM, Surrey S, Schwartz E, Salmon JE,
McKenzie SE. A naturally occurring mutation in FcgRIIA: a Q to K127 change
confers unique IgG binding properties to the R131 allelic form of the receptor.
Blood 1998; 91:656–662.
Ory PA, Clark MA, Kwoh EE, Clarkson SB, Goldstein IM. Sequences of comple-
mentary DNAs that encode the NA1 and NA2 forms of Fc receptor III on
human neutrophils. J Clin Invest 1989; 84:1688–1691.
Osborne JM, Chacko GW, Brandt JT, Anderson CL. Ethnic variation in frequency of
an allelic polymorphism of human FcgRIIa determined with allele specific
oligonucleotide probes. J Immunol Methods 1994; 173:207–217.
Parren PWHI, Warmerdam PAM, Boeije LCM, Arts J, Westerdaal NAC, Vlug A,
Capel PJA, Aarden LA, van de Winkel JGJ. On the interaction of IgG sub-
classes with the low affinity Fc gamma RIIa (CD32) on human monocytes,
neutrophils, and platelets. Analysis of a functional polymorphism to human
IgG2. J Clin Invest 1992; 90:1537–1546.
Pfueller SL, David R. Different platelet specificities of heparin-dependent plate-
let aggregating factors in heparin-associated immune thrombocytopenia. Br J
Haematol 1986; 64:149–159.
sera from heparin-induced thrombocytopenia patients. Blood 1998; 91:549–
554.
Pouplard C, May MA, Iochmann S, Amiral J, Marchand M, Gruel Y. Antibodies
to platelet factor 4–heparin after cardiopulmonary bypass in patients anti-
coagulated with unfractionated heparin or a low-molecular-weight heparin:
clinical implications for heparin-induced thrombocytopenia. Circulation 1999;
99:2530–2536.
Pouplard C, Iochmann S, Renard B, Herault O, Gruel Y. Induction of monocyte tissue
factor expression by antibodies to heparin-platelet factor 4 complexes developed
in heparin-induced thrombocytopenia. Blood 2001; 97:3300–3302.
Qiu WQ, de Bruin D, Brownstein BH, Pearse R, Ravetch JV. Organization of the
human and mouse low-affinity FcgR genes: duplication and recombination.
Science 1990; 248:732–735.
Rappaport EF, Cassel DL, Walterhouse DO, McKenzie SE, Surrey S, Keller MA,
Schreiber AD, Schwartz E. A soluble form of the human Fc receptor FcgRIIa:
cloning, transcript analysis and detection. Exp Hematol 1993; 21:689–696.
Rascu A, Repp R, Westerdaal NAC, Kalden JR, van de Winkel JGJ. Clinical rele-
vance of Fcg receptor polymorphisms. Ann NY Acad Sci 1997; 815:282–295.
Regnault V, de Maistre E, Carteaux JP, Gruel Y, Nguyen P, Tardy B, Lecompte T.
Platelet activation induced by antibodies to interleukin-8. Blood 2003; 101:
1419–1421.
Reilly MP, Taylor SM, Hartman NK, Arepally GM, Cines DB, Poncz M, McKen-
zie SE. Heparin-induced thrombocytopenia/thrombosis in a transgenic mouse
model requires human platelet factor 4 and platelet activation through Fcg-
RIIA. Blood 2001; 98:2442–2447.
248 Denomme
Roberts B, Rosato FE, Rosato EF. Heparin: a cause of arterial emboli? Surgery 1964;
55:803–808.
Roitt I, Brostoff J, Male D, Eds. Immunology. London: Gower Medical Publishing,
1985.
Rosenfeld SI, Looney RJ, Leddy JP, Phipps DC, Abraham GN, Anderson CL.
Human platelet Fc receptor for immunoglobulin G Identification as a 40,000-
molecular-weight membrane protein shared by monocytes. J Clin Invest 1985;
76:2317–2322.
Rosenfeld SI, Ryan DH, Looney RJ, Anderson CL, Abraham GN, Leddy JP.
Human Fcg receptors: stable inter-donor variation in quantitative expression
on platelets correlates with functional responses. J Immunol 1987; 144:3920–
3927.
Rubinstein E, Boucheix C, Worthington RE, Carroll RC. Anti-platelet antibody
interactions with Fcg receptor. Semin Thromb Hemost 1995; 21:10–22.
Saffroy R, Bachelot-Loza C, Fridman WH, Aiach M, Teullaud J-L, Rendu F. Plasma
levels of soluble Fcg receptors II (sCD32) and III (sCD16) in patients with
heparin-induced thrombocytopenia. Thromb Haemost 1997; 78:970–971.
Salem HH, Van der Weyden MB. Heparin-induced thrombocytopenia. Variable
platelet-rich plasma reactivity to heparin-dependent aggregating factor. Pathol-
ogy 1983; 15:297–299.
Salmon JE, Edberg JC, Kimberly RP. Fcg receptor III on human neutrophils. J Clin
Invest 1990; 85:1287–1295.
Salmon JE, Edberg JC, Brogle NL, Kimberly RP. Allelic polymorphism of human Fcg
receptor IIA and Fcg receptor IIIB. Independent mechanisms for differences in
human phagocyte function. J Clin Invest 1992; 89:1274–1281.
Sanders LAM, van de Winkel JGJ, Rijkers GT, Voorhorst-Ogink MM, de Haas M,
Capel PJA, Zegers BJM. Fc gamma receptor IIa (CD32) heterogeneity in
patients with recurrent bacterial respiratory tract infections. J Infect Dis 1994;
170:854–861.
Sedmak DD, Davis DH, Singh U, van de Winkel JG, Anderson CL. Expression of IgG
Fc receptor antigens in placenta and on endothelial cells in humans. An im-
munohistochemical study. Am J Pathol 1991; 138:175–181.
Sims PJ, Wiedmer T, Esmon CT, Weiss HJ, Shattil SJ. Assembly of the platelet
prothrombinase complex is linked to vesiculation of the platelet plasma mem-
brane. Studies in Scott syndrome: an isolated defect in platelet procoagulant
activity. J Biol Chem 1989; 264:17049–17057.
Su K, Wu J, Edberg JC, McKenzie SE, Kimberly RP. Genomic organization of
classical human low-affinity Fcg receptor genes. Genes Immun 2002; 3(suppl 1):
S51–S56.
Suh JS, Malik MI, Aster RH, Visentin GP. Characterization of the humoral response
in heparin-induced thrombocytopenia. Am J Hematol 1997; 54:196–201.
Taylor SM, Reilly MP, Schreiber AD, Chien P, Tuckosh JR, McKenzie SE. Throm-
bosis and shock induced by activating antiplatelet antibodies in human Fcg-
RIIA transgenic mice: the interplay among antibody, spleen, and Fc receptor.
Blood 2000; 96:4254–4260.
Tax WJM, Williams HW, Reckers PPM, Capel PJA, Keone RAP. Polymorphism in
mitogenic effect of IgG1 monoclonal antibodies against T3 antigen on human T
cells. Nature 1983; 304:445.
Tax WJM, Hermes FFM, Willems RW, Capel JA, Koene RAP. Fc receptors for
mouse IgG 1 on human monocytes: polymorphism and role in antibody-
induced T cell proliferation. J Immunol 1984; 133:1185–1189.
Tomer A. A sensitive and specific functional flow cytometric assay for the diagnosis of
Tomiyama Y, Kunicki TJ, Zipf TF, Ford SB, Aster RH. Response of human platelets
to activating monoclonal antibodies: importance of FcgRII (CD32) phenotype
and level of expression. Blood 1992; 80:2261–2268.
Untch B, Ahmad S, Messmore HL, Schultz CL, Ma Q, Hoppensteadt DA, Walenga
JM, Fareed J. Development of a non-human primate sub-clinical model of
heparin-induced thrombocytopenia: platelet response to human anti-heparin-
platelet factor 4 antibodies. Thromb Res 2002; 106:149–156.
van de Winkel JGJ, Capel PJA. Human IgG Fc receptor heterogeneity: molecular
aspects and clinical implications. Immunol Today 1993; 14:215–221.
plexed with heparin or bound to endothelial cells. J Clin Invest 1994; 93:81–
88.
Warkentin TE, Sheppard JI. Generation of platelet-derived microparticles and pro-
coagulant activity by heparin-induced thrombocytopenia IgG/serum and other
IgG platelet agonists: a comparison with standard platelet agonists. Platelets
1999; 10:319–326.
Warkentin TE, Hayward CPM, Smith CA, Kelly PM, Kelton JG. Determinants of
donor platelet variability when testing for heparin-induced thrombocytopenia. J
Lab Clin Med 1992; 120:371–379.
Kelton JG. Sera from patients with heparin-induced thrombocytopenia gen-
erate platelet-derived microparticles with procoagulant activity: an explanation
for the thrombotic complications of heparin-induced thrombocytopenia. Blood
1994; 84:3691–3699.
Warkentin TE, Levine MN, Hirsh J, Horsewood P, Roberts RS, Tech M, Gent M,
Kelton JG. Heparin-induced thrombocytopenia in patients treated with low-
molecular-weight heparin or unfractionated heparin. N Engl J Med 1995; 332:
1330–1335.
Warmerdam PAM, van de Winkel JGJ, Gosselin EJ, Capel PJA. Molecular basis for
a polymorphism of human Fcg receptor II (CD32w). J Exp Med 1990; 172:19–
25.
Warmerdam PAM, van de Winkel JGJ, Vlug A, Westerdaal NAC, Capel PJA. A
single amino acid in the second domain of the human Fcg receptor II is critical
for human IgG2 binding. J Immunol 1991; 147:1338–1343.
250 Denomme
Wu J, Edberg JC, Redecha PB, Bansal V, Guyre PM, Coleman K, Salmon JE,
Kimberly RP. A novel polymorphism of FcgRIIIa (CD16) alters receptor
function and predisposes to autoimmune disease. J Clin Invest 1997; 100:1059–
1070.
Yanaga F, Poole A, Asselin J, Blake R, Schieven GL, Clark EA, Law CL, Watson SP.
Syk interacts with tyrosine-phosphorylated proteins in human platelets acti-
vated by collagen and cross-linking of the Fc gamma-IIA receptor. Biochem J
1995; 313:471–478.
10
Immune Vascular Injury in Heparin-
Induced Thrombocytopenia
Gowthami M. Arepally
Duke University Medical Center, Durham, North Carolina, U.S.A.
Mortimer Poncz
University of Pennsylvania School of Medicine and Children’s Hospital
of Philadelphia, Philadelphia, Pennsylvania, U.S.A.
Douglas B. Cines
University of Pennsylvania School of Medicine, Philadelphia,
Pennsylvania, U.S.A.
I. INTRODUCTION
The most important complication of HIT is thrombosis. Clinically overt

arterial or venous thrombi have been observed in 50% or more of thrombo-
cytopenic patients with HIT in some series (see Chaps. 3 and 4), a frequency
that far exceeds any other drug-induced immune platelet disorder. The
propensity for thrombosis has been attributed to platelet activation through
FcgIIA receptors by IgG-containing complexes that comprise heparin and
platelet factor 4 (PF4) (see Chap. 9). Strong support for this notion comes
from recent studies involving mice transgenic for human FcgIIA and human
PF4 (Reilly et al., 2001) (see Chap. 9). However, the ‘‘nonpermissive’’ FcgIIA
phenotype affords limited protection against thrombosis (Arepally et al.,
1997; Denomme et al., 1997; Bachelot-Loza et al., 1998; Carlsson et al., 1998),
perhaps because of the high incidence of IgG1 antibodies in HIT patients
(Denomme et al., 1997). Furthermore, the occasional patient in whom only
IgM or IgA HIT antibodies are detected (Amiral et al., 1996) suggests that
additional factors, acting at the level of the platelet or elsewhere, must
251
252 Arepally et al.
contribute to the thrombotic diathesis. In addition, thrombi usually occur at a

limited number of vessel sites, typically large arteries and veins (Boshkov et
al., 1993) (see Chap. 3), even though circulating HIT antibodies and activated
platelets can be readily detected in asymptomatic patients. These consider-
ations point to injury to the local vascular milieu in influencing the clinical
expression of disease.
There are several other reasons to suspect that immune complex–
mediated injury of the vasculature may contribute to the thrombotic com-
plications of HIT. First, the endothelium helps maintain the balance between
blood fluidity and clotting. Second, the endothelium expresses heparan
sulfate-containing proteoglycans that help regulate coagulation and contrib-
ute to the metabolism of PF4. Third, antiendothelial cell antibodies have been
identified in patients with other disorders characterized by thrombosis and
thrombocytopenia. And fourth, there is some direct evidence that patients
with HIT form antibodies that recognize heparin–PF4 complexes on the
endothelium, and thereby promote prothrombotic reactions.
II. THE ENDOTHELIUM IN HEMOSTASIS
The role of the endothelium in regulating blood fluidity and trafficking of

circulating hematopoietic cells has been the subject of recent reviews (Bas-
senge, 1996; Bombeli et al., 1997; Cadroy et al., 1997; Lüscher and Barton,
1997; Cines et al., 1998; Aird, 2003). Endothelial cells (ECs) express a variety
of factors that inhibit coagulation. These include soluble substances, such as
nitric oxide and prostacyclin (acting to inhibit platelet activation), and tissue-
type plasminogen activator (t-PA, acting to promote fibrinolysis), among
many others. Endothelial cell surface–bound molecules with anticoagulant
activity include heparan sulfate–containing proteoglycans (see later discus-
sion), thrombomodulin, complement regulatory proteins, as well as receptors
for activated C, urokinase, and plasminogen.
Unperturbed ECs also do not express several moieties that promote
platelet and leukocyte adhesion, such as endothelial leukocyte adhesion
molecule (ELAM), P-selectin, and platelet-activating factor (PAF). These
can be induced, however, when the cells are stimulated by agonists, such as
cytokines or endotoxin (Drake et al., 1993), or when the cells are injured by
immune factors, atherosclerosis, or shear stress. Additionally, ECs exposed to
such factors express a reduced content of heparan sulfate, internalize and
degrade activated protein C, elaborate tissue factor, and secrete abundant
plasminogen activator inhibitor, each of which may promote thrombus
formation (Cines et al., 1998). Histochemical studies of the endothelium in
Immune Vascular Injury in HIT 253
murine models of inflammation have confirmed many of these observations,

which have been predicated on cell culture data (Fries et al., 1993), affirming
the notion that the endothelium undergoes a multifaceted change from an
antithrombotic to a procoagulant phenotype in response to injury.
Also relevant to the pathogenesis of HIT is the remarkable heteroge-
neity of ECs, within and among different vascular beds, owing to genetic
differences and acquired changes in phenotype (for review see Cines et al.,
1998). For example, only a small fraction of ECs constitutively express t-PA
or urokinase-type plasminogen activator in vivo (Levin and Del Zoppo,
1994), whereas a different subset express tissue factor when exposed to
endotoxin (Drake et al., 1993). ECs from different organs express tissue-
specific promotes that regulate the expression of von Willebrand factor
(vWF) in vivo (Aird et al., 1997). ECs also show regional variation in the
synthesis of prostacyclin, expression of leukocyte adhesion molecules and Fcg
receptors, among many other phenotypic differences.
There is also recent evidence to indicate that protein C activation on
macrovascular ECs is mediated predominantly through the protein C recep-
tor, whereas thrombomodulin (TM) may be the dominant system in the
microvasculature (Laszik et al., 1997; Fukudome et al., 1998). TM changes
thrombin from a procoagulant to an anticoagulant enzyme (i.e., TM-bound
thrombin activates the natural anticoagulant zymogen, protein C) (Esmon,
1989). The anticoagulant function of TM in the microvasculature may
contribute to the pathogenesis of warfarin-associated venous limb gangrene
that can complicate HIT. This syndrome has been attributed to the coinci-
dence of persistent thrombin generation and acquired protein C deficiency
that may occur during the first few days of anticoagulation with warfarin
(Warkentin et al., 1997) (see Chaps. 3 and 12).
The behavior of ECs can also be modified during the evolution of
vascular disease. For example, atherosclerotic vessels produce less nitric oxide
in response to a variety of stimuli than do healthy vessels (Gryglewski, 1995).
Atherosclerotic vessels may also undergo alterations in their expression of
glycosaminoglycans (GAGs) (Williams and Tabas, 1995) and an increase in
expression of various cell adhesion molecules (for review see Fuster et al.,
1998). The binding of advanced glycation end products to specific EC
receptors during normal aging and diabetes mellitus increases vascular
permeability, exposing the subendothelial matrix to lipoproteins and other
injurious substances (Schmidt et al., 1994). It is also likely that genetic
variation in EC behavior contributes to the host response to antibody- or
platelet-mediated EC injury, although the methods to identify or monitor
such risk factors remain to be developed. Thus, any inquiry into the reason
why only a subset of patients who develop antiheparin–PF4 antibodies
develop thrombosis, or why thrombi occur at restricted vascular sites, must
254 Arepally et al.
take into consideration the specific attributes of a particular endothelial

vascular bed.
III. IMMUNE ENDOTHELIAL CELL INJURY
Endothelial cell–reactive antibodies have been found in patients suffering

from disorders characterized by vasculitis or thrombosis (for review see
Meroni et al., 1996; Praprotnik et al., 2001). The best-studied example is
allograft rejection, a setting in which there is extensive evidence for anti-
endothelial cell antibodies (AECA), in part directed at carbohydrate antigens
that regulate procoagulant activity in vitro (Saadi and Platt, 1995; Siegel et al.,
1997). AECA have also been identified in patients with hyperacute and acute
Figure 1 Model of HIT antibody interactions with endothelial cells: (1) HIT
antibodies bind to circulating antigen that becomes localized to platelets. (2) Platelet
activation occurs after Fc receptor binding, leading to platelet granule release. ADP
is released from platelet dense granules and PF4 is released from platelet a-granules.
(3) Released PF4 binds to platelets and endothelial cell heparan sulfate (HS), dis-
placing antithrombin (AT) from endothelial cells. (4) Antigen complexes on endo-
thelial cells bind HIT antibodies. (5) HIT antibody binding to endothelial cells leads
to endothelial cell activation, resulting in further platelet activation.
graft rejection, systemic lupus erythematosus (Cines et al., 1984), antiphos-

pholipid antibody syndrome (McCrae et al., 1991), and hemolytic-uremic
syndrome (Leung et al., 1988). What is curious is the extraordinary diversity
of the clinical syndromes associated with AECA. Also of interest, the target
cells used in most assays (i.e., endothelial cells derived from human umbilical
vein [HUVEC]), are not known to be affected by immune vascular injury in
the clinical setting. This suggests that the expression of the target antigen(s) is
highly restricted in vivo, reflecting either regional differences in the compo-
sition of the vascular bed (microvascular or macrovascular) or, perhaps,
indicating distinct responses to injury and inflammation. Alternatively, these
AECA could represent a surrogate marker for other pathogenic antibodies,
or the capacity of the affected vasculature is a critical determinant of whether
thrombosis develops.
Several effects of AECA that could contribute to thrombosis include cell
lysis or apoptosis (Bordron et al., 1998; Nakamura et al., 1998), induction of
cytokine secretion and promotion of leukocyte adhesion (Del Papa et al.,
1997), enhancement of vascular permeability (Saadi and Platt, 1995), accel-
eration of procoagulant reactions (Tannenbaum et al., 1986; Saadi et al.,
1995), and reduction in the expression of heparan sulfate (Platt et al., 1991).
The reason why some antibodies promote thrombosis (e.g., HIT antibodies),
whereas others induce primarily a necrotizing vasculitis, remains unresolved,
but may relate directly to the biological functions of heparin and PF4 (Fig. 1).
IV. HEPARAN SULFATE–CONTAINING PROTEOGLYCANS,

HEPARIN, AND THE ENDOTHELIUM
The expression and anticoagulant function of heparan sulfate-type proteo-

glycans (HSPGs) by ECs may be central to the pathogenesis of vascular
thrombosis in patients with HIT. The biochemistry and function of these
GAGs and proteoglycans to which they bind has been the subject of extensive
study (for review see Ruoslahti, 1988; Rosenberg et al., 1997; Esko and
Lindahl, 2001); the involvement of heparan sulfate in the development of HIT
is considered elsewhere (see Chap. 8). HSPGs expressed by ECs bind
antithrombin in vitro and in vivo, and accelerate the inactivation of thrombin
and factor Xa approximately 20-fold (Marcum et al., 1984), an effect that is
biologically equivalent to 0.1–0.5 U/mL of heparin (Marcum and Rosenberg,
1987). Yet less than 1% of the HSPGs isolated from cultured ECs express
anticoagulant activity (Marcum and Rosenberg, 1984). Active species are
characterized by an approximately twofold enrichment in glucoronyl 3-O-
sulfated glucosamine residues, compared with inactive species (Marcum and
Rosenberg, 1987). Interestingly, targeted deletion of the murine 3-O-sulfo-
256 Arepally et al.
transferase-1 enzyme, the enzyme responsible for generating this anticoagu-

lant modification of HSPGs, does not lead to a prothrombotic phenotype
(HajMohammadi et al., 2003). The physiological mechanisms that control the
synthesis and postsynthetic modifications of HSPG remain an active area of
investigation (Forsberg and Kjellen, 2001).
Microheterogeneity in the composition of HSPG in arteries, veins,
and capillaries has been noted (Lowe-Krentz and Joyce, 1991), but the sig-
nificance of these differences is unknown. Expression of HSPG by ECs
undergoes developmental changes (David et al., 1992), and its composition
varies after the cells are exposed to thrombin (Benezra et al., 1993), homo-
cysteine (Nishinaga et al., 1993), heparin (Nader et al., 1989), wounding and
migration (Kinsella and Wight, 1986), and after induction by activated
platelets (Yahalom et al., 1984), among other stimuli. ECs also bind heparin
(for review see Patton et al., 1995), which alters EC proliferation, matrix
composition, and many other vascular functions. It has also been reported
that antithrombin is displaced from ECs by heparin, and its binding is
inhibited by PF4 (Stern et al., 1985). Whether HIT antibodies promote the
capacity of PF4 to neutralize antithrombin activity has not been reported.
V. PLATELET FACTOR 4 AND THE ENDOTHELIUM
The biochemistry of PF4 and its involvement in HIT is reviewed elsewhere

(see Chap. 7). The metabolism of the protein is regulated by its interactions
with the endothelium. PF4 is stored in the a-granules of platelets as a tetramer
bound to chondroitin sulfate (Barber et al., 1972). The tetramer may dis-
sociate from the glycosaminoglycan as the platelets are activated, but more
likely, dissociation occurs subsequent to binding to EC HSPG, which
contains a higher charge density. [125I]PF4 is cleared from the circulation
with an a-elimination phase approximating 2 min, which primarily represents
binding to the endothelium, and a h-elimination phase approximating 40 min,
corresponding to uptake and degradation, predominantly by hepatocytes
(Rucinski et al., 1986, 1990).
The ECs bind about 50 pmol PF4/105 cells (Rybak et al., 1989). Several
classes of binding sites have been identified, including a high-capacity, low-
affinity site on HSPG, as well as higher-affinity binding sites involving specific
chemokine receptors and certain coagulant proteins (see later discussion).
Binding of PF4 to the endothelium is attenuated by pretreatment with hepa-
rinase (Marcum et al., 1984), and plasma concentrations are increased 10- to
20-fold after heparin is infused intravenously (Dawes et al., 1982). Binding of
PF4 to EC GAGs is electrostatic (Wu and Cohen, 1984) and is independent of
the pentasaccharide involved in the binding of antithrombin (Loscalzo et al.,
1985). The affinity of PF4 binding to ECs is lower than to purified heparin (Kd
= 2 nM vs. 2–3 AM, respectively) (Rybak et al., 1989), consistent with the
biochemical hetrogeneity of vascular matrix. PF4 has a 10- to 100-fold greater
affinity for EC HSPG than does antithrombin (Jordan et al., 1982) and
markedly attenuates its antiprotease cofactor activity on intact vessels (Busch
et al., 1980; Stern et al., 1985).
PF4 also binds to the chemokine receptor, Duffy (Hadley et al., 1994),
which has been identified on ECs in postcapillary venules and in the
splanchnic bed, even in individuals who do not express the antigen on their
erythrocytes (Peiper et al., 1995). The distribution of Duffy on ECs in other
vascular beds is less well studied. The binding site on PF4 for Duffy has not
been deduced, and the capacity of Duffy-bound PF4 to bind heparin or HIT
antibodies has not been reported. PF4 does not appear to bind to any of the
other members of the chemokine receptor family (for review see Baggiolini et
al., 1997), including CXCR4, the only other related receptor yet identified on
ECs (Volin et al., 1998). However, PF4 exposed to leukocyte lysates is cleaved
between Thr-16 and Ser-17 to yield a peptide with 30-fold greater biological
activity than its parent molecule (Gupta et al., 1995). Such proteolysis may
enable modified PF4 to bind to CXCR1 or R2 (formerly called IL-8R-a and -
h, respectively) (for review see Baggiolini et al., 1997), but no evidence has
been presented that ECs express either receptor, nor has binding of modified
PF4 to CXCR4 been shown.
The CXC chemokines that contain a Glu-Leu-Arg (ELR) NH2-termi-
nal sequence induce angiogenesis, whereas those that do not, such as PF4 and
gIP-10, inhibit angiogenesis (Baggiolini and Moser, 1997). The molecular
basis by which PF4 inhibits angiogenesis is unclear. Considering that ECs lack
CXCR1 and CXCR2, the effect of PF4 may be indirect. PF4 interrupts the
binding of heparin-binding proteins, such as vascular endothelial cell growth
factor (VEGF) and basic fibroblast growth factor (hFGF) to their receptors
(Gengrinovitch et al., 1995; Peng et al., 1997). However, PF4 also inhibits
angiogenesis stimulated by a truncated version of VEGF that lacks the hepa-
rin-binding domain (Gengrinovitch et al., 1995). Whether PF4 influences the
behavior of ECs from disordered vessels in other ways, or whether the PF4-
mediated signal transduction through chemokine receptors is modulated by
HIT antibodies, deserves additional study.
PF4 binds preferentially to ECs at sites of angiogenesis (Hansell et al.,
1995). For example, in a hamster cheek pouch model, [125I]PF4 is taken up
preferentially by ECs within the neovasculature where its binding is inhibited
by the related CXC chemokine, gIP-10 (Luster et al., 1995). PF4 inhibits cell
proliferation in vitro (Sharpe et al., 1990; Maione et al., 1991; Gupta and
Singh, 1994) and tumor-induced neovascularization (Sharpe et al., 1990;
Maione et al., 1991) in vivo. Both the heparin-binding domain (Maione et al.,
258 Arepally et al.
1990; Gupta and Singh, 1994) and other portions of the molecule (Maione et
al., 1991; Gupta and Singh, 1994; Gupta et al., 1995; Lecomte-Raclet et al.,
1998) have been implicated.
PF4 also binds to thrombomodulin (TM), a 60.3 kDa protein consti-
tutively expressed on the surface of ECs. Binding of thrombin to TM alters its
substrate specificity, such that proteolytic cleavage of protein C is accelerated
20,000-fold (Esmon, 1989). TM is posttranslationally modified by association
with a chondroitin sulfate A–like GAG, which invests it with the capacity to
bind cationic peptides at physiological pH. The binding of eosinophilic
cationic protein, major basic protein, and histidine-rich glycoprotein to these
GAG residues inhibits the function of TM, whereas the binding of PF4 (but
not h-thromboglobulin or thrombospondin) increases protein C-cofactor
activity 25-fold in a cell-free system (Slungaard and Key, 1994; Dudek et al.,
1997). The results of one recent study suggest that PF4 may exert a
physiogically relevant anticoagulant effect (Slungaard et al., 2003). Addition
of PF4 to cultured endothelial cells accelerates APC generation approximate-
ly 5- to 10-fold depending on vascular origin. Injection of PF4 into primates
infused with thrombin increases APC generation 2- to 3-fold and prolongs the
baseline aPTT. Additional studies should clarify whether HIT antibodies
interfere with the anticoagulant function of PF4 and thereby may predispose
to warfarin-associated venous limb gangrene.
Recent in vitro and in vivo studies suggest potential mechanisms
whereby PF4 may play a vital role in promoting atherogenesis. Human ath-
erosclerotic lesions are invested with PF4 (Pitsilos et al., 2003). PF4 is found
not only along the overlying endothelium, but also in foam cells and acellular
portions of the plaque. In vitro, PF4 binds to the LDL receptor and to
proteoglycans, forming ternary complexes that show limited migration into
clathrin-coated pits, thereby retarding endocytosis and catabolism of LDL
(Sachais et al., 2002). PF4 also binds directly to oxidized LDL, promoting
foam cell formation (Nassar et al., 2003). In mice, activated platelets deposit
PF4 on endothelium and monocytes, potentiating effects of P-selectin on plate-
let-leukocyte aggregate formation and atherosclerotic development (Huo et
al., 2003). Antibodies to heparin–PF4 complexes have recently been identified
as an independent predictor of myocardial infarction at 30 days in patients
presenting with acute coronary ischemic syndromes (Williams et al., 2003).
Thus, HIT antibodies may modify the interactions of PF4 with diseased
endothelium by (1) binding to PF4/proteoglycan complexes in atherosclerotic
lesions, (2) inducing formation of platelet-leukocyte aggregates (Khairy et al.,
2001), or (3) binding to circulating monocytes (Pouplard et al., 2001; Arepally
and Mayer, 2001), thereby increasing local inflammation and stimulating
procoagulant processes.
VI. EVIDENCE FOR ENDOTHELIAL CELL

ANTIBODIES IN HIT
There are limited experimental data to indicate that EC-reactive antibodies or

immune complexes contribute to the development of thrombosis in patients
with HIT. Over 10 years ago, one group reported that sera from essentially all
patients with HIT deposit increased amounts of IgG, IgM, or IgA on
HUVEC (Cines et al., 1987). Binding was reduced when the cells were
pretreated with enzymes that degrade heparin or heparan sulfate, whereas
addition of chondroitinase was without effect. HIT sera induced ECs to
express the procoagulant tissue factor, and the expression of procoagulant
activity was enhanced further in the presence of platelets. These observations
were confirmed and extended by Visentin and colleagues (1994), who
demonstrated that the binding of HIT antibodies of HUVEC was dependent
on PF4, but not on exogenous heparin, in contrast to the requirement for both
to be added for antibody binding to platelets (Fig. 2). This is consistent with
the concept that PF4, released from activated platelets, can form a competent
antigenic complex on the pericellular matrix of the endothelium.
The role of EC-reactive antibodies was also explored in an animal model
that simulates certain aspects of HIT (Blank et al., 1997). Mice injected with
IgG fractions obtained from HIT patients developed anti-idiotypic antibodies
that recognized complexes between human PF4 and heparin. Furthermore,
the anti-idiotypic antibodies competed with the immunizing antibodies for
binding to the antigenic complex. These effects were noted when antibodies
obtained from mice immunized with control IgG were studied (Fig. 3A).
Additionally, mice immunized with HIT–IgG developed thrombocytopenia
when exposed to heparin. Affinity-purified antiheparin–PF4 antibodies
bound to murine endothelioma cells in the presence of PF4, but not h2GPI
(Fig. 3B). Of interest, immunized mice did not develop overt thrombi on
exposure to heparin, possibly because of insufficient circulating PF4, intrinsic
differences in the balance between the procoagulant and fibrinolytic systems
compared with humans, or differences in signal transduction through murine
and human platelet Fcg receptors. However, it is also possible that the
difference lies in a reduced capacity of otherwise healthy mouse ECs to
respond to the procoagulant stimulus induced by these antibodies, compared
with the responsiveness of patients with HIT, who, in the main, comprise a
more elderly population with underlying vascular disorders.
Recent studies suggest that requirements for endothelial cell activation
by HIT antibodies may differ based on the vascular site. According to
preliminary observations by Blank et al. (2002), anti-PF4/heparin IgG can
effect activation of endothelial cells directly in microvascular sites, such as
260 Arepally et al.
Figure 2 Binding of IgG (A–D) and IgM (E) from the plasma of a patient with
HIT to cultured human umbilical vein endothelial cells (HUVEC): (A) Plasma alone;
(B) plasma plus 0.05 U/mL heparin; (C) plasma plus 10 Ag/mL PF4; (D) plasma plus
0.05 U/mL heparin plus 10 Ag/mL PF4; (E) binding of IgM from plasma containing
10 Ag/mL PF4. Both IgG (C) and IgM (E) bound to HUVEC in the presence of PF4
alone. The binding of each was completely inhibited by heparin. (From Visentin et
al., 1994.)
bone marrow. However, for macrovascular endothelial cells, such as

HUVECs, either antibody-mediated platelet activation (Hebert et al., 1998)
or stimulation with cytokines such as TNF-a (Blank et al., 2002) appear to be
necessary for activation of ECs. In this latter setting, sera or IgG from patients
with HIT induced various changes in the behavior of HUVEC in the presence
of platelets or TNF-a, including increased expression of E-selectin, VCAM,
ICAM-1, and tissue factor, release of IL-1h, IL-6, TNF-a, and PAI-1, and
adhesion of platelets or monocytes to the activated ECs. EC activation was
inhibited by SR121566a (a platelet glycoprotein IIb/IIIa antagonist) and, to
some extent, by apyrase and ATPgS, implicating expression of endogenous
Figure 3 Antiendothelial antibodies in sera of mice immunized with HIT-IgG: (A)
Sera from mice immunized with IgG fractions of two patients with heparin-induced
thrombocytopenia, HIT-1 (9) and HIT-2 (x) or IgG from an individual not exposed
to heparin, NC (n), were tested for binding to murine endothelial cells by ELISA.
(B). Antiendothelial cell binding of affinity-purified anti-PF4/heparin IgG from sea
of immunized mice. Affinity-purified IgG (10 Ag/mL) was tested for endothelial cell
binding in the absence or presence of 10 Ag PF4 or h2-glycoprotein-1 (h2-GP I).
(From Blank et al., 1997.)
262 Arepally et al.
platelet fibrin(ogen) and release of ADP in the process. HIT sera/IgG exerted
little effect on the behavior of HUVEC in the absence of prestimulation by
platelets or cytokines, even in the presence of exogenous PF4. The mecha-
nism(s) by which platelet or cytokine activation facilitate binding of HIT
antibodies to cell-associated heparin/PF4 requires further investigation.
VII. IMMUNE VASCULAR INJURY AND MONOCYTE

ACTIVATION IN HIT-ASSOCIATED THROMBOSIS
The studies outlined above support the notion that platelet activation and/or
an inflammatory milieu contribute to endothelial cell dysfunction, predispos-
ing to HIT-associated thrombosis (HITT). This view of HIT as an inflam-
matory disorder has gained experimental support through recent findings of
monocyte activation. HIT plasma or IgG stimulates monocytes to elaborate
tissue factor–dependent procoagulant activity in monocytes (Pouplard et al.,
2001). This procoagulant effect required small amounts of heparin to acti-
vate the monocytes in whole blood, but appeared to be heparin-independent
when isolated mononuclear cells were studied, presumably by exposing cell-
associated proteoglycans (Pouplard et al., 2001). Heparin-independent upreg-
ulation of monocyte tissue factor activity by HITT antibodies was confirmed
in a second study using human and murine antibodies (Arepally and Mayer,
2001). In this latter study, maximal tissue factor expression was detected at 4–
6 hours, suggesting that new synthesis, rather than de-encryption, was
primarily responsible for increased procoagulant activity. Activation of
monocyes likely requires ligation of cellular FcgRII receptors, as signaling
intermediates downstream of this receptor become phosphorylated in the
presence of HITT antibody (Ma and Arepally, 2002). The development of
monoclonal antibodies against PF4 and PF4/heparin (Arepally et al., 2000)
and the availability of a murine model of HITT should help to delineate the
contribution of monocyte activation to the immune pathogenesis of throm-
bocytopenia and thrombosis.
VIII. PERSPECTIVE AND FUTURE DIRECTIONS
Heparin-induced thrombocytopenia continues to pose several enigmas in-

cluding the fundamental issue of how heparin induces the formation of self-
reactive antibodies to a native protein in such a high proportion of immuno-
logically competent individuals (Visentin et al., 1996; Bauer et al., 1997). It
also remains unclear why only a subset of patients with antiheparin–PF4

antibodies develop thrombocytopenia, and fewer still develop thrombosis. It
is possible that characteristics of HIT antibodies, such as their subtype,
specificity, and affinity for portions of the PF4 molecule, may provide some
of the answers. However, it is also likely that part of the propensity for throm-
bosis, and the localization of clotting observed in HIT, relate to antigen ex-
pression and response to injury at the level of the vessel wall itself.
REFERENCES
Aird WC. The role of endothelium in severe sepsis and multiple organ dysfunction
syndrome. Blood 2003; 101:3765–3777.
Aird WC, Edelberg JM, Weiler-Guettler H, Simmons WW, Smith TW, Rosenberg
RD. Vascular bed-specific expression of an endothelial cell gene is programmed
by the tissue microenvironment. J Cell Biol 1997; 138:1117–1124.
Amiral J, Wolf M, Fischer A-M, Boyer-Neumann C, Vissac A-M, Meyer D. Patho-
with heparin-induced thrombocytopenia. Br J Haematol 1996; 92:954–959.
Arepally G, McKenzie SE, Jiang X-M, Poncz M, Cines DB. FcgRIIA H/R131
Arepally GM, Kamei S, Park KS, Kamei K, Li ZQ, Liu W, Siegel DL, Kisiel W, Cines
DB, Poncz M. Characterization of a murine monoclonal antibody that mimics
heparin-induced thrombocytopenia antibodies. Blood 2000; 95:1533–1540.
cytopenia stimulate monocytic cells to express tissue factor and secrete inter-
leukin-8. Blood 2001; 98:1252–1254.
the FcgRIIa-Arg/His-131 polymorphism in heparin-induced thrombocytope-
Baggiolini M, Moser B. Blocking chemokine receptors. J Exp Med 1997; 186:1189–
1191.
Baggiolini M, Dewald B, Moser B. Human chemokines: an update. Annu Rev
Immunol 1997; 15:675–705.
Barber AJ, Käser-Glanzmann R, Jakábová M, Lüscher EF. Characterization of
a chondroitin 4-sulfate proteoglycan carrier for heparin neutralizing activity
(platelet factor 4) released from human blood platelets. Biochim Biophys Acta
1972; 286:312–329.
Bassenge E. Endothelial function in different organs. Prog Cardiovasc Dis 1996; 39:
209–228.
McNulty S, Amiral J, Hauck WW, Edie RN, Mannion JD. Prevalence of hepa-
264 Arepally et al.
rin-associated antibodies without thrombosis in patients undergoing cardio-

pulmonary bypass. Circulation 1997; 95:1242–1246.
Benezra M, Vlodavsky I, Ishai-Michaeli R, Neufeld G, Bar-Shavit R. Thrombin-
induced release of active basic fibroblast growth factor–heparan sulfate com-
plexes from subendothelial extracellular matrix. Blood 1993; 81:3324–3331.
Blank M, Cines DB, Arepally G, Eldor A, Afek A, Shoenfeld Y. Pathogenicity of
human anti-platelet factor 4 (PF4)/heparin in vivo: generation of mouse anti-
PF4/heparin and induction of thrombocytopenia by heparin. Clin Exp Immu-
nol 1997; 108:333–339.
Blank M, Shoenfeld Y, Tavor S, Praprotnik S, Boffa MC, Weksler B, Walenga MJ,
Amiral J, Eldor A. Anti-platelet factor 4/heparin antibodies from patients with
heparin-induced thrombocytopenia provoke direct activation of microvascular
endothelial cells. Int Immunol 2002; 14:121–129.
Bombeli T, Mueller M, Haeberli A. Anticoagulant properties of the vascular endo-
thelium. Thromb Haemost 1997; 77:408–423.
Bordron A, Dueymes M, Levy Y, Jamin C, Leroy J-P, Piette J-C, Shoenfeld Y,
Youinou PY. The binding of some human antiendothelial cell antibodies in-
duces endothelial cell apoptosis. J Clin Invest 1998; 101:2029–2035.
Boshkov LK, Warkentin TE, Hayward CP, Andrew M, Kelton JG. Heparin-induced
thrombocytopenia and thrombosis: clinical and laboratory studies. Br J Hae-
matol 1993; 84:322–328.
Busch C, Dawes J, Pepper DS, Wasteson A. Binding of platelet factor 4 to cultured
human umbilical vein endothelial cells. Thromb Res 1980; 19:129–137.
Cadroy Y, Diquélou A, Dupouy D, Bossavy JP, Sakariassen KS, Sié P, Boneu B. The
thrombomodulin/protein C/protein S anticoagulant pathway modulates the
thrombogenic properties of the normal resting and stimulated endothelium.
Arterioscler Thromb Vasc Biol 1997; 17:520–527.
NAC, Kiefel V, van der Winkel JGL, Graincher A. Heparin-induced
thrombocytopenia: new insights into the impact of the FcgRIIa–R-H131 poly-
morphism. Blood 1998; 92:1526–1531.
Cines DB, Lyss AP, Reeber M, Bina M, DeHoratius RJ. Presence of complement-
fixing anti-endothelial cell antibodies in systemic lupus erythematosus. J Clin
Invest 1984; 73:611–625.
Cines DB, Pollak ES, Buck CA, Loscalzo J, Zimmerman GA, McEver RP, Pober JS,
Wick TM, Konkle BA, Schwartz BS, Barnathan ES, McCrae KR, Hug BA,
Schmidt A-M, Stern DM. Endothelial cells in physiology and in the patho-
physiology of vascular disorders. Blood 1998; 91:3527–3561.
David G, Bai XM, van der Schueren B, Cassiman JJ, van den Berghe H. Develop-
mental changes in heparan sulfate expression: in situ detection with MAbs.
J Cell Biol 1992; 119:961–975.
Dawes J, Pumphrey CW, McLaren KM, Prowse CV, Pepper DS. The in vivo release of
human platelet factor 4 by heparin. Thromb Res 1982; 27:65–76.
Del Papa N, Guidali L, Sala A, Buccellati C, Khamashta MA, Ichikawa K, Koike T,

Balestrieri G, Tincani A, Hughes GRV, Meroni PL. Endothelial cells as target
for antiphospholipid antibodies. Human polyclonal and monoclonal anti-h2-
glycoprotein I antibodies react in vitro with endothelial cells through adherent
h2-glycoprotein I and induce endothelial activation. Arthritis Rheum 1997; 40:
551–561.
Denomme GA, Warkentin TE, Horsewood P, Sheppard J-AI, Warner MN, Kelton
JG. Activation of platelets by sera containing IgG1 heparin-dependent anti-
bodies: an explanation for the predominance of the FCgRIIa ‘‘low responder’’
(his131) gene in patients with heparin-induced thrombocytopenia. J Lab Clin
Med 1997; 130:278–284.
Drake TA, Cheng J, Chang A, Taylor FB Jr. Expression of tissue factor,
thrombomodulin, and E-selectin in baboons with lethal Escherichia coli sepsis.
Am J Pathol 1993; 142:1458–1470.
Dudek AZ, Pennell CA, Decker TD, Young TA, Key NS, Slungaard A. Platelet factor
4 binds to glycanated forms of thrombomodulin and to protein C. A potential
mechanism for enhancing generation of activated protein C. J Biol Chem 1997;
272:31785–31792.
Esko JD, Lindahl U. Molecular diversity of heparan sulfate. J Clin Invest 2001;
108:169–173.
Eslin DE, Zhang C, Samuels KJ, Gewirtz JE, Cines DB, Poncz M, Kowalska MA.
Platelet Factor 4 (PF4) has a in vivo role in arterial thrombosis: implications for
heparin therapy [abstr]. Blood 2002; 100:124a.
Esmon CT. The roles of protein C and thrombomodulin in the regulation of blood
coagulation. J Biol Chem 1989; 264:4743–4746.
Forsberg E, Kjellén L. Heparan sulfate: lessons from knockout mice. J Clin Invest
2001; 108:175–180.
Fries JWU, Williams AJ, Atkins RC, Newman W, Lipscomb MF, Collins T. Expres-
sion of VCAM-1 and E-selectin in an in vivo model of endothelial activation.
Am J Pathol 1993; 143:725–737.
Fukudome K, Ye X, Tsuneyoshi N, Tokunaga O, Sugawara K, Mizokami H, Kimoto
M. Activation mechanism of anticoagulant protein C in large blood vessels in-
volving the endothelial cell protein C receptor. J Exp Med 1998; 187:1029–1035.
Fuster V, Poon M, Willerson JT. Learning from the transgenic mouse. Endothelium,
adhesive molecules, and neointimal formation. Circulation 1998; 97:16–18.
Gengrinovitch S, Greenberg SM, Cohen T, Gitay-Goren H, Rockwell P, Maione TE,
Levi B-Z, Neufeld G. Platelet factor-4 inhibits the mitogenic activity of
VEGF121 and VEGF165 using several concurrent mechanisms. J Biol Chem
1995; 270:15059–15065.
Goldberg ID, Stemerman MB, Handin RI. Vascular permeation of platelet factor 4
after endothelial injury. Science 1980; 209:611–612.
Gryglewski RJ. Endothelial nitric oxide, prostacyclin (PGI2) and tissue plasminogen
activator (t-PA): alliance or neutrality? Pol J Pharmacol 1995; 47:467–472.
Gupta SK, Singh JP. Inhibition of endothelial cell proliferation by platelet factor-4 in-
volves a unique action on S phase progression. J Cell Biol 1994; 127:1121–1127.
266 Arepally et al.
Gupta SK, Hassel T, Singh JP. A potent inhibitor of endothelial cell proliferation is
generated by proteolytic cleavage of the chemokine platelet factor 4. Proc Natl
Acad Sci USA 1995; 92:7799–7803.
Hadley TJ, Lu Z-H, Wasniowska K, Martin AW, Peiper SC, Hesselgesser J, Horuk R.
Postcapillary venule endothelial cells in kidney express a multispecific chemo-
kine receptor that is structurally and functionally identical to the erythroid iso-
form, which is the Duffy blood group antigen. J Clin Invest 1994; 94:985–991.
HajMohammadi S, Enjyoji K, Princivalle M, Christi P, Lech M, Beeler D, Rayburn H,
Schwartz JJ, Barzegar S, de Agostini Al, Post MJ, Rosenberg RD, Shworak
NW. Normal levels of anticoagulant heparan sulfate are not essential for normal
hemostasis. J Clin Invest 2003; 111:989–999.
Hansell P, Maione TE, Borgström P. Selective binding of platelet factor 4 to regions of
active angiogenesis in vivo. Am J Physiol 1995; 269:H829–H836.
Hebert JM, Savi P, Jeske WP, Walenga JM. Effect of SR121566A, a potent GP IIb–
IIIa antagonist, on the HIT serum/heparin-induced platelet-mediated activa-
tion of human endothelial cells. Thromb Haemost 1998; 80:326–331.
Huo Y, Schober A, Forlow SB, Smith DF, Hyman MC, Jung S, Littman DR, Weber
C, Ley K. Circulating activated platelets exacerbate atherosclerosis in mice
deficient in apolipoprotein E. Nat Med 2003; 9:61–67.
Jordan RE, Favreau LV, Braswell EH, Rosenberg RD. Heparin with two binding sites
for antithrombin or platelet factor 4. J Biol Chem 1982; 257:400–406.
Khairy M, Lasne D, Brohard-Bohn B, Aiach M, Rendu F, Bachelot-Loza C. A new
approach in the study of the molecular and cellular events implicated in heparin-
induced thrombocytopenia. Formation of leukocyte-platelet aggregates.
Thromb Haemost 2001; 85:1090–1096.
Kinsella MG, Wight TN. Modulation of sulfated proteoglycan synthesis by bovine
aortic endothelial cells during migration. J Cell Biol 1986; 102:679–687.
Laszik Z, Mitro A, Taylor FB Jr, Ferrell G, Esmon CT. Human protein C receptor is
present primarily on endothelium of large blood vessels: implications for the
control of the protein C pathway. Circulation 1997; 96:3633–3640.
Lecomte-Raclet L, Alemany M, Sequeira-Le Grand A, Amiral J, Quentin G, Vissac
AM, Caen JP, Han ZC. New insights into the negative regulation of hema-
topoiesis by chemokine platelet factor 4 and related peptides. Blood 1998;
91:2772–2780.
Leung DYM, Moake JL, Havens PL, Kim M, Pober JS. Lytic anti-endothelial cell
antibodies in haemolytic-uraemic syndrome. Lancet 1988; 2:183–186.
Levin EG, Del Zoppo GJ. Localization of tissue plasminogen activator in endothelium
of a limited number of vessels. Am J Pathol 1994; 144:855–861.
Loscalzo J, Melnick B, Handin RI. The interaction of platelet factor four and
glycosaminoglycans. Arch Biochem Biophys 1985; 240:446–455.
Lowe-Krentz LJ, Joyce JG. Venous and aortic porcine endothelial cells cultured under
standard conditions synthesize heparan sulfate chains which differ in charge.
Anal Biochem 1991; 193:155–163.
Lüscher TF, Barton M. Biology of the endothelium. Clin Cardiol 1997; 20(suppl 2):II-
3–II-10.
Luster AD, Greenberg SM, Leder P. The IP-10 chemokine binds to a specific cell
surface heparan sulfate site shared with platelet factor 4 and inhibits endothelial
cell proliferation. J Exp Med 1995; 182:219–231.
Ma AD, Arepally G. HIT antibodies and monocyte signaling [abstr]. Blood 2002;
100:16a.
Maione TE, Gray GS, Petro J, Hunt AJ, Donner AL, Bauer SI, Carson HF, Sharpe
RJ. Inhibition of angiogenesis by recombinant human platelet factor-4 and
related peptides. Science 1990; 247:77–79.
Maione TE, Gray GS, Hunt AJ, Sharpe RJ. Inhibition of tumor growth in mice by an
analogue of platelet factor 4 that lacks affinity for heparin and retains potent
angiostatic activity. Cancer Res 1991; 51:2077–2083.
Marcum JA, Rosenberg RD. Anticoagulantly active heparin-like molecules from
vascular tissue. Biochemistry 1984; 23:1730–1737.
Marcum JA, Rosenberg RD. Anticoagulantly active heparan sulfate proteoglycan and
the vascular endothelium. Semin Thromb Hemost 1987; 13:464–474.
Marcum JA, McKenney JB, Rosenberg RD. The acceleration of thrombin–
antithrombin complex formation in rat hindquarters via heparinlike molecules
bound to the endothelium. J Clin Invest 1984; 74:341–350.
McCrae KR, DeMichele A, Samuels P, Roth D, Kuo A, Meng Q-H, Rauch J, Cines
DB. Detection of endothelial cell-reactive immunoglobulin in patients with anti-
phospholipid antibodies. Br J Haematol 1991; 79:595–605.
Meroni PL, Del Papa N, Borghi MO. Antiphospholipid and antiendothelial
antibodies. Int Arch Allergy Immunol 1996; 111:320–325.
Nader HB, Buonassisi V, Colburn P, Dietrich CP. Heparin stimulates the synthesis
and modifies the sulfation pattern of heparan sulfate proteoglycan from
endothelial cells. J Cell Physiol 1989; 140:305–310.
Nakamura N, Ban T, Yamaji K, Yoneda Y, Wada Y. Localization of the apoptosis-
inducing activity of lupus anticoagulant in an annexin V-binding subset. J Clin
Invest 1998; 101:1951–1959.
Nassar T, Sachais BS, Akkawi S, Kowalska MA, Bdeir K, Leitersdorf E, Hiss E,
Ziporen L, Aviram M, Cines D, Poncz M, Higazi AA. Platelet factor 4 enhances
the binding of oxidized low-density lipoprotein to vascular wall cells. J Biol
Chem 2003; 278:6187–6193.
Nishinaga M, Ozawa T, Shimada K. Homocysteine, a thrombogenic agent, suppresses
anticoagulant heparan sulfate expression in cultured porcine aortic endothelial
cells. J Clin Invest 1993; 92:1381–1386.
Patton WA II, Granzow CA, Getts LA, Thomas SC, Zotter LM, Gunzel KA, Lowe-
Krentz LJ. Identification of a heparin-binding protein using monoclonal anti-
bodies that block heparin binding to porcine aortic endothelial cells. Biochem J
1995; 311:461–469.
Peiper SC, Wang Z-X, Neote K, Martin AW, Showell HJ, Conklyn MJ, Ogborne K,
Hadley TJ, Lu Z-H, Hesselgesser J, Horuk R. The Duffy antigen/receptor for
chemokines (DARC) is expressed in endothelial cells of Duffy negative indi-
viduals who lack the erythrocyte receptor. J Exp Med 1995; 181:1311–1317.
Peng H, Wen TC, Igase K, Tanaka J, Matsuda S, Aburaya J, Sakanaka M. Sup-
268 Arepally et al.
pression by platelet factor 4 of the myogenic activity of basic fibroblast growth

factor. Arch Histol Cytol 1997; 60:163–174.
Pitsilos S, Hunt J, Mohler ER, Prabhakar AM, Poncz M, Dawicki J, Khalapyan T,
Wolfe ML, Fairman R, Mitchell M, Carpenter J, Golden MA, Cines DB,
Sachais BS. Platelet factor 4 localization in carotid atherosclerotic plaques:
correlation with clinical parameters. Thromb Haemost. In press.
Platt JL, Lindman BJ, Geller RL, Noreen HJ, Swanson JL, Dalmasso AP, Bach FH.
The role of natural antibodies in the activation of xenogenic endothelial cells.
Transplantation 1991; 52:1037–1043.
Pouplard C, Iochmann S, Renard B, Herault O, Colombat P, Amiral J, Gruel Y.
Induction of monocyte tissue factor expression by antibodies to heparin-platelet
factor 4 complexes developed in heparin-induced thrombocytopenia. Blood
2001; 97:3300–3302.
Praprotnik S, Blank M, Meroni PL, Rozman B, Eldor A, Shoenfeld Y. Classification
of anti-endothelial cell antibodies into antibodies against microvascular and
macrovascular endothelial cells: the pathogenic and diagnostic implications.
Arthritis Rheum 2001; 44:1484–1494.
Reilly MP, Taylor SM, Hartman NK, Arepally GM, Sachais B, Cines DB, Poncz M,
McKenzie SE. Heparin-induced thrombocytopenia/thrombosis in a transgenic
mouse model requires human platelet factor 4 and platelet activation through
FcgRIIA [abstr]. Blood 2001; 98:2442–2447.
Rosenberg RD, Shworak NW, Liu J, Schwartz JJ, Zhang L. Heparan sulfate pro-
teoglycans of the cardiovascular system. Specific structures emerge but how is
synthesis regulated? J Clin Invest 1997; 100(suppl):S67–S75.
Rucinski B, Knight LC, Niewiarowski S. Clearance of human platelet factor 4 by liver
and kidney: its alteration by heparin. Am J Physiol 1986; 251:H800–H807.
Rucinski B, Niewiarowski S, Strzyzewski M, Holt JC, Mayo KH. Human platelet
factor 4 and its C-terminal peptides: heparin binding and clearance from the
circulation. Thromb Haemost 1990; 63:493–498.
Ruoslahti E. Structure and biology of proteoglycans. Annu Rev Cell Biol 1988; 4:229–
255.
Rybak ME, Gimbrone MA Jr, Davies PF, Handin RI. Interaction of platelet factor
four with cultured vascular endothelial cells. Blood 1989; 73:1534–1539.
Saadi S, Platt JL. Transient perturbation of endothelial integrity induced by natural
antibodies and complement. J Exp Med 1995; 181:21–31.
Saadi S, Holzknecht RA, Patte CP, Stern DM, Platt JL. Complement-mediated
regulation of tissue factor activity in endothelium. J Exp Med 1995; 182:1807–
1814.
Sachais BS, Kuo A, Nassar T, Morgan J, Kariko K, Williams KJ, Feldman M, Aviram
M, Shah N, Jarett L, Poncz M, Cines DB, Higazi AA. Platelet factor 4 binds to
low-density lipoprotein receptors and disrupts the endocytic machinery, result-
ing in retention of low-density lipoprotein on the cell surface. Blood 2002; 99:
3613–3622.
Schmidt AM, Mora R, Cao R, Yan SD, Brett J, Ramakrishnan R, Tsang TC,
Simionescu M, Stern D. The endothelial cell binding site for advanced glycation
end products consists of a complex: an integral membrane protein and a lacto-

ferrin-like polypeptide. J Biol Chem 1994; 269:9882–9888.
Sharpe RJ, Byers HR, Scott CF, Bauer SI, Maione TE. Growth inhibition of murine
melanoma and human colon carcinoma by recombinant human platelet factor
4. J Natl Cancer Inst 1990; 82:848–853.
Siegel JB, Grey ST, Lesnikoski B-A, Kopp CW, Soares M, am Esch JS II, Bach FH,
Robson SC. Xenogeneic endothelial cells activate human prothrombin. Trans-
plant 1997; 64:888–896.
Slungaard A, Key NS. Platelet factor 4 stimulates thrombomodulin protein C-activa-
ting cofactor activity. A structure-function analysis. J Biol Chem 1994; 269:
25549–25556.
Slungaard A, Fernandez JA, Griffin JH, Key NS, Long JR, Piegors DJ, Lentz SR.
Platelet factor 4 enhances generation of activated protein C in vitro and in vivo.
Blood 2003. In press.
Stern D, Nawroth P, Marcum J, Handley D, Kisiel W, Rosenberg R, Stern K.
Interaction of antithrombin III with bovine aortic segments. Role of heparin in
binding and enhanced anticoagulant activity. J Clin Invest 1985; 75:272–279.
Tannenbaum SH, Finko R, Cines DB. Antibody and immune complexes induce tissue
factor production by human endothelial cells. J Immunol 1986; 137:1532–1537.
with heparin:platelet factor 4 complexes. J Lab Clin Med 1996; 128:376–383.
Volin MV, Joseph L, Shockley MS, Davies PF. Chemokine receptor CXCR4 ex-
pression in endothelium. Biochem Biophys Res Commun 1998; 242:46–53.
Williams KJ, Tabas I. The response-to-retention hypothesis of early atherogenesis.
Arterioscler Thromb Vasc Biol 1995; 15:551–558.
Deliargyris EN, Sane DC. Anti-platelet factor 4/heparin antibodies: an inde-
pendent predictor of 30-day myocardial infarction after acute coronary ischemic
syndromes. Circulation 2003; 107:2307–2312.
Wu V-Y, Cohen MP. Platelet factor 4 binding to glomerular microvascular matrix.
Yahalom J, Eldor A, Fuks Z, Vlodavsky I. Degradation of sulfated proteoglycans in
the subendothelial extracellular matrix by human platelet heparitinase. J Clin
Invest 1984; 74:1842–1849.
11
Laboratory Testing for
Andreas Greinacher
Ernst-Moritz-Amdt University Greifswald, Greifswald, Germany
I. INTRODUCTION
Heparin-induced thrombocytopenia (HIT) is caused by heparin-dependent

antibodies that usually recognize multimolecular complexes of platelet factor
4–heparin (PF4–H). HIT can be viewed as a clinicopathologic syndrome
(Warkentin et al., 1998). This implies that a diagnosis of HIT should be based
on two criteria: (1) clinically evident abnormalities, most commonly, throm-
bocytopenia with or without thrombosis (see Chap. 3), and (2) detection of
HIT antibodies. In some ways, HIT resembles another clinicopathologic
disorder, the antiphospholipid antibody (lupus anticoagulant) syndrome
(Table 1).
Two major classes of assays—activation (functional) and antigen—have
been developed to detect HIT antibodies (Table 2). A third (miscellaneous)
group of assays will be discussed briefly at the end of this chapter.
II. ACTIVATION ASSAYS FOR HIT ANTIBODIES

A. Washed Platelet Assays
The classic washed platelet assay for HIT is the serotonin release assay, or
SRA (Sheridan et al., 1986; Warkentin et al., 1992). This assay was a
271
272 Warkentin and Greinacher
Table 1 Two Clinicopathologic Syndromes: HIT and Antiphospholipid Antibody Syndrome

Heparin-induced
Laboratory features thrombocytopenia Antiphospholipid antibody syndrome
Structure of the major ‘‘Cryptic’’ autoepitope on PF4 ‘‘Cryptic’’ autoepitope on h2-GPI

antigen expressed when complexed expressed when bound to anionic
with heparin (see Chaps. 6–8) phospholipid (Pengo et al., 1995)
Nonspecific nature of Variable reactivity when heparin Variable reactivity when cardiolipina
anionic component substituted by LMWH, substituted by phosphatidylserineb
of antigen danaparoid, pentosan or other molecules (e.g., irradiated
(cross-reactivity) polysulfate, and others plastic)
Sequestered location PF4 within platelet a-granules Anionic phospholipids of inner
of antigen leaflet of bilipid membranes
Functional assay Heparin-dependent activation Prolongation of phospholipid-
of platelets by patient serum dependent coagulation assay by
or plasma patient’s plasma
Inhibition of functional Inhibition by high-dose heparin Inhibition of lupus anticoagulant
assay assays by excess phospholipid
Antigen assay PF4–H–EIA Anticardiolipin EIA
Relative sensitivity of Antigen > functional Antigen > functional
assays for antibodies
Relative specificity of Functional > antigen Functional > antigen
assays for disease state
Information on relative sensitivity and specificity of functional and antigens assays for HIT and
antiphospholipid antibody syndrome are provided elsewhere (Visentin et al., 1994; Ginsberg et al., 1995;
Berube et al., 1998; Warkentin et al., 2000).
Abbreviations: EIA, enzyme immunoassay; h2-GPI, beta2-glycoprotein I; LMWH, low molecular weight
heparin; PF4-H, platelet factor 4-heparin.
a
Cardiolipin is found primarily in the inner leaflet of the mitochondrial membrane.
b
Phosphatidylserine is located in the inner leaflet of platelet membranes; thus, antiphospholipid
antibodies with antiphosphatidylserine activity could be relatively more important in the pathogenesis of
thrombocytopenia.
modification of platelet-washing techniques in use at McMaster University

that resuspended washed platelets in buffer containing physiological concen-
trations of calcium. The purpose was to avoid platelet activation artifacts
associated with low calcium concentrations (Mustard et al., 1972; Kinlough-
Rathbone et al., 1983) (see Chap. 1). The use of washed platelets is also central
to certain other functional assays, such as the heparin-induced platelet acti-
vation (HIPA) assay (Greinacher et al., 1991). Figure 1 summarizes washed
platelet assays for HIT.
Preparation of Platelets for Washed Platelet Assays

1. Collect 8.4 volumes of blood from a normal donor into 1.6 volumes
of acid–citrate–dextrose (ACD).
Laboratory Testing for HIT 273
Table 2 Classification of Laboratory Tests for HIT
Activation (functional) assays

Washed platelet assays
Heparin-induced platelet activation (HIPA) test: visual assessment of platelet aggregation
(Greinacher et al., 1991; Eichler et al., 1999)
Serotonin release assay (SRA): quantitation of 14C-radiolabeled serotonin released from
dense granules of activated platelets (Sheridan et al., 1986); chemical detection of
serotonin also described (Schnell et al., 1998; Harenberg et al., 2000)
ATP release detected by luminography (Stewart et al., 1995)
Platelet microparticle assay: quantitation of platelet-derived microparticles by flow
cytometry (Lee et al., 1996)
Platelets in citrated platelet-rich plasma (c-PRP)
Platelet aggregation test (PAT): assessment of platelet aggregation using conventional
aggregometry (Fratantoni et al., 1975; Chong et al., 1993a)
Annexin V-binding assay: quantitation by flow cytometry of annexin V binding to anionic
phospholipids expressed by activated platelets (Tomer, 1997; Tomer et al., 1999)
Serotonin release detected by flow cytometry (Gobbi et al., 2003)
Antigen assays
Target antigen: platelet factor 4-heparin complexes
PF4-heparin enzyme immunoassay (PF4–H–EIA)
Surface-bound antigen (Amiral et al., 1992)
Fluid-phase antigen (Newman et al., 1998)
Particle gel immunoassay (Meyer et al., 1999; Eichler et al., 2001)
Target antigen: platelet factor 4-polyvinylsulfonate complexes
PF4-polyvinylsulfonate-EIA (Collins et al., 1997; Visentin et al., 2001)
Comment. Aspirin-free normal blood donors whose platelets are

known to respond well to HIT sera should be selected, as there is considerable
heterogeneity to platelet activation by HIT sera among platelets obtained
from different normal individuals (Warkentin et al., 1994). In Hamilton,
platelets from two donors are combined. In Greifswald, platelets from four
different donors selected randomly are prepared and tested individually. ABO
blood group discrepancies do not affect the results of these assays (Greinacher
et al., 1991).
2. Perform differential centrifugation to obtain ACD-anticoagulated
platelet-rich plasma.
Comment. Low-speed centrifugation prepares ACD-anticoagulated
platelet-rich plasma (PRP). Additional ACD (111 AL/mL PRP) is added
(Greifswald) to ensure that the pH of the PRP is sufficiently low (V6.5) to
prevent platelet aggregation that otherwise would occur during platelet
pelleting: the platelet release reaction is triggered by close platelet contact
in low calcium concentrations at physiological pH (Kinlough-Rathbone et
Figure 1 Schematic overview of washed platelet assays: HIT serum causes platelet
activation at therapeutic (0.1–0.3 U/mL) heparin concentrations, but not in the
presence of Fc receptor-blocking monoclonal antibody or high (100 U/mL) heparin
concentrations. Platelet activation by HIT serum is potentiated by ADP release from
platelet dense granules. Various platelet activation endpoints can be used. False-
positive results can be avoided if typical reaction profiles of non-HIT platelet
activation triggers are recognized [e.g., (1) residual thrombin (activation at low but not
high heparin concentrations, including activation in the presence of Fc receptor-
blocking monoclonal antibody; (2) immune complexes (activation at low and high
heparin concentrations, both of which are inhibited by Fc receptor-blocking
monoclonal antibody; and (3) thrombotic thrombocytopenic purpura (TTP) serum
(variable activation in the presence of heparin that is not inhibited by Fc receptor-
blocking monoclonal antibody)]. Abbreviations: ACD, acid–citrate–dextrose; ATP
adenosine triphosphate; PRP, platelet-rich plasma.
al., 1983). If the serotonin release method is used, the PRP is incubated at
37jC for 30 min with [14C]serotonin (0.1 ACi/mL of PRP added from a stock
solution of 50 ACi/mL of [14C]serotonin) (Lee et al., 1996).
3. Wash the platelets by pelleting them from PRP, then gently resus-
pend the platelets in calcium- and magnesium-free Tyrode’s buffer,
pH 6.3, containing glucose (5.6 mmol/L) and apyrase (2.5 U/mL).
Comment. Tyrode’s buffer consists of physiological concentrations of
sodium chloride (NaCl, 137 mmol/L), potassium chloride (2.7 mmol/L),
calcium chloride (CaCl2, 2 mmol/L), magnesium chloride (MgCl2, 1.0
mmol/L), and sodium dihydrogen phosphate (NaH2PO4, 3.3 mmol/L);
however, calcium-free and magnesium-free Tyrode’s is used in this wash
step to avoid activating the coagulation factors and platelets. The low pH
prevents platelets from aggregating during pelleting. Apyrase is an enzyme
that degrades adenine nucleotides (i.e., accumulation of the ADP from the
platelets is prevented). Azide-free bovine serum albumin (3.5 mg/mL) and
hirudin (1 U/mL) are included in the wash buffer in Greifswald, but not
Hamilton, although HEPES (5 mmol/L) is added to this buffer in Hamilton.
Following resuspension, the platelets are incubated for 15 min at 37jC
(Greifswald).
4. Pellet the washed platelets as before, and then gently resuspend the
platelets into calcium- and magnesium-containing Tyrode’s buffer,
pH 7.4, without apyrase or hirudin.
Comment. Following resuspension, the platelets should ‘‘rest’’ for 45
min at 37jC (Greifswald). The final resuspension buffer (Tyrode’s buffer at
physiological pH) contains calcium (2 mmol/L) and magnesium (1 mmol/L
Hamilton; 2 mmol/L Greifswald). The platelet count is adjusted to a
minimum of 300 wells 109/L; thus, after addition of washed platelets (75
AL) to the microtiter wells containing test serum (20 AL) and heparin-buffer (5
AL in Hamilton, 10 AL in Greifswald), the final platelet concentration will be
at least 215 109/L. Apyrase must not be included in this buffer, as the ADP
released during assessment of HIT-induced platelet activation is an important
potentiator of the platelet Fc receptor–mediated platelet activation (Polgár et
al., 1998).
Test Conditions of Heparin-Dependent Platelet Activation:

Perform Platelet Activation Studies Under Various Test
Conditions
Comment. In Hamilton, sera are studied using six different reaction
conditions: (1) buffer; (2) unfractionated heparin (UFH), 0.1 U/mL; (3) UFH,
0.3 U/mL; (4) UFH, 100 U/mL; (5) low molecular weight heparin (LMWH),
enoxaparin, 0.1 U/mL; and (6) UFH, 0.3 U/mL plus a monoclonal antibody
(IV.3) that inhibits platelet Fc receptor–mediated platelet activation. In
Greifswald, routine testing is performed using (1) buffer; (2) LMWH
(reviparin), 0.2 U/mL; (3) UFH, 100 U/mL; and (4) danaparoid, 0.2 U/mL
(to assess cross-reactivity); (5) sometimes LMWH 0.2 U/mL plus IV.3 is
performed to resolve unclear results. In Greifswald, the LMWH preparation
reviparin (Clivarine) is used because of its narrow molecular weight (MW)
range (80% of its chains have molecular mass of 2.4–7.2 kDa, i.e., 4–12
disaccharide units) (Jeske et al., 1997); this results in more consistent
formation of PF4–heparin complexes, enhancing sensitivity of the assay
(Greinacher et al., 1994b). Platelets are incubated with various test and
positive and negative control sera under these various reaction conditions
for up to 45 min (Greifswald) or 60 min (Hamilton). The order of pipetting is
important in optimizing assay results (Eichler et al., 1999) (see Table 3).
Interpretation of the Obtained Test and Control Data

Comment. Several techniques can be used to assess activation of
washed platelets (see Fig. 1). The actual method of detection of platelet
activation is probably less important than the technique of platelet
preparation itself, including the selection of suitable platelet donors.
A positive test result is one in which heparin-dependent platelet
activation occurs at therapeutic concentrations of heparin (0.1–0.3 U/mL)
but is inhibited at very high (100 U/mL) heparin concentrations and in the
presence of platelet Fc receptor–blocking monoclonal antibody. By assessing
activation in the presence of different LMWH compounds or danaparoid,
studies of in vitro cross-reactivity can be performed (discussed subsequently).
It is important to ensure that control HIT sera, including one or more weak
positive controls, react as expected. Given the experience from a workshop on
testing for HIT antibodies (Eichler et al., 1999), we recommend exchange of
weak positive control sera among laboratories for quality control.
Platelet Activation Endpoints

[14C]Serotonin release was the first activation endpoint described using
washed platelets (Sheridan et al., 1986). In this method, the washed platelets
are incubated with test and control serum or plasma and heparin–buffer in
flat-bottomed polystyrene microtiter wells (in duplicate or triplicate), per-
formed on a platelet shaker (shaken, not stirred). After 1 h, the reaction is
halted with 100 AL of 0.5% EDTA in phosphated-buffered saline (PBS). The
microtiter plates are centrifuged at 1000 g for 5 min, and 50 AL of supernatant
Table 3 Pipette Scheme for the HIPA Test
Add last
Add second Add third
Add first LMWH Danaparoid
Laboratory Testing for HIT
UFH Washed platelet (2.1 anti-Xa (2.1 anti-Xa

Heat-inactivated 1050 U/mL suspension U/mL) U/mL)
patient or = 100 U/mL (300,000 Suspension = 0.2 U/mL = 0.2 U/mL
control serum (final) platelets/AL) buffer (final) (final)
Control with buffer 20 AL 75 AL 10 AL

Low heparin concentration 20 AL 75 AL 10 AL
High heparin concentration 20 AL 10 AL 75 AL
Cross-reactivity with danaparoid 20 AL 75 AL 10 AL
The order of pipetting is important. After adding serum to the microtiter plate wells, high heparin concentrations are added to the appropriate
wells: this will disrupt PF4–heparin complexes that may be present in the serum. After adding washed platelets, buffer, LMWH (low
concentrations), and danaparoid (for cross-reactivity testing) are added. If inhibition by monoclonal antibody IV.3 is tested, this reagent is added
before addition of the washed platelet suspension. In Hamilton, the pipetting order for the SRA is (1) addition of buffer-heparin, (2) serum, and (3)
platelets.
277
fluid is transferred to tubes containing scintillation fluid for detection of

[14C]serotonin released during platelet activation.
Carbon-14 is a radioisotope with a long half-life (5730 years) that emits
h-particles (electrons). Laboratories require special licenses to handle radio-
isotopes, thus limiting widespread use of this platelet-activation marker.
However, it is also possible to quantitate serotonin by nonradioactive analysis
(Schnell et al., 1998; Harenberg et al., 2000; Gobbi et al., 2003). Results are
expressed as percentage of serotonin released. This is calculated based on
comparison with maximal possible release (determined following detergent-
induced platelet lysis), adjusted for background release (determined by
quantitating serotonin release from a sample incubated with buffer alone).
Acceptable experiments should have less than 5% background release, with
both buffer and negative control serum testing negative for HIT antibodies.
Aggregation of Washed Platelets. A convenient and useful activation
endpoint—platelet aggregation—was reported in the heparin-induced plate-
let activation (HIPA) assay (Greinacher et al., 1991; Eichler et al., 1999). Test
serum and heparin–buffer are placed in U-bottomed polystyrene microtiter
wells containing two stainless steel spheres, and the platelets are stirred at
approximately 500 rpm, using a magnetic stirrer. At 5-min intervals the wells
are examined against an indirect light source: a change in appearance of the
reaction mixture from turbidity (nonaggregated platelets) to transparency
(aggregated platelets) is a positive result. Although the activation endpoint is
evaluated subjectively, interobserver agreement is good. A further advantage
of this technique is its repeated evaluation of platelet activation over time.
Thus, strong HIT sera that cause the typical activation profile of HIT (i.e.,
activation at low, but not high, heparin concentrations) within 15–30 min are
readily identified. In contrast, such a strong HIT serum might eventually
cause platelet activation even at the high heparin concentration and thus
cause an ‘‘indeterminate’’ reaction pattern (activation at both low and high
heparin concentrations) if activation is assessed at a later time point only.
Occasionally there is interference with visual interpretation (e.g., a lipemic
serum).
Luminography. Stewart et al. (1995) reported luminography to detect
platelet activation, using a commercially available lumiaggregometer. Aden-
osine triphosphate (ATP) is released from platelet dense granules during
platelet activation. In the presence of luciferin-luciferase reagent, a light flash
is generated in the presence of ATP, which is detected and quantitated.
Another group reported similar results using a standard scintillation counter
(Teitel et al., 1996). It is uncertain how the sensitivity and specificity of these
assays compare with other markers of platelet activation.
Platelet-Derived Microparticle Generation. Generation of platelet-de-

rived microparticles occurs when washed platelets are activated by HIT sera
(Warkentin et al., 1994). With use of a fluorescein-labeled anti-GPIba murine
monoclonal antibody, a method for quantitating microparticles using flow
cytometry was reported by Lee and coworkers (1996). Although both
platelets and microparticles bind fluorescein-labeled anti-GPIba monoclonal
antibody, they can be distinguished by their size and scatter parameters using
flow cytometry, with microparticles quantitated in relation to platelet numb-
ers (Lee et al., 1996).
Heat Inactivation of Patient Serum or Plasma

To avoid thrombin-induced platelet activation in buffer containing physio-
logical calcium, steps are taken to inactivate residual thrombin. Thus, plasma
and serum must first be heat inactivated before use in these assays. Heating at
56jC for 30–45 min inactivates thrombin and complement. Fibrin and other
precipitates are removed by high-speed centrifugation (8000 g for 5 min).
More intense heating of serum (63jC for 20 min) forms platelet-activating
immune complexes (Warkentin et al., 1994); thus, if a patient sample shows
heparin-independent platelet activation (indeterminate result), another sam-
ple aliquot should be heat inactivated, and the HIT assay repeated. Often, this
will result in disappearance of the initial artifact that presumably was caused
by too intense heat inactivation. Serum is preferred for use in functional HIT
assays in our laboratories, as serum contains more PF4, thereby facilitating
initial formation of the antigen.
Biological Basis for High Sensitivity of Washed Platelets to

Activation by HIT Antibodies
Table 4 lists differences between using washed platelets and platelets sus-
pended in citrate-anticoagulated plasma to study HIT antibody-mediated
platelet activation. Some of these differences may be important in explaining
the greater sensitivity and specificity of washed platelets to detect HIT
antibodies.
1. Baseline platelet activation, including platelet granule release,
occurs during preparation of washed platelets. This enhances the
platelet-binding capacity for heparin (Horne and Chao, 1989) and
may also increase the availability of PF4 to form the target antigen.
2. Apyrase is used to prevent accumulation of ADP during platelet
washing. This prevents platelets from becoming refractory to sub-
280
Table 4 Comparison Between Citrated Platelet-Rich Plasma and Washed Platelet Assays
Technical aspects Washed platelet assay Platelet-rich plasma assay Comments
Platelet preparation High g centrifugation during Low g centrifugation: less Availability of PF4 may be higher using
washing: increased baseline baseline platelet activation washed platelets (greater formation of
platelet activation PF4–heparin antigen complexes)
Apyrase Apyrase added to wash solution, No apyrase used Apyrase degrades ADP, and prevents its
but not to the final (no wash steps) accumulation; thus, platelet refractoriness
resuspension (reaction) buffer to ADP-mediated potentiation of HIT
serum-induced platelet activation is
avoided by apyrase
Reaction milieu Physiological calcium Low (micromolar) calcium IgG-mediated platelet activation optimal
concentration (2 mmol/L) owing to citrate with physiological calcium concentrations
IgG levels Reduced IgG levels during final Normal plasma IgG levels Reduced inhibition of Fc receptor–mediated
reaction platelet activation by IgG in washed
platelet assays
Plasma protein levels Reduced plasma protein levels Normal plasma protein levels Reduced nonidiosyncratic platelet activation
by heparin using washed platelets (?)
Temperature Room temperature 37jC Unknown significance
Reaction assessment Microtiter plates Conventional aggregometer Many assays performed simultaneously
using microtiter plates
See text for further details on differences between washed platelet and citrated platelet-rich plasma assays (pp. 279–282).
Warkentin and Greinacher
sequent ADP-mediated platelet activation (Ardlie et al., 1970).

Empirically, apyrase grade III (Sigma) is acceptable for use: grades I
and II are too impure, and grades IV and higher are expensive.
3. Physiological calcium concentrations are present when washed
platelets are used. Under these conditions, ADP produces only pri-
mary platelet aggregation. However, as observed by Packham et al.
(1971), traces of immunoglobulin complexes in amounts too low to
cause aggregation themselves will cause secondary aggregation to
occur following the addition of ADP. Recently, the importance of
ADP in mediating HIT antibody-induced platelet activation has
been reported by Polgár and colleagues (1998). Thus, the reaction
conditions that exist when washed platelets are used appear to max-
imize HIT antibody-induced platelet activation because the platelets
retain sensitivity to ADP-mediated platelet activation.
4. Low concentrations of IgG are present in the final washed platelet
reaction mixture: there is a fivefold reduction in IgG compared with
citrated platelet-rich plasma (c-PRP) assays, because only IgG from
the test serum is present in the final reaction mixture. Chong and
colleagues (1993a) showed that high plasma IgG levels in one plate-
let donor’s blood seemed to explain the discrepancy between studies
using donor c-PRP (poor reactivity) and donor washed platelets
(good reactivity). These and other investigators (Greinacher et al.,
1994c) also observed that addition of IgG inhibits HIT serum-
induced activation of washed platelets in a dose-dependent fashion.
5. Low concentrations of fibrinogen and other plasma proteins could
reduce the potential for nonidiosyncratic heparin-induced platelet
activation (Salzman et al., 1980; Chong et al., 1993a). In contrast,
low concentrations of heparin rarely cause significant activation of
washed platelets. It is possible that acute-phase reactant proteins
such as fibrinogen could lead to false-positive activation assays for
HIT using c-PRP.
6. Room temperature conditions are used for washed platelet assays;
in contrast, c-PRP studies are performed at 37jC. Although this is a
major difference between the assays, it is unknown whether there are
advantages or disadvantages of performing washed platelet assays
at room temperature. In Greifswald, all buffers are warmed to 37jC,
and all incubation steps are performed at this temperature; only the
final incubation on the microtiter plates is performed at room
temperature.
7. Multiple serum–platelet reactions in microtiter plates can be per-
formed, and even several hundred reactions studied in parallel.
Quality control is thereby enhanced by the large number of control
and test reaction conditions that can be analyzed, and the long in-
cubation period employed (up to 60 min). The incubation period in
HIT assays should be at least 20–30 min (Stewart et al., 1995).
Quality Control in Washed Platelet Assays for HIT

The variable reactivity of donor platelets to HIT sera is an important issue in
activation assays for HIT. It has long been recognized that inconsistent results
can be obtained using these assays (Salem and van der Weyden, 1983; Pfueller
and David, 1986; Warkentin et al., 1992).
Hierarchical Versus Idiosyncratic Platelet Activation by HIT Sera. The
results of a systematic investigation, summarized in Table 5, showed that both
HIT sera and platelet donors exhibit variable reactivity in a hierarchical,
rather than an idiosyncratic, manner. The strongest reactions were produced
by strong HIT sera against strongly reactive platelet donors. All of the nega-
tive reactions occurred when the weakest sera were mixed with the weakest
platelets. Importantly, no unexpected negative reactions occurred elsewhere
in the 10 10 serum-platelet grid (see Table 5). Furthermore, the relative
ranking of platelet donors appeared to be stable over time, an observation
also reported by Chong and colleagues (1993a) in their study of platelet donor
variability using c-PRP.
The finding of a hierarchical pattern of reactivity has important
implications for quality control in diagnostic testing for HIT using activation
assays. First, it indicates that platelets from certain donors who tend to re-
spond well to HIT sera should be chosen. Second, relatively weak HIT sera
should be included as positive controls (f20–50% serotonin release, or 25-
min lag time in the HIPA).
Heparin-Independent Platelet Activation: Indeterminate Results.
About 5% of test sera or plasma give an indeterminate result in an activation
assay. This is defined as platelet activation that occurs at both therapeutic
(0.1–0.3 U/mL) and supratherapeutic (10–100 U/mL) heparin concentra-
tions. Often an interpretable result is obtained when the assay is repeated
using another heat-inactivated aliquot. This suggests that the first result may
have been an artifact caused by heat-aggregated IgG generated ex vivo.
However, some serum and plasma samples repeatedly demonstrate heparin-
independent platelet activation. Biological explanations include circulating
immune complexes (e.g., systemic lupus erythematosus), high-titer HLA class
I alloantibodies, and, possibly, other platelet-activating factors (e.g., throm-
botic thrombocytopenic purpura). An antigen assay is required for further
investigation when an indeterminate result is consistently obtained.
Table 5 Reactivities of Ten HIT Sera with Platelets from Ten Normal Donors
Normal platelet donors: strongest (P1) to weakest (P10)

HIT sera
(S1–S10) P1 84.3 P2 71.2 P3 68.4 P4 53.6 P5 52.5 P6 41.3 P7 39.7 P8 38.9 P9 36.9 P10 29.9
S1 85.4 ++++ ++++ ++++ ++++ ++++ ++++ +++ ++++ ++ +++
S2 84.4 ++++ ++++ ++++ ++++ ++++ +++ +++ ++++ +++ +++
Laboratory Testing for HIT
S3 69.1 ++++ ++++ ++++ +++ +++ ++ +++ ++ ++ ++

S4 61.0 ++++ ++++ +++ +++ +++ ++ ++ ++ ++ +
S5 56.4 ++++ +++ +++ ++ +++ ++ ++ + + +
S6 50.7 +++ +++ +++ +++ ++ ++ ++ ++ + +
S7 44.1 ++++ +++ +++ + ++ + + + ++
S8 30.1 ++++ ++ +++ + + +
S9 24.2 +++ ++ + ++ +
S10 11.3 ++ + +
Serum samples and platelet donors are ranked from strongest to weakest (S1–S10 and P1–P10, respectively), according to the mean percentage of
[14C]serotonin release when considering all 100 serum–platelet donor pairs (10 pairs corresponding to each HIT serum and each normal platelet
donor). For each serum–platelet donor pair, the individual amount of serotonin release is summarized as follows: 80–100%, ++++; 60–79%
release, +++; 40–59% release, ++; 20–39% release, +; <20% release,–.
Overall, there is a graded pattern of reactivity among the individual reaction pairs that is hierarchical (i.e., there are no unexpected weak or strong
reactions among the pairs). All negative reactions (<20% release) were found in the lower right portion of the table. Conversely, the strongest
reactions (z80% release) were found in the upper left portion of the table.
Source: Warkentin et al., 1992.
283
Inhibition by High Heparin Concentrations. Sheridan and colleagues

(1986) first emphasized that there was a relatively specific activation profile
triggered by HIT sera and plasmas: activation at therapeutic heparin con-
centrations (maximal at 0.1–0.3 U/mL) that progressively diminished with
increasing heparin concentrations, typically falling to background activation
at very high (100 U/mL) heparin concentrations. Classically, a positive test
was defined as greater than 20% serotonin release at 0.1 U/mL heparin and
less than 20% serotonin release at 100 U/mL heparin. These criteria should
not be applied indiscriminately, however. For example, a very strong HIT
serum could produce more than a 90% release at 0.1 U/mL heparin and 25%
release at 100 U/mL heparin. Alternatively, a serum or plasma sample that
was not adequately heat-inactivated could produce a similar reaction profile
(i.e., residual thrombin is inhibited by the high, but not low, heparin con-
centration). The strength of reactivity caused by patient serum can be helpful:
clinically significant HIT antibodies almost always cause more than 50%
serotonin release using optimally reactive platelets (Warkentin et al., 2000;
Warkentin and Heddle, 2003). In the HIPA test, differences in the lag time to
platelet aggregation provide useful information.
Inhibition by Fc Receptor Blockade. Platelet activation by HIT anti-
bodies is inhibited in the presence of a murine IgG2b monoclonal antibody
(IV.3) that recognizes the platelet FcgIIa receptor (Kelton et al., 1988; Chong
et al., 1989) and can be used to enhance test specificity.
Interpretation of Platelet Activation by HIT Serum in the Absence

of Added Heparin
With activation assays, it is not uncommon for HIT serum or plasma to cause
platelet activation, even in the absence of added heparin. Usually, even
greater platelet activation occurs in the presence of added heparin. When
strong serum-dependent platelet activation occurs with buffer and at a 0.1–0.3
U/mL heparin concentration, it is important to ensure that the other reactions
(at 100 U/mL heparin, and 0.1–0.3 U/mL heparin together with Fc receptor
blockade) are as expected. This is because residual thrombin could produce
strong platelet activation in both the absence and presence of low heparin
concentration, thereby causing the potential for a false-positive result.
There are at least two potential explanation for strong platelet activa-
tion in the absence of added heparin. First, there may be residual heparin in
the sample (White et al., 1992; Pötzsch et al., 1996). However, this phenom-
enon can persist despite attempts to remove heparin using binding resins.
Further, heparin-independent platelet activation can be a feature of serum
obtained from patients with ‘‘delayed-onset HIT,’’ in which the presence of
residual heparin is unlikely because onset of thrombocytopenia and throm-
bosis begins several days after the patient’s last exposure to heparin (War-
kentin and Kelton, 2001) (see Chap. 3). A second explanation is that some
HIT antibodies recognize platelet-bound PF4 in the absence of an exogenous
source of heparin, perhaps by PF4 bound to platelet glycosaminoglycans.
Alternatively, as HIT antibodies are heterogeneous, there may be pathogenic
antibody subpopulations that bind relatively well to PF4 even in the absence
of heparin or heparin-like molecules (Newman and Chong, 1999; Amiral et
al., 2000). This phenomenon has implications for the interpretation of tests of
cross-reactivity of LMWH and danaparoid, as discussed later.
Disadvantages of Washed Platelet Assays

The major disadvantage of washed platelet assays to detect HIT antibodies is
that they are technically demanding and labor-intensive. A workshop that
compared a washed platelet assay (the HIPA test) and an antigen assay
showed greater variability in activation assay results among the participating
laboratories (Eichler et al., 1999). Washed platelet activation assays are best
suited for reference laboratories assessing many HIT sera, as this facilitates
acquisition of sufficient technical experience to perform the assay successfully
on a consistent basis. Assay-specific disadvantages include the requirement
for radioactivity (SRA), the use of a subjective, visual endpoint (HIPA), and
expensive equipment (flow cytometry–based assays).
B. Activation Assays Using Citrate-Anticoagulated Blood

The first reports describing the use of normal donor c-PRP to detect platelet
activation caused by HIT serum or plasma appeared in the 1970s (Rhodes et
al., 1973; Fratantoni et al., 1975; Babcock et al., 1976). A ratio of serum (or
plasma) to c-PRP between 0.66 and 1.0 was used (e.g., 200 AL serum added to
200–300 AL c-PRP). No standardized method has evolved, however, although
a survey of 54 laboratories in France (Nguyên et al., 1994) found some
practices to be more common. For example, most laboratories test patient
citrated platelet-poor plasma (c-PPP) rather than heat-inactivated serum.
Variable heparin concentrations are used, most commonly between 0.5 and
1.0 U/mL. The ratio of patient c-PPP to donor c-PRP is usually 1:1, and ABO
discrepancies are usually ignored. About 75% of the laboratories use at least
two platelet donors for diagnostic testing.
Testing for HIT Antibodies Using c-PRP

The following description of the assay taken from Chong and colleagues
(1989, 1993a) has the highest reported sensitivity and specificity among c-PRP
methods. Blood is obtained from normal blood donors whose platelets

respond well to serum or plasma from HIT patients, and c-PRP is prepared.
Testing involves addition of 150 AL of patient heat-inactivated c-PPP or
serum to 340 AL of c-PRP (final platelet concentration, 250–350 109/L) at
37jC. The platelets are monitored for a few minutes to exclude nonspecific
platelet aggregation. After addition of 10 AL heparin-saline, aggregation is
monitored over the next 15 min or until aggregation has occurred. A positive
result is an increase in light transmission of more than 25% above baseline in
the presence of therapeutic-dose heparin (0.5 U/mL) and patient serum or c-
PPP and inhibition of aggregation in the presence of patient serum or plasma
and supratherapeutic-dose heparin (100 U/mL). Use of such a two-point
assay reduced the false-positive rate, as serum or plasma from some patients
without HIT caused platelet aggregation at all heparin concentrations tested.
To ensure that the platelets are functional, platelets are also tested with
collagen (2 Ag/mL). Details on methodology of c-PRP assays are also given
elsewhere (Kapsch and Silver, 1981; Almeida et al., 1998).
Some workers report that platelets from a patient with HIT are
very reactive to heparin-dependent activation by their own serum or plasma
(Kappa et al., 1987; Chong et al., 1993b). Use of autologous c-PRP can
sometimes be limited by the patient’s thrombocytopenia, however. Potential
explanations for the high sensitivity of autologous platelets include persisting
high Fc receptor expression on platelets of patients with acute HIT (Chong et
al., 1993b) and baseline platelet activation (Chong et al., 1994), with the
potential for higher PF4 availability.
Disadvantages of c-PRP Aggregation Assays

Problems with these assays include (1) potential for false-positive interpreta-
tion if heparin produces nonspecific aggregation of donor platelets, an effect
that could be enhanced nonspecifically by proaggregatory factors in the pa-
tient serum or plasma, and (2) risk of false-negative interpretation if HIT
serum–induced platelet aggregation begins even before addition of heparin.
Nonspecific activation of platelets by heparin occurs with some normal
donor c-PRP (Chong et al., 1993a), rendering these donors unsuitable for
diagnostic testing. It is also possible that plasma from very sick patients may
be more likely to cause nonspecific aggregation of platelets in c-PRP in the
presence of heparin (Goodfellow et al., 1998).
An important practical disadvantage is that only a limited number of
platelet aggregation tracings can be performed using conventional aggreg-
ometers. Thus, relatively few reactions with a limited number of patient and
control samples can be evaluated.
Other Assays Using Citrated-Anticoagulated Whole Blood

or Platelet-Rich Plasma
Tomer (1997) reported a c-PRP activation assay for HIT antibodies in which
the activation endpoint is quantitation of binding of fluorescein-labeled re-
combinant annexin V to platelets, as detected using flow cytometry. Annexin
V, a placental protein, interacts with the prothrombinase-binding anionic
phospholipids expressed on the surface of activated platelets and correlates
with platelet procoagulant activity. It is uncertain whether the reaction
conditions employed (e.g., 30-min incubation at 26jC) or the high sensitivity
of annexin V binding (300-fold increase over baseline) overcomes the inherent
limitations of sensitivity observed with other assays using c-PRP. Recently,
Gobbi and colleagues (2003) developed a flow cytometry assay modeled after
that of Tomer (1997), except that loss of serotonin from platelet granules was
used as the platelet activation endpoint.
C. Comparison of Washed Platelet and c-PRP Activation

Assays
It became evident during the mid-1980s that the sensitivity of c-PRP aggre-
gation assays for HIT was relatively poor (Kelton et al., 1984; Pfueller et al.,
1986). Favaloro and colleagues (1992) first compared the c-PRP aggregation
assay with the washed platelet SRA. They observed that only 6 of 13 HIT sera
or plasmas that tested positive in the SRA also tested positive in the c-PRP
aggregation assay. In contrast, no sample was identified that tested positive
only in the aggregation assay. Chong and colleagues (1993a) also found a
higher sensitivity for the SRA method. However, considerable variability in
sensitivity for HIT antibodies among the various platelet donors was seen,
ranging from 39–81% (c-PRP assay) to 65–94% (washed platelet SRA).
Strong evidence in favor of a higher sensitivity for washed platelet
assays was provided by direct comparison using platelets prepared and tested
in parallel that were obtained simultaneously from the same platelet donors
(Greinacher et al., 1994a). Only 23 of 70 HIT sera that tested positive by the
HIPA assay also tested positive using c-PRP aggregation. In contrast, all but
1 of 24 sera testing positive in the c-PRP aggregation assay also tested positive
in the HIPA test.
More recently, Walenga and colleagues (1999) also found a lower
sensitivity of the c-PRP aggregation test compared with the SRA. In con-
trast, Pouplard and co-workers (1999) reported a similarly high sensitivity of
the c-PRP as the SRA (91% vs. 88%), but with a lower specificity (77% vs.
100%).
III. ANTIGEN ASSAYS FOR HIT ANTIBODIES

A. Solid-Phase Enzyme Immunoassay
The solid-phase enzyme immunoassay (EIA) has been described in detail
(Amiral et al., 1992; Visentin et al., 1994; Greinacher et al., 1994a; Amiral
et al., 1995; Horsewood et al., 1996). Methods differ in the way that PF4–
heparin complexes are coated on the microtiter wells. A general scheme is
shown in Fig. 2. In this assay, stoichiometric concentrations of PF4 and
heparin (e.g., 50 AL each of 20 Ag/mL PF4 and 1 U/mL UFH) dissolved in
phosphate buffer are added together to the wells of a microtiter plate and
incubated at 4jC overnight. After washing with phosphate buffer saline–
Tween 20 (PBS–Tw), the wells are ‘‘blocked’’ with a protein-containing so-
lution such as PBS–Tw containing either 10% normal goat serum (NGS) or
20% fetal calf serum, followed by washing with PBS–Tw. To perform the
assay, 50–100 AL of test or control plasma or serum diluted 1:50 in PBS con-
taining 2% NGS is added to duplicate wells for 1 h at room temperature.
After thorough washing with PBS–Tw, bound immunoglobulin is detected by
adding alkaline phosphatase–conjugated goat antihuman immunoglobulin
(e.g., affinity-purified goat antihuman IgG Fc diluted 1:1000 in PBS–Tw-2%
NGS) followed by incubation for 1 h at room temperature. After through
washing, incubation with p-nitrophenyl phosphate in 1 M diethanolamine
buffer is added. After incubation in the dark, the reaction is stopped with 1 N
NaOH, and absorbance is read at 405 nm using an automated microplate
Figure 2 Schematic figure of solid-phase PF4–heparin–EIA.

reader. The upper limit of the normal range is usually set at the mean plus 3
SD obtained using normal sera. Some laboratories set an indeterminate range
for samples that are only minimally above the upper normal range.
B. PF4–Polyvinylsulfonate Antigen Assay

Several negatively charged substances can cause the cryptic autoepitope
within PF4 to become recognizable to HIT antibodies (see Chaps. 6–8).
Indeed, a commercial assay for HIT using PF4 complexed with polyvinylsul-
fonate has been developed (Collins et al., 1997; Visentin et al., 2001).
Sensitivity and reproducibility were high using polyvinylsulfonate that had
been fractionated to a relatively uniform molecular weight (5000 F 500 Da).
Some technical advantages of this assay include the observation that the ratio
of PF4/PVS is not critical (cf. PF4–heparin), with acceptable concentrations
of PVS ranging from 0.1 to 100 Amol/L for a corresponding concentration of
10 Ag/mL PF4. The antigen complex is also stable for long periods.
Table 6 compares the two commercial EIAs. In the laboratory in
Greifswald, discrepant results between the two assays have been observed
in about 15% of patient samples tested.
Some research laboratories perform ‘‘in-house’’ PF4-dependent EIAs
to detect HIT antibodies. An advantage is that an EIA can be used that only
detects HIT antibodies of the IgG class (Warkentin et al., 2000; Lindhoff-Last
et al., 2001; Untch et al., 2002). This improves test specificity, as PF4–heparin-
reactive IgA and IgM class antibodies (which are detected in the two com-
mercial EIAs) are unlikely to cause HIT.
C. Fluid-Phase EIA
The fluid-phase EIA for HIT antibodies (Newman et al., 1998) is an adaption
of a staphylococcal protein A antibody-capture EIA method (Nagi et al.,
1993). By permitting antibody–antigen interactions to occur in a fluid phase,
problems of protein (antigen) denaturation inherent in solid-phase assays are
avoided.
Platelet factor 4 (5% biotinylated) is mixed with an optimal concentra-
tion of heparin, and this antigen mixture is incubated with diluted patient
serum or plasma (Fig. 3). Subsequently, the antigen–antibody mixture is
incubated with protein G-Sepharose in a microcentrifuge tube. Biotinylated
antigen–antibody complexes become bound to the protein G–Sepharose by
antibody Fc, and the complexes, are separated from unbound-antigen by cen-
trifugation and washing. The amount of biotin–PF4–heparin–antibody com-
plexes immobilized to the beads is measured using peroxidase substrate after
initial incubation with streptavidin-conjugated peroxidase.
Table 6 Comparison of Two Commercial Antigen Assays for HIT Antibodies
AsserachromR (Stago) GTI-PF4 EIA (GTI)
Target antigen PF4–heparin complex PF4–polyvinyl sulfonate

complex
Source of PF4 Recombinant PF4 PF4 purified from
outdated platelets
Microwell strips provided Six strips of eight wells, Twelve strips of eight
in three nonresealable wells (in resealable
pouches pouches)
Sample required Plasma (sodium citrate) Serum or plasma 50 AL
200 AL diluted 1:100 diluted 1:50 (=1 AL
(=2 AL plasma per well) serum/plasma per well)
Controls supplied Positive and negative control Positive and negative
lyophylate,a calibrating control seraa
standard
Covers supplied One provided Multiple provided
Incubation times and All at 22jC (f2 h total) 37jC (two steps); then
conditions 22jC (f2 h total)
Plate reader settings 492 nm 405 or 410 nm
Detecting antibody system Goat anti-IgG/A/M Goat anti-IgG/A/M
(peroxidase-conjugated) (alkaline phosphatase-
conjugated)
Reaction stopping solution 3 M sulfuric acid or 1 M 3 M sodium hydroxide
hydrochloric acid
Cutoff from negative Internal control reagent is >0.4 OD (assumes controls
used to calibrate react as expected: pos z
1.8, neg V 0.2)
a
The AsserachromR assay provides lyophylized control sera, whereas handling of control sera in the GTI
assay is similar to handling of the test sera/plasma.
Source: Warkentin, 2000.
The fluid-phase EIA appears to have lower rate of false-positive re-

actions. This may be because in the solid-phase-EIA, nonspecific binding of
IgG to the microtiter wells can occur. Furthermore, the cryptic antigen site of
PF4 can be exposed when the molecule comes into close contact with the
plastic surface, even in the absence of heparin (Newman and Chong, 1999).
The fluid-phase assay avoids these problems by first precipitating all reactive
IgG antibodies, then detecting the antigen specifically bound to the IgG.
Thus, higher concentrations of patient serum or plasma can be tested without
increasing nonspecific reactivity. The advantages of this assay in performing
in vitro cross-reactivity are discussed later. Because antibody is bound using-
Figure 3 Schematic figure of fluid-phase PF4–heparin–EIA.
protein G–Sepharose, IgM and IgA anti-PF4–heparin antibodies are not

detected in this assay.
Wang and coworkers (1999a,b) used protein A to capture IgG anti-
bodies from HIT patient serum. The immobilized antibodies were then in-
cubated with normal serum (presumed to contain PF4) and fluorescence-
labeled heparin. The amount of fluorescence dye bound to the protein A
sepharose was used to detect HIT antibodies. Major drawbacks of this
approach include the initial capturing of IgG other than HIT-IgG, as well
as the unpredictable PF4/heparin ratios.
D. Particle Gel Immunoassay (ID-H/PF4 test)

This assay adopts the ID microcolumn system, in wide use for routine red
blood cell serology. Known as the H/PF4–PaGIA assay (DiaMed, Switzer-
land), red-dyed, high-density polystyrene particles coated with PF4–heparin
complexes serve as the solid phase. To perform this assay, 20 AL of these PF4–
heparin–coated microbeads and 10 AL of patient serum are added to the in-
cubation chamber of the microcolumn card. After a 5-min incubation, the
card is centrifuged. In the presence of anti-PF4–heparin antibodies, the par-
ticles are agglutinated and remain at the top of the column (strong positive),
or dispersed within the gel (weak positive); nonagglutinated particles precip-
itate to the bottom (negative) (Meyer et al., 1999; Eichler et al., 2001). The
assay is technically easy, can be performed rapidly, and is readily automated.
Results are read visually.
Eichler and colleagues (2001) compared this new assay with two func-
tional assays (HIPA test; SRA) and both commercially available solid-phase
PF4-dependent EIAs. In preselected samples, the H/PF4–PaGIA had a sen-
sitivity intermediate between that of the functional and commercial antigen
assays. The specificity appeared to resemble that of the functional assays.
In contrast, Risch and coworkers (2003) found many more sera to test
positive using the H/PF4–PaGIA, compared to a commercial EIA (Asse-
rochromR), among 42 patients sampled 10–18 days following cardiac sur-
gery (69% vs. 26%). Since none of the patients had clinical evidence for
HIT, this suggested the diagnostic specificity of the H/PF4–PaGIA to be far
less than the solid-phase EIA. These authors did not test sera from patients
with HIT, and therefore were unable to assess test sensitivity (Warkentin,
2003b).
The manufacturer’s instructions indicate that the assay is to be read as
‘‘positive’’ (any agglutination within the gel), ‘‘negative’’ (no agglutination)
using neat (undiluted) serum, or ‘‘borderline.’’ However, when a positive or
borderline test result was obtained, Alberio et al. (2003) repeated the assay
with undiluted and serially diluted plasma (up to 1 in 1024) until the result was
negative. The reported titer was the last positive result followed by either
borderline or negative results. Patients judged clinically to have had ‘‘prob-
able’’ or ‘‘highly probable/definite’’ HIT had antibody titers of 4 or more in
39 of 54 (72%) cases, compared with only 2 of 85 (2%) judged ‘‘unlikely’’ to
have had HIT. Further, all 19 of the patient samples that tested positive in a
c-PRP aggregation assay tested positive in the PaGIA (generally, in a titer
of 8 or higher). Among all patients studied, the percentage with associated
thrombotic complications increased from 8% (negative or low titer) to 55%
(positive titer 4–16) to 74% (positive titer 32–256). This study suggests that
reporting quantitatively the results of the HPF4–PaGIA—with a titer of 4 or
more being clinically significant—may increase diagnostic usefulness.
E. Comparison of Activation and Antigen Assays

Both PF4–heparin–EIA and washed platelet activation assays have approx-
imately equal sensitivity for clinical HIT (Greinacher et al., 1994a; Warkentin
et al., 2000). For serum or plasma samples that are known to be positive by
one sensitive washed platelet activation assay (e.g., SRA or HIPA), the
corresponding probability of the PF4–heparin–EIA for confirming the posi-
tive result is at least 75–90% (Greinacher et al., 1994a; Arepally et al., 1995).
Conversely, a similar percentage of referred samples that test positive in the
EIA will also test positive using a washed platelet activation assay (Grei-
nacher et al., 1994a). The sensitivity of both EIA and SRA was even higher
(>90%) for detecting antibodies that caused HIT in prospectively studied
postoperative patients (Warkentin et al., 2000).
Although both antigen and activation assays have similarly high sen-
sitivity for clinical HIT, there is evidence that antigen assays have greater
sensitivity for detecting HIT antibodies not associated with thrombocytope-
nia or other clinical events (Amiral et al., 1995; Arepally et al., 1995; Bauer et
al., 1997; Warkentin et al., 2000) (Fig. 4). Stated another way, the SRA is
more specific for clinical HIT than the antigen assay. The biological expla-
nation for greater specificity of a sensitive activation assay for clinical HIT,
compared with an antigen assay, could relate to the functional heterogeneity
of HIT antibodies against antigenic determinants on PF4, only some of which
activate platelets strongly (Amiral et al., 2000). Data reported by Visentin and
colleagues (1994) also support a higher sensitivity of antigen assays for de-
tecting HIT antibodies. These workers studied 12 HIT plasmas that tested
positive in both SRA and PF4–heparin–EIA. However, at a 1:100 sample
dilution, only 2 of the 12 samples still tested positive in the activation assay. In
contrast, even at a 1:200 dilution, all 12 plasmas still tested positive in the EIA.
Bachelot and colleagues (1998) observed that HIT plasmas that tested only
weakly positive in the PF4–heparin–EIA tended to give negative washed
platelet SRA results when using platelets with the least reactive FcgIIA
receptor genotype, Arg131 (see Chap. 9).
The difference in sensitivity for HIT antibodies between the PF4–hep-
arin–EIA and aggregation studies using c-PRP is considerable. Only about
33–64% of samples that test positive in the PF4–heparin–EIA also test pos-
itive using c-PRP aggregation (Greinacher et al., 1994a; Nguyên et al., 1995;
Rugeri, et al., 1999). Although one laboratory reported a greater sensitivity
using c-PRP aggregation than the EIA (Look et al., 1997), these workers did
not employ a two-point method, and so may have observed false-positive
results using the aggregation assay.
Table 7 summarizes possible explanations for discrepancies in results of
activation and antigen assays for HIT.
IV. INTERPRETATION OF HIT TEST RESULTS
It is important to incorporate clinical information into the interpretation of

any laboratory result for HIT. This is because thrombocytopenia, whether or
not caused by HIT, is common in hospitalized patients receiving heparin, and
because nonpathogenic HIT antibodies are often detected by sensitive assays
in patients who have received heparin for 5 or more days.
Several clinical scoring methods have been described to help estimate

the probability of HIT independently of the HIT antibody test results
(Greinacher et al., 1994a; Pouplard et al., 1997; Warkentin, 2003a; Warkentin
and Heddle, 2003). Some include assessing the platelet count recovery upon
stopping heparin, and so may be more useful when reviewing a case after its
clinical evolution. Chapter 3 provides an example of one scoring system to
estimate the pretest probability of HIT that can be applied at the time of initial
diagnostic assessment.
A. Rapid Versus Typical Onset of Thrombocytopenia

Diagnostic algorithms taking into account the pretest probability of HIT are
shown in Fig. 5A (activation assay screen) and 5B (antigen assay screen). We
have organized the diagnostic approach based on the timing of onset of
thrombocytopenia, either rapid (<5 days) or typical (z5 days) (see Chap. 3).
In general, there are two broad pretest probabilities for patients with
rapid thrombocytopenia: low and high. Patients with low pretest probability
for HIT are those who have not recently been exposed to heparin (thus, they
would not be expected to have circulating HIT antibodies, or to have gen-
erated them so quickly), or who have another good explanation for throm-
bocytopenia. (An important caveat is that sometimes a recent heparin
exposure is not known to the patient or has not been documented in the
medical records.) With a low pretest probability for HIT, either of the sen-
sitive assays for HIT (washed platelet activation assay or antigen assay) can
reliably rule out HIT. However, an unexpected negative result in a patient
with a high pretest probability, or an unexpected positive result in a patient
Figure 4 Comparison of activation and antigen assays for HIT-IgG: analysis of

prospective studies. Quantitative results of an activation assay, the SRA, are shown
on the x-axis (although samples that gave <20% serotonin release are shown without
reference to the actual quantitative result obtained [see box designated <20%]);
quantitative results of the antigen assay (which detected only IgG anti-PF4–H
antibodies) are shown on the y-axis. (A) Orthopedic surgery patients who received
UFH; (B) orthopedic surgery patients who received LMWH; and (C) cardiac surgery
patients who also received postoperative UFH. The arrows indicate the data points
corresponding to the 15 patients who developed clinical HIT (>50% platelet count
fall from the postoperative peak). The data show similarly high sensitivity of the
activation and antigen assays for clinical HIT; however, the activation assay had
higher specificity for clinical HIT. Most sera (13/15, 87%) from patients with clinical
HIT strongly activated platelets (>80% serotonin release). (From Warkentin et al.,
2000.)
Table 7 Discrepancies in Results of HIT Antibody Testing: Possible Explanations and Implications
296
c-PRP Washed platelet

aggregation activation assay Interpretation (assumes clinical Possible explanations
assay (SRA or HIPA) PF4–H–EIA picture compatible with HIT) and implications
Neg Neg Neg Not HIT Continue or resume heparin

Pos Pos Pos HIT Stop heparin and give alternative
anti-coagulant
Pos Neg Neg Not HIT False-positive aggregation assay
(see text); continue or resume
heparin
Pos Pos Neg HIT caused by antibodies against Stop heparin and give alternative
‘‘minor’’ antigen anti-coagulant
Neg Pos Pos HIT (weak-moderate antibodies) Stop heparin and give alternative
anti-coagulant
Neg Neg Pos Weak HIT-IgG antibodies; IgM/IgA Stop heparin and give alternative
anti-PF4–H antibodies anticoagulant if clinical picture
warrants; otherwise, maintain
heparin
Neg Pos Neg Possible thrombin artifact or HIT Repeat assay with a new heat-
caused by antibodies against inactivated serum aliquot; stop
‘‘minor’’ antigen heparin and give alternative
anticoagulant if clinical picture
warrants
Pos Neg Pos HIT (false-negative washed platelet Stop heparin and give alternative
assay) anticoagulant; check washed
platelet activation assay method
Pos, positive test result; Neg, negative test result.
Warkentin and Greinacher
Figure 5 (A) Diagnostic algorithm using a washed platelet activation assay as a

screening test for HIT, either the serotonin-release assay (SRA) or the heparin-induced
platelet activation (HIPA) test. (B) Diagnostic algorithm using PF4–H–EIA as a
screening test for HIT. Abbreviations: IND, indeterminate test result; NEG, negative
test result; POS, positive test result. The asterisk (*) in B indicates that a washed
platelet activation assay could be useful in a patient with moderate pretest probability
for HIT and a positive PF4–heparin–EIA, as a positive activation assay has a higher
specificity for clinical HIT.
Figure 5 Continued.
with a low pretest probability, should lead to repeating the test or perfor-
mance of the complementary activation or antigen assay. Additionally,
further clinical information should be sought (e.g., Has another explanation
for the thrombocytopenia become apparent? Could the patient have had an
unrecognized recent heparin exposure?).
In contrast, for patients with the typical temporal onset of thrombocy-
topenia (i.e., a platelet count fall that begins 5–10 days after beginning heparin
treatment), we believe that, in general, there are two different pretest
probabilities for HIT: moderate and high. Because HIT is a relatively
common explanation for thrombocytopenia that begins during this charac-
teristic time period, it should be considered a plausible diagnosis even if

another possible explanation for thrombocytopenia is identified (hence, a
moderate pretest probability). In a patient without another apparent expla-
nation for thrombocytopenia, or one in whom an unexplained new throm-
botic event has occurred, the pretest probability for HIT would be considered
to be high.
B. Diagnostic Interpretation of Laboratory Results

In patients with a high pretest probability of HIT who have a negative
screening test, the test should be repeated and the complementary activation
or antigen assay should be performed. The diagnosis of HIT is very unlikely if
both activation and antigen assays are negative. In patients with a moderate
pretest probability who have one or more positive tests for HIT, the final
diagnosis may well rest on the overall clinical picture, rather than on the test
result alone (Fig. 6). This conclusion results from two clinical realities: (1)
sensitive HIT assays frequently detect clinically insignificant HIT antibodies
in patients who have received heparin for more than 5 days, and (2) throm-
bocytopenia—whether caused by HIT or not—is common in clinical practice
(see Chap. 4). There is evidence that positive washed platelet activation assays
for HIT have greater diagnostic specificity for clinical HIT (Warkentin et al.,
2000), especially when strong, rapid platelet activation is produced by patient
serum. Regardless, these considerations underscore the importance of con-
ceptualizing HIT as a clinicopathologic syndrome, in which both clinical
information and results of HIT antibody testing are used for diagnosis.
The diagnostic usefulness of certain laboratory tests for HIT is shown
in Fig. 7 (Warkentin, 2003b). Both the SRA and an in-house EIA that detect
only HIT-IgG were very sensitive and specific for clinical HIT in post–
orthopedic surgery patients. The diagnostic usefulness of these assays was
somewhat less in a post–cardiac surgery population. For example, among
post–cardiac surgery patients, the likelihood ratio of a strong-positive SRA
result (90% serotonin release; see open circle in Fig. 7) is about 20. The
likelihood ratio, which is defined as the extent to which a given test result
alters the physician’s estimate of the pretest probability of HIT, is defined as
sensitivity/1 specificity. In this example, the corresponding likelihood ratio
is 0.70/(1 0.965) = 20. Thus, if the physician had estimated a pretest
probability of 50% (odds of 0.5:0.5), then this test result would increase the
posttest probability to more than 95% (0.5:0.5 20:1 = 20:1, or 95.2%). In
contrast, the high sensitivity of this assay to detect clinically important HIT
antibodies (>95%) means that a negative test result lowers the posttest
probability to less than 5%.
Figure 6 Diagnosis of HIT in doubt: Although this 71-year-old patient with mod-
erate pretest probability of HIT had positive laboratory testing for HIT antibodies,
the clinical course casts doubt on the actual role HIT played in the thrombocyto-
penia and thrombotic events. Thrombocytopenia began on day 5 of heparin treat-
ment (nadir, 31 109/L; day 8), together with clinical and laboratory evidence for
septicemia. Laboratory testing for HIT antibodies was strongly positive by activa-
tion assay (SRA, 90% release at 0.1 U/mL heparin, <5% release at 100 U/mL
heparin) and weakly positive by PF4–heparin–EIA: O.D. = 0.709). Clinical evidence
for HIT includes the symptomatic DVT and (possible) pulmonary embolism; how-
ever, the dramatic increase in platelet count during therapeutic-dose heparin treat-
ment and the bacteremia are strong evidence for septicemia, rather than HIT, as an
explanation for the thrombocytopenia.
The diagnostic impact of such a strong-positive SRA result (90% sero-

tonin release) is even greater in post–orthopedic surgery patients, for whom
the corresponding likelihood ratio is about 85 (0.85/1-0.99). As before, a
negative test result essentially rules out HIT.
Although the EIA that detects only HIT-IgG antibodies has lower di-
agnostic specificity than the SRA, it remains a useful assay. The likelihood
ratios for a strong positive test result (e.g., optical density of 1.5) range from
about 10 to 40 for post–cardiac and post–orthopedic surgery patients, respec-
tively. Also, its high sensitivity (>95%) means that a negative test generally
rules out HIT.
Thus, HIT antibody testing is among the most useful of platelet im-
munology assays. For comparison, Fig. 7 also shows the profile of a ‘‘nonin-
formative assay’’ (see line A). This is the profile for various tests of ‘‘platelet-
associated IgG’’ for the diagnosis of autoimmune thrombocytopenia. Certain
glycoprotein-specific platelet antibody tests have operating characteristics
intermediate between those for HIT and a noninformative assay. For
example, the MAIPA (monoclonal antibody immobilization of platelet
antigens) assay has only moderate sensitivity but high specificity for diagnosis
of autoimmune thrombocytopenia (see Chap. 2).
Figure 7 Sensitivity-specificity tradeoffs for diagnosis of HIT (receiver operating

characteristic curve analysis). The arrows indicate various cut-offs between positive
and negative test results, e.g., the open circle indicates 90% serotonin release (post–
cardiac surgery patient) using a washed platelet activation assay (serotonin release
assay, or SRA). The likelihood ratio for HIT for a given positive test result can be
estimated from the graph, using the formula: likelihood ratio = sensitivity/(1
specificity). Thus, for 90% serotonin release (post–cardiac surgery), the estimated
likelihood ratio is 0.7/(1 0.965) = 20. (From Warkentin, 2003b.)
A practical implication of Fig. 7 is that the magnitude of a positive HIT

antibody test provides diagnostically useful information, with a strong
positive result associated with a greater likelihood of a patient having clini-
cal HIT than a weak positive result (Warkentin, 2003b). Similarly, if two
sensitive and complementary tests for HIT antibodies (washed platelet
activation assay, PF4-dependent EIA) both give negative test results, the
diagnosis of HIT is virtually excluded (even in a patient with a high pretest
probability).
V. IN VITRO CROSS-REACTIVITY
A. Cross-Reactivity Using Activation Assays
Cross-reactivity studies have been performed most frequently using activa-
tion assays. However, there are no standard methods for, or even a standard
definition of, in vitro cross-reactivity. In one study of LMWH and danapa-
roid cross-reactivity, an increase in platelet activation in the presence of the
drug over baseline was used to determine cross-reactivity (Warkentin, 1996).
This definition was used to avoid falsely attributing cross-reactivity to drug-
independent platelet activation that is produced by some patients’ sera. The
reason for this definition was the common phenomenon that platelet activa-
tion can be caused by a patient’s serum even in the absence of added heparin.
In the HIPA test, comparison of the lag time to aggregation can be used to
judge cross-reactivity: if a sample shows platelet aggregation with heparinoid
or LMWH earlier than in the presence of buffer, then cross-reactivity is
present. In general, in vitro cross-reactivity with danaparoid is usually
clinically insignificant (Warkentin, 1996; Newman et al., 1998) (see Chap. 14).
Comparison of c-PRP Versus Washed Platelet Assays

Sensitive washed platelet assays generally show almost 100% cross-reactivity
of HIT antibodies for LMWH (Greinacher et al., 1992; Warkentin et al.,
1995a). Indeed, UFH and LMWH are essentially indistinguishable in these
assays. However, very different results have been reported by investigators
using c-PRP assays (Makhoul et al., 1986; Chong et al., 1989; Kikta et al.,
1993; Vun et al., 1996). Here, LMWH consistently shows less cross-reactivity
compared with UFH. It is possible that differences in nonidiosyncratic hep-
arin-induced platelet activation underlie these observations (see Chap. 5):
UFH is more likely to result in weak platelet activation, including some PF4
release, that leads to amplification of the platelet activation response in the
presence of PF4–heparin-reactive HIT antibodies. In contrast, in washed
platelet assays, IgG-mediated platelet activation, but not nonidiosyncratic

heparin-induced platelet activation, occurs.
B. Cross-Reactivity Using Antigen Assays

Although it is theoretically possible to perform a solid-phase EIA to assess
cross-reactivity (Amiral et al., 1995), this is complicated because the antigen
has to be coated as a complex to the solid phase. This problem has been
overcome in a fluid-phase EIA described by Newman and colleagues (1998).
Because this assay detects binding to a defined quantity of labeled PF4-con-
taining antigen, the assay is able to determine in vitro cross-reactivity more
accurately than the solid-phase EIA. These investigators observed an in vitro
cross-reactivity rate of 88% for LMWH; about half the HIT samples reacted
weakly against danaparoid in their study. The fluid-phase EIA has also been
used to show that the antithrombin-binding pentasaccharide, fondaparinux,
does not cross-read with HIT-IgG antibodies (Warkentin et al., 2003).
VI. MISCELLANEOUS ASSAYS FOR HIT ANTIBODIES
A third group of assays detect heparin-dependent immunoglobulin binding to

platelet membranes (Griffiths and Dzik, 1997). None of these assays appear to
be sufficiently reliable to have gained widespread acceptance for the diagnosis
of HIT. This may be related to the fundamental pathogenesis of HIT: the
pathogenic platelet-activating properties of HIT antibodies appear to be me-
diated by relatively few antibodies. Thus, despite producing strong platelet
activation, concomitant evaluation of heparin-dependent IgG binding to
platelets by flow cytometry, for instance, was either no greater than back-
ground (Warkentin et al., 1994) or, in the presence of PF4, only an average of
5.6 times background (Visentin et al., 1994). Furthermore, only 7 of 12 sera
exhibited any detectable IgG binding by flow cytometry, even though all 12
were positive using conventional activation or antigen assays (Visentin et al.,
1994).
Assays have been developed that attempt to measure an increase in
heparin-dependent platelet-bindable immunoglobulin in the presence of HIT
serum (Howe and Lynch, 1985; Lynch and Howe, 1985; Gruel et al., 1991),
but these whole-platelet EIA methods have been superceded by the more
specific PF4–heparin–EIA described earlier. Immunofluorescence assays
have also been reported (Wolf et al., 1983; Silberman and Kovarik, 1987).
A commercially available platelet-bindable IgG assay that uses anti-
IgG-coated indicator red cells, known as the solid-phase red cell adherence
assay (SPRCA), has been developed (Sinor et al., 1990; Sinor and Stone, 1994;
Leach et al., 1994, 1995, 1997). Platelets are coated onto U-shaped microtiter
platelet wells, and either heparin-serum (immune complex method), or
heparin-albumin with subsequent addition of serum after a wash step (hapten
method), is added. After washing, red cells coated with anti-IgG are added to
the wells. Following centrifugation, the appearance of the indicator red cells
on the well bottoms is scored on a 10-point scale, ranging from negative (tight
red cell button) to strongly positive (diffuse red cell binding on the well
bottom). According to Leach and colleagues (1997), both immune complex
and hapten reaction profiles are commonly seen with putative HIT sera
(immune complex > both > hapten). However, only limited comparisons
with conventional HIT assays have been performed, and the diagnostic use-
fulness of these assays is unknown. Disadvantages include the subjective
scoring system, as well as the need to ensure that HLA and platelet-specific
alloantibodies do not interfere with testing.
There have been attempts to identify specific binding of heparin-depen-
dent antibodies to platelet glycoproteins using immunoblotting and immuno-
precipitation (Lynch and Howe, 1985; Howe and Lynch, 1985; Greinacher
et al., 1994d). However, no study has identified a consistent reaction profile
diagnostic of HIT. There is thus no experimental evidence implicating the
involvement of platelet glycoproteins as an immune target in the pathogenesis
of HIT.
ACKNOWLEDGMENTS
Studies described in this chapter were supported by the Heart and Stroke
Foundation of Ontario (operating grants A2449, T2967, B3763, and T4502),
and by the Deutsche Forschungsgemeinschaft Gir 1096/2-1 and Gir 1096/2-
2. Dr. Warkentin was a Research Scholar of the Heart and Stroke Foun-
dation of Canada.
REFERENCES
Alberio L, Kimmerle S, Baumann A, Taleghani BM, Biasiutti FD, Lämmle B. Rapid

determination of anti-heparin/platelet factor 4 antibody titers in the diagnosis of
heparin-induced thrombocytopenia. Am J Med 2003; 114:528–536.
Almedia JI, Coats R, Liem TK, Silver D. Reduced morbidity and mortality rates of
the heparin-induced thrombocytopenia syndrome. J Vasc Surg 1998; 27:309–
316.
Ardlie NG, Packham MA, Mustard JF. Adenosine diphosphate-induced platelet ag-
gregation in suspensions of washed rabbit platelets. Br J Haematol 1970; 19:7–
17.
induced thrombocytopenia. Thromb Haemost 1992; 68:95–96.
4–heparin complexes in heparin-induced thrombocytopenia: a study of 44 cases.
Amiral J, Pouplard C, Vissac AM, Walenga JM, Jeske W, Gruel Y. Affinity purifica-
tion of heparin-dependent antibodies to platelet factor 4 developed in heparin-
induced thrombocytopenia: biological characteristics and effects on platelet
activation. Br J Haematol 2000; 109:336–341.
Arepally G, Reynolds C, Tomaski A, Amiral J, Jawad A, Poncz M, Cines DB. Com-
parison of the PF4/heparin ELISA assay with the 14C-serotonin assay in the
diagnosis of heparin-induced thrombocytopenia. Am J Clin Pathol 1995; 104:
648–654.
Babcock RB, Dumper CW, Scharfman WB. Heparin-induced immune thrombocy-
topenia. N Engl J Med 1976; 295:237–241.
the FcgRIIa–Arg/His-131 polymorphism in heparin-induced thrombocytope-
Berube C, Mitchell L, Silverman E, David M, Saint Cyr C, Laxer R, Adams M, Vegh
P, Andrew M. The relationship of antiphospholipid antibodies to thromboem-
bolic events in pediatric patients with systemic lupus erythematosus: a cross-
sectional study. Pediatr Res 1998; 44:351–366.
Chong BH, Ismail F, Cade J, Gallus AS, Gordon S, Chesterman CN. Heparin-induced
thrombocytopenia: studies with a new molecular weight heparinoid, Org 10172.
Blood 1989; 73:1592–1596.
Chong BH, Burgess J, Ismail F. The clinical usefulness of the platelet aggregation test
for the diagnosis of heparin-induced thrombocytopenia. Thromb Haemost
1993a; 69:344–350.
Chong BH, Pilgrim RL, Cooley MA, Chesterman CN. Increased expression of platelet
IgG Fc receptors in immune heparin-induced thrombocytopenia. Blood 1993b;
81:988–993.
Chong BH, Murray B, Berndt MC, Dunlop LC, Brighton T, Chesterman CN. Plasma
P-selectin is increased in thrombotic consumptive platelet disorders. Blood
1994; 83:1535–1541.
Collins JL, Aster RH, Moghaddam M, Piotrowski MA, Strauss TR, McFarland JG.
Diagnostic testing for heparin-induced thrombocytopenia (HIT): and enhanced
platelet factor 4 complex enzyme linked immunosorbent assay (PF4 ELISA)

[abstr]. Blood 1997; 90(suppl 1):461a.
Eichler P, Budde U, Haas S, Kroll H, Loreth RM, Meyer O, Pachmann U, Pötzsch B,
Schabel A, Albrecht D, Greinacher A. First workshop for detection of heparin-
induced antibodies: validation of the heparin-induced platelet activation test
(HIPA) in comparison with a PF4/heparin ELISA. Thromb Haemost 1999; 81:
625–629.
Eichler P, Raschke R, Lubenow N, Meyer O, Schwind P, Greinacher A. The ID mi-
cro-typing system for detection of heparin-induced antibodies in compari-
son with functional and antigenic assays [abstr]. Ann Hematol 2001; 80(suppl):
A15.
Favaloro EJ, Bernal-Hoyos E, Exner T, Koutts J. Heparin-induced thrombocytopenia
laboratory investigation and confirmation of diagnosis. Pathology 1992; 24:
177–183.
Fratantoni JC, Pollet R, Gralnick HR. Heparin-induced thrombocytopenia: con-
firmation of diagnosis with in vitro methods. Blood 1975; 45:395–401.
Ginsberg JS, Wells PS, Brill-Edwards P, Donovan D, Moffatt K, Johnston M, Stevens
P, Hirsh J. Antiphospholipid antibodies and venous thromboembolism. Blood
1995; 86:3685–3691.
Gobbi G, Mirandola P, Tazzari PL, Ricci F, Caimi L, Cacchioli A, Papa S, Conte R,
Vitale M. Flow cytometry detection of serotonin content and release in resting
and activated platelets. Br J Haematol 2003; 121:892–896.
Goodfellow KJ, Brown P, Malia RG, Hampton KK. A comparison of laboratory tests
for the diagnosis of heparin-induced thrombocytopenia [abstr]. Br J Haematol
1998; 101(suppl 1):89.
diagnosing heparin-associated thrombocytopenia. Thromb Haemost 1991;
66:734–736.
penia: the antibody is not heparin specific. Thromb Haemost 1992; 67:545–
549.
Greinacher A, Amiral J, Dummel V, Vissac A, Kiefel B, Mueller-Eckhardt C. Lab-
oratory diagnosis of heparin-associated thrombocytopenia and comparison of
factor 4/heparin enzyme-linked immunosorbent assay. Transfusion 1994a; 34:
381–385.
Greinacher A, Feigl M, Mueller-Eckhardt C. Crossreactivity studies between sera of
patients with heparin associated thrombocytopenia and a new low molecular
weight heparin, reviparin [letter]. Thromb Haemost 1994b; 72:644–645.
Greinacher A, Liebenhoff U, Kiefel V, Presek P, Mueller-Eckhardt C. Heparin-asso-
ciated thrombocytopenia: the effects of various intravenous IgG preparations
on antibody mediated platelet activation—a possible new indication for high
dose i.v. IgG. Thromb Haemost 1994c; 71:641–645.

Haemost 1994d; 71:247–251.
Griffiths E, Dzik WH. Assays for heparin-induced thrombocytopenia. Transf Med
1997; 7:1–11.
Gruel Y, Rupin A, Darnige L, Moalic-Reverdiau P, Poumier-Gaschard P, Binet C,
Bardos P, Leroy J. Specific quantitation of heparin-dependent antibodies for the
diagnosis of heparin-associated thrombocytopenia using an enzyme-linked
immunosorbent assay. Thromb Res 1991; 62:377–387.
Harenberg J, Huhle G, Giese C, Wang LC, Feuring M, Song XH, Hoffman U. De-
termination of serotonin release from platelets by enzyme immunoassay in the
diagnosis of heparin-induced thrombocytopenia. Br J Haematol 2000; 109:182–
186.
1989; 74:238–243.
Howe SE, Lynch DM. An enzyme-linked immunosorbent assay for the evaluation
of thrombocytopenia induced by heparin. J Lab Clin Med 1985; 105:554–559.
Jeske W, Fareed J, Eschenfelder V, Iqbal O, Hoppensteadt D, Ahsan A. Biochemical
and pharmacologic characteristics of reviparin, a low-molecular-mass heparin.
Semin Thromb Hemost 1997; 23:119–128.
induced platelet activation in sixteen surgical patients: diagnosis and manage-
ment. J Vasc Surg 1987; 5:101–109.
Kapsch D, Silver D. Heparin-induced thrombocytopenia with thrombosis and
hemorrhage. Arch Surg 1981; 116:1423–1427.
Kelton JG, Sheridan D, Brain H, Powers PJ, Turpie AG, Carter CJ. Clinical useful-
ness of testing for a heparin-dependent platelet-aggregating factor in patients
with suspected heparin-associated thrombocytopenia. J Lab Clin Med 1984;
103:606–612.
925–930.
Kikta MJ, Keller MP, Humphrey PW, Silver D. Can low molecular weight heparins
and heparinoids be safely given to patients with heparin-induced thrombocy-
topenia syndrome? Surgery 1993; 114:705–710.
Kinlough-Rathbone RL, Packham MA, Mustard JF. Platelet aggregation. In: Harker
LA, Zimmerman TS, eds. Methods in Hematology: Measurements of Platelet
Function. Edinburgh: Churchill Livingstone, 1983:64–91.
Lee DP, Warkentin TE, Denomme GA, Hayward CPM, Kelton JG. A diagnostic test
for heparin-induced thrombocytopenia: detection of platelet microparticles
using flow cytometry. Br J Haematol 1996; 95:724–731.
Leach MF, Ammann J, AuBuchon JP. Comparison of solid-phase and standard
techniques for detection of drug-dependent platelet antibodies [abstr]. Trans-
fusion 1994; 34(suppl):17S.
Leach MF, Cooper LK, AuBuchon JP. Detection of drug-dependent platelet anti-
bodies by use of solid-phase red cell adherence techniques. Immunohematology
1995; 11:143–149.
Leach MF, Cooper LK, AuBuchon JP. Detection of drug-dependent, platelet-reactive
antibodies by solid-phase red cell adherence assays. Br J Haematol 1997; 97:
755–761.
Lindhoff-Last E, Gerdsen F, Ackermann H, Bauersachs R. Determination of heparin–
platelet factor 4–IgG antibodies improves diagnosis of heparin–induced throm-
bocytopenia. Br J Haematol 2001; 113:886–890.
Look KA, Sahud M, Flaherty S, Zehnder JL. Heparin-induced platelet aggregation
vs. platelet factor 4 enzyme-linked immunosorbent assay in the diagnosis of
heparin-induced thrombocytopenia–thrombosis. Am J Clin Pathol 1997; 108:
78–82.
Lynch DM, Howe SE. Heparin-associated thrombocytopenia: antibody binding spe-
cificity to platelet antigens. Blood 1985; 66:1176–1181.
Makhoul RG, Greenberg CS, McCann RL. Heparin-induced thrombocytopenia and
thrombosis: a serious clinical problem and potential solution. J Vasc Surg 1986;
4:522–528.
Meyer O, Salama A, Pittet N, Schwind P. Rapid detection of heparin-induced platelet
antibodies with particle gel immunoassay (ID-HPF4). Lancet 1999; 354:1525–
1526.
Mustard JF, Perry DW, Ardlie NG, Packham MA. Preparation of suspensions of
washed platelets from humans. Br J Haematol 1972; 22:193–204.
Nagi PK, Ackermann F, Wendt H, Savoca R, Bosshard HR. Protein A antibody-
capture ELISA (PACE): an ELISA format to avoid denaturation of surface-
adsorbed antigens. J Immunol Methods 1993; 158:267–276.
Newman PM, Swanson RL, Chong BH. Heparin-induced thrombocytopenia: IgG
binding to PF4–heparin complexes in the fluid phase and cross-reactivity with
low molecular weight heparin and heparinoid. Thromb Haemost 1998; 80:292–
297.
Nguyên P, Lecompte T, and Groupe d’Etude sur l’Hémostase et la Thromboses
(GEHT) de la Société Francßaise d’Hématologie. Nouv Rev Fr Hematol 1994;
36:353–357.
Nguyên P, Droullé C, Potron G. Comparison between platelet factor 4/heparin
complexes ELISA and platelet aggregation test in heparin-induced thrombo-
cytopenia [letter]. Thromb Haemost 1995; 74:793–810.
Packham MA, Guccione MA, Perry DW. ADP does not release platelet granule
contents in a plasma-free system [abstr]. Fed Proc 1971; 30:201.
Pengo V, Biasiolo A, Fior MG. Autoimmune antiphospholipid antibodies are directed
against a cryptic epitope expressed when h2-glycoprotein I is bound to a suitable
surface. Thromb Haemost 1995; 73:29–34.
Pfueller SL, David R. Different platelet specificities of heparin-dependent platelet
aggregating factors in heparin-induced thrombocytopenia. Br J Haematol 1986;
64:149–159.

554.
Pötzsch B, Keller M, Madlener K, Müller-Berghaus G. The use of heparinase
improves the specificity of crossreactivity testing in heparin-induced thrombo-
cytopenia [letter]. Thromb Haemost 1996; 76:1118–1122.
Pouplard C, Amiral J, Borg JY, Vissac AM, Delahousse B, Gruel Y. Differences
in specificity of heparin-dependent antibodies developed in heparin-induced
thrombocytopenia and consequences on cross-reactivity with danaparoid
sodium. Br J Haematol 1997; 99:273–280.
Pouplard C, Amiral J, Borg JY, Laporte-Simitsidis S, Delahousse B, Gruel Y.
Decision analysis for use of platelet aggregation test, carbon 14-serotonin
release assay, and heparin-platelet factor 4 enzyme-linked immunosorbent assay
for diagnosis of heparin-induced thrombocytopenia. Am J Clin Pathol 1999;
111:700–706.
Risch L, Bertschmann W, Heijnen IAFM, Huber AR. A differentiated approach to
assess the diagnostic usefulness of a rapid particle gel immunoassay for the
detection of antibodies against heparin-platelet factor 4 in cardiac surgery pa-
tients. Blood Coagul Fibrinolysis 2003; 14:99–106.
Rugeri L, Bauters A, Trillot N, Susen S, Decoen C, Watel A, Jude B. Clinical use-
fulness of combined use of platelet aggregation test and anti PF4-H antibodies
elisa test for the diagnosis of heparin induced thrombocytopenia. Hematology
1999; 4:367–372.
Salem HH, van der Weyden MB. Heparin induced thrombocytopenia. Variable
platelet-rich plasma reactivity to heparin dependent platelet aggregating factor.
Pathology 1983; 15:297–299.
Schnell MK, Giordano KJ, Henry M, Munizza M, Nance S, Murphy S, Warkentin
TE. Diagnosis of heparin-induced thrombocytopenia (HIT): comparison of
methods. Transfusion 1998; 38(suppl):98S.
Silberman S, Kovarik P. Heparin-induced thrombocytopenia: use of indirect immuno-
fluorescence. Ann Clin Lab Sci 1987; 17:106–110.
Sinor LT, Stone DL. Serological confirmation of heparin-induced thrombocytopenia.
Clin Hemost Rev 1994; 8:9–10.
Sinor LT, Stone DL, Plapp FV, Rachel JM, Thompson KS. Detection of heparin-IgG
immune complexes on platelets by solid phase red cell adherence assays.
Immuno-correspondence 1990; 4:1–6.
Stewart MW, Etches WS, Boshkov LK, Gordon PA. Heparin-induced thrombocy-
topenia: an improved method of detection based on lumi-aggregometry. Br J
Haematol 1995; 91:173–177.
Teitel JM, Gross P, Blake P, Garvey MB. A bioluminescent adenosine nucleotide

release assay for the diagnosis of heparin-induced thrombocytopenia [letter].
Tomer A. A sensitive and specific functional flow cytometric assay for the diagnosis of
Tomer A, Masalunga C, Abshire TC. Determination of heparin-induced thrombo-
cytopenia: a rapid flow cytometric assay for direct demonstration of antibody-
mediated platelet activation. Am J Hematol 1999; 61:53–61.
required for detection of antibodies associated with heparin-induced thrombo-
cytopenia/thrombosis. J Lab Clin Med 2001; 138:22–31.
Vun CH, Evans S, Chong BH. Cross-reactivity study of low molecular weight heparin
and heparinoid in heparin-induced thrombocytopenia. Thromb Res 1996;
81:525–532.
Walenga JM, Jeske WP, Fasanella AR, Wood JJ, Ahmad S, Bakhos M. Laboratory
diagnosis of heparin-induced thrombocytopenia. Clin Appl Thrombosis/
Hemostasis 1999; 5(suppl 1):S21–S27.
Wang L, Huhle G, Malsch R, Hoffmann U, Song X, Harenberg J. Determination of
heparin-induced IgG antibody by fluorescence-linked immunofiltration assay
(FLIFA). J Immunol Meth 1999a; 222:93–99.
Wang L, Huhle G, Malsch R, Hoffmann U, Song X, Harenberg J. Determination of
heparin-induced IgG antibody in heparin-induced thrombocytopenia type II.
Eur J Clin Invest 1999b; 29:232–237.
Warkentin TE. Danaparoid (Orgaran) for the treatment of heparin-induced throm-
bocytopenia (HIT) and thrombosis: effects on in vivo thrombin and cross-linked
fibrin generation, and evaluation of the clinical significance of in vitro cross-
reactivity (XR) of danaparoid for HIT-IgG [abstr]. Blood 1996; 88(suppl 1):
626a.
Warkentin TE. Laboratory testing for heparin-induced thrombocytopenia. J Thromb
Thrombolysis 2000; 10:S35–S45.
Br J Haematol 2003a; 121:535–555.
thrombocytopenia [letter]. Arch Pathol Lab Med 2003b; 127:783.
Warkentin TE, Heddle NM. Laboratory diagnosis of immune heparin-induced throm-
bocytopenia. Curr Hematol Rep 2003; 2:148–157.
Warkentin TE, Hayward CPM, Smith CA, Kelly PM, Kelton JG. Determinants of
platelet variability when testing for heparin-induced thrombocytopenia. J Lab
Clin Med 1992; 120:371–379.

1994; 84:3691–3699.
Warkentin TE, Levine MN, Hirsh J, Horsewood P, Roberts RS, Gent M, Kelton
JG. Heparin-induced thrombocytopenia in patients treated with low-molec-
ular-weight heparin or unfractionated heparin. N Engl J Med 1995; 332:1330–
1335.
topenia. Blood 2000; 96:1703–1708.
after elective hip or knee replacement surgery in patients receiving antithrom-
botic prophylaxis with fondaparinux or enoxaparin [abstr]. Blood, 2003. In
press.
White MM, Siders L, Jennings LK, White FL. The effect of residual heparin on
the interpretation of heparin-induced platelet aggregation in the diagnosis of
heparin-associated thrombocytopenia [letter]. Thromb Haemost 1992; 68:88.
Wolf H, Nowack H, Wick G. Detection of antibodies interacting with glycosamino-
glycan polysulfate in patients treated with heparin or other polysulfated glyco-
saminoglycans. Int Arch Allergy Appl Immunol 1983; 70:157–163.
12
Pseudo–Heparin-Induced
Thrombocytopenia
I. INTRODUCTION
A. The Concept of Pseudo-HIT
Heparin-induced thrombocytopenia (HIT) is strongly associated with life-
and limb-threatening venous and arterial thrombosis, including pulmonary
embolism, venous limb gangrene, and large vessel arterial occlusion. How-
ever, HIT is by no means a unique explanation for the combination of throm-
bocytopenia and thrombosis (Table 1).
In these pseudo-HIT disorders, thrombocytopenia usually occurs early
during the course of heparin treatment. This could reflect the prothrombotic
process associated with the patient’s primary diagnosis. Alternatively, hepa-
rin could exacerbate the platelet count fall by nonimmune proaggregatory
effects on platelets (see Chap. 5). If the patient previously received heparin,
physicians might consider HIT in the differential diagnosis of the platelet
count fall.
However, one pseudo-HIT syndrome in particular closely resembles
even the typical day 5–10 timing of thrombocytopenia characteristic of HIT:
adenocarcinoma-associated disseminated intravascular coagulation (DIC).
In these patients, the fall in platelet count begins soon after stopping heparin
treatment. Because the patients usually will have received heparin for 5–10
days to treat adenocarcinoma-associated thrombosis, the timing of the onset
of thrombocytopenia closely resembles immune HIT. Furthermore, the fre-
313
314 Warkentin
Table 1 Pseudo-HIT Disorders Characterized by Thrombocytopenia and Thrombosis
Pathogenesis of thrombocytopenia
Pseudo-HIT disorder and thrombosis Timing
Adenocarcinoma DIC secondary to procoagulant material(s) Latea

produced by neoplastic cells
Pulmonary embolism Platelet activation by clot-bound thrombin Earlyb or latec
Diabetic ketoacidosis Hyperaggregable platelets in ketoacidosis (?) Earlyd
Antiphospholipid Multiple mechanisms described, including Early
antibody syndrome platelet activation by antiphospholipid
antibodies (?)
Thrombolytic therapy Platelet activation by thrombin bound to Earlye
fibrin degradation products (?)
Septicemia-associated Symmetrical peripheral gangrene secondary Early
purpura fulminans to DIC with depletion of protein C and
antithrombin
Infective endocarditis Infection-associated thrombocytopenia; Early
ischemic events secondary to septic emboli
Paroxysmal nocturnal Platelets susceptible to complement-mediated Early
hemoglobinuria damage; platelet hypoproduction
Posttransfusion ‘‘Pseudospecific’’ alloantibody-mediated Late
purpura (PTP) platelet destruction (exception: bleeding,
not thrombosis)
These pseudo-HIT disorders can mimic HIT by causing thrombocytopenia and thrombosis in association
with heparin treatment. An exception is PTP, which causes bleeding, but not thrombosis; however, PTP can
resemble HIT because both disorders usually occur about a week after major surgery requiring blood and
postoperative heparin. The pseudo-HIT disorders can be categorized based on whether the onset of throm-
bocytopenia is typically ‘‘early’’ (<5 days) or ‘‘late’’ (z5 days) in relation to the heparin.
a
See Fig. 1 for an example of pseudo-HIT caused by adenocarcinoma-associated DIC.
b
See Fig. 3 for early thrombocytopenia associated with pulmonary embolism.
c
See Fig. 4 for late thrombocytopenia associated with pulmonary embolism.
d
See Fig. 5 for early thrombocytopenia associated with diabetic ketoacidosis.
e
See Fig. 6 for early thrombocytopenia caused by thrombolytic therapy.
quent occurrence of new or progressive thrombosis in this setting also sug-

gests HIT.
This chapter draws attention to those clinical disorders that can mimic
and, thereby, be confused with HIT. This is not a trivial distinction: whereas
heparin is contraindicated in patients with HIT, it often is the optimal
treatment of patients with pseudo-HIT. Second, the close clinical parallels
between HIT and certain pseudo-HIT disorders can provide insights into the
pathogenesis of thrombosis. For example, the recognition that venous limb
gangrene can complicate metastatic adenocarcinoma, and the clinical paral-
Pseudo–HIT 315
lels with a similar syndrome in HIT patients, suggests that a common factor
(coumarin anticoagulation) may play a crucial pathogenic role in both
disorders (Warkentin, 2001). Likewise, similarities between HIT and the
lupus anticoagulant syndrome suggest that they could also share common
pathogenic mechanisms (Arnout, 1996, 2000; Gruel, 2000).
II. PSEUDO-HIT SYNDROMES

A. Adenocarcinoma
Mucin-producing adenocarcinoma is an important cause of venous and
arterial thrombosis that occurs in association with thrombocytopenia. In
these patients, DIC is often the explanation for the thrombocytopenia. The
diagnosis is suggested by reduced fibrinogen levels (or prolonged thrombin
time), elevated prothrombin time, and elevated cross-linked (D-dimer) fibrin
degradation products (or a positive protamine sulfate ‘‘paracoagulation’’
test).
Adenocarcinoma-associated DIC can strongly resemble HIT (Fig. 1).
Typically, a patient presents with idiopathic deep vein thrombosis (DVT),
sometimes with mild to moderate thrombocytopenia. During treatment with
therapeutic-dose unfractionated or low molecular weight heparin (LMWH)
the platelet count rises, likely because of improved control of DIC by the
heparin. In my experience, this often dramatic rise in the platelet count during
heparin treatment of ‘‘idiopathic’’ DVT is a clinically useful marker for
adenocarcinoma-associated DIC. During the 5- to 10-day period of heparin
treatment with overlapping warfarin anticoagulation, no problems are en-
countered. However, there is rapid recurrence of thrombocytopenia within
hours or days of discontinuing the heparin, despite apparent therapeutic
anticoagulation with warfarin, during which time the patient develops new or
progressive venous, or even arterial, thrombosis. Thus, the onset of throm-
bocytopenia and thrombosis may occur within the 5- to 10-day ‘‘window’’
that suggests HIT.
Venous Limb Gangrene Complicating Adenocarcinoma

The venous thrombotic events complicating adenocarcinoma include DVT,
phlegmasia cerulea dolens, and even venous limb gangrene (Everett and
Jones, 1986; Adamson and Currie, 1993). Clinical and laboratory parallels
between HIT and adenocarcinoma suggest that, paradoxically, coumarin
treatment could contribute to the pathogenesis of venous gangrene in these
patients through a disturbance in procoagulant-anticoagulant balance (War-
kentin, 1996, 2001). Figure 2 summarizes the proposed pathogenesis of this
316 Warkentin
Figure 1 Pseudo-HIT: Adenocarcinoma with thrombocytopenia and phlegmasia

cerulea dolens after stopping administration of unfractionated heparin (UFH). The
late presentation of thrombocytopenia suggested HIT, prompting use of an alternative
anticoagulant (ancrod). Heparin was restarted when HIT antibodies were not detected
by serotonin-release assay (SRA). Subsequently, discontinuation of heparin led to
recurrence of thrombocytopenia and warfarin-associated phlegmasia cerulea dolens
(repeat of pseudo-HIT cycle). Abbreviations: DVT, deep venous thrombosis; INR,
international normalized ratio; PE, pulmonary embolism.
syndrome from the perspective of the characteristic clinical triad of venous

limb gangrene: (1) thrombocytopenia caused by HIT or adenocarcinoma-
associated DIC; (2) acute DVT with acral (distal) microvascular thrombosis;
and (3) a supratherapeutic international normalized ratio (INR) associated
with coumarin therapy.
Venous limb gangrene appears to result from failure of the protein C
anticoagulant pathway to down-regulate thrombin generation within the
microvasculature (Warkentin 1996; Warkentin et al., 1997; see Chap. 3).
Here, the elevated INR may represent a surrogate marker for marked
reduction in functional protein C levels (by a parallel reduction in factor
VII); the thrombocytopenia is a surrogate marker for uncontrolled thrombin
generation associated either with HIT or adenocarcinoma (see Fig. 2). As
Pseudo–HIT 317
Figure 2 The pathogenesis of warfarin-associated venous limb gangrene is shown in

relation to its typical clinical triad—supratherapeutic international normalized ratio
(INR), microvascular thrombosis, and thrombocytopenia. The central paradox is per-
sisting formation of thrombin despite markedly depleted plasma factor VII level,
which is paralleled by severely depleted protein C activity, leading to impaired down-
regulation of thrombin generation in the microvasculature and, consequently, micro-
vascular thrombosis. (From Warkentin, 2001.)
venous limb gangrene occurs in a limb with preceding active DVT, this
suggests that local factors, such as direct extension of thrombosis, as well as
exacerbation of distal thrombosis by venous stasis, contribute to large- and
small-vessel thrombosis characteristic of this syndrome.
Venous thrombosis complicating adenocarcinoma, especially when
complicated by DIC or severe venous ischemia or necrosis, should be treated
with heparin, rather than warfarin or other coumarin anticoagulants. Rever-
sal of warfarin anticoagulation (with vitamin K and plasma infusion, but not
with prothrombin complex concentrates, as these do not contain sufficient
protein C) and prompt control of DIC with heparin could salvage a limb with
severe phlegmasia, or limit damage in a patient with venous gangrene. An
effective agent often is LMWH (Prandoni, 1997; Lee et al., 2003). I recom-
318 Warkentin
mend intermittent monitoring using antifactor Xa levels, because some

patients with heparin resistance require high doses of LMWH to achieve
therapeutic anticoagulation.
Ironically, one of the problems of heparin in these patients is its efficacy:
thus, if heparin is discontinued for any reason, rapid recurrence of thrombo-
cytopenia and thrombosis can result. Figure 1 shows an example in which
thrombocytopenia and pulmonary embolism occurred (day 21) when heparin
was held for a few hours to permit a liver biopsy to diagnose metastatic
carcinoma. I have also observed a patient with lung adenocarcinoma in whom
heparin was held to permit limb amputation; postanesthesia, the patient was
aphasic (intraoperative stroke). Often, recurrent thrombosis is as ‘‘malig-
nant’’ as the cancer itself.
B. Pulmonary Embolism
Mild thrombocytopenia is common in patients with pulmonary embolism.
Sometimes the thrombocytopenia is severe and associated with laboratory
markers of DIC (Stahl et al., 1984; Mustafa et al., 1989) (Fig. 3). The throm-
bocytopenia presumably results from thrombin-induced platelet activation.
Large thromboemboli within the high-flow pulmonary vessels may act as a
reservoir for clot-bound thrombin that is relatively protected from inhibition
by antithrombin-dependent inhibitors (Weitz et al., 1990). This view is
indirectly supported by the observation that thrombocytopenia commonly
occurs in patients with pulmonary embolism, but not in patients with DVT
alone (Monreal et al., 1991; Warkentin et al., 2003a). Further, increased
heparin clearance has been demonstrated in experimental pulmonary embo-
lism (Chiu et al., 1977).
Because HIT is also strongly associated with pulmonary embolism
(Warkentin et al., 1995, 2003a), a diagnostic and therapeutic dilemma results
when a patient presents with pulmonary embolism and thrombocytopenia 5
or more days after surgery managed with postoperative heparin prophylaxis
(Fig. 4). Initiating therapeutic heparin could have catastrophic consequences
for the patient who has circulating HIT antibodies, although in sufficient
doses it is effective for a patient with pulmonary embolism and DIC without
HIT. Because these two possibilities cannot be readily distinguished on
clinical grounds alone, one should manage such a patient with an alternative
anticoagulant until the results of HIT antibody testing become available
(Warkentin, 2000).
C. Diabetic Ketoacidosis
Diabetic ketoacidosis (DKA) can be associated with acute thromboembolic
complications. Evidence for in vivo platelet activation was observed in one
Pseudo–HIT 319
Figure 3 Pseudo-HIT secondary to pulmonary embolism and DIC: An obese, 50-

year-old man with paraplegia was admitted for recurrent hypotension. He initially
received twice-daily (b.i.d.) subcutaneous (sc), unfractionated heparin (UFH) for anti-
thrombotic prophylaxis, as the initial diagnosis was septicemia. Deep vein thrombosis
(DVT) and pulmonary embolism (PE) were then diagnosed (Dx), and therapy changed
to intravenous UFH, 1200 U/h. The platelet count fell over 4 days to a nadir of 30
109/L; danaparoid sodium (DS) was given because of concern over possible HIT (there
was a remote history of previous heparin use). An echocardiogram showed large right
atrial thrombus (likely representing a leg vein embolus), and the patient was
transferred to a cardiac surgical center. The platelet count fall was judged too rapid to
be HIT (see Chap. 3), a viewpoint supported by negative testing for HIT antibodies by
serotonin-release assay (SRA) and PF4–heparin enzyme-linked immunosorbent assay
(EIA). UFH administration was restarted in higher doses with antifactor Xa
monitoring to overcome heparin resistance. Recurrent hypotension occurred when
the right atrial thrombus embolized; full hemodynamic and platelet count recovery
occurred following tissue plasminogen activator (t-PA) administration, followed by
UFH, then low molecular weight heparin (LMWH), and (later) warfarin treatment.
The patient was well at 3-year follow-up, without evidence of carcinoma.
Figure 4 Pseudo-HIT associated with pulmonary embolism versus HIT: (A) A
patient developed a platelet count fall from 387 to 159 109/L (59% fall) that began on
day 7 of UFH prophylaxis following orthopedic surgery. Pulmonary embolism (PE)
was diagnosed by pulmonary angiography on postoperative day 11. The platelet count
fell during initial intravenous heparin therapy, rising only when sufficient UFH was
given (2360 U/h) to overcome ‘‘heparin resistance’’ (as shown by subtherapeutic acti-
vated partial thromboplastin times, aPTTs). HIT antibodies were not detectable (day
12), either by serotonin-release assay (SRA, <5% release) or PF4–heparin–EIA (op-
tical density, 0.149; negative, <0.450). (B) A platelet count profile similar to that seen
in Fig. 4A also occurred in a patient who developed a platelet count fall from 378 to 161
109/L (57% fall) that began on day 7 after cardiac surgery in which unfractionated
heparin (UFH) was given for cardiopulmonary bypass (CPB). The platelet count
recovered on therapeutic-dose danaparoid. Only one clinical clue pointed to the diag-
nosis of HIT: erythematous skin lesions at the UFH injection sites were also observed
on day 7 (not shown on figure). Testing for HIT antibodies was strongly positive in the
SRA (98% release at 0.1 U/mL heparin; 0% release at 100 U/mL heparin and at 0.1 U/
mL heparin in the presence of Fc receptor–blocking monoclonal antibody). The
similar platelet count profiles between these patients illustrate the difficulty in
determining on clinical grounds whether postoperative PE is caused by HIT or not.
Pseudo–HIT 321
study of 10 patients who had elevated plasma levels of platelet factor 4 and h-
thromboglobulin during DKA that resolved following recovery (Campbell et
al., 1985) Evidence for activation of coagulation includes elevated fibrin
degradation products and reduced antithrombin (Paton, 1981). Figure 5
illustrates a patient with ‘‘white clots’’ in the femoral artery, leading to
amputation, who was initially thought to have HIT. However, HIT antibody
testing and subsequent clinical events proved that the patient did not have
HIT as the initial explanation for this dramatic clinical presentation of
thrombocytopenia and thrombosis complicating DKA (although HIT oc-
curred later in the clinical course). I am also aware of a patient with essential
thrombocythemia who developed postoperative DKA, thrombocytopenia,
and bilateral lower limb artery thrombosis that occurred too early during
UFH prophylaxis (days 2–3) to have been caused by immune HIT. A similar
example of early-onset severe thrombocytopenia and arterial thrombosis
resulting in amputation of an arm was reported in a patient with diabetic
ketoacidosis and adult respiratory distress syndrome (ARDS) (Phillips et al.,
1994). Although the authors suggested HIT secondary to heparin ‘‘flushes’’ as
the diagnosis, pseudo-HIT seems more likely based upon the temporal
features of the case, as well as the negative laboratory testing for HIT
antibodies.
D. Antiphospholipid Antibody Syndrome or Lupus

Anticoagulant Syndrome
Clinical Features
Antiphospholipid antibodies can be detected either as ‘‘lupus anticoagu-
lants’’ or as anticardiolipin antibodies (Asherson et al., 1989; Ginsberg et al.,
1995) (see Chap. 11). Antiphospholipid antibody syndrome (APLAS) is char-
acterized by increased risk for thrombosis and recurrent fetal loss; limb or
intra-abdominal vein thrombosis, cerebral venous (dural sinus) thrombosis,
nonatheromatous arterial thrombosis, cardiac valvulitis, and microvascular
thrombosis (e.g., acrocyanosis, ‘‘blue toe syndrome’’ digital ulceration or
gangrene, livedo reticularis) are described (Hojnik et al., 1996; Gibson et al.,
1997). Many patients have thrombocytopenia (Morgan et al., 1993; Galli et
al., 1996), which is typically mild and intermittent. The explanation for throm-
bocytopenia is uncertain: some patients have platelet-reactive autoantibodies
(Galli et al., 1994; Lipp et al., 1998), but platelet-activating effects of IgG are
also suspected.
The explanation for the prothrombotic tendency of APLAS is also
elusive. A multifactorial pathogenesis is likely, because the antibodies recog-
nize complexes of negatively charged phospholipids with many different pro-
322 Warkentin
Figure 5 Pseudo-HIT during diabetic ketoacidosis (DKA), later complicated by

HIT: A 27-year-old man developed rapid onset of thrombocytopenia and white clots in
the left femoral artery (at a femoral artery catheter site) during management of DKA
that included prophylactic-dose unfractionated heparin (UFH). HIT was suspected
erroneously on the basis of a possible previous remote heparin exposure (gastric sur-
gery 10 years earlier). The patient underwent two embolectomies as well as treat-
ment with urokinase and intravenous (iv) danaparoid. The patient developed a second
platelet count fall during danaparoid treatment that began on day 6 in relation to the
initial course of UFH. Tests for HIT antibodies changed from negative (serotonin-
release assay [SRA]: days 1 and 4, serotonin release <5%) to positive (days 9 and 12,
serotonin release 92% and 80%, respectively). By PF4–heparin–EIA (set up to detect
IgG antibodies), the day 1 sample also was negative (O.D., 0.262; negative, <0.450),
the day 4 sample was weakly positive (0.804), and the day 9 and 12 samples were
strongly positive (1.863 and 1.002, respectively). Although the possibility of in vivo
cross-reactivity of danaparoid with the HIT antibodies is suggested by the throm-
bocytopenia and progression of limb ischemia, the platelet count subsequently rose
during danaparoid treatment, and no additional thromboembolic events occurred. In
vitro cross-reactivity was detected on the day 9, but not the day 12, blood sample.
Pseudo–HIT 323
tein cofactors, such as h2-glycoprotein I (h2GP I), prothrombin, protein C,

protein S, and annexin V (Galli, 1996; Triplett, 1996) (see Chap. 2). Indeed,
interference with endothelial cell function, impaired fibrinolysis, disturbances
in protein C anticoagulant pathway activities, and antibody-mediated platelet
activation have all been described (for review see Petri, 1997; Gruel, 2000;
Arnout and Vermylen, 2003).
Parallels Between APLAS and HIT

Table 2 lists some common features of APLAS and HIT. Both clinicopath-
ologic disorders are characterized by thrombocytopenia, a paradoxical risk
for venous and arterial thrombosis, and associated antibodies that can be
detected by either functional or antigen assays (see Chap. 11): Moreover, for
both APLAS and HIT, positive functional assays are more strongly associ-
ated with thrombosis than positive antigen assays (Ginsberg et al., 1995;
Table 2 Clinical Parallels Between HIT and APLAS
Heparin-induced Antiphospholipid antibody

thrombocytopenia (HIT) syndrome (APLAS)
Thrombotic paradox Thrombosis despite Thrombosis despite

thrombocytopenia prolonged coagulation
tests (F thrombocytopenia)
Spectrum of Venous z arterial Venous z arterial thrombosis;
thrombotic events thrombosis; adrenal adrenal infarction, dural
infarction, dural sinus thrombosis
sinus thrombosis
Severity of Mild to moderate Mild to moderate
thrombocytopenia thrombocytopenia thrombocytopenia
Laboratory diagnosis (1) Platelet activation (1) Lupus anticoagulant
by (1) functional or assays (e.g., serotonin (i.e., prolonged
(2) antigen assays release assay, phospholipid-dependent
heparin-induced coagulation assay in
platelet activation test); presence of patient plasma);
(2) platelet factor (2) h2GP I-dependent
4–heparin–EIA anticardiolipin–EIA
Pathogenesis Platelet activation Uncertain pathogenesis:
by platelet Fc receptors; immune platelet activation
endothelial activation and endothelial injury
by immune injury are possible factors
Further laboratory parallels between HIT and APLAS are discussed in Chap. 11.
Abbreviations: h2GP I, h2 glycoprotein I; EIA, enzyme-linked immunosorbent assay.
324 Warkentin
Warkentin et al., 2000; Galli et al., 2003). The parallels between these
disorders led Arnout (1996) to hypothesize that IgG-mediated platelet
activation could explain thrombosis in APLAS. Supportive experimental
data include the observations that antiphospholipid antibodies enhance
platelet activation induced by other agonists (Martinuzzo et al., 1993).
Furthermore, Arvieux et al. (1993) observed that murine monoclonal anti-
bodies reactive against h2GP I induced platelet activation in the presence of
subthreshold concentrations of ADP and epinephrine, an effect dependent on
binding to platelet FcgIIa receptors. However, other workers were unable to
demonstrate enhanced platelet activation in the presence of IgG antiphos-
pholipid antibodies (Shi et al., 1993; Ford et al., 1998) or showed no role for
platelet FcgIIa receptors (Lutters et al., 2001; Jankowski et al., 2003).
Thrombocytopenia in Patients with APLAS Receiving Heparin

In retrospective studies, Auger and colleagues (1995) reported that platelet
counts typically fell by about 50% in patients with chronic thromboembolic
disease and the lupus anticoagulant who were treated with heparin. Neither
timing of the onset of thrombocytopenia nor results of specific antigen or
activation assays for HIT antibodies were reported, so it remains uncertain
whether these patients had (immune) HIT. It is possible that nonidiosyncratic
platelet activation caused by heparin could increase the thrombocytopenic
potential of antiphospholipid antibodies in the absence of HIT antibodies.
Alternatively, some patients with APLAS may have low levels of circulating
HIT antibodies even in the absence of previous heparin exposure (Lasne et al.,
1997; Martinuzzo et al., 1999). Recently, this view has gained some experi-
mental support. Bourhim and colleagues (2003) showed that affinity-purified
IgM anti-h2GP I from a patient with APLAS gave a positive reaction in PF4-
dependent EIAs. Further, mice actively immunized with the purified IgM
anti-h2GP I generated anti-h2GP I antibodies (via an idiotype–anti-idiotype
mechanism) that also cross-reacted with PF4–heparin.
E. Thrombolytic Therapy
Acute thrombocytopenia is common in patients treated with streptokinase,
especially when combined with heparin (Balduini et al., 1993) (Fig. 6). This
could represent a direct, activating stimulus of heparin on platelets that
perhaps is exacerbated by procoagulant effects of thrombolytic therapy. For
example, fibrin degradation products generated by thrombolytic agents bind
and protect thrombin from inhibition by heparin (Weitz et al., 1998). Such a
mechanism could explain thrombocytopenia after use of any thrombolytic
drug.
Pseudo–HIT 325
Figure 6 Pseudo-HIT associated with thrombolytic therapy. A 72-year-old man

developed moderate thrombocytopenia shortly after receiving streptokinase and hep-
arin, which resolved following discontinuation of heparin. The early onset of throm-
bocytopenia, as well as the negative testing for HIT-IgG using the platelet serotonin
release assay, was consistent with pseudo-HIT. Abbreviations: i.v., intravenous; p.o.,
per os; s.c., subcutaneous. (From Warkentin and Kelton, 1994.)
However, some investigators have reported that plasma containing

anti-streptokinase antibodies can activate platelets through their Fc receptors
in the presence of streptokinase (Vaughan et al., 1988; Lebrazi et al., 1995;
Regnault et al., 2003). Thus, high-titer antistreptokinase antibodies found in
some normal individuals could explain the occasional occurrence of throm-
bocytopenia and thrombosis following treatment with streptokinase.
F. Septicemia-Associated Purpura Fulminans

Septicemia complicated by DIC occasionally results in progressive ischemia
and necrosis of fingers or hands and toes or feet, producing a syndrome of
symmetrical peripheral gangrene also known as purpura fulminans (Knight et
al., 2000). The association with DIC suggests that increased thrombin
generation in vivo, together with severe consumption and depletion of natural
anticoagulant factors (e.g., protein C, protein S, antithrombin), leads to
326 Warkentin
dysregulated fibrin deposition in the microvasculature. Other contributing

factors can include hypotension or shock, pharmacological vasoconstriction
(e.g., dopamine, epinephrine, norepinephrine) (Winkler and Trunkey, 1981;
Hayes et al., 1992), vessel injury from invasive catheters, impaired hepatic
synthesis of natural anticoagulants (e.g., vitamin K deficiency), postsplenec-
tomy status, or congenital deficiency of natural anticoagulants. Rarely,
purpura fulminans occurs several weeks after varicella infection, usually
because of autoantibodies reactive against protein S (Smith and White, 1999).
Meningococcemia in particular is often complicated by peripheral tissue
necrosis that seems to parallel the severity of protein C depletion (Fijnvan-
draat et al., 1995). Recent trials suggest that protein C replacement therapy
improves the natural history of this infection (Smith and White, 1999; White
et al., 2000). Other infections that sometimes are complicated by symmetrical
peripheral gangrene include septicemia secondary to pneumococcus (Johan-
sen and Hansen Jr., 1993), Escherichia coli (Rinaldo and Perez, 1982), Hae-
mophilus influenzae type b (Hayes et al., 1992), and Capnocytophaga
canimorsus (Kullberg et al., 1991), among others. Sometimes severe systemic
inflammatory response syndromes, such as ARDS, in the absence of demon-
strable infection, can be complicated by limb necrosis (Bone et al., 1976).
Acquired antithrombin deficiency in such patients with ARDS could be
associated with thrombosis (Owings et al., 1996).
The development of acral tissue ischemia or necrosis in a thrombocy-
topenic, septic patient receiving heparin may suggest HIT. Although a
common therapeutic response to such a diagnostic dilemma might be to stop
heparin pending results of diagnostic testing for HIT antibodies, this could
result in further ischemic injury, because anticoagulants might help prevent
microvascular thrombosis (White et al, 2000). Furthermore, alternative non-
heparin anticoagulants could be relatively contraindicated in a patient with
significant renal or hepatic dysfunction. Thus, a reasonable treatment ap-
proach might well include continued heparin if clinical judgment posited a
higher likelihood of septicemia, rather than HIT, as the cause of the
microvascular thrombosis.
Only a small minority of septic patients develop acral limb ischemia or
necrosis. Many, however, develop thrombocytopenia, with or without labo-
ratory evidence for DIC. The predominant explanation for increased platelet
destruction in sepsis is uncertain, but appears to involve the underlying
inflammatory host response (Aird, 2003a,b). Since hospitalized septic patients
frequently are exposed to heparin, diagnostic confusion with HIT can result.
Low protein C levels correlate with poor outcomes in sepsis (Yan et al., 2001),
and recombinant human activated protein C (drotrecogin, Xigris) has been
shown to reduce mortality in septic patients (Bernard et al., 2001). It is
possible that this therapy might reduce risk of limb ischemia from microvas-
cular thrombosis in this patient population. A potential dilemma is that septic
Pseudo–HIT 327
patients with severe thrombocytopenia (<30 109/L) were excluded in the

clinical trials because of the bleeding potential of drotrecogin; however, as
relative and absolute efficacy was greatest in the patients with the most severe
sepsis, it has been suggested that otherwise eligible patients with such severe
thrombocytopenia be considered as candidates for drotrecogin following
platelet transfusion (Warkentin et al., 2003b).
G. Infective Endocarditis
Infective endocarditis is frequently complicated by thrombocytopenia. These
patients are also at risk for septic emboli manifesting as thrombotic or
hemorrhagic stroke, myocardial infarction, renal infarction, or even acute
limb ischemia (de Gennes et al., 1990). Thus, the profile of macrovascular
thrombosis and thrombocytopenia characteristic of HIT can be mimicked,
especially as heparin is often used to anticoagulate patients with septic
endocarditis (Delahaye et al, 1990). Microembolization leading to multiple
small infarcts or microabscesses, in such organs as muscles, adrenal glands,
and spleen, is an additional feature of endocarditis (Ting et al., 1990) that is
not seen in HIT.
H. Paroxysmal Nocturnal Hemoglobinuria

Paroxysmal nocturnal hemoglobinuria (PNH) is a clonal myeloid disorder
characterized by an acquired defect in the X-linked phosphatidylinositol
glycan class A (PIG-A) gene, leading to loss of cell surface glycosylphospha-
tidylinositol (GPI)-anchored proteins (for review see Rosse, 1997). Loss of the
complement-regulating GPI-linked surface proteins, decay-accelerating fac-
tor and membrane attack complex inhibitory factor, causes the red cells to be
exquisitely sensitive to complement-mediated hemolysis. Some patients have
thrombocytopenia, and an increased risk for unusual, life-threatening venous
thrombotic events, such as hepatic vein thrombosis, occurs in some patients.
Thus, the clinical profile of HIT potentially can be mimicked. The thrombo-
cytopenia could be related either to decreased platelet production or to
complement-mediated formation of procoagulant platelet-derived micropar-
ticles (Wiedmer et al., 1993).
I. Posttransfusion Purpura
Posttransfusion purpura (PTP) is a rare syndrome characterized by severe
thrombocytopenia and mucocutaneous bleeding that begins 5–10 days after
blood transfusion, usually red cell concentrates. More than 95% of affected
patients are older women, in keeping with its pathogenesis of an anamnestic
recurrence of platelet-specific alloantibodies in women previously sensitized
328 Warkentin
by pregnancy. Destruction of autologous platelets is believed to result from

the pseudospecificity of the alloimmune response, e.g., the high-titer anti-
HPA-la alloantibodies (the most frequent cause of the syndrome) probably
somewhat recognize the autologous HPA-1b alloantigen (see Chap. 2).
Because both PTP and HIT typically occur about a week after surgery
managed with perioperative blood transfusions and postoperative heparin
prophylaxis, a diagnostic dilemma can arise (Lubenow et al., 2000). A useful
clinical clue is the presence or absence of petechiae: PTP almost invariably is
characterized by this hallmark of severe thrombocytopenia, whereas patients
with HIT generally do not develop petechiae, even if they have very severe
thrombocytopenia. The presence of high titers of platelet-reactive alloanti-
bodies suggests PTP.
III. RECOGNITION AND TREATMENT OF PSEUDO-HIT
Many patients with pseudo-HIT can be distinguished from HIT because of

the early onset of thrombocytopenia (see Table 1). Unless the patient received
heparin within the past 100 days, the early platelet count fall is strong evidence
against HIT (Warkentin and Kelton, 2001; Lubenow et al., 2002) (see Chap.
3). These patients with pseudo-HIT should be further anticoagulated with
heparin.
However, for patients with adenocarcinoma-associated DIC, or post-
operative pulmonary embolism, in whom the platelet count fall can occur
after 5 days of heparin treatment, the diagnosis will be initially uncertain. As
heparin can cause catastrophic complications if HIT is the underlying cause,
and as alternative anticoagulants (danaparoid, lepirudin, or argatroban) are
available in most countries, treatment with one of these agents before ob-
taining results of HIT antibody testing should be considered. For patients
with adenocarcinoma without HIT antibodies, longer-term management is
often more successful with LMWH or UFH than with warfarin (Prandoni,
1997; Lee et al., 2003).
A. Pseudo-HIT Complicated by HIT

Heparin-induced thrombocytopenia is a relatively common complication of
heparin therapy. It may be even more common in patients who have baseline
platelet activation and PF4 release, as occurs in adenocarcinoma-associated
DIC or diabetic ketoacidosis. Therefore, a patient with early thrombocyto-
penia attributable to a pseudo-HIT disorder may subsequently develop clin-
ically significant HIT antibodies (see Fig. 5). Another example is that of a
patient with lung cancer and DVT who developed a platelet count rise during
Pseudo–HIT 329
intravenous heparin therapy, followed by recurrent thrombocytopenia and,

ultimately, venous limb gangrene during anticoagulation with warfarin and
ancrod (Fig. 7). In this situation, one might have expected platelet count
recovery during a second course of heparin. However, an intravenous heparin
challenge resulted in worsening of thrombocytopenia, and the patient had a
strong positive assay for HIT antibodies. These patient cases emphasize the
Figure 7 Pseudo-HIT complicated by HIT: A 78-year-old man, with right proximal

lower limb deep venous thrombosis (DVT) and thrombocytopenia, developed prog-
ressive platelet count increase during therapeutic-dose UFH treatment. Recurrent
thrombocytopenia developed after UFH was stopped and when the patient was anti-
coagulated with warfarin. A liver biopsy on day 9 showed metastatic adenocarcinoma
(primary lung neoplasm), and adenocarcinoma-associated disseminated intravascular
coagulation (DIC) was diagnosed. However, a heparin challenge produced a further
platelet count fall; HIT antibody testing was strongly positive (serotonin-release assay
[SRA]: 88% serotonin release at 0.1 U/mL heparin; <15% release at 0 and 100 U/mL
heparin). Subsequently, the patient developed new left-sided DVT, as well as venous
gangrene of the left foot during treatment with warfarin and ancrod (peak INR = 3.8).
Although the clinical course was initially identical with pseudo-HIT (rising platelet
count on heparin therapy; abrupt platelet count fall after heparin administration was
stopped), the subsequent heparin-induced fall in the platelet count, and strong positive
HIT test results, indicate the patient also had HIT.
330 Warkentin
importance of a high clinical suspicion for HIT even in complex situations for
which other explanations for thrombocytopenia are present. The wider
availability of assays for HIT antibodies should help clinicians better diag-
nose and manage patients who develop thrombocytopenia and thrombosis
during or shortly following heparin treatment.
REFERENCES
Adamson DJA, Currie JM. Occult malignancy is associated with venous thrombosis
unresponsive to adequate anticoagulation. Br J Clin Pract 1993; 47:190–191.
Aird WC. The role of the endothelium in severe sepsis and the multiple organ
dysfunction syndrome. Blood 2003a; 101:3765–3777.
Aird WC. The hematologic system as a marker of organ dysfunction in sepsis. Mayo
Clin Proc 2003b; 78:869–881.
Arnout J. The pathogenesis of the antiphospholipid antibody syndrome: a hypothesis
based on parallelisms with heparin-induced thrombocytopenia. Thromb Hae-
most 1996; 75:536–541.
Arnout J. The role of h2-glycoprotein I-dependent lupus anticoagulants in the patho-
genesis of the antiphospholipid syndrome. Verh K Acad Geneeskd Belg 2000;
62:353–372.
Arnout J, Vermylen J. Current status and implications of autoimmune antiphospho-
lipid antibodies in relation to thrombotic disease. J Thromb Haemost 2003;
1:931–942.
Arvieux J, Roussel B, Pouzol P, Colomb MG. Platelet activating properties of murine
monoclonal antibodies to beta2-glycoprotein I. Thromb Haemost 1993; 70:
336–341.
Asherson RA, Khamashta MA, Ordi-Ros J, Derksen RH, Machin SJ, Barquinero J,
Outt HH, Harris EN, Vilardell-Torres M, Hughes GR. The ‘‘primary’’ anti-
phospholipid syndrome: major clinical and serological features. Medicine 1989;
68:366–374.
Auger WR, Permpikul P, Moser KM. Lupus anticoagulant, heparin use, and throm-
bocytopenia in patients with chronic thromboembolic pulmonary hypertension:
a preliminary report. Am J Med 1995; 99:392–396.
Balduini CL, Noris P, Bertolino G, Previtali M. Heparin modifies platelet count and
function in patients who have undergone thrombolytic therapy for acute
myocardial infarction [letter]. Thromb Haemost 1993; 69:522–532.
Bernard GR, Vincent JL, Laterre PF, LaRosa SP, Dhainaut JF, Lopez-Rodriguez A,
Steingrub JS, Garber GE, Helterbrand JD, Ely EW, Fisher JC Jr. Efficacy and
safety of recombinant human activated protein C for severe sepsis. N Engl J
Med 2001; 344:699–709.
Bone RC, Francis PB, Pierce AK. Intravascular coagulation associated with the adult
respiratory distress syndrome. Am J Med 1976; 61:585–589.
Pseudo–HIT 331
Bourhim M, Darnige L, Legallais C, Arvieux J, Cevallos R, Pouplard C, Vigaya-

lakshmi MA. Anti-h2-glycoprotein I antibodies recognizing platelet factor
4–heparin complex in antiphospholipid syndrome in patient substantiated
with mouse model. J Mol Recognit 2003; 16:125–130.
Campbell RR, Foster KJ, Stirling C, Mundy D, Reckless JPD. Paradoxical platelet
behaviour in diabetic ketoacidosis. Diabetic Med 1985; 3:161–164.
Chiu HM, van Aken WG, Hirsh J, Regoeczi E, Horner AA. Increased heparin clear-
ance in experimental pulmonary embolism. J Lab Clin Med 1977; 90:204–215.
De Gennes C, Souilhem J, Du LTH, Chapelon C, Raguin G, Wechsler B, Blétry O,
Godeau P. Embolie artérielle des membres au cours des endocardites infec-
tieuses sur valves natives. Presse Méd 1990; 19:1177–1181.
Delahaye JP, Poncet P, Malquarti V, Beaune J, Garé JP, Mann JM. Cerebrovascular
accidents in infective endocarditis: role of anticoagulation. Eur Heart J 1990;
11:1074–1078.
Everett RN, Jones FL Jr. Warfarin-induced skin necrosis. A cutaneous sign of
malignancy? Postgrad Med 1986; 79:97–103.
Fijnvandraat K, Derkx B, Peters M, Bijlmer R, Sturk A, Prins MH, van Deventer
SJH, ten Cate JW. Coagulation activation and tissue necrosis in meningococcal
septic shock: severely reduced protein C levels predict a high mortality. Thromb
Haemost 1995; 73:15–20.
Ford I, Urbaniak S, Greaves M. IgG from patients with antiphospholipid syndrome
binds to platelets without induction of platelet activation. Br J Haematol 1998;
102:841–849.
Galli M. Non beta2-glycoprotein I cofactors for antiphospholipid antibodies. Lupus
1996; 5:388–392.
Galli M, Daldossi M, Barbui T. Anti-glycoprotein Ib/IX and IIb/IIIa antibodies in
patients with antiphospholipid antibodies. Thromb Haemost 1994; 71:571–575.
Galli M, Finazzi G, Barbui T. Thrombocytopenia in the antiphospholipid syndrome.
Br J Haematol 1996; 93:1–5.
Galli M, Luciani D, Bertolini G, Barbui T. Lupus anticoagulants are stronger risk
factors for thrombosis than anticardiolipin antibodies in the antiphospholipid
syndrome: a systematic review of the literature. Blood 2003; 101:1827–1832.
Gibson GE, Su WP, Pittelkow MR. Antiphospholipid syndrome and the skin. J Am
Acad Dermatol 1997; 36(pt 1):970–982.
Ginsberg JS, Wells PS, Brill-Edwards P, Donovan D, Moffatt K, Johnston M, Stevens
P, Hirsh J. Antiphospholipid antibodies and venous thromboembolism. Blood
1995; 86:3685–3691.
Gruel Y. Antiphospholipid syndrome and heparin-induced thrombocytopenia:
update on similarities and differences. J Autoimmun 2000; 15:265–268.
Hayes MA, Yau EH, Hinds CJ, Watson JD. Symmetrical peripheral gangrene: asso-
ciation with noradrenaline administration. Intensive Care Med 1992; 18:433–
436.
Hojnik M, George J, Ziporen L, Shoenfeld Y. Heart valve involvement (Libman-Sacks
endocarditis) in the antiphospholipid syndrome. Circulation 1996; 93:1579–
1587.
332 Warkentin
Jankowski M, Vreys I, Wittevrongel C, Boon D, Vermylen J, Hoylaerts MF, Arnout J.

Thrombogenicity of h2-glycoprotein I dependent antiphospholipid antibodies
in a photochemically induced thrombosis model in the hamster. Blood 2003;
101:157–162.
Johansen K, Hansen ST Jr. Symmetrical peripheral gangrene (purpura fulminans)
complicating pneumococcal sepsis. Am J Surg 1993; 165:642–655.
Knight TT Jr, Gordon SV, Canady J, Rush DS, Browder W. Symmetrical peripheral
gangrene: a new presentation of an old disease. Am Surg 2000; 66:196–199.
Kullberg BJ, Westendorp RGJ, van’t Wout JW, Meinders AE. Purpura fulminans and
symmetrical peripheral gangrene caused by Capnocytophaga canimorsus (for-
merly DF-2) septicemia—a complication of dog bite. Medicine (Balt) 1991;
70: 287–292.
Lasne D, Saffroy R, Bachelot C, Vincenot A, Rendu F, Papo T, Aiach M, Piette J-C .
Tests for heparin-induced thrombocytopenia in primary antiphospholipid
syndrome [letter]. Br J Haematol 1997; 97:939.
Lebrazi J, Helft G, Abdelouahed M, Elalamy I, Mirshahi M, Samama MM, Lecompte
T. Human anti-streptokinase antibodies induce platelet aggregation in an Fc
receptor (CD32) dependent manner. Thromb Haemost 1995; 74:938–942.
Lee AY, Levine MN, Baker RI, Bowden C, Kakkar AK, Prins M, Rickle FR, Julian
JA, Haley S, Kovacs MJ, Gent M. Low-molecular-weight heparin versus a
coumarin for the prevention of recurrent venous thromboembolism in patients
with cancer. N Engl J Med 2003; 349:109–111.
Lipp E, von Felten A, Sax H, Müller D, Berchtold P. Antibodies against platelet
glycoproteins and antiphospholipid antibodies in autoimmune thrombocyto-
penia. Eur J Haematol 1998; 60:283–288.
Lubenow N, Eichler P, Albrecht D, Carlsson LE, Kothmann J, Rossocha W, Hahn M,
Quitmann H, Greinacher A. Very low platelet counts in post-transfusion pur-
pura falsely diagnosed as heparin-induced thrombocytopenia. Report of four
cases and review of literature. Thromb Res 2000; 100:115–125.
to initial use or reexposure to heparin. Chest 2002; 122:37–42.
Lutters BC, Meijers JC, Derksen RH, Arnout J, de Groot PG. Dimers of h2-glyco-
protein I mimic the in vitro effects of h2-glycoprotein I-anti-h2-glycoprotein I
antibody complexes. J Biol Chem 2001; 276:3060–3067.
Martinuzzo ME, Maclouf J, Carreras LO, Lévy-Toledano S. Antiphospholipid anti-
bodies enhance thrombin-induced platelet activation and thromboxane for-
mation. Thromb Haemost 1993; 70:667–671.
Martinuzzo ME, Forastiero RR, Adamczuk Y, Pombo G, Carreras LO. Antiplate-
let factor 4–heparin antibodies in patients with antiphospholipid antibodies.
Thromb Res 1999; 95:271–279.
Monreal M, Lafoz E, Casals A, Ruı́z J, Arias A. Platelet count and venous throm-
boembolism. A useful test for suspected pulmonary embolism. Chest 1991;
100:1493–1496.
Morgan M, Downs K, Chesterman CN, Biggs JC. Clinical analysis of 125 patients
with the lupus anticoagulant. Aust NZ J Med 1993; 23:151–156.
Pseudo–HIT 333
Mustafa MH, Mispireta LA, Pierce LE. Occult pulmonary embolism presenting with
thrombocytopenia and elevated fibrin split products. Am J Med 1989; 86:490–
491.
Owings JT, Bagley M, Gosselin R, Romac D, Disbrow E. Effect of critical injury on
plasma antithrombin activity: low antithrombin levels are associated with
thromboembolic complications. J Trauma 1996; 41:396–405.
Paton RC. Haemostatic changes in diabetic coma. Diabetologia 1981; 21:172–177.
Petri M. Pathogenesis and treatment of the antiphospholipid antibody syndrome. Adv
Rheumatol 1997; 81:151–177.
Phillips DE, Payne DK, Mills GM. Heparin-induced thrombotic thrombocytopenia.
Ann Pharmacother 1994; 28:43–46.
Prandoni P. Antithrombotic strategies in patients with cancer. Thromb Haemost 1997;
78:141–144.
Regnault V, Helft G, Wahl D, Czitrom D, Vuillemenot A, Papouin G, Roda L,
Danchin N, Lecompte T. Antistreptokinase platelet-activating antibodies are
common and heterogeneous. J Thromb Haemost 2003; 1:1055–1061.
Rinaldo JE, Perez H. Ischemic necrosis of both lower extremities as a result of the
microembolism syndrome complicating the adult respiratory distress syndrome
caused by Escherichia coli pneumonia and septicemia. Am Rev Respir Dis 1982;
126:932–935.
Rosse WF. Paroxysmal nocturnal hemoglobinuria as a molecular disease. Medicine
1997; 76:63–93.
Shi W, Chong BH, Chersterman CN. h2-Glycoprotein I is a requirement for anti-
cardiolipin antibodies binding to activated platelets: differences with lupus anti-
coagulants. Blood 1993; 81:1255–1262.
Smith OP, White B. Infectious purpura fulminans: diagnosis and treatment. Br J
Haematol 1999; 104:202–207.
Stahl RL, Javid JP, Lackner H. Unrecognized pulmonary embolism presenting as
disseminated intravascular coagulation. Am J Med 1984; 76:772–778.
Ting W, Silverman NA, Arzouman DA, Levitsky S. Splenic septic emboli in endo-
carditis. Circulation 1990; 82(suppl 5):IV105–IV109.
Triplett DA. Lupus anticoagulants/antiphospholipid-protein antibodies: the great
imposters. Lupus 1996; 5:431–435.
Vaughan DE, Kirshenbaum JM, Loscalzo J. Streptokinase-induced, antibody-me-
diated platelet aggregation: a potential cause of clot propagation in vivo. J Am
Coll Cardiol 1988; 11:1343–1348.
Warkentin TE. Heparin-induced thrombocytopenia: IgG-mediated platelet activation
platelet microparticle generation, and altered procoagulant/anticoagulant bal-
ance in the pathogenesis of thrombosis and venous limb gangrene complicating
heparin-induced thrombocytopenia. Transfusion Med Rev 1996; 10:249–258.
Warkentin TE. Venous thromboembolism in heparin-induced thrombocytopenia.
Curr Opin Pulm Med 2000; 6:343–351.
Warkentin TE. Venous limb gangrene during warfarin treatment of cancer-associated
deep venous thrombosis. Ann Intern Med 2001; 135:589–593.
334 Warkentin

N Engl J Med 2001; 344:1286–1992.
Heparin-induced thrombocytopenia in patients treated with low-molecular
topenia. Blood 2000; 96:1703–1708.
Warkentin TE, Roberts RS, Hirsh J, Kelton JG. An improved definiton of immune
Warkentin TE, Aird AC, Rand J. Platelet-endothelial interaction: sepsis, HIT, and
antiphospholipid syndrome. Hematology (Am Soc Hematol Educ Program)
2003b. In press.
Weitz JI, Hudoba M, Massel D, Maraganore J, Hirsh J. Clot-bound thrombin is
protected from inhibition by heparin-antithrombin III but is susceptible to
inactivation by antithrombin III-independent inhibitors. J Clin Invest 1990;
86:385–391.
Weitz JI, Leslie B, Hudoba M. Thrombin binds to soluble fibrin degradation pro-
ducts where it is protected from inhibition by heparin-antithrombin but
susceptible to inactivation by antithrombin-independent inhibitors. Circulation
1998; 97:544–552.
White B, Livingstone W, Murphy C, Hodgson A, Fafferty M, Smith OP. An open-
label study of the role of adjuvant hemostatic support with protein C replace-
ment therapy in purpura fulminans-associated meningococcemia. Blood 2000;
96:3719–3724.
Wiedmer T, Hall SE, Ortel TL, Kane WH, Rosse WF, Sims PJ. Complement-induced
vesiculation and exposure of membrane prothrombinase sites in platelets of
paroxysmal nocturnal hemoglobinuria. Blood 1993; 82:1192–1196.
Winkler MJ, Trunkey DD. Dopamine gangrene. Am J Surg 1981; 142:588–589.
Yan SB, Helterbrand JD, Hartman DL, Wright TJ, Bernard GR. Low levels of pro-
tein C are associated with poor outcome in severe sepsis. Chest 2001; 120:915–
922.
13
Treatment of Heparin-Induced
Thrombocytopenia: An Overview
Andreas Greinacher
I. INTRODUCTION
Heparin-induced thrombocytopenia (HIT) presents a unique situation: hep-

arin causes the very problems its use was intended to prevent, namely, such
complications as pulmonary embolism, stroke, and limb gangrene. Further-
more, several treatment paradoxes pose serious management pitfalls (Table
1). This chapter summarizes our treatment approach, with emphasis on prac-
tical management issues. We view HIT as a syndrome of increased thrombin
generation. Accordingly, we emphasize the use of rapidly acting anticoagu-
lant drugs that control thrombin generation in HIT.
This chapter is not the outcome of a formal consensus conference, as
defined elsewhere (McIntyre, 2001). Nevertheless, we have used an evidence-
based approach to frame our recommendations, based upon the Sixth
American College of Chest Physicians (ACCP) Consensus Conference on
Antithrombotic Therapy (Hirsh et al., 2001a,b; Guyatt et al., 2001). Accord-
ing to these guidelines, the recommendation to use (or not use) a particular
treatment is based on the trade-off between the expected benefits on the one
hand, and the risks on the other. Thus, based upon the evidence, as well as our
own experience, when we concluded that benefits of a particular treatment
outweighed risks, we recommended the treatment. If we were quite certain the
335
336 Greinacher and Warkentin
Table 1 Treatment Paradoxes of HIT Management
Paradoxical effect
Treatment for HIT of treatment Comments
Discontinue heparin High frequency of Use an alternative, rapidly acting

thrombosis despite anticoagulanta when heparin is
stopping heparin stopped because of suspected
HIT
Coumarin (e.g., warfa- High frequency of Control thrombin generation
rin, phenprocoumon) thrombosis; potential for with alternative anticoagulanta
warfarin-induced venous and await partial or full
limb gangrene syndrome resolution of HIT before
starting coumarin for
longer-term control of
thrombosis
Low molecular weight High frequency of Although LMWH is less likely
heparin (LMWH) thrombocytopenia than unfractionated heparin
or thrombosis when (UFH) to cause HIT, LMWH
given to patients is relatively likely to maintain
with acute HIT or worsen acute HIT caused
by UFH
Low-dose danaparoid High frequency of High (therapeutic)-dose
(antithrombotic thrombosis when danaparoid recommended
prophylaxis) low-dose danaparoidb for patients with isolated HIT
given to patients with (or HIT-thrombosis)
isolated HIT
Platelet transfusions May increase risk for Spontaneous bleeding is
platelet-mediated uncommon even in severe HIT;
thrombosis thus, prophylactic platelet
transfusions are relatively
contraindicated
Vena cava filters May increase risk for Vena cava filters should be
inferior vena cava avoided in acute HIT; if used,
thrombosis, DVT, or concomitant anticoagulation
pulmonary embolism should be given, if possible
a
Rapidly acting alternative parenteral anticoagulants, such as danaparoid, lepirudin, and argatroban, are
discussed in Table 2.
b
Low-dose danaparoid (750 U two or three times a day) is approved for prevention of thrombosis in
acute HIT in some jurisdictions.
Treatment of HIT 337
evidence favored the recommendation, a grade 1 recommendation was made.

If we were less certain of the trade-off between benefits and risks, a weaker
recommendation of grade 2 was made.
We also assessed the methodological quality of the studies supporting
the recommendations, also using the ACCP guidelines: grade A: randomized
trials without important limitations; grade B: randomized trials with impor-
tant limitations; grade C+: no randomized trials but observations from other
trials can be unequivocably extrapolated, or there is overwhelming evidence
from observation studies; and grade C: observation studies.
Regarding studies of HIT, there is only one small randomized trial
(Chong et al., 2001), and this study had methodological flaws such as non-
blinded assessment of treatment outcomes. Hence, we have no grade A and
only one grade B recommendation. Grade C recommendations are based
upon observation studies. Regarding HIT, this includes prospective cohort
treatment studies with historical controls (Greinacher et al., 1999a,b; 2000;
Lewis et al., 2001, 2003); case-control series (Warkentin et al., 1997); and large
case series (e.g., Magnani, 1993, 1997; Warkentin and Kelton, 1996; Wallis
et al., 1999). Thus, our recommendations are graded as follows, with the
implications of the recommendation shown:
Grade 1B and Grade 1C+: strong recommendations; can apply to
most patients in most circumstances.
Grade 1C: strong or intermediate-strength recommendation; may
change when stronger evidence is available.
Grade 2C: weak recommendation; other alternatives may be equally
reasonable.
As no studies have directly compared the three major treatment options for
HIT (danaparoid, lepirudin, argatroban), any recommendation for use of one
of these drugs does not imply any consistent advantage over one of the others.
However, there are important pharmacokinetic differences among these
drugs, which might very well favor one in the particular circumstances of
an individual patient situation (see also Chaps. 14–19).
A. Disclaimer
There are several challenging aspects to treating patients with HIT. In par-
ticular, these patients are not clinically homogeneous: they represent a com-
plex mix of varying initial indication for heparin, location and severity of
HIT-associated thrombosis, and, not infrequently, dysfunction of one or more
vital organs. This presents difficulties both for performing clinical studies
as well as in the application of treatment recommendations for individual
patients. Furthermore, there are important differences among countries in

the approval or availability status of certain recommended treatment ap-
proaches. The treatment recommendations we make cannot thus be indiscrim-
inately applied to all patients with suspected HIT.
A further practical problem is that the major treatment options for HIT
include relatively new and, for some physicians, unfamiliar or even unap-
proved anticoagulant agents. (Indeed, the approval of lepirudin by the
European Union in May 1997 as a treatment for a thrombosis that compli-
cates HIT represented the first time an anticoagulant drug obtained by
recombinant technology became available for clinical use.) This presents
extra challenges to physicians and also to laboratories asked to monitor
anticoagulant treatment effects, as the treatment ‘‘learning curve’’ may occur
in emergency situations. Also, immediate results of reliable laboratory tests
for HIT are usually unavailable. Difficult management decisions may be
needed amid diagnostic uncertainty: a diagnosis of HIT that seems obvious in
retrospect may not have been so clear during its early evolution.
As an iatrogenic illness that occurs unexpectedly, often in a setting of
antithrombotic prophylaxis, medicolegal aspects must be considered (see
Chaps. 21 and 22). Thus, once HIT is entertained as part of a differential
diagnosis, we suggest that physicians carefully document the various diag-
nostic and treatment considerations as events unfold.
As one of those common, rare diseases [We acknowledge Prof. R. Hull
(Calgary, Canada) for his description of HIT as a ‘‘common, rare disease.’’]
that physicians only occasionally manage, and only rarely enter into clinical
studies, we need to acknowledge that no final answer for treatment is likely to
emerge. Therefore, even in this third edition, this chapter should be viewed as
a basis for further discussion and study of the treatment of HIT patients.
II. NONIMMUNE HEPARIN-ASSOCIATED

THROMBOCYTOPENIA
In some patients, especially those with comorbid conditions associated with

platelet activation (burns, anorexia nervosa), heparin treatment can result in a
transient decrease in platelet count (Burgess and Chong., 1997; Reininger et
al., 1996) (see Chap. 5). Unfractionated heparin (UFH) activates platelets
directly (Salzman et al., 1980), an effect observed less frequently with low
molecular weight heparin (LMWH) (Brace and Fareed, 1990). Known as
nonimmune heparin-associated thrombocytopenia (nonimmune HAT), this
direct proaggregatory effect of heparin occurs predominantly in patients
receiving high-dose, intravenous UFH therapy. Typically, platelet counts de-

crease within the first 1–2 days of treatment, and then recover over the next 3–
4 days. There are no data indicating that these patients are at increased risk for
adverse outcomes, including thrombosis. Indeed, it is possible that inap-
propriate discontinuation of heparin for nonimmune HAT could increase the
risk for thrombosis owing to the underlying clinical condition for which the
heparin is being given.
Recommendation. Heparin should not be discontinued in patients
clinically suspected of having nonimmune heparin-associated thrombo-
cytopenia (grade 2C).
III. THERAPY OF (IMMUNE) HIT

A. Pathogenesis of HIT: Treatment Implications
Heparin-induced thrombocytopenia is caused by antibodies that usually
recognize multimolecular complexes of platelet factor 4 (PF4) and heparin.
HIT can be viewed as a syndrome of in vivo thrombin generation that results
from the activation of platelets, endothelium, monocytes, and coagulation
pathways (Fig. 1; see color insert) (see Chaps. 5–10) (Warkentin and Kelton,
1994; Greinacher, 1995; Warkentin, 1997, 2003; Warkentin et al., 1998).
Given this model of pathogenesis, therapy for acute HIT should focus
on the following issues: (1) interruption of the immune response (e.g.,
discontinuation of heparin); (2) rapid reduction of increased thrombin
generation; and (3) treatment of HIT-associated thrombosis. In most patients
with HIT, effective pharmacological therapy for thrombosis will involve an
agent that rapidly controls thrombin generation, although in some situations
additional adjunctive treatments may be necessary (e.g., surgical thromboem-
bolectomy).
A newly recognized treatment issue involves patients with detectable
anti-PF4–heparin antibodies, but no platelet count reduction or other clinical
evidence of HIT. With increased testing for HIT antibodies, it is now clear
that many patients develop HIT antibodies without developing clinical HIT
(see Chaps. 4 and 11). In these patients it seems acceptable to continue heparin
treatment, but to monitor the platelet counts carefully (‘‘watch-and-wait’’
strategy).
Two recent retrospective studies found that patients with acute coro-
nary syndrome who either had HIT antibodies at presentation (Williams et
al., 2003) or developed HIT antibodies following heparin treatment (Mattioli
et al., 2000) had a higher frequency of vascular events during follow-up that
Figure 1 Pathogenesis of HIT; a central role for thrombin generation: HIT-IgG
antibodies bind to several identical epitopes on the same antigen complex, thus form-
ing immune complexes that become localized to the platelet surface. The IgG immune
complexes can cross-link the platelet FcgRIIa receptors, resulting in FcgRIIa-depen-
dent platelet activation (Kelton et al., 1988). The GP IIb/IIIa complex is not required
for platelet activation (Greinacher et al., 1994a). The activated platelets trigger a
cascade of events that ultimately lead to activation of the coagulation pathways,
resulting in thrombin generation. Activated platelets release their a-granule proteins
(Chong et al., 1994), including PF4, leading to formation of more multimolecular
PF4–heparin complexes, setting up a vicious cycle of platelet activation, triggering
even more platelet activation (Greinacher, 1995). The activated platelets bind fibrin-
ogen, recruit other platelets, and begin to form a primary clot. During shape change,
procoagulant, platelet-derived microparticles are released, providing a phospholipid
surface for amplifying thrombin generation (Warkentin et al., 1994). The released PF4
also binds to endothelial cell heparan sulfate, forming local antigen complexes to
which HIT antibodies bind (Cines et al., 1987; Visentin et al., 1994; Greinacher et al.,
1994b). Tissue factor expression on activated endothelial cells and monocytes (Are-
pally and Mayer, 2001; Pouplard et al., 2001) further enhances thrombin generation.
(See color insert.)
ranged from 1 month to 1 year, despite the absence of apparent HIT in any
patient. The implications of ‘‘subclinical’’ HIT antibody seroconversion on
influencing subsequent cardiovascular morbidity and mortality awaits pro-
spective study.
B. Discontinuation of Heparin for Clinically Suspected HIT

Numerous case reports describe the occurrence of new, progressive, or
recurrent thromboembolic events during continued or repeated use of heparin
in patients with acute HIT. Moreover, the thrombocytopenia usually persists
if the administration of heparin is not stopped. Thus, all heparin treatment
should be discontinued in patients strongly suspected of having HIT, and
usually substituted by another anticoagulant (discussed subsequently), while
awaiting results of HIT antibody testing.
Recommendation. All heparin administration should be discontinued in
patients clinically suspected of having (immune) HIT (grade 1C+).
The routine use of heparin (e.g., line flushing) is pervasive in hospitals.

Thus, based on our experience it can be helpful to institute methods to reduce
the risk for inadvertent heparin use in hospitalized patients with HIT.
Recommendation. A clearly visible note should be placed above the
patient’s bed stating ‘‘NO HEPARIN: HIT’’ (grade 2C).
Not infrequently, patients in whom heparin administration has been
stopped because of clinically suspected HIT subsequently are found to have
negative laboratory tests for HIT antibodies. In our experience, it is reason-
able and safe to restart heparin therapy in these patients, provided the
intervening clinical events are consistent with an alternative explanation for
thrombocytopenia (see Chaps. 2, 3, and 12), and provided the laboratory has
adequately excluded the presence of HIT antibodies (see Chap. 11).
Recommendation. Heparin can be safely restarted in patients proved not
to have HIT antibodies by sensitive activation or antigen assay (grade
2C).
C. Anticoagulation of the HIT Patient with Thrombosis

The Need for Anticoagulation of HIT-Associated Thrombosis
Heparin-induced thrombocytopenia is a strong, independent risk factor for
venous and arterial thrombosis (Warkentin et al., 1995, 2003; Girolami et al.,
2003). HIT can be complicated by thrombosis in several ways: (1) a preceding
thrombosis, leading to the heparin treatment that caused HIT; (2) new,
progressive, or recurrent thrombosis resulting from HIT itself; or (3) for both
reasons.
For a HIT patient with thrombosis in whom heparin administration has
been discontinued, there is, nevertheless, a very high risk for subsequent
thrombosis. This was shown in two historically controlled prospective treat-
ment cohort studies (Greinacher et al., 1999a,b), in which the incidence of
thrombotic events ranged from 5 to 10% per patient day (see Chap. 15). This
high event rate (6.1% per day in the meta-analysis) occurred after stopping
heparin therapy and after laboratory confirmation of HIT, but before insti-
tution of alternative anticoagulation with lepirudin (mean period of treat-
ment delay; 1.7 days) (Greinacher et al., 2000). This experience suggests that
alternative anticoagulant therapy should not be delayed for results of HIT
antibody testing in patients strongly suspected of having HIT.
Anticoagulants Evaluated for Treatment of HIT

Current treatment of HIT focuses on agents that rapidly control thrombin
generation (Warkentin et al., 1998; Hirsh et al., 1998, 2001b) (Fig. 2). Table 2
lists the available evidence on efficacy for three such agents: danaparoid,
lepirudin, and argatroban (listed in the order the drugs became available).
Only one randomized controlled trial for the management of HIT has
been performed: this study compared danaparoid with dextran for the
treatment of HIT-associated thrombosis (Chong et al., 2001; Ortel and
Chong, 1998) (see Chap. 14). As the study was small and open-label, we have
listed it as a level 1B, rather than level 1A, recommendation.
Recommendation. Therapeutic-dose anticoagulation with a rapidly acting
anticoagulant, e.g., danaparoid (grade 1B), lepirudin (grade 1C+), or
argatroban (grade 1C), should be given to a patient with thrombosis
complicating acute HIT. Treatment should not be delayed pending la-
boratory confirmation in a patient strongly suspected of having HIT.
Pharmacological and Pharmacokinetic Considerations:

Danaparoid, Lepirudin, and Argatroban
The lack of prospective comparative studies between danaparoid, lepirudin,
and argatroban precludes definitive conclusions about relative efficacy and
safety. However, there are several pharmacological and pharmacokinetic
differences that physicians should consider when determining which drug may
be preferred in an individual patient (Tables 3–5). For example, in a patient
with vital organ or limb ischemia or infarction who might need urgent surgical
intervention, an agent with a short half-life may be preferred. On the other
hand, in a patient with venous thromboembolism in whom an uncomplicated
Figure 2 Thrombin generation and fibrin formation in acute HIT. (A) Thrombin
generation, as assessed by thrombin–antithrombin (TAT) complexes, is markedly in-
creased in acute HIT (mean, 55 ng/mL; normal,<4.1 ng/mL). Whereas danaparoid
reduces thrombin generation in these patients, the defibrinogenating snake venom,
ancrod, does not. (B) Levels of cross-linked fibrin degradation products (D-dimer) are
increased in patients with acute HIT (mean, f4–5 Ag/mL; normal, <0.5 Ag/mL).
Whereas danaparoid reduces D-dimer levels, ancrod increases their levels. Baseline (B)
samples were obtained at diagnosis of HIT and before treatment with danaparoid or
ancrod; subsequent values are shown for day 1 (D1, 1–24 h postinitiation of treat-
ment), day 2 (25–48 h), and day 3 (49–72 h). *p < 0.001, **p < 0.002. (From War-
kentin, 1998.)
overlap with (longer-term) warfarin anticoagulation is anticipated, or who

requires outpatient treatment by subcutaneous injections, danaparoid may
have certain advantages. Argatroban (which undergoes hepatobiliary clear-
ance) is suited for patients with renal insufficiency, as dose reduction is
generally not required (cf. lepirudin). Other factors to consider include drug
availability to, and prior experience of, the physician; availability and
turnaround time of laboratory monitoring; and so on.
Other Drugs that Reduce Thrombin Generation in HIT

Other drugs with antithrombin activity described anecdotally as treatment
for HIT include bivalirudin (Nand, 1993; Reid and Alving, 1994; Chamberlin
et al., 1995; Campbell et al., 2000; Francis et al., 2003) and the glycosamino-
glycan agent dermatan sulfate (Agnelli et al., 1994). Theoretically, the
heparin-binding pentasaccharide fondaparinux should not cause HIT or
cross-react with HIT antibodies (Elalamy et al., 1995; Greinacher et al.,
Table 2 Rapidly Acting Anticoagulants for the Treatment of HIT: Main Studies Supporting Efficacy
344
Drug Mechanism of action Studies
Danaparoid sodium Mixture of glycosaminoglycans . Randomized controlled trial comparing danaparoid

(Orgaran, formerly with anti-Xa activity H plus warfarin versus dextran 70 plus warfarin
known as Org 10172 anti-IIa (antithrombin) activity (Chong et al., 2001).
and Lomoparan . Compassionate release program (Magnani, 1993,
1997).
. Retrospective comparison of danaparoid and
ancrod (Warkentin, 1996).
Recombinant hirudin Recombinant protein derived . HAT-1: prospective cohort study with historical
(lepirudin, Refludan) from leeches that directly controls (Greinacher et al., 1999a).
inactivates thrombin . HAT-2: prospective cohort study with historical
controls (Greinacher et al., 1999b).
. Meta-analysis of HAT-1 and HAT-2 studies
. HAT-3: prospective cohort study with historical
controls (Eichler et al., 2002).
. Meta-analysis of HAT-1, HAT-2, and HAT-3
studies (Lubenow et al., 2002a).
. Retrospective comparison of lepirudin (prospective
cohort) and danaparoid (contemporaneous controls)
(Farner et al., 2001).
. Postmarketing surveillance (Lubenow et al., 2002b).
Argatroban (Novastana) Small molecule, direct . Arg-911: prospective cohort study with historical
thrombin inhibitor controls (Lewis et al., 2001).
. Arg-915: prospective cohort study with historical
controls (Lewis et al., 2003).
a
Argatroban is marketed in the United States by its generic name (with a capital A). In some other countries, it is marketed by its trademark. Novastan
Greinacher and Warkentin
(see Chap. 16).

Table 3 Main Characteristics of Danaparoid Sodium
Mechanism of action,
pharmacokinetics Monitoring Undesirable effects Comments
Treatment of HIT
Catalyzes the inactivation Anti-Xa levels during Cross-reactivity (XR) with Anticoagulant effect
of factor Xa by AT, treatment by an HIT antibodies: in vitro depends on adequate
and of thrombin (IIa) amidolytic assay XR usually not associated AT levels
by AT and HCII using danaparoid with adverse effects; patients Does not significantly
Bioavailability after sc reference curve should be monitored for prolong the aPTT, ACT,
injection f100%; peak Monitoring recommended in in vivo XR (unexplained PT/INR (does not
anti-Xa levels, 4–5 h patients with: (1) significant platelet count fall, progressive interfere with monitoring
after injection renal impairment; (2) body new TECs); in vivo XR is of overlapping oral
(Danhof et al., 1992) weight <45 kg or >110 kg; estimated to occur in f3% anticoagulants)
Mean plasma distribution (3) life- or limb-threatening of patients (Magnani, 1993) Reduce dosage if serum
time following iv bolus, thrombosis; (4) unexpected Bleeding complications in creatinine >265 Amol/L
f2.3 h bleeding; (5) critically ill or compassionate-release No antidote: In case of
Plasma t1/2 of anti-Xa unstable patient study (Ortel and Chong, overdosage, stop the
activity, 17–28 h 1998): fatal (0.9%), major drug and treat bleeding
(mean, 25 h); t1/2 of nonfatal bleeding (6.5%); with blood products as
anti-IIa activity, 2–4 h no major bleeds in RCT indicated
(Danhof et al., 1992) (Chong, 1996)
Skin hypersensitivity: rare
Abbreviations: ACT, activated clotting time; aPTT, activated partial thromboplastin time; AT, antithrombin; HCII, heparin cofactor II; iv,
intravenous; PT/INR, prothrombin time/international normalized ratio; RCT, randomized controlled trial; sc, subcutaneous; t1/2, drug half-life.
345
346
Table 4 Main Characteristics of the r-Hirudin, Lepirudin

Direct, noncovalent, aPTT during treatment; a Development of antihirudin f40% of patients develop antihirudin
irreversible inhibitor more precise monitoring antibodies in f40% of antibodies on day 5 or later of
of free and clot-bound is possible by the ECT patients. In about 3% of treatment; in only f5% of these
thrombin (see Chaps. 15, 17, 19) patients, these antibodies patients is a dose reduction or
Bioavailability after Daily aPTT monitoring enhance the anticoagulant increase needed; risk of anaphylactic
sc injection, f100%; is recommended in all effect of hirudin, and reactions post–iv bolus
peak effect, 2–3 h patients (see Comments require a substantial dose No major effect on PT/INR
Mean plasma distribution re: antihirudin antibodies) reduction. (Greinacher et al., 2000)
time after iv bolus, f2 h Monitoring by ECT Anaphylactic reactions: Reduce dosage if serum creatinine
Mean plasma t1/2, 1.3 h; recommended: f0.015% (first exposure) >120 Amol/L (see Table 3 in
t1/2 greatly prolonged 1. During cardiopulmo- f0.15% (reexposure) Chap. 15)
in renal failure (f200 h nary bypass surgery associated with iv bolus No antidote: In case of overdosage,
in nephrectomized 2. Unexpected bleeding injection (Greinacher stop the drug and treat bleeding
patients) Monitoring by quantitative et al., 2003) with blood products as indicated
hirudin EIA recommended: Allergic reactions: very rare (hemofiltration with a high-flux
1. Decreased prothrombin Skin hypersentivity: very rare membrane is a possible treatment
levels (Lindhoff-Last Bleeding complications in for life-threatening bleeding)
et al., 2000) HIT patients in prospective
studies: major bleeding in
two prospective studies, 13.4,
17% (see Chap. 15)
Abbreviations: aPTT, activated partial thromboplastin time; ECT, ecarin-clotting time; EIA, enzyme immunoassay; sc, subcutaneous; iv, intravenous;
t1/2 drug half-life.
Greinacher and Warkentin
Table 5 Main Characteristics of the Direct Thrombin Inhibitor, Argatroban
Treatment of HIT
Direct, noncovalent, reversible aPTT during treatment; no data No reported side effects besides Only iv use of argatroban has been
inhibitor of free and clot- exist as to whether more pre- bleeding complications (num- tested in HIT
bound thrombin cise monitoring at higher doses ber of patients treated is Reduce dosage by 75% in case of
f50% of the drug is would be achieved using too low to rule out possibi- liver impairment
plasma protein bound; other methods, such as ECT lity of rare side effects) No dose reduction in renal failure
Steady state is reached Target INR is >4.0 when No antidote: in case of overdosage
1–3 h after starting warfarin is overlapped with or severe bleeding, stop the drug
iv infusion argatroban (however, following and treat bleeding with blood
Mean plasma t1/2 is discontinuation of argatroban, products as indicated
40–50 min; t1/2 is the usual target INR of Argatroban prolongs the INR and
prolonged 4- to 2.0–3.0 applies during further requires a strategy adopted to
5-fold in moderate warfarin treatment) the INR reagent used for
liver impairment overlapping treatment with
warfarin (see Chap. 16)
Abbreviations: aPTT, activated partial thromboplastin time; ECT, ecarin-clotting time; INR, international normalized ratio; iv, intravenous; t1/2, drug
half-life.
347
1995; Amiral et al., 1997; Turpie et al., 2001), but there is no reported
experience with this agent for HIT.
D. Anticoagulation of the HIT Patient Without Thrombosis

Approximately 50% of patients with HIT do not have a new HIT-associated
thrombosis at the time HIT is first clinically suspected on the basis of
thrombocytopenia alone (Warkentin and Kelton, 1996; Greinacher et al.,
1999a,b). In a retrospective cohort study of 62 such patients with ‘‘isolated
thrombocytopenia,’’ the subsequent 30-day cumulative thrombotic event rate
was high (52.8%) (see Fig. 2 in Chap. 4). The rate of thrombosis was similar in
the two largest patient subgroups: patients treated with discontinuation of
heparin therapy alone (20/36, 56%) and patients treated with substitution of
warfarin for heparin (10/21, 48%). Overall, 6 of the 62 patients developed
pulmonary embolism (2 fatal); another patient who died suddenly may also
have had a fatal pulmonary embolism.
In a subsequent large retrospective cohort study of serologically con-
firmed HIT performed by Wallis and coworkers (1999), a 38% thrombotic
event rate was observed in patients with isolated HIT managed by cessation of
heparin. Further, early cessation of heparin was not associated with a reduc-
tion in the rate of thrombosis. Thus, the high thrombotic event rates observed
in these retrospective cohort studies are consistent with the experience of the
prospective treatment cohort studies that also observed a high rate of throm-
bosis soon after the diagnosis of HIT (Greinacher et al., 1999a,b, 2000).
The very high initial thrombotic event rates (5–10%/day over the first 1–
2 days) observed in these prospective and retrospective studies suggest that
many patients may have had subclinical DVT at the time that HIT was first
suspected. The data support the recommendation that antithrombotic treat-
ment for HIT be started before serological confirmation is received, even in
patients without clinical evidence of thrombosis.
In a retrospective analysis of patients with isolated HIT comparing
treatment with danaparoid and lepirudin, it was observed that patients who
received prophylactic-dose danaparoid (750 U sc b.i.d. or t.i.d.) had a trend to
a higher rate of thrombosis than patients treated with lepirudin (0.1 mg/kg
b.w./h, aPTT-adjusted) (Farner et al., 2001). In contrast, patients with HIT-
associated thrombosis had similar outcomes when treated with therapeutic
doses of either drug. This indicates that therapeutic, rather than prophylactic,
doses of danaparoid may be more effective for patients with isolated HIT
(Farner et al., 2001; Warkentin, 2001).
We usually prescribe an alternative anticoagulant in therapeutic doses
in this situation. Prophylactic-dose anticoagulation is a reasonable option in
HIT patients judged to be at higher risk for bleeding complications, as is
regular screening for venous thrombosis without anticoagulation in a patient

at very high bleeding risk. Thrombocytopenia itself should not be considered
a contraindication to anticoagulation in patients with HIT, as petechiae and
other spontaneous hemorrhagic manifestations are not usually seen in these
patients (see Chap. 3). However, if the platelet count is less than 20 109/L
and bleeding signs, but not thrombosis, are observed, then alternative
diagnoses such as posttransfusion purpura or other drug-dependent immune
thrombocytopenic disorders should be considered (see Chap. 2).
Recommendation. Alternative anticoagulation with an appropriate anti-
coagulant, such as danaparoid, lepirudin, or argatroban, should be con-
sidered in patients with clinically suspected HIT even in the absence of
symptomatic thrombosis. Anticoagulation should be continued at least
until recovery of the platelet counts to a stable plateau. Patients should
undergo imaging studies for lower limb DVT, especially those at highest
risk for venous thromboembolism, such as postoperative patients (grade
1C+).
E. Longer-Term Anticoagulant Management of the HIT

Patient with Thrombosis
Acute HIT by itself is not an indication for longer-term anticoagulation (i.e.,
3–6 months). However, HIT-associated thrombosis, or the underlying disease
itself, often is. For long-term control of thrombosis, oral anticoagulants of the
coumarin class (e.g., warfarin or phenprocoumon) are the treatment of
choice. Generally, it takes at least 5 days of oral anticoagulant therapy before
therapeutic functional hypoprothrombinemia is achieved (Harrison et al.,
1997). It is important that thrombin generation be controlled in patients with
acute HIT before initiation of coumarin treatment, particularly in patients
with severe HIT-associated DVT, because coumarin-induced venous limb
gangrene is a potential outcome (Warkentin et al., 1997) (see Chap. 3). It is
our practice to postpone starting administration of coumarin anticoagulants
until therapeutic anticoagulation is achieved with danaparoid, lepirudin, or
argatroban and substantial platelet count recovery has occurred (at least
z100 109/L).
Recommendation. The drug of choice for longer-term anticoagulation of
HIT patients is an oral anticoagulant of the coumarin class (e.g., war-
farin or phenprocoumon). However, in a patient with acute HIT, oral
anticoagulant therapy should be delayed until the patient is adequately
anticoagulated with a rapidly acting parenteral anticoagulant, and ide-
ally not until there has been substantial platelet count recovery (at least
z100 109/L). Oral anticoagulants should be started in low main-
tenance doses (e.g., V5 mg warfarin), with at least 5 days of overlap with
the parenteral anticoagulant (including at least 2 days in the target-

therapeutic range). If applicable, oral or intravenous vitamin K should
be given to reverse coumarin anticoagulation in a patient recognized as
having acute HIT after coumarin has been commenced (grade 1C).
Besides avoiding the risk of coumarin-induced necrosis, there are other
reasons for postponing coumarin anticoagulation in a patient with acute HIT.
For example, coumarin will increase a patient’s activated partial thrombo-
plastin time (aPTT). Since the aPTT is used to monitor the anticoagulant
effect of the direct thrombin inhibitors, the patient is at risk of receiving
insufficient dosing of the direct thrombin inhibitor if coumarin has been given.
For unknown reasons, the direct thrombin inhibitors have varying
effects upon the PT/INR (prothrombin time/international normalized ratio),
which are both agent-specific and patient-specific. For example, argatroban
has been reported to increase the INR more than lepirudin and bivalirudin
(argatroban > bivalirudin > lepirudin) (Gosselin et al., 2003). This compli-
cates the issue of coumarin overlap somewhat more with argatroban than the
other direct thrombin inhibitors (see Chap. 16). Another reason, therefore, to
postpone coumarin therapy in patients receiving concomitant therapy with
direct thrombin inhibitors is that it avoids the potential for a physician to
conclude incorrectly that the patient is adequately anticoagulated with
coumarin (because of the greater increase in the INR during cotherapy),
and thus to discontinue the direct thrombin inhibitor prematurely. Once the
anticoagulant effect of the direct thrombin inhibitor has dissipated within a
few hours, the circumstances favoring microvascular thrombosis (and, hence,
coumarin-induced necrosis) might well be present, i.e., ongoing thrombin
generation from acute HIT, warfarin-induced protein C depletion, and active
deep venous thrombosis (Smythe et al., 2002; Srinivasan et al., 2003).
In case of coumarin overdose and severe bleeding during the first 3
months after an episode of HIT, prothrombin complex concentrates should
only be used with extreme caution to ‘‘reverse’’ coumarin anticoagulation.
This is because these concentrates contain heparin and have caused recurrent
thrombocytopenia and thrombosis in patients with circulating HIT anti-
bodies (Greinacher et al., 1992).
Recommendation. Prothrombin complex concentrates should not be
used to reverse coumarin anticoagulation in a patient with acute or
recent HIT unless bleeding is otherwise unmanageable (grade 2C).
F. Reexposure of the HIT Patient to Heparin

Heparin Reexposure of the Patient with Acute or Recent HIT
Deliberate or accidental readministration of heparin to a patient with acute
or recent HIT can cause an abrupt platelet count fall, sometimes complicated
by thrombosis or acute systemic reactions (see Chap. 3). Accordingly,

deliberate heparin rechallenge for diagnostic purposes is not recommended,
especially because sensitive assays for HIT antibodies are available. This is a
strong recommendation because the diagnostic usefulness of laboratory
assays for HIT has been established in controlled studies (see Chap. 11).
Recommendation. Deliberate reexposure to heparin of a patient with
acute or recent HIT for diagnostic purposes is not recommended.
Rather, the diagnosis should be confirmed by testing acute patient se-
rum or plasma for HIT antibodies using a sensitive activation or antigen
assay (grade 1C+).
Heparin Reexposure of the Patient with a History of Remote HIT

The HIT antibodies are usually not detectable 3 months after an episode of
HIT (Warkentin and Kelton, 2001). There are few data describing the clinical
and serological outcomes of patients with previously documented HIT in the
remote past (arbitrarily, >3 months ago, or sooner, if HIT antibodies have
disappeared). One patient who developed fatal HIT on day 15 of UFH treat-
ment had a history of HIT complicated by thrombosis 6 years earlier (Gruel
et al., 1990). However, several patients with previous remote HIT have been
observed in whom repeat heparin use caused neither HIT nor HIT antibody
formation (Pötzsch et al., 2000; Warkentin and Kelton, 2001).
Because there are acceptable alternative anticoagulant options for most
prophylactic and therapeutic indications, both UFH and LMWH should be
avoided in patients with a previous history of HIT. As discussed in the fol-
lowing section, however, there are special circumstances, such as cardiac or
vascular surgery, during which it is reasonable to use heparin for a patient
with a previous history of HIT, provided certain precautions are taken.
Recommendation. Heparin should not be used for antithrombotic
prophylaxis or therapy in a patient with a previous history of HIT,
except under special circumstances (e.g., cardiac or vascular surgery)
(grade 2C).
IV. HIT IN SPECIAL CLINICAL SITUATIONS

A. Cardiopulmonary Bypass or Vascular Surgery
Management of the Patient with Acute or Recent HIT
For patients with acute HIT who require heart surgery, or with recent HIT
and persistence of circulating HIT antibodies, it is possible to use alternative
anticoagulants during cardiopulmonary bypass (CPB) (for review, see War-
kentin and Greinacher, 2003). Three options for such patients are bivalirudin,
lepirudin, and danaparoid (listed in order of preference). Unfortunately, the
lack of a specific antidote, the need for special intraoperative monitoring, and
other considerations mean that none is ideal for managing CPB. Another
approach is to administer heparin together with a potent antiplatelet agent,
e.g., tirofiban (GPIIb/IIIa antagonist) or epoprostenol (prostacyclin ana-
logue). This special topic of managing cardiac surgery patients with acute or
previous HIT is discussed in detail in Chap. 19, as well as in relation to specific
anticoagulant agents in Chaps. 14–17.
Danaparoid and lepirudin have also been used to provide intraoperative
anticoagulation, as well as to ‘‘flush’’ blood vessels during vascular surgery in
patients with acute HIT.
Recommendation. Alternative anticoagulation should be used for heart
or vascular surgery in a patient with acute or recent HIT with detectable
HIT antibodies. Either bivalirudin, lepirudin, or danaparoid are ap-
propriate alternatives for intraoperative anticoagulation, provided that
appropriate, rapid-turnaround laboratory monitoring and blood prod-
uct support to manage potentially severe bleeding complications are
available. Another approach is to give heparin together with a potent
antiplatelet agent (grade 2C).
Management of the Patient Following Disappearance of HIT

Antibodies
The drawbacks of alternative anticoagulants for CPB provide a rationale for
the use of heparin in two groups of patients with a previous history of HIT: (1)
a patient with a history of HIT, but who no longer has circulating HIT anti-
bodies detected by sensitive laboratory assay; and (2) a patient with acute or
recent HIT who requires elective heart surgery. In the latter situation, it is
reasonable to delay cardiac surgery until HIT antibodies become undetect-
able, which usually occurs in a few weeks or months (Warkentin and Kelton,
2001; Warkentin and Greinacher, 2003).
It is feasible to give UFH for cardiac or vascular surgery in a patient
with a previous history of HIT, provided that HIT antibodies are not detect-
able at the time of surgery (Olinger et al., 1984; Smith et al., 1985; Makhoul
et al., 1987; Pötzsch et al., 2000; Warkentin and Kelton, 2001; Warkentin and
Greinacher, 2003). We recommend that heparin be avoided completely both
before surgery (to prevent restimulation of HIT antibodies) and after surgery
(thus making HIT unlikely even if HIT antibodies are reformed). Current
evidence suggests that there is a minimum time to formation of clinically
significant HIT antibodies of 5 days even in patients who have a previous
history of HIT (Cadroy et al., 1994; Warkentin and Kelton, 2001; Lubenow
et al., 2002c). The patient should receive routine doses of UFH for the surgical
procedure itself. Preoperative anticoagulation (e.g., for heart catheterization)
and postoperative antithrombotic prophylaxis can be achieved with a non-
heparin agent such as danaparoid (750 U b.i.d.-t.i.d.) or r-hirudin (15 mg
b.i.d. s.c.) (Eriksson et al., 1997) (see Chap. 14).
Recommendation. In a patient with a previous history of HIT, heart or
vascular surgery can be performed using heparin, provided that HIT
antibodies are absent (by sensitive assay) and heparin use is restricted to
the surgical procedure itself (grade 1C).
B. HIT During Pregnancy

There are a few reports describing HIT during pregnancy (Meytes et al., 1986;
Henny et al., 1986; Copplestone and Oscier, 1987; Calhoun and Hesser 1987;
van Besien et al., 1991; Greinacher et al., 1993a). Danaparoid has been used in
at least 31 pregnant women using dosing schedules similar to those in
nonpregnant patients. Danaparoid does not cross the placenta, based on
cord blood assessment (see Chap. 14).
Lepirudin, bivalirudin, argatroban, danaparoid, and fondaparinux are
category B drugs, i.e., indicating absence of fetal damage in certain high-
dose animal studies, but limited (if any) human data. Only one report de-
scribes the use of lepirudin during pregnancy (Huhle et al., 2000). Danaparoid
and fondaparinux (Lagrange et al., 2002) do not appear to cross the pla-
centa, whereas hirudin can cross the placenta in low doses (Markwardt et al.,
1988) and has caused embryopathy in rabbits given high doses of hirudin
(Lubenow and Greinacher 2000). Further, a zebrafish model reveals that
thrombin plays a role in embryogenesis (Jagadeeswaran et al., 1997). Thus,
danaparoid and fondaparinux may be preferable for treatment of HIT during
(early) pregnancy.
Recommendation. If available, danaparoid (and possibly fondaparinux)
is preferred for parenteral anticoagulation of pregnant patients with
HIT, or in those who have a previous history of HIT (grade 2C).
C. Treatment of HIT in Children

There are only a few reports describing the management of HIT in children
(Oriot et al., 1990; Potter et al., 1992; Murdoch et al., 1993; Klement et al.,
1996; Schiffmann et al., 1997; Ranze et al., 1999) (see also Chap. 20); there-
fore, no clear treatment recommendations can be made. Single cases sug-
gest that lepirudin, argatroban, and danaparoid can be used successfully
in these children. The dosing schedules for adults (appropriately weight-

adjusted for the child) can be used as a guideline, but careful monitoring is
recommended.
V. ADJUNCTIVE THERAPIES
A. Medical Thrombolysis
Thrombocytopenia is not a contraindication to thrombolytic therapy in
patients with HIT. Streptokinase (Fiessinger et al., 1984; Cohen et al.,
1985; Bounameaux et al., 1986; Cummings et al., 1986; Mehta et al., 1991),
urokinase (Leroy et al., 1985; Krueger et al., 1985; Clifton and Smith, 1986),
and tissue plasminogen activator (t-PA) (Dieck et al., 1990; Schiffmann et al.,
1997) have been used both systemically and by local infusion (Quinones-
Baldrich et al., 1989). In patients at high bleeding risk, an ultra-low–dose t-PA
(2 mg/h over 12 h) was successfully applied without bleeding complications
(Olbrich et al., 1998). As thrombin generation is not inhibited by thrombol-
ysis, concomitant nonheparin anticoagulation should be given, in reduced
dose, until the fibrinolytic effects have waned.
Recommendation. Regional or systemic pharmacological thrombolysis
should be considered as a treatment adjunct in selected patients with
limb-threatening thrombosis or pulmonary embolism with severe
cardiovascular compromise (grade 2C).
B. Surgical Thromboembolectomy
Vascular surgery is often needed to salvage an ischemic limb threatened by
HIT-associated acute arterial thromboembolism involving large arteries
(Sobel et al., 1988). When performing vascular surgery during acute HIT, it
is appropriate to maintain anticoagulation at least in the lower therapeutic
range, if possible, before, during, and after surgery, until platelet count
recovery. In patients with latent HIT (i.e., no longer thrombocytopenic, but
with clinically significant levels of HIT antibodies still present), the intensity
of anticoagulation depends on the perceived risk of vessel (or graft) occlu-
sion. In patients at high risk of occlusion (e.g., surgery involving below-knee
vessels), the patient should be therapeutically anticoagulated before vessel
clamping (in addition to receiving intraoperative flushes with anticoagu-
lant), with therapeutic anticoagulation maintained for several days after
surgery. In surgery involving larger vessels, the use of intraoperative flushes
alone, followed by postoperative prophylactic-dose anticoagulation, might
be sufficient.
Either danaparoid or lepirudin can provide intraoperative anticoagu-

lation. One author (AG) uses one of the following solutions to flush the vessel
postembolectomy: (1) lepirudin, 0.1 mg/mL saline (one 20-mg ampule in 200
mL saline), using up to 250 mL in a normal-weight patient, and assessing the
aPTT before giving more lepirudin to avoid overdosage (the lepirudin flushes
thus can achieve therapeutic intraoperative anticoagulation; see pp. 425–426);
(2) danaparoid, 3 anti-Xa U/mL (one 750 U ampule in 250 mL saline), using
up to 50 mL in a normal-weight patient (this small flush dose is used because
systemic anticoagulation is achieved by giving a 2250 U bolus of danaparoid
preoperatively (see p. 377).
Recommendation. Surgical thromboembolectomy is an appropriate
adjunctive treatment for selected patients with limb-threatening large-
vessel arterial thromboembolism. Thrombocytopenia is not a contra-
indication to surgery. An alternative anticoagulant to heparin should be
used for intraoperative anticoagulation (grade 1C).
C. Intravenous Gammaglobulin
In vitro, both intact IgG as well as its Fc fragments inhibit HIT antibody-
induced platelet activation, an effect that depends somewhat on the method of
immunoglobulin preparation (Greinacher et al., 1994a) (see Chap. 9). Case
reports describe rapid increase in the platelet counts after high-dose intrave-
nous (iv) IgG (Vender et al., 1986; Frame et al., 1989; Nurden et al., 1991;
Grau et al., 1992; Prull et al., 1992; Warkentin and Kelton, 1994). The
possibility that ivIgG treatment interrupts platelet activation by HIT anti-
bodies provides a rationale for its use as an adjunct to anticoagulant therapy
in certain life- or limb-threatening situations. The dose should be 1 g/kg body
weight per day for 2 consecutive days.
Recommendation. ivIgG is a possible adjunctive treatment in selected
patients requiring rapid blockade of the Fc receptor-dependent platelet-
activating effects of HIT antibodies (e.g., management of patients with
cerebral venous thrombosis, severe limb ischemia, or very severe throm-
bocytopenia) (grade 2C).
D. Plasmapheresis
Plasmapheresis has been associated with successful treatment outcomes in
uncontrolled studies of patients with severe HIT (Vender et al., 1986; Bouvier
et al., 1988; Nand and Robinson 1988; Thorp et al., 1990; Manzano et al.,
1990; Brady et al., 1991; Poullin et al., 1998). Whether this is due to removal
of HIT antibodies or pathogenic immune complexes, or even correction of
acquired natural anticoagulant deficiencies by normal plasma replacement, is

unresolved. For example, a patient with warfarin-induced acquired protein C
deficiency and severe venous limb ischemia may have benefited from correc-
tion of the protein C deficiency with apheresis using plasma replacement
(Warkentin et al., 1997).
Recommendation. Plasmapheresis, using plasma as replacement fluid,
may be a useful adjunctive therapy in selected patients with acute HIT
and life- or limb-threatening thrombosis who are suspected or proved to
have acquired deficiency of one or more natural anticoagulant proteins
(grade 2C).
E. Antiplatelet Agents
Dextran
Dextran in high concentrations inhibits platelet function and fibrinogen
polymerization. It also inhibits HIT antibody-mediated platelet aggregation
(Sobel et al., 1986). However, a prospective randomized trial (Chong et al.,
2001) (see Chap. 14) showed that in patients with severe HIT-associated
thrombosis, dextran was less effective therapy than danaparoid. It is unknown
whether dextran would provide additional clinical benefit if combined with
another anticoagulant. Neither of us uses dextran for the management of HIT.
Recommendation. Dextran should not be used as primary therapy for
acute HIT complicated by thrombosis (grade 1B).
Acetylsalicylic Acid and Dipyridamole

Both acetylsalicylic acid (aspirin, ASA) and dipyridamole have been used in
HIT patients with variable success (Janson et al., 1983; Makhoul et al., 1986;
Kappa et al., 1987, 1989; Laster et al., 1989; Gruel et al., 1991; Hall et al.,
1992; Almeida et al., 1998). Sometimes the platelet count appeared to rise
promptly with the application of antiplatelet therapy (Warkentin, 1997).
However, HIT antibodies are potent platelet activators, and their effect
cannot always be blocked in vitro by ASA or dipyridamole. These antiplatelet
agents may be used as adjunctive therapy, but one should note that they can
cause prolonged platelet inhibition. Because there is little information on the
interactions of nonheparin anticoagulants with antiplatelet agents, in HIT
combined use should probably be restricted to patients judged to be at high
risk for arterial thromboembolism.
Recommendation. Antiplatelet agents, such as aspirin, may be used as
adjuncts to anticoagulant therapy of HIT, particularly in selected pa-
tients at high risk for arterial thromboembolism. The possible benefit in

preventing arterial thrombosis should be weighed against the potential
for increased bleeding (grade 2C).
Platelet Glycoprotein IIb/IIIa Inhibitors

Several platelet glycoprotein (GP) IIb/IIIa inhibitors are now available that
potently block fibrinogen binding to platelets. They also can reduce thrombin
generation by inhibiting the exposure of procoagulant phospholipid surfaces
on platelets (Pedicord et al., 1998; Keularts et al., 1998, Hérault et al., 1998).
In vitro, GP IIb/IIIa antagonists inhibit platelet aggregation (Hérault et al.,
1997), endothelial cell activation (Herbert et al., 1998), and platelet micro-
particle generation (Mak et al., 1998) by HIT antibodies. However, Fc
receptor–dependent platelet activation by HIT antibodies is independent of
the GP IIb/IIIa complex (Greinacher et al., 1994a); therefore, GP IIb/IIIa
inhibitors do not inhibit platelet granule release (Tsao et al., 1997; Polgár
et al., 1998). As these agents do not have a direct anticoagulant effect, they
probably need to be combined with an anticoagulant (danaparoid, lepirudin,
or argatroban) to treat HIT. Because there are no data available on the inter-
action of these newer anticoagulants with the GP IIb/IIIa inhibitors, and
because a synergistic effect on bleeding is likely, combined use for the manage-
ment of HIT should be considered experimental. Theoretically, synthetic GP
IIb/IIIa inhibitors with a short half-life could be safer than agents with a long
half-life (e.g., abciximab)
Recommendation. GP IIb/IIIa inhibitors should be considered as
experimental treatment in HIT and used with caution if combined with
anticoagulant drugs (grade 2C).
VI. CAVEATS FOR THE TREATMENT OF HIT

A. Low Molecular Weight Heparin
Low molecular weight heparin is less likely than UFH to cause HIT antibody
formation as well as clinical HIT (Warkentin et al., 1995, 2003). Furthermore,
LMWH binds less avidly to platelets than does UFH (Greinacher et al.,
1993b). With functional assays employing platelet-rich plasma, several
investigators reported a reduced cross-reactivity of HIT antibodies with
LMWH compared with UFH (Ramakrishna et al., 1995; Slocum et al.,
1996; Vun et al., 1996); however, with sensitive washed platelet functional
assays, the cross-reactivity rate of LMWH is nearly 100% (Greinacher et al.,
1992; Warkentin et al., 1995) (see Chap. 11).
Owing to the unavailability of other anticoagulant options during the

1980s, LMWH preparations were often used in Europe for further parenteral
anticoagulation of HIT patients. No prospective cohort studies are available,
but case reports (Roussi et al., 1984; Leroy et al., 1985; Vitoux et al., 1986;
Gouault-Heilmann et al., 1987; Bauriedel et al., 1988; Kirchmaier and
Bender, 1988) and a review (Reuter, 1987) suggest that LMWH may benefit
some patients. Other case series, however, clearly show that LMWH is asso-
ciated with disastrous complications in HIT patients (Horellou et al., 1984;
Leroy et al., 1985; Gouault-Heilmann et al., 1987; Greinacher et al., 1992;
Kleinschmidt et al., 1993). Unfortunately, no laboratory assay reliably pre-
dicts these differing treatment responses.
Treatment of HIT with LMWH is frequently unsuccessful. Of eight
consecutive HIT patients who received LMWH, thrombocytopenia persisted
in all, and new thromboembolic events occurred in two patients (Greinacher
et al., 1992). After LMWH became available in North America, a similar
experience was observed in seven HIT patients treated with LMWH (War-
kentin, 1997). A recent study has also shown a relatively high risk of adverse
outcomes of treating HIT with LMWH (Ranze et al., 2000).
Recommendation. LMWH should not be used to treat patients with
acute HIT (grade 1C).
B. Oral Anticoagulants (Vitamin K Antagonists)

Although oral anticoagulants, such as warfarin, phenprocoumon, and other
coumarin agents, are an important part of the longer-term management of
patients with HIT-associated thrombosis, they are ineffective, and potentially
dangerous, when given as single therapy, or in combination with ancrod, in
patients with acute HIT (Warkentin et al., 1997) (see Chap. 3). In patients
with active DVT, oral anticoagulants may cause thrombosis to progress to
involve even the microvasculature, leading to coumarin-induced venous limb
gangrene. This syndrome appears to result from a transient disturbance in
procoagulant–anticoagulant balance: increased thrombin generation associ-
ated with HIT remains high during early warfarin treatment, while simulta-
neously there is severe, acquired deficiency in the natural anticoagulant
protein C. Although high doses of oral anticoagulants may be more likely
to cause this syndrome, even relatively low doses that produce a rise in the
INR to higher than 4.0 can cause limb gangrene in some patients. Thus,
warfarin and phenprocoumon should always be given in combination with an
agent that reduces thrombin generation in patients with acute HIT. Further-
more, it is prudent to delay coumarin anticoagulation until the HIT is ade-
quately controlled by the parenteral anticoagulant, as judged by improved or

stabilized thrombotic signs and partial or complete platelet count recovery.
Recommendation. Oral anticoagulants are contraindicated in patients
with acute HIT, unless combined with an agent that reduces thrombin
generation, (grade 1C+).
C. Ancrod
Ancrod is a defibrinogenating enzyme obtained from the venom of the
Malayan pit viper. The thrombin-like enzyme cleaves fibrinopeptide A from
fibrinogen but, in contrast with thrombin, does not proteolyze fibrinopeptide
B (Bell, 1997). On the basis of uncontrolled studies, ancrod has been used
successfully to treat several patients with HIT, primarily in Canada (Teasdale
et al., 1989; Cole et al., 1990; Demers et al., 1991).
However, ancrod does not inhibit thrombin generation, and in HIT
patients it even appears to increase thrombin generation initially (Fig. 2)
(Warkentin, 1998). Animal models indicate that under special clinical circum-
stances, such as septicemia, ancrod contributes to enhanced fibrin deposition
(Krishnamurti et al., 1993). These data could help explain why some patients
have developed venous limb gangrene during combined treatment with
ancrod and warfarin (Warkentin et al., 1997; Gupta et al., 1998) (i.e., in-
creased thrombin generation during ancrod treatment could contribute to
the disturbance in procoagulant–anticoagulant balance during warfarin ther-
apy that has been hypothesized to explain venous limb gangrene) (Warken-
tin et al., 1997). In a retrospective nonblinded comparison, ancrod appeared
to be less effective than danaparoid in one medical community (Warkentin,
1996).
The manufacturer discontinued ancrod in 2002.
D. Platelet Transfusions
Usually there is no need to treat thrombocytopenia with platelet transfusions,
as patients with HIT rarely bleed spontaneously. Indeed, platelet transfusions
should be avoided because the transfused platelets can be activated by the
same immune mechanisms as the patient’s own platelets. Anecdotal experi-
ence describes thrombotic events soon after platelet transfusions given to
patients with acute HIT (Babcock et al., 1976; Cimo et al., 1979). Several
consensus conferences (Contreras, 1998; Hirsh et al., 2001b; British Commit-
tee for Standards in Haematology, 2003) stated that thrombotic thrombo-
cytopenic purpura (TTP) and HIT are two disorders in which prophylactic
platelet transfusions are not recommended because of the risk of precipitating

thrombosis.
Recommendation. Prophylactic platelet transfusions are relatively
contraindicated in patients with acute HIT (grade 2C).
Therapeutic platelet transfusions are appropriate for patients with HIT

who develop severe hemorrhage, particularly if the heparin administration
has been discontinued for more than a day.
E. Vena Cava Filters

Vena cava (Greenfield) filters are sometimes used to manage patients judged
to be at high risk for life-threatening pulmonary embolism. However, their
use can be complicated by massive vena cava thrombosis, including the renal
veins, and other serious progression of venous thromboembolism, especially
if pharmacological anticoagulation is not given (Sobel et al., 1988; Jouanny et
al., 1993). In our opinion, these devices should rarely be used in patients with
acute HIT.
REFERENCES
Agnelli G, Iorio A, De Angelis V, Nenci GG. Dermatan sulphate in heparin-induced

thrombocytopenia [letter]. Lancet 1994; 344:1295–1296.
Almeida JI, Coats R, Liem TK, Silver D. Reduced morbidity and mortality rates of the
heparin-induced thrombocytopenia syndrome. J Vasc Surg 1998; 27:309–316.
Amiral J, Lormeau JC, Marfaing-Koka A, Vissac AM, Wolf M, Boyer-Neumann C,
Tardy B, Herbert JM, Meyer D. Absence of cross-reactivity of SR90107A/
ORG31540 pentasaccharide with antibodies to heparin–PF4 complexes devel-
oped in heparin-induced thrombocytopenia. Blood Coagul Fibrinolysis 1997;
8:114–117.
cytopenia stimulate monocytic cells to express cells to express tissue factor and
secrete interleukin-8. Blood 2001; 98:1252–1254.
Babcock RB, Dumper CW, Scharfman WB. Heparin-induced thrombocytopenia. N
Engl J Med 1976; 295:237–241.
Bauriedel G, Gerbig H, Riess H, Samtleben W, Steinbeck G. Heparin-induzierte
Thrombozytopenie. Weiterbehandlung mit niedermolekularem Heparin.
Munch Med Wochenschr 1988; 8:133–134.
Bell WR Jr. Defibrinogenating enzymes. Drugs 1997; 54(suppl 3):18–31.
Bounameaux H, de Moerloose P, Schneider PA, Leuenberger A, Krähenbühl B,
Bouvier CA. Thrombose arterielle femorale associee a une thrombopenie induit
par l’heparin. Schweiz Med Wochenschr 1986; 116:1576–1579.
Bouvier JL, Lefevre P, Villain P, Elias A, Durand JM, Juhan I, Serradimigni A.

Treatment of serious heparin-induced thrombocytopenia by plasma exchange:
report on 4 cases. Thromb Res 1988; 51:335–336.
Brace LD, Fareed J. Heparin-induced platelet aggregation. II. Dose/response rela-
tionships for two low molecular weight heparin fractions (CY 216 and CY 222).
Thromb Res 1990; 59:1–14.
Brady J, Riccio JA, Yumen OH, Makary AZ, Greenwood SM. Plasmapheresis: a
therapeutic option in the management of heparin-associated thrombocytopenia
with thrombosis. Am J Clin Pathol 1991; 96:394–397.
British Committee for Standards in Haematology, Blood Transfusion Task Force.
Guidelines for the use of platelet transfusions. Br J Haematol 2003; 122:10–23.
Cadroy Y, Amiral J, Raynaud H, Brunel P, Mazaleyrat A, Sauer M, Sie P. Evolution
of antibodies anti-PF4/heparin in a patient with a history of heparin-induced
thrombocytopenia reexposed to heparin (letter). Thromb Haemost 1994; 72:
783–784.
Calhoun BC, Hesser JW. Heparin-associated antibody with pregnancy: discussion of
two cases. Am J Obstet Gynecol 1987; 156:964–966.
Campbell KR, Mahaffey KW, Lewis BE, Weitz JI, Berkowitz SD, Ohman EM, Califf
RM. Bivalirudin in patients with heparin-induced thrombocytopenia under-
going percutaneous coronary intervention. J Invas Cardiol 2000; 12(suppl F):
14F–19F.
Chamberlin JR, Lewis B, Leya F, Wallis D, Messmore H, Hoppensteadt D, Walenga
JM, Moran S, Fareed J, McKiernan T. Successful treatment of heparin-associ-
ated thrombocytopenia and thrombosis using Hirulog. Can J Cardiol 1995;
11:511–514.
Chong BH, Murray B, Berndt MC, Dunlop LC, Brighton T, Chesterman CN. Plasma
P-selectin is increased in thrombotic consumptive platelet disorders. Blood
1994; 83:1535–1541.
Chong BH, Gallus AS, Cade JF, Magnani H, Manoharan A, Oldmeadow M, Arthur
C, Rickard K, Gallo J, Lloyd J, Seshadri P, Chesterman CN. Prospective ran-
domised open-label comparison of danaparoid with dextran 70 in the treatment
of heparin-induced thrombocytopenia with thrombosis. A clinical outcome
study. Thromb Haemost 2001; 86:1170–1175.
Cimo PL, Moake JL, Weinger RS, Ben-Menachem Y, Khalil KG. Heparin-induced
thrombocytopenia: association with a platelet aggregating factor and arterial
thromboses. Am J Hematol 1979; 6:125–133.
Clifton GD, Smith MD. Thrombolytic therapy in heparin-associated thrombocyto-
penia with thrombosis. Clin Pharm 1986; 5:597–601.
Cole CW, Fournier LM, Bormanis J. Heparin-associated thrombocytopenia and
thrombosis: optimal therapy with ancrod. Can J Surg 1990; 33:207–210.
Cohen JI, Cooper MR, Greenberg CS. Streptokinase therapy of pulmonary emboli
with heparin-associated thrombocytopenia. Arch Intern Med 1985; 145:1725–
1726.
Contreras M. The appropriate use of platelets: an update from the Edinburgh Con-
sensus Conference. Br J Haematol 1998; 101(suppl 1):10–12.
Copplestone A, Oscier DG. Heparin-induced thrombocytopenia in pregnancy. Br J
Haematol 1987; 65:248.
Cummings JM, Mason TJ, Chomka EV, Pouget JM. Fibrinolytic therapy of acute
myocardial infarction in the heparin thrombosis syndrome. Am Heart J 1986;
112:407–409.
Danhof M, de Boer A, Magnani HN, Stiekema JCJ. Pharmacokinetic considerations
on Orgaran (Org 10172) therapy. Hemostasis 1992; 22:73–84.
Demers C, Ginsberg JS, Brill-Edwards P, Panju A, Warkentin TE, Anderson DR,
Turner C, Kelton JG. Rapid anticoagulation using ancrod for heparin-induced
Dieck JA, Rizo-Patron C, Unisa A, Mathur V, Massumi GA. A new manifestation
and treatment alternative for heparin-induced thrombosis. Chest 1990; 98:
1524–1526.
Eichler P, Lubenow N, Greinacher A. Results of the third prospective study of treat-
ment with lepirudin in patients with heparin-induced thrombocytopenia (HAT)
Elalamy I, Lecrubier C, Potevin F, Abdelouahed M, Bara L, Marie JP, Samama M.
Absence of in vitro cross-reaction of pentasaccharide with the plasma heparin-
dependent factor of twenty-five patients with heparin-associated thrombocyto-
penia. Thromb Haemost 1995; 74:1384–1385.
Eriksson BI, Wille-Jorgensen P, Kalebo P, Mouret P, Rosencher N, Bosch P, Baur M,
Ekman S, Bach D, Lindbratt S, Close P. A comparison of recombinant hirudin
with a low-molecular-weight heparin to prevent thromboembolic complications
after total hip replacement. N Engl J Med 1997; 337:1329–1335.
Farner B, Eichler P, Kroll H, Greinacher A. A comparison of danaparoid and lepirudin
in heparin-induced thrombocytopenia. Thomb Haemost 2001; 85:950–957.
Fiessinger JN, Aiach M, Rocanto M, Debure C, Gaux JC. Critical ischemia during
heparin-induced thrombocytopenia. Treatment by intra-arterial streptokinase.
Thromb Res 1984; 33:235–238.
Frame JN, Mulvey KP, Phares JC, Anderson MJ. Correction of severe heparin-
associated thrombocytopenia with intravenous immunoglobulin. Ann Intern
Med 1989; 111:946–947.
Francis JL, Drexler JL, Gwyn G, Moroose R. Bivalirudin, a direct thrombin inhibitor,
most 2003; 1(suppl 1):P1909.
Baggio G, Fabris F, Girolami A. The incidence of heparin-induced throm-
bocytopenia in hospitalized medical patients treated with subcutaneous un-
fractionated heparin: a prospective cohort study. Blood 2003; 101:2955–2959.
Gosselin RC, Dager WE, King JH, Janatpour KA, Mahackian KA, Larkin EC,
Owings JT. Effect of direct-thrombin-inhibitors: bivalirudin, lepirudin, and

argatroban, on prothrombin time measurements [abstr]. Am Coll Clin Pharm
Annual Meeting, Nov. 2–5, 2003.
Gouault-Heilmann M, Huet Y, Adnot S, Contant G, Bonnet F, Intrator L, Payen D,
Levent M. Low molecular weight heparin fractions as an alternative therapy in
heparin-induced thrombocytopenia. Haemostasis 1987; 17:134–140.
Grau E, Linares M, Olaso MA, Ruvira J, Sanchis J. Heparin-induced thrombocy-
topenia—response to intravenous immunoglobulin in vivo and in vitro. Am J
Hematol 1992; 39:312.
penia: the antibody is not heparin-specific. Thromb Haemost 1992; 67:545–
549.
Greinacher A, Eckhardt T, Mussmann J, Mueller-Eckhardt C. Pregnacy complicated
by heparin-associated thrombocytopenia: management by a prospectively in
vitro selected heparinoid (Org 10172). Thromb Res 1993a; 71:123–126.
membrane by the negative charge of highly sulfated oligosaccharides. Br J
Haematol 1993b; 84:711–716.
Greinacher A, Liebenhoff U, Kiefel V, Presek P, Mueller-Eckhardt C. Heparin-
associated thrombocytopenia: the effects of various intravenous IgG prepara-
tions on antibody mediated platelet activation—a possible new indication for
high dose i.v. IgG. Thromb Haemost 1994a; 71:641–645.
Greinacher A, Poetzsch B, Amiral J, Dummel V, Eichner A, Mueller-Eckhard C.
Haemost 1994b; 71:247–251.
safe and effective anticoagulation in patients with the immunologic type of
heparin-induced thrombocytopenia: a prospective study. Circulation 1999a; 99:
73–80.
Greinacher A, Janssens U, Berg G, Böck M, Kwasny H, Kemkes-Matthes B, Eichler
P, Völpel H, Pötzsch B, Luz M. Lepirudin (recombinant hirudin) for parenteral
anticoagulation in patients with heparin-induced thrombocytopenia. Circula-
tion 1999b; 100:587–593.
Greinacher A, Eichler P, Lubenow N, Kwasny H, Luz H. Heparin-induced thrombo-
cytopenia with thromboembolic complications: meta-analysis of two prospec-
Greinacher A, Lubenow N, Eichler P. Anaphylactic and anaphylactoid reactions
associated with lepirudin in patients with heparin-induced thrombocytopenia.
Circulation 2003; 108:2062–2065.
Gruel Y, Lang M, Darnige L, Pacouret G, Dreyfus X, Leroy J, Charbonnier B. Fatal
effect of re-exposure to heparin after previous heparin-associated thrombocy-
topenia and thrombosis. Lancet 1990; 336:1077–1078.
Gruel Y, Lermusiaux P, Lang M, Darnige L, Rupin A, Delahousse B, Guilmot JL,
Leroy J. Usefulness of antiplatelet drugs in the management of heparin-asso-
ciated thrombocytopenia and thrombosis. Ann Vasc Surg 1991; 5:552–555.
macother 1998; 32:55–59.
Guyatt G, Schunëmann H, Cook D, Jaeschke R, Pauker S, Bucher H. Grades of
recommendation for antithrombotic agents. Chest 2001; 119(suppl):3S–7S.
Clin Nephrol 1992; 38:86–89.
Harrison L, Johnston M, Massicotte MP, Crowther M, Moffat K, Hirsh J. Com-
parison of 5-mg and 10-mg loading doses in initiation of warfarin therapy. Ann
Intern Med 1997; 126:133–136.
Henny CHP, ten Cate H, ten Cate JW, Prummel MF, Peters M, Büller HR.
Thrombosis prophylaxis in an AT III deficient pregnant women: application of
a low molecular weight heparinoid. Thromb Haemost 1986; 55:301.
Hérault JP, Lalé A, Savi P, Pflieger AM, Herbert JM. In vitro inhibition of heparin-
induced platelet aggregation in plasma from patients with HIT by SR 121566, a
newly developed Gp IIb/IIIa antagonist. Blood Coagul Fibrinolysis 1997;
8:206–207.
Hérault JP, Peyrou V, Savi P, Bernat A, Herbert JM. Effect of SR121566A, a potent
GP IIb–IIIa antagonist on platelet-mediated thrombin generation in vitro and
in vivo. Thromb Haemost 1998; 79:383–388.
Herbert JM, Savi P, Jeske WP, Walenga JM. Effect of SR 121566A, a potent GP IIb–
IIIa antagonist, on the HIT serum/heparin-induced platelet mediated activation
of human endothelial cells. Thromb Haemost 1998; 80:326–331.
Hirsh J, Warkentin TE, Raschke R, Granger C, Ohman EM, Dalen JE. Heparin and
low-molecular-weight heparin. Mechanisms of action, pharmacokinetics, dosing
considerations, monitoring, efficacy, and safety. Chest 1998; 114:489S–510S.
Hirsh J, Dalen JE, Guyatt G. The sixth (2000) ACCP guidelines for antithrombotic
therapy for prevention and treatment of thrombosis. Chest 2001a; 119(suppl):
1S–2S.
Granger C, Ohman EM, Dalen JE. Heparin and low-molecular-weight heparin.
safety. Chest 2001b; 119(suppl):64S–94S.
Horellou MH, Conard J, Lecrubier C, Samama M, Roque-D’Orbcastel O, de Fenoyl
O, Di Maria G, Bernadou A. Persistent heparin-induced thrombocytopenia
despite therapy with low molecular weight heparin. Thromb Haemost 1984;
51:134.
Huhle G, Geberth M, Hoffmann U, Heene DL, Harenberg J. Management of heparin-
associated thrombocytopenia in pregnancy with subcutaneous r-hirudin. Gy-
necol Obstet Invest 2000; 49:67–69.
Janson PA, Moake JL, Garpinito C. Aspirin prevents heparin-induced platelet aggre-
gation in vivo. Br J Haematol 1983; 53:166–168.
Jouanny P, Jeandel C, Laurain MC, Penin F, Cuny G. Thrombopénie a l’héparine
et filtre cave. Difficultés du traitement. J Mal Vas 1993; 18:320–322.
Kappa JR, Horn MK III, Fisher CA, Cottrell ED, Ellison A, Addonizio VP Jr.
Efficacy of iloprost (ZK36374) versus aspirin in preventing heparin-induced
platelet activation during cardiac operations. J Thorac Cardiovasc Surg 1987;
97:405–413.
Kappa JR, Fisher CA, Addonizio VP. Heparin-induced platelet activation: the role of
thromboxane A2 synthesis and the extent of granule release in two patients. J
Vasc Surg 1989; 9:574–579.
925–930.
Keularts IMLW, Béguin S, de Zwaan C, Hemker HC. Treatment with a GPIIb/III
antagonist inhibits thrombin generation in platelet rich plasma from patients.
Kirchmaier CM, Bender N. Heparin-induzierte Thrombozytopenie mit arterieller und
venöser Thrombose. Inn Med 1988; 15:174–178.
Kleinschmidt S, Ziegenfuss T, Seyfert UT, Greinacher A. Septisch toxisches Herz
Kreislauf Versagen als Folge einer Heparin-induzierten Thrombozytopenie mit
‘‘White Clot Syndrome’’. Anaesthesiol Intensivmed Notfallmed Schmerzther
1993; 28:58–60.
Klement D, Rammos S, von Kries R, Kirschke W, Kniemeyer HW, Greinacher A.
Heparin as a cause of thrombus progression. Heparin-associated thrombocy-
topenia is an important differential diagnosis in paediatric patients even with
normal platelet counts. Eur J Pediatr 1996; 155:11–14.
Krishnamurti C, Bolan C, Colleton CA, Reilly TM, Alving BM. Role of plasminogen
activator inhibitor-1 in promoting fibrin deposition in rabbits infused with an-
crod or thrombin. Blood 1993; 82:3631–3636.
Krueger SK, Andreas E, Weinand E. Thrombolysis in heparin-induced thrombocy-
topenia with thrombosis. Ann Intern Med 1985; 103:159.
Lagrange F, Vergnes C, Brun JL, Paolucci F, Nadal T, Leng JJ, Saux MC, Banwarth B.
Absence of placental transfer of pentasaccharide (fondaparinux, arixtra) in the
dually perfused human cotyledon in vitro. Thromb Haemost 2002; 87:831–835.
Laster JL, Elfrink R, Silver D. Reexposure to heparin of patients with heparin-asso-
ciated antibodies. J Vasc Surg 1989; 9:677–681.
Leroy J, Leclerc MH, Delahousse B, Guerois C, Foloppe P, Gruel Y, Toulemonde F.
Treatment of heparin-associated thrombocytopenia and thrombosis with low
molecular weight heparin (CY 216). Semin Thromb Hemost 1985; 11:326–329.
Lewis BE, Wallis DE, Berkowitz SD, Matthai WH, Fareed J, Walenga JM, Barth-
olomew J, Sham R, Lerner RG, Zeigler ZR, Rustagi PK, Jang IK, Rifkin SD,
Moran J, Hursting MJ, Kelton JG, for the ARG-911 Study Investigators.
Circulation 2001; 103:1838–1843.
Lewis BE, Wallis DE, Leya F, Hursting MJ, Kelton JG. Argatroban anticoagulation
in patients with heparin-induced thrombocytopenia. Arch Intern Med 2003;
163:1849–1856.
Lindhoff-Last E, Piechottka GP, Rabe F, Bauersachs R. Hirudin determination in
plasma can be strongly influenced by the prothrombin level. Thromb Res 2000;
100:55–60.
Lubenow N, Greinacher A. Heparin-induced thrombocytopenia. Recommendations
for optimal use of recombinant hirudin. BioDrugs 2000; 14:109–125.
Lubenow N, Eichler P, Leitz T, Greinacher A. Meta-analysis of three prospective
studies of lepirudin in the prevention of thrombosis in patients with heparin-
induced thrombocytopenia. Blood [abstr] 2002a; 100(suppl 1):501a–502a.
Lubenow N, Eichler P, Greinacher A. Results of a large drug monitoring program
confirms the safety and efficacy of Refludan (lepirudin) in patients with immune-
mediated heparin-induced thrombocytopenia [abstr]. Blood 2002b; 100(suppl
1):502a.
to initial use or reexposure to heparin. Chest 2002c; 122:37–42.
Magnani HN. Orgaran (danaparoid sodium) use in the syndrome of heparin-induced
thrombocytopenia. Platelets 1997; 8:74–81.
Mak KH, Kottke-Marchant K, Brooks LM, Topol EJ. In vitro efficacy of platelet
glycoprotein IIb/IIIa antagonist in blocking platelet function in plasma of pa-
tients with heparin-induced thrombocytopenia. Thromb Haemost 1998; 80:
989–993.
Makhoul RG, Greenberg CS, McCann RL. Heparin-associated thrombocytopenia
and thrombosis: a serious clinical problem and potential solution. J Vasc Surg
1986; 4:522–528.
Makhoul RG, McCann RL, Austin EH, Greenberg CS, Lowe JE. Management of
patients with heparin-associated thrombocytopenia and thrombosis requiring
cardiac surgery. Ann Thorac Surg 1987; 43:617–621.
Manzano L, Yebra M, Vargas JA, Barbolla L, Alvarez-Mon M. Plasmapheresis in
heparin-induced thrombocytopenia and thrombosis [letter]. Stroke 1990; 21:
1236.
Markwardt F, Fink G, Kaiser B, Klöcking HP, Nowak G, Richter M, Stürzebecher
J. Pharmacological survey of recombinant hirudin. Pharmazie 1988; 43:202–
207.
Mattioli AV, Bonetti L, Sternieri S, Mattioli G. Heparin-induced thrombocytopenia
in patients treated with unfractionated heparin: prevalence of thrombosis in a 1
year follow-up. Ital Heart J 2000; 1:39–42.
McIntyre K. Medicolegal implications of the Consensus Conference. Chest 2001;

119(suppl):337S–343S.
Mehta DP, Yoder EL, Appel J, Bergsman KL. Heparin-induced thrombocytopenia
and thrombosis: reversal with streptokinase. A case report and review of litera-
ture. Am J Hematol 1991; 36:275–279.
Meytes D, Ayalon H, Virag I, Weisbort Y, Zakut H. Heparin-induced thrombocy-
topenia and recurrent thrombosis in pregnancy. A case report. J Reprod Med
1986; 31:993–996.
Murdoch IA, Beattie RM, Silver DM. Heparin-induced thrombocytopenia in chil-
dren. Acta Paediatr 1993; 82:495–497.
Nand S, Robinson JA. Plasmapheresis in the management of heparin-associated
thrombocytopenia with thrombosis. Am J Hematol 1988; 28:204–206.
Nand S. Hirudin therapy for heparin-associated thrombocytopenia and deep venous
thrombosis. Am J Hematol 1993; 43:310–311.
Nurden AT, Laroche-Traineau J, Jallu V, Broult J, Durrieu C, Besse P, Brossel
C, Hourdillé P. Heparin-induced thrombocytopenia: observation of the na-
ture of the antibody activities and on the use of gammaglobulin concentrates
in a patient with thrombotic complications. [abstr]. Thromb Haemost 1991;
65:796.
Olbrich K, Wiersbitzky M, Wacke W, Eichler P, Zinke H, Schwock M, Möx B, Kraatz
G, Motz W, Greinacher A. Atypical heparin-induced thrombocytopenia
complicated by intracardiac thrombus, effectively treated with ultra-low-dose
rt-PA lysis and recombinant hirudin (lepirudin). Blood Coagul Fibrinolysis
1998; 9:273–277.
Oriot D, Wolf M, Wood C, Brun P, Sidi D, Devictor D, Tchernia G, Huault G.
Thrombopénie sévère induite par l’héparine chez un nourrisson porteur d’une
myocardite aiguë. Arch Fr Pediatr 1990; 47:357–359.
Ortel TL, Chong BH. New treatment options for heparin-induced thrombocytopenia.
Semin Hematol 1998; 35:26–34.
Pedicord DL, Thomas BE, Mousa SA, Dicker IB. Glycoprotein IIb/IIIa receptor
antagonists inhibit the development of platelet procoagulant activity. Thromb
Res 1998; 90:247–258.
554.
child. J Pediatr 1992; 121:135–138.
N Engl J Med 2000; 343:515.
Poullin P, Pietri P, Lefevre P. Heparin-induced thrombocytopenia with thrombosis:
successful treatment with plasma exchange [letter]. Br J Haematol 1998; 102:

630–631.
Pouplard C, Iochmann S, Renard B, Herault O, Colombat P, Amiral J, Gruel Y.
Induction of monocyte tissue factor expression by antibodies to heparin-platelet
factor 4 complexes developed in heparin-induced thrombocytopenia. Blood
2001; 97:3300–3302.
Prull A, Nechwatal R, Riedel H, Mäurer W. Therapie des Heparin-induzierten
Thrombose–Thrombozytopenie Syndroms mit Immunglobulinen. Dtsch Med
Wochenschr 1992; 117:1838–1842.
Quinones-Baldrich WJ, Baker JD, Busuttil RW, Machleder HI, Moore WS. Intra-
operative infusion of lytic drug for thrombotic complications of revascularisa-
tion. J Vasc Surg 1989; 10:408–417.
Ramakrishna R, Manoharan A, Kwan YL, Kyle PW. Heparin-induced thrombocyto-
penia: cross-reactivity between standard heparin, low molecular weight heparin,
dalteparin (Fragmin) and heparinoid, danaparoid (Orgaran). Br J Haematol
1995; 91:736–738.
in paediatric patients: a review of the literature and a new case treated with
Ranze O, Eichner A, Lubenow N, Kempf R, Greinacher A. The use of low-molecular-
weight heparins in heparin-induced thrombocytopenia (HIT): a cohort study
[abstr]. Ann Hematol 2000; 79(suppl 1):P198.
Reid T III, Alving BM. Hirulog therapy for heparin-associated thrombocytopenia and
deep venous thrombosis. Am J Hematol 1994; 45:352–353.
Thromb Res 1996; 81:641–649.
Reuter HD. Niedermolekulares Heparin in der Therapie der Heparininduzierten
Thrombozytopenie. Med Klin 1987; 82:115–118.
Roussi JH, Houboyan LL, Goguel AF. Use of low-molecular-weight heparin in hep-
arin-induced thrombocytopenia with thrombotic complications. Lancet 1984;
1:1183.
Schiffmann H, Unterhalt M, Harms K, Figulla HR, Völpel H, Greinacher A.
Erfolgreiche Behandlung einer Heparin-induzierten Thrombozytopenie Typ II
im Kindesalter mit rekombinantem Hirudin. Monat Kinderheilkd 1997; 145:
606–612.
patients with heparin-induced thrombocytopenia syndrome. J Vasc Surg 1996;
23:839–849.
Smith JP, Walls JT, Muscato MS, McCord ES, Worth ER, Curtis JJ, Silver D.
Extracorporeal circulation in a patient with heparin-induced thrombocytope-
nia. Anaesthesiology 1985; 62:363–365.
Smythe MA, Warkentin TE, Stephens JL, Zakalik D, Mattson JC. Venous limb
gangrene during overlapping therapy with warfarin and a direct thrombin

inhibitor for immune heparin-induced thrombocytopenia. Am J Hematol 2002;
71:50–52.
Sobel M, Adelman B, Greenfield LJ. Dextran 40 reduces heparin-mediated platelet
aggregation. J Surg Res 1986; 40:382–387.
Sobel M, Adelman B, Szentpeterey S, Hofmann M, Posner MP, Jenvey W. Surgical
management of heparin-associated thrombocytopenia. Strategies in the treat-
ment of venous and arterial thromboembolism. J Vasc Surg 1988; 8:395–401.
Srinivasan AF, Rice L, Bartholomew JR, Rangaswamy C, La Perna L, Thompson JE,
Murphy S, Baker KR. Warfarin-induced skin necrosis and venous limb
gangrene in the setting of heparin-induced thrombocytopenia. Arch Intern Med
2003. In press.
Teasdale SJ, Zulys VJ, Mycyk T, Baird RJ, Glynn MF. Ancrod anticoagulation for
cardio-pulmonary bypass in heparin-induced thrombocytopenia and thrombo-
sis. Ann Thorac Surg 1989; 48:712–713.
Thorp D, Canty A, Whiting J, Dart G, Lloyd JV, Duncan E, Gallus A. Plasma
exchange and heparin-induced thrombocytopenia. Prog Clin Biol Res 1990;
337:521–522.
Tsao PW, Forsythe MS, Mousa SA. Dissociation between the anti-aggregatory and
anti-secretory effects of platelet integrin (aIIbh2 (GPIIb/IIIa) antagonists, c7E3
and DMP728. Thromb Res 1997; 89:137–146.
Turpie AGG, Gallus AS, Hoek JA, for the Pentasaccharide Investigators. A synthetic
pentasaccharide for the prevention of deep-vein thrombosis after total hip re-
placement. N Engl J Med 2001; 344:619–625.
Van Besien K, Hoffman R, Golichowski A. Pregnancy associated with lupus anti-
coagulant and heparin-induced thrombocytopenia: management with a low
molecular weight heparinoid. Thromb Res 1991; 62:23–29.
Vender JS, Matthew EB, Silverman IM, Konowitz H, Dau PC. Heparin-associated
thrombocytopenia: alternative managements. Anesth Analg 1986; 65:520–522.
Visentin GP, Ford SE, Scott PJ, Aster RH. Antibodies from patients with heparin-
Vitoux JF, Mathieu JF, Roncato M, Fiessinger JN, Aiach M. Heparin-associated
thrombocytopenia treatment with low-molecular weight heparin. Thromb Hae-
most 1986; 55:37–39.
Vun CM, Evans S, Chong BH. Cross-reactivity study of low molecular weight hepa-
rins and heparinoid in heparin-induced thrombocytopenia. Thromb Res 1996;
81:525–532.
Wallis DE, Workman KL, Lewis BE, Steen L, Pifarre R, Moran JF. Failure of early
Med 1999; 106:629–635.
reactivity (XR) of danaparoid for HIT–IgG [abstr]. Blood 1996; 88(suppl 1): 626a.
Warkentin TE. Heparin-induced thrombocytopenia. Pathogenesis, frequency, avoid-

ance and management. Drug Safety 1997; 17:325–341.
Br J Haematol 2003; 121:535–555.
entirety in Ann Thorac Surg 2003, Dec. In press.]
Am J Med 1996; 101:502–507.
N Engl J Med 2001; 344:1286–1292.
1994; 84:3691–3699.
Deliargyris EN, Sane DC. Anti-platelet factor 4/heparin antibodies. An inde-
pendent predictor of 30-day myocardial infarction after acute coronary ischemic
syndromes. Circulation 2003; 107:2307–2312.
14
Danaparoid for the Treatment
of Heparin-Induced
Thrombocytopenia
Beng Hock Chong

St. George Hospital and University of New South Wales, Kogarah,
New South Wales, Australia
Harry N. Magnani
Medical Consultant, Oss, The Netherlands
I. INTRODUCTION
In patients with heparin-induced thrombocytopenia (HIT), cessation of hep-

arin is mandatory. Thereafter, an alternative anticoagulant is usually needed
for the treatment of HIT-associated venous or arterial thrombosis, for the
prevention of thrombosis in isolated HIT, or for other indications (Chong,
1995; Hirsh et al., 2001; Warkentin 2001) (see Chap. 13). Currently, three
antithrombotic agents, danaparoid sodium, recombinant hirudin (r-hirudin),
and argatroban, have been approved in various countries for the treatment of
patients with HIT. The clinical use of danaparoid in HIT patients will be
reviewed in this chapter, and the other treatments in Chaps. 15 and 16.
Danaparoid (Orgaran, Organon NV, The Netherlands; formerly known as
Org 10172 and Lomoparan) has been used extensively to treat HIT patients.
Worldwide, many thousands of patients have been successfully treated with
this drug.
371
372 Chong and Magnani
A. Chemistry, Pharmacology, Pharmacodynamics,

and Pharmacokinetics
Chemistry
Although danaparoid is often referred to as a low molecular weight (LMW)
heparinoid (implying that it has heparin-like activity), there are, in fact, sub-
stantial differences in the chemistry, pharmacology, and pharmacokinetics
between danaparoid and both unfractionated heparin (UFH) and low mo-
lecular weight heparin (LMWH).
Danaparoid consists of a mixture of LMW glycosaminoglycans;
namely, heparan sulfate (84%), dermatan sulfate (12%), and chondroitin
sulfate (4%) (Meuleman, 1992). A small proportion of the heparan sulfate
molecules have high affinity for antithrombin (AT, formerly known as anti-
thrombin III) (Meuleman, 1992; Ofosu, 1992). Danaparoid has an average
molecular mass of approximately 6000 Da. It does not contain heparin or
heparin fragments, and differs in chemical composition from heparin in that
the repeating disaccharide subunits in heparan sulfate, its principal constit-
uent, are predominantly glucuronic acid and N-acetyl-glucosamine, whereas
in heparin, they are mostly iduronic acid and glucosamine-N-sulfate (Gordon
et al., 1990) (Fig. 1). Compared with LMWH, danaparoid has a lower degree
of sulfation and a lower charge density. These two factors play an important
role in its binding (or more precisely, its lack of binding) to plasma proteins
and platelets, which is particularly relevant for the pharmacological profile of
the drug (Casu, 1991) (see Chap. 8).
Figure 1 Comparison of predominant disaccharide structure of heparin with

danaparoid. The low molecular weight (LMW) heparin disaccharide is mostly (left)
glucosamine-N-sulfate and (right) iduronic acid, whereas danaparoid’s principal con-
stituent, heparan sulfate, is predominantly (left) N-acetyl-glucosamine and (right) glu-
curonic acid. The degrees of sulfation (sulfate groups per disaccharide unit) for heparin
and danaparoid are approximately 2.0–2.5 and 1.0–1.5, respectively (see Chap. 8).
Danaparoid for Treatment of HIT 373
Pharmacology
Danaparoid exerts its antithrombotic effects predominantly by inhibition of
factor Xa; it has only minimal antifactor IIa (antithrombin) activity. Its ratio
of antifactor Xa (anti-Xa) to antithrombin (anti-IIa) is 22:1 or higher, which is
considerably greater than that of LMWH (2:1–4:1) and UFH (1:1) (Meule-
man et al., 1982; Gordon et al., 1990; Meuleman, 1992). Its inhibition of
factor Xa is mediated by AT, and its minor effect on thrombin by both AT and
heparin cofactor II. Its highly selective inhibition of factor Xa confers on this
drug the advantage of a linear inhibitory effect on thrombin generation and
fibrin formation. Another advantage is that danaparoid does not interfere
with normal platelet function (Meuleman et al., 1982; Mikhailidis et al., 1984,
1987; Meuleman, 1987). Thus, unlike UFH, danaparoid does not interfere
with platelet accretion to experimental thrombi, although thrombus growth is
markedly reduced by its prevention of fibrin accretion. Similarly, danaparoid
has minimal effects on formation of the platelet-dependent hemostatic plug
(Meuleman, 1992). These beneficial characteristics of danaparoid contribute
to the high therapeutic index (i.e., favorable benefit/risk ratio) of this drug.
Pharmacodynamics and Pharmacokinetics

Danaparoid has a pharmacokinetic profile different from that of UFH or
LMWH. It is well-absorbed after subcutaneous administration in humans,
with its bioavailability approaching 100% (Stiekema et al., 1989; Danhof et
al., 1992). In comparison, the bioavailability of LMWH is 87–92%, and that
of UFH only 15–20% (Skoutakis, 1997). Danaparoid’s plasma anti-Xa levels
peak 4–5 h following subcutaneous injection (Danhof et al., 1992). Unlike
UFH, it is not neutralized by plasma proteins, such as platelet factor 4 (PF4)
and histidine-rich glycoprotein; hence, after subcutaneous or intravenous ad-
ministration, more predictable plasma levels of the drug are obtained.
Danaparoid is eliminated mainly by the kidneys. It has a relatively long
plasma anti-Xa half-life (t1/2) of about 25 h. Plasma t1/2 values of anti-IIa
activity and thrombin generation-inhibiting activity are much shorter, rang-
ing from 2 to 4 h and 3 to 7 h, respectively (Bradbrook et al., 1987; Stiekema
et al., 1989; Danhof et al., 1992). In patients with impaired renal function,
the drug has a tendency to accumulate, and the dose should be reduced
accordingly.
Danaparoid’s metabolism is not affected by hepatic cytochrome P-450,
nor does it affect hepatic or renal handling of other drugs. It has no significant
effect on the pharmacodynamics and pharmacokinetics of coumarin anti-
coagulants. Its pharmacokinetics are not modified by age or body weight;
hence, dose adjustments are usually unnecessary in the elderly or in over-
weight patients (Stiekema et al., 1989; Danhof et al., 1992).
Similar to LMWH, protamine chloride only very minimally neutralizes

the anticoagulant activity of danaparoid. There is no effective antidote for
the drug (Stiekema et al., 1989), and in severe bleeding, the drug should be
stopped and blood product replacement given, as indicated on clinical
grounds. There is limited evidence that plasmapheresis can accelerate elim-
ination of the drug (Schmahl et al., 1997), but this option may not be practical
in an unstable patient.
II. CLINICAL USE OF DANAPAROID

A. Clinical Use of Danaparoid in Disorders Other Than HIT
Danaparoid has been studied for the prophylaxis and treatment of venous
thromboembolism in several controlled clinical studies of routine (i.e., non-
HIT) patients. These trials confirmed the efficacy of danaparoid as an anti-
thrombotic agent and, in some trials, it was even more effective than other
standard antithrombotic agents. Six prospective, randomized, controlled,
and assessor-blind studies showed that danaparoid is more effective than
warfarin, dextran, or low-dose UFH plus dihydroergotamine in preventing
deep vein thrombosis (DVT) after total hip replacement or hip fracture
surgery, and compared favorably with LMWH in patients undergoing total
hip replacement surgery (Bergqvist et al., 1991; Gerhart et al., 1991; Hoek et
al., 1992; Leyvraz et al., 1992; Organon report, 1994; Gent et al., 1996). In
addition, there are also prospective controlled and uncontrolled studies, as
well as case reports, demonstrating the efficacy of danaparoid in the treatment
of DVT, acute thrombotic stroke, and disseminated intravascular coagula-
tion complicating promyelocytic leukemia, as well as DVT prophylaxis and
the prevention of fibrin deposition on the dialysis membrane during hemo-
dialysis (Henny et al., 1983; Nieuwenhuis and Sixma, 1986; Cade et al., 1987;
Biller et al., 1989; von Bonsdorff et al., 1990; Gallus et al., 1993; de Valk et al.,
1995).
B. Clinical Use of Danaparoid in Patients with HIT

Danaparoid has been used extensively to treat patients with HIT (Chong and
Magnani, 1992; Magnani, 1993, 1997). After the diagnosis of HIT and dis-
continuation of heparin administration, patients often require an alterna-
tive anticoagulant for any one of the following indications: (1) treatment of
a recent or new thrombosis; (2) prophylaxis of venous thromboembolism;
(3) anticoagulation for cardiopulmonary bypass (CPB) surgery or periph-
eral arterial surgery; (4) anticoagulation of hemodialysis or hemofiltration;
(5) cardiac catheterization or coronary angioplasty; or (6) maintenance of
intravascular catheter patency. The rationale for the use of danaparoid to

treat patients with HIT for these indications is that danaparoid is an effective
antithrombotic-anticoagulant agent, as shown by the results of controlled
trials (Skoutakis, 1997). Furthermore, it has a specific inhibitory effect on HIT
antibody-induced platelet aggregation (Chong et al., 1989). Additionally,
unlike LMWH, it has a low frequency of in vitro cross-reactivity with HIT
antibodies (Makhoul et al., 1986; Chong et al., 1989; Greinacher et al., 1992;
Kikta et al., 1993; Vun et al., 1996).
The largest clinical experience with the use of danaparoid in the
treatment of patients with HIT is in the compassionate-use (named patient)
program organized by the manufacturer (Organon NV, Oss, The Nether-
lands) (Magnani, 1993, 1997). From 1981 to 1997, over 750 patients were
treated under this program for the various foregoing indications (Ortel and
Chong, 1998). The duration of treatment ranged from 1 day to 3.5 years, and
the posttreatment follow-up was 3 months. Interim, updated reports of this
program have been published (Chong and Magnani, 1992; Magnani, 1993,
1997). The overall success rate, defined as platelet count recovery without
new, progressive, or recurrent thrombosis during the danaparoid treatment
period or thrombotic death during 3 months follow-up, and in the absence of
any adverse effect necessitating treatment cessation, has been 91–94%, as
judged by the local physician-investigators. However, as this definition does
not include nonthrombotic death, the overall mortality observed in the pro-
gram was 18%, including both the treatment and a 3-month posttreatment
follow-up period. Most patients in this program received danaparoid for the
treatment of acute thromboembolism, often in the setting of severe illness,
such as renal or multisystem organ failure. Besides this compassionate-release
program, other studies supporting the efficacy of danaparoid therapy for
acute HIT include a randomized controlled trial comparing danaparoid with
dextran (Chong et al., 2001), as well as a retrospective analysis comparing
danaparoid and lepirudin (Farner et al., 2001), among other studies.
Treatment of Venous and Arterial Thromboembolism

Patients with HIT frequently have one or more acute thromboses, which may
have occurred before the development of HIT, as a complication of HIT itself,
or both (Warkentin and Kelton, 1996). Venous thrombosis complicates HIT
more often than does arterial occlusion. Indeed, in the compassionate-use
program, the ratio of venous to arterial thrombosis was 2:1, with some
patients having both types of thrombosis (Ortel and Chong, 1998). In HIT
patients, nevertheless, after cessation of UFH administration, the acute
thrombosis requires continuation of antithrombotic therapy. Currently,
either danaparoid, r-hirudin, or argatroban is believed to be effective in this
situation (Warkentin et al., 1998). These agents have in common the capacity
to inhibit thrombin generation, either by inhibition of factor Xa (danaparoid)
or by direct inhibition of thrombin (r-hirudin, argatroban).
In the compassionate-use program, it was recommended that HIT
patients with acute thrombosis receive intravenous danaparoid administered
as a bolus of 2500 U, followed by an infusion of danaparoid at 400 U/h for 4 h,
followed by 300 U/h for 4 h, and then 150–200 U/h for at least 5 days, aiming
for a plasma anti-Xa level of 0.5–0.8 anti-Xa U/mL. Table 1 describes a
similar protocol that takes into account the amount of danaparoid per
marketed ampule (750 anti-Xa U/ampule), as well as certain initial bolus
dose adjustments based on body weight. Danaparoid is also effective when
administered subcutaneously (de Valk et al., 1995): in this situation, the
equivalent 24-h actual or estimated intravenous dose is given in two to three
divided doses by subcutaneous injection over a 24-h period. For example,
2250 U (3 ampules) every 12 h by subcutaneous injection is approximately
equal to 190 U/h by intravenous infusion given over 24 h. In the compas-
sionate-use program, 464 patients with acute thromboembolism were treated
with danaparoid, with efficacy judged to be over 90% (Ortel and Chong,
1998).
Danaparoid treatment for HIT patients also proved efficacious in a
prospective, randomized, controlled clinical study (Chong et al., 2001). In this
trial, HIT patients with an acute thrombosis (venous, arterial, or both) were
randomized to receive either danaparoid plus warfarin, or dextran 70 plus
warfarin. Dextran is a glucose polymer, with an average molecular mass of
70 kDa. It is a weak antithrombotic agent that has been used to prevent DVT
in postoperative patients (Aberg and Rausing, 1978; Bergqvist, 1980). It is
known to block HIT antibody-induced platelet aggregation in vitro (Sobel et
al., 1986), and it has been suggested as a potentially useful drug for the
treatment of HIT. The reason for its use in the control group was that dextran
70 was the only rapidly acting antithrombotic drug available for the treatment
of HIT-associated thrombosis in Australia at study commencement in 1988.
The danaparoid treatment regimen was slightly different from that of
the compassionate-use program. Danaparoid was given as a bolus of 2400 U,
followed by an infusion of 400 U/h for 2 h, 300 U/h for 2 h, and then 200 U/h
for 5 days. In the dextran 70 arm, patients received dextran, 1 L on day 1, and
then 500 mL/day from days 2 to 5. In both treatment arms, the patients also
received warfarin, with doses adjusted to an INR of 2–4; the warfarin was
continued for 3 months. Patients were also stratified at randomization,
depending on the severity of their thrombosis, using predefined criteria.
Resolution of thrombocytopenia showed a nonsignificant trend in fa-
vor of danaparoid over dextran 70. Among the patients stratified as having
‘‘mild’’ thrombosis, a slightly higher percentage of thromboembolic events
Table 1 Danaparoid Dosing Schedules in HIT Patients

Clinical indication Danaparoid dosing schedule
Venous thromboembolism
Prophylaxis (prior HIT) 750 U sc, b.i.d. or t.i.d.
Prophylaxis (acute HIT) Treatment doses (see below) may be appropriate for prophylaxis
of acute HIT (see pp. 16, 348–349, 379)
Venous or arterial thrombo- 2250 U iv bolusa followed by 400 U/h for 4 h, 300 U/h for 4 h,
embolism: Treatment then 150–200 U/h for z5 days, aiming for a plasma anti-Xa
(either prior or acute HIT) level of 0.5–0.8 U/mL
Subcutaneous administration schedule: 1500–2250 U sc b.i.d. (given
almost 100% bioavailability, 2250 U sc b.i.d. is approximately
equal to an iv infusion rate of 200 U/h)
Embolectomy or other Preoperative: 2250 U iv bolusa; intraoperative flushes: 750 U in
peripheral vascular surgery 250 mL saline, using up to 50 mL (see p. 355); postoperative:
750 U sc t.i.d. (low-risk patients) or 150–200 U/h (high-risk
patients) beginning at least 6 h after surgery
Hemodialysis (on alternate 3750 U iv before 1st and 2nd dialyses; 3000 U for 3rd dialysis;
days) then 2250 U for subsequent dialyses, aiming for plasma anti-Xa
level of <0.3* U/mL predialysis, and 0.5–0.8 U/mL during
dialysis (see also Chap. 18).
Hemofiltration 2250 U iv bolus, followed by 600 U/h for 4 h, then 400 U/h for
4 h, then 200–400 U/h aiming for a plasma anti-Xa level of
0.5–1.0 U/mL (see also Chap. 18).
Cardiopulmonary bypass 125 U/kg iv bolus after thoracotomy; 3 U/mL in priming fluid of
surgery (CPB) apparatus; 7 U/kg/h iv infusion commencing after CPB hookup,
and continued until 45 min before expectation of stopping CPB
(see also Chap. 19)
Cardiac catheterization Preprocedure: 2250 U iv bolus (3000 U if 75–90 kg and 3750 U
if >90 kg)
Percutaneous transluminal Preprocedure: bolus as per foregoing; postprocedure: 150–200 U/h
coronary angioplasty for 1–2 days after PTCA (or until removal of balloon pump)
(PTCA) or intra-aortic
balloon pump
Catheter patency 750 U in 50 mL saline, then 5–10 mL per port, or as required
Pediatric dosage considerations Prophylaxis: 10 U/kg sc b.i.d.
Treatment: 30 U/kg b.w., iv bolus, then 1.2–2.0 U/kg b.w./h
depending upon severity of thrombosis
Abbreviations: b.w., body weight; b.i.d., twice daily; iv, intravenous: t.i.d., three times daily.
Compatibility with intravenous solutions: Danaparoid is compatible for dilution with the following
solutions: saline, dextose, dextrose–saline, Ringer’s, lactated Ringer’s, 10% mannitol.
Preparation of solution for infusion: One option is to add four ampules containing 3000 U (i.e., 750 anti-
Xa U/0.6 mL ampule) of danaparoid to 300 mL of intravenous solution (i.e., a solution that comprises 10
U danaparoid per milliliter of intravenous solution: thus, an infusion rate of 40 mL/h corresponds to a
dose of 400 U/h: 20 mL/h to a dose of 200 U/h, and so on.
a
Adjust iv danaparoid bolus for body weight:<60 kg, 1500 U; 60–75 kg, 2250 U; 75–90 kg, 3000 U; >90
kg, 3750 U.
treated with danaparoid (83%) improved compared with those who received
dextran 70 (73%). In contrast, a substantial and significant difference in
treatment outcome occurred in patients with ‘‘serious’’ thrombosis: 88% of
danaparoid-treated thromboembolic events recovered, compared with 44%
of those treated with dextran 70. These data suggest that the use of an effective
anticoagulant to treat thromboembolism associated with HIT is particularly
important in those with more severe disease.
Recently, treatment outcomes in patients with HIT who received either
danaparoid or lepirudin have been compared (Farner et al., 2001). Although
not a randomized trial, this study had important strengths. First, all patients
had serologically confirmed HIT, with about 70% having thrombosis at the
time of study entry. Second, all patients met the same inclusion and exclu-
sion criteria, had similar baseline characteristics, and were treated during the
same time period (25 months ending April 1996). Third, many patients were
studied (danaparoid, n = 126; lepirudin, n = 175). Furthermore, patients
were subdivided into those treated with prophylactic or therapeutic doses.
This study suggests that both danaparoid and lepirudin have similar
efficacy for treatment of HIT-associated thrombosis when given in therapeutic
doses: the day 42 success rate was about 80% for either agent, when failure
was defined as the cumulative event rate of a composite endpoint of new
thrombosis, death, or limb loss. When evaluating the single endpoint of new
thrombosis in those patients who received therapeutic doses of study drug,
danaparoid and lepirudin also showed similar efficacy (90.6 vs. 92.1%;
p = 0.74). Moreover, safety analysis of all patients (regardless of dose
received) showed significantly fewer major bleeds with danaparoid (2.5 vs.
10.4%; p = 0.009). These data suggest that the favorable therapeutic index of
danaparoid extends even to patients with a serious prothrombotic disorder
such as acute HIT complicated by thrombosis.
Overlapping Oral Anticoagulants with Danaparoid

In the treatment of HIT patients, it is important to inhibit thrombin
generation adequately with danaparoid until the acute thrombosis is well
controlled. Furthermore, it generally takes at least 5 days of warfarin therapy
before therapeutic functional hypoprothrombinemia is achieved (Harrison
et al., 1997). Thus, even when oral anticoagulant therapy is begun soon after
starting danaparoid, at least 5 days of danaparoid therapy are usually
required. Many danaparoid-treated HIT patients also receive overlapping
warfarin treatment, because oral anticoagulants are usually preferred when
at least 3–6 months of further anticoagulation is indicated because of venous
or arterial thromboembolism. Warfarin administration can be started safely
together with danaparoid in patients who do not have severe HIT-associated
thrombosis. However, in HIT patients with severe or extensive thrombosis, it
is prudent to delay administration of warfarin until the thrombotic process is

controlled and substantial resolution of the thrombocytopenia has occurred.
This caveat is based on the observation that warfarin may aggravate the
thrombotic process during the first few days of its administration, by reducing
levels of the natural anticoagulant protein C, particularly when there is
uncontrolled thrombin generation (Warkentin, 1996a; Warkentin et al.,
1997; Pötzsch et al., 1996) (see Chaps. 3 and 12). Warfarin does not neutralize
activated coagulation factors, and thus sufficient time must pass before its
antithrombotic effects are achieved through reduction in the vitamin K–
dependent procoagulant factors, particularly prothrombin. Danaparoid
therapy usually is discontinued when the INR is at, or near, the target ther-
apeutic range, and provided the acute thrombosis appears controlled on clin-
ical grounds. Danaparoid does not interfere with INR measurements during
oral anticoagulant therapy.
Prophylaxis During Acute HIT

In contrast to the comparable efficacy of danaparoid and lepirudin when
used in therapeutic doses to treat HIT-associated thrombosis (discussed
previously), lepirudin appeared somewhat more effective than danaparoid
when prophylactic-dose regimens were compared for preventing the single
endpoint of new thrombosis (91.4 vs. 81.4%; p = 0.138); this difference was
larger, and reached statistical significance, when the composite endpoint (new
thrombosis, limb amputation, or death) was examined (Fig. 2). However,
superior efficacy of lepirudin came at a price: for patients without thrombosis
at baseline (most of whom thus received prophylactic-dose therapy), lepirudin
was associated with a trend to more major bleeding events than danaparoid
(16.3 vs. 2.9%; p = 0.075) (Farner et al., 2001). Kodityal and colleagues
(2003) reported 5 patients who developed new thromboses while receiving
relatively low doses of danaparoid (usually, 1250 U every 12 hours by
subcutaneous injection).
These data support current treatment recommendations (Warkentin,
2001) (see Chaps. 1 and 13) that therapeutic doses of danaparoid are appro-
priate for most patients with HIT, whether they have HIT-associated throm-
bosis or just isolated HIT. A comparable situation exists with the two
thrombin inhibitors available to treat HIT: argatroban is approved in the
United States in therapeutic doses, both for HIT-associated thrombosis and
isolated HIT (see Chap. 16). And, although there exists a lower-dose
(prophylactic) regimen for lepirudin to manage isolated HIT (see Chap. 15),
the use of anticoagulant monitoring to adjust the infusion rate means that
most patients eventually receive doses that approach the therapeutic regimen
(Warkentin, 2001).
Figure 2 Comparison of the outcomes of HIT patients, with and without throm-
boembolic complications (TEC), before the start of alternative anticoagulation; dana-
paroid vs. lepirudin: Time-to-event analysis of the incidences of a combined endpoint
(new thromboembolic events, limb amputation, death; maximum, one endpoint per
patient) up to day 42. Among patients without TEC at baseline (most of whom were
treated with a prophylactic-dose regimen), there was a significantly higher incidence of
the combined endpoint among patients treated with danaparoid, compared with lepi-
rudin ( p = 0.02, long-rank test). This suggests that the approved, prophylactic-dose
regimen for danaparoid (750 U b.i.d. or t.i.d. by subcutaneous injection, without
anticoagulant monitoring) may be relatively less effective for managing patients with
isolated HIT, compared with the prophylactic-dose regimen for lepirudin (initial in-
travenous infusion rate, 0.10 mg/kg/h; subsequently, dose adjusted by aPTT). In
contrast, the combined endpoint did not differ significantly between danaparoid and
lepirudin for patients with TEC at baseline, suggesting that therapeutic (treatment)
doses of danaparoid (see Table 1) have similar efficacy as does therapeutic-dose
lepirudin (see Table 2 in Chap. 15). Also shown is the number of patients at risk on the
starting day and at subsequent 7-day intervals. (From Farner et al., 2001.)
Prophylaxis of Venous Thromboembolism

Patients with a previous history of HIT may require an alternative anticoag-
ulant to prevent venous thromboembolism if they are in a high-risk situation,
such as following surgery. UFH cannot be used, particularly during the first 1
or 2 months after the onset of HIT when HIT antibodies still circulate.
Thereafter, although HIT antibodies are usually undetectable, and the risk of
HIT is possibly relatively low (Warkentin and Kelton, 2001), most physicians
are understandably reluctant to readminister heparin in this situation.
Danaparoid is an effective and convenient drug for the prevention of
venous thromboembolism in patients with prior HIT. In the compassionate-
use program, 390 patients received danaparoid, 750 U by subcutaneous
injection, usually twice daily for DVT prophylaxis for many postoperative
settings, including following general, gynecological, neurological, cancer, and
organ transplant surgery. A high rate of success was observed (Magnani,
1997; Ortel and Chong, 1998).
Prophylaxis of Arterial Thromboembolism

Danaparoid has been used to prevent arterial thromboembolism in patients
undergoing various vascular operations, including peripheral artery bypass
graft surgery, embolectomy, and endarterectomy. In these patients, it was
given as a preoperative intravenous bolus of 2500 U, and in some it was also
administered postoperatively. Given as an intravenous bolus of 2500 U
immediately before the procedure, danaparoid has also been used to provide
antithrombotic cover for percutaneous coronary angioplasty, with or without
stenting, and for insertion of inferior vena cava filters and intra-aortic balloon
devices.
Anticoagulation for Cardiopulmonary Bypass Surgery

Patients with acute HIT, or recent previous HIT with persisting HIT anti-
bodies, may need to undergo cardiac surgery. UFH is contraindicated for these
patients, necessitating an alternative anticoagulant for use during cardiopul-
monary bypass surgery (CPB). After successful experiments in dogs (Henny
et al., 1985a), danaparoid has been used since 1985 for CPB in these patients
(Magnani, 1993; Wilhelm et al., 1996; Christiansen et al., 1998; Fernandes
et al., 2000; Olin et al., 2000). A report summarizes the experience of 47 eval-
uable patients who underwent CPB using danaparoid (Magnani et al., 1997).
The initial recommended dosing schedule consisted of 8750 U by
intravenous bolus postthoracotomy (with the bolus dose changed to 5,000
and 10,000 U for patients weighing <60 and >90 kg, respectively) plus 7500 U
danaparoid given in the priming fluid of the CPB machine. Further intra-
operative booster injections were recommended up to an hourly basis, if

necessary, because of fibrin clot formation, but not closer than 45 minutes
to the expected end of CPB. In the 47 reported patients who received dana-
paroid for CPB under the compassionate-use program, the recommended
dosing schedule was not strictly followed in all patients. Some received in-
appropriate or wrongly timed booster doses (too near to surgical closure) and
others received substantially higher doses in erroneous attempts to obtain the
same activated-clotting time (ACT) prolongation as with UFH, even though
anticoagulant-effective danaparoid doses do not prolong the ACT (Gitlin et
al., 1998).
Despite suboptimal dosing in some patients, cardiac surgery neverthe-
less could be successfully completed in 45 of the 47 patients (Magnani et al.,
1997). In two patients, the operations had to be abandoned because of
formation of ‘‘clots’’ in the operative field or in the CPB circuit. In another
16 patients, intraoperative clots were seen, but dissolved on booster doses of
danaparoid. No patients in this series had detectable in vitro cross-reactivity;
thus, intraoperative clot formation is unlikely to be attributable to a mani-
festation of in vivo cross-reactivity. Intraoperative clot formation did not
result in clinically evident thrombosis in any of the patients.
Increased postoperative bleeding was a significant problem with the use
of danaparoid for CPB (Magnani et al., 1997). Eleven (22%) of the patients
suffered severe postoperative blood loss, and another 33% had mild-to-
moderate bleeding. Possible reasons for the excessive bleeding included (1)
excessive danaparoid dosing owing to ad hoc protocol modifications; (2)
administration of danaparoid booster doses too close to the end of surgery;
(3) use of blood salvage techniques that returned anticoagulant-containing
blood to the patient; and (4) the long plasma half-life of danaparoid and lack
of an antidote to neutralize its anticoagulant effect (Skoutakis, 1997). It is also
theoretically possible that suboptimal anticoagulation during CPB would
paradoxically be associated with greater postoperative bleeding, as greater in
vivo thrombin generation during CPB might lead to greater postoperative
blood loss from secondary hyperfibrinolysis. This issue is raised because the
empirically modified protocol subsequently described may, therefore, not
necessarily be associated with reduction in bleeding outcomes.
On review of this experience, it was hypothesized that unnecessarily
high drug doses contributed to postoperative bleeding. Supporting evidence
included the observation that no severe bleeding was observed in patients who
received a total dose of danaparoid of 16,250 U or less (i.e., V250 U/kg)
(Magnani et al., 1997). Therefore, a new dosing regimen (see Table 1) was
developed that delivers no more than 232 U/kg of the drug. Continuous
intraoperative danaparoid infusion is also recommended, which may reduce
the need for a further drug bolus shortly before wound closure, as well as
provide therapeutic drug levels throughout CPB.
Disappointingly, the new regimen does not appear to have reduced the
frequency of severe postoperative bleeding. This new dosing regimen was
used by Olin and coworkers (2000) to manage five patients with acute or prior
HIT for CPB: four of the patients experienced prolonged bleeding requiring
reexploration and extensive transfusions, and two patients had clots identi-
fied in the surgical field. Other investigators later amended the protocol to
continue danaparoid until completion of CPB (rather than stopping the
anticoagulant 45 min before the expected end of pump run) because of
problems with clotting in the CPB circuit or in the operative field (Fernandes
et al., 2000).
Advances in surgical method may permit other treatment approaches
in selected patients. For example, the off-pump (‘‘beating heart’’) technique
does not utilize CPB, and thus a far lower dose of danaparoid may be feasible
for intraoperative anticoagulation. This approach was used successfully to
perform multiple coronary artery bypass grafting in a patient with acute HIT
and unstable angina (Warkentin et al., 2001). A relatively low target plasma
antifactor Xa level (0.6 U/mL) was used, rather than the conventional level
(>1.5 U/mL) sought during CPB (see Chap. 19).
A randomized, double-blind comparison of danaparoid (n = 34) with
heparin (n = 37) for off-pump coronary artery bypass grafting in non-HIT
patients showed a nonsignificant trend to greater postoperative blood loss
(mean, 264 mL) but a significant increase in patients exposed to homologous
blood (53% vs. 27%) with danaparoid. However, clinical outcomes appeared
similar, and the authors concluded that danaparoid could be a valuable
option in patients undergoing off-pump surgery when heparin is contra-
indicated (Carrier et al., 2002).
Other approaches for managing CPB or off-pump surgery in patients
with acute or previous HIT are discussed further in Chaps. 13 and 19.
Hemodialysis and Hemofiltration

Danaparoid has been used to anticoagulate patients with HIT requiring
hemodialysis or hemofiltration, in one of several clinical settings (Henny et
al., 1983, 1985b). First, patients with chronic or acute renal failure undergoing
regular hemodialysis with UFH occasionally develop HIT, thus necessitating
use of an alternative anticoagulant for subsequent dialyses. Second, very ill
patients in intensive care settings who develop HIT not uncommonly have a
need for anticoagulation during hemodialysis or hemofiltration (Wester et al.,
2000; Lindhoff-Last et al., 2001). Third, danaparoid is useful for patients who
experience difficulty in undergoing hemodialysis or hemofiltration with UFH
because of repeated deposition of fibrin on the dialysis or hemofiltration
membranes (Burgess and Chong, 1997). A change from UFH to danaparoid
often allows continuation of hemodialysis or hemofiltration without further
incident. This problem, which may be secondary to UFH-induced platelet

aggregation and microthrombus formation, despite absence of HIT anti-
bodies, may be a manifestation of nonimmune heparin-associated thrombo-
cytopenia, although significant thrombocytopenia usually does not occur
(Burgess and Chong, 1997).
Because danaparoid is cleared renally, the drug accumulates in the
blood of patients with renal failure undergoing hemodialysis or hemofiltra-
tion with the heparinoid. A reduction in the dose of danaparoid generally is
necessary for the second or third procedure after 2 or 3 days. For the first
hemodialysis, an intravenous bolus of 3750 U (2250 U if the body weight is
<55 kg) is usually administered. A plasma anti-Xa level should be per-
formed before the second and subsequent dialysis. The drug dose should
be appropriately reduced (usually to 2250 or 3000 U), depending on
whether hemodialysis is performed daily or every second day. The aim is
to maintain a plasma anti-Xa level between 0.5 and 0.8 U/mL. For hemo-
filtration, the dosing regimen is similar to the intravenous infusion regimen
used for the treatment of venous thrombosis (see Table 1). To avoid
excessive accumulation of the drug in the blood, the infusion rate may be
reduced on the second or subsequent day, according to the plasma anti-Xa
levels.
Reviews of danaparoid use in heparin-intolerant patients have identi-
fied 228 cases of HIT patients with acute or chronic renal failure (Magnani,
in preparation; Gallus and Magnani, in preparation). Clinical outcome data
show that danaparoid provides efficacious and safe anticoagulation, having
been used for up to 4 years for routine hemodialysis (three times per week),
and has also been found to extend the lifespan of the hemofilters (van Eps
et al., 2000; Lindhoff-Last et al., 2001). Nevertheless, these patients had a
relatively high mortality (26.3%) associated with severe comorbidity (the
mortality was 51.2% in the subset of 77 patients requiring hemofiltration,
usually in the setting of multiple organ failure).
Use in Children and Pregnant Women

Danaparoid has been used in only a very small number of pediatric and
pregnant patients (Gill and Kovacs, 1997; Ranze et al., 1999, 2001; Saxon
et al., 1999; Macchi et al., 2000). Its major use in these patients has been for the
treatment of HIT. Sixteen children were treated with danaparoid for various
indications, including maintenance of catheter patency, renal dialysis, cardiac
surgery, and thrombosis. The treatment was successful in all except one
patient in whom DVT occurred during danaparoid use associated with
formation of HIT antibodies. In general, it was noted that children, partic-
ularly infants, often required higher doses of danaparoid than adults on a
weight-adjusted basis.
Danaparoid is known to have been used in 31 pregnancies in women

with HIT. All had a history of thrombosis, either acute or during previous
pregnancy. Danaparoid was begun mainly in the first or third trimesters, and
continued for up to 34 weeks. In 19 pregnancies no significant problems
occurred. In the remaining 12, new thrombosis occurred in 6 (successfully
treated by increasing danaparoid dosing in 3); one of the other 3 patients with
new thrombosis developed HIT antibodies. Two maternal major bleeds
occurred, one due to abruptio placenta (fatal bleeding in a Jehovah’s witness
refusing blood transfusion) and the other due to placenta previa (fatal
cardiopulmonary complications). In another patient danaparoid was stopped
because of in vitro and in vivo evidence of danaparoid cross-reactivity. Two
fetal and one infant death occurred in mothers with systemic lupus ery-
thematosus or antiphospholipid antibody syndrome. One of these was further
complicated by onset of HELLP syndrome (hemolysis, elevated liver en-
zymes, low platelets), which began 6 weeks into treatment with danaparoid.
In 4 infants, testing of cord plasma revealed no anti-Xa activity. In four
breast milk samples obtained from mothers continuing danaparoid postpar-
tum, anti-Xa activity was virtually undetectable. Thus, the main antithrom-
botic subfraction of danaparoid does not cross the placenta, and the tiny
amounts which appear in the breast milk are probably hydrolyzed in the
infant’s stomach (Lindhoff-Last and Bauersachs, 2002). No increase in
postpartum bleeding in danaparoid-treated pregnant mothers was reported.
The doses of danaparoid used in these pregnant patients were the same as
those used in nonpregnant women for the same clinical situations.
C. Laboratory Monitoring
Measurement of plasma anti-Xa levels using an amidolytic assay can be used
for the laboratory monitoring of danaparoid’s anticoagulant action. The
heparinoid does not significantly prolong the activated partial thromboplas-
tin time (APTT), prothrombin time (INR), or activated-clotting time (ACT),
except at very high doses. Hence, these assays cannot be used for laboratory
monitoring of danaparoid. However, monitoring is not required in many
clinical situations. The drug has a bioavailability of almost 100% after
subcutaneous administration, and because of its lack of plasma protein
interaction, predictable plasma levels are usually obtained with subcutaneous
or intravenous use. However, laboratory monitoring is recommended in the
following clinical settings: (1) patients with substantial renal impairment; (2)
patients with unusually low or high body weight; (3) patients with life- or
limb-threatening thrombosis; (4) patients with unexpected bleeding; and (5)
critically ill or unstable patients.
It must be emphasized that for any assay of danaparoid-associated
plasma anti-Xa activity, the standard calibration curve must be constructed
using danaparoid, and not UFH or even LMWH (Laposata et al., 1998).
Because of its lack of interaction with plasma heparin-binding proteins,
danaparoid gives a dose-response relation different from these heparins,
and plasma anti-Xa levels during danaparoid treatment will be overestimated
if a LMWH standard curve is used. Indeed, there are differences in the stated
therapeutic range among these various glycosaminoglycan anticoagulants
(UFH, 0.2–0.4 U/mL by protamine titration; UFH, 0.3–0.7 anti-Xa U/mL;
LMWH, 0.6–1.0 U/mL; danaparoid, 0.5–0.8 anti-Xa U/mL) (Hirsh et al.,
1998; Laposata et al., 1998; Warkentin et al., 1998). In some clinical treatment
settings using danaparoid, it might be advisable to aim for a lower anti-Xa
level (e.g., about 0.3 U/mL for a patient judged to have a high risk of bleed-
ing); sometimes, a higher target anti-Xa level should be sought (e.g., about 1.0
U/mL for a patient with life- or limb-threatening venous or arterial throm-
bosis).
Anti-Xa levels are determined using a chromogenic assay (i.e., a method
similar to that performed for monitoring LMWH treatment). A standard
reference curve must be constructed using various dilutions of danaparoid
(e.g., 1.6, 1.0, 0.5, 0.3, and 0 U/mL danaparoid, diluted in pooled normal
platelet-poor plasma). Control plasma samples are prepared by adding
known quantities of danaparoid to normal pooled plasma aliquots (assuming
100% recovery of the known quantity of danaparoid added) in three different
concentrations approximating treatment situations (e.g., 0.2, 0.7, and 1.25
U/mL, corresponding to low-, mid-, and high-control danaparoid levels).
Aliquots stored at 70jC are stable indefinitely if used only once, without
refreezing and rethawing.
D. Cross-Reactivity of HIT Antibodies with Danaparoid

As danaparoid consists of a mixture of glycosaminoglycans (mainly heparan
sulfate), it is not surprising that a small percentage of antibodies from HIT
patients do cross-react with the drug. The cross-reactivity rate is low (mean,
9.6%) if aggregation studies using citrated platelet-rich plasma are performed
(Makhoul et al., 1986; Kikta et al., 1993; Ramakrishna et al., 1995; Vun et al.,
1996), but higher if more sensitive washed platelet activation (Warkentin,
1996b; Koster et al., 2000) or a fluid-phase antigen assay (Newman et al.,
1998) is used (see Chap. 11). Because activation assays are dependent on the
donor platelets used for testing, we have used highly reactive donor platelets
under standardized conditions using platelet aggregation, to investigate the
cross-reactivity of HIT antibodies for danaparoid and LMWH. We found
cross-reactivity rates of 7% with danaparoid and 83–89% with LMWH (Vun
et al., 1996). However, with the sensitive fluid-phase antigen assay, we
observed a higher cross-reactivity rate of 50% with danaparoid, and again,
a much higher rate (88%) with LMWH (Newman et al., 1998) (Fig. 3).
Importantly, even when in vitro reactivity with danaparoid is observed, it is

generally weak and quantitatively less than seen with LMWH.
The in vitro cross-reactivity of the HIT antibodies with danaparoid does
not appear to be clinically significant. We have investigated the clinical
significance of in vitro cross-reactivity in 21 patients treated with the LMW
heparinoid (Newman et al., 1998). The 8 patients who tested positive with
danaparoid by the fluid-phase enzyme immunoassay, but negative by the
[14C]serotonin-release washed platelet assay, recovered with resolution of
their thrombocytopenia and thrombosis, in a fashion similar to the 11
Figure 3 Cross-reactivity of HIT-IgG antibodies with PF4 complexed to heparin-

like anticoagulants: The fluid-phase enzyme-linked immunoassay was used to assess
the degree to which IgG present in HIT sera or plasma bound to PF4 alone, PF4–
heparin, PF4–dalteparin (Fragmin), PF4–enoxaparin (Clexane), or PF4–danaparoid
(Orgaran). The positive cutoff (dashed line) is 3 standard deviations above the mean
(log transformed) absorbance of the normal samples (triangles) using PF4–heparin.
The binding of normal antibodies is indicated by triangles. Circles indicate HIT
samples that have been positive (closed circle), negative (open circle), or not tested
(speckled circle) in a functional assay with the corresponding drug. (From Newman
et al., 1998.)
patients who did not manifest in vitro cross-reactivity to danaparoid by both

assays. Two patients tested positive by both activation and antigen assays: in
1 patient, both thrombocytopenia and pulmonary embolism resolved during
danaparoid treatment. However, in the other patient, thrombocytopenia and
extensive thrombosis persisted despite danaparoid therapy. It is unclear
whether this unusual patient course represented a specific danaparoid treat-
ment failure, as the patient’s subsequent clinical course was characterized by
consistent failure for all of the antithrombotic therapies used (Fig. 4).
Warkentin (1996b) also evaluated the clinical significance of in vitro
cross-reactivity with danaparoid in 29 HIT patients who had been treated
with the heparinoid for HIT. This investigator found no difference in clinical
outcomes, or in the time to platelet count recovery, between the two patient
groups. However, there are isolated anecdotal reports of unfavorable clinical
outcomes in HIT patients treated with danaparoid (Tandy-Poncet et al., 1995;
Insler et al., 1997; Muhm et al., 1997). Cross-reactive antibodies were not
always investigated in these studies. It remains possible that unrelated clinical
factors might have caused the fall in the platelet count in these patients.
Figure 4 Serial platelet counts of representative HIT patients treated with

danaparoid: Solid black bar shows duration of heparin administration, striped bar
indicates danaparoid therapy, open bar shows warfarin therapy. (a) Typical profile
of the 11 patients who were negative in both fluid-phase EIA and functional assay.
(b) Typical profile of the eight patients who were positive in the fluid-phase, but
negative in a functional assay. (c) Profile of patient A, one of the two patients who
were positive in both types of assay. She recovered during a short course of
danaparoid. (d) Profile of patient B, the other patient positive in both assays. The
profile indicates the course of HIT following transfer to a major hospital. Despite
treatment with many antithrombotic agents, he eventually died following major
thrombosis. (From Newman et al., 1998.)
Additionally, it may not be possible to distinguish clinical cross-reactivity to

danaparoid with the natural course of a severe episode of HIT (Warkentin,
1998; Baumgärtel et al., 2000).
Overall, on the balance of current evidence, it would appear that in vitro
cross-reactivity with danaparoid is not associated with adverse clinical out-
comes in most patients. Accordingly, many physicians do not obtain cross-
reactivity test results before instituting danaparoid therapy. This approach
seems reasonable. It is based on several considerations: the overall low risk
of in vivo cross-reactivity, the low predictivity of in vitro cross-reactivity
testing, the lack of standardized test methods, and the potential for adverse
clinical events during withholding of treatment pending the results of cross-
reactivity testing. However, cross-reactivity testing should be performed in
patients who develop new, progressive, or recurrent thrombocytopenia or
thrombosis during treatment with danaparoid.
E. Adverse Effects
Severe bleeding, the most serious adverse effect of danaparoid, rarely occurs
except in patients who are treated with very high doses of the drug, or in those
who develop drug accumulation (renal failure), or who have additional hemo-
static or vascular defects. However, serious bleeding occurred in a significant
number of patients who had undergone CPB with danaparoid (Magnani et al.,
1997; Westphal et al., 1997, Fernandes et al., 2000; Olin et al., 2000). In con-
trast, bleeding was not seen in the randomized trial in which HIT patients
with venous or arterial thromboses received danaparoid (Chong, 1996). Com-
pared with CPB patients, these patients underwent less intense anticoagula-
tion and did not suffer from the additional insults of CPB and chest incision.
Skin hypersensitivity reactions have been reported with danaparoid, but
these are rare (Magnani, 1993). Osteoporosis (an important complication of
prolonged UFH treatment) was not detected in any danaparoid-treated pa-
tients in the compassionate-use program, even in those treated for more than
3 months. Despite the issue of in vitro cross-reactivity with danaparoid, it is
noteworthy that new-onset immune-mediated thrombocytopenia has never
been reported with this agent.
F. Availability of Danaparoid
Table 2 lists the countries in which danaparoid has been approved for the
treatment of HIT, either with or without associated thrombosis. In some
countries in which danaparoid is approved for DVT prophylaxis, physicians
have the legal option to prescribe danaparoid for HIT (i.e., for ‘‘off-label’’ use
in a nonapproved indication) (see Chaps. 19 and 20). Danaparoid is no longer
Table 2 Countries in Which Danaparoid Is Approved for Clinical Usea
Heparin-induced
DVT prophylaxis thrombocytopenia
Country Perioperativeb Poststroke Prophylaxisc Treatment
North America
Canada X X X X
United States Xd
Europe
Austria X X X X
Belgium X X X
Denmark X X X
Finland X X
France Xe X X
Germany X X
Great Britain X X
Greece X
Ireland X X X
Italy X
Luxembourg X X X
Netherlands Xf Xf X X
Norway X X
Portugal X X X
Sweden X X X X
Switzerland X X X
Australasia and Africa
Australia X X
Japang
Korea X
New Zealand X X X X
South Africa X X
a
Danaparoid is no longer marketed in some of these countries, e.g., United States, Great
Britain, Norway.
b
Orthopedic and general surgery only (unless otherwise indicated); approval includes starting
danaparoid 1–4 h preoperatively (except for U.S.).
c
Approved dose of 750 U b.i.d.–t.i.d. may be too low for acute HIT (see pp. 16, 348–349, 379).
d
Elective hip surgery only.
e
Orthopedic and cancer surgery only.
f
Approval modified to facilitate approval for HIT in Finland and Germany.
g
Approved only for treatment of DIC.
marketed in some countries (e.g., United States [since April 2002], United
Kingdom, Norway).
III. CONCLUSION
Danaparoid is a safe and effective anticoagulant for the prevention or

treatment of venous or arterial thrombosis in HIT patients, regardless of
the presence of antibody cross-reactivity with the drug. It is also efficacious as
an anticoagulant for hemodialysis-hemofiltration and cardiac surgery
employing CPB. With the use of danaparoid for the cardiac surgery, serious
postoperative bleeding can occur.
ACKNOWLEDGMENTS
Some of the studies described in this chapter were supported by a program

grant from the National Health and Medical Research Council of Australia.
REFERENCES
Aberg M, Rausing A. The effect of dextran 70 on the structure of ex vivo thrombi.

Thromb Res 1978; 12:1113–1122.
Baumgärtel MW, Eichler P, Glöckner WM, Ranze O, Greinacher A. Heparin-induced
thrombocytopenia (HIT): in vitro and in vivo cross-reactivity to danaparoid
sodium and successful treatment with recombinant hirudin (lepirudin) [letter].
Eur J Haematol 2000; 65:148–149.
Bergqvist D. Prevention of postoperative deep vein thrombosis in Sweden. Results
of a survey. World J Surg 1980; 4:489–495.
Bergqvist D, Kettunen K, Fredin H, et al. Thromboprophylaxis in patients with hip
fractures: a prospective, randomized, comparative study between Org 10172
and dextran 70. Surgery 1991; 109:617–622.
Biller J, Massey EW, Marler JR, et al. A dose escalation study of Org 10172 (low
molecular weight heparinoid) in stroke. Neurology 1989; 39:262–265.
Bradbrook ID, Magnani HN, Moelker HCT, Morrison PJ, Robinson J, Rogers HJ,
Spector RG, Van Dinther T, Wijnand H. Org 10172: a low molecular weight
heparinoid anticoagulant with a long half-life in man. Br J Clin Pharmacol 1987;
23:667–675.
Cade JF, Wood M, Magnani HN, Westlake GW. Early clinical experience of a new
heparinoid, Org 10172, in prevention of deep venous thrombosis. Thromb Res
1987; 45:497–503.
Carrier M, Robitaille D, Perrault LP, Pellerin M, Page P, Cartier R, Bouchard D.

Heparin versus danapariod in off-pump coronary bypass grafting: results of a
prospective randomized clinical trial. J Thorac Cardiovasc Surg 2003; 125:325–
329.
Casu B. Structural features of chondroitin sulphate, dermatan sulphate and heparan
sulphate. Semin Thromb Haemost 1991; 17(suppl 1):9–14.
Chong BH. Heparin-induced thrombocytopenia. Br J Haematol 1995; 89:431–439.
Chong BH, Magnani HN. Orgaran in heparin-induced thrombocytopenia. Haemo-
stasis 1992; 22:85–91.
Chong BH, Ismail F, Cade J, Gallus AS, Gordon S, Chesterman CN. Heparin-induced
thrombocytopenia: studies with a new molecular weight heparinoid, Org 10172.
Blood 1989; 73:1592–1596.
Chong BH, Gallus AS, Cade JF, Magnani H, Manoharan H, Oldmeadow M, Arthur
C, Rickard K, Gallo J, Lloyd J, Seshadri P, Chesterman CN. Prospective ran-
domised open-label comparison of danaparoid and dextran 70 in the treatment
of heparin-induced thrombocytopenia and thrombosis. Thromb Haemost 2001;
86:1170–1175.
Christiansen S, Geiger A, Splittgerber FH, Reidemeister JC. Coronary artery bypass
grafting in a patient with type II heparin associated thrombopenia. Cardiovasc
Surg 1998; 6:90–93.
Danhof M, de Boer A, Magnani HN, Stiekema JCJ. Pharmacokinetic considerations
on Orgaran (Org 10172). Hemostasis 1992; 22:73–84.
de Valk HW, Banga JD, Wester JWJ, Brouwer CB, van Hessen MWJ, Meuwissen
OJAT, Hart HC, Sixma JJ, Nieuwenhuis HK. Comparing subcutaneous dana-
paroid with intravenous unfractionated heparin for the treatment of venous
thromboembolism. A randomized controlled trial. Ann Intern Med 1995; 123:
1–9.
Farner B, Eichler P, Kroll H, Geinacher A. A comparison of danaparoid and lepirudin
in heparin-induced thrombocytopenia. Thromb Haemost 2001; 85:950–957.
Fernandes P, Mayer R, MacDonald JL, Cleland A, Hay-McKay C. Use of danaparoid
sodium (Orgaran) as an alternative to heparin sodium during cardiopulmonary
bypass: a clinical evaluation of six cases. Perfusion 2000; 15:531–539.
Gallus A, Cade J, Ockelford P, Hepburn S, Maas M, Magnani H, Bucknall T, Stevens
J, Porteious F. Orgaran (Org 10172) or heparin for preventing venous throm-
bosis after elective surgery for malignant disease? A double-blind, randomised,
multicentre comparison. Thromb Haemost 1993; 70:562–567.
Gent M, Hirsh J, Ginsberg JS, Powers PJ, Levine MN, Geerts WH, Jay RM, Leclerc J,
Neemeh JA, Turpie AG. Low-molecular-weight heparinoid Orgaran is more
effective than aspirin in the prevention of venous thromboembolism after
surgery for hip fracture. Circulation 1996; 93:80–84.
Gerhart TN, Yett HS, Robertson LK, Lee MA, Smith M, Salzman EW. Low-
molecular-weight heparinoid compared with warfarin for prophylaxis of deep-
vein thrombosis in patients who are operated on for fracture of the hip. A
prospective, randomized trial. J Bone Joint Surg 1991; 73A:494–502.
Gill J, Kovacs MJ. Successful use of danaparoid in treatment of heparin-induced
during twin pregnancy. Obstet Gycenol 1997; 90:648–650.
Gitlin SD, Deeb GM, Yann C, Schmaier AH. Intraoperative monitoring of dana-
paroid sodium anticoagulation during cardiovascular operations. J Vasc Surg
1998; 27:568–575.
Gordon DL, Linhardt R, Adams HP. Low-molecular-weight heparins and heparin-
oids and their use in acute or progressing ischaemic stroke. Clin Neurophar-
macol 1990; 13:522–543.
Greinacher A, Michels I, Muller-Eckhardt C. Heparin-associated thrombocytopenia:
the antibody is not heparin specific. Thromb Haemost 1992; 67:545–549.
Harrison L, Johnston M, Massicotte MP, Crowther M, Moffat K, Hirsh J. Com-
parison of 5-mg and 10-mg loading doses in initiation of warfarin therapy. Ann
Intern Med 1997; 126:133–136.
Henny CP, ten Cate H, ten Cate JW, Surachno S, van Bronswijk H, Wilmink JM,
Ockelford PA. Use of a new heparinoid as anticoagulant during acute haemo-
dialysis of patients with bleeding complications. Lancet 1983; 1:890–893.
Henny CP, ten Cate H, ten Cate JW, Moulijn AC, Sie TH, Warren P, Buller HR. A
randomized blind study comparing standard heparin and a new low molecular
weight heparinoid in cardiopulmonary bypass surgery in dogs. J Lab Clin Med
1985a; 106:187–196.
Henny CP, ten Cate H, Surachno S, Stevens P, Buller HR, den Hartog M, ten Cate JW.
The effectiveness of a low molecular weight heparinoid in chronic intermittent
haemodialysis. Thromb Haemost 1985b; 54:460–462.
low-molecular-weight heparin. Mechanisms of action, pharmacokinetics, dosing
considerations, monitoring, efficacy, and safety. Chest 1998; 114:489S–510S.
Hoek JA, Nurmohamed MT, Hamelynck KJ, Marti RK, Knipscheer HC, ten Cate H,
Büller HR, Magnani HN, ten Cate JW. Prevention of deep vein thrombosis
following total hip replacement by low molecular weight heparinoid. Thromb
Haemost 1992; 67:28–32.
Insler SR, Kraenzler EJ, Bartholomew JR, Kottke-Marchant K, Lytle B, Starr NJ.
Thrombosis during the use of the heparinoid Organon 10172 in a patient with
heparin-induced thrombocytopenia. Anesthesiology 1997; 86:495–498.
Kikta MJ, Keller MP, Humphrey PW, Silver D. Can low molecular weight heparins
and heparinoids be safely given to patients with heparin-induced thrombocy-
topenia syndrome? Surgery 1993; 114:705–710.
Kodityal S, Manhas AH, Udden M, Rice L. Danaparoid for heparin-induced throm-
bocytopenia: an analysis of treatment failures. Eur J Haematol 2003; 71:1–5.
Koster A, Meyer O, Hausmann H, Kuppe H, Hetzer R, Mertzlufft F. In vitro cross-
reactivity of danaparoid sodium in patients with heparin-induced thrombocy-
topenia type II undergoing cardiovascular surgery. J Clin Anesth 2000; 12:324–
327.
Laposata M, Green D, Van Cott EM, Barrowcliffe TW, Goodnight SH, Sosolik RC.
College of American Pathologists Conference Therapy. The clinical use and
laboratory monitoring of low-molecular-weight heparin, danaparoid, hirudin

and related compounds, and argatroban. Arch Pathol Lab Med 1998; 122:799–
807.
Leyvraz P, Bachmann F, Bohnet J, et al. Thromboembolic prophylaxis in total hip
replacement: a comparison between the low molecular weight heparinoid Lo-
moparan and heparin-dihydroergotamine. Br J Surg 1992; 79:911–914.
Lindhoff-Last E, Bauersachs R. Heparin-induced thrombocytopenia—alternative
anticoagulation in pregnancy and lactation. Semin Thromb Hemost 2002; 28:
439–445.
Lindhoff-Last E, Betz C, Bauersachs R. Use of a low molecular weight heparinoid
(danaparoid sodium) for continuous renal replacement therapy in intensive
care unit patients. Clin Appl Thromb/Hemost 2001; 7:300–304.
Macchi L, Sarfati R, Guicheteau M, Chamlian V, Pourrat O, Gruel Y, Magnin G,
Brizard A, Boinot C. Thromboembolic prophylaxis with danaparoid (Orgaran)
in a high-thrombosis-risk pregnant woman with a history of heparin induced
thrombocytopenia (HIT) and Widal’s disease. Clin Appl Thromb Hemost 2000;
6:187–189.
Magnani HN. Orgaran (danaparoid sodium) use in the syndrome of heparin-induced
thrombocytopenia. Platelets 1997; 8:74–81.
Magnani HN, Beijering RJR, ten Cate JW, Chong BH. Orgaran anticoagulation for
cardiopulmonary bypass in patients with heparin-induced thrombocytopenia.
In: Pifarre R, ed. New Anticoagulants for the Cardiovascular Patient. Phil-
adelphia: Hanley & Belfus, 1997:487–500.
Makhoul RG, Greenberg CS, McCann RL. Heparin-induced thrombocytopenia and
thrombosis: a serious clinical problem and potential solution. J Vasc Surg 1986;
4:522–528.
Meuleman DG, Hobbelen PMJ, Van Dedem G, Moelker HCT. A novel anti-
thrombotic heparinoid (Org 10172) devoid of bleeding inducing capacity: a
survey of pharmacological properties in experimental animal models. Thromb
Res 1982; 27:353–363.
Meuleman DG. Synopsis of the anticoagulant and antithrombotic profile of the
low molecular weight heparinoid Org 10172 in experimental models. Thromb
Haemost 1987; 58:376–380.
Mikhailidis DP, Barradas MA, Mikhailidis AM, Magnani H, Dandona P.
Comparison of the effect of a conventional heparin and a low molecular weight
heparinoid on platelet function. Br J Clin Pharmacol 1984; 17:43–48.
Mikhailidis DP, Fonseca VA, Barradas MA, Jeremy JY, Dandona P. Platelet ac-
tivation following intravenous injection of a conventional heparin: absence of
effect with a low molecular weight heparinoid (Org 10172). Br J Clin Pharmacol
1987; 24:415–424.
Muhm M, Claeys L, Huk I, Koppensteiner R, Kyrle PA, Minar E, Stumpflen A,
Ehringer H, Polterauer P. Thromboembolic complications in a patient with
heparin-induced thrombocytopenia (HIT) showing cross-reactivity to a low

molecular weight heparin-treatment with Org 10172 (Lomoparan). Wien Klin
Wochenschr 1997; 109:128–131.
Newman PM, Swanson RL, Chong BH. IgG binding to PF4–heparin complexes in the
fluid phase and cross-reactivity with low molecular weight heparin and
heparinoid. Thromb Haemost 1998; 80:292–297.
Nieuwenhuis HK, Sixma JJ. Treatment of disseminated intravascular coagulation in
acute promyelocytic leukemia with low molecular weight heparinoid Org 10172.
Cancer 1986; 58:761–764.
Ofosu FA. Anticoagulant mechanisms of Orgaran (Org 10172) and its fraction with
high affinity to antithrombin III (Org 10849). Haemostasis 1992; 22:66–72.
Olin DA, Urdaneta F, Lobato EB. Use of danaparoid during cardiopulmonary bypass
in patients with heparin-induced thrombocytopenia. J Cardiothorac Vasc
Anesth 2000; 14:707–709.
Org 10172 hip joint replacement report, research protocol 004–023. West Orange, NJ:
Organon Inc., 1994.
Ortel TL, Chong BH. New treatment options for heparin-induced thrombocytopenia.
Semin Hematol 1998; 35(suppl 5):26–34.
Pötzsch B, Unrig C, Madlener K, Greinacher A, Müller-Berghaus G. APC resistance
and early onset of oral anticoagulation are high thrombotic risk factors in
patients with heparin-associated thrombocytopenia (HAT) [abstr]. Ann
Hematol 1996; 72(suppl 1):A6.
Ramakrishna R, Manoharan A, Kwan YL, Kyle PW. Heparin-induced thrombocy-
topenia: cross-reactivity between standard heparin, low molecular weight
heparin, dalteparin (Fragmin) and heparinoid (Orgaran). Br J Haematol 1995;
91:736–738.
Ranze O, Rakow A, Ranze P, Eichler P, Greinacher A, Fusch C. Low-dose dana-
paroid sodium catheter flushes in an intensive care infant suffering from hep-
arin-induced thrombocytopenia. Pediatr Crit Care Med, 2001; 2:175–177.
Saxon BR, Black MD, Edgell D, Noel D, Leaker MT. Pediatric heparin-induced
thrombocytopenia: management with danaparoid (Orgaran). Ann Thorac Surg
1999; 68:1076–1078.
Schmahl TE, Ganjoo AK, Harloff MG. Orgaran (Org 10172) for cardiopulmonary
bypass in heparin-induced thrombocytopenia: role of adjunctive plasmaphe-
resis. J Cardiothorac Vasc Anesth 1997; 11:262–263.
Skoutakis VA. Danaparoid in the prevention of thrombo-embolic complications. Ann
Pharmacother 1997; 31:876–887.
Sobel M, Adelman B, Greenfield LJ. Dextran 40 reduces heparin-mediated platelet
aggregation. J Surg Res 1986; 40:382–387.
Stiekema JC, Wijnand HP, van Dinther TG, Moelker HCT, Dawes J, Vinchenzo A,
Toeberich H. Safety and pharmacokinetics of the low molecular weight
heparinoid Org 10172 administered to healthy elderly volunteers. Br J Clin
Pharmacol 1989; 27:39–48.
Tardy/Poncet B, Mahul P, Beraud AM, Favre JP, Tardy B, Guyotat D. Failure of

Orgaran therapy in a patient with a previous heparin-induced thrombocyto-
penia. Br J Haematol 1995; 90:69–70.
Van Eps RS, Wester JPJ, de Ruiter FE, Girbes ARJ. Danaparoid sodium in
continuous veno-venous hemofiltration in patients with multiple organ
dysfunction syndrome [abstr]. Neth J Med 2000; 56:A40–A41.
Von Bonsdorff M, Stiekema J, Harjanne A, Alapiessa U. A new low molecular weight
heparinoid Org 10172 as anticoagulant in hemodialysis. Int J Artif Organs 1990;
13:103–108.
Vun CH, Evans S, Chong BH. Cross-reactivity study of low molecular weight heparins
and heparinoid in heparin-induced thrombocytopenia. Thromb Res 1996;
81:525–532.
Warkentin TE. Heparin-induced thrombocytopenia: IgG-mediated platelet activa-
tion, platelet microparticle generation, and altered procoagulant/anticoagulant
balance in the pathogenesis of thrombosis and venous limb gangrene compli-
cating heparin-induced thrombocytopenia. Transf Med Rev 1996a; 10:249–258.
reactivity of danaparoid for HIT-IgG [abstr]. Blood 1996b; 88:626a.
Warkentin TE. Limitation of conventional treatment options for heparin-induced
Am J Med 1996; 101:502–507.
N Engl J Med 2001; 344:1286–1292.
towards consensus. Thromb Haemost 1998; 79:1–7.
Warkentin TE, Dunn GL, Cybulsky IJ. Off-pump coronary artery bypass of grafting
for acute heparin-induced thrombocytopenia. Ann Thorac Surg, 2001; 72:1730–
1732.
Westphal K, Martens S, Strouhal U, Matheis G. Heparin-induced thrombocytopenia
type II: perioperative management using danaparoid in a coronary artery
bypass patient with renal failure. Thorac Cardiovasc Surg 1997; 45:318–320.
Wester JPJ, Stolk M, Geers ABM, Vincent HH, Haas FJLM, Biesma DH, Veth G,
Leusink JA, Wiltink HH. Danaparoid sodium in CAVHD in seriously ill
patients with heparin-induced thrombocytopenia [abstr]. Neth J Med 2000;
56:A44.
Wilhelm MJ, Schmid C, Kekecioglu D, Mollhoff T, Ostermann H, Scheld HH.
Cardiopulmonary bypass in patients with heparin-induced thrombocytopenia
using Org 10172. Ann Thorac Surg 1996; 61:920–924.
15
Lepirudin for the Treatment of
Andreas Greinacher
I. INTRODUCTION
With its approval by the European Medical Evaluation Agency (EMEA) in

1997 and by the U.S. Food and Drug Administration (FDA) in 1998 for the
treatment of heparin-induced thrombocytopenia (HIT) complicated by
thrombosis, lepirudin (Refludan, Schering AG, Berlin; licensed to Berlex
Laboratories, Montville, NJ [US, Canada], and to Pharmion, Cambridge,
UK [all other countries]) became the first direct thrombin inhibitor (DTI)
available for this indication. The scientific literature has documented experi-
ence with lepirudin in more than 8500 patients. The total number of patients
exposed to this anticoagulant (until 2003) is estimated at between 35,000 and
60,000 worldwide.
II. HIRUDIN AND ITS DERIVATIVES

A. Chemistry
Hirudin, the most potent natural thrombin inhibitor identified to date, is a 65-
amino-acid polypeptide (molecular mass, approximately 7 kDa) produced by
the parapharyngeal glands of the medicinal leech, Hirudo medicinalis. The
NH2 terminal part of the molecule (residues 1–39) is stabilized by three
disulfide bridges integral to its function. The COOH-terminal moiety (resi-
dues 40–65) is highly acidic. In the three-dimensional structure of hirudin
(Clore et al., 1987; Sukumaran et al., 1987), three areas are distinguished: a
397
398 Greinacher
central core (residues 3–30, 37–46, 56–57), a ‘‘finger’’ (residues 31–36), and a
loop (residues 47–55). Hirudin is very stable at extremes of pH (1.5–13.0) and
at high temperatures (>90jC). It is soluble in water, but insoluble in alcohol
or acetone. The isoelectric point of hirudin is approximately 4.
Hirudins for therapeutic use are now produced by recombinant bio-
technology, using the yeast Saccharomyces cerevisiae, yielding recombinant
hirudin (r-hirudin). Lepirudin, a desulfatohirudin, differs from natural hiru-
din by lacking the sulfate group at Tyr-63 and also has an NH2-terminal
leucine residue in place of the isoleucine. Although such structural differences
result in a 10-fold reduction in the dissociation constant of r-hirudin, as
compared with natural hirudin, r-hirudins remain highly selective inhibitors
of thrombin, with an inhibition constant for thrombin in the picomolar range
(Stone and Hofsteenge, 1986).
B. Pharmacology
Lepirudin acts independently of the cofactors antithrombin and heparin
cofactor II (Markwardt, 1992) and forms tight, noncovalent 1:1 complexes
with thrombin. Interacting with both binding sites, lepirudin is a bivalent
inhibitor of thrombin (cf. argatroban, a univalent DTI) (Fig. 1). Lepirudin
inhibits all the biological activities of thrombin.
Three amino acids (residues 46–48) near the NH2-terminus of hirudin
bind to the active site cleft on thrombin, while the core of the hirudin molecule
closes off the active site pocket of thrombin. The COOH-terminal tail of
hirudin interacts with the fibrinogen anion-binding site, helping to block
thrombin-catalyzed fibrinogen cleavage. Hirudin inhibits the feedback loop
whereby thrombin enhances its own generation via activation of factors Va
and VIIIa (Kaiser and Markwardt, 1986; Pieters et al., 1989). In addition to
inhibiting free thrombin, hirudin inhibits clot-bound thrombin (Hogg and
Jackson, 1989; Weitz et al., 1990) and thrombin bound to fibrin split products
(Weitz et al., 1998). In contrast, heparin-antithrombin complexes are unable
to access and inactivate clot-bound thrombin. This important difference
between hirudin and heparin might explain why hirudin is more effective
than heparin in dissolving mural thrombi in experimental models (Meyer
et al., 1998). Hirudin shows virtually no interaction with plasma proteins
(Glusa and Markwardt, 1990), and its activity is standardized in thrombin
inhibitory units (TIU): 1 TIU is the amount of hirudin inhibiting 1 U of
thrombin at 37jC. The specific activity of lepirudin is 16,000 TIU/mg.
C. Pharmacokinetics
Lepirudin is administered parenterally. Studies of plasma pharmacokinetics
in healthy subjects reveal a two-compartment model. The initial plasma half-
Lepirudin in HIT Treatment 399
Figure 1 Schematic representation of the thrombin molecule and its inhibition by

hirudin, bivalirudin (formerly, Hirulog), hirugen, and argatroban: (1) active-site
pocket; (2) fibrinogen-binding site. The active-site pocket catalyzes most of the
functions of the thrombin molecule, whereas the fibrinogen-binding exosite mediates
the binding of thrombin to fibrinogen. Hirudin is a 7000 Da (7 kDa) protein composed
of 65 amino acids, which binds to the active-site pocket and the fibrinogen-binding
exosite of thrombin, i.e., it is a bivalent direct thrombin inhibitor. Bivalirudin is a small
synthetic peptide (20 amino acids) designed also to block both of these sites on throm-
bin. Hirugen, a synthetic peptide, mimics the binding site of fibrinogen to thrombin,
thereby inhibiting binding of thrombin to fibrinogen and, therefore, fibrinogen
cleavage by thrombin. The arginine derivative argatroban binds competitively to only
the active binding site pocket of thrombin. Hirugen and argatroban are univalent direct
thrombin inhibitors. (Adapted from Hermann et al., 1997.)
life (t1/2a) of lepirudin is 8–12 min, after which it is distributed in the

extracellular space. Only 20% of lepirudin is found in the plasma, while the
remaining 80% is in the extravascular compartment (Glusa, 1998). Lepirudin
is not transported into the cerebrospinal fluid or breast milk (Refludan
Package Insert; Lindhoff-Last et al., 2000a).
The terminal plasma elimination half-life (t1/2h) ranges from 0.8 to 1.7
h (mean, f1.3 h, or 80 min) following intravenous (iv) injection of bolus
doses of 0.01–0.5 mg/kg and 1.1–2.0 h following continuous iv infusions over
6 h. Maximum activated partial thromboplastin time (aPTT) ratios occur
about 10 min after iv bolus, 3–6 h following 6-h continuous iv infusion, and
2–3 h following subcutaneous (sc) administration (Table 1). During iv
infusion, therapeutic levels are usually reached within 30–60 min.
400 Greinacher
Table 1 Maximum aPTT Ratios (Mean vs. Baseline Values) After Lepirudin
Administration
aPTT ratio
Dose group (mode of application),
(mg/kg body weight) iv bolus 6 h iv infusion sc
Single-dose studies
0.01 1.3 ND ND
0.02 1.5 ND ND
0.04 1.9 ND ND
0.05 ND ND 1.2
0.07 2.3 ND ND
0.10 2.3 1.4 1.4
0.15 ND 1.7 1.6
0.20 2.1 1.7 1.9
0.30 2.4 ND ND
0.35 ND ND 1.9
0.50 2.8 ND 2.0
Multiple-dose studies
0.1 every 24 h 2.4 ND ND
0.1 every 12 h 2.2 ND ND
0.5 every 24 h ND ND 2.3
0.5 every 12 h ND ND 2.1
The maximum ratios of mean aPTT values versus baseline are shown. For iv bolus application,
6-h continuous iv infusion, and sc application, maximum ratios were usually reached at 0–10
min, 3–6 h, and 2–3 h after first application, respectively. This is in good correlation with the
peak plasma lepirudin concentrations achieved with these different modes of application.
Generally, maximum aPTT ratios increased with higher lepirudin doses, with maximum values
of 2.1–2.4 prolongation of baseline for repeated doses above 0.1 mg/kg body weight.
Abbreviations: aPTT, activated partial thromboplastin time; iv, intravenous; ND, not done; sc,
subcutaneous.
The approved dose for lepirudin in patients with HIT and acute
thrombosis (with normal renal function) is an iv bolus of 0.40 mg/kg,
followed by an iv infusion of 0.15 mg/kg/h (Table 2). However, in patients
without massive, life-threatening thrombosis, and especially in elderly
patients, I recommend that the bolus be omitted and an initial infusion rate
of 0.10 mg/kg/h commenced so as to avoid overdosage in case of unrecog-
nized renal insufficiency. The infusion rate should then be adjusted according
to aPTT after 4 h.
Renal clearance (160–200 mL/min for an adult with normal body
surface area of 1.73 m2) and degradation account for approximately 90%
of the systemic clearance of lepirudin. The t1/2h of r-hirudin lengthens with
deterioration of renal function (Markwardt, 1989; Nowak et al., 1991, 1992,

1997; Vanholder et al., 1994, 1997); in nephrectomized patients, it can be up to
120 h (Wittkowsky and Kondo, 2000; Dager and White, 2001; Fischer, 2002;
Shepherd, 2002).
A clinically important observation is that renal blood flow decreases
during anesthesia, so that the elimination half-life is prolonged to 3–5 h. If
lepirudin is used intraoperatively, the dose should be reduced by 30–50%, and
close monitoring is mandatory.
With sc administration, bioavailability is nearly 100%. Dose-ranging
studies have shown that its concentration in the blood reaches 0.3–0.5 Ag/mL
after a sc dose of lepirudin of 0.5 mg/kg and about 0.7 Ag/mL after a sc dose of
0.75 mg/kg, making twice-daily injections effective (Schiele et al., 1994; Huhle
et al., 2000a; Nowak, 2001). When administered sc, this drug is usually in-
jected into an abdominal skin fold and reaches its peak concentration after 2–
3 h. Lepirudin has been administered sc for long-term prophylaxis in HIT
after the acute disease has been controlled (Huhle et al., 2000a). In one patient
the drug was safely administered sc twice daily for 8 months for antithrom-
botic therapy in the setting of malignant disease. Lepirudin has been
administered sc as an adjunct to streptokinase in patients with acute myo-
cardial infarction (MI) (Neuhaus et al., 1999) and in the outpatient manage-
ment of acute MI (Begelman and Deitcher, 2002).
D. Tests for Monitoring Anticoagulation

Numerous tests have been evaluated for monitoring anticoagulation pro-
duced by DTIs, ranging from the ubiquitous aPTT to the newer ecarin
clotting time (ECT) and enzyme-linked immunosorbent assay (ELISA)
techniques for directly measuring the hirudin concentration (Hafner et al.,
2002).
The aPTT is a global coagulation assay and is the current method of
choice for monitoring lepirudin therapy in most situations. In patients who
require higher levels of plasma hirudin and aPTT values above f70 s (depend-
ing on the reagent), the hirudin concentration–aPTT curve flattens, and the
correlation between these parameters diminishes. Because the sensitivities of
different aPTT reagents vary, each laboratory should generate its own stan-
dard curve by spiking normal pooled plasma samples with 0.25, 0.50, 0.75,
1.0, 1.25, 1.5, and 2.0 Ag/mL lepirudin (Fig. 2).
Unlike global tests for measuring clotting time, the ECT monitors
prolongation of clotting time caused by thrombin inhibition alone (Callas
et al., 1995; Nowak and Bucha, 1996; Pötzsch et al., 1997a,b; Koster et al.,
2000a; Fabrizio, 2001; de Denus and Spinler, 2002; Liu et al., 2002). Ecarin,
which is obtained from snake venom, catalyzes the cleavage of prothrombin
402
Table 2 Dosing Schedules for Lepirudin Treatment of Patients with HIT
Bolusa,b IV infusiona,b Target aPTT ratioc
HIT with isolated Noned 0.10 mg/kg b.w./hd,e 1.5–2.5

thrombocytopenia (dose (0.5–0.8 mg/mL)
regimen B in HAT trials)
HIT and thrombosis (dose (0.40 mg/kgd) 0.15 mg/kg b.w./hd 1.5–2.5
regimen A1 in HAT trials) (0.6–1.4 mg/mL)
Thrombosis prophylaxis in 15 mg sc b.i.d.f —
patients with a history of HIT
HIT with thrombosis and 0.20 mg/kg b.w. ivd 0.10 mg/kg b.w./hd 1.5–2.5
concomitant thrombolysis
(dose regimen A2 in HAT
trials)
Renal dialysis every alternate 0.10 mg/kg b.w. iv — 2.0–2.5
day predialysis
Continuous venovenous — 0.005 mg/kg b.w./h 1.5–2.5
hemofiltration (CVVH) (initial rate)
PCI (Mehta et al., 2002); UA 0.40 mg/kg b.w iv 0.15 mg/kg b.w./h 1.5–2.5
or acute MI without ST
Greinacher
elevation (OASIS-2, 1999)

Vascular surgery 0.40 mg/kg b.w. iv 0.10 mg/kg/h 1.5–2.5
(Hach-Wunderle, 2001)
Vascular surgery (intraoperative use up to 250 mL — —
vessel flushes) (0.1 mg/mL solution)
Postoperative anticoagulation — 0.10 mg/kg b.w./h 1.5–2.5
Cardiac surgery using CPB 0.25 mg/kg b.w. ivd 0.20 0.50 mg/mina,g Monitored by ECT:
(dose regimen C in HAT mg/kg b.w. in the >2.5 Ag/mL before
trials) (see also Chap. 19) priming fluid start of CPB; 3.5–4.5
Ag/mL during CPBh
Lepirudin in HIT Treatment
Repeat aPTT determinations should be made 4–6 h after any dose adjustment. Abbreviations: aPTT, activated partial thromboplastin time; b.w.,
body weight; CPB, cardiopulmonary bypass; ECT, ecarin clotting tim; iv, intravenous; MI, myocardial infarction; PCI, percutaneous coronary
intervention; UA, unstable angina.
a
A maximum body weight of 100 kg should be used for dose calculations.
b
Adjust for renal insufficiency.
c
The ratio is based on comparison with the normal laboratory mean aPTT. If Actin FS or Neothromtin reagents are used, the aPTT target range is
usually 1.5–3.0.
d
Used in the HAT-1, HAT-2, and HAT-3 trials.
e
This is the author’s recommended starting dose in all HIT patients, unless life- or limb-threatening thrombosis is present.
f
Tested in a prospective, randomized trial after orthopedic surgery (Eriksson et al., 1996, 1997).
g
Stop 15 min before end of CPB; put 5 mg into CPB after disconnection to avoid clotting of pump.
h
The target lepirudin level pre-CPB (>2.5 Ag/mL) is lower than the level sought during CPB (3.5–4.5 Ag/mL) because of the addition of lepirudin to
the pump priming fluid (0.2 mg/kg body weight).
403
404 Greinacher
to meizothrombin (Kornalik and Blombäck, 1975; Novoa and Seegers, 1980;

Nishida et al., 1995). Meizothrombin is biologically similar to thrombin, ex-
cept that it cleaves fibrinogen much more slowly than thrombin. The interac-
tion of meizothrombin with hirudin, however, is similar to that of thrombin.
Thus, when all the hirudin present in a blood sample has been neutralized by
meizothrombin, thrombin will no longer be inhibited, and clotting will occur
(see Chap. 19).
The ECT shows a linear correlation to lepirudin plasma levels over a
wide range. At present this assay is recommended for monitoring antico-
agulation when higher concentrations of lepirudin are used. It is mandatory
for monitoring of lepirudin during cardiopulmonary bypass (CPB) surgery.
Limitations of Functional Monitoring Tests

Results obtained with aPTT or ECT may be inaccurate in patients whose
plasma is depleted of prothrombin (e.g., severe liver disease) or in patients
with fibrinogen depletion (e.g., post-thrombolysis, hemodilution during CPB)
(Lindhoff-Last et al., 2000b; de Denus and Spinler, 2002). In the ECT, this can
be overcome by addition of normal plasma 1:1 to the assay (Koster et al.,
2000a).
ELISAs can be used to measure the concentration of lepirudin in
plasma. These assays are independent of prothrombin concentrations. Plasma
concentations for lepirudin are 0.2–0.4 mg/mL for thrombosis prophylaxis,
0.5–0.8 mg/mL for isolated HIT, and 0.6–1.4 mg/mL for HIT and thrombosis.
E. Dose Adjustments
Generally, in lepirudin-treated patients, laboratory values to monitor the anti-
coagulant effect should be obtained prior to treatment, 4 h after the start of iv
infusion, and 4 h after every change in dose. For most patients, the primary
anticoagulation parameter used should be the aPTT, and testing should be
performed at least once daily during treatment with lepirudin. If the target
range is exceeded, the infusion should be stopped for 2 h and restarted at a
50% lower dose once the therapeutic range has been reached (Greinacher
et al., 1999a,b). When the dose is subtherapeutic, the infusion rate should be
increased by 20%.
Figure 2 Lepirudin standard curve. This curve was generated using seven normal
plasmas spiked with various concentrations of lepirudin (Ag/mL) using reagent Actin
FS and the BCS analyzer (Dade-Behring, Germany). Note that incremental changes in
activated partial thromboplastin time (aPTT) are much smaller as the dose-response
curve flattens at greater plasma lepirudin concentrations.
406 Greinacher
Renal Impairment
Lepirudin has been studied in patients with varying degrees of renal impair-
ment. It can be used safely and effectively in reduced dose (Table 3). In case of
transient renal failure close monitoring of aPTT is mandatory. To avoid
overdosing due to compensated renal insufficiency, I recommend avoiding the
initial lepirudin bolus (unless severe thrombosis is present) and starting the
lower infusion rate of 0.10 mg/kg/h iv, with subsequent adjustments accord-
ing to aPTT.
Transitioning to Warfarin
In patients with HIT, warfarin (or other coumarins) should be initiated only
after platelet levels have normalized. Further, no loading dose of warfarin
should be given. To cover the initial prothrombotic effects of warfarin,
therapeutic levels of lepirudin (aPTT ratio = 1.5–2.5) should be maintained
for at least 4–5 days after the initiation of oral anticoagulation. In case of a
rapid increase in the international normalized ratio (INR), prothrombin
levels should be used for dose finding.
F. Reversal/Removal of Lepirudin
Bleeding is an important and potentially severe consequence of hirudin
treatment (Antman, 1994; Neuhaus et al., 1994; Frank et al., 1999). As with
all DTIs, no specific antidote is available. In a patient with minor bleeding and
normal renal function, stopping the drug may suffice, since the drug concen-
tration drops quickly. However, when bleeding is life-threatening or the
patient has renal failure, cessation alone may not be adequate.
Table 3 Dosing Schedule for Lepirudin in Patients with HIT and Renal
Impairment
Creatinine clearance Serum creatinine, Adjusted iv infusion rate

(mL/min) mg/dL (Amol/L) (% of original dose [see Table 2])a
45–60 1.6–2.0 (141–177) 50

30–44 2.1–3.0 (178–265) 25
15–29 3.1–6.0 (266–530) 10
<15 >6.0 (>530) 0.005 mg/kg/h b.w. iv
(adjusted for aPTT)
Abbreviations: aPTT, activated partial thromboplastin time; b.w., body weight; iv, intravenous.
a
No initial bolus of lepirudin is given.
Hemodialysis or hemofiltration can reduce plasma levels of lepirudin

(Riess et al., 1995). However, only some filters are effective, e.g., polysulfone
F80 (Fresenius, Germany) (Frank et al., 1999; Bucha et al., 1999). Variable
efficacy of filters in removing lepirudin could explain conflicting results (Van-
holder et al., 1997). Clinical data are limited, and hemofiltration is not always
a practical option in emergency situations.
Hirudin overdosage may also be treated pharmacologically by admin-
istration of desmopressin (Ibbotson et al., 1991; Butler et al. 1993; Bove et al.,
1996), or von Willebrand factor (vWF), or vWF-containing factor VIII con-
centrates (Dickneite et al., 1996, 1998). Irami and coworkers (1995) described
a patient in whom r-hirudin–induced bleeding was treated by administration
of prothrombin complex concentrates, a method previously used in animal
models (Diehl et al., 1995). However, since these concentrates can contain
heparin, they could be dangerous for a patient with acute HIT. Recombinant
FVIIa is another ‘‘panhemostatic’’ treatment option. Meizothrombin is an-
other potential antidote, but it is not available for use in humans (Nowak and
Bucha, 1995).
G. Clinical Use of Lepirudin

Besides its use in patients with HIT, lepirudin has been investigated exten-
sively in controlled clinical trials for acute coronary syndromes (n > 14,000),
such as MI (Antmann, 1994; Neuhaus et al., 1994) and unstable angina pec-
toris (Rupprecht et al., 1995; Organization to Assess Strategies for Ischemic
Syndromes [OASIS-2], 1999); and in pilot studies for prophylaxis and treat-
ment of deep venous thrombosis (Parent et al., 1993; Schiele et al., 1997). In
patients undergoing dialysis (see Chap. 18) or cardiac surgery (see Chap. 19),
there is observational evidence indicating safe and effective use of lepirudin.
Results of three prospective clinical trials with lepirudin and an exten-
sive postmarketing drug monitoring study in HIT patients treated in the
‘‘real-world’’ setting are described in the next section.
III. CLINICAL STUDIES WITH LEPIRUDIN IN HIT

A. Three Prospective Clinical Trials: HAT-1, -2, and -3
Three prospective studies with lepirudin for HIT were designated Heparin-
Associated Thrombocytopenia (HAT)-1, -2, and -3 (Greinacher et al., 1999a,
b; Lubenow and Greinacher, 2002; Eichler et al., 2002). There was no ap-
proved nonheparin alternative anticoagulant during the 3 years in which the
HAT studies were conducted (March 1994 to April 1996), and thus for ethical
reasons, a placebo control was not appropriate. The HAT studies therefore
408 Greinacher
included comparisons of clinical outcomes with a historical control group

treated before lepirudin became available.
Meta-analysis of HAT-1 and -2 was performed to evaluate patients
given lepirudin for treatment of HIT with thrombosis. A second meta-anal-
ysis of the HAT-1, -2, and -3 studies was performed to evaluate the effects of
lepirudin in patients with HIT and isolated thrombocytopenia (‘‘isolated
HIT’’). In addition, an observational study termed the drug-monitoring
program (DMP) was carried out to determine the effects of lepirudin in a
large cohort of patients treated in routine clinical settings.
Objectives
The three HAT trials examined whether lepirudin administered iv to patients
with serologically confirmed HIT would safely reduce the risk of new arterial
or venous thrombosis, limb amputations, and death. The laboratory objective
was to determine whether the drug would allow an increase in the platelet
count in thrombocytopenic patients or maintain the baseline platelet val-
ues (in nonthrombocytopenic patients), while providing effective anticoagu-
lation. The latter was defined as a prolongation of the aPTT by 1.5- to 2.5-fold
over baseline values with no more than two dose increases. (Note: If Actin FS
or Neothromtin reagents were used, the aPTT target range was a 1.5- to 3.0-
fold prolongation.)
Patients
Patients were eligible for study if their platelet count fell by more than 50% or
to fewer than 100 109/L or if they exhibited new thrombosis while receiving
heparin. A strict criterion for study entry was laboratory confirmation of the
clinical diagnosis of HIT by the heparin-induced platelet activation (HIPA)
test (Greinacher et al., 1991; Eichler et al., 1999) (see Chap. 11).
Clinical outcomes included a composite endpoint (new thrombosis,
limb amputation, death) as well as each individual endpoint. Clinical events
that occurred between diagnosis and start of treatment with lepirudin were
included, as were all clinical events that occurred up to day 14 after stopping
lepirudin treatment. Clinical outcomes for lepirudin were compared with a
historical control group treated conventionally by Kaplan-Meier time-to-
event analysis, beginning at laboratory confirmation of HIT for lepirudin-
treated patients and one day after laboratory confirmation for controls.
Laboratory response was defined as (1) the maintenance of an on-
treatment aPTT ratio higher than 1.5 in at least 80% of measurements and
requiring no more than two dose increases and (2) an increase in the platelet
count to more than 30% from the nadir and to more than 100 109/L by day
10 of lepirudin treatment (thrombocytopenic patients), or maintenance of
normal platelet counts on days 3 and 10 (nonthrombocytopenic patients).
Historical Control Group

The historical control patients (n = 120) also had a diagnosis of HIT con-
firmed by a positive HIPA test in our laboratory (Greinacher et al, 1999b).
They were treated according to hospital protocol with danaparoid (n = 36),
oral anticoagulants (e.g., phenprocoumon [n = 27]), no anticoagulation (n =
23), or miscellaneous treatments (e.g., aspirin [n = 5], low molecular weight
heparin [n = 8], or thrombolytics [n = 4]). Incomplete data for 17 patients in
the control group precluded treatment assignment.
B. HAT-1 Study
The HAT-1 study involved 82 patients with confirmed HIT: 51 patients were
assigned to dose regimen A1, 5 to regimen A2, 18 to regimen B, and 8 to
regimen C (Table 4) (Greinacher et al., 1999a). The median duration of treat-
ment was 10 days (range 3–47 days) for regimen A1, 9 days (7–29 days) for A2,
15 days (2–58 days) for B, and 9 days (3–25 days) for C.
Efficacy Outcomes
Compared with the control group, the lepirudin-treated group had signifi-
cantly lower rates of the combined endpoint of new thrombosis, limb am-
putation, and death at day 35 (25.4% vs. 52.1%; p = 0.014). This represented
a 51.2% reduction in risk for the combined endpoint (Table 5). Similarly, the
Table 4 Summary of Treatment Regimens in Four Large Studies of Lepirudin

for HIT
Lepirudin studies
Lepirudin treatment group HAT-1 HAT-2 HAT-3 DMP Total
HIT with thrombosis (regimen A1) 51 65 98 496 710

HIT with thrombosis treated with 5 4 12 — 9
concomitant thrombolysis
(regimen A2)
Prophylaxis in isolated HIT 18 43 84 612 757
(regimen B)
CPB surgery (regimen C) 8 — 10 — 18
Miscellaneous (including sc dosing) — — 1 221 222
Total 82 112 205 1329 1716
Dosing regimens A1, A2, B, and C are described in Table 2.
Abbreviations: CPB, cardiopulmonary bypass; DMP, drug-monitoring program; HAT, heparin-
associated thrombocytopenia trial; sc, subcutaneous.
410
Table 5 Incidences of Composite and Individual Clinical Endpoints in the Individual Clinical Studies and Their Meta-analyses
(see also Fig. 3)
Historical Risk reduction

Study (Ref.) and endpointa Observation period Lepirudin (%)b control (%)b p-value (%)
HAT-1 (Greinacher et al., 1999a) Dx of HIT to d35 n = 71 n = 120

Composite 25.4 52.1 0.014 51.2
New thrombosis 18.4 32.1 0.270 42.7
Limb amputation 5.7 8.2 0.783 30.5
Death 8.6 22.3 0.071 61.4
HAT-2 (Greinacher et al., 1999b) Dx of HIT to d35 n = 95
Composite 30.9 52.1 0.120 40.7
New thrombosis 17.4 32.1 0.260 45.8
Death 10.5 22.3 0.210 52.9
HAT-3 (Eichler et al., 2002) Dx of HIT to d35 n = 191
Composite 26.2 52.1 0.002 49.7
New thrombosis 9.9 32.1 <0.001 69.2
Death 13.6 22.3 0.73 39.0
HIT with thrombosis, meta-analysis Start of lep to d35 n = 113 n = 75
(Greinacher et al., 2000)
Composite 21.3 47.8 0.004 55.4
New thrombosis 10.1 27.2 0.005 62.9
Limb amputation 6.5 10.4 N/S 37.5
Greinacher
Death 8.9 17.6 N/S 49.4

HIT with thrombosis, DMP Start to end of lep n = 496
(Lubenow et al., 2002b) Rx + 1d
Composite 21.9
New thrombosis 5.2
Limb amputation 5.8
Death 10.9
Isolated HIT, meta-analysis Start to end of lep Rx n = 111
(Lubenow et al., 2002a)
Composite 9.0
New thrombosis 2.7
Lepirudin in HIT Treatment
Limb amputation 2.7

Death 4.5
Isolated HIT, DMP Start to end of lep n = 612
(Lubenow et al., 2002b) Rx + 1d
Composite 15.7
New thrombosis 2.1
Limb amputation 1.3
Death 12.3
a
Each patient could contribute only once to the composite endpoint (any one of the individual endpoints of new thrombosis, limb amputation, or
death).
b
Number of patients eligible for comparison.
Abbreviations: d, day; DMP, drug-monitoring program; Dx, diagnosis; lep, lepirudin; Rx, treatment.
411
412 Greinacher
incidence of each of the individual endpoints at day 35 was reduced: new

thrombosis (18.4% vs. 32.1%; p = 0.27), limb amputation (5.7% vs. 8.2%; p
= 0.78), and deaths (8.6% vs. 22.3%; p = 0.07). Causes of death were heart
failure (n = 3), sepsis (n = 2), and multiorgan failure (n = 1). None of the
deaths was judged to be an adverse effect of lepirudin. In 88.7% of the lepi-
rudin-treated patients, platelet counts increased to greater than 100 109/L
within 10 days (Greinacher et al., 1999a).
Compared with the historical control group, the cumulative frequency
of new thrombosis, limb amputation, and death was lower in the lepirudin
group at all time points after laboratory confirmation of HIT (Fig. 3). The
adjusted risk ratio for lepirudin-treated patients relative to historical controls
was 0.508 (95% CI 0.29–0.892).
Safety Outcomes
At 4 weeks the cumulative rate of bleeding events was not statistically dif-
ferent in the lepirudin group compared to control (39.6% vs. 35.3%; p =
0.60). During the study, 27 patients (32.9%) experienced one bleeding event,
with 11 patients (13.4%) suffering 15 major bleeding events. Of those, 8 were
at invasive sites and 7 were spontaneous. There were no significant differences
between groups in the frequency of bleeding events requiring transfusion
(9.9% vs. 9.1%; p = 0.59). There were no intracerebral or fatal hemorrhages
in the lepirudin group.
C. HAT-2 Study
The HAT-2 study involved 112 patients with confirmed HIT: 65 patients were
assigned to dose regimen A1, 4 to regimen A2, and 43 to regimen B (Table 4)
(Greinacher et al., 1999b). The overall median duration of treatment was 11
days (range 0–104 days); for regimen A1, it was 13 days (0–104 days); for A2,
10 days (1–58 days); and for B, 8 days (1–67 days).
Efficacy Outcomes
The average combined outcome rate (expressed per patient-day) markedly
decreased from the pretreatment period (5.1%) to the periods during (1.5%)
and after (0.6%) lepirudin treatment. Platelet count recovery was achieved in
87 of 94 (92.6%) evaluable patients, with median platelet counts increasing
about fourfold over the first 10 days.
Compared with the historical control group, the cumulative frequency
of the composite endpoint (new thrombosis, limb amputation, death) was
lower in the lepirudin group at all time points after laboratory confirmation
of HIT (Table 5). At 5 weeks, the frequencies were 30.9% (95% CI 21.0–40.7)
Figure 3 Time-to-event analyses of efficacy and safety endpoints in the HAT-1 and
HAT-2 studies (combined) in comparison with the historical control group. (A)
Composite endpoint (new thromboembolic complication, limb amputation, or death);
(B) new thromboembolic complication; (C) limb amputation; (D) death; and (E) major
bleeding.
414 Greinacher
for the lepirudin group and 52.1% (95% CI 40.4–63.9) for the historical
control group ( p = 0.12). Lepirudin-treated patients fared somewhat better
than historical controls at 5 weeks for the individual outcomes of new
thromboses (17.4% vs. 32.1%; p = 0.26), and death (10.5% vs. 22.3%;
p = 0.21) but not for limb amputation (10.0% vs. 8.2%; p = 0.43). The
adjusted risk ratio for lepirudin-treated patients relative to historical controls
was 0.709 (95% CI 0.44–1.14; p = 0.15). Causes of death included multi-
organ failure (n = 3), sepsis (n = 2), heart failure (n = 2), pulmonary
embolism, ventricular fibrillation, shock, and apnea (n = 1 each). None of
the deaths were judged to be related to adverse effects of lepirudin.
Safety Outcomes
At 35 days, the cumulative frequency of bleeding events was 44.6% (95% CI
33.8–55.4) in the lepirudin group and 27.2% (95% CI 16.3–38.0) in the con-
trol group ( p = 0.0001 by log-rank test). There were no significant differences
between the lepirudin and control groups, however, in the frequency of
bleeding events requiring transfusion (12.9% vs. 9.1%; p = 0.23). Most se-
vere bleeding occurred at invasive sites. The frequency of serious spontaneous
bleeding, including gastrointestinal hemorrhages (2.1% for lepirudin vs.
5.0% for control) and pulmonary hemorrhages (2.1% for lepirudin vs.
1.7% for control), was low and not significantly different. No cerebral hem-
orrhages occurred in the lepirudin group.
D. HAT-3 Study
A third prospective trial, HAT-3, was the largest and involved 205 patients: 98
patients were assigned to dose regimen A1, 12 to regimen A2, and 84 to
regimen B (Eichler et al., 2002). Ten patients received lepirudin for CPB
(regimen C), and one received lepirudin by the sc route. Fifteen patients were
enrolled twice. For the efficacy parameters only the first treatment cycle was
calculated. For safety analysis, especially allergic reactions, all treatment cy-
cles were included.
Median treatment duration was 10 days over all treatment groups, 9
days (range 1–197 days) for regimen A1, 12 days (5–21 days) for regimen A2,
10 days (1–47 days) for regimen B, and 7 days (1–37 days) for regimen C.
Efficacy Outcomes
The results of HAT-3 confirmed the efficacy of lepirudin in HIT, as outcomes
were similar to HAT-1 and HAT-2. Of the 84 patients with HIT and isolated
thrombocytopenia, 9 (10.7%) experienced new thrombosis, 7 (8.3%) under-
went limb amputation, and 11 (13.1%) died. Since patients were counted only
once if multiple events occurred, the incidence of the combined endpoint was
27.4% (23/84). Of the 110 patients with HIT and thrombosis (treatment reg-
imens A1 and A2), 19 (17.3%) experienced new thromboses, 6 (5.5%) un-
derwent limb amputation, and 16 (14.5%) died; in this group, the combined
endpoint rate was 31.8% (35/98).
The average combined event-rate per day, as observed before, during,
and after lepirudin treatment, supports the previous observation that heparin
cessation alone may not prevent serious complications in HIT. The mean
1.7-day delay in initiating therapy for clinically suspected HIT while await-
ing laboratory confirmation was associated with severe, even life-threaten-
ing clinical consequences, with an average event-rate per day of 5.6%. This
dropped to 0.87% during treatment and was 0.77% after cessation of lepir-
udin therapy.
Compared to the historical control, from laboratory confirmation until
end of the observation period, the combined endpoint was markedly reduced
(26.2% vs. 52.1%; p = 0.002) primarily due to a reduction in new thrombosis
(9.9% vs. 32.1%; p = 0.0002).
Safety Outcomes
The overall incidence of major bleeding was 19.5%, with 5 fatal outcomes
(2.4%). One patient who received concomitant thrombolysis experienced
intracranial hemorrhage. Of the 205 patients, 8 (3.9%)—7 in the HIT plus
thrombosis group and 1 in the HIT group—experienced 11 episodes of
allergic reactions to lepirudin, with 4 associated with an antibody to lepirudin.
There were no cases of anaphylaxis.
E. Meta-Analysis of HAT-1 and HAT-2: Patients with HIT

and Thrombosis
A meta-analysis of HAT-1 and HAT-2 was performed to determine the effi-
cacy and safety of lepirudin in 113 patients with HIT complicated by
thrombosis (Greinacher et al., 2000). As in the HAT-1, -2, and -3 studies,
the risk for new thrombotic complications (per day) was highest between
diagnosis of HIT and start of treatment: 34.1% of all events occurred during
this pretreatment period (Fig. 4).
Efficacy Outcomes
When outcomes were assessed from the start of lepirudin treatment, the
combined endpoint for new thrombosis, limb amputation, and death was
416 Greinacher
significantly lower in the lepirudin-treated patients (n = 113) than in the

controls (n = 91) (21.3% vs. 47.8%; p = 0.004). This difference was primar-
ily due to a reduction in the number of new thrombosis (10.1% vs. 27.2%;
p = 0.005). Incidences of limb amputation (6.5% vs. 10.4%) and death
(8.9% vs. 17.6%) were also lower in the lepirudin group than in the historical
controls (Table 5).
Safety Outcomes
There were no fatal or intracranial bleeds but, compared with the controls, the
cumulative incidence of bleeding was higher in the lepirudin group than in the
control group (42.0% vs. 23.6%; p = 0.001), and more lepirudin-treated
patients experienced bleeding requiring transfusion (18.8% vs. 7.1%; p =
0.02).
One of the more important points to emerge from the meta-analysis was
the relationship of aPTT ratios with lepirudin safety and efficacy. For low
aPTT ratios (<1.5), the incidence of the combined endpoint was not signif-
icantly reduced compared to the control (RR = 0.86; p = 0.72). In addition,
the risk of bleeding was not significantly greater in the lepirudin group than
in the control group (RR = 1.57; p = 0.42). At medium aPTT ratios (1.5–
2.5), efficacy was significantly greater for the lepirudin-treated patients than
for the controls (RR = 0.42; p = 0.009), but there was also an increased risk
of bleeding (RR = 3.21; p = 0.0003). At higher aPTT ratios (>2.5), the
efficacy of lepirudin was not enhanced, but there was an even greater risk of
bleeding.
F. Meta-Analysis of HAT-1, -2, -3: Patients with Isolated HIT

Each of the HAT trials examined the effects of lepirudin in patients with HIT
and isolated thrombocytopenia and in HIT patients with thrombosis. The
Figure 4 Average combined event rate (new thromboembolic complication, limb

amputation, death) per day in the HAT-1 and HAT-2 studies (n = 113 patients with
thromboembolic complications at baseline). The bar width indicates the mean
duration of the observation period (days) and is shown for three time periods: before,
during, and after lepirudin therapy. The high average event rate (0.061 [6.1%] event
per day during a mean period of 1.7 days) from diagnosis of HIT until start of lepirudin
therapy indicates that cessation of heparin alone is insufficient to prevent HIT-
associated thrombosis, thus warranting treatment with an alternative anticoagulant if
HIT is strongly suspected. (From Greinacher et al., 2000.)
418 Greinacher
meta-analysis of HAT-1 and HAT-2 showed that lepirudin was effective in

patients with HIT plus thrombosis. In a meta-analysis of HAT-1, -2, and -3,
the safety and efficacy of lepirudin in patients with HIT in the absence of
known thrombosis was demonstrated (Lubenow et al., 2002a). This meta-
analysis included 111 patients treated according to dosing regimen B; of these,
17 had a history of HIT (latent HIT), not acute HIT. Patients with recent
thrombosis at baseline were excluded from this analysis. Mean duration of
treatment was 13.5 days (range 1–68 days).
Efficacy Outcomes
Of the 111 patients in this study during treatment with lepirudin, 3 (2.7%)
experienced new thromboses, 3 (2.7%) underwent limb amputation, and 5
(4.5%) deaths occurred. Most of the deaths were related to underlying
disease, not to HIT or treatment with lepirudin. Since patients were counted
only once if multiple events occurred, the incidence of the combined endpoint
was 10/111 (9.0%). The median platelet count rebounded to 150 109/L
within 4 days of beginning lepirudin treatment.
Safety Outcomes
Episodes of major bleeding occurred in 16/111 (14.4%) of patients in this
meta-analysis. aPTT ratios above 2.5 were associated with an increased risk of
bleeding, and bleeding rates were significantly lower in patients with aPTT
<60. Nearly all patients with bleeding complications had impaired renal
function. Antihirudin antibodies were detected in 36 of the 108 (33.3%) eval-
uable patients. There were no differences in adverse events or outcomes be-
tween patients with and without antihirudin antibodies. No anaphylaxis was
observed.
G. Postmarketing Drug Monitoring Program

Preliminary results from 1329 patients treated with lepirudin in a drug-
monitoring program (DMP) were recently made available (Lubenow et al.,
2002b): 496 patients had HIT and thrombosis and 612 patients had isolated
HIT. This postmarketing study evaluated the same clinical endpoints as were
used in HAT-1, -2, and -3. In this DMP, lepirudin could be started immedi-
ately upon clinical diagnosis of HIT, thus avoiding the inherent delay
awaiting laboratory confirmation. A total of 382 (77.0%) of the 496 patients
with HIT and thrombosis were positive in the HIPA test, while 406 (66.3%) of
the 612 patients with isolated HIT were positive in the HIPA test.
Efficacy Outcomes
In the routine clinical settings of the DMP, lepirudin-treated patients with
isolated HIT and HIT with thrombosis had the lowest incidence of all clinical
endpoints reported with any agent. The incidence of the combined clinical
endpoint in the 496 patients with HIT and thrombosis was 21.9%: 26 patients
(5.2%) experienced new thrombosis, 29 (5.8%) underwent limb amputation,
and 54 patients (10.9%) died. The largest cause of death was multiorgan
failure (23/54 patients [42.6%]), emphasizing the serious underlying medical
condition of these patients. The incidence of new thrombosis in this study
(5.2%) was lower than that observed in the HAT-1 and -2 meta-analysis
(10.1%). This may be due to physicians’ increased clinical experience with
lepirudin, as illustrated by the decision to begin lepirudin treatment immedi-
ately upon clinical diagnosis of HIT, thereby improving efficacy and safety
outcomes.
The combined endpoint of new thrombosis, limb amputation, and
death occurred in 96 (15.7%) of the 612 patients with isolated HIT; 13 pa-
tients (2.1%) experienced new thrombosis, 8 (1.3%) underwent limb ampu-
tation, and 75 patients (12.3%) died. These event rates are the lowest reported
for any agent used to treat HIT. As seen in the group of patients with HIT plus
thrombosis, the largest cause of death in this group was multiorgan failure
(39/75 patients, 52.0%).
The overall mortality rate due to new thrombosis in the group of 1108
patients treated with regimen A1 or B (thus, excluding patients receiving ‘‘mis-
cellaneous’’ treatments) (see Table 4) was low (15 patients, or 1.4%). Efficacy
variables in the DMP were even more favorable than those seen in the meta-
analyses of the HAT studies. This DMP thus confirms the efficacy of lepiru-
din in routine clinical practice for both the prophylaxis and the treatment of
thromboembolism in patients with HIT.
There were no differences in the mean infusion rates in patients with
HIT and thrombosis (0.12 mg/kg/h) and those with isolated HIT (0.11 mg/kg/
h) in the DMP. As lepirudin dose is adjusted based on aPTT, the major
difference between the two regimens is the initial bolus in HIT patients with
acute thrombosis. However, as discussed earlier, in my view the bolus should
be avoided in most situations to prevent overdosing.
Safety Outcomes
In the DMP incidence of bleeding was greatly decreased when compared to
the HAT clinical trials. In the group of 496 patients with HIT plus thrombo-
sis, there were 27 (5.4%) major bleeding episodes, and among the 612 patients
with isolated HIT, 36 (5.9%) had major bleeding. Allergic reactions were re-
420 Greinacher
ported in 4 (0.8%) patients in the HIT plus thrombosis group and in 1 (0.2%)
patient with isolated HIT. No anaphylaxis was reported.
The decreased incidence of bleeding events in the DMP most likely is
attributed to physicians’ greater experience with administering lepirudin and
monitoring its effects.
H. Comparison with Other Treatments for HIT

Mortality rates in patients with HIT have been approximately 20–30% for
more than a decade (King and Kelton, 1984; AbuRahma et al., 1991;
Warkentin and Kelton, 1996; Nand et al., 1997). Notably, these rates are
two to three times higher than those observed in the HAT studies. The HAT
trials demonstrate the necessity for prompt initiation of alternative anti-
coagulation following the discontinuation of heparin. In addition to lepir-
udin, other drugs with antithrombin activity (e.g., argatroban) or antifactor
Xa activity (e.g., danaparoid) may be appropriate for further parenteral
anticoagulation in patients with HIT (see Chaps. 13, 14, 16, 17).
Comparisons of the results of the various clinical trials of agents used to
treat HIT need to be interpreted with caution, since there have been no direct
comparative trials and the studies employed are of somewhat different
designs. Trials of lepirudin and argatroban, however, utilized similar clinical
endpoints and historical controls for comparison. The most obvious differ-
ences between the lepirudin and the argatroban trials are (1) the need for
laboratory confirmation of HIT in the lepirudin trials; (2) treatment duration,
which was consistently longer than 10 days in the lepirudin-treated patients
but less than 6 days in the argatroban-treated patients (potential to increase
apparent efficacy and also bleeding with lepirudin); (3) the observation
period, which started at the time of diagnosis in lepirudin-treated patients
compared with the time of treatment initiation in the argatroban trials
(potential to underestimate the efficacy of lepirudin); and (4) a considerable
proportion of patients in the historical control group of the HAT trials had
been treated with danaparoid (potential to underestimate the efficacy of
lepirudin). To allow a more direct comparison, we reanalyzed the data of
the HAT trials as per the argatroban trials, i.e., analyzing those events
occurring from start of active treatment only (Table 5).
The rates for the combined endpoint were consistently lower in the
lepirudin trials than in the two argatroban studies. They were 9.0% in the
HAT-1, -2, -3 meta-analysis and 15.7% in the DMP during treatment with
lepirudin, as compared to 25.6% and 28.0% in the two argatroban trials, for
patients with isolated HIT. (However, the observation period for argatroban
also included a follow-up period, whereas the lepirudin data were obtained
during the treatment period only.) For patients with HIT and thrombosis, the
combined endpoint occurred in 21.3% (meta-analysis, including the post-

treatment follow-up period) and 21.9% (DMP, treatment period only) with
lepirudin, and in 43.8% and 41.5% (including the follow-up period) of
patients treated in the two argatroban trials.
Because of the often-critical condition of the patient population inves-
tigated, the rate of deaths observed for both DTIs is not likely to be attri-
butable to treatment failure and may vary considerably with different patient
populations. As the argatroban trials included many patients who most likely
did not have HIT, the death rate associated with HIT might have been
overestimated, as non-HIT patients with a decrease of platelet count are often
very sick (e.g., septicemia, disseminated intravascular coagulation). Death
rates were 4.5% (meta-analysis) and 12.3% (DMP) for patients with isolated
HIT treated with lepirudin, but 18.1% and 23.1% in those treated in the two
argatroban trials. In patients with HIT complicated by thrombosis, death
rates with lepirudin were 8.9% (meta-analysis) and 10.9% (DMP), and 18.0%
and 23.1% in the argatroban trials.
Limb amputation occurred in patients with isolated HIT in 2.7% (meta-
analysis) and 1.3% (DMP) when treated with lepirudin, and in 1.9% and
4.2% in those treated with argatroban. The amputation rates were higher in
HIT with thrombosis; 6.5% (meta-analysis) and 5.8% (DMP) in those treated
with lepirudin, and 11.1% and 14.8% in those treated with argatroban.
There was also a difference in the incidences of new thrombosis, which is
most likely the most important parameter for assessment of an alternative
anticoagulant in HIT. In those with isolated HIT, it was 2.7% (meta-analysis)
and 2.1% (DMP) with lepirudin, and 6.9% and 5.8% in those treated with
argatroban. In HIT with thrombosis it was 10.1% (meta-analysis) and 5.2%
(DMP) for lepirudin, and 14.6% and 13.1% for argatroban-treated patients.
The risk of major bleeding seems to be higher in lepirudin-treated
patients when comparing total numbers. In those with isolated HIT, it was
14.4% (meta-analysis) and 5.9% (DMP), but 3.1% and 5.3% in those treated
with argatroban. In patients with HIT and thrombosis, maj or bleeding
occurred in lepirudin-treated patients in 18.8% (meta-analysis) and 5.4%
(DMP) and in 11.1% and 6.1% of the patients in the argatroban trials.
However, the differences in treatment duration are important for assessing the
bleeding risk. In fact, when calculating the risk for major bleeds per treatment
day, for HIT patients with thrombosis it was 0.020 major bleeds per patient
day during active treatment in the HAT trials and 0.019 in the argatroban
trials, respectively.
Because there are no prospective data comparing lepirudin and dana-
paroid for treatment of HIT, we retrospectively compared 126 danaparoid-
treated patients with 175 lepirudin-treated patients who fulfilled the same
inclusion and exclusion criteria (Farner et al., 2001). In the patients with HIT
422 Greinacher
without thromboembolic complications at baseline, a time-to-event analysis

showed that the cumulative risk of the combined endpoint was higher in
danaparoid-treated patients than in the lepirudin-treated patients ( p = 0.02
by log rank test; hazard ratio [HR] = 2.9 [95% CI 1.1–7.6]; p = 0.027). This
was due primarily to an increased incidence of new thromboembolic compli-
cations (20% [95% CI 8.4–36.9] for danaparoid vs. 6.3% [95% CI 1.3–17.2]
for lepirudin; p = 0.087). Of note, patients with isolated HIT usually received
only prophylactic-dose danaparoid. In contrast, HIT patients with throm-
bosis at baseline, and who were therefore treated with a therapeutic-dose
regimen of danaparoid, had a similar outcome as patients receiving lepirudin
( p = 0.913).
The major conclusions of these comparisons are: (1) HIT patients seem
to benefit from a longer treatment period with an alternative anticoagulant,
with 10 days better than 5 days; (2) the prophylactic-dose regimen of
danaparoid (750 U sc two or three times daily) approved in the European
Union for HIT with isolated thrombocytopenia appears to be suboptimal.
Thus, patients with acute HIT require anticoagulation generally in therapeu-
tic doses (see also Chap. 13).
Potentially, the short half-lives of lepirudin and argatroban are advan-
tageous in patients who require dose changes, such as those with planned
invasive procedures. As argatroban is mainly eliminated hepatically, its half-
life is independent of renal function. The long half-life of danaparoid may be
advantageous in other clinical settings, such as in prophylaxis and early
mobilization of patients.
I. Antibody Formation
Because hirudin is a protein obtained from a nonhuman species, lepirudin can
induce antibody production in humans. Antibodies are induced by both iv
therapeutic-dose and sc prophylactic-dose use (Greinacher et al., 2003a).
Antihirudin antibodies have been detected in 44–74% of patients treated with
lepirudin (Huhle et al., 1998; Song et al., 1999; Eichler et al., 2000). Of 196
HIT patients treated with lepirudin for 5 or more days, 44% developed
antihirudin antibodies of the IgG class (Eichler et al., 2000). These antibodies
were not associated with an increase in thrombin-antithrombin (TAT)
complexes (Fig. 5). None of these patients developed allergic reactions to
lepirudin. Antibody formation occurred as early as day 4 and peaked at days
8–9 (Eichler et al., 2000).
Antilepirudin antibodies can extend the half-life of lepirudin (Liebe et
al., 2002), most likely by reduced renal filtration of lepirudin-antilepirudin
complexes (Fig. 6); in about 2–3% of patients with antilepirudin antibodies,
an inhibitory effect is seen (Huhle et al., 2001; Fischer et al., 2003). The
Figure 5 Thrombin-antithrombin (TAT) complex concentrations in relation to

antihirudin antibody formation. TAT complex concentrations did not differ between
antihirudin antibody–positive (., solid lines) and antihirudin antibody–negative (o,
dotted lines) patients (median and 25 and 75% quartiles are given).
biological effects of antilepirudin antibodies on anticoagulation can be easily

compensated by changes in the lepirudin dose. Thus, ongoing daily aPTT
measurements are recommended during lepirudin treatment, even when
stable anticoagulation has been observed during the first 5 days.
J. Allergic Reactions
Lepirudin administration during prospective studies in patients with HIT was
associated with a low incidence of allergic events, as well as during the much
larger clinical trials in patients with acute coronary syndromes. Among the
adverse events reported were eczema, rash, pruritus, hot flushes, fever, chills,
urticaria, bronchospasm, cough, stridor, dyspnea, angioedema (variously of
the face, tongue, larynx), and injection-site reactions. Any causal relationship
of lepirudin to these adverse events is unclear.
To date, of 35,000–60,000 patients treated with lepirudin, nine patients
were judged to have had severe anaphylaxis in close temporal association with
424 Greinacher
Figure 6 This 53-year-old woman was admitted to the hospital because of an ankle
fracture. She received low molecular weight heparin for 10 days, but was switched to
unfractionated heparin because of a distal deep vein thrombosis (DVT). Ten days later
she presented with proximal DVT, pulmonary embolism, and a rapid fall in platelet
count from more than 200 to 12 109/L. She was switched to intravenous (iv) lep-
irudin (schedule A1). After normalization of platelet counts, she received overlapping
oral anticoagulation (phenprocoumon), with lepirudin stopped when the INR reached
2.0. Antihirudin antibodies were first detected on day 7; at the same time, the aPTT
increased despite a stable hirudin dosage of 0.05 mg/kg b.w. per hour.
lepirudin use (Greinacher et al., 2003b). All reactions occurred within minutes
of iv bolus lepirudin administration, with four fatal outcomes (three acute
cardiorespiratory arrests, one hypotension-induced MI). In these four cases, a
previous uneventful treatment course with lepirudin was identified (1–12
weeks earlier). In an additional patient with nonfatal anaphylaxis (who did
not receive a bolus), we found high-titer IgG antilepirudin antibodies. Since
lepirudin has been used in approximately 35,000 patients, the risk of ana-
phylaxis is estimated at 0.015% (5/32,500) in first-exposure and 0.16% (4/
2500) in reexposed patients (assuming 7.5% reexposure frequency). We and
others (Bircher et al., 1996) demonstrated high titer antihirudin antibodies of
the IgG class, but not of the IgE class in patients with hirudin-associated
anaphylaxis. IgG-dependent anaphylaxis likely is Fc receptor-mediated and
related to infusion dose. Thus, besides reducing bleeding risk, avoiding iv

bolus administration of lepirudin should also reduce the risk of severe ana-
phylactic reactions.
IV. LEPIRUDIN TREATMENT IN OTHER CLINICAL

SETTINGS
A. Acute Coronary Syndromes and Percutaneous Coronary
Intervention
Due in part to their ability to inhibit clot-bound thrombin, DTIs have also
been investigated as anticoagulants for acute coronary syndromes (ACS) and
percutaneous coronary intervention (PCI). Lepirudin was examined in large
numbers of patients with unstable angina or suspected acute MI without ST
segment elevation in the OASIS-1 (OASIS Investigators, 1997) (n = 909) and
OASIS-2 (n = 10,141) trials (OASIS Investigators, 1999). These trials
concluded that lepirudin is superior to heparin in preventing ischemic out-
comes. A meta-analysis of 11 ACS trials involving over 35,000 patients
revealed a 15% reduction in death or MI when bivalent DTIs (lepirudin or
bivalirudin) were used to treat ACS patients, compared with heparin (DTI
Trialists’ Collaborative Group, 2002). A retrospective subset analysis of the
OASIS 2 trial examined the benefit of lepirudin in 117 ACS patients
undergoing PCI within the first 72 hours (Mehta et al., 2002). Lepirudin
was superior to heparin in reducing the risk of death or MI at 96 hours ( p =
0.036) and 35 days ( p = 0.02). Based on this evidence, lepirudin should be
considered a treatment option in ACS patients with HIT.
B. Cardiopulmonary Bypass and Vascular Surgery

Lepirudin was initially used to manage CPB patients in the HAT studies
(Riess et al., 1995, 1996) but has also been used successfully by other
investigators for CPB (Warkentin and Greinacher, 2003). It is now accepted
that lepirudin is a suitable alternative for anticoagulation during CPB in
patients with acute HIT, provided that ECT monitoring is performed (Koster
et al., 1998, 2000a,b; Johnston et al., 1999; Follis and Schmidt, 2000; Latham
et al., 2000; Longrois et al., 2000). Neither the activated clotting time (ACT)
nor the aPTT is appropriate for monitoring r-hirudin plasma levels in such
high-dose situations (see Chap. 19).
Koster and colleagues (2000b) used lepirudin instead of heparin in 57
patients who had clinically diagnosed HIT and required CPB. The primary
diagnoses included coronary artery disease (n = 27, including 8 cases of MI),
426 Greinacher
valvular heart disease (n = 14), combined coronary artery and valvular

disease (n = 9), thoracic aortic aneurysms (n = 4), ventricular septal defect
resulting from infarction (n = 2), and atrial tumor (n = 1). In that study,
anticoagulation was monitored with ECT, and lepirudin was maintained in
the range of 3–4 Ag/mL. The dose requirement for CPB was 0.016–0.035 Ag/
kg/min, with concurrent 24-h blood drainage of 50–2200 mL. Elimination of
the drug at the conclusion of CPB was augmented through modified zero-
balanced ultrafiltration and forced diuresis. However, drug removal was
dependent on the prevailing renal function. Four patients with impaired renal
function showed prolonged elimination and bleeding. Of the 57 patients, 54
achieved full recovery and showed no signs of thromboembolism over a 6-
month follow-up. Three patient deaths were unrelated to perioperative
management.
For patients undergoing vascular surgery, the dosage of lepirudin
should be adjusted for the risk for reocclusion (Hach-Wunderle, 2001). In
patients with a low risk of reocclusion (e.g., in the aortic, iliac, and carotid
arteries), a bolus of 0.4 mg/kg (reduced in case of renal insufficiency) is given
just before the vessel is clamped and is followed postoperatively by either an
aPTT-adjusted infusion of 0.1 mg/kg/h or 15 mg injected sc b.i.d. In patients
with an increased risk for enhanced reocclusion (e.g., undergoing calf-vessel
reconstruction or bypass), a preoperative bolus of lepirudin (0.4 mg/kg [less in
case of renal impairment]) should be administered, followed by a postoper-
ative infusion of 0.15 mg/kg/h, aPTT-adjusted, for at least 3–4 days. For
intraoperative flushing of the vessel during vascular surgery, up to 250 mL
(0.1 mg/mL solution) of lepirudin can be used. As patients with acute HIT are
at high risk for new thromboembolic complications, therapeutic levels of
anticoagulation should be achieved before surgery and maintained after
surgery, at least until platelet counts are normalized (see Chap. 13).
C. Hemodialysis
Hirudin was the first anticoagulant to be used for hemodialysis, as performed
by Haas (1924) in Germany. Because native hirudin preparations were crude
and supply of leeches insufficient, hirudin was replaced by heparin to prevent
clotting during dialysis. Currently, more published reports describe lepirudin
for hemodialysis than the other DTIs (see Chap. 18).
Management of these patients requires careful dosing and frequent
monitoring. HIT patients with transient renal failure are difficult to manage
with lepirudin, because substantial dose adjustments are necessary, depend-
ing on the extent of renal failure. To reduce bleeding risk, we prefer admin-
istering a continuous iv infusion, starting at 0.005 mg/kg/h, with adjustments
made according to the aPTT, while others use intermittent iv boluses of

0.005–0.01 mg/kg (Fischer et al., 1999; Kern et al., 1999).
D. Lepirudin in Pregnancy
Data on the treatment of HIT during pregnancy are limited. In general, the
use of lepirudin during pregnancy is not recommended, as it crosses the
placenta. Zebrafish experiments indicate that thrombin has an important role
in early embryogenesis and that inhibition by lepirudin may cause cell regu-
lation defects (Jagadeeswaran et al., 1997). Experiments in rabbits showed
a fetal hirudin plasma concentration that was 1/60 that of the maternal
concentration (Markwardt et al., 1988), and embryotoxic effects were seen in
rabbits at high, but not low, doses (30 vs. 1–10 mg/kg/day, respectively) (Ber-
lex Laboratories, data on file).
Reports on the use of lepirudin in pregnancy are sparse (Lindhoff-Last
and Bauersachs, 2002). A pregnant woman with systemic lupus erythemato-
sus who was treated with dalteparin developed HIT at week 25. Her platelet
count dropped from 230 to 59 109/L, after which she was treated with
lepirudin (15 mg sc twice daily), with aPTT and ECT used to monitor her
dosage. Following delivery by cesarean section, she experienced no postpar-
tum bleeding complications, and treatment with lepirudin was continued for
several weeks thereafter (Huhle et al., 2000b). Another pregnant woman with
lupus anticoagulant and HIT was successfully treated for 36 weeks with
lepirudin.
A case report described a breastfeeding woman diagnosed with HIT
who was treated with sc lepirudin, 50 mg twice daily (Lindhoff-Last et al.,
2000a). No lepirudin was detected in her breast milk, although plasma levels
were within therapeutic range. Neither bleeding nor thrombosis occurred in
mother or infant.
Lepirudin and danaparoid are each classified by the FDA as pregnancy
category B, based on limited animal data. However, danaparoid does not
cross the placenta, and it has been used for prophylaxis and therapy of HIT
during pregnancy (Greinacher et al., 1993; Dager and White, 2002) (see
Chaps. 13 and 14).
E. Lepirudin in Children
Although rare in children, HIT is important in the differential diagnosis of
thrombocytopenia or unexplained thrombosis in the presence of heparin
administration (Ranze et al., 1999). Because of the rarity of HIT and its
clinical heterogeneity in pediatric patients, it is difficult to design a standard-
428 Greinacher
ized dosage protocol for lepirudin. Accordingly, current therapeutic recom-

mendations are based on anecdotal experience. Given that children usually
have normal renal function, the short half-life of lepirudin presents an
advantage in the event of bleeding complications or the need for invasive
procedures. However, the dose required may range between 0.05 and 0.22 mg/
kg/h, depending on comorbidity and renal function (Schiffmann et al., 1997;
Deitcher et al., 2002; Nguyen et al., 2003) (see Chap. 20).
V. CONCLUSION
The r-hirudin lepirudin is a DTI that provides rapid and effective anti-
coagulation and significantly reduces the risk of thrombosis in patients with
HIT, including those with isolated thrombocytopenia. Fewer than 10% of all
patient groups with HIT developed a new thrombosis after start of active
treatment. The drug also reduced risk of limb amputation and death.
Published data on lepirudin include more than 8500 treated patients. Of
these, about 1500 patients were treated for HIT (the largest experience with a
DTI). An additional 7300 patients received lepirudin for ACS and PCI.
Lepirudin is given parenterally by iv infusion or sc injection. Recom-
mended lepirudin dosage schedules have been established (Table 2). Lepir-
udin has a short half-life, which presents an advantage if invasive surgical
procedures are indicated. However, its elimination strongly depends on renal
function. Bolus dosing should be avoided, especially in elderly patients, to
avoid overdosing. Lepirudin can be used safely and effectively in patients with
renal impairment by appropriate dosing according to serum creatinine and
regular monitoring. Lepirudin also allows for a safe and uncomplicated
transition to warfarin.
The most common adverse event in the prospective clinical trials was
bleeding. No antidote exists for the DTIs. Excess lepirudin can be removed by
hemofiltration, but clinical data are limited. Daily monitoring of aPTT is
recommended with dosage adjustments made as needed to maintain the target
aPTT value. Routine monitoring with ECT should be performed in high-dose
situations, such as those required during CPB.
Besides the 399 patients with HIT treated in prospective trials, an
additional 1329 patients received lepirudin for HIT in a postmarketing
surveillance study. Data on these patients, collected under routine clinical
conditions, showed the lowest incidence of the clinical endpoints of death,
new thrombosis, and amputations, with risk reductions exceeding those
reported in the prospective clinical trials. Even more importantly, the inci-
dence of major bleeding was low. These differences support the assumption
that outcomes in patients with HIT can be substantially improved by im-
mediately stopping heparin and starting lepirudin when HIT is strongly sus-
pected on clinical grounds, without awaiting results of antibody testing, and

that the bleeding risk has been reduced substantially as physicians have
learned to handle this agent.
The experience with lepirudin in the prospective trials and the DMP
has now been extended to about 35,000–60,000 patients, including HIT as
well as ACS, PCI, CPB, and deep vein thrombosis. This far exceeds the avail-
able data on any other therapy in HIT and also gave insights in the frequency
of rare adverse effects associated with lepirudin treatment, such as anaphy-
laxis. The results of these trials and the DMP demonstrate that lepirudin is
highly effective in reducing the risk of the potentially devastating complica-
tions of HIT.
ACKNOWLEDGMENTS
The HAT studies were performed jointly by the combined clinical research
team of Behring-Werke AG, Hoechst, and Aventis. The laboratory studies
were supported by Deutsche Forschungsgemeinschaft GR 1096-2/2 and 2/3,
and 2/4. Analysis of the HAT-3 data and the HAT 1-2-3 meta-analysis was
supported by a grant from Berlex Laboratories (Montville, NJ) and
Pharmion (Cambridge, UK). The study on a comparison of danaparoid and
lepirudin was supported by Organon NV (Oss, The Netherlands)
and Thiemann/Celltech (Essen, Germany). The assistance of Theresia Lietz
in data analysis is highly appreciated.
REFERENCES
AbuRahma AF, Boland JP, Witsberger T. Diagnostic and therapeutic strategies of

white clot syndrome. Am J Surg 1991; 162:175–179.
Antman EM. Hirudin in acute myocardial infarction: safety report from the Throm-
bolysis and Thrombin Inhibition in Myocardial Infarction (TIMI) 9A trial.
Circulation 1994; 90:1624–1630.
Begelman SM, Deitcher SR. Outpatient management of venous thromboembolic dis-
ease with subcutaneous lepirudin: a case report. J Thromb Thrombolysis 2002;
13:183–185.
Bircher AJ, Czendlik CH, Messmer SL, Muller P, Howard H. Acute urticaria caused
by subcutaneous recombinant hirudin: evidence for an IgG-mediated hyper-
sensitivity reaction. J Allergy Clin Immunol 1996; 98:994–996.
Bove CM, Casey B, Marder VJ. DDAVP reduces bleeding during continued hirudin
administration in the rabbit. Thromb Haemost 1996; 75:471–475.
Bucha E, Nowak G, Czerwinski R, Thieler H. R-hirudin as anticoagulant in regular
hemodialysis therapy: finding of therapeutic R-hirudin blood/plasma concen-
trations and respective dosages. Clin Appl Thromb Hemost 1999; 5:164–170.
430 Greinacher
Butler KD, Dolan SL, Talbot MD, Wallis RB. Factor VII and DDAVP reverse the
effect of recombinant desulphatohirudin (CGB 39393) on bleeding in the rat.
Callas DD, Hoppensteadt DA, Iqbal O, Rubsamen K, Fareed J. Ecarin clotting time
(ECT) is a reliable method for the monitoring of hirudins, argatroban, efegatran
and related drugs in therapeutic and cardiovascular indications [abstr]. Blood
1995; 86(10 suppl 1):866a.
Clore GM, Sukumaran DK, Nilges M, Zarbock J, Gronenborn AM. The confor-
mation of hirudin in solution. A study using nuclear magnetic resonance, dis-
tance geometry and restrained molecular dynamics. EMBO J 1987; 6:529–553.
Dager WE, White RH. Use of lepirudin in patients with heparin-induced throm-
bocytopenia and renal failure requiring hemodialysis. Ann Pharmacother 2001;
35:885–890.
Dager WE, White RH. Treatment of heparin-induced thrombocytopenia. Ann Phar-
macother 2002; 36:489–503.
de Denus S, Spinler SA. Clinical monitoring of direct thrombin inhibitors using the
ecarin clotting time. Pharmacotherapy 2002; 22:433–435.
Deitcher SR, Topoulos AP, Bartholomew JR, Kichuk-Chrisant MR. Lepirudin anti-
coagulation for heparin-induced thrombocytopenia. J Pediatr 2002; 140:264–
266.
Dickneite G, Friesen HJ, Kumpe G, Reers M. Reduction of rH induced bleeding in
pigs by the administration of von Willebrand factor. Platelets 1996; 7:283–290.
Dickneite G, Nicolay U, Friesen HJ, Reers M. Development of an anti-bleeding agent
for recombinant hirudin induced skin bleeding in the pig. Thromb Haemost
1998; 80:192–198.
Diehl KH, Römisch J, Hein B, Jessel A, Ronneberger H, Paques EP. Investigation of
activated prothrombin complex concentrate as potential hirudin antidote in
animal models. Haemostasis 1995; 25:182–192.
Direct Thrombin Inhibitors Trialists’ Collaborative Group. Direct thrombin inhibi-
tors in acute coronary syndrome: principal results of a meta-analysis based on
individual patients’ data. Lancet 2002; 359:294–302.
Eichler P, Budde U, Haas S, Kroll H, Loreth RM, Meyer O, Pachmann U, Potzsch B,
Schabel A, Albrecht D, Greinacher A. First workshop for detection of heparin-
induced antibodies: validation of the heparin-induced platelet activation test
(HIPA) in comparison with a PF4/heparin ELISA. Thromb Haemost 1999;
81:625–629.
Eichler P, Friesen HJ, Lubenow N, Jaeger B, Greinacher A. Antihirudin antibodies
in patients with heparin-induced thrombocytopenia treated with lepirudin:
incidence, effects on aPTT, and clinical relevance. Blood 2000; 96:2373–2378.
Eichler P, Lubenow N, Greinacher A. Results of the third prospective study of treat-
ment with lepirudin in patients with heparin-induced thrombocytopenia (HIT)
Eriksson BI, Ekman S, Kälebo P, Zachrisson B, Bach D, Close P. Prevention of deep-
vein thrombosis after total hip replacement: direct thrombin inhibition with
recombinant hirudin, CGP 39393. Lancet 1996; 347:635–639.
Eriksson BI, Wille-Jorgensen P, Kalebo P, Mouret P, Rosencher N, Bosch P, Baur M,

Ekman S, Bach D, Lindbratt S, Close P. A comparison of recombinant hirudin
with a low-molecular-weight heparin to prevent thromboembolic complications
after total hip replacement. N Engl J Med 1997; 337:1329–1335.
Fabrizio MC. Use of ecarin clotting time (ECT) with lepirudin therapy in heparin-
induced thrombocytopenia and cardiopulmonary bypass. J Extra Corpor Tech-
nol 2001; 33:117–125.
Farner B, Eichler P, Kroll H, Greinacher A. A comparison of danaparoid and lepiru-
din in heparin-induced thrombocytopenia. Thromb Haemost 2001; 85:950–957.
Fischer KG. Hirudin in renal insufficiency. Semin Thromb Hemost 2002; 28:467–482.
Fischer KG, van de Loo A, Böhler J. Recombinant hirudin (lepirudin) as anticoa-
gulant in intensive care patients treated with continuous hemodialysis. Kidney
Int 1999; 56(suppl 72):S46–S50.
Fischer KG, Liebe V, Hudek R, Piazolo L, Haase KK, Borggrefe M, Huhle G. Anti-
hirudin antibodies alter pharmacokinetics and pharmacodynamics of recombi-
nant hirudin. Thromb Haemost 2003; 89:973–982.
Follis F, Schmidt CA. Cardiopulmonary bypass in patients with heparin-induced
thrombocytopenia. Ann Thorac Surg 2000; 70:2173–2181.
Frank RD, Farber H, Stefanidis I, Lanzmich R, Kierdorf HP. Hirudin elimination by
hemofiltration: a comparative in vitro study of different membranes. Kidney Int
1999; 72(suppl):S41–S45.
Glusa E. Pharmacology and therapeutic applications of hirudin, a new anticoagulant.
Kidney Int 1998; 64(suppl):S54–S56.
Glusa E, Markwardt F. Platelet functions in recombinant hirudin-anticoagulated
blood. Haemostasis 1990; 20:112–118.
diagnosing heparin-associated thrombocytopenia. Thromb Haemost 1991; 66:
734–736.
Greinacher A, Eckhardt T, Mußmann J, Mueller-Eckhardt C. Pregnancy complicated
by heparin associated thrombocytopenia: management by a prospectively in
vitro selected heparinoid (Org10172). Thromb Res 1993; 71:123–127.
safe and effective anticoagulation in patients with heparin-induced thrombo-
cytopenia: a prospective study. Circulation 1999a; 99:73–80.
Greinacher A, Janssens U, Berg G, Böck M, Kwasny H, Kemkes-Matthes B, Eichler
P, Völpel H, Pötzsch B, Luz M for the Heparin-Associated Thrombocytopenia
Study (HAT) investigators. Lepirudin (recombinant hirudin) for parenteral
tion 1999b; 100:587–593.
bocytopenia with thromboembolic complications: meta-analysis of 2 prospec-
Greinacher A, Eichler P, Albrecht D, Strobel U, Pötzsch B, Eriksson BI. Antihirudin
432 Greinacher
antibodies following low-dose subcutaneous treatment with desirudin for

thrombosis prophylaxis after hip-replacement surgery: incidence and clinical
relevance. Blood 2003a; 101:2617–2619.
Greinacher A, Eichler P, Lubenow N. Anaphylactic reactions associated with lep-
irudin in patients with heparin-induced thrombocytopenia (HIT). Circulation
2003; 108:2062–2065.
Haas G. Über Versuche der Blutauswaschung am Lebenden mit Hilfe der Dialyse.
Klin Wochenschr 1924; 4:13–14.
Hach-Wunderle V. Hirudin in der Gefaesschirurgie. In: Greinacher A, ed. Hirudin in
der vaskulaeren Medizin. Bremen: Uni-Med Verlag, 2001:76–77.
Hafner G, Roser M, Nauck M. Methods for the monitoring of direct thrombin
inhibitors. Semin Thromb Hemost 2002; 28:425–430.
Hermann JPR, Kutryk MJV, Serruys PW. Clinical trials of direct thrombin inhibitors
during invasive procedures. Thromb Haemost 1997; 78:367–376.
Hogg PJ, Jackson CM. Fibrin monomer protects thrombin from inactivation by
heparin-antithrombin III: implications for heparin efficacy. Proc Natl Acad Sci
USA 1989; 86:3619–3623.
Huhle G, Song X, Wang LC, Hoffman U, Harenberg J. Generation and disappearance
of antihirudin antibodies during treatment with r-hirudin. Fibrinol Proteol
1998; 12(suppl 2):91–113.
Huhle G, Hoffmann U, Hoffmann I, Liebe V, Harenberg JF, Heene DL. A new ther-
apeutic option by subcutaneous recombinant hirudin in patients with heparin-
induced thrombocytopenia type II: a pilot study. Thromb Res 2000a; 99:325–
334.
Huhle G, Geberth M, Hoffmann U, Heene DL, Harenberg J. Management of heparin-
associated thrombocytopenia in pregnancy with subcutaneous r-hirudin. Gyne-
col Obstet Invest 2000b; 49:67–69.
Huhle G, Liebe V, Hudek R, Heene DL. Anti-r-hirudin antibodies reveal clinical
relevance through direct functional inactivation of r-hirudin or prolongation of
r-hirudin’s plasma halflife. Thromb Haemost 2001; 85:936–938.
Ibbotson SH, Grant PJ, Kerry R, Findlay VS, Prentice CRM. The influence of
infusions of 1-desamino-8-D-arginine vasopressin (DDAVP) in vivo on the
anticoagulant effect of recombinant hirudin (CGP 39393) in vitro. Thromb
Haemost 1991; 65:64–66.
Irami MS, White HJ Jr, Sexon RG. Reversal of hirudin-induced bleeding diathesis by
prothrombin complex concentrate. Am J Cardiol 1995; 75:422–423.
Jagadeeswaran P, Liu YC, Eddy CA. Effects of hirudin (thrombin specific inhibitor) in
zebrafish embryos: a developmental role for thrombin. Blood Cells Mol Dis
1997; 23:410–414.
Johnston N, Jessen ME, DiMaio M, Douglass DS. The emergency use of recombinant
hirudin in cardiopulmonary bypass. J Extra Corpor Thrombosehnol 1999; 31:
211–215.
Kaiser B, Markwardt F. Antithrombotic and haemorrhagic effects of synthetic and
naturally occurring thrombin inhibitors. Thromb Res 1986; 43:613–620.
Kern H, Ziemer S, Kox WJ. Bleeding after intermittend or continuous r-hirudin
during CVVH. Intensive Care Med 1999; 25:1311–1314.
King DJ, Kelton JG. Heparin-associated thrombocytopenia. Ann Intern Med 1984;
100:535–540.
Kornalik F, Blombäck B. Prothrombin activation induced by ecarin—a prothrombin
converting enzyme from Echis carinatus venom. Thromb Res 1975; 6:57–63.
Koster A, Kuppe H, Hetzer R, Sodian R, Crystal GJ, Mertzlufft F. Emergent cardio-
pulmonary bypass in five patients with heparin-induced thrombocytopenia type
II employing recombinant hirudin. Anesthesiology 1998; 89:777–780.
Koster A, Hansen R, Grauhan O, Hausmann H, Bauer M, Hetzer R, Kuppe H,
Mertzlufft F. Hirudin monitoring using the TAS ecarin clotting time in patients
with heparin-induced thrombocytopenia type II. J Cardiothorac Vasc Anesth
2000a; 14:249–252.
Koster A, Hansen R, Kuppe H, Hetzer R, Crystal GJ, Mertzlufft F. Recombinant
hirudin as an alternative for anticoagulation during cardiopulmonary bypass in
patients with heparin-induced thrombocytopenia type II: a 1-year experience in
57 patients. J Cardiothorac Vasc Anesth 2000b; 14:243–248.
Latham P, Revelis AF, Joshi GP, DiMaio JM, Jessen ME. Use of recombinant hirudin
in patients with heparin-induced thrombocytopenia with thrombosis requiring
cardiopulmonary bypass. Anaesthesiology 2000; 92:263–266.
Liebe V, Brückmann M, Fischer KG, Haase KK, Borggrefe M, Huhle G. Biological
relevance of anti-recombinant hirudin antibodies–results from in vitro and in
vivo studies. Semin Thromb Hemost 2002; 28:483–489.
Lindhoff-Last E, Bauersachs R. Heparin-induced thrombocytopenia-alternative anti-
coagulation in pregnancy and lactation. Semin Thromb Haemost 2002; 28:439–
445.
Lindhoff-Last E, Willeke A, Thalhammer C, Nowak G, Bauersachs R. Hirudin treat-
ment in a breastfeeding woman [letter]. Lancet 2000a; 355:467–468.
Lindhoff-Last E, Piechottka GP, Rabe F, Bauersachs R. Hirudin determination in
plasma can be strongly influenced by the prothrombin level. Thromb Res 2000b;
100:55–60.
Liu H, Fleming NW, Moore PG. Anticoagulation for patients with heparin-induced
thrombocytopenia using recombinant hirudin during cardiopulmonary bypass.
J Clin Anesth 2002; 14:452–455.
Longrois D, de Maistre E, Bischoff N, Dopff C, Meistelman C, Angioi M, Lecompte T.
Recombinant hirudin anticoagulation for aortic valve replacement in heparin-
induced thrombocytopenia. Can J Anaesth 2000; 47:255–260.
Lubenow N, Greinacher A. Hirudin in heparin-induced thrombocytopenia. Semin
Thromb Hemost 2002; 28:431–438.
Lubenow N, Eichler P, Leitz T, Greinacher A. Meta-analysis of three prospective
studies of lepirudin in the prevention of thrombosis in patients with heparin-
induced thrombocytopenia [abstr]. Blood 2002a; 100(suppl 1):501a–502a.
Lubenow N, Eichler P, Greinacher A. Results of a large drug-monitoring program
confirms the safety and efficacy of Refludan (lepirudin) in patients with im-
mune-mediated heparin-induced thrombocytopenia (HIT) [abstr]. Blood 2000b;
100(suppl 1):502a.
Markwardt F. Development of hirudin as an antithrombotic agent. Semin Thromb
Hemost 1989; 15:269–282.
434 Greinacher
Markwardt F. Hirudin: the promising antithrombotic. Cardiovasc Drug Rev 1992; 10:
211–232.
Markwardt F, Fink G, Kaiser B, Klocking HP, Nowak G, Richter M, Sturzebecher J.
Pharmacological survey of recombinant hirudin. Pharmazie 1988; 43:202–207.
Mehta SR, Eikelboom JW, Rupprecht H-.J, Lewis BS, Natarajan MK, Yi C, Pogue J,
Yusuf S. Efficacy of hirudin in reducing cardiovascular events in patients with
acute coronary syndrome undergoing early percutaneous coronary interven-
tion. Eur Heart J 2002; 23:117–123.
Meyer BJ, Badimon JJ, Chesebro JH, Fallon JT, Fuster V, Badimon L. Dissolution of
mural thrombus by specific thrombin inhibition with r-hirudin: comparison
with heparin and aspirin. Circulation 1998; 97:681–685.
Nand S, Wong W, Yuen B, Yetter A, Schmulbach E, Gross Fisher S. Heparin-induced
thrombocytopenia with thrombosis: incidence, analysis of risk factors, and
clinical outcomes in 108 consecutive patients treated at a single institution. Am J
Hematol 1997; 56:12–16.
Neuhaus KL, von Essen R, Tebbe U, Jessel A, Heinrichs H, Maurer W, Doring W,
Harnjanz D, Kotter V, Kalhammer E, et al. Safety observations from the pilot
phase of the randomized r-Hirudin for Improvement of Thrombolysis (HIT-III)
study: a study of the Arbeitsgemeinschaft Leitender Kardiologischer Kranken-
hausärzte (ALKK). Circulation 1994; 90:1638–1642.
Neuhaus KL, Molhoek GP, Zeymer U, Tebbe U, Wegschieder K, Schroder R, Camez
A, Laarman GJ, Grollier GM, Lok DJ, Kuckuck H, Lazarus P. Recombinant
hirudin (lepirudin) for the improvement of thrombolysis with streptokinase in
patients with acute myocardial infarction: results of the HIT-4 trial. J Am Coll
Cardiol 1999; 34:966–973.
Nguyen TN, Gal P, Ransom JL, Carlos R. Lepirudin use in a neonate with heparin-
induced thrombocytopenia. Ann Pharmacother 2003; 37:229–233.
Nishida S, Fujita T, Kohno N, Atoda H, Morita T, Takeya H, Kido I, Paine MJ,
Kawabata S, Iwanaga S. cDNA cloning and deduced amino acid sequence of
prothrombin activator (ecarin) from Kenyan Echis carinatus venom. Biochem-
istry 1995; 34:1771–1778.
Novoa E, Seegers WH. Mechanisms of alpha-thrombin and beta-thrombin-E for-
mation: use of ecarin for isolation of meizothrombin 1. Thromb Res 1980; 18:
657–668.
Nowak G. Clinical monitoring of hirudin and direct thrombin inhibitors. Semin
Thromb Hemost 2001; 27:537–541.
Nowak G, Bucha E. Prothrombin conversion intermediate effectively neutralizes toxic
levels of hirudin. Thromb Res 1995; 80:317–325.
Nowak G, Bucha E. Quantitative determination of hirudin in blood and body fluids.
Semin Thromb Haemost 1996; 22:197–202.
Nowak G, Bucha E, Gööck T, Prasa D, Thieler H. Pharmakokinetik von Hirudin bei
gestörter Nierenfunktion. Haemostaseologie 1991; 11:152–157.
Nowak G, Bucha E, Gööck T, Thieler H, Markwardt F. Pharmacology of r-hirudin in
renal impairment. Thromb Res 1992; 66:707–715.
Nowak G, Bucha E, Brauns I, Czerwinski R. Anticoagulation with r-hirudin in regular
haemodialysis with heparin-induced thrombocytopenia (HIT II). The first long

term application of r-hirudin in a haemodialysis patient. Wien Klin Wochenschr
1997; 109:354–358.
Organization to Assess Strategies for Ischemic Syndromes (OASIS) Investigators.
Comparison of the effects of two doses of recombinant hirudin compared with
heparin in patients with acute myocardial ischemia without ST elevation: a pilot
study. Circulation 1997; 96:769–777.
Organization to Assess Strategies for Ischemic Syndromes (OASIS-2) Investigators.
Effects of recombinant hirudin (lepirudin) compared with heparin on death,
myocardial infarction, refractory angina, and revascularisation procedures in
patients with acute myocardial ischemia without ST elevation: a randomised
trial. Lancet 1999; 353:429–438.
Parent F, Bridey F, Dreyfus M, Musset D, Grimon G, Duroux P, Meyer D, Simon-
neau G. Treatment of severe thromboembolism with intravenous hirudin (HBW
023): an open pilot study. Thromb Haemost 1993; 70:386–388.
Pieters J, Lindhout T, Hemker HC. In situ-generated thrombin is the only enzyme
that effectively activates factor VIII and factor V in thromboplastin-activated
plasma. Blood 1989; 74:1021–1024.
Pötzsch B, Madlener K, Seelig C, Riess CF, Greinacher A, Müller-Berghaus G.
Monitoring of r-hirudin anticoagulation during cardiopulmonary bypass—
assessment of the whole blood ecarin clotting time. Thromb Haemost 1997a;
77:920–925.
Pötzsch B, Hund S, Madlener K, Unkrig C, Müller-Berghaus G. Monitoring of re-
combinant hirudin: assessment of a plasma-based ecarin clotting time assay.
Thromb Res 1997b; 86:373–383.
Refludan Package Insert. Monville, NJ: Berlex Laboratories, 2002.
Riess FC, Löwer C, Seelig C, Bleese N, Kormann J, Müller-Berghaus G, Pötzsch B.
Recombinant hirudin as a new anticoagulant during cardiac operations instead
of heparin: successful for aortic valve replacement in man. J Thorac Cardiovasc
Surg 1995; 110:265–267.
Riess FC, Pötzsch B, Bader R, Bleese N, Greinacher A, Löwer C, Madlener K, Müller-
Berghaus G. A case report on the use of recombinant hirudin as an anticoa-
gulant for cardiopulmonary bypass in open heart surgery. Eur J Cardio-thorac
Surg 1996; 10:386–388.
Rupprecht HJ, Terres W, Özbek C, Luz M, Jessel A, Hafner G, vom Dahl J, Kromer
EP, Prellwitz W, Meyer J. Recombinant hirudin (HBW 023) prevents troponin
T release after coronary angioplasty in patients with unstable angina. J Am Coll
Cardiol 1995; 26:1637–1642.
Schiele F, Vuillemenot A, Kramarz P, Kieffer Y, Soria J, Soria C, Camez A, Mirshahi
MC, Bassand JP. A pilot study of subcutaneous recombinant hirudin (HBW
023) in the treatment of deep vein thrombosis. Thromb Haemost 1994; 71:558–
562.
436 Greinacher
Schiele F, Lindgaerde F, Eriksson H, Bassand JP, Wallmark A, Hansson PO, Grollier

G, Sjo M, Moia M, Camez A, Smyth VWalker M for the International Mul-
ticenter Hirudin Study Group. Subcutaneous recombinant hirudin (HBW 023)
versus intravenous sodium heparin in treatment of established acute deep vein
thrombosis of the legs: a multicentre prospective dose-ranging randomized trial.
Schiffmann H, Unterhalt M, Harms K, Figulla HR, Völpel H, Greinacher A. Erfol-
greiche Behandlung einer Heparin-induzierten Thrombozytopenie Typ II im
Kindesalter mit rekombinantem Hirudin. Monatsschr Kinderheildk 1997; 145:
606–612.
Shepherd MF. Dosage of lepirudin in renal failure. Am J Health Syst Pharm 2002;
59:77–78.
Song X, Huhle G, Wang L, Hoffmann U, Harenberg J. Generation of anti-hirudin an-
tibodies in heparin-induced thrombocytopenic patients treated with r-hirudin.
Circulation 1999; 100:1528–1532.
Stone S, Hofsteenge J. The kinetics of the inhibition of thrombin by hirudin. Bio-
chemistry 1986; 25:4622–4628.
Sukumaran DK, Clare GM, Presus A, Zarbock J, Gronenborn AM. Proton nuclear
magnetic resonance study of hirudin: resonance assignment and secondary
structure. Biochemistry 1987; 26:333–338.
Vanholder RC, Camez AA, Veys NM, Soria J, Mirshahi M, Soria C, Ringoir S.
Recombinant hirudin: a specific thrombin inhibiting anticoagulant for hemo-
dialysis. Kidney Int 1994; 45:1754–1759.
Vanholder R, Camez A, Veys N, Van Loo A, Dhondt AM, Ringoir S. Pharma-
cokinetics of recombinant hirudin in hemodialyzed end-stage renal failure pa-
tients. Thromb Haemost 1997; 77:650–655.
gery. Ann Thorac Surg 2003;76:638–648. [Corrected article to be reprinted in its
Am J Med 1996; 101:502–507.
Weitz JI, Hudoba M, Massel D, Maraganore J, Hirsh J. Clot-bound thrombin is
protected from inhibition by heparin-antithrombin III but is susceptible to in-
activation by antithrombin III-independent inhibitors. J Clin Invest 1990; 86:
385–391.
Weitz JI, Leslie B, Hudoba M. Thrombin binds to soluble fibrin degradation products
where it is prothromboseted from inhibition by heparin-antithrombin but sus-
ceptible to inactivation by antithrombin-independent inhibitors. Circulation
1998; 97:544–552.
Wittkowsky AK, Kondo LM. Lepirudin dosing in dialysis-dependent renal failure.
Pharmacotherapy 2000; 20:1123–1128.
16
Argatroban Therapy in
Bruce E. Lewis
Loyola University Medical Center, Maywood, Illinois, and Catholic
Health Partners, Chicago, Illinois, U.S.A.
Marcie J. Hursting
Clinical Science Consulting, Potomac, Maryland, U.S.A.
I. INTRODUCTION
Heparin-induced thrombocytopenia (HIT) is an immune-mediated syndrome

characterized by thrombocytopenia, which can be isolated or associated with
thrombotic events (see Chaps. 1–4). Complications of HIT span a spectrum of
venous and arterial thromboembolic events, including deep venous throm-
bosis, pulmonary embolism, myocardial infarction, thrombotic stroke, and
limb artery occlusion requiring amputation (Warkentin, 2003). The elimina-
tion of all heparin sources and the initiation of alternate anticoagulation
are recommended for treating patients with HIT (Hirsh et al., 2001). Not
infrequently, thrombosis is the first manifestation of HIT. Even among
patients with isolated thrombocytopenia who are managed by heparin ces-
sation alone, approximately 25–50% develop new thrombosis (Warkentin
and Kelton, 1996; Wallis et al., 1999; Lewis et al., 2003). Furthermore, many
patients with HIT require ongoing anticoagulation for underlying medical
conditions.
Two direct thrombin inhibitors, argatroban and r-hirudin (lepirudin)
(see Chap. 15), are approved in the United States for use as anticoagulants in
patients with HIT. The specific indications of these agents differ somewhat.
437
438 Lewis and Hursting
Lepirudin is indicated as anticoagulation for patients having HIT with

associated thromboembolic disease in order to prevent further thromboem-
bolic complications (Refludank Prescribing Information, U.S., 1998). Arga-
troban is indicated as an anticoagulant for prophylaxis or treatment of
thrombosis in patients with HIT (Argatroban Prescribing Information,
U.S., 2002). In other words, while both agents are approved for treating
HIT patients with thrombosis, only argatroban is approved for managing
HIT patients with isolated thrombocytopenia. Argatroban is also approved
in Canada as an anticoagulant for patients with HIT who, in the opinion of
their attending physician, require anticoagulation, and in the United States as
an anticoagulant for patients with, or at risk for, HIT undergoing percuta-
neous coronary intervention (PCI).
In this chapter, the clinical pharmacology of argatroban is reviewed,
together with its clinical utility as an anticoagulant in patients with HIT.
Historically, argatroban was initially known worldwide as ‘‘MD-805’’ and
then under the trademark ‘‘Novastan.’’ In 2000, the U.S. Food and Drug
Administration disallowed the trademark Novastan in the United States
because of potential similarities with other named products. Hence, argatro-
ban (generic name) is now marketed in the United States without a trademark
and under the name ‘‘Argatroban’’ (with a capital A). However, the trade-
mark Novastan continues to be used in some other countries.
II. ARGATROBAN
A. Chemical Description
Argatroban is a synthetic direct thrombin inhibitor derived from L-arginine
(Okamoto and Hijikata, 1981; Kikumoto et al., 1984). Its chemical structure is
shown in Fig. 1. Argatroban (molecular weight, 526.66) consists of a mixture
of 21-(R) and 21-(S) stereoisomers in a ratio of approximately 65:35 (Rawson
et al., 1993). There is no interconversion between these stereoisomers.
B. Clinical Pharmacology
Mechanism of Action
Argatroban is a potent and selective inhibitor of thrombin (Okamoto and
Hijikata, 1981; Kikumoto et al., 1984). Argatroban was developed using the
approach of rational drug design through the mimicry of substrates of
thrombin. It displays an inhibitory constant (Ki) of 0.04 Amol/L for thrombin
and has little or no effect on related serine proteases (K1 values of 5 Amol/L for
Argatroban Therapy in Heparin-Induced Thrombocytopenia 439
Figure 1 Chemical structure of argatroban. Its chemical name is 1-[5-[(aminoimino

methyl)amino]-1-oxo-2-[[(1,2,3,4-tetrahydro-3-methyl-8-quinolinyl)sulfonyl]amino]-
pentyl]-4-methyl-2-piperidinecarboxylic acid, monohydrate, and its molecular weight
is 526.66.
trypsin, 210 Amol/L for factor Xa, and 800 Amol/L for plasmin) (Kikumoto
et al., 1984). Argatroban exerts its anticoagulant effects in the absence of
any cofactor by inhibiting thrombin-catalyzed or -induced reactions, such as
fibrin formation, the activation of factors V, VIII, and XIII, and platelet
aggregation (Okamoto and Hijikata-Okunomiya, 1993).
Argatroban effectively inhibits free and clot-bound thrombin (Berry
et al., 1994; Hantgan et al., 1998). Argatroban is over 500-fold more potent
than r-hirudin in its relative ability to inhibit clot-bound versus free thrombin
(Berry et al., 1994). The lower molecular weight of argatroban, compared
with hirudin, may allow it better accessibility to thrombin incorporated
within the clot. The ability to inhibit effectively clot-bound thrombin may
be of particular benefit in treating hypercoagulable states such as HIT and
also in reducing extension of existing thromboses.
Structural studies have shown that argatroban binds tightly to throm-
bin by inserting the dual hydrophobic moieties on its arginine backbone into
deep clefts near the thrombin active site (Banner and Hadvary, 1993). Thus,
physiological substrates of thrombin are sterically hindered from access to the
catalytic pocket of thrombin. Figure 2 (see color insert) shows a model of the
interaction between argatroban and thrombin. This interaction is reversible,
unlike the irreversible interaction between r-hirudin and thrombin (Chap. 15).
The combination of reversible binding and a short elimination half-life (see
next subsection) may improve the ability to control anticoagulation in the
intensive care setting.
Figure 2 Model of the interaction between argatroban and thrombin. (See color
insert.)
Distribution, Metabolism, and Excretion

Argatroban distributes mainly in the extracellular fluid, as evidenced by a
steady state volume of distribution of 174 mL/kg (Swan and Hursting, 2000).
It is 54% serum protein-bound (Tatsuno et al., 1986).
Unlike other commercially available direct thrombin inhibitors, arga-
troban undergoes no significant renal clearance. The main route of metabo-
lism is hydroxylation and aromatization of the 3-methyltetrahydroquinoline
ring in the liver (Izawa et al., 1986). In vitro, the human liver microsomal
cytochrome P450 3A4/5 (CYP3A4/5) catalyzes the formation of each of the
four known metabolites. In plasma, unchanged argatroban is the major
component, while the primary metabolite (M1), which has three- to fivefold
less activity than argatroban, is present at concentrations that are 0–20% of
that of the parent drug (Ahsan et al., 1997). The other metabolites have not
been detected in plasma or feces and are found only in very low quantities in
urine. These data, together with the lack of effect of erythromycin, a potent
CYP3A4/5 inhibitor, on argatroban pharmacokinetics (Tran et al., 1999)
suggest that CYP3A4/5-mediated metabolism is not an important pathway
in vivo.
The plasma clearance rate of argatroban is approximately 5.1 mL/min/
kg for infusion doses up to 40 Ag/kg/min in healthy volunteers (Swan et al.,
2000). Its elimination half-life is 39–51 min (Swan and Hursting, 2000),
which is somewhat less than the 1–2 h half-life of r-hirudin (Vanholder et al.,
1997). Argatroban is excreted primarily in the feces, presumably by biliary
secretion.
Pharmacokinetic–Pharmacodynamic Relationship
The pharmacokinetic and pharmacodynamic profiles of intravenously ad-
ministered argatroban are consistent with an anticoagulant agent that is
predictable, has a fast onset of action, and is rapidly eliminated (Swan and
Hursting, 2000; Swan et al., 2000).
The anticoagulant effects of argatroban are routinely monitored using
the activated partial thromboplastin time (aPTT). Higher levels of antico-
agulation, such as that required during interventional procedures, are moni-
tored using the activated clotting time (ACT). Argatroban also increases in a
dose-dependent fashion the prothrombin time (PT)/ International Normal-
ized Ratio (INR), thrombin time (TT), and ecarin clotting time (ECT) (Naga-
sawa et al., 1981; Clark et al., 1991; Walenga et al., 1999a; Swan et al., 2000;
Sheth et al., 2001). High-performance liquid chromatography (Rawson et al.,
1993; Walenga et al., 1999a) and liquid chromatography/tandem mass spec-
trometry (Tran et al., 1999) methods for measuring plasma argatroban are
described but are not practical (or needed) for routine clinical monitoring.
Immediately upon starting argatroban infusion, anticoagulant effects
are produced as plasma argatroban concentrations begin to rise. Steady-
state levels of both drug and anticoagulant effect typically are attained within
1–3 h (faster when a loading bolus is administered) and maintained with low
intra- and intersubject variability until the infusion is discontinued or the
dosage adjusted. Plasma drug concentrations increase proportionally with
doses up to 40 Ag/kg/min and are well correlated with steady-state antico-
agulant effects. The relationship at steady state between argatroban dose up
to 10 Ag/kg/min, plasma argatroban concentration, and anticoagulant effect
(aPTT) is shown in Fig. 3. On stopping infusion, plasma argatroban concen-
trations decline rapidly (half-life of 39–51 min), and anticoagulant effects
return to pretreatment values with similar effect half-lives (Swan et al., 2000).
Special Populations
Age, gender, and renal function exert no clinically significant effects on the
pharmacokinetics or pharmacodynamics of argatroban. Patients with mod-
erate hepatic impairment (Child–Pugh score>6), compared with healthy
volunteers, have an approximate fourfold decrease in drug clearance (to 1.5
mL/min/kg) and an approximate threefold increase in elimination half-life (to
152 min) (Swan and Hursting, 2000). Owing to the decreased clearance, a
Figure 3 Relationship at steady state between argatroban dose, plasma argatroban

concentration, and anticoagulant effect (aPTT). Mean (SEM) steady-state plasma
argatroban concentrations and aPTT values are for healthy subjects (n = 9) ad-
ministered intravenous argatroban at doses between 1.25 and 10 Ag/kg/min. (Data
from Swan et al., 2000.)
fourfold downward adjustment in argatroban dosage is required for individ-

uals with moderate hepatic impairment. No adjustment in initial argatroban
dosage is needed for patients with renal impairment. In contrast, r-hirudin
requires dosage adjustment for patients with renal impairment but not for
patients with hepatic impairment (Refludank Prescribing Information, U.S.,
1998). In patients without hepatic dysfunction undergoing PCI, the pharma-
cokinetic values of argatroban are similar to those reported in healthy volun-
teers, and argatroban clearance is unaffected by age, gender, or race (Cox
et al., 2003).
Drug-Drug Interactions
No pharmacokinetic or pharmacodynamic drug interactions have been
demonstrated between argatroban and aspirin (Clarke et al., 1991), erythro-
mycin (Tran et al., 1999), acetaminophen, digoxin, or lidocaine (Inglis et al.,
2002). In practice, argatroban coadministered with these frequently used
medications should require no dosage adjustments.
No pharmacokinetic interactions have been demonstrated between
argatroban and warfarin (Brown and Hursting, 2002). However, because
argatroban is a direct thrombin inhibitor, the concomitant use of argatroban

and warfarin results in prolongation of the PT/INR beyond that produced by
warfarin alone (Hursting et al., 1999; Sheth et al., 2001). Cotherapy compared
with warfarin monotherapy exerts no additional effect on vitamin K–depen-
dent factor X levels (Sheth et al., 2001). Hence the previously established
(‘‘traditional’’) relationship between INR and bleeding risk is altered during
combination therapy. Guidelines for monitoring the transition from arga-
troban to warfarin anticoagulation are presented in Section IV.
Argatroban and a variety of drugs, including the glycoprotein (GP) IIb/
IIIa antagonists abciximab, eptifibatide, and tirofiban, have been evaluated
for chemical or physical/visual compatibility at concentrations commonly
used in practice. This is important for supporting their simultaneous admin-
istration via Y-site injection. Argatroban and eptifibatide or tirofiban are
chemically and physically compatible (Patel, 2002). Argatroban and abcix-
imab (Patel, 2002), fentanyl citrate, midazolam hydrochloride, morphine
sulfate, dopamine hydrochloride, dobutamine hydrochloride, phenylephrine
hydrochloride, atropine sulfate, hydrocortisone sodium succinate, meto-
prolol tartrate, diphenhydramine hydrochloride, verapamil hydrochloride,
norepinephrine bitartrate, or diltiazem hydrochloride (Hartman et al., 2002)
are physically/visually compatible; their chemical stability remains to be
established.
C. Other Distinguishing Features

Lack of Cross-Reactivity with HIT Antibody
In contrast with low molecular weight heparin and danaparoid, the direct
thrombin inhibitors, including argatroban, hirudin, and bivalirudin, bear no
resemblance to heparin, do not cross-react with HIT antibodies, and have not
been associated with potentiation of HIT (Walenga et al., 1996).
Lack of Drug-Specific Antibody

Prolonged or repeated exposure to argatroban does not result in the gener-
ation of antibodies that alter its anticoagulant activity (Walenga et al., 2002).
That is, there is no enhancement or suppression of anticoagulant response
among individuals receiving prolonged or repeated administration. This has
been shown in healthy volunteers, HIT patients, and HIT patients undergoing
PCI, and in the postmarketing safety surveillance of over 4800 patients
treated in Japan between 1991 and 1998 with argatroban anticoagulation
(Walenga et al., 2002). In contrast, approximately 50% of r-hirudin (lepir-
udin)-treated patients develop drug-specific antibodies (Song et al., 1999;
Eichler et al., 2000) (see Chap. 15), and levels of these antibodies increase
upon reexposure to lepirudin (Harenberg et al., 2000). Among those who

develop antihirudin IgG antibodies during treatment, the anticoagulant ac-
tivity of lepirudin is mildly enhanced in approximately 45% of patients and
suppressed in approximately 6% of patients, requiring careful monitoring
and probable dosage adjustment (Eichler et al., 2000).
D. Reversal of Argatroban
Argatroban has a gentle dose-response relationship that offers a wide margin
of safety during dose titration (see Fig. 3). However, as with any anticoag-
ulant, bleeding is a major safety concern. Excessive anticoagulation, with or
without bleeding, may be controlled by discontinuing argatroban or decreas-
ing its infusion dose. Anticoagulant parameters generally return to baseline
within 2–4 h after discontinuation of argatroban (Swan et al., 2000; Swan and
Hursting, 2000). This reversal takes longer (at least 6 h and up to more than
20 h) in patients with hepatic impairment (Swan and Hursting, 2000).
There is no specific antidote to argatroban available. If life-threatening
bleeding occurs and excessive plasma levels of argatroban are suspected,
argatroban should be discontinued immediately, and the patient should be
provided symptomatic and supportive therapy. Argatroban can be cleared,
albeit slowly, using high flux dialysis membranes (Murray et al., 2003), which
suggests a possible means to facilitate its removal if needed urgently. Recom-
binant factor VIIa has been suggested as a possible pharmacologic agent for
treating severe bleeding in this setting (Alving, 2002), although this remains to
be evaluated.
E. Clinical Use of Argatroban

In addition to studies conducted in patients with HIT, argatroban has been
evaluated in patients with acute myocardial infarction (Théroux, 1997; Jang
et al., 1999; Vermeer et al., 2000), unstable angina pectoris (Gold et al., 1993),
peripheral arterial obstructive disease (Matsuo et al., 1995), or stroke (Koba-
yashi and Tazaki, 1997; LaMonte, 2003), and in patients undergoing PCI
(Herrman et al., 1996; Jang et al., 2003), and patients with unstable angina
pectoris (Gold et al., 1993) or hemodialysis (Murray et al., 2003). In each of
these populations, argatroban produces predictable anticoagulant effects and
is generally safe and well tolerated. In Japan, argatroban is approved for use
in nonlacunar stroke, chronic arterial occlusion, and hemodialysis of patients
with acquired or congenital antithrombin deficiency. Argatroban is also ap-
proved in Korea for use in chronic arterial occlusion and acute cerebral
thrombosis. A prospective, controlled study of argatroban with tissue plas-
minogen activator in acute ischemic stroke is currently underway in the
United States.
In a study in patients with unstable angina (Gold et al., 1993), cessation

of argatroban was associated with ‘‘rebound’’ thrombin generation, as
measured by increased plasma levels of thrombin–antithrombin complex
(TAT) at 2 h after infusion cessation compared with the pretreatment baseline
values, and early dose-related recurrence of angina. However, there was no
rebound in plasma levels of fibrinopeptide A, which is a marker of thrombin
activity. Also, since argatroban displaces thrombin from antithrombin (Eidt
et al., 1989) and since argatroban is largely eliminated within 2 h of cessation
of infusion, an increase in TAT at the end of the infusion is reasonably
expected (Willerson and Casscells, 1993). Furthermore, reactivation of angina
after the termination of heparin occurs (Théroux et al., 1992), and heparin had
been administered in this study within 4 h of argatroban therapy. Rather than
true ‘‘rebound’’ phenomena, the renewed angina after argatroban infusion
may simply represent its recurrence after withdrawal of thrombin inhibition in
some refractory patients (Willerson and Casscells, 1993). Notably, all patients
studied (n = 43) were free of angina during argatroban therapy.
III. ARGATROBAN THERAPY OF HIT

A. Overview of Studies
The efficacy and safety of argatroban anticoagulant therapy in patients with
clinically diagnosed HIT has been evaluated in the following prospective,
multicenter, open-label studies:
ARG-911, a historical controlled study
ARG-915, a follow-on study that also used the historical control
group from ARG-911 as comparator
ARG-915X, a Phase III extension of study ARG-915 that allowed
physicians continued access to argatroban while it was under regu-
latory review
Study ARG-911 has been reported in full (Lewis et al., 2001). Topline data
from study ARG-915 (without its extension) as well as safety summaries from
ARG-911 plus ARG-915 appear in the product’s labeling information
(Argatroban Prescribing Information, U.S., 2002). Outcomes of patients
with acute HIT from study ARG-915 plus its extension, together simply
referred to as ‘‘Argatroban-915,’’ have also been reported in full (Lewis et al.,
2003). Across these studies, 754 patients received argatroban therapy on 809
separate occasions (Lewis et al., 2000).
When these studies were conducted between 1995 and 1998, no ap-
proved alternative agent was available for use as an active comparator, and a
randomized, placebo-controlled design was deemed unethical; thus historical
controls were used for comparison. The studies were similar in design with
regard to objectives, inclusion and exclusion criteria, the argatroban dosing

regimen, and assessments. In each study, patients were assigned at enrollment
to one of two prospectively defined study arms: HIT (with isolated thrombo-
cytopenia) or HIT with thrombosis (also referred to as ‘‘HIT with thrombosis
syndrome’’ or ‘‘HITTS’’). The overall study design is presented in Fig. 4.
Study Objectives
The objective of study ARG-911 was to evaluate the use of argatroban as
an anticoagulant for the prophylaxis of thrombosis in HIT patients and
the treatment of HIT patients with thrombosis. Similarly, the objective of
studies ARG-915 and ARG-915X was to evaluate the safety and efficacy of
argatroban in HIT patients, with or without thrombosis, requiring anti-
coagulation.
Study Population
Adult patients were eligible if they had a clinical diagnosis of HIT with or
without thrombosis. HIT was defined as a platelet count <100 109/L, or a
50% decrease in the platelet count after initiation of heparin therapy, with no
apparent explanation other than HIT. Patients with a documented history of
a positive HIT antibody test who needed anticoagulation were also eligible for
the HIT study arm in the absence of thrombocytopenia. Patients were
excluded if they had an unexplained aPTT greater than 2 times control at
baseline, documented coagulation disorder or bleeding diathesis unrelated to
HIT, a lumbar puncture within the prior 7 days, or a history of previous
Figure 4 Schematic of the study design for ARG-911, ARG-915, and ARG-915X.
Patients with a clinical diagnosis of HIT with or without thrombosis were eligible.
The starting dose of argatroban, 2 Ag/kg/min, was titrated to achieve an aPTT 1.5–
3.0 times the baseline aPTT (not to exceed 100 s). Outcomes over a 37-day period
were compared with those of a historical control group.
aneurysm, hemorrhagic stroke, or recent (within 6 months) thrombotic stroke

unrelated to HIT. Reentry of patients into studies ARG-915 and ARG-915X
was allowed, although outcomes from initial entries only were included in the
primary analyses to avoid potential bias.
The historical control group of ARG-911 consisted of patients at the
participating centers who met the same inclusion–exclusion criteria for the
study and who were seen prior to the initiation of the study. Controls were
treated according to the local standard of practice at the time of HIT
diagnosis, with typical treatments being heparin discontinuation and/or oral
anticoagulation (Lewis et al., 2001).
Treatment
The treatment group received an initial dose of argatroban 2 Ag/kg/min via
continuous intravenous infusion. The aPTT was measured at least 2 h later,
and dosage was adjusted (up to 10 Ag/kg/min, maximum) until the aPTT was
1.5–3 times the baseline aPTT value (not to exceed 100 s). The aPTT was
measured daily and 2 h after each dosage adjustment. Patients remained on
argatroban for up to 14 days, until the underlying condition resolved or
appropriate anticoagulation was provided with other agents.
Assessments
The primary efficacy assessment was a composite endpoint of all-cause death,
all-cause amputation, or new thrombosis within a 37-day study period.
Additional analyses included the evaluation of event rates for the components
of the composite endpoint and death due to thrombosis. Secondary efficacy
endpoints included the achievement of adequate anticoagulation (i.e., an
aPTT >1.5 times baseline) and resolution of thrombocytopenia (i.e., platelet
count >100 109/L or z1.5 times baseline by study day 3).
Major bleeding was defined as overt and associated with a hemoglobin
decrease z2 g/dL, that led to a transfusion of z2 units, or that was intra-
cranial, retroperitoneal, or into a major prosthetic joint. Other overt bleeding
was considered minor.
B. ARG-911
In study ARG-911, 304 patients having clinically diagnosed HIT (n = 160) or
HIT with thrombosis (n = 144) received argatroban at a mean dose of 2.0 Ag/
kg/min for an average of 6 days. This study also enrolled 193 historical
controls (HIT, n = 147; HIT with thrombosis, n = 46). Although not required
for enrollment, laboratory confirmation of HIT antibodies occurred in 57%
of the argatroban-treated patients and 77% of controls; the remaining in-
dividuals were either never tested or had a negative result (Lewis et al., 2001).
Efficacy
As seen in Table 1, the composite endpoint was reduced significantly in
argatroban-treated patients versus controls with HIT (25.6% vs. 38.8%, p =
0.014). In HIT with thrombosis, the composite endpoint occurred in 43.8% of
argatroban-treated patients compared with 56.5% of controls ( p = 0.13).
Significant between-group differences by time-to-event analysis of the com-
posite endpoint favored argatroban treatment in HIT ( p = 0.010, hazard
ratio = 0.60; 95% CI, 0.40–0.89) (Fig. 5a) and HIT with thrombosis ( p =
0.014, hazard ratio = 0.57; 95% CI, 0.36–0.90) (Fig. 5b).
Argatroban therapy, compared with controls, significantly reduced
death due to thrombosis in each study arm (HIT, p = 0.005; HIT with
thrombosis, p < 0.001). There were no between-group differences in all-cause
mortality. The incidence of amputation (as the most severe outcome) was
similar between groups. Argatroban therapy also significantly reduced the
percentage of patients experiencing new thrombosis in each study arm (HIT,
p < 0.001; HIT with thrombosis, p = 0.044).
Argatroban-treated patients achieved therapeutic aPTTs generally at
first measure (i.e., within 4–5 h of starting therapy) and maintained these
levels throughout infusion. Resolution of thrombocytopenia occurred by day
3 in 53% of argatroban-treated patients with HIT and 58% of patients having
HIT with thrombosis. Compared with controls, argatroban-treated patients
had a significantly more rapid rise in platelet counts.
Safety
Major bleeding occurred in 6.9% (21/304) of argatroban-treated patients,
compared with 6.7% (13/193) of historical controls. In each group, there were
two fatal bleeding events. One patient experienced a fatal intracranial
hemorrhage 4 days after discontinuation of argatroban and following uroki-
nase and warfarin therapy; one historical control also experienced a fatal
intracranial hemorrhage. Minor bleeding rates were similar between the
groups (41%). The most common adverse events among argatroban-treated
patients with HIT or HIT with thrombosis, respectively, were diarrhea (11%)
and pain (9%).
C. Argatroban-915
A total of 418 patients with acute HIT (n = 189) or HIT with thrombosis (n =
229) were prospectively treated with argatroban in study ARG-915 or its
extension (together referred to as ‘‘Argatroban-915’’) (Lewis et al., 2003). The
mean argatroban dose was 1.8 Ag/kg/min, and the mean duration of therapy
Table 1 Comparisons of Argatroban-Treated Patients with Historical Controls in ARG-911
HIT, n (%) HIT with thrombosis, n (%)
Control Argatroban Control Argatroban

Parameter (n = 147) (n = 160) p (n = 46) (n = 144) p
Composite endpointa 57 (38.8) 41 (25.6) 0.014 26 (56.5) 63 (43.8) 0.13

Odds ratio = 0.54 (95% CI, 0.33–0.88) Odds ratio = 0.60 (95% CI, 0.31–1.17)
Components by severityb
Death (all causes) 32 (21.8) 27 (16.9) 0.31 13 (28.3) 26 (18.1) 0.15
Amputation (all causes) 3 (2.0) 3 (1.9) 1.00 4 (8.7) 16 (11.1) 0.79
New thrombosis 22 (15.0) 11 (6.9) 0.027 9 (19.6) 21 (14.6) 0.49
Death due to thrombosis 7 (4.8) 0 (0.0) 0.005 7 (15.2) 1 (0.7) <0.001
Any new thrombosisc 33 (22.4) 13 (8.1) <0.001 16 (34.8) 28 (19.4) 0.044
a
All-cause death, all-cause amputation, or new thrombosis within 37-day study period.
b
Severity ranking: all-cause death > all-cause amputation > new thrombosis; patients with multiple outcomes counted once.
c
Argatroban Therapy in Heparin-Induced Thrombocytopenia
Patient counted only once if multiple events occurred.

Source: Lewis et al., 2001.
449
Figure 5 Time to first event for the composite endpoint through day 37 in study
ARG-911. Significant differences in favor of argatroban therapy were detected in
(a) the HIT study arm (argatroban group, n = 160; historical controls, n = 147) and
(b) the HIT with thrombosis study arm (argatroban group, n = 144; historical con-
trols, n = 46). (Data from Lewis et al., 2001.)
was 6 days. Comparisons were made with 185 historical controls with acute
HIT with or without thrombosis (obtained from ARG-911).
Efficacy
Efficacy results (Table 2) were confirmatory and supportive of those from
ARG-911. There were significant improvements in the composite endpoint
for argatroban-treated patients versus controls among those with HIT
(28.0% vs. 38.8%, p = 0.04) or HIT with thrombosis (41.5% vs. 56.5%,
p = 0.07). Argatroban treatment was significantly favored, compared with
control, by time-to-event analysis of the composite endpoint in HIT ( p =
0.02, hazard ratio = 0.64, 95% CI, 0.43–0.93) or HIT with thrombosis ( p =
0.008, hazard ratio = 0.56, 95% CI, 0.36–0.87).
Consistent with ARG-911, the positive benefits on the composite
endpoint were driven in main part by significant reductions in new thrombosis
( p < 0.001 in each study arm) (Table 2). There were no significant between-
group differences in all-cause mortality or amputation. Argatroban therapy
significantly reduced the incidence of death due to thrombosis in patients
having HIT with thrombosis ( p = 0.008).
Similar, predictable aPTT responses occurred in patients with HIT or
HIT with thrombosis. The target aPTT was typically achieved by first
assessment, and mean aPTT values remained generally constant throughout
the infusion. Platelet counts recovered more rapidly in argatroban-treated
patients than controls ( p < 0.001 for each study arm).
Safety
Major bleeding rates were not different between argatroban-treated patients
and controls in either study arm (Table 2). Twenty-four (5.7%) argatroban-
treated patients experienced major bleeding, including a single fatal event in a
patient hospitalized for rectal bleeding and who received urokinase. No
patient experienced an intracranial hemorrhage. Minor bleeding rates were
not different between the groups and were similar to those in ARG-911.
D. Patients with a History of HIT Requiring Acute

Anticoagulation
Across the studies of argatroban in HIT, 76 patients with a serologically
confirmed history of HIT and requiring acute anticoagulation were prospec-
tively administered argatroban on a total of 87 occasions (Matthai et al.,
2001). The most common admitting diagnoses were deep venous thrombosis
or pulmonary embolism, chest pain or acute coronary syndrome, and arterial
peripheral vascular disease. A therapeutic aPTT was achieved in 95% of the
452
Table 2 Comparisons of Argatroban-Treated Patients and Historical Controls with Acute HIT in Argatroban-915
HIT, n (%) HIT with thrombosis, n (%)
Control Argatroban Control Argatroban

Outcome (n = 139) (n = 189) pd (n = 46) (n = 229) pd
Composite Endpointa 54 (38.8) 53 (28.0) 0.04 26 (56.5) 95 (41.5) 0.07

Death (all causes)b 29 (20.9) 36 (19.0) 0.78 13 (28.3) 53 (23.1) 0.45
Death due to thrombosis 6 (4.3) 1 (0.5) 0.04 7 (15.2) 6 (2.6) 0.002
Amputation (all causes)b 4 (2.9) 8 (4.2) 0.57 5 (10.9) 34 (14.8) 0.64
New thrombosisb 32 (23.0) 11 (5.8) <0.001 16 (34.8) 30 (13.1) <0.001
Major bleedingc 12 (8.6) 10 (5.3) 0.27 1 (2.2) 14 (6.1) 0.48
Minor bleedingc 57 (41.0) 59 (31.2) 0.08 19 (41.3) 87 (38.0) 0.74
a
All-cause death, all-cause amputation, or new thrombosis within 37-day study period.
b
Outcome categories are not mutually exclusive; within a given category, a patient is counted only once if >1 event.
c
Patients with >1 event are counted only once.
d
Significance level of p < 0.05 for the primary endpoint (composite) and bleeding and p < 0.0125 for secondary endpoints (components of
composite, and death due to thrombosis).
Lewis and Hursting
patients. Adverse outcome rates were significantly less than those of patients
with active HIT (e.g., composite endpoint rates of 16.1% vs. 36.3%, p <
0.001) and were comparable with rates that have been reported for similar
patients without HIT receiving standard therapy. Among 42 treatment
courses in 40 patients with a history of HIT who had fully recovered from
their initial episode of HIT, had a normal platelet count, and had no exposure
to heparin or other parenteral anticoagulants during their hospitalization,
none had a bleeding or new thromboembolic event (Matthai et al., 2004).
Argatroban therefore provided effective anticoagulation, upon initial and
repeat exposure, in patients with a history of HIT requiring acute anti-
coagulation for a variety of indications.
E. Argatroban Reexposure
Across the prospective studies of HIT, 55 patients underwent therapy with
argatroban on more than one occasion. The argatroban dosing and duration
were similar between these patients (repeat group) and patients upon their
first exposure (initial group, n = 754). Event rates in the repeat group were
less than with those in the initial group for the composite endpoint (20% vs.
34%), new thrombosis (3.6% vs. 11.1%), and major bleeding (3.6% vs.
6.6%). The patients reexposed to argatroban had no allergic reactions or
apparent differences, relative to the initial group, in adverse experiences.
Hence, argatroban is well tolerated upon reexposure, while providing effec-
tive antithrombotic therapy for HIT (Lewis et al., 2000).
F. Discussion of Prospective Studies of Argatroban in HIT

Consistently in these studies, argatroban therapy, compared with historical
controls, produced significant benefits in clinical outcomes in patients having
HIT with or without thrombosis. Specifically, argatroban, relative to control,
was effective in reducing the composite of death, amputation, or new
thrombosis, lowering mortality from thrombosis and preventing new throm-
botic events—without increasing bleeding.
Patients were entered into these studies having a clinical diagnosis of
HIT, and laboratory confirmation of HIT was not required for their treat-
ment. This study design simulated the ‘‘real world’’ of managing HIT, wherein
it may be important to initiate alternative anticoagulation in a clinically
suspected HIT patient prior to laboratory confirmation of HIT (due perhaps
to limited availability and/or a long turnaround time for the laboratory test).
In ARG-911, HIT antibodies were demonstrated in the majority of patients
and controls. However, individuals with a clinical diagnosis of HIT were also
studied who lacked serologically confirmed HIT antibodies and in whom
thrombocytopenia may have been due to sepsis, cancer, or other causes.

Argatroban therefore is an effective antithrombotic agent in clinically diag-
nosed, albeit not always laboratory-confirmed, HIT. Argatroban also pro-
vided effective anticoagulation in patients with a history of HIT who required
acute anticoagulation for a variety of indications.
Across the studies, the overall major bleeding rate was 6.6% for
argatroban-treated patients, similar to that (6.7%) of the historical control
(Lewis et al., 2000). No patient experienced an intracranial hemorrhage while
on argatroban therapy. By comparison, the major bleeding rate for lepirudin-
treated HIT patients is 14–17% (Greinacher et al., 1999a,b). However, these
direct thrombis inhibitors remain to be compared directly, and conclusions
about their relative safety profiles cannot be reached.
These studies supported the approval of argatroban in the United States
as an anticoagulant for the prophylaxis or treatment of thrombosis in patients
with HIT, and the protocol-specified dosing regimen used in the studies was
adopted as the recommended dosing schedule. The dosing regimen evaluated
was initially selected in consideration of the established relationships between
the aPTT and adequate, safe levels of anticoagulation. Prolongation of the
aPTT to 1.5 times control is the typical minimum target for heparin anti-
coagulation in the prevention and treatment of thromboembolic disease, and
downward titration of heparin is recommended when aPTTs are approxi-
mately 3 times control (Hirsh, 1991). Also, aPTTs greater than 100 s are asso-
ciated with significant increases in major hemorrhage when another direct
thrombin inhibitor, hirudin, is used as adjunctive therapy to thrombolysis
(Antman, 1994). Hence, an argatroban dose was selected for study that was
anticipated to prolong the aPTT 1.5–3 times baseline (i.e., 2 Ag/kg/min, up to
10 Ag/kg/min), and also a maximum aPTT of 100 s was recommended to help
ensure patient safety. The mean steady-state aPTT in healthy subjects
receiving argatroban 10 Ag/kg/min is 86 s (Swan et al., 2000). That finding
contributed to the selection of 10 Ag/kg/min as a reasonable upper dose limit
for prolonging the aPTT to no more than 3 times control, without exceeding
100 s. Although these studies did not evaluate argatroban at doses greater
than 10 Ag/kg/min, experience in healthy volunteers (Swan et al., 2000) and
HIT patients undergoing PCI indicates that doses up to 40 Ag/kg/min may be
used safely to achieve higher levels of anticoagulation, if desired.
IV. PRACTICAL ASPECTS OF ARGATROBAN DOSING

AND MONITORING
A. Duration of Therapy
Guidelines for treating patients with HIT (Hirsh et al., 2001) recommend that
anticoagulation with an alternative parenteral agent such as argatroban
should be continued at least until platelet count recovery. This approach

should help avoid thrombotic events resulting from premature discontinua-
tion of anticoagulation in a patient with profound hypercoagulability sec-
ondary to HIT. In patients with HIT in study Argatroban-915, mean platelet
counts were >100 109/L after 2 days of therapy and >150 109/L after 4
days of therapy (Lewis et al., 2003).
B. Dosing and Dosage Adjustments

For either the prophylaxis or treatment of thrombosis in HIT patients, the
recommended initial dose of argatroban is 2 Ag/kg/min (Table 3). Because
argatroban clearance is reduced in patients with hepatic impairment, a re-
duced initial dose, i.e., 0.5 Ag/kg/min, is recommended for patients with at
least moderate hepatic impairment, e.g., a Child-Pugh score >6 (Pugh et al.,
1973). Retrospective studies suggest that routine liver function tests, such as
serum bilirubin and alkaline phosphatase, may be useful for refining indi-
vidualized dosing in patients with hepatic dysfunction (Pippis et al., 2002;
Smith et al., 2002); however, this requires further investigation. No initial
dosage adjustment is required in patients with renal impairment, which is an
important advantage in managing patients with renal dysfunction.
The initial dose should be adjusted, as needed, to achieve a target aPTT
1.5–3 times the baseline value. The choice of aPTT reagent does not materially
affect assessment of argatroban therapy (Francis and Court, 2002). Approx-
imately 1 in 6 patients in study ARG-911 maintained their initial argatroban
dose for the duration of therapy, indicating that dosage adjustment is often
unnecessary (Verme-Gibboney and Hursting, 2003). When dosage adjust-
ment is necessary, the patient’s current dose, aPTT, and clinical status (e.g.,
hepatic function) should be considered. A reasonable increment for most
patients is 0.5 Ag/kg/min. Smaller increments (e.g., 0.2 Ag/kg/min) are ap-
propriate when dosing is already reduced for reasons such as hepatic im-
pairment (Verme-Gibboney and Hursting, 2003). It has been suggested that
downward dosage adjustments required in some patients may be associated
with decreased liver perfusion (Reichert et al., 2003). At substantially higher
argatroban doses, such as used during PCI, increments of 5 Ag/kg/min are
recommended (see Sec. V.A).
C. Conversion to Warfarin Anticoagulation

Warfarin is not recommended as sole anticoagulant therapy during acute
HIT. For patients with HIT requiring long-term oral anticoagulation, the
initiation of warfarin should be delayed until substantial recovery of the
platelet count has occurred (at least >100 x 109/L), the patient is adequately
Table 3 Dosing Schedules for Argatroban Treatment of Patients with HIT (Approved
Indications)
Monitoring and
Clinical use Bolusa IV infusiona adjusting therapy
Prophylaxis or — 2 Ag/kg/min (For hepatically Dose adjusted (not to

treatment of impaired patients, reduce exceed 10 Ag/kg/min)
thrombosisb,c initial dose.d Patients to achieve steady state
with renal insufficiency aPTT 1.5–3.0 times the
require no initial dosage baseline value (not to
adjustment.) exceed 100 s)e,f,g
Percutaneous 350 Ag/kg 25 Ag/kg/min Infusion dose adjusted
coronary (given over (15–40 Ag/kg/min) to
intervention 3–5 min) achieve an ACT 300–
(PCI)b,h,i 450 s; additional bolus
doses of 150 Ag/kg may
be given as neededj,k
Abbreviations: HIT, heparin-induced thrombocytopenia; IV, intravenous; aPTT, activated partial throm-
boplastin time; ACT, activated clotting time; PCI, percutaneous coronary intervention.
a
Based on patient’s body weight.
b
Includes patients with active HIT who have isolated thrombocytopenia or associated thrombosis, as well
as patients with a documented history of HIT who are no longer thrombocytopenic but require anti-
coagulation.
c
Argatroban is approved in the United States as an anticoagulant for prophylaxis or treatment of
thrombosis in patients with HIT, and in Canada as an anticoagulant in patients with HIT who, in the
opinion of their attending physician requires anticoagulation.
d
For patients with moderate hepatic impairment, an initial dose of 0.5 Ag/kg/min is recommended.
e
The aPTT should be checked at least 2 h after the initiation of argatroban or any dosage change.
f
For patients in studies ARG-911 and ARG-915, the mean F SEM dose of argatroban was 1.9 F 0.1 Ag/
kg/min.
g
For transferring a patient to warfarin anticoagulant therapy: After substantial resolution of throm-
bocytopenia, initiate warfarin therapy using the expected daily dose of warfarin (do not use a loading
dose) while maintaining argatroban infusion. At least 5 days of warfarin therapy are required to lower
functional prothrombin concentrations to a therapeutic, steady state level. For monitoring the conversion
to warfarin during coadministration of argatroban at doses up to 2 Ag/kg/min, see text and Fig. 7.
h
Argatroban is approved in the United States as an anticoagulant in patients with or at risk for HIT
undergoing PCI. Argatroban has not been evaluated in hepatically impaired patients undergoing PCI.
These recommendations do not consider the combination use of argatroban with glycoprotein IIb/IIIa
antagonists, wherein lower doses of argatroban (e.g., 250–300 Ag/kg bolus followed by infusion of 15 Ag/kg/
min) have been shown to provide effective anticoagulation with an acceptable bleeding risk (Jang et al.,
2003).
i
Includes percutaneous transluminal coronary angioplasty (balloon angioplasty), stent implantation, and
atherectomy; oral aspirin 325 mg should be given 2–24 h prior to PCI.
j
The ACT should be checked 5–10 min following the initial bolus dose and after any additional bolus dose
or change in the infusion rate. In studies ARG-216, ARG-310, and ARG-311, the majority of patients
required only one bolus dose during the interventional procedure, and the mean F SEM dose of
argatroban was 23.1 F 0.7 Ag/kg/min.
k
After the procedure, the sheaths should be removed no sooner than 2 h after discontinuing argatroban
and when the ACT is <160 s.
anticoagulated with an alternative parenteral agent such as argatroban, and

the patient is clinically improving (Hirsh et al., 2001). Because warfarin should
be avoided during acute HIT (Hirsh et al., 2001; Smythe et al., 2002; Srinvasan
et al., 2003), if warfarin has already been initiated before HIT is diagnosed, it
may be prudent in some circumstances to administer vitamin K to reverse
warfarin’s effects. During the transition period from argatroban to warfarin
anticoagulation, both agents are used concurrently, and their overlap should
be continued for at least 4–5 days. A loading dose (>5 mg) of warfarin should
not be used; rather, warfarin should be initiated using modest, anticipated
daily maintenance doses (V5 mg), with careful monitoring. Ideally, the INR
should be within the target therapeutic range for at least the last 2 days of
overlap (discussed subsequently). During the overlap period, argatroban is
monitored using the aPTT. These guidelines are important to ensure contin-
uous anticoagulation and avoid prothrombotic effects of initiating warfarin
during acute HIT, e.g., warfarin-induced venous limb gangrene or skin
necrosis syndromes (see Chap. 3).
As mentioned in Section II, because argatroban is a direct thrombin
inhibitor, concomitant use of argatroban and warfarin prolongs the PT/INR
beyond that produced by warfarin alone (Hursting et al., 1999; Sheth et al.,
2001). In the clinical studies of argatroban therapy in HIT, the majority of
patients transferred to warfarin therapy for continued anticoagulation and
there was no evidence of systematic underdosing or overdosing of warfarin
(Hiatt et al., 2001). The method of transition from argatroban to warfarin was
not specified in the protocols, and a variety of approaches were employed.
Warfarin was generally started at the expected daily maintenance dose and
continued for 3–4 days. In one method, the argatroban infusion was then
discontinued for 4 h, if deemed clinically acceptable, prior to the INR being
checked in the absence of argatroban effect. If the INR was in the therapeutic
range, argatroban was permanently discontinued, and warfarin therapy was
continued. In order to improve this procedure and avoid the possibility that
the patient would be unprotected before the desired level of warfarin anti-
coagulation was achieved, a study in healthy subjects was conducted to
develop guidelines for monitoring this transition (Sheth et al., 2001).
Results from Sheth et al. (2001) indicate that the relationship between
the INR on cotherapy and the INR on warfarin monotherapy can be used to
interpret the INR during the transition period. Specifically, INRs on cother-
apy increase linearly with INRs on warfarin monotherapy. The slope of this
relationship is sensitive to the argatroban dose and the thromboplastin
reagent used, particularly, its International Sensitivity Index (ISI). For
argatroban doses of 1–2 Ag/kg/min, prediction errors for monotherapy INRs
from cotherapy INRs are sufficiently low (F0.4 units) to allow for clinically
reliable estimations of a monotherapy INR from a cotherapy INR. However,
at argatroban doses greater than 2 Ag/kg/min, monotherapy INRs cannot be

reliably predicted from cotherapy INRs.
Figure 6 presents the relationships between INRs on cotherapy and
INRs on warfarin monotherapy, for argatroban doses of 1–2 Ag/kg/min and
thromboplastin reagents with ISIs between 0.88 and 2.13. For each of these
combinations of argatroban dose and thromboplastin, a cotherapy INR
greater than 4 is predicted to be related to a monotherapy INR between
approximately 2.0 and 3.0, i.e., in the therapeutic range for warfarin mono-
Figure 6 Relationships between the INR during warfarin/argatroban cotherapy

and the INR during warfarin monotherapy, by argatroban dose and the Interna-
tional Sensitivity Index (ISI) of the thromboplastin reagent used. Two sets of lines
are superimposed due to similarity in slope and intercept. Prediction errors are V
F0.4 INR units.
therapy. In the United States, commercially available thromboplastins with

ISI values greater than 2.2 are rare. In general, therefore, following at least 4–
5 days of coadministration of warfarin and argatroban at doses up to 2 Ag/kg/
min, argatroban can be discontinued when the cotherapy INR is greater than
4 (and ideally has been for 2 days). Upon cessation of argatroban, it would be
prudent to check the INR 4–6 hours later, when the effect of argatroban is
negligible, to ensure an actual therapeutic value reflective of warfarin mono-
therapy. For coadministration of warfarin and argatroban at doses greater
than 2 Ag/kg/min, the argatroban dose should be temporarily (4–6 h) reduced
to 2 Ag/kg/min. Then the procedure described above for predicting the
warfarin monotherapy INR from the cotherapy INR at doses up to 2 Ag/
kg/min can be followed. These guidelines are summarized in Fig. 7.
Of additional note, factor assays that are insensitive to argatroban
interference, such as the two-stage chromogenic factor X assay (Hoppen-
Figure 7 Guidelines for conversion from argatroban to oral anticoagulant therapy

with warfarin. Warfarin should be initiated only once there has been substantial
resolution of the thrombocytopenia, and beginning only with anticipated main-
tenance doses (V5/mg). Argatroban and warfarin should be overlapped for at least 5
days before discontinuing argatroban. Ideally, the INR should be within the target
therapeutic range (i.e., >4.0 during cotherapy for argatroban doses up to 2 Ag/kg/
min) for at least the last 2 days of overlap.
steadt et al. 1997; Sheth et al., 2001), may be useful for supplemental moni-
toring during argatroban and warfarin cotherapy, if desired.
D. Conversion to Phenprocoumon or Acenocoumarol

Anticoagulation
The pharmacologic interactions between argatroban and the oral anticoagu-
lants phenprocoumon or acenocoumarol are comparable to those described
for argatroban and warfarin. Guidelines for the conversion from argatroban
to phenprocoumon or acenocoumarol are similar to those for the conversion
to warfarin (Breddin et al., 2002).
V. ARGATROBAN FOR HIT PATIENTS IN SPECIAL

CLINICAL CIRCUMSTANCES
A. Percutaneous Coronary Intervention
Argatroban is the only agent approved in the United States as an anticoag-
ulant for patients with, or at risk of, HIT who undergo PCI.
Clinical Studies
Argatroban has been evaluated in three multicenter, open-label prospective
studies in patients with HIT undergoing PCI, including percutaneous trans-
luminal coronary angioplasty, stent implantation, or rotational atherectomy.
The studies (ARG-216, ARG-310, and ARG-311) were similar in design with
respect to eligibility criteria, argatroban dosing regimen, and main outcome
assessments, and their pooled analysis has been reported (Lewis et al., 2002).
Among these studies, 91 patients with HIT underwent 112 PCIs on
argatroban anticoagulation (Lewis et al., 2002). Patients received 325 mg oral
aspirin 2–24 h before PCI. In the catheterization laboratory, patients received
intravenous argatroban at 25 Ag/kg/min (initial bolus dose of 350 Ag/kg)
titrated to achieve an ACT of 300–450 s during PCI (mean infusion dose, 23
Ag/kg/min). Additional bolus doses of 150 Ag/kg to achieve or maintain the
target ACT were allowed, though usually not needed. Target ACT values
were achieved typically within 10 min of initiating argatroban and were
maintained throughout the infusion. When argatroban was discontinued
after the procedure, ACTs rapidly returned to baseline.
Primary efficacy endpoints were subjective assessment of the satisfac-
tory outcome of the procedure and adequate anticoagulation, which occurred
in 94.5% and 97.8%, respectively, of patients undergoing their initial PCI
with argatroban (n = 91) (Table 4). Death (no patients), myocardial in-
farction (4 patients), and revascularization at 24 h after PCI (4 patients)
occurred in 7 (7.7%) patients. Other efficacy endpoints were also consistent
with argatroban enabling a satisfactory outcome (Table 4). One patient (1%)
experienced major periprocedural bleeding (nonfatal retroperitoneal hemor-
rhage). No unsatisfactory outcomes occurred during repeat PCIs with arga-
troban (n = 21; mean separation of 150 days from the initial PCI). Overall,
the clinical outcomes compared favorably with those reported historically for
heparin anticoagulation during PCI.
In a separate multicenter prospective study of 101 patients (including 1
patient with HIT) undergoing PCI, reduced doses of argatroban were
evaluated in combination with the GPIIb/IIIa antagonists abciximab (n =
99) or eptifibatide (n = 2) (Jang et al., 2003). Patients received argatroban as
an initial bolus of 250 Ag/kg followed by an infusion of 15 Ag/kg/min, and
additional boluses of 150 Ag/kg were allowed, if necessary, to achieve a target
ACT of 275–325 s. This target ACT was reached in 94 patients. Death (no
patients), myocardial infarction (3 patients), and urgent revascularization
at 30 days (2 patients) occurred in 3 (3%) patients. Two additional patients
had cardiac symptoms and elevated troponin without significant creatine ki-
nase elevation. There were 2 major bleeding events (1 retroperitoneal, 1 groin
Table 4 Efficacy Assessments in HIT Patients Undergoing PCI Using

Argatroban Anticoagulation
Number of patients with

outcome/total n (%)
Outcome Initial group Repeat group
Satisfactory outcome 86/91 (94.5%) 21/21 (100%)

of procedurea
Adequate anticoagulationa 89/91 (97.8%) 21/21 (100%)
Lack of major acute 89/91 (97.8%) 21/21 (100%)
complicationsb
Angiographic successc 86/88 (97.7%) 20/20 (100%)
Clinical successd 86/88 (97.7%) 20/20 (100%)
a
Primary, subjective outcomes.
b
No death, emergent coronary artery bypass graft surgery, or Q-wave myocardial infarction
during argatroban infusion or 24 h of its cessation (or discharge, whichever came first).
c
Final stenosis of <50% in at least one lesion attempted, for patients with angiographic data
available.
d
Angiographic success plus the lack of major acute complications.
hematoma). Hence, argatroban in combination with a GPIIb/IIIa antagonist

provides adequate anticoagulation with an acceptable bleeding risk. Addi-
tional studies are underway to refine this dosing regimen (e.g., evaluating a
greater initial bolus dose of 300 Ag/kg).
Argatroban Dosing and Monitoring During PCI

Dosing recommendations for patients with, or at risk for, HIT undergoing
PCI (Table 3) are based on those used during studies ARG-216, ARG-310,
and ARG-311 (Lewis et al., 2002). Argatroban should be started at an
infusion dose of 25 Ag/kg/min and a bolus of 350 Ag/kg given over 3–5 min.
The ACT should be checked 5–10 min after the bolus dose is completed. If the
ACT is >300 s, the PCI may proceed. If the ACT is <300 s, an additional
bolus dose of 150 Ag/kg should be given and the infusion dose increased to 30
Ag/kg/min. If, however, the ACT is >450 s after the initial bolus, then the
infusion dose should be reduced to 15 Ag/kg/min. After any additional bolus
or dosage adjustment, the ACT should be checked again after 5–10 min to
confirm the patient attained a therapeutic ACT. During a prolonged proce-
dure, additional ACTs should be obtained every 20–30 min. For patients
requiring anticoagulation after the procedure, argatroban infusion may be
continued at a reduced dose such as that recommended for the prophylaxis or
treatment of thrombosis in HIT.
These dosing recommendations do not take into consideration the
possible combination use of GPIIb/IIIa antagonists. In that setting, lower
doses of argatroban are prudent, e.g., bolus dose of 250–300 Ag/kg followed
by an infusion dose of 15 Ag/kg/min (Jang et al., 2003).
High doses of argatroban should be avoided in HIT patients who
require PCI and have clinically significant hepatic disease, including labora-
tory evidence such as aspartate aminotransferase or alanine aminotransferase
at least three times the upper limit of normal. Argatroban use during PCI has
not been studied in such patients (Lewis et al., 2002; Jang et al., 2003).
A practical issue associated with monitoring heparin anticoagulation
during PCI is that ACTs from the most commonly used automated systems
(the Hemochron and HemoTec systems) are not interchangeable (Avendano
and Ferguson, 1994). In contrast, this is not an issue with argatroban anti-
coagulation. Argatroban equally prolongs the Hemochron ACT and Hemo-
Tec ACT (Iqbal et al., 2002), and investigators in the PCI trials effectively
used whichever ACT method was available at their sites to monitor anti-
coagulation.
In reported studies of argatroban anticoagulation use during PCI
(Lewis et al., 2002; Jang et al., 2003), sheaths were removed when the ACT
was less than 160 s.
B. Peripheral Intervention
Case reports describe the successful use of argatroban anticoagulation in
patients with HIT during renal stent implant (Lewis et al., 1997) and carotid
stent implant (Lewis et al., 1998). The argatroban dose and target ACT values
were the same as those recommended for PCI in the absence of GPIIb/IIIa
inhibition.
C. Argatroban in Hemodialysis
Argatroban has been used successfully for anticoagulation during hemodial-
ysis in patients with HIT (Matsuo et al., 1988, 1990; Koide et al., 1995; Reddy
et al., 2002; Mihindu et al., 2002), including a patient with comorbid hepatic
failure (Dager and White, 2003) (see also Chap. 18). In the latter case,
argatroban effectively prevented clotting in the dialyzer circuit, and consistent
with reduced argatroban clearance in patients with hepatic impairment, the
aPTT and INR were slow to recover after stopping argatroban. The effective
use of argatroban for the maintenance of catheter or graft patency between
dialysis treatments has also been reported (Mihindu et al., 2002). Although
guidelines are available for argatroban use in dialysis-dependent patients with
HIT at a major medical center (O’Shea et al., 2003), the safety and efficacy of
argatroban in patients with HIT undergoing hemodialysis have not been fully
evaluated in a clinical trial.
Argatroban administration by bolus alone, infusion alone, or bolus plus
infusion has been evaluated in a prospective cross-over study of 12 patients
with end-stage renal disease undergoing chronic hemodialysis (Murray et al,
2003). Target ACTs during dialysis were 140–180% of the baseline value. The
most satisfactory intradialysis anticoagulation was achieved using a steady-
state infusion of argatroban (2 Ag/kg/min begun f4 h before dialysis), or a
250 Ag/kg bolus dose at the start of dialysis followed by a continuous 2 Ag/kg/
min infusion. In 38 separate hemodialysis sessions, no dialysis membrane
required changing, and one (2.6%) session was shortened owing to circuit
clotting that occurred after 3 h of hemodialysis. There were no bleeding
events. Although confirmatory studies are required, it is anticipated that sim-
ilar dosing regimens may be adequate for inpatients with HIT already at
steady-state argatroban levels or outpatients with a history of HIT who re-
quire hemodialysis.
In support of the general safety of argatroban as an anticoagulant
during hemodialysis in patients with HIT, 54 argatroban-treated patients
with HIT underwent hemodialysis in studies ARG-911, ARG-915, and ARG-
915X. These protocols made no recommendations regarding argatroban
dosing during hemodialysis. For the patients who did, versus did not, undergo
hemodialysis, there were no significant between-group differences in either the

primary efficacy endpoint or major bleeding.
Argatroban is approved in Japan as an anticoagulant for hemodialysis
in patients with congenital or acquired antithrombin deficiency. The recom-
mended infusion dose for argatroban in antithrombin-deficient patients
undergoing hemodialysis (Novastan Prescribing Information, Japan, 1998)
is generally similar to that which has been used for HIT patients undergoing
hemodialysis (Matsuo et al., 1988, 1990; Koide et al., 1995).
D. Stroke
The effect of argatroban anticoagulation on stroke in HIT has been retro-
spectively evaluated using case records from studies of argatroban for HIT
(LaMonte et al., 2003). Stroke occurred in 3.0% of 1005 individuals with HIT
and was a significant predictor of death in argatroban-treated patients and
historical controls (odds ratio z3 for each group). Almost all strokes (33/35,
94%) were ischemic, consistent with the prothrombotic nature of HIT. Stroke
occurred most often in females, in those with a traditional risk factor for
stroke, in patients with more severe thrombocytopenia, and within 2 weeks of
HIT presentation. Compared with controls, argatroban therapy for HIT was
associated with reduced frequency of new stroke (1.8% vs. 4.7%, p = 0.032)
and stroke-associated mortality (1.0% vs. 3.1%, p = 0.036). These benefits
were achieved without increased intracranial or major bleeding. These data
highlight the importance of considering HIT in the differential diagnosis when
stroke occurs, particularly in hospital in-patients.
In a randomized, double-blind clinical study of patients treated with
argatroban vs. placebo within 12 h of ischemic stroke onset, there were no
significant between-group differences in intracranial hemorrhage or major
bleeding rates (LaMonte, 2003). Although not conducted in patients with
HIT, the study further supports the safety of argatroban anticoagulation in
patients with stroke.
As mentioned previously, argatroban is approved in Japan for use in
nonlacunar stroke and in Korea for use in acute cerebral thrombosis.
E. Cardiopulmonary Bypass Surgery or Extracorporeal

Membrane Oxygenation
Cardiopulmonary bypass (CPB) surgery using argatroban anticoagulation
has been performed successfully in dogs (Walenga et al., 1997) and pigs
(Ahmad et al., 2001). In addition, there have been case reports of HIT patients
in whom argatroban anticoagulation has been used successfully during off-
pump coronary artery bypass surgery (Arnoletti and Whitman, 1999; Ide et al.,
2001; Kieta et al., 2003; Ohno et al., 2003), and during (Edwards et al., 2003)
or immediately before and after CPB (Lubenow et al., 2003). Levels of
anticoagulation used during off-pump coronary bypass surgery tend to par-
allel those used during angioplasty. However, in the experience to date with
argatroban anticoagulation in HIT patients in this setting, infusion doses have
been generally similar to those recommended for the prophylaxis or treatment
of thrombosis in HIT, with target ACTs >200 s (Ide et al., 2001; Ohno et al.,
2003) or twice the baseline value (Kieta et al., 2003). A consistently safe and
effective dose to support CPB surgery in humans has not been established.
In vitro studies indicate that argatroban is at least as effective as heparin
in preventing thrombin generation in extracorporeal membrane oxygenation
(ECMO) circuits (Yonekawa et al., 2002). The successful use of argatroban
anticoagulation in ECMO has been described, including an adult patient with
HIT (Johnston et al., 2003) and two neonates, including one in whom ECMO
was continued for 78 days (Kawada et al., 2000). In each case, ACTs were
typically maintained >200 s, although further study is required to establish
dosing recommendations.
F. Other Hypercoagulability States

Antithrombin (AT) deficiency is a congenital hypercoagulability disorder
wherein an alternative to heparin may be required. Argatroban (2 Ag/kg/min)
has been used successfully in a patient with burn-related severe acquired AT
deficiency who failed heparin (Gorman et al., 2001). However, no formal
studies have been conducted. As mentioned, argatroban is approved in Japan
as anticoagulation for hemodialysis in patients with congenital or acquired
AT deficiency.
The effective use of direct thrombin inhibitors, including argatroban,
has been described in patients with disseminated intravascular coagulation
(DIC), including patients with low levels of AT or with suspected HIT
(Kumon et al., 1984; Mukundan and Zeigler, 2002). The data, albeit limited,
provide evidence that argatroban can improve DIC, and also that DIC in a
patient with HIT should not preclude use of argatroban.
G. Argatroban in Pregnant, Nursing, or Postpartum Women

Argatroban anticoagulation in pregnant or nursing women has not been
studied. Teratology studies in rats reveal no evidence of impaired fertility or
fetal harm due to argatroban (Argatroban Prescribing Information, 2002).
Because animal reproductive studies are not always predictive of human
response, it is recommended that the drug be used during pregnancy only if
clearly needed.
Argatroban is detected in rat milk (Iida et al., 1986). It is unknown

whether argatroban is excreted in human milk, although many drugs are.
Hence, it is recommended that a decision be made either to discontinue nurs-
ing or discontinue the drug.
H. Argatroban in Geriatric or Pediatric Patients

The pharmacokinetic parameters of argatroban are similar between young
adult and elderly volunteers (Swan and Hursting, 2000), and no dosage ad-
justment is required for the elderly. Further, the effectiveness of argatroban for
HIT was not influenced by patient age (Lewis et al., 2001). Across the pivotal
studies evaluating argatroban for HIT, patient age ranged from 17 to 91 years.
Data on argatroban use in pediatric patients are limited. Argatroban
anticoagulation has been used successfully in at least three pediatric patients
with HIT for the prophylaxis or treatment of thrombosis or cardiac cathe-
terization (Boshkov et al., 2002, 2003) and at least two neonates during
ECMO (Kawada et al., 2000). A study to establish the pharmacokinetics of
argatroban in children is underway in the United States.
VI. CONCLUSION
Argatroban, a synthetic direct thrombin inhibitor, is an effective anticoagu-

lant with a predictable dose-response effect. This agent offers several theo-
retical advantages as an anticoagulant for patients with HIT: it inhibits free
and bound thrombin, it does not cross-react with HIT antibodies, and its
anticoagulant effects are rapidly active and also rapidly reversible. Further,
upon prolonged or repeated administration, argatroban is well tolerated, with
no alteration in anticoagulant response and no induction of drug-specific
antibodies.
In clinical studies, argatroban therapy, compared with historical con-
trols, improves outcomes of HIT, particularly thrombosis and its sequelae,
without increasing bleeding risk. Argatroban also provides safe and effective
anticoagulation in patients with a history of HIT requiring acute antico-
agulation. No introcranial hemorrhage has occurred during argatroban
infusion in over 900 patients with HIT, including many with stroke, who
have received argatroban during clinical trials. These benefits are achieved
when argatroban is administered intravenously at 2 Ag/kg/min, titrated to
achieve an aPTT 1.5–3.0 times baseline. Although no initial dosage adjust-
ment is required for patients with renal impairment, an initial dose of 0.5 Ag/
kg/min is recommended for hepatically impaired patients. Also in clinical
studies, argatroban at higher doses (25 Ag/kg/min, titrated to achieve an ACT
of 300–450 s) provides safe and adequate anticoagulation in HIT patients

during PCI. Lower doses of argatroban in combination with a GPIIb/IIIa
antagonist also provide safe and adequate anticoagulation during PCI. The
recommended dosing schedules for the approved indications of argatroban in
the United States (i.e., as an anticoagulant for the prophylaxis or treatment of
thrombosis in patients with HIT and during PCI for patients with or at risk
for HIT) are summarized in Table 3.
Although data are limited, patients with HIT have also successfully un-
dergone hemodialysis, off-pump coronary bypass surgery, and ECMO using
argatroban anticoagulation. Furthermore, although studied in patients with-
out HIT, argatroban dosing regimens tailored to meet the needs of the inpa-
tient or outpatient safely provide intradialysis anticoagulation. Argatroban
therefore offers a versatile therapeutic option for the management of patients
with HIT in diverse clinical settings.
REFERENCES
Ahmad S, Rusterholtz L, Jain S, Iqbal O, Ahsan A, Strony J, Walenga JM, Fareed J.

Anticoagulant dosage optimization of argatroban in a porcine model of
cardiopulmonary bypass surgery [abstr]. Blood 2001; 98(part 1):47a.
Ahsan A, Ahmad S, lqbal O, Schwarz R, Knappenberger G, Joffrion J, Becker JC,
Messmore H. Comparative studies on the biochemical and pharmacological
properties of a major metabolite of argatroban (MI): potential clinical impli-
cations [abstr]. Thromb Haemost 1997; 78(suppl 2):370.
Alving BM. How I treat heparin-induced thrombocytopenia and thrombosis. Blood
2003; 101:31–37.
Antman EM, for the TIMI 9A investigators. Hirudin in acute myocardial infarction,
safety report from the Thrombolysis and Thrombin Inhibition in Myocardial
Infarction (TIMI) 9A trial. Circulation 1994; 90:1624–1630.
Argatroban Prescribing Information. GlaxoSmithKline, Research Triangle Park,
April 2002.
Arnoletti JP, Whitman GJR. Heparin-induced thrombocytopenia in coronary bypass
surgery. Ann Thorac Surg 1999; 68:576–578.
Avendano A, Ferguson JJ. Comparison of Hemochron and HemoTec activated coag-
ulation time target values during percutaneous transluminal coronary angio-
plasty. J Am Coll Cardiol 1994; 23:907–910.
Banner DW, Hadvary P. Crystallographic analysis at 3.0-Å resolution of the binding
to human thrombia of four active site-directed inhibitors. J Biol Chem 1991;
266:20085–20093.
Berry CN, Girardot C, Lecoffre C, Lunven C. Effects of the synthetic thrombin in-
hibitor argatroban on fibrin- or clot-incorporated thrombin: comparison with
heparin and recombinant hirudin. Thromb Haemost 1994; 72:381–386.
Boshkov LK, Thomas G, Kirby A, Shen I, Swanson V, Burch G, Ungerleider R.

Pharmacokinetics of argatroban infusion in a 6 month old congenital cardiac
patient with previously diagnosed heparin-induced thrombocytopenia [abstr].
Blood 2002; 100(part 1):269a.
Boshkov LK, Ibsen L, Kirby A, Ungerleider R, Shen I. Report of argatroban in-
fusions for heparin-induced thrombocytopenia diagnosed by functional assay
in 2 congenital cardiac surgery patients, a neonate and a 5 month old [abstr]. J
Thromb Haemost 2003; 1(suppl 1):P1495.
Brown PM, Hursting MJ. Lack of pharmacokinetic interactions between argatroban
and warfarin. Am J Health Syst Pharm 2002; 59:2078–2083.
Clark RJ, Mayo G, Fitzgerald GA, Fitzgerald DJ. Combined administration of aspirin
and a specific thrombin inhibitor in man. Circulation 1991; 83:1510–1518.
Cox DS, Aluri J, Minthorn E, Holdbrook F, Parchman LF, Fossler MJ. Pharma-
cokinetics and pharmacodynamics of argatroban in combination with a GP
IIb/IIIa antagonist in patients undergoing percutaneous coronary intervention
[abstr]. Clin Pharmacol Therap 2003; 73:P32.
Dager WE, White RH. Argatroban anticoagulation for heparin-induced thrombo-
cytopenia in hepato-renal failure and CVVHD. Ann Pharmacother (published
online DOI 10.1345/aph. 1D010: 17 July, 2003).
Edwards JT, Hamby JK, Worrall NK. Successful use of argatroban as a heparin
substitute during cardiopulmonary bypass: heparin-induced thrombocytope-
nia in a high-risk cardiac surgical patient. Ann Thorac Surg 2003; 75:1622–
1624.
Eichler P, Friesen HJ, Lubenow N, Jaeger B, Greinacher A. Antihirudin antibodies
in patients with heparin-induced thrombocytopenia treated with lepirudin-
incidence effects on aPTT and clinical relevance. Blood 2000; 96:2373–2378.
Eidt JF, Allison P, Noble S, Ashton J, Golino P, McNatt J, Buja LM, Willerson JT.
Thrombin is an important mediator of platelet aggregation in stenosed canine
coronary arteries with endothelial injury. J Clin Invest 1989; 84:18–27.
Francis JL, Court N. Effect of reagent choice on the aPTT response to argatroban
[abstr]. Blood 2002; 100(part 2):122b.
Gold HK, Torres FW, Garabedian HD, Werner W, Jang I-K, Khan A, Hagstrom
JN, Yasuda T, Leinbach RC, Newell JB, Bovill EG, Stump DC, Collen D.
Evidence for a rebound coagulation phenomenon after cessation of a 4-hour
infusion of a specific thrombin inhibitor in patients with unstable angina
pectoris. J Am Coll Cardiol 1993; 21:1039–1047.
Gorman R, Gordan L, Zumberg M, Kitchens C. Successful use of argatroban as an
anticoagulant in burn-related severe acquired antithrombin III deficiency after
heparin failure. Thromb Haemost 2001; 86:1596–1597.
P, Mueller-Velten HG, Pötzsch B, for the HIT investigators group. Recom-
binant hirudin (lepirudin) provides safe and effective anticoagulation in pa-
tients with heparin-induced thrombocytopenia: a prospective study. Circulation
1999a; 99:73–80.
Greinacher A, Janssens U, Berg G, Böck M, Kwasay H, Kemkes-Matthes B, Eichler P,
Völpel H, Pötzsch B, Luz M, for the Heparin-Associated Thrombocytopenia

Study (HAT) investigators. Lepirudin (recombinant hirudin) for parenteral
tion 1999b; 100:587–593.
Hantgan RR, Jerome WG, Hursting MJ. No effect of clot age or thrombolysis on
argatroban’s inhibition of thrombin. Blood 1998; 92:2064–2074.
Harder S, Hentig NV, Walenga JM, Osakabe M, Watanabe H, Breddin HK. Tran-
sition of treatment with i.v. argatroban to phenprocoumon and acenocouma-
rol [abstr]. J Thromb Haemost 2003; 1(suppl 1):1902.
Harenberg J, Huhle G, Wang LC, Hoffman U, Song XH. Re-exposure to recombinant
(r)-hirudin in antihirudin antibody–positive patients with a history of heparin-
induced thrombocytopenia. Br J Haematol 2000; 109:360–363.
Hartman CA, Baroletti SA, Churchill WW, Patel P. Visual compatibility of
argatroban with selected drugs. Am J Health Syst Pharm 2002; 59:1784–1785.
Herrman J-P, Suryapranata H, den Heijer P, Gabriël L, Kutryk MJB, Serruys PW.
Argatroban during percutaneous transluminal coronary angioplasty: results of
a dose-verification study. J Thromb Thrombolysis 1996; 3:367–375.
Hiatt BK, Macfarlane DE, Hursting MJ. Transition from argatroban to warfarin in
patients with heparin-induced thrombocytopenia [abstr]. Blood 2001; 98(part
2):91b.
Hirsh J. Heparin. N Engl J Med 1991; 324:1565–1574.
safety. Chest 2001; 119:64S–94S.
Hoppensteadt DA, Kahn S, Fareed J. Factor X values as a means to assess the extent
of oral anticoagulation in patients receiving antithrombin drugs. Clin Chem
1997; 43:1786–1788.
Hursting MJ, Zehnder JL, Joffrion JL, Becker JC, Knappenberger GD, Schwarz RP.
The International Normalized Ratio during concurrent warfarin and arga-
troban anticoagulation: differential contributions of each agent and effects of
the choice of thromboplastin used. Clin Chem 1999; 45:409–412.
Ide H, Fujiki T, Sato M, Endo H, Imamura K, Sudo K. Off-pump coronary artery
bypass for a heparin-allergic patient. Jpn J Thorac Cardiovasc Surg 2001;
49:250–254.
Iida S, Komatsu T, Sato T, Hayashi K, Inokuchi T. Pharmacokinetic studies of
argatroban. (MD-805) in rats: excretion into milk and foeto-placental transfer.
Jpn Pharmacol Ther 1986; 14:229–235.
Inglis AML, Sheth SB, Hursting MJ, Tenero DM, Graham AM, DiCicco R.
Investigation of the interaction between argatroban and acetaminophen,
lidocaine or digoxin. Am J Health Syst Pharm 2002; 59:1258–1266.
Iqbal O, Ahmad S, Lewis BE, Walenga JM, Rangel Y, Fareed J. Monitoring of
argatroban in ARG310 study: potential recommendations for its use in
interventional cardiology. Clin Appl Thromb Hemost 2002; 8:217–224.
Izawa O, Katsuki M, Komatsu T, Iida S. Pharmacokinetic studies of argatroban
(MD-805) in human: concentrations of argatroban and its metabolites in plasma

urine and feces during and after drip intravenous infusion. Jpn Pharmacol Ther
1986; 14:251–263.
Jang I-K, Brown DFM, Giugliao RP, Anderson HV, Losordo D, Nicolau JC, Dutra
OP, Bazzino O, Viamonte VM, Norbady R, Liprandi S, Massey TJ, Dinsmore
R, Schwarz RP, and the MINT investigators. A multicenter, randomized study
of argatroban versus heparin as adjunct to tissue plasminogen activator (TPA)
in acute myocardial infarction: myocardial infarction with Novastan and TPA
(MINT) trial. J Am Coll Cardiol 1999; 33:1879–1885.
Jang I-K, Lewis BE, Matthai WH, Kleiman NS. Combination of a direct thrombin
inhibitor, argatroban, and glycoprotein IIb/IIIa inhibitor is effective and safe
in patients undergoing percutaneous coronary intervention [abstr]. J Am Coll
Cardiol 2003; 91(suppl):68A.
Johnston N, Wait M, Huber L. Argatroban in adult extracorporeal membrane oxy-
genation. J Extra Corpor Technol 2002; 34:281–284.
Kawada T, Kitagawa H, Hoson M, Okada Y, Shiomura J. Clinical application of
argatroban as an alternative anticoagulant for extracorporeal circulation. He-
matol Oncol Clin North Am 2000; 14:445–457.
Kieta DR, McCammon AT, Holman WL, Nielson VG. Hemostatic analysis of a
patient undergoing off-pump coronary artery bypass surgery with argatroban
anticoagulation. Anesth Analg 2003; 96:956–958.
Kikumoto R, Tamao Y, Tezeka T, Tonomura S, Hara H, Ninomiya K, Hijikata A,
Okamoto S. Selective inhibition of thrombin by (2R, 4R)-4-methyl-1-[N2-[(3-
methyl-1,2,3,4-tetrahydro-8-quinolinyl)sulfonyl]-L-arginyl)]-2-piperidinecar-
boxylic acid. Biochemistry 1984; 23:85–90.
Kobayashi W, Tazaki Y. Effect of the thrombin inhibitor argatroban in acute cere-
bral thrombosis. Semin Thromb Hemost 1997; 23:531–534.
Koide M, Yamamoto S, Matsuo M, Suzuki S, Arima N, Matsuo T. Anticoagulation
for heparin-induced thrombocytopenia with spontaneous platelet aggregation
in a patient requiring haemodialysis. Nephrol Dial Transplant 1995; 10:2137–
2140.
Kumon K, Tanaka K, Nakajima N, Naito Y, Fujita T. Anticoagulation with a syn-
thetic thrombin inhibitor after cardiovascular surgery and for treatment of dis-
seminated intravascular coagulation. Crit Care Med 1984; 12:1039–1043.
LaMonte MP. Results of ARGIS-1: argatroban injection for acute ischemic stroke
[abstr]. Stroke 2003; 34:246.
LaMonte MP, Brown PM, Berens KL. Reduction in stroke-associated mortality in
patients with heparin-induced thrombocytopenia after treatment with the
direct thrombin inhibitor argatroban [abstr]. Stroke 2002; 33:382.
Lewis BE, Grassman ED, Wrona L, Rangel Y. Novastan anticoagulation during
renal stent implant in a patient with heparin-induced thrombocytopenia.
Lewis BE, Rangel Y, Fareed J. The first report of successful carotid stent implant
using argatroban anticoagulation in a patient with heparin-induced thrombo-
cytopenia and thrombosis syndrome. Angiology 1998; 49:61–67.
Lewis BE, Wallis DE, Zehnder JL, Barton JC, for the ARG-911/915/915X inves-
tigators. Argatroban reexposure in patients with heparin-induced thrombocy-
topenia [abstr]. Blood 2000; 96(Part 1):52a.
Lewis BE, Wallis DE, Berkowitz SD, Matthai WH, Fareed J, Walenga JM, Bar-
tholomew J, Sham R, Lerner RG, Zeigler ZR, Rustagi PK, Jang I-K, Rifkin SD,
Moran J, Hursting MJ, Kelton JG, for the ARG-911 Study Investigators.
Argatroban anticoagulant therapy in patients with heparin-induced thrombo-
cytopenia. Circulation 2001; 103:1838–1843.
Lewis B, Matthai WH, Cohen M, Moses JW, Hursting MJ, Leya F, for the ARG-
216/310/311 investigators. Argatroban anticoagulation during percutaneous
coronary intervention in patients with heparin-induced thrombocytopenia.
Catheter Cardiovasc Interv 2002; 57:177–184.
Lewis BE, Wallis DE, Leya F, Hursting MJ, Kelton JG, for the ARG-915 inves-
tigators. Argatroban anticoagulation in patients with heparin-induced throm-
bocytopenia. Arch Intern Med 2003; 163:1849–1856.
Lubenow N, Selleng S, Wollert HG, Eichler P, Mullejans B, Greinacher A. Heparin-
induced thrombocytopenia and cardiopulmonary bypass: perioperative arga-
troban use. Ann Thorac Surg 2003; 75:577–579.
Matsuo T, Chikahira Y, Yamada T, Nakao K, Ueshima S, Matsuo O. Effect of
synthetic thrombin inhibitor (MD805) as an alernative drug on heparin induced
thrombocytopenia during hemodialysis. Thromb Res 1988; 52:165–171.
Matsuo T, Yamada T, Yamanshi T, Ryo R. Anticoagulant therapy with MD805 of a
hemodialysis patient with heparin-induced thrombocytopenia. Thromb Res
1990; 58:663–666.
Matsuo T, Kario K, Matsuda S, Yamaguchi N, Kakishita E. Effect of thrombin
inhibition on patients with peripheral arterial obstructive disease: a multi-
center clinical trial of argatroban. J Thromb Thrombolysis 1995; 2:131–136.
Matthai WH, Hursting MJ, Lewis BE. Argatroban use in patients with a history of
heparin-induced thrombocytopenia who require acute anticoagulation [abstr].
Blood 2001; 98(part 2):45a.
Matthai WH, Hursting MJ, Lewis BE, Kelton JG. Argatroban anticoagulation in
patients with a history of heparin-induced thrombocytopenia. Submitted for
publication, 2004.
Mihindu JCL, Weinhold JR, Huntington N. Use of argatroban anticoagulation in
extracorporeal circulation and central venous catheter during hemodialysis in
patients with suspected heparin induced thrombocytopenia [abstr]. J Am Soc
Nephrol 2002; 13:189.
Mukundan S, Zeigler ZR. Direct antithrombin agents ameliorate disseminated intra-
vascular coagulation in suspected heparin-induced thrombocytopenia throm-
bosis syndrome. Clin Appl Thromb Hemost 2002; 8:287–289.
Murray P, Reddy B, Grossman E, Hammes M, Trevino S, Ferrell J, Tang I, Khorana
A, Fosbinder T, Swan SK. Safety and tolerability of argatroban anticoagulation
in patients with end-stage renal disease undergoing hemodialysis: a prospective
study [abstr]. J Thromb Haemost 2003; 1(suppl 1):P1911.
Nagasawa H, Fukutake K, Hada M, Takahashi E, Natsubara Y, Samori T, Ike-
matsu S, Kitahara T, Ukita M, Fujimaki M, Fukutake K. Phase one study of

synthetic antithrombin agent (MD-805) single and multiple administration
studies. Jpn J Clin Pharmacol Ther 1981; 12:359–375.
NovastanR Prescribing Information. Mitsubishi Chemical Corporation. Tokyo, Ja-
pan. March, 1998.
Ohno H, Higashidate M, Yokosuka T. Argatroban as an alternative anticoagulant
for patients with heparin allergy during coronary bypass surgery. Heart Ves-
sels 2003; 18:40–42.
Okamoto S, Hijikata A. Potent inhibition of thrombin by the newly synthesized
arginine derivative no. 805. The importance of stereostructure of its hydropho-
bic carboxamide portion. Biochem Biophys Res Commun 1981; 101:440–446.
Okamoto S, Okunomiya-Hijikata A. Synthetic selective inhibitors of thrombin.
Meth Enzymol 1993; 222:328–340.
O’Shea SI, Ortel TL, Kovalik EC. Alternative methods of anticoagulation for
dialysis-dependent patients with heparin-induced thrombocytopenia. Semin
Dial 2003; 16:61–67.
Patel KH. Stability and compatibility of argatroban and glycoprotein IIb/IIIa
receptor antagonists during simulated Y-site administration [abstr]. Int Pharm
Abs, Am Soc Health Sys Pharm Midyear Clin Meeting 2002; 37:P-47E.
Pippis EJ, Holdbrook FK, Hardwick JP, Hursting MJ. Serum total bilirubin with
alkaline phosphatase as means for refining argatroban dosing in patients with
hepatic impairment [abstr]. Int Pharm Abs, Am Soc Health Sys Pharm Mid-
year Clin Meeting 2002; 37:P-101E.
Pugh RNH, Murray-Lyon IM, Dawson JL, Pietroni MC, Williams R. Transection
of the oesophagus for bleeding oesophageal varices. Br J Surg 1973; 60:646–
649.
Rawson TE, VanGorp KA, Yang J, Kogan TP. Separation of 21-(R)- and 21-(S)-
argatroban: solubility and activity of the individual diastereoisomers. J Pharm
Sci 1993; 82:672–673.
Reddy BV, Grossman EJ, Nahlik L, Trevino S, Murray PT. Argatroban antico-
agulation during renal replacement therapy in patients with heparin-induced
thrombocytopenia [abstr]. J Am Soc Nephrol 2002; 13:604a.
Refludank Prescribing Information. Hoechst Marion Roussel, Kansas City, MO.
March, 1998.
Reichert MG, MacGregor DA, Kincaid EH, Dolinski SY. Excessive argatroban
anticoagulation for heparin-induced thrombocytopenia. Ann Pharmacother
2003; 37:652–654.
Sheth SB, DiCicco RA, Hursting MJ, Montague T, Jorkasky DK. Interpreting the
International Normalized Ratio (INR) in individuals receiving argatroban and
warfarin. Thromb Haemost 2001; 85:435–440.
Smith AL, Bair AK, Stalder BA. Can the optimal dosing of argatroban be
determined based upon degree of hepatic insufficiency? [abstr]. Int Pharm Abs,
Am Soc Health Sys Pharm Midyear Clin Meeting 2002; 37:P-323R.
Smythe MA, Warkentin TE, Stephens JL, Zakalik D, Mattson JC. Venous limb gan-
grene during overlapping therapy with warfarin and a direct thrombin inhibi-
tor for immune heparin-induced thrombocytopenia. Am J Hematol 2002; 71:

50–52.
Song X, Huhle G, Wang L, Hoffman U, Harenberg J. Generation of anti-hirudin an-
tibodies in heparin-induced thrombocytopenic patients treated with r-hirudin.
Circulation 1999; 100:1528–1532.
Srinvasan AF, Rice L, Bartholomew JR, Rangaswamy C, La Perna L, Thompson
JE, Murphy S, Baker KR. Warfarin-induced skin necrosis and venous limb
gangrene in the setting of heparin-induced thrombocytopenia. Arch Intern
Med 2000. In press.
Swan SK, Hursting MJ. The pharmacokinetics and pharmacodynamics of argatroban:
effects of age, gender, and hepatic or renal dysfunction. Pharmacotherapy 2000;
20:318–329.
Swan SK, St. Peter JV, Lambrecht LJ, Hursting MJ. Comparison of anticoagulant
effects and safety of argatroban and heparin in healthy subjects. Pharmaco-
therapy 2000; 20:756–770.
Tatsuno J, Komatsu T, Iida S. Pharmacokinetic studies of argatroban (MD-805):
protein binding and blood cell binding. Jpn Pharmacol Ther 1986; 14(suppl
5):243–249.
Théroux P. The Argatroban Myocardial Infarction (MINT) study. Scientific Session
News of the American College of Cardiology 1997; 15:6.
Théroux P, Waters D, Lam J, Juneau M, McCans J. Reactivation of unstable angina
after the discontinuation of heparin. N Engl J Med 1992; 327:141–145.
Tran JQ, DiCicco RA, Sheth SB, Tucci M, Pend L, Jorkasky DK, Hursting MJ,
Benincosa LJ. Assessment of the potential pharmacokinetic and pharmacody-
namic interactions between erythromycin and argatroban. J Clin Pharmacol
1999; 39:513–519.
Vanholder R, Camez A, Veys N, Van Loo A, Dhondt AM, Ringoir S.
Pharmacokinetics of recombinant hirudin in hemodialyzed end-stage renal
failure patients. Thromb Haemost 1997; 77:650–655.
Verme-Gibboney CN, Hursting MJ. Argatroban dosing in patients with heparin-
Vermeer F, Vahanian A, Fels PW, Besse P, Muller E, Van de Werf F, Fitzgerald D,
Darius H, Puel J, Garrigou D, Simoons ML for the ARGAMI Study Group.
Argatroban and alteplase in patients with acute myocardial infarction: the
ARGAMI study. J Thromb Thrombolysis 2000; 10:233–240.
Walenga JM, Koza MJ, Lewis BE, Pifarré R. Relative heparin-induced thrombo-
cytopenic potential of low molecular weight heparins and new antithrombotic
agents. Clin Appl Thrombosis/Hemostasis 1996; 2(suppl 1):S21–S27.
Walenga JM, Koza MJ, Terrell MR, Lonchyna V, Arcidi J, Pifarré R. Experimental
evaluation of argatroban for cardiopulmonary bypass. In: Pifarré R, ed. New
Anticoagulants for the Cardiovascular Patient. Philadelphia: Hanley & Belfus,
1997:251–263.
Walenga JM, Fasanella AR, Iqbal O, Hoppensteadt DA, Ahman S, Wallis DE,
Bakhos M. Coagulation laboratory testing patients treated with argatroban.
Semin Thromb Hemost 1999; 25(suppl 1):61–66.
Walenga JM, Ahmad S, Hoppensteadt DA, Iqbal O, Hursting MJ, Lewis BE.
Argatroban therapy does not generate antibodies that alter its anticoagulant
activity in patients with heparin-induced thrombocytopenia. Thromb Res
2002; 105:401–405.
Wallis DE, Workman DL, Lewis BE, Steen L, Pifarré R, Moran JF. Failure of early
Med 1999; 106:629–635.
Br J Haematol 2003; 121:535–555.
Am J Med 1996; 101:502–507.
Willerson JT, Casscells W. Thrombin inhibitors in unstable angina: rebound or
continuation of angina after argatroban withdrawal? J Am Coll Cardiol 1993;
21:1048–1051.
Yonekawa KE, Nagakawa PA, Nugent DJ, Young G. Argatroban is as effective as
heparin at preventing thrombin generation in extracorporeal membrane oxy-
genation circuits [abstr]. Blood 2002; 11(part 2):127b.
17
Bivalirudin for the Treatment of
John R. Bartholomew
Cleveland Clinic Foundation, Cleveland, Ohio
I. INTRODUCTION
The treatment for heparin-induced thrombocytopenia (HIT) has undergone

major changes over the past decade. Until recently, clinicians had few alter-
natives for treating this potentially devastating syndrome. Fortunately, with
the development of several new anticoagulants, physicians have a number of
novel treatment options. These include the heparinoid danaparoid (see Chap.
14) and the direct thrombin inhibitors (DTIs), including recombinant hirudin
(r-hirudin) (e.g., lepirudin) (see Chap. 15) and the small molecule DTI ar-
gatroban (see Chap. 16) (Alving, 2003; Chong, 2003; Warkentin, 2003). More
recently, another DTI, bivalirudin (Angiomax), has been approved for use in
percutaneous coronary angioplasty (PCI). This anticoagulant also has been
used ‘‘off-label’’ for treatment of HIT.
Bivalirudin is a hirulog, i.e., one of a group of drugs designed from the
structure of hirudin (analogue of hirudin). It was developed in the early 1990s
by the Biogen Corporation (Cambridge, MA), and was originally known as
BG8967 or Hirulog. The U.S. Food and Drug Administration (FDA) man-
dated a name change to avoid confusion with Humalog (recombinant human
insulin) when The Medicines Company (Parsippany, NJ) acquired licensure
for bivalirudin in 1997. The name was then changed to Angiomax.
Bivalirudin has also been used with favorable results in several ‘‘on-
pump’’ and ‘‘off-pump’’ cardiac surgery cases in patients with HIT. A
recently completed clinical trial in New Zealand compared bivalirudin with
heparin (with protamine reversal) in non-HIT patients requiring off-pump
475
476 Bartholomew
coronary artery bypass (OPCAB) surgery (Merry et al., 2004). It is currently

under investigation (phase II and III multicenter trials) as an alternative
anticoagulant for both on-pump and off-pump cardiac surgery. EVOLU-
TION, a randomized trial, will compare bivalirudin with heparin (and
protamine reversal) in non-HIT patients, and CHOOSE (not randomized)
will use bivalirudin as an alternative anticoagulant for patients with HIT.
II. BIVALIRUDIN
A. Chemistry
Bivalirudin is a small synthetic 20-amino-acid peptide that is a specific and
reversible inhibitor of thrombin (Parry et al., 1994) (Fig. 1). Although it is an
analogue of hirudin, its amino acid sequence is considerably shorter. Bivali-
rudin unites a carboxy-terminal segment of 12 amino acids (dodecapeptide)
derived from native hirudin (residues 53–64) to an active site-binding tetra-
peptide sequence (D-Phe-Pro-Arg-Pro) at its amino terminal (Maraganore et
al., 1990; Nawarskas and Anderson, 2001; White and Chew, 2002). Four gly-
cine residues bridge these two segments together. The amino-terminal segment
has a high affinity and specificity for binding to the active site of thrombin
(Fareed et al., 1999; Sciulli and Mauro, 2002), while the carboxy terminal
Figure 1 The structure of bivalirudin. (Top) Bivalirudin is comprised of 20 amino

acids, with an N-(amino-)terminal D-Phe-Pro-Arg-Pro (F-P-R-P) region that binds
with high affinity to the active site region of thrombin; a (gly)4 (G4) ‘‘spacer’’ region;
and a C-(carboxy-)terminal Asn-Gly-Asp-Phe-Glu-Glu-Ile-Pro-Glu-Glu-Tyr-Leu
dodecapeptide (N-G-D-F-E-E-I-P-E-E-Y-L) that binds to the fibrinogen-binding
region (exosite 1) of thrombin. The 11 C-terminal amino acids (shaded circles)
correspond exactly to the 53- to 64-amino-acid sequence of lepirudin. Highly-specific,
non-competitive binding between bivalirudin and thrombin results. (Not shown is the
heparin-binding region [exosite 2] of thrombin.) However, proteases (including other
thrombin molecules [not shown]) can cleave the Arg3-Pro4 of bivalirudin, leading to
loss of antithrombin activity. (Bottom) Initially, there is bivalent binding of
bivalirudin to thrombin, as shown. Following cleavage at Arg3-Pro4, the N-terminal
sequence of bivalirudin no longer binds to thrombin, leaving the residual C-terminal
dodecapeptide with greatly reduced binding affinity for exosite 1 of thrombin. Thus,
the bivalirudin remnant transforms to a competitive inhibitor of thrombin. Other
substrates, e.g., fibrinogen, can compete with, and displace, bivalirudin, thus allowing
thrombin to resume its prohemostatic functions. Abbreviations: Arg (R), arginine;
Asn (N), asparagine; Asp (D), aspartic acid; Glu (E), glutamic acid; Gly (G), glycine;
Ile (I), isoleucine; Leu (L), leucine; Phe (F), phenylalanine; Pro (P), proline; Tyr (Y),
tyrosine.
Bivalirudin in HIT Treatment 477
478 Bartholomew
binds to the fibrinogen recognition site of thrombin at exosite 1 (Thiagarajan

and Wu, 1999; Reed and Bell, 2002). One difference between bivalirudin and
hirudin is that the binding of bivalirudin to the active site of thrombin is
transient, whereas with lepirudin, irreversible thrombin-hirudin complexes
are formed (Weitz and Hirsh, 1998; Nawarskas and Anderson, 2001).
Bivalirudin is produced by solid-phase peptide synthesis (Maraganore
et al., 1990). Its molecular mass is 2180 Da. Bivalirudin has no structural
similarity to heparin.
B. Pharmacology
Bivalirudin is a bivalent DTI, i.e., it binds two distinct regions of thrombin:
the active (catalytic) site and the fibrinogen-binding site. Moreover, like
lepirudin and argatroban, bivalirudin binds to both free (soluble) and clot-
bound (fibrin-bound) thrombin. It forms a 1:1 stoichiometric complex that
neutralizes thrombin during coagulation and thrombus formation (Maraga-
nore and Adelman, 1996). Thus, bivalirudin inhibits proteolytic cleavage of
fibrinogen, thrombin-mediated activation of factors V and VIII, and throm-
bin-induced platelet activation.
Bivalirudin (unlike lepirudin) is a reversible inhibitor of thrombin (Fig.
1). It acts initially as a noncompetitive inhibitor, rendering thrombin inactive.
Circulating proteases (including other thrombin molecules) slowly cleave
bivalirudin near the amino-terminal end (between arg3-pro4), thus eventually
releasing the amino-terminal segment from the active site region of throm-
bin (Bates and Weitz, 1998; Carswell and Plosker, 2002; Reed and Bell, 2002;
Sciulli and Mauro, 2002). This allows thrombin to resume catalytic function.
As mentioned, bivalirudin also inhibits thrombin by the binding of its
carboxy-terminal segment to the fibrinogen-binding site on thrombin. This
occurs at the same time that the amino-terminal segment attaches to the active
site, thus resulting in dual blockage with complete inhibition of thrombin’s
multiple activities (Sciulli and Mauro, 2002). Once the amino-terminal moiety
of bivalirudin is cleaved, however, the carboxy-terminal region acquires low-
affinity, weakly competitive binding properties. Fibrinogen can now displace
the bivalirudin remnant from thrombin and align itself over the active site to
be converted to fibrin (Parry et al., 1994).
Bivalirudin is not inactivated by platelet factor 4, nor does it require any
cofactor for its activity. It does not bind to red blood cells or proteins other
than thrombin.
C. Pharmacokinetics
Bivalirudin has predictable pharmacokinetics, and exhibits a linear dose-
response relationship when given by the intravenous (iv) route to healthy
volunteers with normal renal function. Its half-life is approximately 25–36

min (Fox et al., 1993; Robson, 2000; Robson et al., 2002). Peak bivalirudin
plasma concentrations after a 15-min iv infusion are related to dose and occur
within 5 min of completing the infusion (Fig. 2).
Bivalirudin has a volume of distribution of 0.24 L/kg and a clearance
rate of approximately 3.4 mL/min/kg (Fox et al., 1993). It is cleared from
plasma by both renal mechanisms and cleavage by plasma proteases. Bivali-
rudin undergoes glomerular filtration, secretion in the proximal convoluted
tubule, and reabsorption in the distal convoluted tubule. The peptides are
then further degraded within the intracellular lysosomes (Robson, 2000;
Robson et al., 2002). In a study by Fox and colleagues (1993), only 20% of
bivalirudin was recovered in the urine.
Clearance of bivalirudin is accomplished predominately by proteolytic
cleavage within plasma and elsewhere and accounts for approximately 80%
of the drug’s metabolism (Fox et al., 1993; Scatena, 2000; Robson et al., 2002;
Warkentin and Greinacher, 2003). Indeed, proteolysis of bivalirudin appears
to result mainly from (nonbivalirudin-inhibited) thrombin, thus providing a
mechanism of degradation that is independent of specific organ function
(Bates and Weitz, 2000; Koster et al., 2002b). This results in degradation to
individual amino acids and small, inactive peptide fragments (Carswell and
Plosker, 2002).
Patients with renal insufficiency need dose adjustments for bivalirudin,
according to their degree of impairment (Table 1). In a study of 45 patients
with normal to severe renal disease, Robson (2000) found that patients with
normal kidneys (glomerular filtration rate [GFR] >90 mL/min) and mildly
impaired renal disease (GFR = 60–89 mL/min) had similar renal clearance
levels and required no dose adjustments. The clearance rate was reduced by
45% in individuals with moderate renal impairment (GFR = 30–59 mL/min)
and by 68% in persons with severe renal impairment (GFR <30 mL/min). In
dialysis-dependent patients, the clearance rate was reduced by 77% (Robson,
2000; Robson et al., 2002). Dose adjustments are thus recommended for
patients with moderate or severe kidney dysfunction and for individuals on
dialysis (Irvin et al., 1999; Robson, 2000; Robson et al., 2002). The half-life of
bivalirudin in patients with severe renal impairment is prolonged (about 1 h).
In dialysis patients, the half-life is approximately 3.5 h (Nawarskas and
Anderson, 2001).
D. Pharmacodynamics
Bivalirudin produces an immediate effect after iv administration. It causes
prolongation of the prothrombin time (PT)/international normalized ratio
(INR), activated clotting time (ACT), the activated partial thromboplastin
time (aPTT), and the thrombin time (TT) (Fox et al., 1993; Lidon et al., 1993;
480 Bartholomew
Sharma et al., 1993; Topol et al., 1993). Although there is some interindivid-
ual variability, a dose of bivalirudin given as an infusion of 0.2 mg/kg/h in-
creased the aPTT from 27 to 62 s in one study, while an infusion rate of 1.0
mg/kg/h resulted in an average aPTT of 98 s in another group of patients
(Lidon et al., 1993).
The INR is also prolonged somewhat during bivalirudin infusion. In 54
healthy volunteers, a dose of 0.05–0.6 mg/kg of bivalirudin given over 15 min
iv increased the INR to between 1.25 and 2.43 (Fox et al., 1993). In a study by
Lidon and coworkers (1993), the PT was prolonged to between 12 and 16 s
with a dose of 0.2 mg/kg/h, while Francis and colleagues (2003; and
unpublished data) recently reported the mean INR to be 1.51 (range 1.26–
2.08) in 40 patients with suspected HIT treated with bivalirudin. Two recent
abstracts also mention a slight prolongation in the INR (Bufton et al.,
2002a,b). Although the increase in the INR seems not to be as great as with
the DTI argatroban, physicians should be aware of DTI-coumarin interac-
tions during overlapping therapy (see Chap. 13).
Bivalirudin decreases fibrinopeptide A levels (a marker of fibrinogen
cleavage) in patients with coronary artery disease (Cannon et al., 1993; Ren et
al., 1997). It may also increase the bleeding time in some patients (Topol et al.,
1993).
Bivalirudin does not inhibit platelet activation or aggregation directly,
but it has been shown to inhibit thrombin-mediated platelet aggregation
without affecting adenosine 5V-diphosphate (ADP) or collagen-mediated
platelet activation (Weitz and Maraganore, 2001; Wiggins et al., 2002; Witt-
kowsky, 2002) (Fig. 3). This antiplatelet effect may make bivalirudin more
useful than heparin in the platelet-rich environment of acute coronary syn-
dromes, where both platelet activation and thrombin formation play signifi-
cant roles.
The anticoagulant effects of bivalirudin reverse rapidly, with coagula-
tion times returning to baseline within 1–2 h after stopping the infusion (Fox
et al., 1993).
E. Dosage
Bivalirudin is approved for iv administration only. It has not yet been ap-
proved for use in HIT, and therefore no dosing guidelines for this indication
Figure 2 Prolongation of the activated partial thromboplastin time (aPTT) by

increasing doses of bivalirudin. Bivalirudin was given by iv infusion over 15 min to five
groups (4 subjects each) in doses ranging from 0.05 mg/kg per 15 min up to 0.6 mg/kg
per 15 min. Each series of data points represents the mean of four study subjects.
(From Fox et al., 1993.)
482 Bartholomew
Table 1 Bivalirudin Dosage Adjustments for Percutaneous Coronary

Intervention Based on Renal Function
Infusion rate for

Renal function Bivalirudin Initial 4 h (mg/kg/h)
(glomerular filtration clearance Half-life bolus (% reduction in
rate, mL/min) (mL/min/kg) (min) (mg/kg) infusion dose)
Normal renal function 3.4–4.6 25 1 2.5

(z90 mL/min)
Mild renal impairment 3.4–4.9 22 1 2.5 (no
(60–89 mL/min) adjustment)
Moderate renal 2.5–2.7 34 1 2.0 (20%
impairment (30–59 reduction)
mL/min)
Severe renal impairment 1.5–2.8 57 1 1.0 (60%
(10–29 mL/min) reduction)
Dialysis-dependent 1.0 210 1 0.25 (90%
patients reduction)
Source: Data shown upon which bivalirudin dosing recommendations (for PCI) are based are
available in Robson (2000) and Robson et al. (2002).
are established (Dager and White, 2002). In trials to date, however, several
dosing regimens have been reported. In three patients with venous and arterial
thrombosis treated with bivalirudin for HIT, Chamberlin and associates
(1994) used doses ranging from 0.05 to 0.2 mg/kg/h. Their goal was to
maintain a therapeutic aPTT greater than 50 s. Bufton and coworkers
(2002a) used an average dose of 0.27 mg/kg/h in one patient who received
bivalirudin for over 2 months.
Francis and colleagues (2003; and unpublished data) have used bivali-
rudin in 40 patients with clinically suspected HIT. Only two patients
were given iv boluses. Initial infusion rates usually ranged from 0.15 to
0.20 mg/kg/h; the overall mean infusion rate was 0.165 mg/kg/h (maximum
0.38 mg/kg/h). The target aPTT was a 1.5- to 2.5-fold prolongation of the
Figure 3 The effect of bivalirudin on thrombin-induced platelet aggregation. Bivali-

rudin completely inhibits thrombin-induced platelet aggregation at concentrations
about 1/500 that of therapeutic doses achieved during PCI, without significant effect
on platelet aggregation by collagen or adenosine diphosphate (ADP). (From Witt-
kowsky, 2002.)
484 Bartholomew
baseline aPTT value. Thus, a reasonable regimen might be to start at 0.15

mg/kg/h (no initial bolus), with subsequent adjustments according to aPTT.
The dose recommended in the ‘‘Anticoagulant Therapy with Bivali-
rudin to Assist in the Performance of Percutaneous Coronary Intervention in
Patients with Heparin-induced Thrombocytopenia’’ (ATBAT) trial was a
bolus of 1.0 mg/kg followed by an infusion of 2.5 mg/kg/h for 4 h. This dose
was later changed to a bolus of 0.75 mg/kg followed by a 1.75 mg/kg/h in-
fusion over 4 h, based on data from the CACHET and REPLACE-1 trials
(Mahaffey, 2001; Lincoff et al., 2002a).
Bivalirudin is approved for PCI in patients with unstable angina.
Currently, the recommended dose for patients with (near) normal renal func-
tion is a bolus of 1.0 mg/kg followed immediately by a continuous 4-h infusion
at 2.5 mg/kg/h (Sciulli and Mauro, 2002). The bolus is given just prior to
angioplasty. After completing the 4-h infusion, additional bivalirudin may be
given at a rate of 0.2 mg/kg/h for up to 20 h.
Bivalirudin infusion should be reduced 20% in patients with moderate
renal impairment and by 60% in patients with severe renal impairment (GFR
<30 mL/min) (Robson et al., 2002). In dialysis-dependent patients the dose is
reduced by 90% and ideally should be given when the patient is off dialysis
(Sciulli and Mauro, 2002) (Table 1).
Allie et al. (2003) used doses similar to the modified ATBAT trial dose
for percutaneous transluminal angioplasty (PTA) of the renal and iliac
arteries. Following a 0.75 mg/kg iv bolus, bivalirudin was subsequently given
by infusion (1.75 mg/kg/h infusion) until completion of the procedure.
Bivalirudin is not FDA approved for this indication.
For patients given bivalirudin for OPCAB, Merry and colleagues (2004)
selected a 0.75 mg/kg bolus followed by a 1.75 mg/kg/h infusion. In the
multicenter CHOOSE and EVOLUTION trials for cardiac surgery (see Sec.
IV.C), for those patients undergoing OPCAB surgery the same dosing will be
given, with the option to increase or decrease the infusion in 0.25 mg/kg/h
increments (or to administer additional 0.1–0.5 mg/kg boluses) to maintain
the ACT over 300 s.
For cardiac surgery using cardiopulmonary bypass (CPB), i.e., on-
pump surgery, about two- to threefold greater levels of anticoagulation are
required, compared with OPCAB. In the multicenter CHOOSE (HIT
patients) and EVOLUTION (non-HIT patients) trials, for patients undergo-
ing CPB, an iv bolus dose of 1.0 mg/kg will be given, followed by a 2.5 mg/
kg/h iv infusion, and 50 mg bivalirudin will be added pre-CPB to the pump
circuit volume. (For details regarding this protocol, including important tech-
nical considerations for the cardiac surgeon and cardiac anesthesiologist, see
Chap. 19; see also Warkentin and Greinacher, 2003.)
F. Administration
Bivalirudin is administered iv and produces a rapid anticoagulant effect (Fig.
2). In several small trials, however, it has also been given by subcutaneous (sc)
injection. In contrast to its rapid clearance following iv injection, its antico-
agulant effects are sustained for several hours following sc administration
(Fox et al., 1993) (Fig. 4). The peak anticoagulant effect occurred between 1
and 2 h after sc administration in a study of human volunteers, with detectable
plasma levels measured up to 6 h postinjection. The aPTT was prolonged
from 150 F 19.4% to 176 F 19.4% of the baseline value, and the INR
increased from 1.18 F 0.05 to 1.48 F 0.17 (Fox et al., 1993). Urinary excretion
of the drug was complete by 8–12 h.
A number of drugs commonly used in patients undergoing PCI have
been tested for Y-site compatibility with bivalirudin. Testing was for short-
term mixing, rather than longer-term interactions (4 h). Drugs found to be
compatible with bivalirudin included abciximab, dexamethasone, digoxin,
diphenhydramine, dobutamine, dopamine, epinephrine, eptifibatide, esmo-
lol, furosemide, heparin, lidocaine, morphine, nitroglycerin, potassium chlo-
ride, sodium bicarbonate, tirofiban, and verapamil (The Medicines
Company, 2001; Reed and Bell, 2002). Table 2 lists nine drugs found to cause
haze formation or gross precipitation, which thus should not be administered
in the same line as bivalirudin.
Drug-drug interaction studies have been performed with the thienopy-
ridine derivative ticlopidine, the glycoprotein (GP) IIb/IIIa inhibitors abcix-
imab, eptifibatide and tirofiban, and low molecular weight heparin and
unfractionated heparin (Reed and Bell, 2002). No pharmacodynamic inter-
actions occurred between bivalirudin and these agents. In patients undergoing
PCI, use of bivalirudin in conjunction with heparin, warfarin, or thrombolytic
therapy has been associated with increased risk of bleeding (The Medicines
Company, 2001). Aspirin was associated with a mild increase in bleeding
times in patients receiving bivalirudin infusions when compared to placebo.
These changes were not felt to be clinically significant (Fox et al., 1993).
G. Monitoring
The PT/INR, ACT, aPTT, and TT all rise linearly with increases in the dose of
bivalirudin. The ACT is generally used to monitor bivalirudin in patients
undergoing PCI, while the aPTT has been used in patients treated for HIT and
other non-PCI indications. Currently, the ECT (see Chap. 19) is recom-
mended for monitoring during on-pump cardiac surgery (CPB). Dosing in
PCI generally aims to maintain the ACT above 300 or 350, while in patients
486 Bartholomew
undergoing aPTT monitoring, the target range generally is a 1.5- to 2.5-fold

increase in the baseline aPTT (Chew et al., 2001).
The ACT and aPTT have limitations, and questions regarding their
adequacy for monitoring DTI therapy remain (Pötzsch et al., 1997; Koster et
al., 2000, 2003a; Despotis et al., 2001; de Denus and Spinler, 2002). As a result,
several new tests have been developed for monitoring these anticoagulants.
Cho and colleagues (2003a) have used a thrombin inhibitor manage-
ment (TIM) point-of-care test based upon the ECT for monitoring the
anticoagulant effects of bivalirudin. This test (TIM-ECT) was developed by
PharmaNetics, Inc. (Morrisville, NC) and evaluated in 64 consecutive pa-
tients who underwent nonemergent PCI. The study compared the TIM-ECT
to two available ACT methods with an antifactor IIa assay for monitoring
bivalirudin anticoagulation. In their study, the TIM-ECT correlated better
with bivalirudin levels than either of the ACT methods.
Koster and colleagues (2003a) and Pötzsch and coworkers (1997) have
also questioned the validity of using the ACT for monitoring DTI therapy.
For cardiac surgery requiring CPB, they recommend intraoperative moni-
toring utilizing the ECT. In a case report, the ECT was maintained between
400 and 450 s (Koster et al., 2003a). A close relationship between the ECT and
bivalirudin concentrations was noted, but not with the ACT. However, Merry
et al. (2004) have suggested that the ACT can be used for off-pump procedures
when using lower bivalirudin concentrations.
As with other anticoagulants, monitoring is not reliable if the patient
has a lupus anticoagulant, low fibrinogen levels, elevated fibrinogen-fibrin
degradation products, or if the plasma contains heparin (Reid and Alving,
1993). In these situations, other tests including high-performance liquid chro-
matography, immunoassays, and chromogenic assays may be superior.
Although such assays have been used to measure levels of various DTIs
(Griessbach et al., 1985; Bichler et al., 1991; Spannagl et al., 1991; Walenga et
al., 1991), these assays are not widely available.
Reid and Alving (1993) developed a quantitative thrombin time (QTT)
in which bivalirudin (or hirudin) levels are measured using patient plasma (or
whole blood) mixed with human fibrinogen solution, with the clotting time
measured after adding human thrombin. The concentration of bivalirudin (or
hirudin) is then determined by comparison with a standard curve that is
generated by adding known concentrations of bivalirudin to pooled normal
plasma.
Figure 4 Prolongation of the activated partial thromboplastin time (aPTT) by

administration of subcutaneous bivalirudin. Three groups of study subjects containing
four subjects each were given an increasing dose of sc bivalirudin. Each line represents
the mean value of four study subjects. (From Fox et al., 1993.)
488 Bartholomew
Table 2 Drugs Incompatible

with Bivalirudin
Alteplase
Amiodarone hydrochloride
Amphotericin B
Chlorpromazine hydrochloride
Diazepam
Prochlorperazine edisylate
Reteplase
Streptokinase
Vancomycin hydrochloride
Source: Reed and Bell, 2002; The
Medicines Company, 2001.
H. Reversal
There is no specific antagonist to bivalirudin. If renal function is normal,
bivalirudin is eliminated rapidly, and its anticoagulant effect clears within a
few hours after discontinuing the infusion. Kaplan and Francis (2002) have
suggested that recombinant factor VIIa and desmopressin may be of benefit if
bleeding occurs. Bivalirudin can be removed by hemodialysis (Irvin et al.,
1999).
Koster and colleagues (2003b) recently demonstrated that large
amounts of bivalirudin can be removed by hemofiltration and plasmaphere-
sis. They utilized five different hemofilters in an in vitro study (conditions
mimicking CPB) and observed a correlation between pore size and elimina-
tion rate. In their study, 65% of bivalirudin was removed using a hemofilter
with a large pore size (65,000 Da) (Mintech Hemocor HPH 700, Minneapolis,
MN), an amount comparable to that eliminated with a plasmapheresis filter
system (69%). This represents a 50% improvement over the amount of
lepirudin that can be removed through filtration (moreover, lepirudin filtra-
tion correlates poorly with pore size). These authors suggest that hemofiltra-
tion using appropriate filters may be useful for routine management of
patients who receive bivalirudin for cardiac surgery.
I. Adverse Effects
Bleeding is the major adverse effect of bivalirudin and occurs more commonly
in patients with renal impairment. Injection site pain has been reported in
individuals given sc bivalirudin (Fox et al., 1993). Mild headache, diarrhea,
Table 3 Adverse Events Occurring in >5% of Patients

Receiving Bivalirudin or Heparin in Randomized Clinical
Trials
Bivalirudin Heparin
Event (n = 2161) (n = 2151)
Cardiovascular
Hypotension 262 (12%) 371 (17%)
Hypertension 135 (6%) 115 (5%)
Bradycardia 118 (5%) 164 (8%)
Gastrointestinal
Nausea 318 (15%) 347 (16%)
Vomiting 138 (6%) 169 (8%)
Dyspepsia 100 (5%) 111 (5%)
Genitourinary
Urinary retention 89 (4%) 98 (5%)
Miscellaneous
Back pain 916 (42%) 944 (44%)
Pain 330 (15%) 358 (17%)
Headache 264 (12%) 225 (10%)
Injection site pain 174 (8%) 274 (13%)
Insomnia 142 (7%) 139 (6%)
Pelvic pain 130 (6%) 169 (8%)
Anxiety 127 (6%) 140 (7%)
Abdominal pain 103 (5%) 104 (5%)
Fever 103 (5%) 108 (5%)
Nervousness 102 (5%) 87 (4%)
nausea, and abdominal cramps have also been reported (Fox et al., 1993). In
the Hirulog Angioplasty Study (HAS) (now known as the Bivalirudin
Angioplasty Trial [BAT]), the most frequent adverse effects included back
pain, nausea, hypotension, pain, and headache. Approximately 5–10% of
patients reported insomnia, hypertension, vomiting, anxiety, dyspepsia, bra-
dycardia, abdominal pain, fever, nervousness, pelvic pain, and pain at the in-
jection site (Bittl et al., 1995; Sciulli and Mauro, 2002) (Table 3).
III. CLINICAL USE OF BIVALIRUDIN (NON-HIT PATIENTS)

A. Treatment of Deep Vein Thrombosis
Bivalirudin is not approved for the treatment of venous thromboembolism.
However, it has been evaluated in animal models of venous and arterial throm-
490 Bartholomew
bosis and in one study involving humans. In a rat model of venous thrombosis
using injections of tissue thromboplastin combined with stasis, the adminis-
tration of bivalirudin demonstrated a dose-dependent interruption of throm-
bus formation (Maraganore et al., 1991).
Ginsberg et al. (1994a) studied iv and sc injections of bivalirudin in 10
patients with calf-vein thrombosis to determine if single injections could
inhibit thrombin generation in a sustained fashion. Prothrombin fragment
(F1+2) levels were used as an index of thrombin generation. Significant
reductions in F1+2 levels were noted at 6 h postinjection, but by 24 h levels
had increased significantly. These workers speculated that higher doses, more
frequent sc injections, or prolonged infusion was required to achieve ongoing
inhibition.
B. Prevention of Deep Vein Thrombosis

Bivalirudin has been evaluated for prevention of deep vein thrombosis (DVT)
in patients undergoing hip or knee surgery. In a phase II, open-label, dose-
optimization study of 222 patients, sc bivalirudin was given beginning 12–24 h
postoperatively for up to 14 days or until hospital discharge (Ginsberg et al.,
1994b). Five dose regimens were used, ranging from 0.3 mg/kg twice a day to
1.0 mg/kg three times a day (Table 4). Patients were evaluated for the oc-
currence of symptomatic DVT or pulmonary embolism (PE) within 72 h of
discontinuing bivalirudin, and assessment of distal or proximal DVT by
venography was performed on day 14 or just prior to discharge. Two patients
suffered PE while three patients had major bleeding. The rate of DVT ranged
from 59% in the lowest-dose regimen to only 17% in the highest-dose regi-
men (1.0 mg/kg three times a day). Proximal DVT occurred in only 2% of pa-
tients in the highest-dose regimen. Bleeding rates were low (<5%) with all
regimens.
C. Percutaneous Coronary Intervention

Bivalirudin has been studied for several cardiology indications, including
most prominently percutaneous coronary intervention (PCI), but also other
nonintervention cardiac situations (Table 5). Bivalirudin has been approved
by the FDA for use in patients with unstable angina undergoing PCI. To date,
over 80,000 patients have been treated with bivalirudin, with nearly 12,000
subjects enrolled in comparative trials with heparin for PCI. Bivalirudin is a
safe and effective alternative to heparin in this patient population.
The first clinical study using bivalirudin for coronary angioplasty was
reported by Topol and coworkers (1993) in a multicenter, open-label, dose-
Table 4 Efficacy and Safety of Bivalirudin in Preventing Deep Vein Thrombosis

After Major Hip or Knee Surgery
Bivalirudin dosing regimen
1.0 mg/kg 1.0 mg/kg

every 12 h for every 8 h
Efficacy or 0.3 mg/kg 0.6 mg/kg 3 d, then 0.6 mg/ 1.0 mg/kg (high-dose
safety endpoint every 12 h every 12 h kg every 12 h every 12 h regimen)
n 17 54 40 20 46
Overall DVT rate 10 (59%) 23 (43%) 16 (40%) 7 (35%) 8 (17%)a
Proximal DVT rate 7 (41%) 9 (17%) 6 (15%) 4 (20%) 1 (2%)b
Pulmonary embolism 0 2 (4%) 0 0 0
Major bleeding 0 1 (2%) 1 (3%) 0 1 (2%)
Minor bleeding 0 2 (4%) 0 1 (5%) 0
Venous thrombosis was documented by bilateral venography or by the occurrence of pulmonary

embolism. Of the 222 patients enrolled in the study, 177 patients had technically adequate bilateral
venography or clinically documented pulmonary embolism and were considered in the analysis of
efficacy. Major bleeding was defined as a fall in hemoglobin level of >2 g/dL or transfusion of >2 units
of blood. All other clinically overt bleeding was classified as minor.
Abbreviations: DVT, deep vein thrombosis.
a
Significantly lower overall DVT rate compared with the first four regimens combined: 8/46 (17%) vs.
56/131 (43%); p < 0.05
b
Significantly lower proximal DVT rate compared with the first regimens combined: 1/46 (2%) vs. 26/
131 (20%); p < 0.01.
Source: Ginsberg et al., 1994b.
finding trial of 258 patients. The encouraging results led to larger studies of
patients requiring urgent angioplasty because of unstable or postinfarction
angina, the Hirulog (bivalirudin) Angioplasty Study (HAS) (for review, see
Nawarskas and Anderson, 2001). The primary endpoint was in-hospital
death, myocardial infarction (MI), or abrupt vessel closure within 24 h of
initiating PCI, or rapid clinical deterioration of cardiac origin. In the original
publication, no statistically significant difference in the primary endpoint was
noted between bivalirudin and heparin (Bittl et al., 1995), causing the sponsor
(Biogen) to abandon further drug development.
Subsequently, The Medicines Company reanalyzed the trial data (in-
cluding an additional 214 patients analyzed by intention-to-treat principle
who were not included in the per-protocol analysis initially reported). In this
study, renamed as Bivalirudin Angioplasty Trial (BAT), the frequency of
secondary endpoints (including death or MI, and major hemorrhage) were
found to be significantly reduced with bivalirudin. Bivalirudin was at least as
effective as heparin in preventing ischemic complications in patients who
underwent angioplasty for unstable angina and included fewer episodes of
492 Bartholomew
Table 5 Major Clinical Studies Using Bivalirudin in Cardiac Patients (PCI and
Non-PCI Indications)
Study acronym
or description Trial (Ref.)
PCI indications
Dose-finding study Multicenter, open-label study (Topol et al., 1993)
HAS Hirulog (Bivalirudin) Angioplasty Study
(Bittl, 1995; Bittl et al., 1995)
BAT Bivalirudin Angioplasty Trial (Bittl et al., 2001)a
CACHET Comparison of Abciximab Complications with
Hirulog Ischemic Events Trial (Lincoff et al.,
2002b)
REPLACE-1 Randomized Evaluation in PCI Linking Angiomax
to Reduced Clinical Events (REPLACE)-1 Trial
(Lincoff et al., 2002a)
REPLACE-2 Randomized Evaluation in PCI Linking Angiomax
to Reduced Clinical Events (REPLACE)-2 Trial
(Lincoff et al., 2003)
Angiomax in Practice Cho et al. (2003b)
Registry
Non-PCI (unstable angina or acute MI)
TIMI-7 Thrombin Inhibition in Myocardial Ischemia-7
(Fuchs and Cannon, 1995)
HERO-1 Hirulog Early Reperfusion/Occlusion-1
(White et al., 1997)
HERO-2 Hirulog Early Reperfusion/Occlusion-2
(White, 2001)
a
Bittl et al. (1995) reported the first study (combining two randomized, controlled trials)
comparing bivalirudin against heparin for PCI; this study, subsequently called the Bivalirudin
Angioplasty Trial (BAT), was later reanalyzed (including data from an additional 214 pa-
tients) (Bittl et al., 2001).
major hemorrhage, retroperitoneal bleeding, and need for blood transfusion

(Topol et al., 1993; Bittl et al., 1995, 2001; Campbell et al., 2000a; Antman and
Braunwald, 2001).
CACHET (phases A, B, and C) evaluated the combination of bivalir-
udin plus the provisional use of a GP IIb/IIIa inhibitor (abciximab) in
comparison to heparin and abciximab in patients undergoing balloon angio-
plasty and stenting. Bivalirudin was found to be safe and effective with stents
and was associated with a lower combined incidence of death, MI, revascu-
larization or major hemorrhage at 7 days (Nawarskas and Anderson, 2001;
Lincoff et al., 2002b; Sciulli and Mauro, 2002).
In the REPLACE-1 trial, heparin was compared to bivalirudin in pa-

tients undergoing coronary stenting with any one of the GP IIb/IIIa inhibitors
(at the discretion of the physician) in 1056 patients. The combined endpoint of
death, MI, or revascularization showed a trend toward a reduction in bivali-
rudin-treated patients at 48 h (Lincoff et al., 2002a).
The REPLACE-2 trial was a randomized, double-blind, active-con-
trolled trial of 6010 patients who received bivalirudin with provisional use of
GPIIb/IIIa blockage or heparin with planned GP IIb/IIIa inhibition. Bivali-
rudin was found to be superior to heparin alone and as effective as heparin
plus GP IIb/IIIa inhibition for ischemic protection (Lincoff et al., 2003). A
significant reduction in the incidence of bleeding and thrombocytopenia were
also noted.
Bivalirudin may be a suitable substitute for heparin in patients with
chronic renal disease who require PCI because its clearance is primarily
determined by proteolysis and not by renal excretion (Robson et al., 2002).
ACT monitoring is recommended in patients with chronic renal disease. The
dose of bivalirudin should be reduced in accordance with the degree of renal
impairment (Table 1) (Robson, 2000; Robson et al., 2002).
D. Unstable Angina and Acute Myocardial Infarction

Some of the largest experience with bivalirudin is with patients who have had
an acute myocardial infarction (MI) or unstable angina. Two open-label, un-
controlled trials were performed to evaluate the efficacy and tolerability of
bivalirudin in patients with unstable angina. Sharma et al. (1993) utilized a
5-day infusion of bivalirudin in patients with unstable angina. Their primary
endpoints included death, development of an MI, or the need for coronary
intervention. Lidon and coworkers (1993) studied 55 patients in patients with
unstable angina in a dose-ranging study. As a result of favorable findings in
these two trials, the Thrombin Inhibition in Myocardial Infarction (TIMI) 7
trial comparing four different doses of bivalirudin in combination with aspirin
was performed in over 400 patients (Fuchs and Cannon, 1995). These trials
suggested that there is a role for bivalirudin in the management of unstable
angina.
A number of trials have evaluated the concomitant use of bivalirudin in
patients who received streptokinase and aspirin for an acute MI. Lidon et al.
1994) compared bivalirudin to heparin in 45 patients who suffered an acute
MI, while Theroux and colleagues (1995) utilized this same strategy in 68
patients. Higher early patency rates and a lower incidence of serious hemor-
rhage were noted (Nawarskas and Anderson, 2001).
The Hirulog Early Reperfusion/Occlusion (HERO) trial randomized
412 patients with acute MI to receive low-dose bivalirudin, high-dose bivali-
494 Bartholomew
rudin, or heparin (White et al., 1997). Bivalirudin was found to be more

effective than heparin in producing early patency rates at a reduced risk for
bleeding.
The HERO-2 trial randomized 17,073 patients who received streptoki-
nase to heparin or bivalirudin for 48 h in patients who presented with an acute
ST-elevation MI. Bivalirudin did not reduce mortality compared to heparin,
but was associated with a 30% reduction in repeat MI, without significant
increase in severe or life-threatening bleeding (White, 2001).
A meta-analysis by the Direct Thrombin Inhibitor Trialists’ Collabo-
rative Group (2002) based on individual patients’ data reported on 11 studies
(35,970 patients) receiving either heparin or DTI therapy (relative number of
patients treated: hirudin > bivalirudin > argatroban > inogatran >
efegatran). Overall, DTI therapy appeared to be superior over heparin for
the prevention of MI in patients with acute coronary syndromes (although the
larger number of patients treated with hirudin meant that this DTI contrib-
uted most to the overall result reported). Bivalirudin was associated with a
56% reduction in major bleeding risk.
E. Percutaneous Transluminal Angioplasty

There is limited experience using bivalirudin in the performance of PTA
involving the renal or other peripheral arteries. Allie et al. (2003) performed
180 renal and 75 iliac artery PTAs for patients with severe arterial disease
using bivalirudin as the only anticoagulant. Procedural success was achieved
in 100% of patients, and no adverse thrombotic events were reported. The
authors did note a decrease in sheath removal time, time to ambulation, and
length of hospital stay. A decrease in vascular access complications was seen.
Bivalirudin was felt to be a safe and reasonable alternative.
F. Off-Pump Coronary Artery Bypass Surgery

Merry et al. (2004) compared bivalirudin to unfractionated heparin for
OPCAB surgery in a semi-open label (surgeon-blinded), prospective study
of 100 patients (half receiving bivalirudin). The primary endpoint was 12-h
blood loss, and secondary endpoints were ischemic complications and cor-
onary artery patency at 12 weeks. No deaths were reported. The ACT took
longer to return to normal after stopping bivalirudin, when compared to the
heparin group (which received protamine reversal). Total blood loss was
similar in both groups, however. An intriguing (and potentially important)
finding was that graft patency was improved in the patients receiving
bivalirudin.
G. Pregnancy and Nursing Mothers

No evidence for impaired fertility or harm to the fetus has been attributed to
bivalirudin in teratogenicity studies performed on rats and rabbits using
higher doses than recommended for human use (The Medicines Company,
2001). There are no well-controlled studies in pregnant women. Caution is
advised when giving bivalirudin to nursing women, as it is not known whether
bivalirudin crosses the placenta or whether it is excreted in breast milk
(Carswell and Plosker, 2002).
H. Other Potential Uses

Bivalirudin has also been studied in animal models for its potential role in
both surgical and interventional fields. Its antithrombotic effects were first
studied in a baboon carotid endarterectomy model (Kelly et al., 1992). In later
studies using endarterectomized rats, significant decreases in platelet deposi-
tion with bivalirudin were shown using 111Indium-labeled platelets (Hamelink
et al., 1995) and scanning electron microscopy (Jackson et al., 1996).
Bivalirudin has also been studied for prevention of vascular restenosis
in a rat carotid artery injury model. Xue and associates (2000, 2001) found
that bivalirudin reduced platelet deposition on denuded intima. Platelet-
derived growth factor levels were also decreased following bivalirudin in-
fusion. The authors suggested that balloon catheter injury–induced neointima
formation might be suppressed by bivalirudin.
Bivalirudin has been administered to rabbits following balloon injury
and reduces vascular restenosis in the femoral artery of angioplasty-injured,
diet-induced atherosclerotic rabbits (Sarembock et al., 1996). These studies
support the possible role of thrombin in restenosis.
In contrast to the above study, Kranzhofer et al. (1999) administered
bivalirudin to rabbits over 3 days immediately after balloon injury to the
abdominal aorta and right iliac artery. Markers of inflammation, including
intercellular adhesion molecule-1, macrophage colony-stimulating factor,
tumor necrosis factor, and interleukin-1h, were examined by immunohisto-
chemistry. These workers found that bivalirudin did not acutely reduce
vascular smooth muscle cell proliferation or inflammation postangioplasty.
They did not rule out other mechanisms by which thrombin inhibition could
prevent restenosis.
Bivalirudin has also been shown to reduce thrombin-generated increase
in levels of plasminogen activator inhibitor-1 (PAI-1) in cultured baboon
aortic smooth muscle cells (Ren et al., 1997). Elevated levels of PAI-1 have
been found in patients with coronary artery disease (Hamsten et al., 1985;
Francis et al., 1988; Sakata et al., 1990), and numerous authors have sug-
496 Bartholomew
gested their role in the development of atherosclerosis and thrombosis (Ren et

al., 1997). Bivalirudin may potentially prevent intravascular thrombogenesis
through inhibition of thrombin-induced PAI-1 production (Ren et al., 1997;
Shen et al., 1998).
Bivalirudin has also been studied in a rat model of endotoxemia and
found to increase survival rate in one (but not the other) study (Cicala et al.,
1995; Itoh et al., 1996). Bivalirudin reduced endotoxin-induced thrombocy-
topenia, leukopenia, and fibrinogen consumption, suggesting a possible fu-
ture therapeutic role in sepsis (Cicala et al., 1995).
IV. BIVALIRUDIN FOR THE TREATMENT OF HIT

A. Miscellaneous Studies
Data on the use of bivalirudin in the treatment of HIT is limited. Chamberlin
et al. (1994) reported three patients who received bivalirudin for HIT. One
patient was treated for 8 days due to bilateral lower extremity DVTs and
recurrent PE with a positive heparin-induced platelet aggregation test, while
the other two patients received bivalirudin for arterial ischemia due to HIT.
One patient required an above-the-knee amputation and was given bivali-
rudin (for 12 days) to prevent loss of the other limb, while the other patient
had worsening peripheral arterial disease and underwent angioplasty of his
right superficial femoral artery using bivalirudin anticoagulation.
In another study, a total of 39 patients with HIT were treated with
bivalirudin (Berkowitz, 1999a; Campbell et al., 2000a; Gladwell, 2002).
Seventeen patients had acute HIT, while 22 had previous HIT. Patients were
treated for a variety of indications (Table 6). There were 4 deaths (10%),
all due to complications from HIT. Revascularization was successful in all
but one patient (94%) who had PCI. The one failure was attributed to an
unapproachable lesion. Two of the patients required intra-aortic balloon
pumps, while another two underwent successful coronary artery bypass
surgery. Bleeding complications were usually minor.
At the recent Congress of the International Society on Thrombosis and
Haemostasis (ISTH), Francis and colleagues (2003; and unpublished data)
presented their experience using bivalirudin to treat 40 patients with sus-
pected HIT (32 tested positive for HIT antibodies; 16 had thrombosis pre-
ceding bivalirudin treatment). The mean bivalirudin infusion rate (using a
target aPTT 1.5- to 2.5-fold greater than baseline) was 0.165 mg/kg/h (max-
imum 0.38 mg/kg/h). Bivalirudin was given for a mean of 8.7 days, with
transition to warfarin (mean overlap 5 days) performed in 35 of 40 patients.
The authors noted minimal increase in the INR on bivalirudin alone (mean
increase 0.31). Minor bleeding was seen in only a few patients. Antithrom-
Table 6 Indications for Bivalirudin Use in Patients with HIT
Indication Number treated (%)
Percutaneous coronary intervention (PCI) 17 (44)

Thrombosis 5 (13)
Coronary artery bypass grafting 4 (10)
Pulmonary embolism and deep vein thrombosis 4 (10)
Intra-aortic balloon pump 2 (5)
Cardiac catheterization 2 (5)
Unstable angina 1 (3)
Deep vein thrombosis 1 (3)
Pulmonary thromboendarterectomy 1 (3)
Aortic reconstructive surgery 1 (3)
Femoral bypass grafting 1 (3)
____
Total 39
Source: Campbell et al., 2000a.
botic efficacy was believed by the authors to be acceptable, but detailed results
are forthcoming.
There is one case report of long-term (2-month) use of bivalirudin to
treat serologically confirmed HIT complicated by recurrent left leg ischemia
and arterial thrombosis while on low molecular weight heparin (Bufton et al.,
2002a). The patient received a continuous infusion of bivalirudin (22 mg/h)
using a CADD pump.
Table 7 summarizes theoretical advantages of bivalirudin as a treat-
ment for HIT.
B. Bivalirudin for Percutaneous Coronary Intervention

The ATBAT trial has recently been completed (Campbell et al., 2000b;
Mahaffey et al., 2003). This was a prospective, open-label study to evaluate
the safety and efficacy of bivalirudin in patients with acute HIT or a past
history of HIT undergoing PCI. The primary endpoint was major bleeding
within 48 h after completion of the bivalirudin infusion (1.0 mg/kg/h iv bolus
followed by 2.5 mg/kg/h by iv infusion for 4 h). This dose was later changed
to a 0.75 mg/kv/iv bolus followed by a 1.75 mg/kg/h infusion for 4 h. Sec-
ondary endpoints included event rates for components of the primary end-
point and the ACT, aPTT, and platelet counts (at baseline, pre-PCI/post-PCI,
and prior to discharge). Clinical success was defined as procedural success
without death, emergency bypass surgery, or q-wave MI. Only one of 52 pa-
tients required a blood transfusion (1 U), and procedural and clinical success
498 Bartholomew
were achieved in 98% and 96% of the patients, respectively. There were no
abrupt closures, nor was thrombus formation reported during or after PCI.
One patient died of cardiac arrest about 46 h after successful PCI.
C. Bivalirudin for Cardiac Surgery

Bivalirudin has been used off-label for cardiac surgery in a number of patients
with acute or previous HIT. Both ‘‘on-pump’’ and ‘‘off-pump’’ experience has
been reported. Except for a recent trial comparing bivalirudin with heparin
(Merry et al., 2004), experience has been anecdotal.
Off-Pump Coronary Artery Bypass Surgery

Spiess et al. (2002) recently reported a case of OPCAB surgery using
bivalirudin for intraoperative anticoagulation in a patient with HIT. Bivali-
rudin was chosen over hirudin because of the patient’s renal insufficiency.
The patient received a 0.75 mg/kg bolus followed by a continuous infusion of
1.75 mg/kg/h. Monitoring was performed using both the ACT and the ECT.
The infusion rate was increased to 2.0 mg/kg/h when the ACT fell below 300 s.
Table 7 Theoretical Advantages of Bivalirudin for Treatment of HIT
Feature of bivalirudin Comment
Short half-life (25–36 min) Avoids need for initial iv bolus; rapid reversal
of anticoagulation (useful if patient develops
bleeding or if used for intraoperative
anticoagulation)a
Predominant enzymic Minor renal excretion (20%) means that risk of
metabolism overdosing in renal failure less than with
lepirudin; less risk of postoperative bleeding
(compared with lepirudin) if used for in-
traoperative anticoagulation (in case of
postoperative renal insufficiency)b
Minimal effect on PT/INR Simplifies transition to oral anticoagulation
(compared with argatroban)
Low immunogenicity Reduced risk of allergy and anaphylaxis
(compared with lepirudin)
Abbreviation: iv, intravenous.
a
Possible disadvantages of a short half-life include need for frequent sc administration (e.g.,
three or four times daily) and rapid loss of anticoagulation (with risk of rebound thrombosis)
if prematurely discontinued in patient with acute HIT.
b
Possible disadvantage of enzymic metabolism includes loss of anticoagulant action in stagnant
blood (implications for cardiac anesthesiology) (see Chap. 19).
The patient required 2 units of packed red blood cells intraoperatively and
underwent surgical reexploration 8 h postoperatively because of increased
chest tube drainage. No bleeding was found, and the patient’s recovery was
otherwise uneventful.
On-Pump (Cardiopulmonary Bypass) Cardiac Surgery

Vasquez et al. (2002) utilized bivalirudin for anticoagulation during CPB in a
patient with infective endocarditis in whom HIT was suspected. Bivalirudin
was administered as a 1.25 mg/kg bolus followed by a continuous iv infusion
of 2.75 mg/kg/h, with the dose adjusted upwards to maintain an ACT of
between 500 and 600 s. The patient required 4 units of packed red blood cells
and recovered uneventfully. No clot formation was observed in the CPB
circuit, and postoperative bleeding was considered acceptable.
Davis and coworkers (2003) reported the use of bivalirudin for a patient
with a history of HIT who required CPB for aortic valve replacement and
coronary artery bypass grafting. The authors used a 50 mg bolus of bivali-
rudin followed by an infusion of 1.5–1.75 mg/kg/h and found that adequate
anticoagulation was obtained. The patient required 3 units of packed red
blood cells within the first 6 h postoperatively, plus 7 units of fresh frozen
plasma, 2 sets of platelets, and an additional unit of packed red blood cells
within the first 24 h after surgery. The recovery was otherwise uneventful,
although the authors did note clot formation in the cell-saving device and
recommended adding bivalirudin to the pump circuit when using this device.
The ACT steadily declined after stopping the bivalirudin infusion.
Koster et al. (2003a) reported the use of bivalirudin in a patient with
HIT, a history of severe anaphylactic reaction to heparin, and renal insuffi-
ciency in a patient requiring repeat open-heart surgery. They used the ECT for
monitoring anticoagulation during CPB. The authors gave the patient a 1.5
mg/kg bolus of bivalirudin followed by 2.5 mg/kg/h continuous infusion. The
infusion was increased to 5.0 mg/kg/h to maintain the ECT between 400 and
450 s. The patient required 2 units of red blood cells and 4 units of plasma and
recovered uneventfully.
Current Studies of Bivalirudin During Cardiac Surgery

Based on these reports, and buoyed by the experience with bivalirudin in
OPCAB surgery from New Zealand (Merry et al., 2004), four phase II and III
multicenter trials utilizing bivalirudin in patients with HIT (CHOOSE) and
comparing bivalirudin to heparin with protamine reversal (EVOLUTION)
are underway. The rationale for using bivalirudin in these settings includes its
direct thrombin inhibition without the requirement of a cofactor, its rapid,
dose-dependent prolongation of the ACT, its short half-life, lack of structural
500 Bartholomew
similarity to heparin (thus, no cross-reactivity with PF4/heparin-dependent

antibodies), avoidance of protamine use (and its potentially-severe adverse
reactions), no need for dose reduction in mild renal impairment, and an ability
to ‘‘reverse’’ its anticoagulant effect through hemofiltration, all of which make
it a drug potentially superior to heparin in cardiac surgery. Further, there is
the potential to avoid HIT antibody formation and, consequently, postoper-
ative HIT.
V. ANTIBIVALIRUDIN ANTIBODIES
Bivalirudin is a relatively small polypeptide and thus is expected to lack

significant antigenicity (Fenton et al., 1998). In a study of plasma samples
from 7 patients, no evidence for antibody formation (IgG, IgM, or IgE) was
found (Fox et al., 1993) with plasma samples obtained at 7 and 14 days after iv
administration. There was also no evidence for changes in the pharmacoki-
netics or pharmacodynamics of bivalirudin in their study. One patient
exhibited antibody titers of greater than 1:2,000 in the assay prior to
administration of bivalirudin, although no explanation was given.
In another review of 494 bivalirudin-treated patients from 9 different
studies, 11 subjects initially tested positive for antibivalirudin antibodies
(Berkowitz, 1999b). However, 9 of these were found to be false positives on
repeat testing. The remaining two (who could not be retested) did not develop
any allergic or anaphylactic reactions. In clinical trials of bivalirudin per-
formed from 1993 to 1995, only 1 of 3639 patients (0.03%) experienced an
allergic reaction considered by the investigator to be related to study drug.
Since bivalirudin shares an 11-amino-acid sequence with hirudin, it
is at least theoretically possible that patients with antilepirudin antibodies
resulting from treatment with lepirudin could cross-react with bivalirudin.
Recently, Eichler and colleagues (2004) found that 22 of 43 (51%) sera con-
taining antilepirudin antibodies showed reactivity in vitro against bivalirudin.
This suggests that if bivalirudin is used in patients previously treated with
lepirudin, extra caution should be used, e.g., careful anticoagulant monitor-
ing, as antilepirudin antibodies sometimes influence pharmacokinetics.
VI. COST ANALYSIS WITH BIVALIRUDIN
Bivalirudin is the only anticoagulant associated with lower rates of both is-
chemic and bleeding complications compared to heparin in studies of PCI.
These complications are associated with increased morbidity and mortality
and also higher costs (Lauer, 2000; Compton, 2002). Bivalirudin may also be
associated with shorter hospital stay, use of fewer closure devices, lower in-
cidence of hematoma formation, earlier sheath removal, and more selective
use of the GP IIb/IIIa inhibitors. Potential savings are also possible in pa-
tients treated for HIT by reducing its devastating and costly thrombotic
complications.
VII. CONCLUSION
Bivalirudin is a unique new anticoagulant with a number of potential

applications. The FDA has approved it for use in PCI. It has extensive
experience in patients with unstable angina and MI. Data are accumulating on
using bivalirudin for HIT, and it may prove especially useful as an alternative
anticoagulant in patients with acute HIT requiring cardiac surgery (Warken-
tin and Greinacher, 2003). Additionally, based on the OPCAB experience
from New Zealand (Merry et al., 2004), bivalirudin has the potential to
become the anticoagulant of choice for heart surgery (and thus avoid HIT in
the post–cardiac surgery setting). Its short half-life, unique metabolism
(enzymic) and low immunogenicity provide it with distinct advantages over
other DTIs. In addition, its reversible thrombin inhibition may be associated
with decreased bleeding risk. Finally, although there are no antidotes avail-
able, the potential for reversibility with hemofiltration (which can be used
routinely in the postcardiac surgery setting) adds to its attractiveness.
REFERENCES
Allie DE, Lirtzman MD, Wyatt CH, Keller VA, Khan MH, Khan MA, Fail PS,
Hebert CJ, Ellis SD, Mitran E, Chaisson G, Stagg S Jr, Allie AA, Walker CM.
Bivalirudin as a foundation anticoagulant in peripheral vascular disease: a safe
and feasible alternative for renal and iliac interventions. J Invasive Cardiol 2003;
15:334–342.
Alving BM. How I treat heparin-induced thrombocytopenia and thrombosis. Blood
2003; 101:31–37.
Antman EM, Braunwald E. A second look at bivalirudin. Am Heart J 2001; 142:929–
931.
Bates SM, Weitz JI. Direct thrombin inhibitors for treatment of arterial thrombosis:
potential differences between bivalirudin and hirudin. Am J Cardiol 1998;
82:12–18P.
Bates SM, Weitz JI. The mechanism of action of thrombin inhibitors. J Invasive
Cardiol 2000; 12(suppl F):27F–32F.
Berkowitz SD. Bivalirudin in heparin-induced thrombocytopenia (HIT) or heparin-
502 Bartholomew
induced thrombocytopenia and thrombosis syndrome (HITTS) patients [abstr].

Blood 1999a; 94(suppl 1):101b.
Berkowitz SD. Antigenic potential of bivalirudin [abstr]. Blood 1999b; 94(suppl 1):
102b.
Bichler J, Siebeck M, Maschler R, Pelzer H, Fritz H. Determination of thrombin-
hirudin complex in plasma with an enzyme-linked immunoabsorbent assay.
Bittl JA. Comparative safety profiles of hirulog and heparin in patients undergoing
coronary angioplasty. The Hirulog Angioplasty Study Investigators. Am Heart
J 1995; 130:658–665.
Bittl JA, Strony J, Brinker JA, Ahmed WH, Meckel CR, Chaitman BR, Maraganore J,
Deutsch E, Adelman B. Treatment with bivalirudin (Hirulog) as compared with
heparin during coronary angioplasty for unstable or postinfarction angina.
Hirulog Angioplasty Study Investigators. N Engl J Med 1995; 333:764–769.
Bittl JA, Chaitman BR, Feit F, Kimball W, Topol EJ. Bivalirudin versus heparin
during coronary angioplasty for unstable or postinfarction angina: final report
reanalysis of the Bivalirudin Angioplasty Study. Am Heart J 2001; 142:952–959.
Bufton MG, Rubin WD, Springhorn ME, Miller CL, Senuty EJ, Gannon MK.
Bivalirudin effect on the INR and experience with prolonged inpatient and
outpatient anticoagulation with bivalirudin for treatment of leg ischemia and
arterial thrombosis due to HIT-TS [abstr]. Blood 2002a; 100:124b.
Bufton MG, Rubin WD, Penz JF. The effect of bivalirudin on the INR and pro-
thrombin time when used as a peri-operative anticoagulant in an obese patient
with panniculitis and a history of HIT-TS [abstr]. Blood 2002b; 100:124b.
Campbell KR, Mahaffey KW, Lewis BE, Weitz JI, Berkowitz SD, Ohman EM, Califf
RM. Bivalirudin in patients with heparin-induced thrombocytopenia under-
going percutaneous coronary intervention. J Invasive Cardiol 2000a; 12(suppl
F):14F–19F.
Campbell KR, Wildermann N, Janning C, Lewis B, Kelton J, Green D, Kottke-
Marchant K, Berkowitz SD, Mahaffey KW. Bivalirudin during percutaneous
coronary intervention in patients with heparin-induced thrombocytopenia:
interim results of the ATBAT Trial [abstr]. Am J Cardiol 2000b; 86(suppl 1):73i–
74i.
Cannon CP, Maraganore JM, Loscalzo J, McAllister A, Eddings K, George D, Selwyn
AP, Adelman B, Fox I, Braunwald E, Ganz P. Anticoagulant effects of hirulog,
a novel thrombin inhibitor, in patients with coronary artery disease. Am J
Cardiol 1993; 71:778–782.
Carswell CI, Plosker GL. Bivalirudin: a review of its potential place in the man-
agement of acute coronary syndromes. Drugs 2002; 62:841–870.
Chamberlin JR, Lewis B, Leya F, Wallis D, Messmore H, Hoppensteadt D, Walenga
JM, Moran S, Fareed J, McKiernan T. Successful treatment of heparin-asso-
ciated thrombocytopenia and thrombosis using Hirulog. Can J Cardiol 1994;
11:511–514.
Chew DP, Bhatt DL, Lincoff AM, Moliterno DJ, Brener SJ, Wolski KE, Topol EJ.
Defining the optimal activated clotting time during percutaneous coronary
intervention: aggregate results from 6 randomized, controlled trials. Circulation

2001; 103:961–966.
Cho L, Kottke-Marchant K, Lincoff AM, Roffi M, Reginelli JP, Kaldus T, Moliterno
DJ. Correlation of point-of care ecarin clotting time versus activated clotting
time with bivalirudin concentrations. Am J Cardiol 2003a; 91:1110–1113.
Cho L, Chew DP, Moliterno DJ, Roffi M, Ellis SG, Franco I, Bajzer C, Bhatt DL,
Dorosti K, Simpfendorder C, et al. Safe and efficacious use of bivalirudin for
percutaneous coronary intervention with adjunctive platelet glycoprotein IIb/
IIIa receptor inhibition. Am J Cardiol 2003b; 91:742–743.
Chong BH. Heparin-induced thrombocytopenia. J Thromb Haemost 2003; 1:1471–
1478.
Cicala C, Bucci MR, Maraganore JM, Cirino G. Hirulog effect in rat endotoxin shock.
Life Sci 1995; 57:307–313.
Compton A. A practical cost analysis of bivalirudin. Pharmacotherapy 2002; 22:119S–
127S.
Dager WE, White RH. Treatment of heparin-induced thrombocytopenia. Ann Phar-
macother 2002; 36:489–503.
Davis Z, Anderson R, Short D, Garber D, Valgiusti A. Favorable outcome with
bivalirudin anticoagulation during cardiopulmonary bypass. Ann Thorac Surg
2003; 75:264–265.
de Denus S, Spinler SA. Clinical monitoring of direct thrombin inhibitors using the
ecarin clotting time. Pharmacotherapy 2002; 22:433–435.
Despotis GJ, Hogue CW, Saleem R, Bigham M, Skubas N, Apostolidou I, Qayam A,
Joist JH. The relationship between hirudin and activated clotting time: impli-
cations for patients with heparin-induced thrombocytopenia undergoing car-
diac surgery. Anesth Analg 2001; 93:28–32.
Direct Thrombin Inhibitor Trialists’ Collaborative Group. Direct thrombin inhibitors
in acute coronary syndromes: principal results of a meta-analysis based on
individual patients’ data. Lancet 2002; 359:294–302.
Eichler P, Lubenow N, Strobel U, Greinacher A. Antibodies against lepirudin are
polyspecific and recognize epitopes on bivalirudin. Blood 2004. In press.
Fareed J, Lewis BE, Callas DD, Hoppensteadt DA, Walenga JM, Bick RL. Anti-
thrombin agents: the new class of anticoagulant and antithrombotic drugs. Clin
Appl Thromb Hemost 1999; 5(suppl 1):S45–S55.
Fenton JW II, Ofosu FA, Brezniak DV, Hassouna HI. Thrombin and antithrom-
botics. Semin Thromb Hemost 1998; 24:87–91.
Fox I, Dawson A, Loynds P, Eisner J, Findlen K, Levin E, Hanson D, Mant T,
Wagner J, Maraganore J. Anticoagulant activity of Hirulogk, a direct throm-
bin inhibitor, in humans. Thromb Haemost 1993; 69:157–163.
Francis RB Jr, Kawanishi D, Baruch T, Mahrer P, Rahimtoola S, Feinstein DI.
Impaired fibrinolysis in coronary artery disease. Am Heart J 1988; 115:776–780.
Francis J, Drexler A, Gwyn G, Moroose R. Bivalirudin, a direct thrombin inhibitor, in
the treatment of heparin-induced thrombocytopenia [abstr]. J Thromb Haemost
2003; 1(suppl):1909.
Fuchs J, Cannon CP. Hirulog in the treatment of unstable angina. Results of the
504 Bartholomew
Thrombin Inhibition in Myocardial Ischemia (TIMI) 7 trial. Circulation 1995;

92:727–733.
Ginsberg JS, Nurmohamed MT, Gent M, MacKinnon B, Stevens P, Weitz J, Mara-
ganore J, Hirsh J. Effects on thrombin generation of single injections of Hi-
rulog in patients with calf vein thrombosis. Thromb Haemost 1994a; 72:523–
525.
Ginsberg JS, Nurmohamed MT, Gent M, MacKinnon B, Sicurella J, Brill-Edwards P,
Levine MN, Panju AA, Powers P, Stevens P, Turpie AGG, Weitz J, Buller HR,
ten Cate JW, Neemeh J, Adelman B, Fox I, Maraganore J, Hirsh J. Use of
Hirulog in the prevention of venous thrombosis after major hip or knee surgery.
Circulation 1994b; 90:2385–2389.
Gladwell TD. Bivalirudin: a direct thrombin inhibitor. Clin Ther 2002; 24:38–58.
Griessbach U, Sturzebecher J, Markwardt F. Assay of hirudin in plasma using a
chomogenic thrombin substrate. Thromb Res 1985; 37:347–350.
Hamelink JK, Tang DB, Barr CF, Jackson MR, Reid TJ, Gomez ER, Alving BM.
Inhibition of platelet deposition by combined hirulog and aspirin in a rat carotid
endarterectomy model. J Vasc Surg 1995; 21:492–498.
Hamsten A, Wiman B, de Faire U, Blomback M. Increased plasma levels of a rapid
inhibitor of tissue plasminogen activator in young survivors of myocardial
infarction. N Engl J Med 1985; 313:1557–1563.
Irvin W, Sica D, Gehr T, McAllister A, Rogge M, Charenkavanich S, Adelman B.
Pharmacodynamics (PD) and kinetics (PK) of bivalirudin (BIV) in renal failure
(RF) and hemodialysis (HD) [abstr]. Clin Pharmacol Ther 1999; 65:202.
Itoh H, Cicala C, Douglas GJ, Page CP. Platelet accumulation induced by bacterial
endotoxin in rats. Thromb Res 1996; 83:405–419.
Jackson MR, Reid TJ, Tang DB, O’Donnell SD, Gomez ER, Alving BM. Anti-
thrombotic effects of hirulog in a rat carotid endarterectomy model. J Surg Res
1996; 60:15–22.
Kaplan KL, Francis CW. Direct thrombin inhibitors. Semin Hematol 2002; 39:187–
196.
Kelly AB, Maraganore JM, Bourdon P, Hanson SR, Harker LA. Antithrombotic
effects of synthetic peptides targeting various functional domains of thrombin.
Proc Natl Acad Sci USA 1992; 89:6040–6044.
Mertzlufft F. Hirudin monitoring using TAS ecarin clotting time in patients
2000; 14:249–252.
Koster A, Chew D, Grundel M, Bauer M, Kuppe H, Spiess BD. Bivalirudin monitored
with the ecarin clotting time for anticoagulation during cardiopulmonary
bypass. Anesth Analg 2003a; 96:383–386.
Koster A, Chew D, Grundel M, Hausmann H, Grauhan O, Kuppe H, Spiess BD. An
assessment of different filter systems for extracorporeal elimination of bival-
irudin: an in vitro study. Anesth Analg 2003b; 96:1316–1319.
Kranzhofer R, Maraganore JM, Baciu R, Libby P. Systemic thrombin inhibition by
Hirulog does not alter medial smooth muscle cell proliferation and inflamma-
tory activation after vascular injury in the rabbit. Cardiovasc Drugs Ther 1999;
13:429–434.
Lauer MA. Cost analysis of bivalirudin in percutaneous coronary intervention. J
Invasive Cardiol 2000; 12(suppl F):37F–40F.
Lidon RM, Theroux P, Juneau M, Adelman B, Maraganore J. Initial experience with a
direct antithrombin, Hirulog, in unstable angina. Anticoagulant, antithrom-
botic, and clinical effects. Circulation 1993; 88:1495–1501.
Lidon RM, Theroux P, Lesperance J, Adelman B, Bonan R, Duval D, Levesque J. A
pilot, early angiographic patency study using a direct thrombin inhibitor as
adjunctive therapy to streptokinase in acute myocardial infarction. Circulation
1994; 89:1567–1572.
Lincoff AM, Bittl JA, Kleiman NS, Kereiakes DJ, Harrington RA, Sarembook IJ,
Jackman JD, Mehta S, Maierson EF, Chew DP, Topol EJ. The REPLACE 1
Trial: a pilot study of bivalirudin versus heparin during percutaneous coronary
intervention with stenting and GP IIb/IIIa blockade [abstr]. J Am Coll Cardiol
2002a; 39(suppl A):16A.
Lincoff AM, Kleiman NS, Kottke-Marchant K, Maierson ES, Maresh K, Wolski KE,
Topol EJ. Bivalirudin with planned or provisional abciximab versus low-dose
heparin and abciximab during percutaneous coronary revascularization: results
of the Comparison of Abciximab Complications with Hirulog for Ischemic
Events Trial (CACHET). Am Heart J 2002b; 143:847–853.
Lincoff AM, Bittl JA, Harrington RA, Feit F, Kleiman NS, Jackman JD, Sarembock
IJ, Cohen DJ, Spriggs D, Ebrahimi R, Keren G, Carr J, Cohen EA, Betriu A,
Desmet W, Kereiakes DJ, Rutsch W, Wilcox RG, deFeyter PJ, Vahanian A,
Topol EJ. Bivalirudin and provisional glycoprotein IIb/IIIa blockade compared
with heparin and planned glycoprotein IIb/IIIa blockade during percutaneous
coronary intervention: REPLACE-2 randomized trial. JAMA 2003; 289:853–
863.
Mahaffey KW. Anticoagulation for acute coronary syndromes and percutaneous
coronary intervention in patients with heparin-induced thrombocytopenia.
Curr Cardiol Rep 2001; 3:362–370.
Mahaffey KW, Lewis BE, Wildermann NM, Berkowitz SD, Oliverio RM, Turco MA,
Shalev Y, Lee PV, Traverse JH, Rodriguez AR, Ohman EM, Harrington RA,
ATBAT Investigators. J Invasive Cardiol 2003. In press.
Maraganore JM, Adelman BA. Hirulog: a direct thrombin inhibitor for management
of acute coronary syndromes. Coron Artery Dis 1996; 7:438–448.
Maraganore JM, Bourdon P, Jablonski J, Ramachandran KL, Fenton JW II. Design
and characterization of hirulogs: a novel class of bivalent peptide inhibitors of
thrombin. Biochemistry 1990; 29:7095–7101.
Maraganore T, Oshima FA, Sugitachi A. Comparison of anticoagulant and anti-
thrombotic activities of hirulog-1 and argatroban (MD-805) [abstr]. Thromb
Haemost 1991; 65:651.
Merry AF, Raudkivi P, White HD, Nand P, Middleton N, McDougall J, Mills BP,
Frampton CM. Bivalirudin versus heparin and protamine in off pump coronary
artery bypass surgery. Ann Thorac Surg 2004. In press.
506 Bartholomew
Nawarskas JJ, Anderson JR. Bivalirudin: a new approach to anticoagulation. Heart

Dis 2001; 3:131–137.
Parry MA, Maraganore JM, Stone SR. Kinetic mechanism for the interaction of
Hirulog with thrombin. Biochemistry 1994; 33:14807–14814.
Pötzsch B, Hund S, Madlener K, Unkrig C, Muller-Berhaus G. Monitoring of re-
combinant hirudin: assessment of a plasma-based ecarin clotting time assay.
Thromb Res 1997; 86:373–383.
Reed MD, Bell D. Clinical pharmacology of bivalirudin. Pharmacotherapy 2002; 22:
105S–111S.
Reid TJ III, Alving BM. A quantitative thrombin time for determining levels of hiru-
din and Hirulog. Thromb Haemost 1993; 70:608–616.
Ren S, Fenton JW II, Maraganore JM, Angel A, Shen GX. Inhibition by Hirulog-1 of
generation of plasminogen activator inhibitor-1 from vascular smooth-muscle
cells induced by thrombin. J Cardiovasc Pharmacol 1997; 29:337–342.
Robson R. The use of bivalirudin in patients with renal impairment. J Invasive Cardiol
2000; 12(suppl F):33F–36F.
Robson R, White H, Aylward P, Frampton C. Bivalirudin pharmacokinetics and
pharmacodynamics: effect of renal function, dose, and gender. Clin Pharmacol
Ther 2002; 71:433–439.
Sakata K, Kurata C, Taguchi T, Suzuki S, Kobayashi A, Yamazaki N, Rydzewski A,
Takada Y, Takada A. Clinical significance of plasminogen activator inhibitor in
patients with exercise-induced ischemia. Am Heart J 1990; 120:831–838.
Sarembock IJ, Gertz SD, Thome LM, McCoy KW, Ragosta M, Powers ER,
Maraganore JM, Gimple LW. Effectiveness of hirulog in reducing restenosis
after balloon angioplasty of atherosclerotic femoral arteries in rabbits. J Vasc
Res 1996; 33:308–314.
Scatena R. Bivalirudin: a new generation antithrombotic drug. Exp Opin Invest Drugs
2000; 9:1119–1127.
Sciulli TM, Mauro VF. Pharmacology and clinical use of bivalirudin. Ann Pharma-
cother 2002; 36:1028–1041.
Sharma GVRK, Lapsley D, Vita JA, Sharma S, Coccio E, Adelman B, Loscalzo J.
Usefulness and tolerability of hirulog, a direct thrombin-inhibitor, in unstable
angina pectoris. Am J Cardiol 1993; 72:1357–1360.
Shen GX, Ren S, Fenton JW II. Transcellular signaling and pharmacological mod-
ulation of thrombin-induced production of plasminogen activator inhibitor-1 in
vascular smooth muscle cells. Semin Thromb Hemost 1998; 24:151–156.
Spannagl M, Bichler J, Birg A, Lill H, Schramm W. Development of a chromogenic
substrate assay for the determination of hirudin in plasma. Blood Coagul
Fibrinol 1991; 2:121–127.
Spiess BD, DeAnda A, McCarthy A, Yeatman D, Harness HL, Katlaps G. Off pump
CABG in a patient with HITT anticoagulated with bivalirudin: a case report
[abstr]. Anesth Analg 2002; 93:SCA70.
The Medicines Company. Angiomax (bivalirudin) package insert. Cambridge, MA,
2001.
Theroux P, Perez-Villa F, Waters D, Lesperance J, Shabani F, Bonan R. Randomized
double-blind comparison of two doses of Hirulog with heparin as adjunctive

therapy to streptokinase to promote early patency of the infarct-related artery in
acute myocardial infarction. Circulation 1995; 91:2132–2139.
Thiagarajan P, Wu KK. Mechanisms of antithrombotic drugs. Adv Pharmacol 1999;
46:297–324.
Topol EJ, Bonan R, Jewitt D, Sigwart U, Kakkar VV, Rothman M, de Bono D,
Ferguson J, Willerson JT, Strony J, Ganz P, Cohen MD, Raymond R, Fox I,
Maraganore J, Adelman B. Use of a direct antithrombin, hirulog, in place of
heparin during coronary angioplasty. Circulation 1993; 87:1622–1629.
Vasquez JC, Vichiendilokkul A, Mahmood S, Baciewicz FA Jr. Anticoagulation with
bivalirudin during cardiopulmonary bypass in cardiac surgery. Ann Thorac
Surg 2002; 74: 2177–2179.
Walenga JM, Hoppensteadt D, Koza M, Pifarre R, Fareed J. Comparative studies
on various assays for the laboratory evaluation of r-hirudin. Semin Thromb
Hemost 1991; 17:103–112.
Br J Haematol 2003; 121:535–555.
gery. Ann Thorac Surg 2003; 638–648. [Corrected article to be reprinted in its
Weitz JI, Hirsh J. New antithrombotic agents. Chest 1998; 114(suppl):715S–727S.
Weitz J, Maraganore JM. The thrombin-specific anticoagulant, bivalirudin, com-
pletely inhibits thrombin-mediated platelet aggregation [abstr]. Am J Cardiol
2001; 88(suppl 5A):83G.
White HD. Thrombin-specific anticoagulation with bivalirudin versus heparin in
patients receiving fibrinolytic therapy for acute myocardial infarction: the
HERO-2 randomised trial. Lancet 2001; 358:1855–1863.
White HD, Chew DP. Bivalirudin: an anticoagulant for acute coronary syndromes
and coronary interventions. Expert Opin Pharmacother 2002; 3:777–788.
White HD, Aylward PE, Frey MJ, Adgey AAJ, Nair R, Hillis WS, Shalev Y, Brown
MA, French JK, Collins R, Maraganore J, Adelman B. Randomized, double-
blind comparison of hirulog versus heparin in patients receiving streptokinase
and aspirin for acute myocardial infarction (HERO). Circulation 1997; 96:
2155–2161.
Wiggins BS, Spinler S, Wittkowsky AK, Stringer KA. Bivalirudin: a direct thrombin
inhibitor for percutaneous transluminal coronary angioplasty. Pharmacother-
apy 2002; 22:1007–1018.
Wittkowsky AK. The role of thrombin inhibition during percutaneous coronary
intervention. Pharmacotherapy 2002; 22:97S–104S.
Xue M, Fenton JW II, Shen GX. Hirulog-1 reduces expression of platelet-derived
growth factor in neointima of rat carotid artery induced by balloon catheter
injury. J Vasc Res 2000; 37:82–92.
Xue M, Ren S, Welch S, Shen GX. Hirulog-like peptide reduces balloon catheter
injury induced neointima formation in rat carotid artery without increase in
bleeding tendency. J Vasc Res 2001; 38:144–152.
18
Hemodialysis in Heparin-Induced
Thrombocytopenia
Karl-Georg Fischer
University Hospital Freiburg, Freiburg, Germany
I. HEPARIN-INDUCED THROMBOCYTOPENIA
IN HEMODIALYSIS PATIENTS
Because unfractionated heparin (UFH) is the major anticoagulant in hemo-

dialysis (HD), it is important to define the potential role of heparin-induced
thrombocytopenia (HIT) in contributing to morbidity and mortality in
patients with dialysis-dependent renal failure. Of 154 patients newly treated
with HD, 6 (3.9%) were clinically suspected of having developed HIT because
of a fall in the platelet count accompanied by clotting of the dialyzer and
extracorporeal circuit (Yamamoto et al., 1996). The clinical diagnosis was
confirmed by the detection of HIT antibodies in all but one patient. Only one
patient developed organ damage from thrombosis (myocardial infarction and
stroke). All six patients were switched to an alternative anticoagulant and did
not suffer from thromboembolic events in the follow-up period. Compared
with the incidence of HIT of 2.7% found in 332 hip surgery patients treated
with UFH (Warkentin et al., 1995), the incidence of HIT in acute hemodialy-
sis patients thus appears to be similar, regardless of the underlying cause of
renal dysfunction (Finazzi and Remuzzi, 1996).
Greinacher and colleagues (1996) performed a cross-sectional study of
165 patients undergoing hemodialysis using UFH and identified 7 (4.2%)
patients as having HIT antibodies using a sensitive activation assay for HIT;
however, there was no difference in the incidence of thromboembolic events
in HIT antibody–positive patients, when compared with HIT antibody–
509
510 Fischer
negative patients. They considered alternative anticoagulants to be justified

only if clinical symptoms of HIT occurred.
Similar findings were reported that used an antigen assay for HIT anti-
bodies (platelet factor 4 [PF4]–heparin enzyme immunoassay [EIA]) in four
other cross-sectional studies of HD patients. In patients undergoing HD using
UFH, the frequency of HIT IgG antibodies varied: 0% (0/45) (de Sancho et
al., 1996), 2.3% (3/128) (Boon et al., 1996), 2.8% (2/170) (Sitter et al., 1998)
and 6% (3/50) (Luzzatto et al., 1998). For patients undergoing HD using low
molecular weight heparin (LMWH), the frequency in one study was 0.3%
(1/133) (Boon et al., 1996). Thrombocytopenia was usually not observed in
patients who formed HIT antibodies, and none of the patients developed
bleeding or thrombosis.
Tentative conclusions suggested by these studies are that only a few
patients who form HIT antibodies in association with HD develop clinical
events and that these are more likely to be clotting of the dialyzer and extra-
corporeal circuit than symptomatic thrombosis affecting the patient. It is also
possible that the risk of clinical HIT is higher in patients undergoing short-
term HD (the population studied by Yamamoto et al., 1996) than in patients
in the long-term phase of HD (as per the remaining studies). Anecdotal case
reports of HIT complicating HD also seem frequently to include patients
undergoing short-term HD (Matsuo et al., 1989; Hall et al., 1992; Nowak
et al., 1997; Gupta et al., 1998).
II. CLINICAL PRESENTATION OF HIT IN HEMODIALYSIS

PATIENTS
The diagnosis of HIT and respective management decisions should be pri-

marily based on clinical criteria (Lewis et al., 1997). A further consideration
in HD patients is that the procedure of HD itself is associated with a relative
decrease in platelet count, even when so-called biocompatible dialyzer mem-
branes are used (Beijering et al., 1997; Schmitt et al., 1987). Furthermore, the
fall in platelet count in HD patients developing HIT may be only moderate
(Matsuo et al., 1997).
The occurrence of fibrin formation, or even frank clotting of the extra-
corporeal circuit, despite apparent sufficient anticoagulation, should lead to a
strong suggestion of possible HIT (Koide et al., 1995). One of the most serious
complications, occlusion of vascular access (the ‘‘Achilles’ heel’’ of HD), may
also indicate HIT, and it has been described both for native fistulae as well
as prosthetic grafts (Hall et al., 1992; Laster et al., 1989). However, vascular
Hemodialysis in HIT 511
access thrombosis is not frequently associated with HIT. Of 88 HD patients

prospectively evaluated for the presence of HIT antibodies, 18 (20%) had a
prior history of access thrombosis, but only one patient (1.1%) without a
history of graft thrombosis tested positive for HIT antibodies (O’Shea et al.,
2002). Severe skin necrosis, even in the presence of normal platelet count, has
been reported in association with the presence of HIT antibodies in patients
after both short- and long-term HD (Bredlich et al., 1997; Leblanc et al.,
1994).
Rarely, patients can develop HIT after years of regular long-term main-
tenance HD. Tholl et al. (1997) reported on a patient developing HIT fol-
lowing surgery after 9 years of long-term intermittent HD performed with
UFH. In this patient, an anaphylactic reaction to heparin, accompanied by a
platelet count fall, led to the diagnosis of HIT. It is possible that the surgery
itself contributed to HIT antibody formation, as the highest reported rates of
HIT are in postoperative patients receiving UFH (see Chap. 4).
Unfortunately, HD complications associated with HIT are not very
specific. Thus, the clinician must consider other factors that could compro-
mise patency of the extracorporeal circuit (e.g., low blood flow, high ultra-
filtration rate, excess turbulence within the circuit, or foam formation with
blood-air interfaces in the drip chambers). The quality of the vascular access
plays a crucial role in this. Other patient-related factors include low arterial
blood pressure, high hematocrit, and the need for intradialytic blood trans-
fusion or lipid infusion (Hertel et al., 2001). In addition to insufficient anti-
coagulation, these factors should be ruled out first as the underlying causes
of clotting within the extracorporeal circuit before HIT is considered in the
differential diagnosis. Given the long-term implications of labeling HD
patients as having HIT, laboratory testing for HIT antibodies should be
performed when HIT is clinically suspected (O’Shea et al., 2003).
III. MANAGEMENT OF HEMODIALYSIS IN HIT PATIENTS

A. Discontinuation of Heparin Treatment
As HIT is frequently associated with potentially life-threatening thrombotic
events (Warkentin et al., 1995; Warkentin and Kelton, 1996), discontinuation
of heparin treatment and initiation of adequate alternative anticoagulation is
generally considered mandatory (Warkentin et al., 1998). Thus, heparin must
not be added to any flushing solution, and no heparin-coated systems can be
used. Indeed, heparin flushes and heparin-coated devices can both initiate and
sustain HIT (Moberg et al., 1990; Kadidal et al., 1999).
512 Fischer
B. Unsuitable Approaches
Low Molecular Weight Heparin
LMWH is not recommended as an alternative anticoagulant. In vitro tests for
HIT antibodies show a high degree of cross-reactivity between UFH and
LMWH (Greinacher et al., 1992b; Vun et al., 1996). Furthermore, in vivo
cross-reactivity manifesting as persistent or recurrent thrombocytopenia or
thrombosis during LMWH treatment of HIT appears to be common (Grei-
nacher et al., 1992a; Horellou et al., 1984; Roussi et al., 1984). Because non-
heparin anticoagulants are available, LMWH should not be used even if in
vitro cross-reactivity is reported to be negative.
Regional Heparinization
Regional heparinization is defined as application of heparin at the inlet of
the extracorporeal circuit and its neutralization by protamine at the outlet
of the circuit. However, its use in HIT is problematic because of the potential
for heparin ‘‘contamination’’ of the patient, as well as for heparin ‘‘rebound
anticoagulation’’ (recurrence of heparin anticoagulation owing to shorter
half-life of protamine compared with heparin) (Blaufox et al., 1966). More-
over, direct injurious effects of protamine on the clotting cascade can occur.
Consequently, this regimen is not recommended for HD of patients with HIT.
Aspirin
Acetylsalicylic acid has been used as an antiplatelet agent together with con-
tinued anticoagulation with UFH for HD of patients with HIT (Hall et al.,
1992; Janson et al., 1983; Matsuo et al., 1989). This approach is not recom-
mended for several reasons: (1) protection against heparin-induced platelet
activation may be incomplete or absent, as aspirin’s effects on blocking the
thromboxane-dependent pathway of platelet activation does not reliably in-
hibit platelet activation by HIT antibodies (Kappa et al., 1987; Polgár et al.,
1998); (2) the bleeding risk of uremic patients is increased; and (3) theoret-
ically, it may lead to induction of persistently high levels of HIT antibodies.
Hemodialysis Without Anticoagulant

Hemodialysis without an anticoagulant (Romao et al., 1997) is not adequate
for maintenance HD. Without anticoagulation, the artificial surfaces become
coated, first by plasma proteins, followed by adhesion and activation of
platelets, with accompanying activation of the coagulation cascade (Basmad-
jian et al., 1997). This will markedly reduce dialysis quality in removal of fluid
and solutes long before clotting of the circuit is visible. Moreover, this ap-
proach may aggravate HIT. However, in patients at high risk of bleeding

(e.g., owing to hepatic disorders or multiorgan failure, or those requiring
surgery), temporary hemodialysis without anticoagulant may be appropriate.
C. Adequate Anticoagulants for Hemodialysis

in HIT Patients
Patients with renal failure show plasma hypercoagulability as well as uremic
platelet defects, both of which can be worsened by HD (Ambühl et al., 1997;
Sreedhara et al., 1995; Vecino et al., 1998). Therefore, selection of an appro-
priate anticoagulant in HD patients who also suffer from HIT is difficult.
Reports on specific anticoagulant strategies in HIT are anecdotal. Large
studies, especially those comparing different anticoagulant regimens, are
lacking. Therefore, no treatment recommendations based on level A or B
evidence can be provided. Furthermore, because UFH is the routine antico-
agulant in use for HD, considerable additional time, effort, and costs are
usually required to manage a new anticoagulant for HD, especially during
initial use. Ideally, therefore, a center should try to gain experience with a
single appropriate alternative anticoagulant for management of these diffi-
cult patients. Fear of inducing bleeding should not be used to justify under-
anticoagulation, with the potential risk for thrombotic complications.
Danaparoid Sodium
Danaparoid sodium (Orgaran, formerly known as Org 10172) is the alterna-
tive anticoagulant that has been most widely used for management of HD in
patients with HIT (Chong and Magnani, 1992; Greinacher et al., 1992a, 1993;
Henny et al., 1983; Magnani, 1993; Neuhaus et al., 2000; Ortel et al., 1992;
Roe et al., 1998; Tholl et al., 1997; Wilde and Markham, 1997). However,
danaparoid has been withdrawn from certain markets, such as the United
States and the United Kingdom (see Chap. 14). Some of its characteristics
require specific attention:
1. The anticoagulant activity of danaparoid can be monitored only by
measurement of antifactor Xa levels based on a danaparoid calibration curve;
however, many laboratories do not routinely perform these assays. Except for
an emergency situation, such as when HIT is strongly suspected and danapa-
roid is the only available alternative, HD should not be performed without
monitoring the antifactor Xa activity to evaluate the dose required for ade-
quate anticoagulation. Once the optimal dose is identified, it can often be used
without alteration for several subsequent HD sessions, provided no bleeding
or inappropriate clotting occurs and no surgical intervention is scheduled.
Periodic measurement of antifactor Xa activity to validate the appropriate
514 Fischer
dosage of danaparoid is recommended. For maintenance HD without com-

plications, single determination of pre-HD antifactor Xa activity probably
suffices. If there are concerns about adequate or excess anticoagulation, then
monitoring of levels at three time points is appropriate (e.g., 30–60 min pre-
HD, 30 min after beginning HD, and just before completion).
2. Regarding the pharmacokinetics of danaparoid, renal excretion ac-
counts for approximately 40–50% of total plasma clearance; accordingly,
diminished clearance of antifactor Xa activity occurs in hemodialysis patients
(Danhof et al., 1992). The elimination half-life of the antifactor Xa activity
(about 24 h in healthy individuals) (Danhof et al., 1992) may reach as high as
4 days (unpublished observations of the author). Thus, significant antifac-
tor Xa levels can be detected in patients undergoing HD with danaparoid
even during the interdialytic interval. Whether this yields clinical benefit, such
as decreased risk of thrombosis or greater maintenance of vascular access,
is unknown. An increase in interdialytic bleeding episodes has not been
reported.
3. Given its pharmacokinetics, danaparoid is given by initial bolus in
intermittent HD, which normally is sufficient to prevent clotting within the
extracorporeal circuit during the procedure. Danaparoid anticoagulation
may also be required in critically ill patients on continuous renal replacement
therapy. In 13 consecutive intensive care patients clinically suspected to have
HIT, danaparoid was administered by initial bolus followed by continuous
infusion (Lindhoff-Last et al., 2001). This regimen was sufficient to prevent
clotting within the extracorporeal circuit both in continuous venovenous
hemofiltration (8 patients) and in continuous venovenous hemodialysis (5
patients), respectively (Lindhoff-Last et al., 2001). Thromboembolic compli-
cations did not occur. Despite a mean danaparoid infusion rate of approxi-
mately 140 U/h, which is markedly reduced compared to the recommendation
of the manufacturer, major bleeding was observed in 6 of 13 patients (which
could be explained by disseminated intravascular coagulation in 5 patients).
However, HIT was confirmed by antibody detection in only 2 patients.
Thrombocytopenic patients not having the prothrombotic state of acute
HIT likely are at increased bleeding risk. Therefore, dosing of danaparoid
in intensive care patients should be based on the individual patient’s risk of
bleeding versus thrombosis. With regard to invasive procedures, the long half-
life of danaparoid should be considered.
4. No antidote to danaparoid exists. Recently, we evaluated hemofil-
tration as a potential means to rapidly reduce danaparoid plasma concentra-
tion. Whereas five different high-flux hemodialyzer membranes did not allow
for danaparoid filtration, a plasmapheresis membrane was capable of remov-
ing danaparoid from the blood compartment (unpublished results). Hence,
plasmapheresis may be a way to reduce danaparoid levels in situations of
overdosing or bleeding. Again, careful dosing of danaparoid is important to

avoid bleeding.
5. HIT antibodies potentially cross-react with danaparoid. Although
the respective clinical risk has been claimed to be less than 5% (Warkentin
et al., 1998), individual patients, nevertheless, may be severely threatened if
this condition occurs. This may be especially true for maintenance HD pa-
tients, who would be exposed to danaparoid repeatedly. As positive in vitro
cross-reactivity is of uncertain clinical significance (Warkentin, 1996; Wilde
and Markham, 1997; Newman et al., 1998), attention should focus on platelet
count monitoring. A further fall in platelet count, or new fibrin deposits and
clot formation within the extracorporeal circuit after application of danapa-
roid, may indicate clinically relevant cross-reactivity. To differentiate in vivo
cross-reactivity from ‘‘under-anticoagulation’’ owing to insufficient dosage,
determination of antifactor Xa levels and HIT antibody cross-reactivity
studies are needed.
Table 1 lists dose recommendations for use of danaparoid for HD as
provided by the manufacturer. The recommendations should be considered as
guidelines and not followed uncritically in any individual patient. If applied
with appropriate care, danaparoid provides adequate anticoagulation for HD
of HIT patients with a favorable benefit/risk ratio, even during long-term use.
Recombinant Hirudin (r-Hirudin)

Native hirudin was the first anticoagulant used for HD over 75 years ago
(Haas, 1925). In recent years, interest in its use for HD has redeveloped
because of the availability of recombinant preparations, as well as the clinical
need for managing patients with HIT. A preparation of r-hirudin, lepirudin
(Refludan or HBW023), has been used successfully in humans for antico-
agulation of both intermittent (Bucha et al., 1999a; Nowak et al., 1992, 1997;
Steuer et al., 1999; Vanholder et al., 1994; Van Wyk et al., 1995) and contin-
uous HD (Fischer et al., 1999; Schneider et al., 2000; Saner et al., 2001; Vargas
Hein et al., 2001).
For use of r-hirudin anticoagulation in HD, some aspects should be
specifically addressed. For further information on recombinant hirudin in
renal insufficiency, the reader is referred to a recent review (Fischer, 2002):
1. As there is repetitive exposure to r-hirudin when used for regular,
intermittent HD, immunogenicity of r-hirudin is of particular interest. Ini-
tially, r-hirudin appeared to be a weak immunogen (Bichler et al., 1991).
However, recent studies revealed frequent development of antihirudin anti-
bodies (AHAb) in patients receiving lepirudin for more than 5 days (Huhle
et al., 1999, 2001; Song et al., 1999; Eichler et al., 2000). In addition, allergic
reactions to r-hirudin were reported (Huhle et al., 1998; Eichler et al., 2000).
516
Table 1 Alternative Anticoagulation for Hemodialysis and Hemofiltration of Patients with HIT
Continuous Monitoring Target
Agent Dialysis Procedure Bolus infusion parametera range
Danaparoid Intermittent HD Before first 2 HDs 3750 (2500)b,c — Anti-Xa activity 0.5–0.8d,e
sodium (Orgaran) (every 2nd day) Subsequent HD Predialytic anti-Xa
activity
(U/mL)d,f
<0.3 3000 (2000)b,c —
0.3–0.35 2500 (1500)b,c —
0.35–0.4 2000 (1500)b,c —
>0.4 0g —
Intermittent HD 1st HD 3750 (2500)b,c — Anti-Xa activity 0.5–0.8d,e
(daily) 2nd HD 2500 (2000)b,c —
Subsequent HD See above —
Continuous HD/HF Initial bolus 2500 (2000)b,c
First 4 h — 600 (600)c,h Anti-Xa activity 0.5–1.0d
Next 4 h — 400 (400)c,h
Subsequently — 200–600g,h,i
(150–400)c,g,h,i
Lepirudin Intermittent HD 0.08–0.15j,k,l — aPTT ratiom,n 2–2.5e,o
(Refludan) (every 2nd day) Hirudin conc.p 0.5–0.8e,q
Continuous HDr,s Initial bolus 0.01j,k,l,t — aPTT ratiom,n 1.5–2.0u,v
Subsequent boluses 0.005–0.01j,k,l,t —
Alternatively — 0.005–0.01t,w,x
Argatroban Intermittent HDr 0.1j 0.1–0.2x aPTT ratiom 1.5–3.0o
(Novastany) (every 2nd day)
Many of the approaches listed have not been formally studied, and none is approved. Treatment examples are given based on a limited number of cases successfully treated with
the respective regimen. The different anticoagulants thus cannot be uncritically applied in the dosages given here. The choice of anticoagulant should depend on the experience
Fischer
of the center and the anticoagulant monitoring available. Doses for danaparoid as given by the manufacturer.
Abbreviations: HD, hemodialysis; HF, hemofiltration; conc., concentration; Xa, factor Xa; aPTT, activated partial thromboplastin time.
a
Monitoring the condition of the dialyzer after a HD session as well as the time required for termination of bleeding of the fistula should also be included.
b
Dosage for bolus given in anti-Xa units (U).
c
Dosage in parentheses for patients with body weight <55 kg.
d
Target range given in anti-factor Xa U/mL.
e
Peak activity determined after about 30 min of HD; this level is not required throughout the whole HD session.
f
Determination 30–60 min before start of the respective HD session.
g
If fibrin deposition in the dialyzer or clots in the extracorporeal circuit occur, give bolus of 1500 anti-Xa U.
h
Hemodialysis in HIT
Dosage given in anti-Xa U/h (infusion).

i
Maintenance dosage dependent on actual anti-Xa activity; determination every 12 h (provided that no bleeding or clotting occurs).
j
Dosage given in mg/kg body weight for hemodialysis performed with polysulfone high-flux hemodialyzers.
k
The dosage required to reach the target range may vary, e.g., owing to residual renal function or the type of dialyzer used (see text).
l
If larger doses are needed to achieve the target range or to avoid clotting of the extracorporeal circuit, changing to another type of dialyzer may be helpful.
m
The aPTT ratio is determined using the mean of the laboratory normal range; according to the literature, alternative tests such as ecarin-clotting time or chromogenic assays
appear superior for monitoring.
n
It is unclear which test is best suited to monitor anticoagulation with r-hirudin, as no test has been prospectively evaluated in HD patients.
o
A peak aPTT of 100 s should not be exceeded.
p
Determination in plasma by chromogenic assays.
q
Target range given in Ag/mL.
r
The agent has not been formally studied in continuous HD procedures.
s
This approach has been successfully performed in several patients in our center without adverse events.
t
Dosage given for anuric patients; in case of polyuria a higher dosage may be required; the daily dosage may vary significantly among patients.
u
As patients requiring continuous procedures often are at increased risk of bleeding, a lower aPTT is preferred (50–70 s).
v
To be initially controlled every 4–6 h to avoid overdosage, especially in patients at risk of bleeding.
w
In our experience, a continuous infusion is more often associated with bleeding events.
x
Dosage given in mg/kg body weight/h.
y
In the United States, argatroban is marketed under the name Argatroban.
517
518 Fischer
r-Hirudin is increasingly used for alternative anticoagulation in HD. Here,

repetitive application of r-hirudin in patients on an intermittent maintenance
HD regimen is likely to favor both induction and boostering of an immune
response against the drug. As prospective studies evaluating sufficient num-
bers of HD patients on r-hirudin anticoagulation for the generation of AHAb
are lacking, the incidence of AHAb and related adverse clinical events in this
patient population remain to be elucidated.
Studies of HIT patients treated with lepirudin suggest that AHAb some-
times reduce renal lepirudin clearance (Huhle et al., 1999; Eichler et al., 2000).
Indeed, marked reduction of renal lepirudin clearance due to monoclonal
AHAb has recently been demonstrated in rats with normal renal function
(Fischer et al., 2003). This was accompanied by a significant increase of both
maximal plasma concentration and area under the curve of the alternative
anticoagulant when compared to non–AHAb-treated animals. In chronic re-
nal failure patients undergoing HD this may not be an issue. However, even
small reductions in residual renal function have been shown to account for
relevant prolongation of r-hirudin decay in plasma (Bucha et al., 1999a; Van-
holder et al., 1997). Further reduction of renal r-hirudin clearance due to
AHAb thus may influence r-hirudin dosing in these patients.
In acute renal failure requiring HD treatment for a prolonged period,
reduction of renal r-hirudin clearance attributable to AHAb may be more
relevant. Here, in patients suffering from multiorgan failure, the r-hirudin
dosage required for sufficient anticoagulation was reduced significantly com-
pared with the dosage needed in patients with normal renal function. In addi-
tion, r-hirudin dosage varied markedly depending on the residual renal
function (Fischer et al., 1999). AHAb are likely to reduce further the amount
of r-hirudin required, and thus may complicate anticoagulation in this chal-
lenging patient population.
The animal study also showed a significant decrease in the volume of
distribution of lepirudin at steady state in the presence of AHAb (Fischer
et al., 2003). Hence, even if further reduction of renal r-hirudin clearance
owing to AHAb was negligible, major alterations in r-hirudin plasma concen-
tration could still occur.
2. There remains debate as to which laboratory parameter is best
suited for monitoring r-hirudin treatment. Initial studies addressing this in
HD patients yielded conflicting results (Vanholder et al., 1994, 1997; van
Wyk, 1995). However, it now appears that the ecarin clotting time (ECT)
(Nowak and Bucha, 1996) and chromogenic substrate assays (Griessbach
et al., 1985) measure the r-hirudin plasma concentration with adequate pre-
cision over a wide concentration range and correlate well with each other
(Hafner et al., 2000, 2002). However, as these tests are often not available,
monitoring of r-hirudin anticoagulation is usually performed with the acti-
vated partial thromboplastin time (aPTT). A meta-analysis of two lepirudin

treatment trials for HIT revealed the optimal aPTT ratio for reducing clinical
thromboembolic complications to be between 1.5 and 2.5, which was associ-
ated with only a moderately increased bleeding risk (Greinacher et al., 2000).
Control of r-hirudin treatment by the aPTT is problematic: there is con-
siderable assay variability among patients and different aPTT reagents (Nur-
mohamed et al., 1994; Hafner et al., 2000; Lubenow and Greinacher, 2000).
In contrast to the foregoing tests, correlation between aPTT and plasma
r-hirudin concentration is not linear over a broad concentration range. In-
stead, linear correlation is observed only with r-hirudin concentrations up to
0.5 Ag/mL (Nowak and Bucha, 1996), a concentration often insufficient for
HD. Above this concentration, the correlation between aPTT and r-hirudin
concentration is poor (Nowak and Bucha, 1996; Hafner et al., 2000), es-
pecially for aPTT values of more than 70 s (Lubenow and Greinacher, 2000).
Nevertheless, because of its wide availability, aPTT monitoring of r-hirudin
treatment is likely to remain common. If available, ECT or chromogenic as-
says are preferred.
3. The elimination of r-hirudin is markedly prolonged in renal im-
pairment. Nowak and colleagues (1992) reported elimination half-lives of up
to 316 h in dialysis patients. Vanholder and coworkers (1997) found a
prolongation of r-hirudin half-life by a factor of 31 in HD patients compared
with healthy controls. Both studies showed a correlation between the residual
creatinine clearance and the r-hirudin clearance, in that a minor improvement
in creatinine clearance resulted in a shorter elimination half-life of r-hirudin.
This was confirmed in a recent study of HD patients repetitively anticoagu-
lated with r-hirudin (Bucha et al., 1999a). As with HD patients treated with
danaparoid, r-hirudin–treated patients remain anticoagulated during the
interdialytic interval (Nowak et al., 1997). Because various organs may me-
tabolize hirudin (Grötsch and Hropot, 1991), other factors affecting meta-
bolic clearance of hirudin may be present in patients with end-stage renal
failure.
In patients suffering from acute renal failure, further deterioration or
partial recovery of renal function frequently occurs (Fischer et al., 1999).
Hence, r-hirudin anticoagulation should be closely monitored in these
patients for timely dose adjustments. Preferably, r-hirudin should be given
in repeated small boluses, rather than administered continuously, to minimize
bleeding risk (Fischer et al., 1999; Kern et al., 1999). For the same reason, use
of r-hirudin permeable high-flux hemodialyzers for patients with HD-depen-
dent acute renal failure is recommended, especially as the patients often need
vessel punctures, biopsies, or surgical interventions.
Given the prolonged half-life of r-hirudin in renal impairment, use of
polyethylene glycol-hirudin (molecular mass 17 kDa), which has an even
520 Fischer
greater elimination half-life compared with uncoupled r-hirudin (Pöschel

et al., 2000), does not seem appropriate for HD, as bleeding risk likely would
be increased.
4. Pharmacokinetics of r-hirudin are also influenced by the type of
dialyzer used. The pharmacology of r-hirudin (molecular mass f7 kDa; vol-
ume of distribution 0.2–0.25 L/kg b.w.; low protein binding) should favor its
elimination by high flux hemodialyzers with a nominal cutoff point of ap-
proximately 60 kDa. Indeed, recent studies confirm high-flux hemodialyzers
to be permeable to r-hirudin, whereas most of the low-flux hemodialyzers
tested appear to be r-hirudin–impermeable (Bucha et al., 1999b; Frank et al.,
1999, 2002; Benz et al., 2000; Koster et al., 2000). However, high-flux hemo-
dialyzers vary considerably in their capacity to filter r-hirudin (Fischer, 2002;
Willey et al., 2002). Further, a specific type of hemophan low-flux dialyzer has
been reported to show high permeability for r-hirudin (Nowak et al., 1997),
whereas a specific type of polysulfone high-flux dialyzer, with a cutoff point of
approximately 50 kDa, did not filter r-hirudin from the circulation (Van-
holder et al., 1997). Thus, knowledge of the actual filtration characteristics
for r-hirudin of a given type of hemodialyzer improves safety of treatment
with r-hirudin in HD.
5. r-Hirudin overdosing or unexpected drug accumulation can lead to
severe bleeding (Fischer et al., 2000; Kern et al., 1999; Müller et al., 1999). In
this situation, r-hirudin can be removed from the circulation using hemofil-
tration (Bauersachs, 1999; Fischer et al., 2000). However, several hours may
be needed to lower r-hirudin plasma levels by 50%, even at high ultrafiltration
rates. Thus, careful r-hirudin dosing is of utmost importance. In the presence
of AHAb, hemofiltration may no longer suffice to eliminate r-hirudin (Fischer
et al., 2003). Here, plasmapheresis may be the only means to clear r-hirudin
from the circulation. Preliminary studies in animals suggest that a possible
future treatment might be use of certain AHAb with r-hirudin–neutralizing
capacity (Liebe et al., 2001).
Table 1 lists dosing recommendations for use of lepirudin for HD. The
recommendations should be considered as guidelines and not followed un-
critically in any individual patient. In summary, r-hirudin is a valid alterna-
tive anticoagulant for HD procedures in HIT patients, but it should be used
with caution and careful monitoring.
Argatroban
Argatroban (Novastan; MD-805) is a potent arginine-derived, synthetic, cat-
alytic site-directed thrombin inhibitor lacking antiplatelet and antifibrinolytic
activities (Koide et al., 1995; Matsuo et al., 1992). This agent is approved in
the United States and Canada as a treatment for HIT (see Chaps. 1, 13, and
16). It does not cross-react with HIT antibodies. Apart from an even better
relative ability to inhibit fibrin-bound versus soluble thrombin (Berry et al.,
1996; Lunven et al., 1996), the principal advantages of argatroban over
heparin are similar to r-hirudin (Markwardt, 1991; Matsuo et al., 1992).
However, argatroban is metabolized primarily by the liver, and its half-life is
only moderately extended in patients with renal insufficiency.
After argatroban proved to be a valuable anticoagulant in HD (Matsuo
et al., 1986), it was applied successfully to HIT patients undergoing this
procedure (Koide et al., 1995; Matsuo et al., 1992). In the studies ARG-911,
ARG-915, and ARG-915X, no differences in the primary efficacy endpoint or
in bleeding were observed in 54 HD patients compared to non-HD patients
being treated with comparable doses of argatroban. A recent prospective
crossover study of 12 maintenance HD patients showed three different arga-
troban dosing regimens (bolus alone, infusion alone, or bolus plus infusion) to
be safe and well tolerated (Murray et al., 2003) (see also Chap. 16). Arga-
troban also proved effective and safe in HD patients with antithrombin
deficiency (Ota et al., 2003). Whether anticoagulation with argatroban alone
is always sufficient to prevent clotting in the extracorporeal circuit is unclear:
in one HD patient treated with argatroban, marked spontaneous platelet
aggregation occurred, perhaps due to HIT together with additional platelet
activation known to occur in HD (Koide et al., 1995). Because platelet aggre-
gation could not be suppressed by argatroban alone in this patient, aspirin was
added to achieve patency of the extracorporeal circuit.
As nonspecific inactivation of argatroban may occur in blood, periodic
monitoring of its anticoagulant activity is recommended (Matsuo et al., 1992);
for example, by measuring the aPTT (Koide et al., 1995; Matsuo et al., 1992)
or the ECT (Berry et al., 1998).
Argatroban appears to be at least as well suited as r-hirudin for anti-
coagulation of HIT patients requiring HD. Its predominant hepatic elimina-
tion theoretically favors argatroban for alternative anticoagulation in renal
failure.
Oral Anticoagulation
For HIT patients requiring long-term anticoagulation, orally active agents
are usually given. Although coumarins interfere with the clotting cascade,
fibrin formation within the extracorporeal circuit is not always sufficiently
blocked. In these cases, additional low-dose intravenous anticoagulation with
UFH is usually given for regular maintenance HD. However, in HIT patients
requiring HD, alternative low-dose anticoagulation has not been formally
studied. The need for additional intravenous anticoagulation depends on the
increase of the INR, which should be checked regularly before HD. Priming
522 Fischer
of the extracorporeal circuit by addition of a compatible anticoagulant to the

filling solution with subsequent washout before start of the respective HD
session may be of value in diminishing the risk of ‘‘overanticoagulation.’’
D. Other Approaches
Dermatan Sulfate
Dermatan sulfate is a natural glycosaminoglycan that selectively inhibits both
soluble and fibrin-bound thrombin through potentiation of endogenous
heparin cofactor II. It does not interfere with platelet function. Dermatan
sulfate has been used successfully to anticoagulate patients with HIT (Agnelli
et al., 1994), and has also been applied successfully as an anticoagulant for
HD (Boccardo et al., 1997).
Nafamostat Mesilate
Nafamostat mesilate (FUT-175), a synthetic nonspecific serine protease
inhibitor with a short half-life, has been evaluated for regional hemodialysis
in patients at risk of bleeding (Akizawa et al., 1993). It has also been applied
occasionally to HIT patients on HD (Koide et al., 1995). However, owing to
significant clot formation at the dialyzer outlet, despite a twofold prolonga-
tion of aPTT, reported both in HIT and non-HIT patients (Koide et al., 1995;
Matsuo et al., 1993; Takahashi et al., 2003), this anticoagulant cannot cur-
rently be recommended for HD of HIT patients.
Prostacyclin
Prostacyclin (PGI2, epoprostenol), a potent antiplatelet agent with a short
half-life, has been evaluated both as a substitute for, and as an adjunct to
standard heparin for HD of patients with acute or chronic renal insufficiency
(Turney et al., 1980; Smith et al., 1982; Samuelsson et al., 1995). Adverse
effects, such as nausea, vomiting, and hypotension, can be avoided by dose
reduction, use of bicarbonate- instead of acetate-containing dialysate, or
infusion of the drug at the inlet of the extracorporeal circuit. Because of its
mode of action, prostacyclin cannot inhibit activation of coagulation during
HD (Rylance et al., 1985; Novacek et al., 1997). Moreover, in a HIT patient
receiving continuous venovenous HD, prostacyclin was unable to suppress
platelet consumption effectively after heparin had been reinstituted, owing to
a false-negative platelet aggregation assay (Samuelsson et al., 1995). Prosta-
cyclin does not seem to be a suitable antithrombotic agent for HD in HIT.
Whether it may be a useful adjunct in selected cases remains to be clarified.
Regional Citrate Anticoagulation

Anticoagulation by regional citrate is based on the concept of inhibition of
clotting by chelation of ionized calcium, and it was first developed as an
alternative anticoagulant regimen in HD patients at risk of bleeding (Pinnick
et al., 1983). Metabolic alkalosis, hypernatremia, alterations in calcium ho-
meostasis, and hyperalbuminemia are reported side effects that are generally
manageable (Ward and Mehta, 1993; Flanigan et al., 1996; Janssen et al.,
1996). Regional citrate anticoagulation is a valuable approach in experienced
centers. Recently, efficient and safe long-term citrate anticoagulation in a HIT
patient over a period of 9 months was reported (Unver et al., 2002). Regional
citrate anticoagulation is a treatment option only in patients with a history of
HIT as it does not suppress the prothrombotic state in acute HIT.
IV. SUMMARY
An alternative anticoagulant is required for HD in patients with HIT. Ap-

propriate agents would appear to be danaparoid sodium; r-hirudin deriva-
tives, such as lepirudin; or argatroban, as these appear to be able to suppress
clot formation without substantially increasing the bleeding risk. As these
results are based on experience with a limited number of patients, larger
prospective trials are needed to define the best treatment options in this set-
ting. Even today, though, HIT should no longer be a life-threatening problem
for patients requiring dialysis.
REFERENCES
Agnelli G, Iorio A, De Angelis V, Nenci GG. Dermatan sulphate in heparin-induced

thrombocytopenia [letter]. Lancet 1994; 344:1295–1296.
Akizawa T, Koshikawa S, Ota K, Kazama M, Mimura N, Hirasawa Y. Nafamostat
mesilate: a regional anticoagulant for hemodialysis in patients at high risk for
bleeding. Nephron 1993; 64:376–381.
Ambühl PM, Wüthrich RP, Korte W, Schmid L, Krapf R. Plasma hypercoagulability
in haemodialysis patients: impact of dialysis and anticoagulation. Nephrol Dial
Transplant 1997; 12:2355–2364.
Basmadjian D, Sefton MV, Baldwin SA. Coagulation on biomaterials in flowing
blood: some theoretical considerations. Biomaterials 1997; 18:1511–1522.
Bauersachs RM, Lindhoff-Last E, Ehrly AM, Betz C, Geiger H, Hauser IA. Treatment
of hirudin overdosage in a patient with chronic renal failure. Thromb Haemost
1999; 81:323–324.
524 Fischer
Beijering RJR, ten Cate H, Nurmohamed MT, ten Cate JW. Anticoagulants and
extracorporeal circuits. Semin Thromb Hemost 1997; 23:225–233.
Benz K, Nauck M, Beck KH, Böhler J, Fischer KG. Filtration characteristics of
recombinant hirudin by different hemodialyzers in vitro [abstr]. Ann Hematol
2000; 79(suppl 1):A26.
Berry CN, Girardot C, Lecoffre C, Lunven C. Effects of the synthetic thrombin
inhibitor argatroban on fibrin- or clot-incorporated thrombin: comparison with
heparin and recombinant hirudin. Thromb Haemost 1996; 72:381–386.
Berry CN, Lunven C, Girardot C, Lechaire I, Girard D, Charles MC, Ferrari P,
O’Brien DP. Ecarin clotting time: a predictive coagulation assay for the anti-
thrombotic activity of argatroban in the rat. Thromb Haemost 1998; 79:228–
233.
Bichler J, Gemmerli R, Fritz H. Studies for revealing a possible sensitization to hirudin
after repeated intravenous injections in baboons. Thromb Res 1991; 61:39–51.
Blaufox MD, Hampers CL, Merrill JP. Rebound anticoagulation occurring after
regional heparinization for hemodialysis. ASAIO Trans 1966; 12:207–209.
Boccardo P, Melacini D, Rota S, Mecca G, Boletta A, Casiraghi F, Gianese F. Indi-
vidualized anticoagulation with dermatan sulphate for haemodialysis in chronic
renal failure. Nephrol Dial Transplant 1997; 12:2349–2354.
Boon DMS, van Vliet HHDM, Zietse R, Kappers-Klunne MC. The presence of anti-
bodies against a PF4–heparin complex in patients on haemodialysis. Thromb
Haemost 1996; 76:480.
Bredlich RO, Stracke S, Gall H, Proebstle TM. Heparin-associated platelet aggrega-
tion syndrome with skin necrosis during haemodialysis. Dtsch Med Woch-
enschr 1997; 122:328–332.
Bucha E, Nowak G, Czerwinski R, Thieler H. r-Hirudin as anticoagulant in regular
hemodialysis therapy: Finding of therapeutic r-hirudin blood/plasma concen-
trations and respective dosages. Clin Appl Thromb Hemost 1999a; 5:164–170.
Bucha E, Kreml R, Nowak G. In vitro study of r-hirudin permeability through mem-
branes of different haemodialyzers. Nephrol Dial Transplant 1999b; 14:2922–
2926.
Chong BH, Magnani HN. Orgaran in heparin-induced thrombocytopenia. Haemo-
stasis 1992; 22:85–91.
Danhof M, De Boer A, Magnani HN, Stiekema JC. Pharmacokinetic considerations
of Orgaran (Org 10172) therapy. Haemostasis 1992; 22:73–84.
De Sancho M, Lema MG, Amiral J, Rand J. Frequency of antibodies directed against
heparin–platelet factor 4 in patients exposed to heparin through chronic hemo-
dialysis [letter]. Thromb Haemost 1996; 75:695–696.
Eichler P, Friesen HJ, Lubenow N, Jaeger B, Greinacher A. Antihirudin antibodies in
patients with heparin-induced thrombocytopenia treated with lepirudin: inci-
dence, effects on aPTT, and clinical relevance. Blood 2000; 96:2373–2378.
Finazzi G, Remuzzi G. Heparin-induced thrombocytopenia—background and impli-
cations for haemodialysis. Nephrol Dial Transplant 1996; 11:2120–2122.
Fischer KG. Hirudin in renal insufficiency. Semin Thromb Hemost 2002; 28:467–482.
Fischer KG, van de Loo A, Böhler J. Recombinant hirudin (lepirudin) as anticoagu-
lant in intensive care patients treated with continuous hemodialysis. Kidney Int
1999; 56(suppl 72):S46–S50.
Fischer KG, Weiner SM, Benz K, Nauck M, Böhler J. Treatment of hirudin overdose
with hemofiltration [abstr]. Blood Purif 2000; 18:80–81.
Fischer KG, Liebe V, Hudek R, Piazolo L, Haase KK, Borggrefe M, Huhle G. Anti-
hirudin antibodies alter pharmacokinetics and pharmacodynamics of recombi-
nant hirudin. Thromb Haemost 2003; 89:973–982.
Flanigan MJ, Pillsbury L, Sadewasser G, Lim VS. Regional hemodialysis anti-
coagulation: hypertonic tri-sodium citrate or anticoagulant citrate dextrose-A.
Am J Kidney Dis 1996; 27:519–524.
Frank RD, Farber H, Stefanidis I, Lanzmich R, Kierdorf HP. Hirudin elimination by
hemofiltration: a comparative in vitro study of different membranes. Kidney Int
1999; 56(suppl 72):S41–S45.
Frank RD, Farber H, Lazmich R, Floege J, Kierdorf HP. In vitro studies on hiru-
din elimination by haemofiltration: comparison of three high-flux membranes.
Nephrol Dial Transplant 2002; 17:1957–1963.
Greinacher A, Drost W, Michels I, Leitl J, Gottsmann M, Kohl HJ, Glaser M,
Mueller-Eckhardt C. Heparin-associated thrombocytopenia: successful therapy
with the heparinoid Org 10172 in a patient showing cross-reaction to LMW
heparins. Ann Hematol 1992a; 64:40–42.
nia: the antibody is not heparin specific. Thromb Haemost 1992b; 67:545–549.
Greinacher A, Philippen KH, Kemkes-Matthes B, Möckl M, Mueller-Eckhardt C,
Schaefer K. Heparin-associated thrombocytopenia type II in a patient with end-
stage renal disease: successful anticoagulation with the low-molecular-weight
heparinoid Org 10172 during haemodialysis. Nephrol Dial Transplant 1993;
8:1176–1177.
Greinacher A, Zinn S, Wizemann U, Birk W. Heparin-induced antibodies as a risk
factor for thromboembolism and haemorrhage in patients undergoing chronic
haemodialysis. Lancet 1996; 348:764.
Greinacher A, Eichler P, Lubenow N, Kwasny H, Luz M. Heparin-induced thrombo-
cytopenia with thromboembolic complications: meta-analysis of 2 prospective
trials to assess the value of parenteral treatment with lepirudin and its thera-
peutic aPTT range. Blood 2000; 96:846–851.
Griessbach U, Sturzebecher J, Markwardt F. Assay of hirudin in plasma using a
chromogenic thrombin substrate. Thromb Res 1985; 37:347–350.
Grötsch H, Hropot M. Degradation of rDNA hirudin and a-human thrombin hirudin
complex in liver and kidney homogenates from rat. Thromb Res 1991; 64:763–
767.
macother 1998; 32:55–59.
Haas G. Versuche der Blutauswaschung am Lebenden mit Hilfe der Dialyse. Klin
Wochenschr 1925; 4:13–14.
Hafner G, Peetz D, Klingel R, Prellwitz W. Methods for hirudin determination in
plasma. J Lab Med 2000; 24:172–178.
526 Fischer
Hafner G, Roser M, Nauck M. Methods for the monitoring of direct thrombin

inhibitors. Semin Thromb Hemost 2002; 28:425–430.
Clin Nephrol 1992; 38:86–89.
Henny CP, ten Cate H, ten Cate JW, Surachno S, Van Bronswijk H, Wilmink JM,
Ockelford PA. Use of a new heparinoid as anticoagulant during acute haemo-
dialysis of patients with bleeding complications. Lancet 1983; 1:890–893.
Hertel J, Keep DM, Caruana RJ. Anticoagulation. In: Daugirdas JT, Blake PG, Ing
TS, eds. Handbook of Dialysis. 3d ed. Boston: Little, Brown, 2001:182–198.
Horellou MH, Conard J, Lecrubier C, Samama M, Roque-D’Orbcastel O, de Fenoyl
O, Di Maria G, Bernadou A. Persistent heparin induced thrombocytopenia
despite therapy with low molecular weight heparin [letter]. Thromb Haemost
1984; 51:134.
Huhle G, Hoffmann U, Wang L, Bayerl C, Harenberg J. Allergy and positive IgG in a
patient reexposed to r-hirudin [abstr]. Ann Hematol 1998; 76(suppl 1):A97.
Huhle G, Hoffmann U, Song X, Wang LC, Heene DL, Harenberg J. Immunologic
response to recombinant hirudin in HIT type II patients during long-term treat-
ment. Br J Haematol 1999; 106:195–201.
Huhle G, Liebe V, Hudek R, Heene DL. Anti-r-hirudin antibodies reveal clinical
relevance through direct functional inactivation of r-hirudin or prolongation of
r-hirudin’s plasma halflife. Thromb Haemost 2001; 85:936–938.
Janson PA, Moake JL, Carpinito G. Aspirin prevents heparin-induced platelet aggre-
gation in vivo [letter]. Br J Haematol 1983; 53:166–168.
Janssen MJFM, Deegens JK, Kapinga TH, Beukhof JR, Huijgens PC, Van Loenen
AC, Van der Meulen J. Citrate compared to low molecular weight heparin
anticoagulation in chronic hemodialysis patients. Kidney Int 1996; 49:806–813.
Kadidal VV, Mayo DJ, Horne MK. Heparin-induced thrombocytopenia (HIT) due to
heparin flushes: a report of three cases. J Intern Med 1999; 246:325–329.
Kappa JR, Fisher CA, Berkowitz HD, Cottrel ED, Addonizio VP. Heparin-induced
platelet activation in sixteen surgical patients: diagnosis and management. J
Vasc Surg 1987; 5:101–109.
Kern H, Ziemer S, Kox WJ. Bleeding after intermittent or continuous r-hirudin during
CVVH. Intensive Care Med 1999; 25:1311–1314.
Koide M, Yamamoto S, Matsuo M, Suzuki S, Arima N, Matsuo T. Anticoagulation
for heparin-induced thrombocytopenia with spontaneous platelet aggregation
in a patient requiring haemodialysis. Nephrol Dial Transplant 1995; 10:2137–
2140.
Koster A, Merkle F, Hansen R, Loebe M, Kuppe H, Hetzer R, Crystal GJ, Mertzlufft
F. Elimination of recombinant hirudin by modified ultrafiltration during simu-
lated cardiopulmonary bypass: assessment of different filter systems. Anesth
Analg 2000; 91:265–269.
Laster J, Elfrink R, Silver D. Reexposure to heparin of patients with heparin-asso-
ciated antibodies. J Vasc Surg 1989; 9:677–682.
Leblanc M, Roy LF, Legault L, Dufresne LR, Morin C, Thuot C. Severe skin necrosis
associated with heparin in hemodialysis. Nephron 1994; 68:133–137.
Lewis BE, Walenga JM, Wallis DE. Anticoagulation with Novastan (argatroban) in
patients with heparin-induced thrombocytopenia and heparin-induced throm-
bocytopenia and thrombosis syndrome. Semin Thromb Hemost 1997; 23:197–
202.
Liebe V, Piazolo L, Fischer KG, Hudek R, Heene DL, Huhle G. A monoclonal mouse
anti-r-hirudin antibody neutralizes r-hirudin in vivo—potential use as antidote
[abstr]. Ann Hematol 2001; 80(suppl 1):A42.
Lindhoff-Last E, Betz C, Bauersachs R. Use of a low-molecular-weight heparinoid
(danaparoid sodium) for continuous renal replacement therapy in intensive care
patients. Clin Appl Thromb Hemost 2001; 7:300–304.
Lubenow N, Greinacher A. Heparin-induced thrombocytopenia. Recommendations
for optimal use of recombinant hirudin. Biodrugs 2000; 14:109–125.
Lunven C, Gauffeny C, Lecoffre C, O’Brien DP, Roome NO, Berry CN. Inhibition by
argatroban, a specific thrombin inhibitor, of platelet activation by fibrin clot-
associated thrombin. Thromb Haemost 1996; 75:154–160.
Luzzatto G, Bertoli M, Cella G, Fabris F, Zaia B, Girolami A. Platelet count, anti-
heparin/platelet factor 4 antibodies and tissue factor pathway inhibitor plasma
antigen level in chronic dialysis. Thromb Res 1998; 89:115–122.
Markwardt F. Past, present and future of hirudin. Haemostasis 1991; 21:11–26.
Matsuo T, Nakao K, Yamada T, Matsuo O. Effect of a new anticoagulant (MD 805)
on platelet activation in the hemodialysis circuit. Thromb Res 1986; 41:33–41.
Matsuo T, Yamada T, Chikahira Y, Kadowaki S. Effect of aspirin on heparin-induced
thrombocytopenia (HIT) in a patient requiring hemodialysis. Blut 1989; 59:393–
395.
Matsuo T, Kario K, Kodama K, Okamoto S. Clinical application of the synthetic
thrombin inhibitor, argatroban (MD-805). Semin Thromb Hemost 1992; 18:
155–160.
Matsuo T, Kario K, Nakao K, Yamada T, Matsuo M. Anticoagulation with nafa-
mostat mesilate, a synthetic protease inhibitor, in hemodialysis patients with a
bleeding risk. Haemostasis 1993; 23:135–141.
Matsuo T, Koide M, Kario K. Application of argatroban, a direct thrombin inhibitor,
in heparin-intolerant patients requiring extracorporeal circulation. Artif Organs
1997; 21:1035–1038.
Moberg PQ, Geary VM, Sheikh FM. Heparin-induced thrombocytopenia: a possible
complication of heparin-coated pulmonary artery catheters. J Cardiothorac
Anesth 1990; 4:226–228.
Müller A, Huhle G, Nowack R, Birck R, Heene DL, van der Woude FJ. Serious
bleeding in a haemodialysis patient treated with recombinant hirudin. Nephrol
Dial Transplant 1999; 14:2482–2483.
Murray P, Reddy B, Grossman E, Hammes M, Trevino S, Ferrell J, Tang I, Khorana
A, Fosbinder T, Swan SK. Safety and tolerability of argatroban anticoagulation
in patients with end-stage renal disease undergoing hemodialysis: a prospective
study [abstr]. J Thromb Haemost 2003; 1(suppl 1):P1911.
528 Fischer
Neuhaus TJ, Götschel P, Schmugge M, Leumann E. Heparin-induced thrombocy-

topenia type II on hemodialysis: switch to danaparoid. Pediatr Nephrol 2000;
14:713–716.
Newman PM, Swanson RL, Chong BH. IgG binding to PF4–heparin complexes in the
fluid phase and cross-reactivity with low molecular weight heparin and
heparinoid. Thromb Haemost 1998; 80:292–297.
Novacek G, Kapiotis S, Jilma B, Quehenberger P, Michitsch A, Traindl O, Speiser W.
Enhanced blood coagulation and enhanced fibrinolysis during hemodialysis
with prostacyclin. Thromb Res 1997; 88:283–290.
Nowak G, Bucha E. Quantitative determination of hirudin in blood and body fluids.
Semin Thromb Hemost 1996; 22:197–202.
Nowak G, Bucha E, Gööck T, Thieler H, Markwardt F. Pharmacology of r-hirudin in
renal impairment. Thromb Res 1992; 66:707–715.
Nowak G, Bucha E, Brauns I, Czerwinski R. Anticoagulation with r-hirudin in regular
haemodialysis with heparin-induced thrombocytopenia (HIT II). The first long
term application of r-hirudin in a haemodialysis patient. Wien Klin Wochenschr
1997; 109:354–358.
Nurmohamed MT, Berckmans RJ, Morriën-Salomons WM, Berends F, Hommes
DW, Rijnierse JJMM, Sturk A. Monitoring anticoagulant therapy by activated
partial thromboplastin time: hirudin assessment. Thomb Haemost 1994; 72:
685–692.
Ortel TL, Gockermann JP, Califf RM, McCann RL, O’Connor CM, Metzler DM,
Greenberg CS. Parenteral anticoagulation with the heparinoid Lomoparan
(Org 10172) in patients with heparin-induced thrombocytopenia and thrombo-
sis. Thromb Haemost 1992; 67:292–296.
O’Shea SI, Sands JJ, Nudo SA, Ortel TL. Frequency of anti-heparin-platelet factor 4
antibodies in hemodialysis patients and correlation with recurrent vascular
access thrombosis. Am J Hematol 2002; 69:72–73.
O’Shea SI, Ortel TL, Kovalik EC. Alternative methods of anticoagulation for dialysis-
dependent patients with heparin-induced thrombocytopenia. Semin Dial 2003;
16:61–67.
Ota K, Akizawa T, Hirasawa Y, Agishi T, Matsui N. Effects of argatroban as an
anticoagulant for haemodialysis in patients with antithrombin III deficiency.
Pinnick RV, Wiegmann TB, Diederich DA. Regional citrate anticoagulation for
hemodialysis in the patient at high risk for bleeding. N Engl J Med 1983; 308:
258–261.
554.
Pöschel KA, Bucha E, Esslinger HU, Nörtersheuser P, Jansa U, Schindler S, Nowak
G, Stein G. Pharmacodynamics and pharmacokinetics of polyethylene glycol–
hirudin in patients with chronic renal failure. Kidney Int 2000; 58:2478–2484.
Roe SD, Cassidy MJD, Haynes AP, Byrne JL. Heparin-induced thrombocytopenia
(HIT) and thrombosis in a haemodialysis-dependent patient with systemic vas-

culitis. Nephrol Dial Transplant 1998; 13:3226–3229.
Romao JE Jr, Fadil MA, Sabbaga E, Marcondes M. Haemodialysis without anti-
coagulant: haemostasis parameters, fibrinogen kinetic, and dialysis efficiency.
Roussi JH, Houbouyan LL, Goguel AF. Use of low-molecular-weight heparin in
heparin-induced thrombocytopenia with thrombotic complications [letter].
Lancet 1984; 1:1183.
Rylance PB, Gordge MP, Ireland H, Lane DA, Weston MJ. Haemodialysis with
prostacyclin (epoprostenol) alone. Proc Eur Dial Transplant Assoc Eur Renal
Assoc 1985; 21:281–286.
Samuelsson O, Amiral J, Attman P-O, Bennegård K, Björck S, Larsson G, Tengborn
L. Heparin-induced thrombocytopenia during continuous haemofiltration.
Saner F, Hertl M, Broelsch CE. Anticoagulation with hirudin for continuous veno-
venous hemodialysis in liver transplantation. Acta Anaesthesiol Scand 2001;
45:914–918.
Schmitt GW, Moake JL, Rudy CK, Vicks SL, Hamburger RJ. Alterations in hemo-
static parameters during hemodialysis with dialyzers of different membrane
composition and flow design. Platelet activation and factor VIII-related von
Willebrand factor during hemodialysis. Am J Med 1987; 83:411–418.
Schneider T, Heuer B, Deller A, Boesken WH. Continuous haemofiltration with r-
hirudin (lepirudin) as anticoagulant in a patient with heparin induced
thrombocytopenia (HIT II). Wien Klin Wochenschr 2000; 112:552–555.
Sitter T, Spannagl M, Banas B, Schiffl H. Prevalence of heparin-induced PF4-heparin
antibodies in hemodialysis patients. Nephron 1998; 79:245–246.
Smith MC, Danviriyasup K, Crow JW, Cato AE, Park GD, Hassid A, Dunn MJ.
Prostacyclin substitution for heparin in long-term hemodialysis. Am J Med
1982; 73:669–678.
Song X, Huhle G, Wang L, Hoffmann U, Harenberg J. Generation of anti-hirudin
antibodies in heparin-induced thrombocytopenic patients treated with r-hiru-
din. Circulation 1999; 100:1528–1532.
Sreedhara R, Itagaki I, Lynn B, Hakim RM. Defective platelet aggregation in uremia
is transiently worsened by hemodialysis. Am J Kidney Dis 1995; 25:555–563.
Steuer S, Boogen C, Plum J, Deppe C, Reinauer H, Grabensee B. Anticoagulation with
r-hirudin in a patient with acute renal failure and heparin-induced thrombo-
cytopenia. Nephrol Dial Transplant 1999; 14(suppl 4):45–47.
Takahashi H, Muto S, Nakazawa E, Yanagiba S, Masunaga Y, Miyata Y, Tamba K,
Kusano E, Matsuo M, Matsuo T, Asano Y. Combined treatment with nafamo-
stat mesilate and aspirin prevents heparin-induced thrombocytopenia in a
hemodialysis patient. Clin Nephrol 2003; 59:458–462.
Tholl U, Greinacher A, Overdick K, Anlauf M. Life-threatening anaphylactic reaction
following parathyroidectomy in a dialysis patient with heparin-induced throm-
bocytopenia. Nephrol Dial Transplant 1997; 12:2750–2755.
Turney JH, Williams LC, Fewell MR, Parsons V, Weston MJ. Platelet protection and
530 Fischer
heparin sparing with prostacyclin during regular dialysis therapy. Lancet 1980;
2:219–222.
Unver B, Sunder-Plassmann G, Hörl WH, Apsner R. Long-term citrate anticoagu-
lation for high-flux haemodialysis in a patient with heparin-induced thrombo-
cytopenia type II. Acta Med Austriaca 2002; 29:146–148.
Vanholder RC, Camez AA, Veys NM, Soria J, Mirshahi M, Soria C, Ringoir S.
Recombinant hirudin: a specific thrombin inhibiting anticoagulant for hemo-
dialysis. Kidney Int 1994; 45:1754–1759.
Vanholder R, Camez A, Veys N, van Loo A, Dhondt AM, Ringoir S. Phar-
macokinetics of recombinant hirudin in hemodialyzed end-stage renal failure
patients. Thromb Haemost 1997; 77:650–655.
Van Wyk V, Badenhorst PN, Luus HG, Kotzé HF. A comparison between the use of
recombinant hirudin and heparin during hemodialysis. Kidney Int 1995;
48:1338–1343.
Vargas Hein O, von Heymann C, Lipps M, Ziemer S, Ronco C, Neumayer HH,
Morgera S, Welte M, Kox WJ, Spies C. Hirudin versus heparin for antico-
agulation in continuous renal replacement therapy. Intensive Care Med 2001;
27:673–679.
Vecino A, Navarro-Antolin J, Teruél J, Navarro J, Cesar J. Lipid composition of
platelets in patients with uremia. Nephron 1998; 78:271–273.
Vun CM, Evans S, Chong BH. Cross-reactivity study of low molecular weight hepa-
rins and heparinoid in heparin-induced thrombocytopenia. Thromb Res 1996;
81:525–532.
Ward DM, Mehta RL. Extracorporeal management of acute renal failure patients at
high risk of bleeding. Kidney Int 1993; 43:S-237–S-244.
reactivity (XR) of danaparoid for HIT-IgG [abstr]. Blood 1996; 88:626a.
Am J Med 1996; 101:502–507.
Wilde MI, Markham A. Danaparoid. A review of its pharmacology and clinical use in
the management of heparin-induced thrombocytopenia. Drugs 1997; 54:903–
924.
Willey ML, de Denus S, Spinler SA. Removal of lepirudin, a recombinant hirudin, by
hemodialysis, hemofiltration, or plasmapheresis. Pharmacotherapy 2002; 22:
492–499.
Yamamoto S, Koide M, Matsuo M, Suzuki S, Ohtaka M, Saika S, Matsuo T.
Heparin-induced thrombocytopenia in hemodialysis patients. Am J Kidney Dis
1996; 28:82–85.
19
Management of Cardiopulmonary
Bypass Anticoagulation in Patients
with Heparin-Induced
Thrombocytopenia
Bernd Poetzsch
Rheinische Friedrich-Wilhelms-University Bonn, Bonn, Germany
Katharina Madlener
Kerckhoff-Klinik, Bad Nauheim, Germany
I. INTRODUCTION
Immediate cessation of, and avoidance of reexposure to, heparin are impor-
tant principles underlying the management of patients with immune-mediated
heparin-induced thrombocytopenia (HIT) (Chong and Berndt, 1989; War-
kentin et al., 1998). Because further antithrombotic therapy is often necessary
for these patients, several alternative anticoagulant strategies have been
developed (see Chaps. 13–17). However, patients with HIT who require car-
diac surgery present special problems. Considerable activation of the hemo-
static system results when blood is exposed to the artificial surfaces of the
cardiopulmonary bypass (CPB) pump used for most heart surgery, making
high-dose anticoagulation mandatory (Edmunds, 1993; Slaughter et al.,
1994). Heparin is the current anticoagulant of choice for CPB, and there is
relatively little experience with other forms of anticoagulation in this patient
setting. Moreover, any alternative anticoagulant considered for HIT patients
should ideally meet certain requirements. First, the agent should be effective
in minimizing activation of coagulation during CPB. Second, a rapid and
simple method of monitoring its anticoagulating effects should be available to
531
532 Poetzsch and Madlener
avoid inappropriate under- or overanticoagulation. Finally, rapid and com-

plete reversibility of the anticoagulating effects is important to minimize
postoperative bleeding complications. Unfortunately, no existing agent meets
all of these requirements.
II. ALTERNATIVE STRATEGIES FOR CPB

ANTICOAGULATION
A variety of approaches to perform CPB anticoagulation in HIT patients has

been reported, including the use of danaparoid sodium, the thrombin
inhibitors lepirudin, bivalirudin, and argatroban, the defibrinogenating en-
zyme ancrod, and antiplatelet agents, such as aspirin, tirofiban, and iloprost.
These antiplatelet agents were used to prevent platelet activation when heparin
was used in a patient with acute HIT or a history of previous HIT. Experience
with a planned reexposure to heparin to permit CPB in patients with a previ-
ous history of HIT, but who no longer have detectable HIT antibodies at
the time of the subsequent heparin reexposure, will also be discussed.
A. Danaparoid Sodium
The efficacy of danaparoid sodium for CPB anticoagulation was first shown in
a dog model (Henny et al., 1985). Subsequently, this agent was used for pa-
tients with HIT who needed heart surgery (Doherty et al., 1990; Magnani,
1993; Wilhelm et al., 1996). In a retrospective analysis Magnani and co-
workers (1997) summarized the experience in 53 patients with HIT who
underwent CPB using danaparoid for anticoagulation. The patients included
in this study generally received an intravenous (iv) bolus of 8750 antifactor Xa
(anti-Xa) units (U) of danaparoid after thoracotomy. The CPB circuit was
primed with 7500 U. During CPB, booster iv injections (1500 U) were to be
administered up to once hourly if there was visually apparent clot or fibrin
formation. Plasma levels of anti-Xa activity generally were not monitored
during CPB.
With this fixed-dosing schedule, ‘‘clots’’ in the operative field, as an
indicator of inadequate anticoagulation, were observed in 18 (34%) patients.
One patient, reported elsewhere, developed near-fatal thrombosis of the CPB
circuit during weaning from bypass (Grocott et al., 1997). Severe postoper-
ative bleeding, defined as more than 20 U of blood transfused, was noted
in 11 (20%) patients. As a result of these data, the authors recommended
a modified treatment regimen that included a priming dose of 3 U/mL, a
weight-adjusted postthoracotomy iv bolus dose of 125 U/kg body weight
(b.w.), and a constant iv infusion of 7 U/kg b.w. per hour started immediately
CPB Anticoagulation in HIT 533
after institution of the CPB, and stopped 45 min before the expected end of
CPB. Thus, for a 70-kg person undergoing an operation with a CPB time of
2 h and a priming volume of 1500 mL, a total dose of approximately 13,860 U
danaparoid, or 198 U/kg, is recommended by the authors.
However, this revised protocol was developed empirically, with adjust-
ments made based on some of the complications observed using the fixed-dose
protocol (Grocott et al., 1997). Even though the revised protocol means that
many patients would receive a lower dose of danaparoid than with the earlier
fixed-dose regimen, this might not lead to reduced bleeding outcomes.
Paradoxically, less effective anticoagulation during CPB could lead to more
thrombin generation during the procedure potentially leading to even greater
postoperative bleeding because of secondary hyperfibrinolysis, even if the
postoperative danaparoid levels are not high. Indeed, Insler and colleagues
(1997) reported a patient receiving danaparoid for CPB who first developed
clots in the operative field and arterial filter of the CPB, followed by severe
postoperative bleeding requiring surgical reexploration. Regardless of the
explanation for excessive bleeding, even a weight-modified treatment regimen
bears the risk of under- or overanticoagulation, if the anticoagulant effect is
not monitored.
Therefore, we developed a danaparoid-dosing schedule for CPB with
dose adjustments made according to the results of anti-Xa measurements
(Table 1). Unfortunately, although both the activated clotting time (ACT)
and the activated partial thromboplastin time (aPTT) are prolonged by the
higher plasma levels of danaparoid used during CPB, there is no acceptable
linear correlation (Gitlin et al., 1998). Because only the plasma anti-Xa levels
correlate linearly with the plasma levels of danaparoid, we based our schedule
on anti-Xa levels. Similar to the protocol recommended by Magnani and
coworkers (1997), the iv bolus and priming dose are adjusted to body weight
and priming volume, respectively. After beginning CPB, plasma anti-Xa
levels should be maintained at 1.5 F 0.3 U/mL. The continuous infusion is
stopped 30 min before the expected end of bypass. However, in our experi-
ence, even if such an anti-Xa–adjusted danaparoid treatment regimen is used,
increased postoperative bleeding is a problem. Possible explanations for the
increased postoperative bleeding include an ongoing anticoagulant effect of
danaparoid (half-life, approximately 17 h) for which there is no pharmaco-
logical antagonist (Meuleman, 1992), as well as the incomplete inhibition of
thrombin generation during CPB, potentially leading to increased postoper-
ative hyperfibrinolysis.
In all, CPB anticoagulation with danaparoid can lead to successful
outcomes. About three quarters (36 of 47; 77%) of the patients reported by
Magnani and coworkers (1997) were alive 6 weeks after cardiac surgery
with danaparoid. Nevertheless, the disadvantages of danaparoid, including
Table 1 Modified Treatment Protocol for CPB Anticoagulation with

Danaparoid Sodiuma
Initial danaparoid dosing (pre-CPB)

Initial intravenous (iv) danaparoid bolusb: 100 U/kg body weight
Danaparoid added to priming solution: 3000 U
Initial target antifactor Xa level: >1.5 U/mL before start of CPB
Additional danaparoid dosing if pre-CPB plasma anti-Xa level <1.5 U/mL
Anti-Xa level Dosing modification
<1.2 U/mL Give extra 1500 U bolus
1.2–1.5 U/mL Give additional 750 U bolus
Dosing and monitoring while on CPB

Danaparoid infusion rate at start of CPB: 200 U/h
Frequency of anti-Xa level monitoring: Every 15 min
Intraoperative dose adjustments, based on plasma anti-Xa levels:
Anti-Xa level Dosing modification
>1.8 U/mL Stop infusion until anti-Xa level<1.5 U/mL
1.2–1.8 U/mL No change in infusion rate
<1.2 U/mL Increase infusion rate to 300 U/h
<1.0 U/mL Administer additional iv bolus of 3000 U
Special steps toward end of CPB

Stop danaparoid infusion 30 min before anticipated end of CPB
a
This protocol, developed by the authors, is based on the availability of rapid-turnaround
plasma anti-Xa levels obtained intraoperatively. Another protocol that is not based on intra-
operative anti-Xa monitoring, and that results in somewhat higher danaparoid dosing, is given
in Chap. 14.
b
The initial intravenous danaparoid bolus should be given 15–20 min before start of CPB
(generally at the time the surgeon has opened the sternum).
its long-lasting anticoagulant activity that cannot be neutralized, and the

significant difficulties in monitoring its anticoagulant effects in an operating
room setting, render danaparoid a ‘‘second-choice’’ anticoagulant in HIT
patients requiring cardiac surgery.
B. Recombinant Hirudin
Recombinant hirudin (r-hirudin), an anticoagulant naturally produced by the
salivary gland of the leech (Hirudo medicinalis), is now approved in most
countries for clinical use by the iv route. Hirudin is a single-chain polypeptide
of 65 amino acids (7000 Da) that forms a tight 1:1 stoichiometric complex with
thrombin, thereby occupying the putative fibrinogen-binding site and block-
ing the catalytic site of thrombin. As a result, all of the thrombin-catalyzed
procoagulant reactions, such as conversion of fibrinogen to fibrin, activation

of coagulation factors V, VIII, and XIII, and thrombin-induced platelet
activation, are inhibited. Although two hirudins are approved (lepirudin,
desirudin), data on the use in cardiac surgery are only available for lepirudin.
Because of its potent anticoagulant effect, r-hirudin has been studied as
an anticoagulant for use in open heart surgery in both dogs (Walenga et al.,
1991) and pigs (Riess et al., 1997). In both animal models, effective CPB
anticoagulation could be achieved by administration of r-hirudin as a bolus
injection (1 mg/kg b.w.) followed by a continuous infusion of 1 mg/kg b.w.,
started after initiation of CPB, and continuing until end of CPB. In humans,
however, recovery of hirudin in the plasma following body weight–adjusted
dosing shows a high interindividual variability (Koza et al., 1993). Therefore,
a fixed-dose protocol for r-hirudin in the CPB setting bears the risk of both
inadequate anticoagulation and overdosing. Although the latter is compli-
cated by excessive and potentially fatal postoperative bleeds, the former may
result in the occurrence of thromboembolic complications while on pump,
including catastrophic total pump occlusion.
To establish a treatment schedule that is adjusted to the individual’s
response to hirudin, we investigated different monitoring systems for hirudin
plasma levels. Several in vitro and in vivo experiments demonstrated that the
ACT and aPTT were not sufficiently sensitive to monitor hirudin plasma
levels (Pötzsch et al., 1997). However, reliable results were obtained by using
the whole blood ecarin clotting time (ECT) (Pötzsch et al., 1997).
Ecarin is a prothrombin-activating enzyme, derived from the venom of
the snake Echis carinatus, that activates prothrombin to an intermediate
product, meizothrombin (Nishida et al., 1995). Meizothrombin expresses
only moderate clotting activity, but is fully reactive toward, and thus inhibited
by, hirudin. As a result, in r-hirudin–containing plasma, meizothrombin forms
stable 1:1 complexes with r-hirudin. Only when hirudin is neutralized does
clotting become initiated, either by meizothrombin or subsequently generated
thrombin. Ecarin is available from commercial sources.
Table 2 outlines the whole blood ECT method, which we perform using
the KC10a coagulometer (Pötzsch et al., 1997). The method is easily adapt-
Table 2 Whole Blood Ecarin Clotting Time
50 AL citrate-anticoagulated whole blood to be analyzed

+ 50 AL standard normal human plasma
Incubate for 1 min at 37jC
+ 50 AL ecarin solution (20 U/mL) containing 0.025 M calcium chloride
Determination of the clotting time
able to any other coagulometer. A calibration curve is constructed by using

citrate-anticoagulated whole blood spiked with r-hirudin to achieve final
concentrations of 0.5, 1.0, 1.5, 2.0, 3.0, and 4.0 Ag/mL. A reliable ECT
required adequate prothrombin levels, which can be reduced in severely ill
patients and/or by hemodilution after beginning CPB. This problem can be
overcome by mixing patient blood with normal human plasma (1:1).
In the United States and Canada, there is the additional option to use a
commercial ECT method available from PharmaNetics (Morrisville, NC) by
way of a ‘‘humanitarian device exemption’’ (H990012) for the specific situation
of CPB when heparin is contraindicated because of HIT (Koster et al., 2000a).
The assay is performed using a point-of-care methodology (Thrombolytic
Assessment System [TAS]; manufactured by PharmaNetics and marketed as
Rapidpoint Coag by Bayer Diagnostics, Toronto, ON, and Tarrytown, NY).
Practical issues include the time required for obtaining the indemnification
agreement and institutional review board approval (U.S.) or patient-specific
regulatory approval (Canada) (Warkentin and Greinacher, 2003).
Critical levels of r-hirudin during the CPB operation were established in
an in vitro CPB setting and in a first series of HIT patients undergoing cardiac
surgery (Pötzsch et al., 1993; Riess et al., 1995, 1996). Clot formation in the
CPB apparatus was seen at levels of r-hirudin below 1.8 Ag/mL, and in-
creasing levels of fibrinopeptide A (an indicator of thrombin-mediated
fibrinogen cleavage) occurred at r-hirudin plasma levels less than 2.0 Ag/
mL. Based on these results, the therapeutic level of r-hirudin during CPB was
set between 3.5 and 4.5 Ag/mL. Higher intraoperative levels of r-hirudin could
be complicated by a higher postoperative bleeding risk, especially because no
antidote is available.
A treatment protocol based upon the ECT-monitoring of hirudin levels
is given in Table 3. The data obtained from ten patients with HIT, treated with
r-hirudin for heart surgery, demonstrated that stable r-hirudin plasma levels
in the range from 3.5 to 5.0 Ag/mL could be obtained using the ECT-adjusted
treatment schedule (Fig. 1a). Because of the relatively short half-life of r-
hirudin of approximately 1 h, plasma levels of r-hirudin declined rapidly after
stopping its infusion (Fig. 1b). However, in renally impaired patients, r-hiru-
din can accumulate, leading to postoperative bleeding (Koster et al., 2000b).
To date, the clinical data demonstrate that r-hirudin is a suitable
alternative for anticoagulation of CPB in HIT patients. The ECT provides
adequate monitoring and allows an adjusted treatment schedule with appar-
ently minimal risk for thrombotic problems on pump. Because of the
relatively short half-life, plasma levels of r-hirudin decline rapidly after stop-
ping its infusion. As hirudin is almost completely eliminated by the kidney in
humans, patients with impaired renal function may require hemofiltration to
reduce plasma levels of r-hirudin.
Table 3 Treatment Protocol for r-Hirudin (Lepirudin) Anticoagulation

During CPB
Initial lepirudin dosing (pre-CPB)

Initial iv lepirudin bolus: 0.25 mg/kg body weight
Initiate continuous iv infusiona: 30 mL/h (0.5 mg/min.)
Lepirudin added to priming solution: 0.2 mg/kg body weight
Target lepirudin plasma levels:b >2.5 Ag/mL before start of CPB
If<2.5 Ag/mL, give additional
bolus (10 mg)
Lepirudin dosing and monitoring while

on CPB
Frequency of lepirudin level monitoring: every 15 min using ECT
Intraoperative dose adjustments,
based on ECT:
Lepirudin plasma level Dosing modification
>4.5 Ag/mL Reduce infusion rate by 10 mL/h
3.5–4.5 Ag/mL No change in infusion rate
<3.5 Ag/mL Increase infusion rate by 10 mL/h
Special steps toward end of CPB

Stop lepirudin infusion 15 min before anticipated end of CPB.
After disconnection of CPB, administer 5 mg hirudin to the heart–lung machine
to avoid clot formation.
CPB, cardiopulmonary; iv, intravenous.
a
50 mg of lepirudin are dissolved in 50 mL 0.9% sodium chloride
b
The target lepirudin level pre-CPB (>2.5 Ag/mL) is lower than the ones sought during CPB
(3.5–4.5 Ag/mL) because of the addition of lepirudin to the pump circuit volume (0.2 mg/kg
body weight).
C. Bivalirudin and Argatroban

Bivalirudin
Bivalirudin (Angiomax; see Chap. 17) is a 20-amino-acid synthetic peptide
modeled after hirudin. It consists of two peptide fragments connected by
a tetraglycine spacer that recognize thrombin’s fibrinogen binding site
(exosite I) and its catalytic site. Unlike lepirudin, this bivalent interaction
with thrombin is reversible once plasma enzymes (including thrombin itself)
cleave the arg3-pro4 bond on bivalirudin. Its short half-life (25 min) and
predominant enzymic elimination might be advantageous for use in CPB. As
with lepirudin, the ECT is recommended for intraoperative monitoring
during CPB (Koster et al., 2003), although anecdotal success (and some
failure) using ACT monitoring exists. For off-pump cardiac surgery (which
Figure 1 Course of free r-hirudin concentrations in HIT patients (n = 10) treated

with hirudin before, (a) during, and (b) after CPB. Free r-hirudin was measured
using a chromogenic thrombin-based assay as described. (From Pötzsch et al., 1997.)
requires only one-third to one-half the usual concentrations of anticoagulant

compared with CPB), the ACT can be used (Merry et al., 2004).
Bivalirudin has been used successfully off-label for anticoagulation
during off-pump and on-pump cardiac surgery in patients with acute or
previous HIT (Spiess et al., 2002; Vasquez et al., 2002; Davis et al., 2003).
In addition, bivalirudin compared favorably in a randomized trial against
heparin in non-HIT patients undergoing off-pump surgery (Merry et al.,
2003). Bivalirudin was therefore evaluated in a 20-patient pilot study (on-
pump, non-HIT), and is currently under investigation in a phase III multi-
center pivotal trial (in comparison with UFH) for on- and off-pump cardiac
surgery in patients with and without HIT.
An investigational, ECT-adjusted treatment protocol is outlined in
Table 4 (Warkentin and Greinacher, 2003). Bivalirudin’s enzymic metabolism
presents certain practical issues and limitations. Surgical techniques that
allow blood to lie stagnant should be avoided, since local bivalirudin levels
will decrease due to its metabolism by proteases, produced in blood exposed
to wound or foreign surfaces, leading to local clot formation. The presence of
visible thrombus in an area of stagnation, such as in the pericardial cavity,
should not be interpreted by the surgeon as indicative of the need for
additional anticoagulation, as this may only reflect local bivalirudin metab-
olism and not correlate with plasma levels. If blood cardioplegia is used, the
blood should be directly sourced from the circuit, and (after mixing with the
cardioplegia solution) immediately reinfused into the coronary system. For
the same reason, assessment of graft blood flow and testing for leakage should
preferably be performed with albumin and saline solutions or, alternatively,
using blood taken directly from the patient and used immediately for this
assessment. Because hypothermia somewhat reduces the proteolysis of
bivalirudin, the patient’s core temperature should be maintained close to
37jC after coming off CPB (or after completing the final anastomoses in off-
pump procedures) and care taken to maintain body temperature during the
early postoperative period.
Following separation from CPB, the risk that the circuit may clot
rapidly may be even higher than with lepirudin, due to bivalirudin’s shorter
half-life and ongoing metabolism. Thus, provision to continue to recirculate
pump blood following separation from bypass is made by adding a cross-limb
in the bypass circuit at the time of setup, which remains clamped until coming
off bypass. Following clamping of the venous line, this limb is opened and the
contents recirculated. Within 10 min of separation from bypass, in case the
patient might need to return to bypass support, a 50 mg bolus of bivalirudin
should be added to the circuit to prevent clotting, and a 50 mg/h bivalirudin
infusion into the bypass circuit should also be started and continued until
such time as it is clear the patient will not require urgent return to CPB. Once
Table 4 Treatment Protocol for Bivalirudin Anticoagulation During CPB (Under

Investigationa)
Initial bivalirudin dosing (pre-CPB)

Initial iv bivalirudin bolus: 1.5 mg/kg body weight
and initiate continuous iv infusion: 2.5 mg/kg/h (42 Ag/kg/min)
Bivalirudin added to pump circuit volume: 50 mg
Target bivalirudin plasma level: >10 Ag/mL before start of CPB
if<10 Ag/mL, give additional bolus
(0.25 mg/kg) and increase infusion
rate by 0.25 mg/kg/h
Bivalirudin dosing and monitoring while on CPB

Continue iv infusion (adjusted as below): 2.5 mg/kg/h or greater (as above)
Frequency of bivalirudin level monitoring: Every 30 min using ECT
Intraoperative dose adjustments, based on ECT:
Bivalirudin plasma level (ECT)b Dosing modification
>15 Ag/mL (>500 s) Reduce infusion rate by 0.25 mg/kg/h
10–15 Ag/mL (400–500 s) No change in infusion rate
<10 Ag/mL (<400 s) Give additional bolus
(0.25 mg/kg) and increase
infusion rate by 0.25 mg/kg/h
Special steps at end of CPB

Stop bivalirudin infusion at end of CPB, then either:
(A) Within 10 min of stopping bivalirudin infusion: first reinfuse appropriate portion of
pump volume to patient, and then give 50 mg bivalirudin bolus to the circuit to
prevent clotting; start an infusion of 50 mg/h into the circuit only, and continue to
recirculate; any subsequent reinfusion of remaining pump volume to patient should
be processed through a cell saver (which removes >90% of bivalirudin); or
(B) Promptly empty remaining pump volume into cell-saving device (replacing the pump
contents with crystalloid), thus avoiding need for postseparation bivalirudin boluses
to circuit; process blood for reinfusion with cell saver to remove bivalirudin.
CPB, cardiopulmonary bypass; ECT, ecarin clotting time; iv, intravenous.
a
Up-to-date information on protocol amendments are available from the manufacturer of bivalirudin (‘‘1-
800-ANGIOMAX’’; The Medicines Company, Parsippany, NJ).
b
The target bivalirudin concentration (10–15 Ag/mL) corresponds to an ecarin clotting time (ECT) of 400–
500 s using the RapidPoint Coag (Bayer); with other ECT methods, the bivalirudin concentration should
be determined using a calibration curve.
Source: Warkentin and Greinacher, 2003.
postseparation bivalirudin dosing to the circuit has commenced, any remain-

ing pump volume contents intended for reinfusion to the patient should first
be processed using a cell-saving device (‘‘cell saver’’), thus washing away most
of the bivalirudin. Another approach is simply to drain rapidly the contents
of the pump into a cell saver following separation from bypass and replace
the pump contents with crystalloid, and wash the blood, thus avoiding the
possibility of pump clotting without the need to administer additional bi-
valirudin into the pump.
Argatroban
Argatroban is a specific thrombin inhibitor derived from L-arginine. It is a
small molecule (532 Da) that binds reversibly to thrombin. It has a half-life of
about 40–50 min in normal humans. The potential of argatroban to be an
effective anticoagulant in patients with HIT has been documented by the
studies of Lewis and coworkers (1997a,b, 2001, 2003) (see Chap. 16).
Although argatroban has been a successful anticoagulant in a CPB model,
only limited information is available on its use in humans for this purpose.
Furukawa and coworkers report a patient who successfully underwent CPB
anticoagulation with argatroban (Furukawa et al., 2001). Argatroban was
administered as a bolus injection of 0.1 mg/kg followed by a continuous
infusion at 5–10 Ag/kg/min. The ACT was used for monitoring.
D. Platelet Inhibition as a Strategy to Permit Heparinization

for CPB
Another approach described to manage CPB in a patient with HIT is to com-
bine full heparinization with one or more antiplatelet agents. Several groups of
investigators have used iloprost for this situation (Kappa et al., 1985; Long,
1985; Palmer Smith et al., 1985; Addonizio et al., 1987; Kraenzler and Starr,
1988), following the original observation by Olinger and colleagues (1984) that
iloprost inhibited heparin-dependent platelet activation in the presence of HIT
serum. Iloprost is a stable analogue of prostacyclin; thus, it stimulates
adenylate cyclase, resulting in increased platelet cAMP levels, which prevents
platelet activation by various platelet agonists, including HIT antibodies.
Recently this approach has experienced a resurgence with epoprostenol
sodium (Flolan), a freeze-dried preparation of prostacyclin itself (Mertzlufft
et al., 2000; Aouifi et al., 2001). Epoprostenol is approved for use in patients
with primary pulmonary hypertension. Its very short half-life (6 min) means
that continuous iv infusion is necessary. Complete inhibition of heparin-
dependent platelet aggregation by HIT antibodies is generally achieved by
doses ranging from 15 to 30 ng/kg/min. One protocol that does not perform
intraoperative monitoring of platelet aggregation gradually increases epo-

prostenol infusion (in 5 ng/kg/min increments made at 5-min intervals) until
the target rate (30 ng/kg/min) is reached, whereupon standard-dose UFH
anticoagulation is commenced (Aouifi et al., 2001). The epoprostenol infusion
is continued until 15 min following reversal of UFH with protamine. The
major adverse effect is vasodilatation, leading to severe hypotension that
requires intraoperative vasopressors.
Pretreatment of patients with more conventional antiplatelet agents, such
as aspirin and dipyridamole, followed by heparin use, has been used success-
fully in patients with a documented previous history of HIT (Makhoul et al.,
1987). However, such an approach is controversial for a patient with acute
HIT, because in vitro activation of platelets by HIT antibodies is not reliably
inhibited by these relatively weak antiplatelet agents (Kappa et al., 1987).
Koster and coworkers (2000c, 2001a,b) combined the short-acting GP
IIb/IIIa inhibitor tirofiban with heparin for anticoagulation during CPB for
patients with HIT and renal insufficiency; pre- and postoperative anticoag-
ulation with lepirudin was used. Tirofiban is given 10 min before standard-
dose UFH as a 10 Ag/kg bolus followed by 0.15 Ag/kg/min continuous
infusion. The tirofiban infusion is stopped 1 hour before end of surgery.
UFH is neutralized with protamine as per usual. Using this treatment pro-
tocol, no thromboses occurred. However, in patients with severe renal im-
pairment, tirofiban persists in the circulation and can cause major bleeding
refractory to platelet transfusions: three such cases led the manufacturer to
discourage use of this off-label protocol (Warkentin and Greinacher, 2003).
In such patients, extracorporeal elimination of tirofiban (e.g., ultrafiltration
at the end of CPB or modified zero-balanced ultrafiltration after CPB) might
be required.
E. Other Anticoagulant Strategies

The thrombin-like snake venom ancrod (Arvin) is a defibrinogenating agent
that cleaves fibrinopeptide A, but not fibrinopeptide B, from fibrinogen. This
results in formation of fibrinogen–fibrin polymers into an unstable configu-
ration that is susceptible to rapid degradation by plasmin. Ancrod has been
used as a treatment for HIT (Cole and Bormanis, 1988; Demers et al., 1991),
including as an alternative anticoagulant for cardiac surgery requiring CPB
(Zulys et al., 1989; Teasdale et al., 1989). The recommended initial dose is
usually 70 U in normal saline, administered slowly, over at least 6–12 h.
There are some important disadvantages of using ancrod for cardiac
surgery. First, ancrod must be given slowly, because a very rapid infusion can
lead to life-threatening intravascular fibrin deposition. Thus, it is not appro-
priate for emergency situations. Second, it is difficult to determine accurately
the fibrinogen level at the recommended target fibrinogen range (0.2–0.5 g/L).
Furthermore, it is uncertain what fibrinogen level is required, if any, to
prevent clinically important fibrin formation during CPB. Third, reversal of
‘‘anticoagulation’’ requires a blood product, fibrinogen concentrates
(Europe) or cryoprecipitate (North America), to replace fibrinogen. Finally,
ancrod does not inhibit thrombin generation and has even been associated
with increased thrombin generation in some clinical settings, such as acute
HIT (Warkentin, 1998). It is possible that this could lead to thrombotic or
post-CPB hemorrhagic complications when used for the management of
acute HIT. All of the considerations suggest that ancrod is not a suitable
alternative to heparin in the setting of CPB surgery.
III. USE OF HEPARIN FOR CPB IN PATIENTS

WITH A REMOTE HISTORY OF HIT
An intriguing option for patients with a history of HIT, but in whom per-
sisting HIT antibodies can no longer be detected, is to consider reexposure to
heparin for CPB, and to avoid heparin completely both before surgery (e.g., at
heart catheterization) and in the postoperative period. This approach has
been used successfully by some physicians (Makhoul et al., 1987; Pötzsch et
al., 2000; Selleng et al., 2001; Warkentin and Kelton, 2001), and it is based on
the following rationale. First, HIT antibodies are transient, and they usually
are not detectable after 100 days following an episode of HIT (see Chap. 3).
Thus, no immediate problems would be expected in a patient without residual
HIT antibodies whose previous episode of HIT was ‘‘remote’’ (i.e., more than
several months before the need for heart surgery). Second, it appears that a
minimum of 5 days are required before clinically significant levels of HIT
antibodies are generated following any episode of heparin treatment (War-
kentin and Kelton, 2001). In the event that a recurrent immune response to
platelet factor 4–heparin is induced by reexposure to heparin during CPB, it is
unlikely that the newly generated HIT antibodies will contact exogenously
administered heparin. As a consequence, platelet activation by HIT anti-
bodies should not occur, and thus the thrombotic risk should not be increased.
We reported 10 patients with a documented history of HIT, but no detectable
HIT antibodies at the time of the proposed surgery, who thus underwent CPB
anticoagulation with heparin (Pötzsch et al., 2000). In none of the 10 patients
was a thromboembolic complication or prolonged thrombocytopenia ob-
served. Further, no increase in HIT antibody concentrations occurred during
a 10-day follow-up period. These data are in contrast to reports of a rapid
‘‘anamnestic’’ immune response in HIT. However, there is evidence that these
episodes represent acute-onset HIT in a patient who has residual circulating
HIT antibodies resulting from a recent episode of HIT, rather than a rapid
recurrence of HIT antibodies because of immune memory caused by a remote
exposure to heparin (Warkentin and Kelton, 2001).
As outlined in Fig. 2, we recommend that HIT antibody–negative
patients with a history of HIT who require CPB for heart surgery should be
treated according to established heparin protocols. The use of heparin should
be restricted to the operative period itself; if necessary, postoperative anti-
coagulation should be achieved with an alternative anticoagulant (see Chaps.
13–16).
Testing for HIT antibodies in patients with a history of HIT before
anticipated heparin reexposure at heart surgery should be performed using
one or more sensitive tests (see Chap. 11) if this approach is to be considered.
Particularly in cardiac surgical centers where there is limited experience with
nonheparin anticoagulation for CPB, risk-benefit considerations favor a brief
use of heparin for these patients. For example, a patient who developed near-
fatal CPB circuit thrombosis during danaparoid anticoagulation had had
HIT 11 years earlier and had no detectable HIT antibodies at the time
danaparoid was used (Grocott et al., 1997).
Figure 2 Algorithm for decision making for alternative anticoagulation in HIT

patients.
IV. ACUTE HIT BEFORE AND AFTER HEART SURGERY
There are several reasons why a patient with HIT might require urgent heart
surgery, including the result of life-threatening thrombotic complications of
HIT affecting the heart (e.g., acute coronary insufficiency or myocardial
infarction; removal of intracardiac thrombus), or because HIT has compli-
cated the course of a critically ill patient receiving heparin before anticipated
heart surgery (e.g., while awaiting a heart for cardiac transplantation, or
during use of an intra-aortic balloon pump). The latter group of patients
appear to have a relatively high risk of developing HIT (Walls et al., 1992). A
suitable alternative for patients who require urgent coronary revasculariza-
tion during acute HIT might be the use of an off-pump (‘‘beating-heart’’)
strategy. This surgical technique requires lower levels of anticoagulation.
Therefore, the dose of the nonheparin anticoagulant could be markedly
reduced, which could reduce postoperative bleeding. Such an approach using
danaparoid as heparin substitute was recently reported by Warkentin and
colleagues (2001). The authors found a minimum plasma anti-Xa level of 0.6
U/mL sufficient to perform surgery off-pump.
Studies of the frequency of HIT antibody formation (Visentin et al.,
1996; Bauer et al., 1997; Warkentin et al., 2000) following heart surgery
suggest that as many as 15–50% of patients form HIT antibodies, using an
enzyme immunoassay that detects IgG antibodies that recognize platelet
factor 4–heparin complexes (see Chap. 4). With the washed platelet serotonin-
release assay, HIT antibodies are detected in 13–20% of patients (Bauer et al.,
1997; Warkentin et al., 2000). However, despite this high rate of serocon-
version, only about 1–3% of patients who receive further postoperative
anticoagulation with unfractionated heparin develop HIT. Currently, there
is no convincing evidence that patients who form HIT antibodies in the
absence of thrombocytopenia are at increased risk for thrombosis (Bauer
et al., 1997; Trossaërt et al., 1998; Warkentin et al., 2000). However,
postoperative cardiac surgical patients who develop clinical HIT appear to
be at increased risk for both venous and arterial thrombotic events (Walls
et al., 1990; van Dyck et al., 1996; Pouplard et al., 1999).
V. DECISION MAKING FOR ANTICOAGULATION IN HIT

PATIENTS
Given these data and clinical experience, an algorithm has been developed to
assist in determining the need for alternative anticoagulation for CPB in HIT
patients (Fig. 2). After the decision to avoid use of heparin in the CPB setting,
the important remaining question is, which strategy should be chosen? Be-
cause each of the different approaches described here provides specific
Figure 3 Treatment strategies for HIT patients requiring alternative anticoagula-

tion for cardiopulmonary bypass. The estimated risk of bleeding is highest for azo-
temic patients receiving hirudin, intermediate for patients receiving danaparoid, and
lowest for patients with normal renal function receiving hirudin or bivalirudin. Ab-
breviations: CPB, cardiopulmonary bypass; ECT, ecarin clotting time.
advantages and limitations, it is not possible to recommend one treatment

regimen that is applicable for each patient. The algorithm shown in Figure 3
considers typical clinical settings in HIT patients and is constructed to support
the decision in finding which of the different approaches is best suited for the
individual patient.
REFERENCES
Addonizio VP Jr, Fisher CA, Kappa JR, Ellison N. Prevention of heparin-induced

thrombocytopenia during open heart surgery with iloprost (ZK36374). Surgery
1987; 102:796–807.
Aouifi A, Blanc P, Piriou V, Bastien OH, Ffrench P, Hanss M, Lehot JJ. Cardiac
surgery with cardiopulmonary bypass in patients with type II heparin-induced
thrombocytopenia. Ann Thorac Surg 2001; 71:678–683.
heparin-associated antibodies without thrombosis in patients undergoing
cardiopulmonary bypass surgery. Circulation 1997; 95:1242–1246.
Cole CW, Bormanis J. Ancrod: a practical alternative to heparin. J Vasc Surg 1988;
8:59–63.
Davis Z, Anderson R, Short D, Garber D, Valgiusti A. Favorable outcome with
bivalirudin anticoagulation during cardiopulmonary bypass. Ann Thorac Surg
2003; 75:264–265.
Demers C, Ginsberg JS, Brill-Edwards P, Panju A, Warkentin TE, Anderson DR,
Turner C, Kelton JG. Rapid anticoagulation using ancrod for heparin-induced
Doherty DC, Ortel TL, De Bruijn N, Greenberg CS, Van Trigt P III. ‘‘Heparin-free’’
cardiopulmonary bypass: first reported use of heparinoid (Org 10172) to pro-
vide anticoagulation for cardiopulmonary bypass. Anesthesiology 1990; 73:
562–565.
Edmunds LH Jr. Blood-surface interactions during cardiopulmonary bypass. J Car-
diovasc Surg 1993; 8:404–410.
Furukawa K, Ohteki H, Hirahara K, Narita Y, Koga S. The use of argatroban as an
anticoagulant for cardiopulmonary bypass in cardiac operations. J Thorac
Cardiovasc Surg 2001; 122:1255–1257.
Gitlin SD, Deeb GM, Yann C, Schmaier AH. Intraoperative monitoring of dan-
aparoid sodium anticoagulation during cardiovascular operations. J Vasc Surg
1998; 27:568–575.
Grocott HP, Root J, Berkowitz SD, deBruijn N, Landolfo K. Coagulation compli-
cating cardiopulmonary bypass in a patient with heparin-induced thrombo-
cytopenia receiving the heparinoid, danaparoid sodium. J Cardiothorac Vasc
Anesth 1997; 11:875–877.
Henny CP, ten Cate H, ten Cate JW, Moulijn AC, Sie TH, Warren P, Büler HR. A
randomized blind study comparing standard heparin and a new low molecular
weight heparinoid in cardiopulmonary bypass in dogs. J Lab Clin Med 1985;
106:187–196.
Insler SR, Kraenzler EJ, Bartholomew JR, Kottke-Marchant K, Lytle B, Starr NJ.
Thrombosis during use of the heparinoid Organon 10172 in a patient with
heparin-induced thrombocytopenia. Anesthesiology 1997; 86:495–498.
Kappa JR, Horn D, McIntosh CL, Fisher CA, Ellison N, Addonizio VP. Iloprost
(ZK36374), a new prostacyclin analogue, permits open cardiac surgery in pa-
tients with heparin-induced thrombocytopenia. Surg Forum 1985; 36:285–286.
Kappa JR, Cottrell ED, Berkowitz HD, Fisher CA, Sobel M, Ellison N, Addonizio
VP Jr. Carotid endarterectomy in patients with heparin-induced platelet acti-
vation: comparative efficacy of aspirin and iloprost (ZK36374). J Vasc Surg
1987; 5:693–701.
Mertzlufft F. Hirudin monitoring using the TAS ecarin clotting time in patients

2000a; 14:249–252.
Koster A, Pasic M, Bauer M, Kuppe H, Hetzer R. Hirudin as anticoagulant for car-
diopulmonary bypass: importance of reoperative renal function. Ann Thorac
Surg 2000b; 69:37–41.
Koster A, Loebe M, Merztlufft F, Kuppe H, Hetzer R. Cardiopulmonary bypass in a
patient with heparin induced thrombocytopenia II and impaired renal function
using heparin and platelet GP IIb/IIIa inhibitor tirofiban as anticoagulant. Ann
Thorac Surg 2000c; 70:2160–2161.
Koster A, Kukucka M, Bach F, Meyer O, Fischer T, Mertzlufft F, Loebe M, Hetzer R,
Kuppe H. Anticoagulation during cardiopulmonary bypass in patients with
heparin-induced thrombocytopenia type II and renal impairment using heparin
and the platelet glycoprotein IIb-IIIa antagonist tirofiban. Anesthesiol 2001a;
94:245–251.
Koster A, Meyer O, Fischer T, Kuschka M, Krabatsch T, Bauer M, Kuppe H, Hetzer
R. One-year experience with the platelet glycoprotein IIb/IIIa antagonist
tirofiban and heparin during cardiopulmonary bypass in patients with heparin-
induced thrombocytopenia type II. J Thorac Cardiovasc Surg 2001b; 122:1254–
1255.
Koster A, Chew D, Grundel M, Bauer M, Kuppe H, Spiess BD. Bivalirudin monitored
with the ecarin clotting time for anticoagulation during cardiopulmonary-
bypass. Anesth Analg 2003; 96:383–386.
Koza MJ, Walenga JM, Fareed J, Pifarre R. A new approach in monitoring recom-
binant hirudin during cardiopulmonary bypass. Semin Thromb Hemost 1993;
19:90–96.
Kraenzler EJ, Starr NJ. Heparin-associated thrombocytopenia: management of pa-
tients for open heart surgery. Case reports describing the use of iloprost. Anes-
thesiology 1988; 69:964–967.
Lewis BE, Johnson SA, Grassman ED, Wrona LL. Argatroban as an anticoagulant for
coronary procedures in patients with HIT antibody. In: Pifarré R, ed. New
Anticoagulants for the Cardiovascular Patient. Philadelphia: Hanley & Belfus,
1997a:301–308.
Lewis BE, Walenga JM, Pifarré R, Fareed J. Argatroban in the management of
patients with heparin-induced thrombocytopenia and heparin-induced throm-
bocytopenia and thrombosis syndrome. In: Pifarré R, ed. New Anticoagulants
for the Cardiovascular Patient. Philadelphia: Hanley & Belfus, 1997b:223–229.
Lewis BE, Wallis DE, Berkowitz SD, Matthai WH, Fareed J, Walenga JM, Bar-
tholomew J, Sham R, Lerner RG, Zeigler ZR, Rustagi PK, Jang I-K, Rifkin SD,
Moran J, Hursting MJ, Kelton JG for the ARG-911 Study Investigators.
Argatroban anticoagulant therapy in patients with heparin-induced thrombo-
cytopenia. Circulation 2001; 103:1838–1843.
Lewis BE, Wallis DE, Leya F, Hursting MJ, Kelton JG, Argatroban-915 Investigators.
Argatroban anticoagulation in patients with heparin-induced thrombocytope-
nia. Arch Intern Med 2003; 163:1849–1856.
Long RW. Management of patients with heparin-induced thrombocytopenia re-
quiring cardiopulmonary bypass. J Thorac Cardiovasc Surg 1985; 89:950–

951.
Magnani HN, Beijering RJR, ten Cate JW, Chong BH. Orgaran anticoagulation
for cardiopulmonary bypass in patients with heparin-induced thrombocyto-
penia. In: Pifarré R ed. New Anticoagulants for the Cardiovascular Patient.
Philadelphia: Hanley & Belfus, 1997:487–500.
Makhoul RG, McCann RL, Austin EH, Greenberg CS, Lowe JE. Management of
patients with heparin-associated thrombocytopenia and thrombosis requiring
cardiac surgery. Ann Thorac Surg 1987; 43:617–621.
Merry AF, Raudkivi P, White HD, Nand P, Middleton N, McDougall J, Mills BP,
Frampton CM. Bivalirudin versus heparin and protamine in off pump coronary
artery bypass surgery. Ann Thorac Surg 2004. In press.
Mertzlufft F, Kuppe H, Koster A. Management of urgent high-risk cardiopulmonary
bypass in patients with heparin-induced thrombocytopenia type II and co-
existing disorders of renal function: use of heparin and epoprostenol combined
with on-line monitoring of platelet function. J Cardiothorac Vasc Anesth 2000;
14::304–308.
Nishida S, Fujita T, Kohno N, Atoda H, Morita T, Takeya H, Kido I, Paine MJI,
Kawabata S, Iwanaga S. cDNA cloning and deduced amino acid sequence of
prothrombin activator (ecarin) from Kenyan Echis carinatus venom. Biochem-
istry 1995; 34:1771–1778.
Palmer Smith J, Walls JT, Muscato MS, Scott McCord E, Worth ER, Curtis JJ, Silver
D. Extracorporeal circulation in a patient with heparin-induced thrombocyto-
penia. Anesthesiology 1985; 62:363–365.
Pötzsch B, Iversen S, Riess FC, Tzanova N, Seelig C, Nowak G, Müller-Berghaus G.
Recombinant hirudin as an anticoagulant in open-heart surgery: a case report
[abstr]. Ann Hematol 1993; 68(suppl 2):A46.
Pötzsch B, Madlener K, Seelig C, Riess CF, Greinacher A, Müller-Berghaus G. Mo-
nitoring of r-hirudin anticoagulation during cardiopulmonary bypass—assess-
ment of the whole blood ecarin clotting time. Thromb Haemost 1997; 77:920–925.
Antibodies to platelet factor 4-heparin after cardiopulmonary bypass in pa-
tients anticoagulated with unfractionated heparin or a low-molecular-weight
heparin: clinical implications for heparin-induced thrombocytopenia. Circu-
lation 1999; 99:2530–2536.
N Engl J Med 2000; 343:515.
Riess FC, Löwer C, Seelig C, Bleese N, Kormann J, Müller-Berghaus G, Pötzsch B.

Recombinant hirudin as a new anticoagulant during cardiac operations instead
of heparin: successful for aortic valve replacement in man. Thorac Cardiovasc
Surg 1995; 110:265–267.
Riess FC, Pötzsch B, Bader K, Bleese N, Greinacher A, Löwer C, Madlener K, Müller-
Berghaus G. A case report on the use of recombinant hirudin as an anticoag-
ulant for cardiopulmonary bypass in open heart surgery. Eur J Cardiothorac
Surg 1996; 10:386–388.
Riess FC, Poetzsch B, Mueller-Berghaus G. Recombinant hirudin as an anticoagulant
during cardiac surgery. In: Pifarré R, ed. New Anticoagulants for the Car-
diovascular Patient. Philadelphia: Hanley & Belfus, 1997:197–222.
Slaughter TF, LeBleu TH, Douglas JM Jr, Leslie JB, Parker JK, Greenberg CS.
Characterization of prothrombin activation during cardiac surgery by hemo-
static molecular markers. Anesthesiology 1994; 80:520–526.
Spiess BD, DeAnda A, McCarthy A, Yeatman D, Harness HL, Katlaps G. Off pump
CABG in a patient with HIT anticoagulated with bivalirudin: a case report
[abstr]. Anesth Analg 2002; 93:SCA70.
Teasdale SJ, Zulys VJ, Mycyk T, Baird RJ, Glynn MFX. Ancrod anticoagulation for
cardiopulmonary bypass in heparin-induced thrombocytopenia and thrombo-
sis. Ann Thorac Surg 1989; 48, 712–713.
Trossaërt M, Gaillard A, Commin PL, Amiral J, Vissac AM, Fressinaud E. High
incidence of anti-heparin/platelet factor 4 antibodies after cardiopulmonary
bypass surgery. Br J Haematol 1998; 101:653–655.
Van Dyck MJ, Lavenne-Pardonge E, Azerad M-A, Matta AG, Moriau M, Comunale
ME. Thrombosis after the use of heparin-coated cardiopulmonary bypass
circuit in a patient with heparin-induced thrombocytopenia. J Cardiothorac
Vasc Anesth 1996; 10:809–815.
Vasquez JC, Vichiendilokkul A, Mahmood S, Baciewicz FA Jr. Anticoagulation with
bivalirudin during cardiopulmonary bypass in cardiac surgery. Ann Thorac
Surg 2002; 74:2177–2179.
Visentin GP, Malik M, Cyganiak KA, Aster RH. Patients treated with unfraction-
ated heparin during open heart surgery are at high risk to form antibodies re-
active with heparin: platelet factor 4 complexes. J Lab Clin Med 1996; 128:376–
383.
Walenga JM, Bakhos M, Messmore HL, Koza M, Wallock M, Orfei E, Fareed J,
Pifarre R. Comparison of recombinant hirudin and heparin as an anticoagulant
in a cardiopulmonary bypass model. Blood Coagul Fibrinolysis 1991; 2:105–
111.
Walls JT, Curtis JJ, Silver D, Boley TM. Heparin-induced thrombocytopenia in
patients who undergo open heart surgery. Surgery 1990; 108:686–693.
Walls JT, Boley TM, Curtis JJ, Silver D. Heparin-induced thrombocytopenia in
patients undergoing intra-aortic balloon pumping after open heart surgery.
ASAIO J 1992; 38:M574–M576.

N Engl J Med 2001; 344:1286–1292.
Warkentin TE, Sheppard JI, Horsewood P, Simpson PJ, Moore JC, Kelton JG. Impact
of the patient population on the risk of heparin-induced thrombocytopenia.
Blood 2000; 96:1703–1708.
Warkentin TE, Dunn GL, Cybulsky IJ. Off-pump coronary artery bypass grafting
for acute heparin-induced thrombocytopenia. Ann Thorac Surg, 2001; 72:1730–
1732.
Wilhelm MJ, Schmid C, Kececioglu D, Möllhoff T, Ostermann H, Scheld HH.
Cardiopulmonary bypass in patients with heparin-induced thrombocytopenia
using Org 10172. Ann Thorac Surg 1996; 61:920–924.
Zulys VJ, Teasdale SJ, Michel ER, Skala RA, Keating SE, Viger JR, Glynn MFX.
Ancrod (Arvin) as an alternative to heparin anticoagulation for cardiopulmo-
nary bypass. Anesthiology 1989; 71:870–877.
20
in Children
Anne F. Klenner and Andreas Greinacher

I. INTRODUCTION
Heparin-induced thrombocytopenia (HIT) can occur in children, with the

potential for severe venous and arterial thrombotic complications (Table 1).
Unlike adults, few data exist regarding pediatric HIT. Only 68 children have
been reported with HIT between 1990 and 2003, including 12 new cases added
in this chapter involving our medical center for diagnosis and/or treatment
(Table 2).
II. PATHOPHYSIOLOGY
Studies of the pathophysiology of HIT have been performed using adult

blood. In our laboratory, pediatric and adult HIT sera react similarly in var-
ious in vitro assays. Therefore, it seems reasonable to infer that the patho-
physiology of HIT in children resembles that in adults (see Chaps. 5–10).
III. FREQUENCY
Only four studies have addressed the frequency of HIT in children. Spadone
and coworkers (1992) collected cases of suspected HIT in a neonatal intensive
care unit (ICU) between 1988 and 1990. Of 1329 newborns enrolled, about
70% received unfractionated heparin (UFH): either 0.5–1.0 IU/mL added to
553
554 Klenner and Greinacher
Table 1 Clinical Complicationsa of HIT in Children (n = 68)
Complication Percentage Case numbers
Venous thrombosis
Iliac vein 16.2 21, 24, 27, 32, 38, 39, 42,
43, 48, 53, 60
Femoral vein 13.2 21, 27, 34, 35, 36, 43, 48,
54, 58
Inferior vena cava 11.8 9, 24, 35, 38, 43, 48, 51, 58
Pulmonary embolism 10.3 6, 32, 37, 49, 51, 65, 66
Progression of deep-vein 5.9 34, 40, 42, 44
thrombosis
Subclavian vein 5.9 17, 36, 38, 55
Calf vein 5.9 32, 35, 36, 48
Superior vena cava 5.9 12, 13, 54, 57
Jugular vein 2.9 62, 64
Rare: renal vein, arm veins, <2 9, 10, 7, 45, 44
intracranial veins, pulmonary
vein, shunt
Arterial thrombosis
Femoral artery 4.4 11, 23, 59
Iliac artery 2.9 2, 16
Foot arteries 2.9 23, 30
Rare: renal artery <2 54
Others
Intracardiac thrombi 5.9 4, 19, 25, 37
Bleeding 5.9 21, 40, 47, 63
Clots in dialyzer 4.4 6, 65, 66
Neurological deficits 2.9 36, 41
Rare: subdural hematoma, <2 61
thrombosis of pulmonary valve
a
Patients may have had more than one complication. Case numbers refer to Table 2.
central venous or peripheral/umbilical artery catheters or via flushing of

peripheral venous catheters (10 IU/mL UFH-saline every 4 h). In 34 (3.7%)
newborns, HIT was suspected because the platelet count fell to less than 70
109/L or because of new thromboembolic events. In 14 of these 34 infants,
HIT antibodies were detected by platelet aggregation assay (incidence 14/930
= 1.5%). However, this study has several limitations. It is an observational
study without a defined protocol. Differentiating HIT from other causes of
thrombocytopenia or thrombosis is difficult. Further, the specificity of plate-
let aggregation testing for HIT antibodies may be low in ICU patients (see
Chap. 11).
Heparin-Induced Thrombocytopenia in Children 555
In a retrospective cohort study in a pediatric ICU, 57 patients developed

arterial and/or venous thrombosis among 612 children treated with UFH for
more than 5 days (Schmugge et al., 2002). In 14 children (2.3%), HIT was
suspected based on thrombosis and a platelet count below 150 109/L (or
platelet fall exceeding 50%) occurring after 5 or more days of UFH use. In
six patients (1.0%), HIT antibodies were demonstrated by platelet factor 4
(PF4)–dependent enzyme immunoassay (EIA) using adult cutoff values for a
positive assay result (Table 2). The eight other patients with clinically
suspected HIT had antibody levels below adult cutoff (Table 2). Eleven of
the 14 patients had received UFH following cardiac surgery. Four were new-
borns and five others were also under one year of age (mean age 6.5 months).
Newall et al. (2003) retrospectively collected cases of HIT in a tertiary
pediatric hospital. During the 2-year study, 116 patients received UFH over a
7-day period (25 reexposures). HIT was suspected in 4 patients who received
therapeutic-dose UFH and developed a platelet count fall of more than 85%
of the preheparin value. Three of the patients were tested for HIT antibodies,
with one positive result (incidence 1/116 = 0.9%).
Between 1999 and 2002 we performed a prospective, randomized, dou-
ble-blind, placebo-controlled trial in a neonatal ICU to assess the impact of
intravenous (i.v.) UFH on the patency of peripheral venous catheters (Klen-
ner et al., 2003b). Of 213 eligible infants, 108 received UFH (0.5 IU/mL) and
105 received saline for at least 5 days. None developed HIT or HIT antibodies
(assessed by EIA). This suggests that the incidence of HIT is lower in neonatal
ICU patients than previously reported. As mentioned, one reason for this
difference could be false-positive results in the platelet aggregation test per-
formed in an earlier study (Spadone et al., 1992). Another reason could be
differences in patient population. For example, we did not enroll neonates
following cardiac surgery, whereas Schmugge and coworkers (2002) noted
that most of their neonates/infants with HIT had received UFH after cardiac
surgery.
IV. CLINICAL PRESENTATION
Table 2 summarizes 68 children with HIT. Thirteen (19.1%) were newborns,

22 (32.4%) were children aged 3 months to 3 years, 11 (16.2%) were between 4
and 11 years of age, and 22 (32.4%) ranged in age from 12 to 18 years (Fig. 1).
In newborns and young children (under 4 years of age), HIT usually occurred
after cardiac surgery (28/35 = 80%). In contrast, among 22 children aged 12
years or older, HIT complicated the use of UFH given because of preceding
thrombosis in 13 (59.1%) patients, and following use of antithrombotic
prophylaxis in 5 (22.7%); only one older child had undergone cardiac surgery.
Table 2 Characteristics of Children with HIT

Gender, Platelet nadir Thrombocytopenia (TP) and
No. age Diagnosis, procedure ( 109/L) thrombotic complication(s)
A: Confirmed HIT (clinical criteria present, laboratory test positive); all children received UFH except
case 45
1 M, nb Preterm, sepsis <100 TP
2 F, nb Heart surgery 191 Iliac artery
3 M, nb Norwood I, hypoplastic heart 36 TP
4 M, nb ECMO <1 TP, ventricular thrombus
5 F, nb Tetralogy of Fallot, heart 45 TP
surgery
6 F, nb Aortic stenosis, hypoplastic NA TP, clots in circuit, PE, lung
left ventricle, heart surgery hemorrhage, renal vein
7 F, nb CHD, heart surgery 48 TP, clot in shunt
8 F, 2m Tetralogy of Fallot, premature, TP
heart surgery, NEC
9 M, 3m Tricuspid valve atresia, 34 Vena cava, renal vein
Blalock-Taussig shunt
10 M, 5m Hypoplastic left heart, heart TP, left pulmonary vein
surgery, ECMO
11 F, 6m Heart surgery 46 Fem artery, DVT, TP
12 M, 8m Tetralogy of Fallot, heart surgery TP, clot in superior vena cava
13 M, 10m Heart surgery 46 TP, vena cava
14 M, 10m Preterm, VACTERL-syndrome 45 TP
15 M, 12m Renal failure, tetralogy of NA TP
Fallot
16 M, 13m CHD, renal agenesis, heart 46 TP, iliac artery
surgery
17 F, 15m Heart surgery 123 TP, SC
18 M, 15m CHD, heart surgery 10 TP
19 F, 17m Acute myocarditis 80 TP, intracardiac
20 M, 23m Hemofiltration NA TP
21 F, 2y Fontan operation 55 TP, DVT
22 M, 2y Fontan operation 73 TP, heparin resistance
23 M, 3y Tricuspid, pulmonary valve 40 Fem artery, foot gangrene
atresia, Fontan operation
24 M, 4y Double-inlet left ventricle, 25 TP, vena cava, DVT
Fontan operation
25 F, 4y Cardiomyopathy, heart 16 TP, intracardiac, bleeding
transplant
26 F, 4y Lung disease, mechanical 50 TP
ventilation
27 F, 7y Cardiomyopathy 71 TP, DVT
28 F, 8y Turner’s syndrome, sinus vein 190 TP, heparin resistance
thrombosis
29 F, 9y DVT, APS 82 TP
30 M, 10y Leg artery thrombosis, PS def 39 TP, foot arteries
31 M, 11y APS 52 TP
Table 2 Characteristics of Children with HIT

Test for HIT
No. antibodies Treatment Outcome (Ref.)
1 HIPA DS(XR?), +lep ? (Nguyen et al., 2003)

2 HIPA Unknown S (Schmugge et al., 2002)
3 EIA DS, ASA S (Girisch et al., 2002)
4 HIPA None y (Butler et al., 1997)
5 HIPA ASA S (Boshkov et al., 2003a)
6 HIPA Arg y bleeding (Boshkov et al., 2003a)
7 HIPA Arg y (Boshkov et al., 2003a)

8 HIPA ASA, Arg S (Boshkov et al., 2003a)
9 HIPA None S (Murdoch et al., 1993)
10 HIPA ASA, Arg S (Boshkov et al., 2003a)
11 EIA Unknown S (Schmugge et al., 2002)

12 HIPA Arg S (Boshkov et al., 2003a)
14 EIA DS S (Ranze et al., 2001)
15 HIPA DS S (Ranze et al., 1999)
16 EIA,HIPA DS ? (new case)

18 HIPA DS S (new case)
19 HIPA LMWH S (Oriot et al., 1990)
20 ‘‘Screen’’ DS S (Newall et al., 2003)
21 EIA,SRA DS S (Saxon et al., 1999)
22 HIPA,EIA Lep, OAC S (Severin et al., 2002a)
23 EIA,HIPA Platelets Amp below knee (Porcelli et al., 2003)
24 HIPA Platelets y (Porcelli et al., 2003)
25 EIA Lep y (Deitcher et al., 2002)
26 EIA Lep ? (new case)
27 EIA Lep S (Deitcher et al., 2002)

28 EIA Lep, OAC S (Severin et al., 2002b)
29 HIPA, EIA DS, OAC Residual clot (Zöhrer et al., 2001)

30 HIPA DS, OAC Amp digits (Zöhrer et al., 2001)
31 EIA DS S (Bocquet et al., 1999)
Table 2 Continued
32 F, 11y DVT 12 TP, DVT, PE

33 M, 12y Hemodialysis No TP Pos HIT test before kidney
transpl
34 M, 12y DVT, PS def, vena cava filter 80 DVT, TP
35 M, 12y DVT, PE 46 DVT, vena cava, TP
36 M, 13y Closure of Blalock-Taussig shunt 52 TP, stroke, DVT, SC vein
37 F, 13y PE, Intracardiac thrombus, No TP PE, intracardiac
DVT, APS
38 M, 13y Embolization of aorto- 55 SC vein, vena cava, DVT, TP
pulmonary shunts after
lung bleeding
39 M, 13y Meningococcal sepsis 27 TP, DVT
40 F, 14y PE, vena cava, APS 24 TP, progressive thrombosis
41 M, 14y Fem artery, mitral valve tumor 60 TP, amaurosis fugax
42 M, 15y DVT No TP Progressive thrombosis
43 M, 15y DVT, PS def 228 DVT, vena cava
44 F, 15y Sinus vein thrombosis No TP Progressive thrombosis
45 F, 16y DVT 77 TP, arm veins
46 M, 16y Stroke 79 TP, phlebitis, sepsis
47 M, 17y Rhabdomyosarcoma 15 TP
48 M, 17y Head trauma No TP DVT, vena cava
49 F, 18y DVT, PS def ? DVT, PE
50 F, 18y Trauma 38 TP
51 M, 18y Ulcerative colitis, DVT No TP PE, vena cava
B: Probable Hit (clinical criteria present, laboratory test borderline)

52 M, nb Transposition of great vessels 7 TP
C: Unlikely HIT (laboratory test negative)

53 M, nb Heart surgery 55 TP, DVT
54 M, nb Heart surgery 29 TP, DVT, vena cava, renal
artery
55 F, nb Heart surgery 60 TP, SC vein
56 F, nb Hypoplastic left heart NA TP
57 M, 2.5m Heart failure NA TP, vena cava
58 M, 3m Abdominal surgery 43 TP, DVT, vena cava
59 M, 3m Heart surgery 37 TP, fem artery
60 M, 7m Heart surgery 43 TP, DVT
61 F, 7m CHD, heart surgery 86 TP, thrombosis pulmonary
valve, subdural hematoma
62 M, 17m Heart surgery 80 TP, jugular vein
63 F, 2.5y Atrioventricular septal defect, NA TP, bleeding
double outlet right ventricle,
pulmonary stenosis
64 F, 4y Heart surgery 93 TP, jugular vein
Test for HIT

32 HIPA Lep S, hematuria (new case)

33 EIA, HIPA Unknown ? (new case)
34 HIPA DS, OAC Occlusion vena cava (Zöhrer et al., 2001)

35 HIPA, EIA DS, lep Amp forefoot (Schiffmann et al., 1997)
36 SRA, EIA OAC, ASA Stroke, SC vein obstruction (Potter et al.,
1992)
37 EIA DS(XR), +lep S (Zöhrer et al., 2001)

40 HIPA LMWH, OAC S (Murdoch et al., 1993)
41 HIPA DS S (Wilhelm et al., 1996)
42 HIPA DS, r-tPA S, bleeding (Klement et al., 1996)
43 EIA UK, r-tPA, lep, OAC S, residual clots (Severin and Sutor, 2001)
44 HIPA, EIA DS S (new case)
45 HIPA r-tPA, DS, OAC S (new case)
46 HIPA ASA ? (new case)
47 HIPA DS S (Sauer et al., 1998)
48 HIPA Lep S (new case)
49 HIPA Lep, OAC S (new case)
52 Borderline Lep y (Klenner et al., 2003a)

55 EIA unknown S (Schmugge et al., 2002)

56 HIT screen DS, platelets S (Newall et al., 2003)
57 not done DS y (Newall et al., 2003)
61 Not done LMWH, r-tPA, DS y lung hemorrhage (Weigel et al., 1999)

63 HIT screen Platelets S (Newall et al., 2003)

Table 2 Continued
65 M, 10y Hemodialysis, hypoplastic 38 TP, clots in dialyzer

kidney
66 M, 14y Hemodialysis 109 TP, clots in dialyzer
67 F, 17y Atrial thrombus, DVT, 30 TP
PE, APS
68 F, 17y Prophylaxis NA n.a.
Abbr: y, died; Amp, amputation; APS, antiphospholipid syndrome; Arg, argatroban; ASA, aspirin
(acetylsalicylic acid); CHD, congenital heart disease; DS, danaparoid sodium; DVT, deep-vein throm-
bosis; ECMO, extracorporeal membrane oxygenation; EIA, PF4-dependent enzyme immunoassay; F,
female; fem, femoral; HIPA, heparin-induced platelet activation test; Lep, lepirudin; +Lep, switch to
lepirudin after suspected or confirmed cross-reactivity to danaparoid; LMWH, low molecular weight hep-
arin; m, month; M, male; MTHFR, methylene tetrahydrofolate reductase; NA, not available; nb, new-
born; NEC, necrotizing enterocolitis; OAC, oral anticoagulants; PE, pulmonary embolism; PS def, protein
S deficiency; r-tPA, recombinant tissue-plasminogen activator; S, survived; SC, subclavian; TP, throm-
bocytopenia; UAC, umbilical artery catheter; UFH, unfractionated heparin; UK, urokinase; XR,
confirmed cross-reactivity with danaparoid; XR?, suspected cross-reactivity with danaparoid; y, year.
Four patients developed HIT during low-dose UFH given for catheter pat-
ency (cases 1, 14, 47, 52). Hemodialysis or hemofiltration accounted for
UFH use in four (5.9%) patients (cases 20, 33, 65, 66). In 17 of the 68 pa-
tients, the laboratory test for HIT was negative or not performed (cases 53–68,
inclusive).
The most frequent manifestation of HIT in the 68 children was a
decrease in platelet count (85.3%). HIT was associated with thromboembolic
complications in about two thirds of the patients, most commonly involving
iliac and femoral veins, the inferior vena cava, and pulmonary embolism
(Table 1). Less commonly, intracardiac thrombi or neurological events oc-
curred, or clotting of the dialyzer. Only about 10% (7/68) of patients devel-
oped arterial thrombosis. Thus, there is a strong preponderance of venous
thrombosis in pediatric HIT.
Eight (11.8%) of the 68 children died (cases 4, 6, 7, 24, 25, 52, 57, 61) and
three required amputations (cases 23, 30, 35). In four children, only partial
recanalization of thrombosed veins occurred (cases 29, 34, 36, 43).
This summary does not include the 14 newborns reported by Spadone
and colleagues (1992). These workers primarily observed arterial thrombosis,
with at least 11 (78.6%) developing aortic thrombosis (one infant died with-
out imaging studies). Two newborns with thrombosis had normal platelet
counts. Eleven (78.6%) survived, the remaining three developing mesenteric
ischemia. Arterial thrombosis likely was related to umbilical artery catheters
Test for HIT

65 HIPA, EIA LMWH, DS S (Neuhaus et al., 2000)
66 HIPA, EIA LMWH, DS S (Neuhaus et al., 2000)

67 Not done DS S (Weigel et al., 1999)
68 None None S (Boon et al., 1994)
Figure 1 Reasons for preceding heparin therapy in children with HIT. Among the
various age groups, the reasons for heparin therapy that led to HIT varied consid-
erably: whereas newborns and infants usually developed HIT after cardiac surgery,
among teenagers HIT more often complicated the use of heparin during treatment of
thrombosis.
(used in all but one of the 14 neonates). In adults, intravascular catheters are
a risk factor for HIT-associated thrombosis (Hong et al., 2003), but whether
the arterial thrombi observed by Spadone et al. indeed were HIT-related is
unclear.
V. LABORATORY TESTING
As in adults, no data exist to justify routine screening for HIT antibodies

during heparin use in children. Laboratory tests for HIT should be used to
exclude or confirm clinically suspected HIT. During UFH therapy, platelet
counts should be monitored regularly (see Chap. 4), particularly between days
5 and 14 of heparin use, the time when over 90% of HIT patients develop their
platelet count fall (see Chap. 3).
For laboratory testing, functional and antigen tests are available (see
Chap. 11). Commercial antigen assays (EIA) are often used and are especially
appropriate for neonates and infants because small blood volumes are needed
(<100 AL vs. >1 mL for many platelet activation assays). However, the
appropriate cutoff level that defines a positive EIA result suitable for children
is debated. In a retrospective study, Schmugge and coworkers (2002) inves-
tigated cutoff levels for children using a commercial EIA (Asserachrom,
Stago). Among 612 children, HIT was suspected in 14 because of thrombo-
cytopenia and thrombosis. Positive test results (using the adult cutoff) were
seen in 6 of the 14 patients. In the remaining 8 children with suspected HIT,
test results ranged from 26 to 80% of the adult cutoff level, i.e., levels that were
higher than among controls (with wide overlap).
A retrospective analysis performed by Risch and colleagues (2003) of
the same 612 pediatric ICU patients initially reported by Schmugge and
coworkers (2002) addressed whether there was an association between anti-
PF4/heparin antibody levels (measured by EIA) and thrombosis. Ten patients
who developed thrombosis without thrombocytopenia constituted the study
group and were compared with 19 matched controls with neither thrombosis
nor thrombocytopenia. All 29 subjects had lower antibody levels than the
cutoff level recommended for adults. However, median assay results were
significantly higher in the thrombosis patients than in controls (51% vs. 23%
of the manufacturer’s cutoff; p = 0.004). The authors concluded that there
might be an association between anti-PF4/heparin antibody levels and throm-
bosis, even in the absence of thrombocytopenia or a positive test result (by
conventional criteria).
However, in our prospective, randomized, double-blind trial, we
screened 108 newborns receiving UFH for more than 5 days and 105 controls
using the PF4/polyvinyl sulfonate EIA (GTI, Brookfield) (Klenner et al.,
2003b). None of the infants developed HIT antibodies using the adult cutoff
[UFH group: mean optical density (OD), 0.020; maximum, 0.328; saline
group: mean OD, 0.019, maximum, 0.239]. Minor changes in OD (increase >
0.100) occurred in six patients (three in each group) (Fig. 2). Therefore, these
minor increases in OD are unlikely to be related to UFH use and could rep-
resent a nonspecific increase in antibody levels in ill patients (acute phase reac-
tion). Among the subjects receiving placebo, all OD values were below 0.400
(the accepted adult cutoff value), suggesting that this level is also appropriate
for neonates.
The limitations of antigen assays observed in adults likely also apply to
children. Thus, in 5–10% of cases, the antigen assay could be false-negative
Figure 2 Six neonates with rising absorbance levels in platelet factor 4–dependent
enzyme immunoassay (EIA). Six of 213 neonates participating in a randomized, dou-
ble-blind trial comparing heparin with normal saline for maintenance of peripheral
venous catheter patency developed a rise in absorbance of more than 0.100 optical
density (OD) units by PF4/polyvinyl sulfonate EIA (GTI, Brookfield, WI). All OD
values were less than the positive adult cutoff (0.400 OD units). No differences were
observed between patients receiving heparin (solid lines) compared to patients re-
ceiving saline (dotted lines). As the maximum OD in saline controls was 0.239, the
0.400 cutoff seems appropriate also for pediatric patients. (Klenner et al., 2003b.)
if HIT antibodies recognize a non–PF4-dependent antigen (Greinacher et

al., 1994) (see Chaps. 6 and 7). Thus, a functional test for HIT should be per-
formed when HIT remains strongly suspected despite a negative EIA. How-
ever, when the pretest probability of HIT is low, a negative antigen test usually
excludes HIT.
VI. THERAPY OF PEDIATRIC HIT
Numerous case reports describe the occurrence of new or recurrent throm-

boembolic events during continued or repeated use of heparin in adult pa-
tients with acute HIT. Further, thrombocytopenia usually persists if heparin
is not stopped. Thus, all heparin should be discontinued in patients strongly
suspected of having HIT, including heparin ‘‘flushes,’’ heparin-coated cath-
eters, and heparin-containing blood products (Severin et al., 2002b) (see
Chap. 13). Because HIT is a prothrombotic (‘‘hypercoagulability’’) state with
high risk of thromboembolic complications, alternative anticoagulation is
usually required after stopping heparin. In adults, danaparoid, lepirudin, and
argatroban have been studied prospectively. For children, experience with
these agents is anecdotal and heterogeneous (Table 2). Danaparoid use has
been reported in 27 patients (with additional aspirin, thrombolysis, or lepiru-
din given in some cases). In two children, danaparoid was stopped because of
cross-reactivity, with further anticoagulation with lepirudin. Twelve patients
received lepirudin (one combined with thrombolysis and oral anticoagulants).
Five children were treated with low molecular weight heparin (LMWH), and
one with warfarin and aspirin. One infant received aspirin, two were given
argatroban plus aspirin, and three were treated with argatroban alone. Eleven
children received oral anticoagulants. In five children, no anticoagulant was
given.
A. Danaparoid
Danaparoid (see Chap. 14) is a mixture of low molecular weight glycos-
aminoglycans that catalyze the inactivation of factor Xa (FXa) by anti-
thrombin. It has minimal anti-factor IIa activity. Dosing schedules for adults
(appropriately weight-adjusted for the child) can be used for guidance. For
antithrombotic prophylaxis, 10 IU/kg body weight given twice daily by sub-
cutaneous injection is recommended. For therapeutic anticoagulation in pe-
diatric HIT patients, an initial i.v. bolus of 30 IU/kg is followed by continuous
infusion of 1.2–2.0 IU/kg/h (Monagle et al., 2001). The anti-FXa level should
be measured during treatment for optimal dosing. Target levels of anti-FXa
activity are 0.4–0.6 IU/mL for standard and 0.5–0.8 IU/mL for higher dana-
paroid doses (Severin et al., 2002b).
Based on experience in adults, LMWH should not be used to treat acute

HIT in children (see Chap. 13).
B. Lepirudin
Lepirudin (see Chap. 15) is a direct inhibitor of free and clot-bound thrombin
through noncovalent, irreversible binding. In adults with HIT complicated by
thrombosis, the approved dose is an initial bolus of 0.4 mg/kg followed by
continuous i.v. infusion (0.15 mg/kg/h) adjusted by activated partial throm-
boplastin time (aPTT). The usual target aPTT ratio should be 1.5–2.5 times
the normal laboratory mean aPTT. Dosing in children is based on anecdotal
experience. Schiffmann et al. (1997) gave a bolus of lepirudin (0.2 mg/kg)
and a continuous infusion (ranging between 0.1 and 0.7 mg/kg/h) adjusted by
aPTT. Severin et al. (2002b) achieved therapeutic anticoagulation with a
continuous infusion of 0.1 mg/kg/h in a 15-year-old boy, and with an infusion
rate of about 0.15 mg/kg/h in an 8-year-old girl. In an 11-year-old girl, 0.15–
0.22 mg/kg/h was given. In a premature infant, Nguyen and coworkers (2003)
gave a 0.2 mg/kg bolus followed initially by 0.1 mg/kg/h infusion rate; the
dose was adjusted daily based on the aPTT, and 0.03–0.05 mg/kg/hr provided
adequate anticoagulation. Since pharmacokinetics depend largely on renal
function, we recommend starting lepirudin with an i.v. infusion of 0.1 mg/
kg/h (if renal function is normal) and to adjust the dose according to aPTT
4 h later, without initial bolus. This minimizes both risk of overdosing and
anaphylaxis (see Chap. 15).
C. Argatroban
Argatroban (see Chap. 16) is a synthetic direct thrombin inhibitor that binds
reversibly to the active site of thrombin. In adults, the recommended initial
dose of argatroban is 2 Ag/kg/min given by continuous i.v. infusion and
adjusted by aPTT (target range, 1.5–3 times the baseline aPTT). Safety and
efficacy of argatroban in pediatric patients have not been established. Ar-
gatroban has been used in neonates (Boshkov et al., 2003a,b). In one new-
born, an initial bolus (200 Ag/kg) was followed by continuous i.v. infusion (7.5
Ag/kg/min). A 5-month-old infant was also treated with an initial bolus (250
Ag/kg) with subsequent infusion (10 Ag/kg/min).
D. Coumarin
Oral anticoagulants of the coumarin class (warfarin, phenprocoumon) are
not appropriate for therapy of acute HIT (see Chap. 13). HIT patients are at
relatively high risk of developing coumarin-induced skin necrosis and venous
limb gangrene syndromes (see Chap. 3). Therefore, coumarin should be
delayed until the patient is adequately anticoagulated with danaparoid, lepi-

rudin, or argatroban, and platelet counts have substantially recovered.
VII. PREVENTION OF HIT IN CHILDREN
Since the pivotal trial in adult orthopedic patients (Warkentin et al., 1995),
it has become clear that LMWH induces HIT less frequently than does UFH.
In children, HIT appears to occur most often among the very young follow-
ing cardiac surgery and among adolescents given UFH to treat spontaneous
thrombosis (Table 2). Data from Pouplard and colleagues (1999) suggest that
HIT also could occur less often if LMWH rather than UFH is used for anti-
thrombotic prophylaxis post–cardiac surgery. This approach should be inves-
tigated in children.
Similarly, in the second group of at-risk pediatric patients (adolescents
with thrombosis), it seems plausible that the frequency of HIT would be
reduced if LMWH is given instead of UFH. Pharmacokinetic studies of
Table 3 Studies on Efficacy and Safety of LMWH in Children

New Bleeding
Number of thromboembolic
Study/Ref. patients event Minor Major Fatal
Therapeutic doses
Massicotte et al., 1996 23 0 0 2 0
Dix et al., 2000 143 2 26 6 1
Dix et al., 1998 9 0 4 0 0
Nohe et al., 1999 38 0 No data 0 0
Massicotte et al., 2003a 36 2 32 2 0
Total 249 4 62 10 1
(1.6%) (24.9%) (4%) (0.4%)
Prophylactic doses
Dix et al., 2000 30 1 2 0 0
Dix et al., 1998 5 0 2 0 0
Broyer et al., 1991 46 1 5 7 0
Nohe et al., 1999 10 0 No data 0 0
Massicotte et al., 1999 47 No data 0 3 0
Andrew et al., 1992 78 11 0 0 0
Elhasid et al., 2001 41 0 0 0 0
Massicotte et al., 2003b 78 11 48 0 0
Total 335 24 57 10 0
(7.2%) (17%) (3%)
LMWH in infants and children have been conducted for enoxaparin and
reviparin. With both drugs, therapeutic anti-FXa levels are achieved 3–4
hours following the first subcutaneous injection and are similar in both infants
and older children (Albisetti and Andrew, 2002; Massicotte et al., 2003c).
The safety and efficacy of prophylactic and therapeutic doses of LMWH
(enoxaparin, reviparin, dalteparin) in children have been evaluated in clinical
trials for a variety of conditions. LMWH is safe and effective for anticoagu-
lation of infants and children of varying age (Table 3) (Roschitz et al., 2002).
Based on adult experience, substituting UFH with LMWH whenever possi-
ble will likely reduce the risk of HIT in children. However, no data exist in
children to support this supposition.
VIII. SUMMARY
HIT appears to be rare in children. The incidence depends somewhat on

patient age and indication for heparin. Two major pediatric at-risk groups
are apparent: newborns/infants after cardiac surgery (incidence f1%), and
adolescents treated with UFH for spontaneous thrombosis. HIT can be life-
threatening even in children (f12% mortality). Venous thrombosis is the
most frequent HIT-associated complication in children. For laboratory con-
firmation of HIT, antigen assays are most appropriate. Although there are
conflicting data on the optimal laboratory cutoff for antigen assays, a ran-
domized, double-blind clinical trial suggests that the cutoff level established
in adults is also appropriate for children. There are no prospective studies
of alternative anticoagulants in children with HIT. Most available data
are for lepirudin and danaparoid. By substituting LMWH for UFH when-
ever possible, the risk of HIT may be similarly reduced in children as it is in
adults.
REFERENCES
Albisetti M, Andrew M. Low molecular weight heparin in children. Eur J Pediatr 2002;
161:71–77.
Andrew M, Vegh P, Johnston M, Bowker J, Ofosu F, Mitchell L. Maturation of the
hemostatic system during childhood. Blood 1992; 80:1998–2005.
Bocquet R, Blanot S, Dautzenberg MD, Pierre-Kahn A, Carli P. Antiphospholipid
antibody syndrome in pediatric neurosurgery: a hemostasis problem. Ann Fr
Anesth Reanim 1999; 18:991–995.
Boon DM, Michiels JJ, Stibbe J, van Vliet HH, Kappers-Klunne MC. Heparin-
induced thrombocytopenia and antithrombotic therapy. Lancet 1994; 344:1296.
Boshkov LK, Ibsen L, Kirby A, Ungerleider R, Shen I. Heparin-induced thrombo-

cytopenia (HIT) in neonates and very young children undergoing congenital
cardiac surgery: a likely under-recognized complication with significant mor-
bidity and mortality: report of 4 sequential cases [abstr]. J Thromb Haemost
2003a; 1(suppl 1):P1494.
Boshkov LK, Ibsen L, Kirby A, Ungerleider R, Shen I. Report of argatroban infusions
for heparin-induced thrombocytopenia (HIT) diagnosed by functional assay in
2 congenital cardiac surgery patients, a neonate and a 5 month old [abstr]. J
Thromb Haemost 2003b; 1(suppl 1):P1495.
Broyer M, Gagnadoux MF, Sierro A, Fischer AM, Niaudet P. Preventive treatment of
vascular thrombosis after kidney transplantation in children with low molecular
weight heparin. Transplant Proc 1991; 23:1384–1385.
Butler TJ, Sodoma LJ, Doski JJ, Cheu HW, Berg ST, Stokes GN, Lancaster KJ.
Heparin-associated thrombocytopenia and thrombosis as the cause of a fatal
thrombus on extracorporeal membrane oxygenation. J Pediatr Surg 1997; 32:
768–771.
Deitcher SR, Topoulos AP, Bartholomew JR, Kichuk-Chrisant MR. Lepirudin
anticoagulation for heparin-induced thrombocytopenia. J Pediatr 2002; 140:
264–266.
Dix D, Marzinotto V, Monagle P, Freedman M, Saunders EF, Doyle J, Calderwood S,
Massicotte MP. Use of low molecular weight heparin (LMWH) for the man-
agement of venous thromboembolic disease (VTE) in children undergoing bone
marrow transplant (BMT) [abstr]. J Pediatr Hematol/Oncol 1998; 20:366.
Dix D, Andrew M, Marzinotto V, Charpentier K, Bridge S, Monagle P, deVeber G,
Leaker M, Chan AK, Massicotte MP. The use of low molecular weight heparin
in pediatric patients: a prospective cohort study. J Pediatr 2000; 136:439–445.
Elhasid R, Lanir N, Sharon R, Weyl BA, Levin C, Postovsky S, Ben Barak A, Brenner
B. Prophylactic therapy with enoxaparin during L-asparaginase treatment in
children with acute lymphoblastic leukemia. Blood Coagul Fibrinolysis 2001;
12:367–370.
Girisch M, Klinge J, Lischetzki G, Buheitel G. In neonates higher doses orgaran may
be needed to achieve effective doses [abstr]. Ann Hematol 2002; 81:A22.
Greinacher A, Amiral J, Dummel V, Vissac A, Kiefel V, Mueller-Eckhardt C. Labo-
ratory diagnosis of heparin-associated thrombocytopenia and comparison of
factor 4/heparin enzyme-linked immunosorbent assay. Transfusion 1994; 34:
381–385.
Hong AP, Cook DJ, Sigouin CS, Warkentin TE. Central venous catheters and upper-
extremity deep-vein thrombosis complicating immune heparin-induced throm-
bocytopenia. Blood 2003; 101:3049–3051.
Klement D, Rammos S, Kries R, Kirschke W, Kniemeyer HW, Greinacher A. Heparin
as a cause of thrombus progression. Heparin-associated thrombocytopenia is an
important differential diagnosis in paediatric patients even with normal platelet
counts. Eur J Pediatr 1996; 155:11–14.
Klenner AF, Fusch C, Varnholt V, Ringe H, Meyer O, Stiller B, Greinacher A. Hep-
arin-induced thrombocytopenia in pediatrics and its therapy—case-report and

review of literature. Monatsschr Kinderheilkd 2003. In press.
Klenner AF, Fusch C, Rakow A, Kadow I, Beyersdorff E, Eichler P, Wander K, Lietz
T, Greinacher A. Benefit and risk of heparin addition in maintaining peripheral
venous catheters in neonates - a prospective, double-blind, placebo-controlled
trial [abstr]. J Thromb Haemost 2003b; 1(suppl 1):OC423.
Massicotte P, Adams M, Marzinotto V, Brooker LA, Andrew M. Low-molecular-
weight heparin in pediatric patients with thrombotic disease: a dose finding
study. J Pediatr 1996; 128:313–318.
Massicotte MP, Marzinotto V, Julian J, Gent M, Shields K, Szechtman B, Chan AKZ,
Andrew M. Dose finding and pharmacokinetics of prophylactic doses of a low
molecular weight heparin (reviparin) in pediatric patients [abstr]. Blood 1999;
94:27a.
Massicotte P, Julian JA, Gent M, Shields K, Marzinotto V, Szechtman B, Andrew M.
An open-label randomized controlled trial of low molecular weight heparin
compared to heparin and coumadin for the treatment of venous thromboem-
bolic events in children: the REVIVE trial. Thromb Res 2003a; 109:85–92.
Massicotte P, Julian JA, Gent M, Shields K, Marzinotto V, Szechtman B, Chan
AK, Andrew M. An open-label randomized controlled trial of low molecular
weight heparin for the prevention of central venous line-related thrombotic
complications in children: the PROTEKT trial. Thromb Res 2003b; 109:101–
108.
Massicotte P, Julian JA, Marzinotto V, Gent M, Shields K, Chan AK, Szechtman B,
Kohne S, Shepherd S, Bacher P, Andrew M. Dose-finding and pharmacokinetic
profiles of prophylactic doses of a low molecular weight heparin (reviparin-
sodium) in pediatric patients. Thromb Res 2003c; 109:93–99.
Monagle P, Michelson AD, Bovill E, Andrew M. Antithrombotic therapy in children.
Chest 2001; 119:344S–370S.
Murdoch IA, Beattie RM, Silver DM. Heparin-induced thrombocytopenia in
children. Acta Paediatr 1993; 82:495–497.
Neuhaus TJ, Goetschel P, Schmugge M, Leumann E. Heparin-induced thrombocy-
topenia type II on hemodialysis: switch to danaparoid. Pediatr Nephrol 2000;
14:713–716.
Newall F, Barnes C, Ignjatovic V, Monagle P. Heparin-induced thrombocytopenia in
children. J Paediatr Child Health 2003; 39:289–292.
Nguyen TN, Gal P, Ransom JL, Carlos R. Lepirudin use in a neonate with heparin-
Nohe N, Flemmer A, Rumler R, Praun M, Auberger K. The low molecular weight
heparin dalteparin for prophylaxis and therapy of thrombosis in childhood: a
report on 48 cases. Eur J Pediatr 1999; 158(suppl 3):S134–S139.
Oriot D, Wolf M, Wood C, Brun P, Sidi D, Devictor D, Tchernia G, Huault G. Severe
thrombocytopenia induced by heparin in an infant with acute myocarditis. Arch
Fr Pediatr 1990; 47:357–359.
Porcelli R, Moskowitz BC, Cetta F, Graham LC, Godwin JE, Eidem BW, Prechel
MM, Walenga JM. Heparin-induced thrombocytopenia with associated
thrombosis in children after the Fontan operation: report of two cases. Tex
Heart Inst J 2003; 30:58–61.
child. J Pediatr 1992; 121:135–138.
Antibodies to platelet factor 4-heparin after cardiopulmonary bypass in patients
anticoagulated with unfractionated heparin or a low-molecular-weight heparin:
clinical implications for heparin-induced thrombocytopenia. Circulation 1999;
99:2530–2536.
Ranze O, Rakow A, Ranze P, Eichler P, Greinacher A, Fusch C. Low-dose danaparoid
sodium catheter flushes in an intensive care infant suffering from heparin-
induced thrombocytopenia. Pediatr Crit Care Med 2001; 2:175–177.
Risch L, Fischer JE, Schmugge M, Huber AR. Association of anti-heparin platelet
factor 4 antibody levels and thrombosis in pediatric intensive care patients with-
out thrombocytopenia. Blood Coagul Fibrinolysis 2003; 14:113–116.
Roschitz B, Sudi K, Beitzke A, Gamillscheg A, Leschnik B, Muntean W. Sub-
cutaneous low molecular weight heparin (LMWH) versus intravenous unfrac-
tionated heparin (UFH) bolus in pediatric cardiac catheterization [abstr]. Ann
Hematol 2002; 81:A69.
Sauer M, Gruhn B, Fuchs D, Altermann W, Zintl F. Heparin-induced type II
thrombocytopenia within the scope of high dose chemotherapy with subsequent
stem cell rescue. Klin Padiatr 1998; 210:102–105.
Saxon BR, Black MD, Edgell D, Noel D, Leaker MT. Pediatric heparin-induced
thrombocytopenia: management with Danaparoid (orgaran). Ann Thorac Surg
1999; 68:1076–1078.
Schiffmann H, Unterhalt M, Harms K, Figulla HR, Völpel H, Greinacher A. Suc-
cessful treatment of heparin-induced thrombocytopenia type II in childhood
with recombinant hirudin. Monatssch Kinderheilk 1997; 145:606–612.
Schmugge M, Risch L, Huber AR, Benn A, Fischer JE. Heparin-induced throm-
bocytopenia-associated thrombosis in pediatric intensive care patients. Pedia-
trics 2002; 109:E10.
Severin T, Sutor AH. Heparin-induced thrombocytopenia in pediatrics. Semin
Thromb Hemost 2001; 27:293–299.
Severin T, Dittrich S, Zieger B, Kampermann J, Kececioglu D, Sutor AH. HIT II after
Fontan procedure—treatment with Lepirudin [abstr]. Ann Hematol 2002a; 81:
A77.
Severin T, Zieger B, Sutor AH. Anticoagulation with recombinant hirudin and
danaparoid sodium in pediatric patients. Semin Thromb Hemost 2002b; 28:447–
454.
Spadone D, Clark F, James E, Laster J, Hoch J, Silver D. Heparin-induced throm-
bocytopenia in the newborn. J Vasc Surg 1992; 15:306–311.

Weigel B, Laky A, Krishnamurti L. Danaparoid (Orgaran) anticoagulation of pedi-
atric patients with heparin-induced thrombocytopenia (HIT) [abstr]. J Pediatr
Hematol/Oncol 1999; 21:327.
Wilhelm MJ, Schmid C, Kececioglu D, Mollhoff T, Ostermann H, Scheld HH. Car-
diopulmonary bypass in patients with heparin-induced thrombocytopenia using
Org 10172. Ann Thorac Surg 1996; 61:920–924.
Zöhrer B, Zenz W, Rettenbacher A, Covi P, Kurnik K, Kroll H, Grubbauer HM,
Muntean W. Danaparoid sodium (Orgaran) in four children with heparin-
induced thrombocytopenia type II. Acta Paediatr 2001; 90:765–771.
21
Legal Aspects of Heparin-Induced
Thrombocytopenia: U.S.
Perspectives
Kevin M. McIntyre
Harvard Medical School, Boston VA Health Care System, and Brigham and
Women’s Hospital, Boston, Massachusetts, U.S.A.
I. INTRODUCTION
Litigation is a not uncommon occurrence when a patient has suffered harm

because of heparin-induced thrombocytopenia (HIT) due to alleged physician
or nursing ‘‘malpractice’’ (more properly termed professional negligence).
One reason may be the general increase in medicolegal actions (Mello et al.,
2003) including those involving issues of thrombosis and antithrombotic
therapy, especially in the United States (McIntyre, 2001). Regarding HIT in
particular, one explanation could be the legal concept of res ipsa loquitur (‘the
thing speaks for itself’), i.e., the situation of a patient who enters the hospital
for elective surgery and leaves with an amputated limb may itself ‘‘speak’’
persuasively to a jury that someone somehow must have been at fault for such
a tragic outcome to have occurred (Kelton, 1998). Further, the patient may
573
574 McIntyre and Warkentin
have been unaware of such a potential catastrophic outcome, as physicians

usually do not specifically include HIT and its attendant complications when
obtaining informed consent. The tendency for a finding of substandard care to
occur more frequently when harm is severe (‘‘outcome bias’’) exists even
though outcome per se is a poor index of blameworthiness (Caplan et al., 1991;
Runciman et al., 2003).
Another factor is that HIT is a ‘‘common, rare disease’’ (see p. 338),
suggesting that although many thousands of patients probably suffer HIT-
related thrombotic complications each year in the United States, few physi-
cians encounter enough patients to develop proficiency in its diagnosis and
management. This is of particular relevance to HIT, where the use of stan-
dard, common-sense treatment approaches, such as discontinuing heparin
and commencing oral anticoagulants, can paradoxically lead to catastrophic
outcomes, e.g., warfarin-induced venous limb gangrene (see Chaps. 3 and 13).
Further, there is an emerging consensus that appropriate medical treatment
for HIT should include use of an alternative agent such as danaparoid, lepi-
rudin, or argatroban (see Chaps. 13–19). However, many physicians are rela-
tively unfamiliar with these newer drugs, some of which might not even be
stocked in the hospital pharmacy (and thus not be quickly available), or for
which special monitoring assays required in special circumstances may not be
available (e.g., anti-factor Xa level; ecarin clotting time) (see Chaps. 14 and 15).
Few medical centers have access to timely diagnostic testing for the
pathogenic HIT antibodies. Further, there exists a variety of assays, with
varying diagnostic usefulness (see Chap. 11). Plus, even a positive test for HIT
does not necessarily establish the diagnosis, given the high frequency of
subclinical HIT antibody seroconversion in certain patient populations (see
Chaps. 4 and 11). Physicians often must make crucial decisions to give or
withhold anticoagulants amidst considerable diagnostic uncertainty.
Thrombocytopenia is a very common event in hospitalized patients, and
it is often inconsequential (e.g., perioperative hemodilution) or otherwise
expected in the context of the patient’s presenting illness (e.g., septicemia) or
its management (e.g., hemodilution/platelet consumption secondary to intra-
aortic balloon pump or heart surgery utilizing cardiopulmonary bypass). Yet,
in a patient who develops HIT, often in the setting of these other potential
explanations for thrombocytopenia, the initial mild or moderate platelet
count decline that first signaled the onset of HIT could very well be over-
looked, but will later be highlighted during the glaring retrospection of a
medicolegal action.
The purpose of this chapter is to discuss the medicolegal impact of
clinical practice guidelines, and to discuss some specific issues that can arise in
HIT from the perspective of the U.S. medicolegal system. Chapter 22
discusses medicolegal aspects of HIT from a European viewpoint.
U.S. Perspectives on Legal Aspects of HIT 575
II. U.S. MEDICOLEGAL SYSTEM
The key issue in a medicolegal case is whether the physician failed to uphold
the perceived ‘‘standard of care’’ in the context of the facts presented and
expert opinions expressed; and whether such failure, if established, was a
substantial contributing factor in bringing about the plaintiff’s injury. Unless
the legal action is dropped, or an out-of-court settlement is reached, the action
generally progresses to the courtroom (with or without prior judicially
overseen medical review, in some states), where it usually is heard and decided
by a jury.
The ‘‘standard of care,’’ as it applies to a malpractice case, generally
refers to how a qualified and reasonably competent practitioner of the same
generalist or specialist class would perform acting under the same or similar
circumstances. The standard of care is developed for the jury through the
interaction of expert witnesses and the lawyers for plaintiff and defendant
sides (McIntyre, 2001). Depending upon the state, this interaction can consist
of experts’ opinions expressed through written reports, and/or as testimony
expressed in response to questioning that primarily is conducted by the
opposing attorney (deposition).
Given the progressively accumulating scientific data base, there is
controversy as to whether a jury system is appropriate for ever more complex
medical issues (McIntyre, 2001; Mello et al., 2003). The jury system may
function reasonably well when the failure of the physician is clear (e.g.,
removal of the wrong leg). But as the requirement for judgment on the part of
the physician increases, especially in a disorder such as HIT with complex
diagnostic and treatment dilemmas, the jury may increasingly find the medical
issues in particular difficult to understand. Even for interested physicians and
scientists, it can be challenging to follow the exponential expansion of the
scientific literature on HIT, with its implications for an evolving standard of
care based upon this new information. The standard of care to be applied to
an individual case must consider the available information at the time the
treatment was applied, which can differ significantly from the standard at the
time of the subsequent medicolegal action itself. Further, diagnostic uncer-
tainty in specific cases of HIT, as discussed earlier, can be considerable, with
the implication that the standard of care to be applied to that case becomes
increasingly ill-defined and uncertain as well.
While the summary of the medicolegal process put forth above tends to
make the U.S. system appear quite fair and orderly, there are those who
believe the process has serious flaws. On the medical side, the increasing
emphasis on scientifically sound evidence as the basis for the development of a
‘‘standard of care’’ for medical practice may have improved the quality of the
medical care in the United States, as judged by such entities as Physician
Review and Evaluation Boards. However, treatment guidelines developed in

part by the consensus conference process have not resulted in either a decline
in the number of malpractice cases or the size of dollar awards to plaintiffs,
awards that at times appear to be massively inappropriate and costly to a
medical system currently staggering under cost burdens.
A principal reason for this cited repeatedly by critics is the alleged
financial motivation of plaintiffs’ attorneys, who may receive one-third of a
multimillion dollar award. These large fees reflect the ‘‘contingency’’ system,
whereby the legal fees and costs are based upon a percentage of the gain
realized for the client. Such contingency fees make access to legal services
available even to an indigent plaintiff, but tend to emphasize legal actions with
the potential for large awards. Attempts at curtailing huge awards have so far
failed, and the cost burden reflected in the form of rising insurance costs is
ultimately passed on to consumers or their employers.
Another aspect of the medical malpractice system in the United States
appears to be the willingness of some expert witnesses to do or say whatever is
necessary to win the case for the client on whose behalf they are testifying. The
Honorable Richard Thornburgh, former U.S. Attorney General under
President Ronald Reagan, and frequent commentator on legal aspects of
important social problems, has used the term ‘‘junk science’’ to apply to the
‘‘testimony of expert witnesses hired not for their scientific expertise, but for
their willingness, for a price, to say whatever is needed to make the client’s
case’’ (Thornburgh, 1998). The expert witness in a malpractice action is often
well paid, perhaps receiving several thousand dollars for a few hours work,
and it is difficult not to believe that, at least in some circumstances, it is that
expert’s preference that the client on behalf of whom he is testifying would go
on to win the case.
In practice, it is the role of the jury to determine whether or not the
practitioner’s actions met the requirements of the ‘‘standard of care.’’ An
expert witness with impressive credentials and a striking presence can have a
powerful effect on the jury’s decision, by simply indicating that, in his opinion,
the practitioner failed to meet the required ‘‘standard of care.’’ This effect may
be disproportionately increased in a complex case involving considerable
physician judgment.
The ‘‘burden of proof’’ required for the plaintiff to prevail in a
malpractice action is an important additional determinant of outcome. Unlike
criminal cases, where the jury must be convinced ‘‘beyond a reasonable
doubt’’ (i.e., theoretically, at least, to a 99.999% level of certainty) that the
individual on trial is guilty, in a medical malpractice trial, the ‘‘burden of
proof’’ is much lower, requiring only that the jury determine that the phy-
sician, based on ‘‘a preponderance of the evidence,’’ i.e., ‘‘more likely than
not,’’ breached the standard of care; and that such breach was a ‘‘substantial
factor’’ (or alternate terminology, depending upon the state), in causing the
damages or injuries alleged by the plaintiff. Depending upon the jurisdiction,
the experts need to express their opinions ‘‘based on a reasonable degree
of medical probability’’ or ‘‘based upon a reasonable degree of medical cer-
tainty’’ or similar such wording.
III. CLINICAL PRACTICE GUIDELINES AND HIT
There are at least seven clinical practice guidelines that discuss one or more
aspects of platelet count surveillance, diagnosis, or management of HIT,
including the guidelines summarized in this book (Olson et al., 1998; Hirsh et
al., 1998, 2001; Greinacher and Warkentin, 2000, 2001; Warkentin, 2002;
Greinacher et al., 2003; see Chap. 13). Although the major purpose of clinical
practice guidelines is to enhance quality of care, it is also possible that their
existence could contribute to medicolegal risk if recommendations are not
followed (McIntyre, 2001). Of course, it is not possible for any consensus
conference or practice guideline to deal explicitly with all of the complex
issues that can arise in any individual case. Thus, there is a major role for
physician judgment in the context of the individual case itself (McIntyre,
2001). Additionally, as Olson (1995) states, ‘‘Readers themselves must assess
the quality and validity of consensus statements as they do all literature.’’
Finally, it remains unproven whether clinical practice guidelines have a
positive impact on patient care (McIntyre, 2001).
Clinical practice guidelines also can demonstrate that the scientific basis
for many recommendations is preliminary, incomplete, or even contradictory.
Increasingly, attempts are being made formally to grade the strength of the
recommendations. The increasing sophistication of the grading systems has
paralleled the growing scope of the recommendations themselves, as shown
by the evolution of the grading systems used by the successive Consensus
Conferences on Antithrombotic Therapy held under the aegis of the Amer-
ican College of Chest Physicians (Sackett, 1989; Cook et al., 1992, 1995;
Guyatt et al., 1998, 2001). The significance of the strength of a recommen-
dation is that a strong scientific basis theoretically would render more difficult
the defense of a deviation of practice from a guideline based upon such strong
evidence (McIntyre, 2001).
Regarding HIT, most of the recommendations are based upon obser-
vational data (level C evidence), rather than randomized clinical trials that
might lead to higher grade recommendations. Nevertheless, it is possible for a
level 1 recommendation to be derived from observational data, provided that
the evidence is convincing and widely accepted.
IV. SPECIFIC MEDICOLEGAL ISSUES IN HIT

A. Informed Consent
HIT often occurs in patients who receive heparin for ‘‘routine’’ antithrom-
botic prophylaxis. For example, the highest frequency of HIT (3–5%) has
been reported in postoperative orthopedic patients receiving unfractionated
heparin (UFH) (Warkentin et al., 1995, 2003; see Chap. 4). HIT also is fairly
common (1–3%) after cardiac or vascular surgery if postoperative UFH pro-
phylaxis is also given. In these situations, which often involve elective surgery,
the process of informed consent routinely includes important perioperative
and postoperative complications, including severe consequences such as
death or disability from thrombosis, bleeding, infection from blood transfu-
sion, etc. However, in our experience, it appears that most physicians do not
specifically list thrombosis secondary to HIT when obtaining consent. One
obvious explanation is that there is something inherently contradictory about
informing a patient that heparin can sometimes cause severe clots, while at the
same time educating the patient that the heparin is prescribed to prevent clots.
Thus, the argument raised is that the overall ‘‘net’’ effect of heparin is
prevention of thrombosis, and so the issue of HIT is irrelevant. Another fac-
tor is that in most jurisdictions, there is no specific requirement to obtain
consent for an approved medication, although specific informed consent is
required in the context of an experimental therapy involving an unapproved
agent. There is no consensus in the medical community whether informed
consent should include specific mention of HIT and its complications.
B. Platelet Count Surveillance

It is a truism that for a thrombocytopenic disorder the platelet count values,
and the interpretation of these results, take on a central role. Few guidelines
exist, however, as to how frequently platelet counts should be measured in
patients receiving heparin. A recommendation from the College of American
Pathologists (CAP) Conference XXXI stated, ‘‘In patients receiving adjusted-
dose and full-dose heparin, monitor for and evaluate HIT by performing
platelet counts pretreatment and, at least, on alternate days for 14 days
beginning on day 4 of therapy in the naive patient and beginning on day 1 in
patients with prior heparin exposure’’ (Olson et al., 1998).
Recent data, however, suggest that these recommendations should be
reconsidered. For example, HIT occurs most often in postoperative ortho-
pedic patients receiving nonadjusted low- or intermediate-dose unfraction-
ated heparin (10,000–15,000 U per day in divided doses) (Warkentin et al.,
1995, 2000, 2003; Ganzer et al., 1997; Funk et al., 2000). Presumably, at least
alternate-day monitoring should apply to these patients as well. Conversely,
HIT appears to be very rare in medical patients receiving therapeutic-dose low

molecular weight heparin (LMWH); perhaps, platelet count monitoring—
especially in an out-patient setting—is not warranted. Also, the role of prior
heparin exposure in the timing of onset of HIT has recently been clarified: a
rapid fall in platelet count seems to occur only in patients who have recently
been exposed to heparin (within the past 100 days) (Warkentin and Kelton,
2001). Thus, for a patient with ‘‘remote’’ heparin exposure, platelet count
monitoring for HIT presumably could begin on day 4 of heparin treatment.
This information has been incorporated into new recommendations regarding
platelet count surveillance that takes into account the risk of HIT in various
clinical situations (Warkentin, 2002; Greinacher et al., 2003; see also Chap. 4).
Another central issue in some medicolegal cases involving HIT is the
interpretative response of the physician to a particular platelet count profile.
Thrombocytopenia is very common in hospitalized patients, with only a small
minority of thrombocytopenic patients receiving heparin actually having
HIT. Unfortunately, the severity of the platelet count fall is usually not help-
ful in distinguishing HIT from other medical problems. For instance, about
10–15% of patients with serologically proven HIT never develop ‘‘thrombo-
cytopenia,’’ as conventionally defined (platelet fall to less than 150 109/L or
by > 50%). Further, the median platelet count nadir in HIT is about 60
109/L, and there are many other illnesses that cause such a moderate degree of
thrombocytopenia. Although a 50% or greater fall in the platelet count from
the postoperative peak is a fairly sensitive and specific marker for HIT in
surgical patients (Warkentin et al., 2003), other conditions (e.g., septicemia)
also can cause a similar platelet count decline. Regardless of the scientific
‘‘definition’’ of thrombocytopenia applicable to HIT, there is the potential for
considerable disagreement in a medicolegal action as to whether a particular
platelet count sequence should have been considered indicative of HIT, and at
what point in time along the platelet count sequence this possibility should
have been considered.
C. Diagnosis of HIT
Although many cases of HIT can be diagnosed with accuracy on clinical
grounds, there are others in which the diagnosis is not certain, and laboratory
confirmation or refutation is important. Moreover, availability of laboratory
testing is variable and may not have been obtained in a patient, or may have
given results that are at odds with the clinical diagnosis. For example, some
tests have limited sensitivity (e.g., platelet aggregation assays), so a negative
test does not necessarily exclude HIT. Further, the transience of HIT anti-
bodies means that the diagnosis cannot usually be established even when
sensitive tests are performed using blood samples obtained several months
after recovery from acute HIT (see Chap. 3). As discussed in Chap. 12, some
disorders so closely mimic HIT on clinical grounds as to merit the designa-
tion, ‘‘pseudo-HIT.’’ These disorders can cause ‘‘experts’’ on opposite sides
to disagree fundamentally on whether HIT was even present. Ironically,
certain of the pseudo-HIT disorders also have a high risk for thrombosis or
fatal outcomes. Issues of turnaround time and diagnostic usefulness of
particular assays can also be an issue in some medicolegal actions, given the
considerable heterogeneity in availability and type of diagnostic testing
among various medical centers, as well as the evolving medical literature on
the diagnostic implications of a positive or negative test result (see Chap. 11).
D. Treatment of HIT
Medicolegal issues involving the treatment of HIT are complicated by such
issues as treatment paradoxes, use of nonapproved medications, new infor-
mation about the natural history of isolated HIT, as well as the recent
availability of new drugs for prevention and treatment of thrombosis
complicating HIT, all of which have influenced the standard of care.
Treatment ‘‘Paradoxes’’
There are many seemingly counterintuitive treatment paradoxes in the
management of HIT (Warkentin, 2001; see Table 1 in Chap. 13). Many of
these paradoxes are known to specialists, although a generalist practitioner
may not be familiar with them. For example, prophylactic platelet trans-
fusions are considered to be relatively contraindicated in HIT, even when a
patient is severely thrombocytopenic. This is because bleeding is rare in HIT,
and because platelet transfusions at least theoretically might increase the risk
for thrombosis. Thus, a platelet transfusion ordered for a patient with HIT
who suffers a subsequent thrombosis might be regarded as prima facie (‘‘at
first view’’) evidence for malpractice. However, in a complex case of HIT,
mitigating factors could include uncertainty about the cause of thrombocy-
topenia at the time of transfusion, and other factors suggesting a high risk for
bleeding. Further, although published anecdotes suggest a link between
platelet transfusions and subsequent thrombosis, the growing awareness that
the natural history of HIT itself is for thrombosis to occur frequently even in
the absence of platelet transfusions suggests that any causal association
between a platelet transfusion and a subsequent adverse outcome is specula-
tive. Indeed, this view is reflected in the ‘‘weakest’’ grade (level 2C) assigned
this recommendation proscribing platelet transfusions.
Another example of a counterintuitive treatment paradox is the recom-
mendation that LMWH is contraindicated as treatment for HIT (Hirsh et al.,
2001; see Chap. 13). This recommendation reflects the observation that even
though using LMWH will prevent most cases of HIT (Warkentin et al., 1995,
2000, 2003), use of these preparations in a patient who already has established
HIT has a high probability of treatment failure (Ranze et al., 2000). Although
this recommendation against use of LMWH for treating HIT was given a rela-
tively ‘‘high-grade’’ recommendation by the ACCP Consensus Conference
(1C+), an alternate view holds that LMWH therapy might be acceptable
provided that in vitro cross-reactivity testing is negative using platelet aggre-
gation assays (Slocum et al., 1996). However, problems with such an approach
involve treatment delays pending laboratory testing, as well as the availability
of other agents with low (danaparoid) or absent cross-reactivity (argatroban,
lepirudin). Nevertheless, this contrary viewpoint does show that expert opin-
ion potentially can differ substantially regarding treatment recommendations.
Nonapproved Medications
Although three drugs are widely regarded as safe and effective for managing
HIT, differences exist in their approval status among different countries. For
example, although danaparoid is approved in the United States for preven-
tion of deep vein thrombosis following orthopedic surgery, it does not have a
specific indication in the United States for prevention or treatment of HIT-
associated thrombosis. Nevertheless, its former availability on the U.S.
market meant that physicians had the legal right to use the drug for ‘‘off-
label’’ treatment of HIT (Preuss and Conour, 1999): According to the U.S.
Food and Drug Administration (FDA, 1982), ‘‘accepted medical practice
often includes drug use that is not reflected in approved labeling.’’ Danapa-
roid was withdrawn from the U.S. market by the manufacturer in April 2002
(although it remains available in some other countries [see Chap. 14]).
Another example of a drug that has been used ‘‘off-label’’ to treat HIT is
bivalirudin, a hirudin derivative approved in the U.S. and Canada for anti-
coagulation during percutaneous coronary interventions (Francis et al., 2003;
see Chap. 17). For theoretical reasons, it seems likely that the pentasaccharide
anticoagulant, fondaparinux, will be effective for treating HIT, and it is pos-
sible that this agent (now approved in the U.S., Canada, European Union,
and elsewhere for antithrombotic prophylaxis in certain orthopedic settings)
will see increasing ‘‘off-label’’ use for treating or preventing thrombosis in
HIT in the coming years (see also Chaps. 4 and 8).
Argatroban is the only drug currently approved in the United States for
treatment and prevention of thrombosis in patients with HIT (lepirudin
having been approved in the United States only for treatment of thrombosis
complicating HIT). However, protocols do exist for ‘‘off-label’’ treatment of
HIT using either danaparoid or lepirudin for isolated HIT; in the case of a
patient with liver impairment, for example, where use of argatroban is

relatively contraindicated, the non-approved treatment with danaparoid or
lepirudin for a patient with isolated HIT might be safer (see Chaps. 14 and 15).
A further consideration is that previous successful experience with a partic-
ular agent could favor its use again by a physician who may be reluctant to try
a newer drug, with its requisite ‘‘learning curve.’’ In the event that a physician
chooses an ‘‘off-label’’ therapy for HIT, she should be prepared to justify the
treatment chosen.
An Evolving Standard of Care

Advances in scientific knowledge, of course, can and do have dramatic impact
upon the standard of care. This is well illustrated in HIT, in which an
emerging consensus regarding diagnosis and treatment began only within
the past few years (Warkentin et al., 1998). For example, we are aware of legal
cases in which expert opinion held that earlier use of coumarin anticoagula-
tion may have prevented limb loss ‘‘within a reasonable degree of medical
probability’’; ironically, however, subsequent research showed that it actually
had been the coumarin anticoagulant that explained the patient’s limb loss,
via the syndrome of warfarin-induced venous limb gangrene (Warkentin et
al., 1997, 1999). Another example: an expert argued successfully that the
patient’s severe episode of venous thromboembolism was not related to HIT,
since HIT was only associated with arterial thrombosis. Again, new data
emerged in the mid-1990s showing that not only did the spectrum of
thrombosis in HIT include venous thrombosis, but that it actually predomi-
nated over arterial thrombosis (see Chap. 3).
Currently, an important issue in HIT that is undergoing a major
conceptual shift is that of treatment of ‘‘isolated HIT’’, i.e., HIT recognized
on the basis of thrombocytopenia alone. Until recently, it was assumed that
prompt cessation of heparin would avoid most complications of HIT.
However, since 1996, it has become increasingly evident that patients with
serologically confirmed HIT have a high risk for subsequent thrombosis
despite stopping heparin, with thrombotic event rates as high as 10% at two-
day, 40% at one-week, and 50% at one-month follow-up (Warkentin and
Kelton, 1996; Wallis et al., 1999; Warkentin, 2003; see Chaps. 4 and 13).
Indeed, the study by Wallis and colleagues (1999) suggests that the highest
risk for thrombosis might paradoxically occur in those patients in whom
heparin is discontinued relatively soon after onset of thrombocytopenia.
This growing awareness of the unfavorable natural history of isolated
HIT has led to new guidelines recommending that physicians consider using
alternative anticoagulation to these patients (Hirsh et al., 1998, 2001; War-
kentin and Greinacher, 2004; see Chaps. 13–16). Also, in November, 2000, the
U.S. Food and Drug Administration approved a new anticoagulant, arga-

troban, for the treatment of thrombosis in patients with HIT-associated
thrombosis, as well as for the prevention of thrombosis in patients with isolated
HIT. Does a new regulatory approval based upon this scientific context now
signal a new standard of care? Or should sufficient time elapse for it to become
apparent that adoption of this practice has become widespread?
V. SUMMARY
This chapter has summarized some of the medicolegal implications of HIT.

The rapid progress of scientific information relating to HIT, as discussed in
this book, and the implications of these scientific advances for the evolution of
the corresponding standard of care indicate the importance for physicians
who use heparin to be familiar with issues of diagnosis and treatment of this
adverse effect of heparin.
REFERENCES
Caplan RA, Posner KL, Cheney FW. Effect of outcome on physician judgments of
appropriateness of care. JAMA 1991; 265:1957–1960.
Cook DJ, Guyatt GH, Laupacis A, Sackett DL. Rules of evidence and clinical
recommendations on the use of antithrombotic agents. Chest 1992; 102(suppl):
305S–311S [published erratum appears in Chest 1994; 105:647].
Cook DJ, Guyatt GH, Laupacis A, Sackett DL, Goldberg RJ. Clinical recommenda-
tions using levels of evidence for antithrombotic agents. Chest 1995; 108(suppl):
227S–230S.
FDA Drug Bulletin 1982; 12:4–5.
Francis JL, Drexler A, Gwyn G, Moroose R. Bivalirudin, a direct thrombin inhibitor,
most 2003; 1(suppl 1):P1909.
Funk S, Eichler P, Albrecht D, Ganzer D, Strobel U, Lubenow N, Greinacher A.
Heparin-induced thrombocytopenia (HIT) in orthopedic patients: a prospective
cohort trial comparing UFH and LMWH [abstr]. Ann Hematol 2000; 79(suppl
I):A92.
laxe als auslöser thromboembolischer Komplicationen. Eine Untersuchung zur
inzidenz der Heparin-induzierten Thrombozytopenie (HIT) Typ II. Z Orthop
1997; 135:543–549.
Greinacher A, Warkentin TE. Treatment of heparin-induced thrombocytopenia: an
overview. In: Warkentin TE, Greinacher A, eds. Heparin-Induced Thrombo-
cytopenia. New York: Marcel Dekker, Inc., 2000:261–290.
Greinacher A, Warkentin TE. Treatment of heparin-induced thrombocytopenia: an
overview. In: Warkentin TE, Greinacher A, eds. Heparin-Induced Thrombo-

cytopenia. 2d ed. New York: Marcel Dekker, Inc., 2001:291–322.
Greinacher A, Lubenow N, Hinz P, Ekkernkamp A. Heparininduzierte Thrombo-
zytopenie. Dtsch Arztebl 2003; 100:A2220–A2229.
Guyatt GH, Cook DJ, Sackett DL, Eckman M, Pauker S. Grades of recommendation
for antithrombotic agents. Chest 1998; 114(suppl):441S–444S.
Guyatt G, Schunëmann H, Cook D, Jaeschke R, Pauker S, Bucher H. Grades of
recommendation for antithrombotic agents. Chest 2001; 119(suppl):3S–7S.
low-molecular-weight heparin. Mechanisms of action, pharmacokinetics, dos-
ing considerations, monitoring, efficacy, and safety. Chest 1998; 114(suppl):
489S–510S.
Granger C, Ohman EM, Dalen JE. Heparin and low-molecular-weight heparin.
Kelton JG. Dealing with drug-induced thrombocytopenia. Hosp Pract (Off Ed) 1998;
33:37–38, 41–43, 47–48, 54–55, 59–60.
McIntyre K. Medicolegal implications of the consensus conference. Chest 2001;
119(suppl):337S–343S.
Mello MM, Studdert DM, Brennan TA. The new medical malpractice crisis. N Engl J
Med 2003; 348:2281–2284.
Olson CM. Consensus statements: applying structure. JAMA 1995; 273:72–73.
Olson JD, Arkin CF, Brandt JT, Cunningham MT, Giles A, Koepke JA, Witte DL.
College of American Pathologists Conference XXXI on Laboratory Monitor-
ing of Anticoagulant Therapy. Laboratory monitoring of unfractionated hep-
arin therapy. Arch Pathol Lab Med 1998; 122:782–798.
Preuss CF, Conour KP. ‘‘Off-label’’ use: the misuse is in the claim. For the Defense,
Dec 1999; 27 ff
Ranze O, Eichner A, Lubenow N, Kempf R, Greinacher A. The use of low-molecular-
weight heparins in heparin-induced thrombocytopenia (HIT): a cohort study
[abstr]. Ann Hematol 2000; 79(suppl 1):198.
Runciman WB, Merry AF, Tito F. Error, blame, and the law in health care-an anti-
podean perspective. Ann Intern Med 2003; 138:974–979.
Sackett DL. Rules of evidence and clinical recommendations on the use of anti-
thrombotic agents. Chest 1989; 95(suppl 2):2S–4S.
patients with heparin-induced thrombocytopenia. J Vasc Surg 1996; 23:839–843.
Thornburgh R. Junk science: the lawyer’s ethical responsibilities. Fordham Urban
Law 1998; XXV:449–469.
Med 1999; 106:629–635.
thrombocytopenia: recommendations of the College of American Pathologists.
Arch Pathol Lab Med 2002; 126:1415–1423.
Br J Haematol 2003; 121:535–555.
Warkentin TE, Greinacher A. Heparin-induced thrombocytopenia: recognition, treat-
ment, and prevention. Chest 2004. In press.
Am J Med 1996; 101:502–507.
N Engl J Med 2001; 344:1286–1292.
Warkentin TE, Sikov WM, Lillicrap DP. Multicentic warfarin-induced skin necrosis
complicating heparin-induced thrombocytopenia. Am J Hematol 1999; 62:44–
48.
Warkentin TE, Sheppard JI, Horsewood P, Simpson PJ, Moore JC, Kelton JG. Im-
pact of the patient population on the risk for heparin-induced thrombocytope-
nia. Blood 2000; 96:1703–1708.
22
Legal Aspects of Heparin-Induced
Thrombocytopenia: European
Perspectives
Klaus Ulsenheimer
University of Munich, Munich, Germany
I. RISK-BENEFIT ASSESSMENT FOR ANTITHROMBOTIC

PROPHYLAXIS
Antithrombotic prophylaxis has become virtually routine in certain perioper-

ative settings, especially for hospital inpatients judged to be at medium or high
risk for thrombosis. In low-risk situations, such as minor operations in
healthy outpatients, the need for antithrombotic prophylaxis is not uniformly
well established. According to the guidelines of 18 surgical societies (AWMF-
Leitlinien, 2003):
In case of leg trauma, the grade of measures for thrombosis prophylaxis
depends on the severity of the trauma, the degree of immobilization, the
duration for healing, and on individual risk factors for thrombosis. For
each patient immobilized by trauma, the measures for thrombosis pro-
phylaxis have to be adjusted individually based on a benefit (prophylaxis
of thrombosis) to risk (bleeding, HIT) consideration. To date, early mo-
bilization is most effective. In patients with recent trauma and/or lower
limb surgery involving immobilization of a joint, pharmacologic throm-
bosis prophylaxis is recommended, especially in case of additional risk
factors for thrombosis.
Prophylaxis against thrombosis thus involves a medical decision involving

risk-benefit assessment for an individual patient.
587
588 Ulsenheimer
II. CHOICE OF ANTITHROMBOTIC PROPHYLAXIS
The choice of an antithrombotic prophylactic agent leads to the principle of

freedom of choice of therapy, meaning that it is primarily the right of the
physician to decide on the treatment (Bundesgerichtshof, 1982). This right is
limited by the duty to choose the treatment that, benefit being equal, carries
the lowest risk for the patient. Choosing a higher-risk agent has to be justified
by special circumstances in individual cases or a higher chance of successful
treatment (Bundesgerichtshof, 1987). Thus, the doctor violates his or her duty
of care if choice of prophylaxis increases patient risk without medical
justification.
These considerations have implications for choice of prophylaxis for
heparin-induced thrombocytopenia (HIT): If serious side effects, such as
HIT, are very rare and associated complications usually not serious, then use
of heparin prophylaxis is certainly indicated. (This also applies to the more
common entity known as nonimmune heparin-associated thrombocytopenia,
which is a benign event without adverse consequences.) Under these circum-
stances, effective pharmacological prophylaxis using heparin can be justified
to decrease the risk to immobilized patients even in an otherwise low-risk
category.
However, if the frequency of HIT is greater, even as high as 0.5-5% in
certain clinical settings (Greinacher, 1996a; Warkentin et al., 1998) (see Chap.
4), the risk-benefit assessment leads to a different conclusion: prophylaxis
with heparin entails significant risk and may no longer be less dangerous than
no prophylaxis. Whether there are alternate antithrombotic maneuvers
available that are as effective as UFH, but cause HIT less frequently, has to
be defined by the medical community.
The decision for or against prophylaxis thus depends on the overall risk/
benefit ratio, including the frequency and clinical effect of HIT in a particular
patient population, compared with the expected benefit of the heparin in
reducing thrombotic events. If, as current opinion suggests, the benefits of a
particular heparin preparation outweigh the low risk of HIT, then prophy-
laxis using heparin may be justified, even in patients at relatively low risk for
thrombosis. The relative frequencies of HIT in different clinical settings are
discussed in Chap. 4.
It remains debated among medical experts which of the various anti-
coagulants should be given to an individual patient. Recently this topic has
gained increased attention as several new drugs for antithrombotic prophy-
laxis have become available, e.g., hirudin, danaparoid, and fondaparinux.
Current guidelines (AWMF, 2003) do not favor any particular drug or drugs.
However, they state that some agents have a very low (or absent) risk of in-
ducing HIT. This does not mean that the new drugs are the medical standard
European Legal Aspects 589
for antithrombotic prophylaxis. The situation is different, however, if there is

a clear indication for one of these new agents, e.g., a patient with a recent
history of HIT.
According to the Stufenplanverfahren des Bundesinstituts für Arzneimit-
tel und Medizinprodukte (official measurements to reduce drug-associated
risks of the German Federal Institute for Drugs and Medicinal Products), it is
necessary to monitor the platelet count in every patient treated with heparin.
However, there is no medical consensus about the details of the monitoring.
The usual recommendation is one platelet count measurement before com-
mencing the heparin therapy, followed by three measurements a week from
day 5 and one weekly count from day 20 (Greinacher, 1996b). However, as
new information on the different risks of developing HIT among patient
populations emerge from clinical studies, an individualized approach to
platelet monitoring taking into account the particular degree of risk may be
appropriate (see Chap. 4).
III. THE INFORMED PATIENT
Another important legal aspect is the duty of the physician to inform the
patient about the actual risk of thrombosis, including the possible risks and
expected benefits of prophylaxis. This information is divided into two parts:
procedure-related and therapy-related. One example is the current discussion
on pharmacological prophylaxis of thrombosis in outpatients. The Bundes-
gerichtshof (federal court) (1996a,b,c) is of the opinion that the current dis-
cussion in medical science about the danger of thrombosis and the means of
pharmacological prophylaxis in outpatients justify a duty to provide infor-
mation. In these cases, ‘‘patient autonomy demands information about pos-
sible dangers of treatment and the means available to avoid or alleviate such
undesired effects.’’
Information about prophylaxis against thrombosis must include hem-
orrhage; allergic reactions, including HIT and its possible complications; the
need for platelet count monitoring; and, for long-term prophylaxis, osteopo-
rosis. Even if the risk of HIT is very low, information about it must be pro-
vided. Furthermore, information about certain life-threatening consequences
of HIT, such as permanent organ damage and even death, as well as possible
countermeasures against these outcomes, must be communicated. Statistical
probabilities in mathematical terms are not of major importance: they are,
according to the Bundesgerichtshof (1994a), of ‘‘minor importance only.’’
Critical, however, is whether the complications are relatively specific for the
treatment intervention in question. Thus, even extremely rare risks need to be
mentioned if they are known to be associated specifically with an intervention
590 Ulsenheimer
and if their occurrence would have noticeable influence on the patient’s life
and occupation (Bundesgerichtshof 1994b, 1996a; Oberlandesgericht Hamm,
1995). Consequently, there should be no doubt about the duty to inform
about the risks associated with HIT, as these complications are known to be
caused by heparin and, therefore, are specific for this intervention. Indeed, the
court (Oberlandesgericht Celle) ruled that the physician should have in-
formed the patient, at least briefly, about the risk of HIT (Oberlandesgericht
Celle, 2002).
If a doctor recommends using heparin in a low-risk patient outside an
approved indication, the patient must also be informed about this. This is
because in legal terms, approval of a drug is ‘‘like a seal of quality, that—
independent of the actual quality or safety—can be decisive for the patient’s
decision-making in the area where the pharmaceutical law is applicable, so he
has to be informed’’ (Bundesgerichtshof, 1996b).
Physicians are allowed to use drugs to treat diseases for which they have
not been approved, provided the medical necessity arises and there is a ratio-
nale or precedence for its benefit and reasonable safety in the clinical context
(Oberlandesgericht Köln, 1991; Bundessozialgericht, 2002). The regulatory
approval of a drug merely creates a state of confidence; that is, the doctor can
rely on the fact ‘‘that the risk versus benefit ratio is considered favorable in the
light of the evidence provided by the manufacturer and the examination of the
Bundesinstitut für Arzneimittel und Medizinprodukte’’ (Weißauer, 1994).
On the other hand, for a doctor using a nonapproved drug or an ap-
proved drug outside its approved indication, this state of confidence does not
cover him or her in the event of damaging side effects. The physician may then
be required to justify the treatment, for example, by showing that the drug has
been used with a favorable side effect profile, that reputable specialists
recommend its use, and that no alternative treatments are available. This is
not an uncommon situation in the management of complications of HIT
itself, where an alternative anticoagulant, danaparoid sodium, is often used
outside its approved indication (i.e., antithrombotic prophylaxis following
orthopedic surgery), or even without any approval in some countries, for the
treatment or prevention of thrombosis associated with HIT.
The physician must inform the patient about reasonable precautions in
optimizing the safety of a prescribed treatment. This includes educating
outpatients about typical signs and symptoms of therapeutic complications.
For example, informing patients about the possibility and significance of skin
reactions at heparin injection sites as an early manifestation of HIT is
appropriate. The main content of the information given to the patient should
be documented in writing by the physician, as legal protection in the event of a
subsequent adverse event occurring. It is especially prudent to document the
informed consent process if practice outside of usual medical care is contem-

plated, or if the patient refuses recommended treatment, such as antithrom-
botic prophylaxis. Incomplete or lacking documentation can lead to a reversal
of the burden of proof in favor of the patient if, in the event of a pulmonary
embolism or deep vein thrombosis, the doctor has to prove that he had
informed the patient.
IV. THE HARMED PATIENT
The mere violation of approved medical practice or failure in the duty to

obtain informed consent by themselves do not constitute a punishable offence
nor justify claims for compensation. For the physician’s mistake in treating or
informing the patient to become punishable, it must be proved that the patient
was harmed.
If it can be demonstrated that a patient who was not informed about the
risks of HIT would nevertheless have chosen this prophylactic treatment
despite its potential risks, then violation of the duty to obtain informed con-
sent becomes irrelevant. This is no longer true if the patient who suffered HIT
or any other complication can prove that having known about the risks, he or
she would have refused the treatment. In this context, the frequency of HIT
among various patient populations could be a factor determining the
likelihood that a patient would have given informed consent. For example,
a court may determine that a reasonable patient might not have given
informed consent to receive unfractionated heparin if the frequency of HIT
is shown to be about 5% (e.g., postoperative orthopedic patients), particu-
larly if other therapeutic options exist (e.g., low molecular weight heparin or
warfarin).
In case of a treatment error, such as omission to provide antithrombotic
prophylaxis, proof is needed that adequate prophylaxis against thrombosis
would have prevented the damage. In civil cases of alleged malpractice, the
burden of proof required by the plaintiff is relatively low, such as prima facie
evidence. In cases of gross negligence, the burden of proof even is reversed: if,
for example, there has not been any prophylaxis despite constitutional risk
factors and immobilization of a patient, or there was no radiological imaging
performed despite suspicion of thrombosis, the physician would have the
difficult task to prove that the pulmonary embolism or deep vein thrombosis
would have occurred even with prophylaxis.
This required burden of proof in a civil litigation is substantially less
than in a criminal case. In criminal law, one of the basic tenets is that the
benefit of doubt goes in favor of the accused. This means that negligent treat-
592 Ulsenheimer
ment, or lack of treatment when indicated, can only be proved to be causative

in a criminal case if the ‘‘correct’’ treatment (e.g., prophylaxis against throm-
bosis) would have prevented death or damage to the patient with a ‘‘proba-
bility bordering on certainty.’’ This legal term cannot be defined in precise
statistical terms (e.g., 95 or 99%). Rather, by raising a ‘‘reasonable doubt,’’
even a ‘‘high’’ or ‘‘very high’’ probability that an omitted treatment might
otherwise have prevented death or disability would not be sufficient to prove
the causality of a violation of duty in a criminal court (Bundesgerichtshof,
1988). ‘‘Probability bordering on certainty’’ means the exclusion of reason-
able doubts, i.e., doubts which are based upon concrete facts. Given the level
of uncertainty in medical knowledge, the required ‘‘probability bordering on
certainty’’ in a criminal case (e.g., when the physician is accused to have
omitted prophylaxis against thrombosis as a breach of the duty of care)
cannot be postulated.
V. COST OF ANTITHROMBOTIC PROPHYLAXIS
Cost is not an argument against indicated prophylaxis. The less expensive but
equally effective drug should be used, but social law’s duty to work econom-
ically does not legitimize lowering the standard of medical care. Medical duty
for care and the duty to work economically are not mutually exclusive, as
social law acknowledges the need for approved treatment. However, eco-
nomic considerations and price have to give way to aspects of effectiveness of
a drug and its indication. So, if there is an appropriate indication for
prophylaxis for an individual patient’s health, insurance must bear the cost.
REFERENCES
AWMF-Leitlinien. Stationäre und ambulante Thromboseprophylaxe in der Chirurgie

und in der perioperativen Medizin. Register-Nr. 003/001, 2003.
Bundesgerichtshof. Neue Juristische Wochenschr 1982; 35:2121–2122.
Bundesgerichtshof. Neue Juristische Wochenschr 1987; 40:2927.
Bundesgerichtshof. Monatsschrift für Deutsches Recht 1988; 42:100.
Bundesgerichtshof. Neue Juristische Wochenschr 1994a; 47:3012.
Bundesgerichtshof. Neue Juristische Wochenschr 1994b; 47:793 and 3012–3014.
Bundesgerichtshof. Neue Juristische Wochenschr 1996a; 49, 776–777.
Bundesgerichtshof. Monatsschr Dtsch Rechts 1996b; 50:1015–1016.
Bundessozialgericht, Urteil vom 19.03.2002, Az. B1 KR 37/00R, 2002.
Greinacher A. Heparin-induzierte Thrombozytopenien. Internist 1996a; 37:1172–
1178.
Greinacher A. Heparin-induzierte Thrombozytopenie (HIT): Wie kann das Risiko

einer postoperativen Thrombose oder Lungenembolie reduziert werden? Ortho-
paede 1996b; 25:379–384.
Oberlandesgericht Celle. Versicherungsrecht 2002; 52:854–855.
Oberlandesgericht Hamm. Versicherungsrecht 1995; 46:47–48.
Oberlandesgericht Köln. Versicherungsrecht 1991; 42:186–188.
Weißauer W. Anaesthesiol Intensivmed 1994; 35:204–205.
Appendixes
Appendixes 597
APPENDIX 1. TEN CLINICAL ‘‘RULES’’ FOR DIAGNOSING HIT
Rule 1: A thrombocytopenic patient whose platelet count fall began

between days 5 and 10 of heparin treatment (inclusive)
should be considered to have HIT unless proved otherwise
(first day of heparin use is considered ‘‘day 0’’).
Rule 2: A rapid fall in the platelet count soon after starting heparin
is unlikely to represent HIT unless the patient has received
heparin in the recent past, usually within the past 100 days.
Rule 3: A platelet count fall of more than 50% from the postopera-
tive peak between days 5 and 14 after surgery associated
with heparin treatment can indicate HIT even if the platelet
count remains higher than 150 109/L.
Rule 4: Petechiae and other signs of spontaneous bleeding are not
clinical features of HIT, even in patients with very severe
thrombocytopenia.
Rule 5: HIT is associated with a high frequency of thrombosis de-
spite discontinuation of heparin with or without substitution
by coumarin: the initial rate of thrombosis is about 5–10%
per day over the first 1–2 days; the 30-day cumulative risk is
about 50%.
Rule 6: Localization of thrombosis in patients with HIT is strongly
influenced by independent acute and chronic clinical factors,
such as the postoperative state, atherosclerosis, or the loca-
tion of intravascular catheters in central veins or arteries.
Rule 7: In patients receiving heparin, the more unusual or severe a
subsequent thrombotic event, the more likely the thrombosis
is caused by HIT.
Rule 8: Venous limb gangrene is characterized by (1) in vivo throm-
bin generation associated with acute HIT; (2) active deep
vein thrombosis in the limb(s) affected by venous gangrene;
and (3) a supratherapeutic international normalized ratio
(INR) during coumarin anticoagulation. This syndrome can
be prevented by: (1) delaying initiation of coumarin antico-
agulation during acute HIT until there has been substantial
recovery of the platelet count (to at least 100–150 109/L)
while receiving an alternative parenteral anticoagulant (e.g.,
lepirudin, argatroban, danaparoid), and only if the throm-
598 Appendixes
bosis has clinically improved; (2) initiating coumarin in low,

maintenance doses (e.g., 2–5 mg warfarin); (3) ensuring that
both parenteral and oral anticoagulant overlap for at least
5 days, with at least the last 2 days in the target therapeutic
range; and (4) if applicable, physicians should reverse
coumarin anticoagulation with vitamin K in a patient
recognized with acute HIT after coumarin therapy has been
commenced.
Rule 9: Erythematous or necrotizing skin lesions at heparin injection
sites should be considered dermal manifestations of the HIT
syndrome, irrespective of the platelet count, unless proved
otherwise. Patients who develop thrombocytopenia in asso-
ciation with heparin-induced skin lesions are at increased
risk for venous and, especially, arterial thrombosis.
Rule 10: Any inflammatory, cardiopulmonary, or other unexpected
acute event that begins 5–30 min after an intravenous hep-
arin bolus should be considered acute HIT unless proved
otherwise. The postbolus platelet count should be measured
promptly, and compared with prebolus levels, because the
platelet count fall is abrupt and often transient.
Disclaimer: The preceding 10 clinical ‘‘rules’’ have been formulated primarily
for didactic purposes, and are not intended necessarily to imply any standard
of care in relation to their clinical application.
Appendixes 599
APPENDIX 2. ESTIMATING THE PRETEST PROBABILITY OF HIT:

THE FOUR T’S
Points (0, 1, or 2 for each of 4 categories:
maximum possible score = 8)a
2 1 0
Thrombocytopenia Nadir, 20–100 (at Nadir, 10–19 109/L; Nadir, <10

(acute) least 30% fall); or any 30–50% fall; 109/L; or any
or any >50% (or >50% fall <30% fall
fall (nadir z 20) associated with
heart surgery)
Timingb of platelet Clear onset Consistent with day Platelet count fall
count fall, between days 5–10 fall, but not V4 days without
thrombosis, or 5–10; or V1 clear (e.g., missing recent heparin
other sequelae day (if heparin platelet counts); or exposure
(first day of exposure within V1 day (heparin
heparin course past 30 days) exposure within past
=day 0) 31–100 days); or
platelet fall after
day 10
Thrombosis or New thrombosis; Progressive or None
other sequelae skin necrosis; recurrent
(e.g., skin lesions, ASR after iv thrombosis;
ASR) heparin bolus erythematous skin
lesions; suspected
thrombosis (not
yet proven);
asymptomatic
upper-limb DVT
Other cause for No explanation Possible other Definite other
thrombocytopenia for platelet cause is evident cause is
not evident count fall is present
evident
ASR, acute systemic reaction; DVT, deep venous thrombosis.
a
Pretest probability score: 6–8 = high; 4–5 = intermediate; 0–3 = low.
b
First day of immunizing heparin exposure considered day 0; the day the platelet count begins to fall is
considered the day of onset of thrombocytopenia (it generally takes 1–3 more days until an arbitrary
threshold that defines thrombocytopenia is passed). The scoring system shown here has undergone
minor modifications from previously published scoring systems (see Chap. 3).
600 Appendixes
APPENDIX 3. PLATELET COUNT MONITORING FOR HIT
We recommend:
1. Monitoring for typical-onset HIT: stratifying the intensity of
platelet count monitoring for HIT based upon its risk
A. Patients at highest risk for HIT (1–5%) (e.g., postoperative
patients receiving prophylactic-dose UFH after major sur-
gery): monitoring during heparin therapy, at least every second
day from day 4 to day 14.*
Patients receiving therapeutic-dose UFH: platelet count mon-
itoring once dailyy from day 4 to day 14.*
B. Patients at intermediate risk for HIT (0.1–1%) (e.g., medical/
obstetrical patients receiving prophylactic-dose UFH, or post-
operative patients receiving prophylactic-dose LMWH, or
postoperative patients receiving intravascular catheter
‘‘flushes’’ with UFH): monitoring during heparin therapy, at
least every 2 or 3 days from day 4 to day 14,* when practical.z
C. Patients at low risk for HIT (<0.1%) (e.g., medical/obstetrical
patients receiving prophylactic- or therapeutic-dose LMWH,
or medical patients receiving only intravascular catheter
‘‘flushes’’ with UFH): routine platelet count monitoring is
not recommended.§
These are draft recommendation (Seventh American College of Chest Physi-
cians Consensus Conference on Antithrombotic Therapy, September 2003).
Readers should consult the publication (Warkentin and Greinacher, 2004) to
obtain the final recommendations.
2. Monitoring for rapid-onset HIT: for a patient recently exposed to
heparin (within the past 100 days), a repeat platelet count within 24
hours following reinitiation of heparin.
* The crucial time period for monitoring ‘‘typical-onset’’ HIT is between days 4 and 14 (first
day of heparin = day zero), where the highest platelet count from day 4 (inclusive) onwards
represents the ‘‘baseline.’’ Platelet count monitoring can cease before day 14 when heparin is
stopped.
y
Once-daily platelet count monitoring recommended as daily blood draws required for aPTT
monitoring.
z
Frequent platelet count monitoring may not be practical when UFH or LMWH is given to
outpatients.
§
Monitoring as per ‘‘intermediate’’ risk is appropriate if UFH was given before initiating
LMWH.
Appendixes 601
3. When to suspect HIT:

A relative (proportional) platelet count fall of 50% or greater that
is otherwise clinically unexplained should be considered suspicious
for HIT, even if the platelet count nadir remains above 150 109/L.
For any patient who develops thrombosis during or within
several days after heparin therapy, or who develops an unusual
clinical event in association with heparin therapy (e.g., inflamma-
tory or necrotic skin lesions at heparin injection sites, acute sys-
temic reaction post–intravenous heparin therapy), a repeat platelet
count should be measured promptly, and compared with recent
values.
602 Appendixes
APPENDIX 4. TREATMENT RECOMMENDATIONS*

Nonimmune Heparin-Induced Thrombocytopenia
Recommendation. Heparin should not be discontinued in patients clinically
suspected of having nonimmune heparin-associated thrombocytopenia
(grade 2C).
Immune Heparin-Induced Thrombocytopenia

Discontinuation of Heparin for Clinically Suspected HIT
Recommendation. All heparin administration should be discontinued in
patients clinically suspected of having (immune) HIT (grade 1C+).
Recommendation. A clearly visible note should be placed above the patient’s
bed stating ‘‘NO HEPARIN: HIT’’ (grade 2C).
Recommendation. Heparin can be safely restarted in patients proved not to
have HIT antibodies by sensitive activation or antigen assay (grade 2C).
Anticoagulation of the HIT Patient with Thrombosis

Recommendation. Therapeutic-dose anticoagulation with a rapidly acting
anticoagulant, e.g., danaparoid (grade 1B), lepirudin (grade 1C+), or arga-
troban (grade 1C), should be given to a patient with thrombosis complicating
acute HIT. Treatment should not be delayed pending laboratory confirma-
tion in a patient strongly suspected of having HIT.
Anticoagulation of the HIT Patient Without Thrombosis

Recommendation. Alternative anticoagulation with an appropriate antico-
agulant, such as danaparoid, lepirudin, or argatroban, should be considered
in patients with clinically suspected HIT even in the absence of symptomatic
thrombosis. Anticoagulation should be continued at least until recovery of
the platelet counts to a stable plateau. Patients should undergo imaging stud-
ies for lower limb DVT, especially those at highest risk for venous thrombo-
embolism, such as postoperative patients (grade 1C+).
* The grades of recommendation are from Guyatt et al. (2001) and Hirsh et al. (2001) (in Chap.
13) and are described in Chap. 13.
Appendixes 603
Longer-Term Anticoagulant Management of the HIT Patient

with Thrombosis
Recommendation. The drug of choice for longer-term anticoagulation of
HIT patients is an oral anticoagulant of the coumarin class (e.g., warfarin or
phenprocoumon). However, in a patient with acute HIT, oral anticoagulant
therapy should be delayed until the patient is adequately anticoagulated with
a rapidly acting parenteral anticoagulant, and ideally not until there has been
substantial platelet count recovery (at least z100 109/L). Oral anticoagu-
lants should be started in low maintenance doses (e.g., <5 mg warfarin), with
at least 5 days of overlap with the parenteral anticoagulant (including at least
2 days in the target therapeutic range). If applicable, oral or intravenous
vitamin K should be given to reverse coumarin anticoagulation in a patient
recognized as having acute HIT after coumarin has been commenced (grade
1C).
Recommendation. Prothrombin complex concentrates should not be used
to reverse coumarin anticoagulation in a patient with acute or recent HIT
unless bleeding is otherwise unmanageable (grade 2C).
Reexposure of the HIT Patient to Heparin

Heparin Reexposure of the Patient with Acute or Recent HIT
Recommendation. Deliberate reexposure to heparin of a patient with acute
or recent HIT for diagnostic purposes is not recommended. Rather, the
diagnosis should be confirmed by testing acute patient serum or plasma for
HIT antibodies using a sensitive activation or antigen assay (grade 1C+).
Heparin Reexposure of the Patient with a History of Remote HIT
Recommendation. Heparin should not be used for antithrombotic pro-
phylaxis or therapy in a patient with a previous history of HIT, except under
special circumstances (e.g., cardiac or vascular surgery) (grade 2C).
Cardiopulmonary Bypass or Vascular Surgery

Management of the Patient with Acute or Recent HIT
Recommendation. Alternative anticoagulation should be used for heart or
vascular surgery in a patient with acute or recent HIT with detectable HIT
antibodies. Bivalirudin, lepirudin, or danaparoid are appropriate alternatives
for intraoperative anticoagulation, provided that appropriate, rapid-turn-
around laboratory monitoring and blood product support to manage poten-
tially severe bleeding complications are available. Another approach is to give
heparin together with a potent antiplatelet agent (grade 2C).
604 Appendixes
Management of the Patient Following Disappearance of HIT Antibodies

Recommendation. In a patient with a previous history of HIT, heart or vas-
cular surgery can be performed using heparin, provided that HIT antibodies
are absent (by sensitive assay), and heparin use is restricted to the surgical
procedure itself (grade 1C).
HIT During Pregnancy

Recommendation. If available, danaparoid (and possibly daparinux) is pre-
ferred for parenteral anticoagulation of pregnant patients with HIT or in
those who have a previous history of HIT (grade 2C).
Adjunctive Therapies for HIT

Medical Thrombolysis
Recommendation. Regional or systemic pharmacological thrombolysis
should be considered as a treatment adjunct in selected patients with limb-
threatening thrombosis or pulmonary embolism with severe cardiovascular
compromise (grade 2C).
Surgical Thromboembolectomy
Recommendation. Surgical thromboembolectomy is an appropriate adjunc-
tive treatment for selected patients with limb-threatening large-vessel arterial
thromboembolism. Thrombocytopenia is not a contraindication to surgery.
An alternative anticoagulant to heparin should be used for intraoperative
anticoagulation (grade 1C).
Intravenous Gammaglobulin
Recommendation. ivIgG is a possible adjunctive treatment in selected pa-
tients requiring rapid blockade of the Fc receptor–dependent platelet-acti-
vating effects of HIT antibodies (e.g., management of patients with cerebral
venous thrombosis, severe limb ischemia, or very severe thrombocytopenia)
(grade 2C).
Plasmapheresis
Recommendation. Plasmapheresis, using plasma as replacement fluid, may
be a useful adjunctive therapy in selected patients with acute HIT and life- or
limb-threatening thrombosis who are suspected or proved to have acquired
deficiency of one or more natural anticoagulant proteins (grade 2C).
Dextran
Recommendation. Dextran should not be used as primary therapy for acute
HIT complicated by thrombosis (grade 1B).
Appendixes 605
Acetylsalicylic Acid and Dipyridamole

Recommendation. Antiplatelet agents, such as aspirin, may be used as ad-
juncts to anticoagulant therapy of HIT, particularly in selected patients at
high risk for arterial thromboembolism. The possible benefit in preventing
arterial thrombosis should be weighed against the potential for increased
bleeding (grade 2C).
Platelet Glycoprotein IIb/IIIa Inhibitors
Recommendation. GP IIb/IIIa inhibitors should be considered as experi-
mental treatment in HIT and used with caution if combined with anticoag-
ulant drugs (grade 2C).
Caveats for the Treatment of HIT

Low Molecular Weight Heparin (LMWH)
Recommendation. LMWH should not be used to treat patients with acute
HIT (grade 1C).
Oral Anticoagulants (Vitamin K Antagonists)
Recommendation. Oral anticoagulants are contraindicated in patients with
acute HIT, unless combined with an agent that reduces thrombin generation
(grade 1C+).
Ancrod
Recommendation. Ancrod should not be used to treat patients with HIT
(grade 1C).
Platelet Transfusions
Recommendation. Prophylactic platelet transfusions are relatively contra-
indicated in patients with acute HIT (grade 2C).
Disclaimer. The listed recommendations represent the general views of
the editors (as of September 2003) and are intended primarily as an educa-
tional guide for physicians who must treat patients with clinically suspected,
or serologically proven, HIT. The editors wish to emphasize that these recom-
mendations cannot be applied indiscriminately to all clinical situations, for
reasons that can include concomitant clinical factors, diagnostic uncertainty,
availability of alternative anticoagulant options, and the availability and
turnaround times for HIT antibody and anticoagulant monitoring assays.
606 Appendixes
APPENDIX 5. DANAPAROID DOSING SCHEDULES IN HIT

PATIENTS
Venous thromboembolism
Prophylaxis (prior HIT) 750 U sc, b.i.d. or t.i.d.
Prophylaxis (acute HIT) Treatment doses (see below) may be
appropriate for prophylaxis of
acute HIT (see pp. 16, 348–349, 379)
Venous or arterial 2250 U iv bolusa followed by 400 U/h for
thromboembolism: 4 h, 300 U/h for 4 h, then 150–200 U/h
Treatment (either for z5 days, aiming for a plasma
prior or acute HIT) anti-Xa level of 0.5–0.8 U/mL
Subcutaneous administration schedule:
1500–2250 U sc b.i.d. (given almost
100% bioavailability, 2250 U sc b.i.d. is
approximately equal to an iv infusion
rate of 200 U/h)
Embolectomy or other Preoperative: 2250 U iv bolusa,
peripheral vascular intraoperative flushes: 750 U in 250 mL
surgery saline, using up to 50 mL (see p. 355);
postoperative: 750 U sc t.i.d. (low-risk
patients) or 150–200 U/h (high-risk
patients) beginning at least 6 h after
surgery
Hemodialysis (on 3750 U iv before 1st and 2nd dialyses;
alternate days) 3000 U for 3rd dialysis; then 2250 U
for subsequent dialyses, aiming for
plasma anti-Xa level of <0.3 U/mL
predialysis, and 0.5–0.8 U/mL during
dialysis (see also Chap. 18).
Hemofiltration 2250 U iv bolus, followed by 600 U/h for
4 h, then 400 U/h for 4 h, then 200–400
U/h aiming for a plasma anti-Xa level
of 0.5–1.0 U/mL (see also Chap. 18).
Cardiopulmonary 125 U/kg iv bolus after thoracotomy;
bypass surgery 3 U/mL in priming fluid of apparatus;
(CPB) 7 U/kg/h iv infusion commencing after
CPB hookup, and continued until 45
min before expectation of stopping
CPB (see also Chap. 19)
Appendixes 607

Cardiac Preprocedure: 2250 U iv bolus (3000 U if
catheterization 75–90 kg and 3750 U if >90 kg)
Percutaneous Preprocedure: bolus as per foregoing;
transluminal postprocedure: 150–200 U/h for 1–2 days
coronary after PTCA (or until removal of balloon
angioplasty pump)
(PTCA) or
intra-aortic
balloon pump
Catheter patency 750 U in 50 mL saline, then 5–10 mL per
port, or as required
Pediatric dosage Prophylaxis: 10 U/kg sc b.i.d.
considerations Treatment: 30 U/kg b.w., iv bolus,
then 1.2–2.0 U/kg b.w./h
depending upon severity of
thrombosis
Abbreviations: b.w., body weight; b.i.d., twice daily; iv, intravenous; t.i.d., three times daily.
Compatibility with intravenous solutions: Danaparoid is compatible for dilution with the fol-
lowing solutions: saline, dextose, dextrose-saline, Ringer’s, lactated Ringer’s, 10% mannitol.
Preparation of solution for infusion: One option is to add four ampules containing 3000 U (i.e.,
750 anti-Xa U/0.6 mL ampule) of danaparoid to 300 mL of intravenous solution (i.e., a solution
that comprises 10 U danaparoid per milliliter of intravenous solution; thus, an infusion rate of
40 mL/h corresponds to a dose of 400 U/h; 20 mL/h to a dose of 200 U/h, and so on.
a
Adjust iv danaparoid bolus for body weight <60 kg, 1500 U; 60–75 kg, 2250 U; 75–90 kg,
3000 U; >90 kg, 3750 U.
608 Appendixes
APPENDIX 6. DOSING SCHEDULES FOR LEPIRUDIN TREATMENT OF

PATIENTS WITH HIT
Target aPTT
Bolusa,b IV infusiona,b ratioc
HIT with isolated Noned 0.10 mg/kg 1.5–2.5

thrombocytopenia b.w./hd,e
(dose regimen B
in HAT trials)
HIT and thrombosis 0.40 mg/kgd 0.15 mg/kg 1.5–2.5
(dose regimen A1 b.w./hd
in HAT trials)
Thrombosis 15 mg sc b.i.d.f —
prophylaxis in
patients with a
history of HIT
HIT with thrombosis 0.20 mg/kg b.w. ivd 0.10 mg/kg 1.5–2.5
and concomitant b.w./hd
thrombolysis (dose
regimen A2 in
HAT trials)
Renal dialysis every 0.10 mg/kg b.w. iv — 2.0–2.5
alternate day predialysis
Continuous — 0.005 mg/kg 1.5–2.5
venovenous b.w./h
hemofiltration (initial rate)
(CVVH)
PCI; UA or acute 0.40 mg/kg b.w. iv 0.15 mg/kg 1.5–2.5
MI without ST b.w./h
elevation
Vascular surgery 0.40 mg/kg b.w. iv 0.10 mg/kg/h 1.5–2.5
Vascular surgery Use up to 250 mL — —
(intraoperative (0.1 mg/mL solution)
vessel flushes)
Postoperative — 0.10 mg/kg 1.5–2.5
anticoagulation b.w./h
Cardiac surgery 0.25 mg/kg b.w. ivd 0.50 mg/mina,g Monitored by
using CPB (dose (0.20 mg/kg b.w. in ECT: >2.5
regimen C in the priming fluid) Ag/mL before
HAT trials) (see start of CPB;
also Chap. 19) 3.5–4.5 Ag/mL
during CPBh
Repeat aPTT determinations should be made 4–6 h after any dose adjustment.
Appendixes 609
Abbreviations: aPTT, activated partial thromboplastin time; b.w., body weight; CPB, cardiopulmonary
bypass; ECT, ecarin clotting time; iv, intravenous; MI, myocardial infarction; PCI, percutaneous coronary
intervention; UA, unstable angina.
a
A maximum body weight of 100 kg should be used for dose calculations.
b
Adjust for renal insufficiency.
c
The ratio is based on comparison with the normal laboratory mean aPTT. If Actin FS or Neothromtin
reagents are used, the aPTT target range is usually 1.5–3.0.
d
Used in the HAT-1, HAT-2, and HAT-3 trials.
e
This is the author’s recommended starting dose in all HIT patients, unless life- or limb-threatening
thrombosis is present.
f
Tested in a prospective, randomized trial after orthopedic surgery (Eriksson et al., 1996, 1997).
g
Stop 15 min before end of CPB; put 5 mg into CPB after disconnection to avoid clotting of pump.
h
The target lepirudin level pre-CPB (>2.5 Ag/mL) is lower than the level sought during CPB (3.5–4.5 Ag/
mL) because of the addition of lepirudin to the pump priming fluid (0.2 mg/kg b.w.).
610 Appendixes
APPENDIX 7. DOSING SCHEDULE FOR LEPIRUDIN IN

PATIENTS WITH HIT AND RENAL IMPAIRMENT
Creatinine Adjusted iv infusion
clearance Serum creatinine, rate (% of original
(mL/min) mg/dL (Amol/L) dose [see Appendix 6])a
45–60 1.6–2.0 (141–177) 50
30–44 2.1–3.0 (178–265) 25
15–29 3.1–60 (266–530) 10
<15 >6.0 (>530) 0.005 mg/kg/h b.w. iv
(adjusted for aPTT)
Abbreviations: aPTT, activated partial thromboplastin time; b.w., body weight; iv, intravenous.
a
No initial bolus of lepirudin is given.
Appendixes 611
APPENDIX 8. DOSING SCHEDULES FOR ARGATROBAN TREATMENT

OF PATIENTS WITH HIT (APPROVED INDICATIONS)
Monitoring and
Clinical use Bolusa IV infusiona adjusting therapy
Prophylaxis or — 2 Ag/kg/min (For Dose adjusted (not to

treatment of hepatically impaired exceed 10 Ag/kg/min)
thrombosisb,c patients, reduce to achieve steady
initial dose.d state aPTT 1.5–3.0
Patients with renal times the baseline
insufficiency require value (not to exceed
no initial dosage 100 s)e,f,g
adjustment.)
Percutaneous 350 Ag/kg 25 Ag/kg/min Infusion dose adjusted
coronary (given over (15–40 Ag/kg/min) to achieve
intervention 3–5 min) an ACT 300–450 s; additional
(PCI)b,h,i bolus doses of 150 Ag/kg may
be given as neededj,k
Abbreviations: HIT, heparin-induced thrombocytopenia; IV, intravenous; aPTT, activated partial

thromboplastin time; ACT, activated clotting time; PCI, percutaneous coronary intervention.
a
Based on patient’s body weight.
b
Includes patients with active HIT who have isolated thrombocytopenia or associated thrombosis, as well as
patients with a documented history of HIT who are no longer thrombocytopenic but require anti-
coagulation.
c
Argatroban is approved in the United States as an anticoagulant for prophylaxis or treatment of
thrombosis in patients with HIT, and in Canada as an anticoagulant in patients with HIT who in the
opinion of their attending physician requires anticoagulation.
d
For patients with moderate hepatic impairment, an initial dose of 0.5 Ag/kg/min is recommended.
e
The aPTT should be checked at least 2 h after the initiation of argatroban or any dosage change.
f
For patients in studies ARG-911 and ARG-915, the mean F SEM dose of argatroban was 1.9 F 0.1 Ag/
kg/min.
g
For transferring a patient to warfarin anticoagulant therapy: After substantial resolution of throm-
bocytopenia, initiate warfarin therapy using the expected daily dose of warfarin (do not use a loading dose)
while maintaining argatroban infusion. At least 5 days of warfarin therapy are required to lower functional
prothrombin concentrations to a therapeutic, steady state level. For monitoring the conversion to warfarin
during coadministration of argatroban at doses up to 2 Ag/kg/min, see text and Fig. 16.7.
h
Argatroban is approved in the United States as an anticoagulant in patients with or at risk for HIT under-
going PCI. Argatroban has not been evaluated in hepatically impaired patients undergoing PCI. These
recommendations do not consider the combination use of argatroban with glycoprotein IIb/IIIa antago-
nists, wherein lower doses of argatroban (e.g., 250–300 Ag/kg bolus followed by infusion of 15 Ag/kg/min)
have been shown to provide effective anticoagulation with an acceptable bleeding risk (Jang et al., 2003).
i
Includes percutaneous transluminal coronary angioplasty (balloon angioplasty), stent implantation, and
atherectomy; oral aspirin 325 mg should be given 2–24 h prior to PCI.
j
The ACT should be checked 5–10 min following the initial bolus dose and after any additional bolus dose
or change in the infusion rate. In studies ARG-216, ARG-310, and ARG-311, the majority of patients
required only one bolus dose during the interventional procedure, and the mean F SEM dose of
argatroban was 23.1 F 0.7 Ag/kg/min.
k
After the procedure, the sheaths should be removed no sooner than 2 h after discontinuing argatroban
and when the ACT is <160 s.
APPENDIX 9. TIMELINES OF AN EPISODE OF HIT
612
Appendixes
Index
Abciximab, 443, 485, 492 Adenocarcinoma-associated

Abciximab-induced immune thrombo- thrombotic endocarditis, 80
cytopenia, 35 Adenosine diphosphate (ADP), 230
ACCP Consensus Conference on Adenosine triphosphate (ATP), 273,
Antithrombotic Therapy, 335 278
Acetylsalicylic acid, 356–357 ADP, 230
Acquired antibody-mediated platelet Adrenal hemorrhagic infarction,
dysfunction, 29–30 80–81
Acquired anticoagulant deficiency, AECA, 254
81–83 Africa, 30
Acrocyanosis, 321 AITP, see Autoimmune thrombocyto-
Activation assays, 273 penic purpura (AITP)
vs. antigen assays, 293–295 Alpha-methyldopa, 36
citrate-anticoagulated blood, Alteplase, 488
285–287 American College of Chest Physicians
HIT antibodies, 271–287 (ACCP) Consensus Conference
in vitro cross-reactivity, 302–303 on Antithrombotic Therapy,
Acute coronary syndrome, 136 335
anti-PF4-heparin antibodies, 203 Amiodarone, 488
lepirudin, 425 Amiral, Jean, 11–12
Acute myocardial infarction, 444 Amphotericin B, 488
Acute postinfectious autoimmune Ancrod (Arvin), 359
thrombocytopenia, 27 cardiopulmonary bypass,
Acute thrombocytopenia, 25–42 542–543
ADCC, 226 Anemia, warm-type autoimmune
Adenocarcinoma-associated dissemi- hemolytic, 28
nated intravascular coagulation Angina pectoris, 444
(DIC), 313, 314, 315–318 Angiomax, see Bivalirudin (Angiomax)
venous limb gangrene, 315–318 Annexin V-binding assay, 273
613
614 Index
Antibivalirudin antibodies, 500 Antigen assays, 273, 285

Antibodies vs. activation assays, 293–295
antibivalirudin, 500 children, 563–564
anticardiolipin, 321 cross reactivity, 303
antichemokine, preexisting, 172 HIT antibodies, 288–295
antiendothelial cell, 254 Antigen-presenting cells (APC),
antiheparin-PF4, 253 190–191
acute coronary syndrome, 203 Antigens
drug-dependent, 41–42 antibody binding, 170
immunoglobulin-binding assays, fluid-phase, 273
41 glycoproteins, 26
endothelial cells, HIT, 259–262 heparin-dependent, HIT, 165–172
heparin-dependent, pathogenicity, surface-bound, 273
168–171 target, 273
heparin-induced thrombocytopenia Antiheparin-PF4 antibodies, 253
(HIT)-associated, beta-1, acute coronary syndrome, 203
3-glucan sulfates, 207–209 Antiphospholipid antibody syndrome
heparin-induced thrombocytopenia (APLAS), 255, 271, 272, 314,
(HIT)-IgG, 340 321–324
IgG, 155 vs. HIT, 323–324
PF4-H, platelet activation, 170 thrombocytopenia, 324
PF4-H-reactive, 169 Antiplatelets, 356–357
platelet, characterization, 39–42 Antithrombin, 152, 197
platelet-activating, HIT, 7–8 deficiency, 465
platelet-bound characterization, Anti-Xa-inhibiting pentasaccharide,
40–41 135–136
quinidine-dependent, 42 Aortic aneurysms, 37
Antibody-dependent cellular cytotoxic- Aortic valve replacement, 80
ity (ADCC), 226 APC, 190–191
Antibody-independent platelet activa- APLAS, see Antiphospholipid anti-
tion, 152 body syndrome (APLAS)
Anticardiolipin antibodies, 321 APTT, 401
Antichemokine antibodies, 172 Apyrase, 279
Anticoagulants Argatroban-911, 447–448
acquired deficiency, 81–83 Argatroban-915, 448–541
cardiopulmonary bypass, Argatroban (Novastan), 342–343, 347,
531–546 399, 437–467, 581–582
with danaparoid sodium, 378–379 breastfeeding, 465–466
decision making, 545–546 cardiopulmonary bypass, 541
HIT, 344 chemical description, 438
oral, 358–359 chemical structure, 439
hemodialysis, 521–522 children, 466, 565
U.S. approval, 18 clinical pharmacology, 438–443
Antiendothelial cell antibodies clinical use, 444–445
(AECA), 254 conversion to warfarin, 455–460
Index 615
[Argatroban (Novastan)] Autoantigens, cryptic, 166

cross-reactivity, 443 Autoimmune disease, 238
distribution, metabolism and excre- Autoimmune hemolytic anemia, warm-
tion, 440–441 type, 28
dosing and monitoring, 454–460, Autoimmune thrombocytopenia
455 acute postinfectious, 27
drug-drug interactions, 442–443 drug-induced, 36
elderly, 466 idiopathic, 36
hemodialysis, 520–521 Autoimmune thrombocytopenic
HIT, 445–454 purpura (AITP), 25–26,
acute anticoagulation, 451–453 26–30
cardiopulmonary bypass surgery, chronic, 27–28
464–465 clinical manifestations, 27–28
extracorporeal membrane oxygen- diagnosis, 28–29
ation, 464–465 pathogenesis, 26–27
hemodialysis, 463–464 secondary, 27
percutaneous coronary interven- therapy, 29
tion, 460–463 Azathioprine, 29
prospective studies, 453–454
stroke, 464 BAT, 489, 491
isolated HIT, 15–16 Best, Charles H., 1
vs. lepirudin, 418–419 Beta-1,3-glucan sulfates
mechanism of action, 438–439 anticoagulant activity, 207–209
pharmacokinetic-pharmacodynamic HIT-associated antibodies,
relationship, 441 207–209
pregnancy, 465–466 BetaTG, 171
reexposure, 453 Beta-thromboglobulin (betaTG),
reversal, 444 171
special populations, 441–442 Bivalirudin (Angiomax), 343, 399,
therapy duration, 454–455 475–501
Arixtra, see Fondaparinux (Arixtra) administration, 485
Arterial obstructive disease, peripheral, adverse effects, 488–489
444 cardiopulmonary bypass, 537–541
Arterial thromboembolism, 83–84 chemistry, 476–478
danaparoid sodium, 375–376 clinical use, 489–496
dosing schedules, 377 cost, 500–501
prophylaxis, 380 dosage, 481–484
Arvin, 359 HIT, 19, 496–500
cardiopulmonary bypass, 542–543 monitoring, 485–486
Aspirin, 485 pharmacodynamics, 479–481
hemodialysis, 512 pharmacokinetics, 478–479
Asserachrom, 290, 292 pharmacology, 478
Atherogenesis, 258 reversal, 488
ATP, 273, 278 Bivalirudin Angioplasty Trial (BAT),
Atropine, 443 489, 491
616 Index
Black population of Africa, 30 [Children]

Blue toe syndrome, 321 HIT, 91–93, 353–354, 553–567
Bone marrow transplantation, 27 clinical presentation, 555–562
HIT, 93 frequency, 553–555
Breastfeeding, 465–466 laboratory testing, 562–564
Burden of proof in malpractice, pathophysiology, 553
576–577 prevention, 566–567
therapy, 564–566
CACHET, 492 lepirudin, 427–428
Calcium, 152 low molecular weight heparin
Cancer, consumptive thrombohemor- (LMWH), 566–567
rhagic disorders, 37 Chlorpromazine, 488
Carbohydrate-based heparin alterna- Chong, Beng, 9
tives, 209–210 CHOOSE, 499
Carbohydrates, sulfated Chronic autoimmune thrombo-
HIT, 207 cytopenic purpura (AITP),
PF4, 198–202 27–28
Cardiac catheterization, 377 Circulated platelet-rich plasma (c-PRP)
Cardiac surgery, 134–135 activation assays
Cardiopulmonary bypass HIT antibodies, 285–286
ancrod (Arvin), 542–543 vs. washed platelet assays, 287,
anticoagulation, 381–382, 302–303
531–546 aggregation assays, disadvantages,
argatroban, 541 286
HIT, 464–465 CISN, 76, 87–88, 336
bivalirudin (Angiomax), 494, Citrate-anticoagulated blood activa-
498–499, 537–541 tion assays, 285–287
danaparoid sodium, 532–534 Citrate anticoagulation, regional,
dosing schedules, 377 523
HIT, 351–354 Citrated-anticoagulated blood, 287
lepirudin, 425–426 Classic coumarin-induced skin necrosis
platelet inhibition, 541–542 (CISN), 87–88
recombinant hirudin (lepirudin), Clexane, 119
534–536 Clinical heparin-induced thrombocyto-
CDR3, 190 penia (HIT), 204–205
Cellular immune response, 189–191 Clinical practice guidelines, 577
Cerebral venous (dural sinus) thrombo- Clinicopathologic syndrome, 108
sis, 79–80 Commercial low molecular weight
Chemokine receptor, 257 heparin (LMWH), 206
Children Congenital hypercoagulability, 83
antigen assays, 563–564 Consumptive thrombohemorrhagic
argatroban, 466 disorders, 36–38
danaparoid sodium, 384–385 clinical disorders, 36–37
dosing schedules, 377 diagnosis, 37–38
EIA, 562–563 pathogenesis, 36
Index 617
Coronary artery bypass surgery, Deep venous thrombosis (DVT),

see Cardiopulmonary bypass 12
Corticosteroids, 32–33 bivalirudin (Angiomax), 489–490,
Coumarin, 349 490
children, 565–566 danaparoid sodium, 374
conversion to, 455–460 lower limb, 73–75
multiple digital necrosis, 80 upper limb, 75
transitioning to, 406 Dermatan sulfate, 343
venous limb gangrene, 13–14, 75–79, hemodialysis, 522
77 Desmopressin, 407
Coumarin-induced skin necrosis Dexamethasone, 485
(CISN), 76, 87–88, 336 Dextran, 356
C-PRP, see Circulated platelet-rich Diabetic ketoacidosis, 314, 318–321,
plasma (c-PRP) 322
Crohn’s disease, 27 Diazepam, 488
Cross reactivity antigen assays, 303 DIC, see Disseminated intravascular
Cryptic autoantigens, 166 coagulation (DIC)
Cyclic thrombocytopenia, 30 Digoxin, 485
Cytopenia, idiopathic autoimmune, Diltiazem, 443
36 Diphenhydramine, 443, 485
Dipyridamole, 356–357
Dalteparin, 119 Disseminated intravascular coagula-
Danaparoid sodium (Orgaran), 157, tion (DIC), 5, 28, 81–83
183, 209–210, 342–343, 344, adenocarcinoma-associated, 313,
345, 348, 371–391, 581 314, 315–318
adverse effects, 389 venous limb gangrene, 315–318
with anticoagulants, 378–379 argatroban, 465
availability, 391 clinical manifestations, 82
cardiopulmonary bypass, consumptive thrombohemorrhagic
532–534 disorders, 37
chemistry, 372 danaparoid sodium, 374
children, 384–385, 564–565 DIT, 33–36
clinical use, 374–391 drugs inducing, 34
dosing schedules, 377 therapy, 34
hemodialysis, 513–515 Dixon, R.H., 5
HIT Dobutamine, 443, 485
antibodies, 386–389 Dopamine, 443, 485
thrombosis, 12 Drug-dependent antibodies, 41–42
laboratory monitoring, 385–386 immunoglobulin-binding assays, 41
vs. lepirudin, 419–420 Drug-induced autoimmune thrombo-
low-dose, 336 cytopenia, 36
pharmacodynamics, 373–374 Drug-induced immune thrombocy-
pharmacology, 373 topenia (DIT), 33–36
pregnancy, 384–385 drugs inducing, 34
prophylaxis, 379 therapy, 34
618 Index
Dural sinus thrombosis, 79–80, 80 Factor Xa (FXa), 199

DVT, see Deep venous thrombosis Familial thrombocytopenia, 28
(DVT) Fcg11a receptor-mediated platelet acti-
vation, 227–232
Ecarin clotting time (ECT), 401–402 Fcg11a receptor-mediated signal trans-
ECT, 401–402 duction, 230–232
EIA, see Enzyme immunoassay (EIA) FcgR111a, 224–227
ELAM, 252 Arg/His131 polymorphism, 234–235
Elderly, argatroban, 466 HIT, 232–236
ELISA, 186, 401–402 plasma-soluble, 233
Embolectomy, danaparoid sodium, FcgR11a polymorphism
377 Arg/His131 variant, 237
Embolism, pulmonary, 12 autoimmune disease, 238
Endarterectomy, 495 disease, 236–241
Endocarditis HIT, 238–241
adenocarcinoma-associated infectious disease, 238
thrombotic, 80 Fc receptor blockade, washed platelet
infective, 314, 327 assays, 284
Endothelial cell injury, immune, Femoral artery, white clots, 321
254–255 Fentanyl, 443
Endothelial leukocyte adhesion Fibronectin, 152
molecule (ELAM), 252 Flow cytometry, 228–229, 273, 285
Endothelium, 255–256 Fluid-phase antigen, 273
antibodies, HIT, 259–262 Fluid-phase enzyme immunoassay
hemostasis, 252–254 (EIA), 289–291
platelet factor 4, 256–258 Fluid-phase PF4-heparin-enzyme
Enoxaparin, 119 immunoassay (EIA), 291
Enzyme immunoassay (EIA), 107, 132, Fondaparinux (Arixtra), 17, 135–136,
133, 288 210, 211–212, 343, 581
children, 562–563 Furosemide, 485
fluid-phase, 289–291 FXa, 199
fluid-phase PF4-heparin, schematic
figure, 291 GAG, 179–180, 183, 198, 253
Enzyme-linked immunosorbent assay Glucan sulfates
(ELISA), 186, 401–402 anticoagulant activity, 207–209
Epinephrine, 485 HIT-associated antibodies, 207–209
Epoprostenol, 352 Glycoproteins, 27
Eptifibatide, 443, 485 antigens, 26
Esmolol, 485 inhibitor inducing immune thrombo-
Evan’s syndrome, 28 cytopenia, 35
EVOLUTION, 499 Glycosaminoglycans (GAG), 179–180,
Expert witnesses, malpractice, 576 183, 198, 253
Extracorporeal circuit, 84–85 Graft
Extracorporeal membrane oxygena- rejection, 255
tion, 464–465 thrombosis, 84–85
Index 619
Harenberg, Job, 12 [Heparin]

HAS, 489, 491 PF4, 179–180
HAT, 107 platelet activation
frequency variability, 108 antibody-independent, 152
nonimmune, 154, 338–339 platelet-related prohemorrhagic
HAT-1, 407, 409–412 effects, 153–154
efficacy, 409–412 platelets, 149–152
meta-analysis, 415–418 repeat use
safety, 412 HIT, 62–63
HAT-2, 407, 412–414 structure, 198–199
efficacy, 412–414 Heparin-associated thrombocytopenia
meta-analysis, 415–418 (HAT), 107
safety, 414 frequency variability, 108
HAT-3, 407, 414–415 nonimmune, 154, 338–339
meta-analysis, 417–418 Heparin-Associated Thrombocyto-
Hemodialysis penia (HAT)-1, 407, 409–412
argatroban, 444 efficacy, 409–412
HIT, 463–464 meta-analysis, 415–418
danaparoid sodium, 383–384 safety, 412
dosing schedules, 377 Heparin-Associated Thrombocytope-
HIT, 509–523 nia (HAT)-2, 407, 412–414
clinical presentation, 510–511 efficacy, 412–414
management, 511–523 meta-analysis, 415–418
lepirudin, 407, 426–427 safety, 414
without anticoagulant, 512–513 Heparin-Associated Thrombo-
Hemofiltration cytopenia (HAT)-3, 407,
danaparoid sodium, 383–384 414–415
dosing schedules, 377 meta-analysis, 417–418
lepirudin, 407 Heparin-coated devices, 121–122
Hemoglobinuria, paroxysmal Heparin-dependent antibodies,
nocturnal, 314, 327 168–171
Hemolytic uremic syndrome (HUS), Heparin-dependent antigens,
38 165–172
Hemostasis, endothelium, 252–254 Heparin-dependent platelet activation
Heparan sulfate-containing proteo- endpoints, 276–278
glycans (HSPG), 255–256 interpretation, 276
Heparin, 255–256, 485 test conditions, 275–276
cardiopulmonary bypass, Heparin-induced platelet activation
543–544 (HIPA) assay, 272, 273, 277,
causing thrombosis, 1–5 278, 292, 293–295, 298–299
discontinuation, 336 heparin-induced, 272, 273, 277, 278,
discovery, 1 292, 293–295, 298–299
incidental exposure, 63–64 Heparin-induced skin lesions at
nonimmune subcutaneous heparin injection
platelets, 157–158 sites, 85
620 Index
Heparin-induced thrombocytopenia [Heparin-induced thrombocytopenia

(HIT), 54–70, 262–263, see also (HIT)]
Pseudo-heparin-induced FcgR11a polymorphisms, 238–241
thrombocytopenia (HIT) frequency, 107–138
animal models, 235–236 incidental UFH flushes, 120–121
antibodies medical patients, 112–114, 117
activation assays, 271–287 platelet count monitoring, 136–138
antigen assays, 288–295 pregnancy, 119–120
beta-1,3-glucan sulfates, 207–209 surgical patients, 115–116, 117
clinical HIT, 204–205 variability, 108, 110
c-PRP, 285–286 hemodialysis, 509–523
cross-reactivity, 205–209 heparin
danaparoid sodium, 386–389 discontinuation, 70, 341
epitopes recognition, 182–189 repeat use, 62–63, 350–351
heparin, 130–131 variable duration, 122
immune complexes, 202–203 heparin-coated devices, 121–122
low molecular weight heparin heparin-dependent antigens,
(LMWH), 205–207 165–172
PF4-heparin complex, 202–205 heparin dose-dependence, 123
test discrepancies, 296 heparin-platelet interactions,
washed platelets, 279–282 154–157
antibody-containing immune com- history, 1–19
plexes, 204 hypercoagulable state, 70
antibody seroconversion, IgG, 259, 261, 340
population-based studies, immune
132–136 frequency, 111–123
anticoagulants, 344 therapy, 339–351
vs. APLAS, 323–324 immune basis, 4–5
bone marrow transplantation, 93 immune vascular injury, 251–263
cardiac surgery, 134–135 induction, 197
cardiopulmonary bypass, 351–354 laboratory testing, 9–11, 271–304
children, 91–93, 353–354, 553–567 classification, 273
clinical legal aspects, 573–592
HIT antibodies, 204–205 medical thrombolysis, 354
picture, 53–95 molecular immunopathogenesis,
complications, 85–87 179–192
cardiac and neurological, 91 monocyte Fcgamma receptors,
diagnosis, 579–580 236
doubts, 300 natural history, 71–72
sensitivity-specificity tradeoffs, neonates, 91–93
301 nonimmune, 8–9
endothelial cells antibodies, 259–262 type 1 vs. type 2, 9
estimating pretest probability, orthopedic patients, LMWH vs.
93–95 UFH, 118–119
FcgR111a, 232–236 paradoxical thrombosis, 5–8
Index 621
[Heparin-induced thrombocytopenia [Heparin-induced thrombocytopenia

(HIT)] (HIT)]
pathogenesis, 149–158, 339–341 unrecognized, 109
platelet activation, dynamic model, vascular surgery, 351–354
232–233 venous thrombosis, 73–83
platelet factor4-heparin, 11–12 without thrombosis, anticoagulation,
platelet Fc receptor, 223–241 348–349
pregnancy, 91, 353 Heparin-induced thrombocytopenia
prospective studies, 10–11 (HIT)-associated thrombosis
risk reduction, 17–19 (HITT), 262
sequelae, 54 monocytes, 262
subclinical, antibody seroconversion, Heparin induced thrombocytopenia
137 thrombosis syndrome (HITTS),
sulfated carbohydrates, 207 71–72
sulfated polysaccharides, 122, Heparinization, regional, 512
197–213 Heparin mimetics, 212
test Heparin-platelet factor 4 (HPF4), 156,
interpretation, 295–302 171
thrombocytopenia, 54–70 Heparin-platelet interactions, 154–157
defined, 67 Heparin sulfate, 181
delayed onset, 64–65 HIPA assay, see Heparin-induced
diminishing risk, 59 platelet activation (HIPA) assay
platelet count monitoring, 67 Hirudin, 152, 397–407, 399
rapid onset, 59–60 chemistry, 397–398
severity, 66–70 pharmacokinetics, 398–401
timing, 54–65 pharmacology, 398
typical onset, 54–59 Hirudo medicinalis, 13
thromboembolectomy, 354–355 Hirugen, 399
thrombosis, 70–85, 83, 128 Hirulog Angioplasty Study (HAS), 489,
anticoagulation, 341–348 491
frequency, 123–132 HIT, see Heparin-induced thrombocy-
heparin resistance, 90 topenia (HIT)
longer-term anticoagulants, HITTS, 71–72
349–350 HIV, 27
pathogenesis, 72–73 HPF4, 156, 171
treatment, 12–14 H/PF4-PaGIA, 292
without hemorrhage, 70 HSPG, 255–256
without thrombocytopenia, 69–70 Human immunodeficiency virus (HIV),
thrombotic and hemorrhagic mani- 27
festations, 5–7 Human umbilical vein (HUVEC), 255,
thrombotic complications, timing, 260
71–72 HUS, 38
transient antibodies, 60–61 HUVEC, 255, 260
treatment, 335–360, 580–583 Hydrocortisone, 443
paradoxes, 336, 580–581 Hypercoagulability, congenital, 83
622 Index
Idiopathic autoimmune cytopenia, 36 Interferon-gamma-inducible protein

Idiopathic immune thrombocytopenia, (IP-10), 181
27 Interleukin-8 (IL-8), 171, 172, 181, 184,
Idraparinux, 211–212 187, 232–233
IgA, 170, 203 Interleukin-10 (IL-10), 181
IgG, see Immunoglobulin G (IgG) Intravenous heparin bolus, 89–90
IgM, 170, 191, 203, 260 IP-10, 181
IL-8, see Interleukin-8 (IL-8) Isolated heparin-induced thrombocyto-
IL-10, 181 penia (HIT), 582–583
ILP, 118 natural history, 14–15, 128–132
Immune complexes, 202–203 therapeutic-dose anticoagulation,
Immune endothelial cell injury, 16
254–255 treatment, 14–16
Immune heparin-induced thrombo- Isolated limb perfusion (ILP), 118
cytopenia (HIT) ITAM, 226
frequency, 111–123
therapy, 339–351 Kasabach-Merritt syndrome, 28, 37
Immune thrombocytopenia Kelton, John, 10
abciximab-induced, 35
drug-induced, 33–36 Leech, 13
drugs inducing, 34 Legal aspects, 573–592
therapy, 34 Lepirudin, 16, 342–343, 344, 346, 348,
idiopathic, 27 378, 397–429, 437
passive, 33 acute coronary syndrome, 425
Immune vascular injury allergic reactions, 423–425
HIT, 251–263 antibodies, 422–423
HIT-associated thrombosis, 262 anticoagulation tests, 401–405
Immunofluorescence assays, 303 aPTT ratios, 400
Immunoglobulin A (IgA), 170, 203 vs. argatroban, 418–419
Immunoglobulin G (IgG), 203, 260, cardiopulmonary bypass, 425–426,
355 534–536
agonists, 227–232 children, 427–428, 565
antibodies, 155 clinical studies, 407–409
HIT, 6–7 vs. danaparoid sodium, 419–420
plasma, 233–234 dosing, 402, 405–406
posttransfusion purpura (PTP), hemodialysis, 426–427, 515–516
32–33 percutaneous coronary intervention,
Immunoglobulin M (IgM), 170, 191, 425
203, 260 postmarketing drug monitoring pro-
Immunoreceptor tyrosine-based activa- gram, 418–420
tion motif (ITAM), 226 pregnancy, 427
Incidental unfractionated heparin removal of, 406–407
(UFH) flushes, 120–121 thrombosis, 13
Infectious disease, 238 transitioning to warfarin, 406
Infective endocarditis, 314, 327 vascular surgery, 425–426
Informed consent, 578 Lidocaine, 485
Index 623
Litigation, 573 Monocytes

Livedo reticularis, 88–89, 321 Fcg receptors, 236
LMWH, see Low molecular weight thrombosis, 262
heparin (LMWH) Morphine, 443, 485
Low-dose danaparoid, 336 Mustard, Fraser, 10
Lower limb deep venous thrombosis Myeloproliferative disease, 80
(DVT), 73–75
Low molecular weight heparin Nadroparin, 119
(LMWH), 17, 180, 336, Nafamostat mesilate, hemodialysis,
357–358, 580–581 522
antibodies, 205–207 NAP-2, 171, 181, 187
children, 566–567 pre-existing antibodies, 172
commercial characteristics, 206 Negligence, professional, 573
early vs. late onset thrombocyto- Neoantigens, 166
penia, 111 Neonates, HIT, 91–93
hemodialysis, 512 Neutrophil-activating peptide 2
limitations, 197 (NAP-2), 171, 181, 187
vs. UFH, orthopedic patients, pre-existing antibodies, 172
118–119 Nitric oxide, 252
Luminography Nitroglycerin, 485
ATP, 278 Nonidiosyncratic heparin-induced
ATP release, 273 platelet activation, 152–153
Lupus anticoagulants, 321 Nonimmune heparin, platelets,
Lupus anticoagulant syndrome, 271, 157–158
272, 321–324 Nonimmune heparin-associated throm-
Lymphoproliferative disorders, 27 bocytopenia (nonimmune
HAT), 154, 338–339
MAIPA assay, 39–40, 301 Nonimmune heparin-induced throm-
Malpractice, 573 bocytopenia (HIT), 8–9
burden of proof, 576–577 type 1 vs. type 2, 9
expert witnesses, 576 Nonimmune heparin-platelet inter-
standard of care, 575–576 actions, 149–158
McLean, Jay, 1 Novastan, see Argatroban (Novastan)
Medicolegal system, 575–577
Melphalan, 118 Oligosaccharides, 211–212
Meningococcemia, 326 Onyalai, 30
Methyldopa, 36 Oral anticoagulants, 358–359
Metoprolol, 443 hemodialysis, 521–522
Meyer, Dominique, 11–12 Orgaran, see Danaparoid sodium
Microparticles, 228–229 (Orgaran)
platelet-derived, 279
Midazolam, 443 PAF, 252
Monoclonal antibody immobilization PAI-1, 495
of platelet antigens (MAIPA) Paradoxical thrombosis, 5–8
assay, 39–40 Paroxysmal nocturnal hemoglobinuria,
thrombocytopenia, 301 314, 327
624 Index
Particle gel immunoassay (ID-H/PF4 [Platelet activation]

test), 273, 291–293 endpoints, 276–278
Passive immune thrombocytopenia, interpretation, 276
33 test conditions, 275–276
PAT, 273 interpretation by HIT serum in
PBMC, 190 absence of added heparin,
PBP, 181 284–285
PCI, see Percutaneous coronary inter- PF4-H antibodies, 170
vention (PCI) Platelet aggregation assays, 9–10
Pentasaccharides, 210–211 Platelet aggregation test (PAT),
anti-Xa-inhibiting, 135–136 273
Percutaneous coronary intervention Platelet antibodies, 39–42
(PCI) Platelet autoantibodies and alloanti-
argatroban, 444, 460–463 bodies, 39–40
bivalirudin (Angiomax), 490–493, Platelet basic protein (PBP), 181
497–498 Platelet-bound antibodies, 40–41
lepirudin, 425 Platelet count, monitoring, 578–579
Percutaneous transluminal coronary frequency, 136–138
angioplasty (PTCA) Platelet-derived microparticle genera-
bivalirudin (Angiomax), 494 tion, 279
danaparoid sodium, 377 Platelet dysfunction, acquired anti-
Peripheral arterial obstructive disease, body-mediated, 29–30
444 Platelet factor 4 (PF4), 155, 165, 262
Peripheral blood mononuclear cells antigenicity, 166–168
(PBMC), 190 endothelium, 256–258
PF4, see Platelet factor 4 (PF4) heparin, 179–180
PF-4H, see Platelet factor 4-heparin polyvinylsulfonate antigen assay,
(PF4-H) 289
PF4-M2, 184 structure, 198
Phenprocoumon, see Coumarin sulfated carbohydrates, 198–202
Phenylephrine, 443 sulfated polysaccharides, 199–201
Phlegmasia, 14 in vivo, 201–202
Plasma immunoglobulin G (IgG), Platelet factor 4-heparin (PF4-H), 201,
233–234 223, 271
heat inactivation, 279 antibodies, 202–205
Plasmapheresis, 355–356 platelet activation, 170
Plasma-soluble FcgR111a, 233 reactive, 169
Plasminogen activator inhibitor-1 Platelet factor 4-heparin enzyme
(PAI-1) immunoassay (PF4-H-EIA),
bivalirudin (Angiomax), 495 273, 293–295, 298–299
Platelet-activating antibodies, 7–8 thrombocytopenia, 298–299
Platelet-activating factor (PAF), 252 Platelet factor 4-M2 (PF4-M2), 184
Platelet activation Platelet Fc receptor, 223–241
Fcg11a receptor-mediated, 227–232 Platelet glycoprotein IIb/IIIa inhibi-
heparin-dependent tors, 357
Index 625
Platelet inhibition, cardiopulmonary PRP, see Platelet-rich plasma (PRP)

bypass, 541–542 P-selectin, 252
Platelet microparticle assay, 273 Pseudo-heparin-induced thrombocyto-
Platelet-rich plasma (PRP), 273, 275, penia (HIT), 313–330, 580
287 HIT, 328–329
Platelet-rich plasma (PRP) aggregation treatment, 328–330
test, 233–234 PTCA
Platelets bivalirudin (Angiomax), 494
circulated platelet-rich plasma danaparoid sodium, 377
(c-PRP), 273 PTP, see Posttransfusion purpura
heparin, 149–152 (PTP)
nonimmune heparin, 157–158 Pulmonary embolism, 12, 314, 318, 319,
transfusions, 336, 359–360, 580 320
Platelet specific alloantigens, posttrans- Purpura fulminans, septicemia, 314,
fusion purpura (PTP), 31 325–327
Polyanion, 182–186
Polysaccharides, sulfated Quinidine
HIT, 122, 197–213 antibodies, 42
PF4, 201–202 drug-induced immune thrombocyto-
Porcine mucosal unfractionated hepa- penia (DIT), 34
rin (UFH), 117–118 Quinine, drug-induced immune throm-
Postoperative orthopedic patients, bocytopenia (DIT), 34
56–57
Posttransfusion purpura (PTP), 31–33, Raynaud’s phenomenon, 80
314, 327–328 Recombinant hirudin (r-hirudin), see
clinical picture, 32 Lepirudin
diagnosis, 32 Regional citrate anticoagulation,
pathogenesis, 31–32 hemodialysis, 523
therapy, 32–33 Regional heparinization, hemodialysis,
Potassium, 485 512
Pregnancy, 80 Reteplase, 488
argatroban, 465–466 Reviparin (Clexane), 119
bivalirudin (Angiomax), 495 Rheumatoid arthritis, 27
danaparoid sodium, 384–385 r-Hirudin, see Lepirudin
HIT, 91, 353 Rhodes, Glen R., 5
frequency, 119–120 Roberts, Brooke, 3–4
lepirudin, 427
Primary biliary cirrhosis, 27 Sarcoidosis, 27–28
Prochlorperazine, 488 Scintillation counter, 278
Professional negligence, 573 Secondary autoimmune thrombocyto-
Prostacyclin, 152, 252 penic purpura (AITP), 27
hemodialysis, 522 Septicemia
Prosthetic device, 84–85 consumptive thrombohemorrhagic
Protein, 186–189 disorders, 36–37
Protein C receptor, 253 purpura fulminans, 314, 325–327
626 Index
Serotonin release, flow cytometry, Thrombin, 153

273 inhibition, 399
Serotonin release assay (SRA), 10, 107, Thrombin-antithrombin (TAT)
132, 133, 180, 186, 271–272, 273, complexes, 78
285, 292, 293–295 Thrombocytopenia, see also Heparin-
thrombocytopenia, 298–299 associated thrombocytopenia
Serum, heat inactivation, 279 (HAT); Heparin-induced
Sheridan, Dave, 10 thrombocytopenia (HIT);
Silver, Donald, 5, 6 Immune thrombocytopenia
Skin necrosis abciximab-induced immune, 35
classic coumarin-induced, 87–88 acute, 25–42
coumarin-induced, 87–88 acute postinfectious autoimmune,
without coumarin therapy, 88 27
Sodium bicarbonate, 485 APLAS, 324
Solid-phase PF4-heparin-EIA, 288 autoimmune, 27
Solid-phase red cell adherence assay cyclic, 30
(SPRCA), 303–304 early vs. late onset, 110–111
Splenectomy, AITP, 29 familial, 28
SPRCA, 303–304 heparin-induced platelet activation
SRA, see Serotonin release assay (HIPA) assay, 298–299
(SRA) HIPA test, 298–299
Standard of care HIT-IgG, 259
evolving, 582–583 idiopathic immune, 27
malpractice, 575–576 laboratory results, 299–302
Streptokinase, 354, 488 monoclonal antibody immobiliza-
Stroke tion of platelet antigens
argatroban, 444, 464 (MAIPA) assay, 301
danaparoid sodium, 374 passive immune, 33
Subclinical heparin-induced thrombo- PF4-H-EIA, 298–299
cytopenia (HIT), 137 rapid vs. typical onset, 297–299
Sulfated carbohydrates SRA, 298–299
HIT, 207 in vitro cross-reactivity, 302–303
PF4, 198–202 activation assays, 302–303
Sulfated polysaccharides washed platelet assay, 298–299
HIT, 122, 197–213 Thromboembolectomy, 354–355
PF4, 201–202 Thromboembolism, see Arterial
Surface-bound antigen, 273 thromboembolism
Systemic lupus erythematosus, 27, 28, Thromboglobulin (TG), 171
255 Thrombolytic therapy, 314, 324–325
Thrombomodulin, 258
Target antigen, 273 Thrombosis, 83, 128, 251
TAT complexes, 78 frequency, 123–132
T-cell receptor (TCR), 189 heparin resistance, 90
TCR, 189 paradoxical, 5–8
TG, 171 treatment, 12–14
Index 627
Thrombotic thrombocytopenic Venous thromboembolism, 75

purpura (TTP), 28, 38 danaparoid sodium, 374, 375–376
Thromboxane A2, 227 dosing schedules, 377
Tirofiban, 352, 443, 485 prophylaxis, 380
Tissue plasminogen activator (t-PA), Verapamil, 443, 485
252, 354 Viral hepatitis, 27
Tobin, Richard W., 1–2 Vitamin K antagonists, 358–359
Towne, Jonathan, 8 Von Willebrand factor (vWF), 38
T-PA, 252, 354 factor VIII, 407
TTP, 28, 38 platelet function, 153
VWF, 38
UFH, see Unfractionated heparin factor VIII, 407
(UFH) platelet function, 153
Unfractionated heparin (UFH),
165–166, 179, 352 Warfarin, see Coumarin
early vs. late onset thrombo- Warm-type autoimmune hemolytic
cytopenia, 111 anemia, 28
HIT frequency Washed platelet assays, 271–272, 273,
incidental flushes, 120–121 293–295
porcine mucosal, 117–118 vs. c-PRP activation assay, 280, 287,
limitations, 197 302–303
Unstable angina disadvantages, 285
argatroban, 445 heparin-independent platelet activa-
bivalirudin (Angiomax), 493–494 tion, 282
Upper limb deep venous thrombosis hierarchical vs. idiosyncratic platelet
(DVT), 75 activation by HIT sera, 282, 283
Urokinase, 354 inhibition, 284
Urokinase-type plasminogen activator, preparation, 272–273
253 quality control, 282–284
Urticaria, 89 schematic overview, 274
thrombocytopenia, 298–299
Vancomycin, 488 Washed platelets
Vascular restenosis, 495 aggregation, 278
Vascular surgery HIT antibodies, 279–282
HIT, 351–354 Weismann, Roger E., 1–2
lepirudin, 425–426 White clot syndrome, 8
Vena cava filters, 336, 360 Witnesses, expert, 576

10 1 1 463 3921

Uploaded by

Copyright:

Available Formats

10 1 1 463 3921

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

10 1 1 463 3921

Uploaded by

Copyright:

Available Formats

Heparin-Induced

Trademark notice: Product or corporate names may be trademarks or registered

Library of Congress Cataloging-in-Publication Data

This book is printed on acid-free paper.

Distribution and Customer Service

Eastern Hemisphere Distribution

World Wide Web

Copyright n 2004 by Marcel Dekker, Inc. All Rights Reserved.

Current printing (last digit):

PRINTED IN THE UNITED STATES OF AMERICA

Associate Editor, Europe

1. Drug Treatment of Hyperlipidemia, edited by Basil M. Rifkind

ADDITIONAL VOLUMES IN PREPARATION

To my co-workers and students for their contributions and eﬀorts; to Sabine,

Although I used to think heparin-induced thrombocytopenia (HIT) was an

Since the appearance of the second edition of Heparin-Induced Thrombocy-

threatening anaphylaxis, perhaps because of antibodies that have been

Since the ﬁrst edition of Heparin-Induced Thrombocytopenia appeared, there

An anticoagulant turns procoagulant; an antithrombotic causes thrombosis.

and Hemostasis in Amsterdam, focused on improvements and innovations in

Figure 1 The iceberg model of heparin-induced thrombocytopenia as an index for

We would also like to acknowledge the help of many individuals in this

Amiral J, Bridey F, Dreyfus M, Vissac AM, Fressinaud E, Meyer D. Identiﬁcation of

Series Introduction (Samuel Z. Goldhaber) v

1. History of Heparin-Induced Thrombocytopenia 1

2. Diﬀerential Diagnosis of Acute Thrombocytopenia 25

3. Clinical Picture of Heparin-Induced Thrombocytopenia 53

4. Frequency of Heparin-Induced Thrombocytopenia 107

5. Nonimmune Heparin–Platelet Interactions: Implications

6. Heparin-Dependent Antigens in Heparin-Induced

7. Molecular Immunopathogenesis of Heparin-Induced

8. Role of Sulfated Polysaccharides in the Pathogenesis of

9. The Platelet Fc Receptor in Heparin-Induced

10. Immune Vascular Injury in Heparin-Induced

11. Laboratory Testing for Heparin-Induced

12. Pseudo–Heparin-Induced Thrombocytopenia 313

13. Treatment of Heparin-Induced Thrombocytopenia: An

14. Danaparoid for the Treatment of Heparin-Induced

15. Lepirudin for the Treatment of Heparin-Induced

16. Argatroban Therapy in Heparin-Induced

17. Bivalirudin for the Treatment of Heparin-Induced

18. Hemodialysis in Heparin-Induced Thrombocytopenia 509

19. Management of Cardiopulmonary Bypass

20. Heparin-Induced Thrombocytopenia in Children 553

21. Legal Aspects of Heparin-Induced Thrombocytopenia:

22. Legal Aspects of Heparin-Induced Thrombocytopenia:

Susanne Alban, Ph.D. Senior Researcher and Lecturer, Institute of Phar-

Jean Amiral, Ph.D. Scientiﬁc Director, HYPHEN BioMed, Neuville-sur-

Gowthami M. Arepally, M.D. Assistant Professor, Department of Medicine,

Richard H. Aster, M.D. Professor, Department of Medicine, Medical

John R. Bartholomew, M.D. Program Director, Section of Vascular Med-

Beng Hock Chong, M.B.B.S., Ph.D., F.R.C.P.(Glasgow), F.R.A.C.P,

Douglas B. Cines, M.D. Professor, Department of Pathology and Labora-

Gregory A. Denomme, Ph.D. Assistant Professor, Department of Labora-

Scientist, Department of Pathology and Laboratory Medicine, Mount Sinai

Karl-Georg Fischer, M.D. Division of Nephrology and General Medicine,

Andreas Greinacher, M.D. Professor, Institute for Immunology and Trans-

McDonald K. Horne III, M.D. Senior Clinical Investigator, Hematology

Marcie J. Hursting, Ph.D., D.A.B.C.C. Director, Clinical Science Consult-