Practical Transfusion Medicine, 2nd edition

Practical Transfusion Medicine EDITED BY

Michael F. Murphy MD, FRCP, FRCPath Professor of Blood Transfusion Medicine University of Oxford Consultant Haematologist National Blood Service and Department of Haematology The John Radcliffe Hospital, Oxford

Derwood H. Pamphilon MD, MRCPCH, FRCP, FRCPath Consultant Haematologist Institute for Transfusion Sciences National Blood Service Bristol

FOREWORD BY

D.J. Weatherall Second edition

Practical Transfusion Medicine

Practical Transfusion Medicine EDITED BY

Michael F. Murphy MD, FRCP, FRCPath Professor of Blood Transfusion Medicine University of Oxford Consultant Haematologist National Blood Service and Department of Haematology The John Radcliffe Hospital, Oxford

Derwood H. Pamphilon MD, MRCPCH, FRCP, FRCPath Consultant Haematologist Institute for Transfusion Sciences National Blood Service Bristol

FOREWORD BY

D.J. Weatherall Second edition

© 2005 by Blackwell Publishing Ltd Blackwell Publishing, Inc., 350 Main Street, Malden, Massachusetts 02148-5020, USA Blackwell Publishing Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK Blackwell Publishing Asia Pty Ltd, 550 Swanston Street, Carlton, Victoria 3053, Australia The right of the Author to be identified as the Author of this Work has been asserted in accordance with the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher. First published 2001 Second edition 2005 Library of Congress Cataloging-in-Publication Data Practical transfusion medicine / edited by Michael F. Murphy, Derwood H. Pamphilon. — 2nd ed. p. ; cm. Includes bibliographical references and index. ISBN 1-4051-1844-X 1. Blood— Transfusion. [DNLM: 1. Blood Transfusion. 2. Blood Grouping and Crossmatching. 3. Communicable Disease Control. 4. Specimen Handling. WB 356 P8957 2005] I. Murphy, Michael F. (Michael Furber) II. Pamphilon, Derwood H. RM171.P727 2005 615¢.39— dc22 2004016676 ISBN-13: 978-1-4051-184-46 ISBN-10: 1-4051-184-4X A catalogue record for this title is available from the British Library Set in 9.5 on 12 pt Sabon by SNP Best-set Typesetter Ltd., Hong Kong Printed and bound in India by Gopsons Paper Ltd, Noida Commissioning Editor: Maria Khan Production Editor: Rebecca Huxley Production Controller: Kate Charman Project Manager: Richard Lawrence For further information on Blackwell Publishing, visit our website: http://www.blackwellpublishing.com The publisher’s policy is to use permanent paper from mills that operate a sustainable forestry policy, and which has been manufactured from pulp processed using acid-free and elementary chlorine-free practices. Furthermore, the publisher ensures that the text paper and cover board used have met acceptable environmental accreditation standards.

Contents

List of Contributors, vii Foreword, xi Preface to the second edition, xiii Preface to the first edition, xv

Part 1 Basic principles of transfusion 1 Introduction Ian M. Franklin, 3 2 Essential immunology for transfusion medicine Willem H. Ouwehand and Tim B. Wallington, 13 3 Human blood group systems Geoff Daniels, 24 4 Human leucocyte antigens Cristina V. Navarrete, 34 5 Platelet and neutrophil antigens David L. Allen, Geoffrey F. Lucas, Willem H. Ouwehand and Michael F. Murphy, 50

Part 3 Complications of transfusion 13 Haemolytic transfusion reactions Sue Knowles and Geoff Poole, 161 14 Febrile reactions and transfusion-related acute lung injury Michael F. Murphy and Sheila MacLennan, 171 15 Urticarial and anaphylactic reactions David J. Unsworth, 179 16 Bacterial contamination Patricia E. Hewitt, 184 17 Post-transfusion purpura Michael F. Murphy, 191 18 Immunomodulation and graft-versus host disease Lorna M. Williamson and Cristina V. Navarrete, 195 19 Transfusion-transmitted infections Alan D. Kitchen and John A.J. Barbara, 208 20 Variant Creutzfeldt–Jakob disease Marc L. Turner, 229

Part 2 Clinical transfusion practice 6 The effective and safe use of blood components Brian McClelland and Tim Walsh, 67 7 Bleeding associated with trauma and surgery Beverley J. Hunt, 86 8 Prenatal and childhood transfusions Irene Roberts, 97 9 Haematological disease Michael F. Murphy and Simon J. Stanworth, 119 10 Transfusion strategies in organ transplant patients Derwood H. Pamphilon, 132 11 Inherited and acquired coagulation disorders Joanne E. Joseph and Samuel J. Machin, 138 12 Uses of intravenous immunoglobulin David J. Unsworth and Tim B. Wallington, 151

Part 4 Practice in blood centres and hospitals 21 Donors and blood collection Liz Caffrey and Moji Gesinde, 241 22 Blood donation testing and the safety of the blood supply David Wenham and Simon J. Stanworth, 250 23 Production and storage of blood components Lorna M. Williamson and Rebecca Cardigan, 259 24 Medicolegal aspects Patricia E. Hewitt, 274 25 Blood transfusion in hospitals Sue Knowles and Geoff Poole, 280 26 Autologous transfusion Dafydd Thomas, 298 27 Tissue banking Deirdre Fehily and Ruth M. Warwick, 309 v

Contents

28 Cord blood banking Ruth M. Warwick, Sue Armitage and Deidre Fehily, 320 29 Therapeutic apheresis Tim B. Wallington and David J. Unsworth, 328

Part 5 Developments in transfusion medicine 30 Blood substitutes Chris V. Prowse and David J. Roberts, 341 31 Cytokines in transfusion practice Derwood H. Pamphilon, 350 32 Haemopoietic stem cell processing and storage David H. McKenna and Mary E. Clay, 357 33 Haemopoietic stem cell transplantation and immunotherapy Ian M. Franklin, 369 34 Gene therapy Colin G. Steward and Marina Cavazzana-Calvo, 390

vi

35 Recombinant antibodies and other proteins Marion Scott, 403 36 Blood transfusion in a global context David Roberts, Jean-Pierre Allain, Alan Kitchen, Stephen Field and Imelda Bates, 415 37 The design of interventional trials in transfusion medicine Paul Hébert, Alan Tinmouth and Dean Fergusson, 424 38 Getting the most out of the evidence for transfusion medicine Simon J. Stanworth, Susan J. Brunskill and Chris J. Hyde, 436 39 The future of transfusion medicine Walter Sunny Dzik, 445 Index, 457 Colour plates are found between pp. 304 and 305

List of Contributors

Jean-Pierre Allain

Rebecca Cardigan

Dean Fergusson

Division of Transfusion Medicine National Blood Service University of Cambridge Long Road Cambridge, CB2 2PT

National Blood Service Cresent Drive Brentwood Essex CM15 8DP

Centre for Transfusion Research University of Ottawa Ottawa, Ontario K1M 8L6 Canada

David L. Allen

Marina Cavazzano-Calvo

National Blood Service John Radcliffe Hospital Headley Way Headington Oxford, OX3 9DU

Director of Biotherapy Hopital Necker-Enfants Malades 149 Rue des Sevres Cedex 15 Paris 75743 France

Susan Armitage

Mary E. Clay

National Blood Service Cord Blood Bank Deansbrook Road Edgeware Middlesex, HA8 9DB

Department of Laboratory Medicine & Pathology MMC 198 Room D-251 Mayo Building University of Minnesota Medical School 420 Delaware Street SE Minneapolis, MN 55455 USA

Stephen Field National Blood Service North London Colindale Avenue Colindale London, NW9 5BG

Ian M. Franklin

John A.J. Barbara National Blood Service North London Colindale Avenue Colindale London, NW9 5BG

Geoff Daniels Bristol Institute for Transfusion Sciences National Blood Service Southmead Road Bristol BS10 5ND

Imelda Bates Liverpool School of Tropical Medicine Pembroke Place Liverpool L3 5QA

Susan Brunskill National Blood Service John Radcliffe Hospital Headley Way Headington Oxford, OX3 9BQ

Walter Sunny Dzik Massachusetts General Hospital Blood Transfusion Service, J-224 55 Fruit Street Boston, MA 02446 USA

Deirdre Fehily National Transplant Centre Via Giano della Bella 00161 Rome Italy

Department of Medicine University of Glasgow Royal Infirmary Glasgow G31 2ER

Moji Gesinde National Blood Service Leeds Blood Centre Bridle Path Leeds LS15 7TW

Paul Hébert Department of Medicine The Ottawa Hospital/General Campus 501 Smyth Road Room 1812H, Box 201 Ottawa, ON K1H 8L6 Canada

Patricia E. Hewitt National Blood Service North London Colindale Avenue Colindale London, NW9 5BG

Elizabeth Caffrey

Beverly J. Hunt

National Blood Service University of Cambridge Long Road Cambridge CB2 2PT

Department of Haematology & Rheumatology Guy’s and St Thomas’ Foundation Trust Lambeth Palace Road London SE1 7EH

vii

List of Contributors

Christopher J. Hyde

David H. McKenna

David J. Roberts

National Blood Service John Radcliffe Hospital Headley Way Headington Oxford, OX3 9BQ

Clinical Cell Therapy Laboratory University of Minnesota Medical School 1900 Fitch Avenue St Paul, MN 55108 USA

National Blood Service John Radcliffe Hospital Headley Way Headington Oxford, OX3 9BQ

Joanne E. Joseph

Sheila MacLennan

Irene A.G. Roberts

Department of Haematology and Stem Cell Transplantation St Vincent’s Hospital Victoria Street Darlinghurst, NSW 2010 Australia

National Blood Service Leeds Blood Centre Bridle Path Leeds LS15 7TW

Department of Haematology Commonwealth Building, 4th Floor Imperial College Hammersmith Campus Du Cane Road London, W12 0NN

Michael F. Murphy

Marion Scott

National Blood Service John Radcliffe Hospital Headley Way Headington Oxford, OX3 9BQ

Bristol Institute for Transfusion Sciences National Blood Service Southmead Road Bristol BS10 5ND

Cristina V. Navarrete

Simon J. Stanworth

National Blood Service North London Colindale Avenue Colindale London, NW9 5BG

National Blood Service John Radcliffe Hospital Headley Way Headington Oxford, OX3 9BQ

Willem H. Ouwehand

Colin G. Steward

National Blood Service Southmead Road Bristol, BS10 5ND

National Blood Service University of Cambridge Long Road Cambridge CB2 2PT

BMT Unit Royal Hospital for Sick Children Upper Maudlin Street Bristol BS2 8BJ

Samuel J Machin

Derwood H. Pamphilon

Dafydd Thomas

Department of Haematology University College London Medical School 3rd Floor Cecil Fleming House Grafton Way London, WC1E 6DB

Bristol Institute for Transfusion Sciences National Blood Service Southmead Road Bristol BS10 5ND

Morriston Intensive Care Unit Swansea NHS Trust Morriston Hospital Swansea

Brian McClelland

Geoff Poole

Edinburgh and SE Scotland Blood Transfusion Centre National Science Laboratory 21 Ellen’s Glen Road Edinburgh, EH17 7QT

National Blood Service Southmead Road Bristol BS10 5ND

Alan D. Kitchen National Blood Service North London Colindale Avenue Colindale London, NW9 5BG

Susan Knowles Department of Haematology Epson & St Helier University Hospitals NHS Trust Wrythe Lane Carshalton Surrey, SM5 1AA

Geoffrey F. Lucas

Alan Tinmouth

Christopher V. Prowse Scottish National Blood Transfusion Service National Science Laboratory 21 Ellen’s Glen Road Edinburgh EH17 7QT

viii

Centre for Transfusion Research University of Ottawa Ottawa, Ontario K1H 8L6 Canada

List of Contributors

Marc L. Turner

Tim Walsh

David Wenham

Edinburgh and SE Scotland Blood Transfusion Centre Royal Infirmary of Edinburgh 51 Little France Crescent Old Dalkeith Road Edinburgh, EH16 4SA Scotland

Department of Anaesthetics Critical Care and Pain Medicine New Royal Infirmary of Edinburgh Little France Edinburgh EH16 4SU

National Blood Service North London Colindale Avenue Colindale London, NW9 5BG

Ruth M. Warwick

Lorna M. Williamson

National Blood Service Tissue Services Deansbrook Road Edgeware Middlesex, HA8 9DB

Division of Transfusion Medicine National Blood Service University of Cambridge Long Road Cambridge, CB2 2PT

David J. Unsworth National Blood Service Southmead Road Bristol BS10 5ND

Tim B. Wallington National Blood Service Southmead Road Bristol BS10 5ND

ix

Foreword

Although we now take blood transfusion very much for granted, and it is an integral part of clinical practice, the early days of its development were anything but smooth. Indeed, it seems likely that it spawned two of the earliest documented cases of scientific fraud and medical malpractice. In 1654 a Florentine physician, Francesco Folli, claimed that he had invented blood transfusion and even published a book many years later to illustrate the complex equipment which he had used. He subsequently confessed that he had not yet done the experiment and, as far as is known, he never did! The first well-documented blood transfusions were carried out in 1667, in Oxford and Paris. The Oxford experiments were the work of the physician, Richard Lower, who was stimulated by the studies of the architect, astronomer and polymath Christopher Wren, who had invented a series of cannulas for injecting drugs into the veins of animals. Lower’s first successful transfusion was from the cervical artery of one dog into the jugular vein of another, previously exsanguinated. Perhaps stimulated by news of the first transfusion involving a human being in Paris in the same year, two years later Lower injected a small amount of sheep blood into a mildly deranged clergyman before an admiring audience at the Royal Society. The patient survived and claimed to feel better. In the same year a French physician, Jean-Baptiste Denis, began a series of experiments in which he transfused varying amounts of animal blood into patients with mental illnesses. Things seemed to go well until he gave repeated injections of the blood of ‘a gentle calf’ to a lunatic; the patient had a typical transfusion reaction and, although he recovered temporarily, died two months later. Denis’s enemies persuaded the patient’s wife to bring a legal action against him but, in the event, the defence was successful in proving that the man had been poisoned with arsenic by his wife! Readers of this book might wish to remind their counsels of this possibility next time they are

facing legal proceedings for a mismatched transfusion. Although over the next two centuries there was considerable progress in developing better ways of transferring blood from one individual to another, it was the beautifully elegant studies of Karl Landsteiner, carried out just over 100 years ago, that formed the basis for modern immunohaematology and the successful development of blood transfusion. It is a remarkable fact that some of the most important discoveries that have changed medical practice have been based on extremely small-scale and simple experiments; the demonstration of the efficacy of penicillin required only eight mice, four treated and four controls. Landsteiner described blood groups A, B and C (later called O), using serum and red cells from six healthy males; the results were confirmed with sera from 16 other healthy individuals. Group AB was discovered a year after Landsteiner’s classical experiment, the M, N and P groups were reported by Landsteiner and Levine in 1927, and the rhesus system was characterized by Levine and Weiner in the early 1940s. More than 250 red cell antigens have now been described and most belong to one of 29 systems. Many of the genes that regulate these systems have been identified and much is known about the structure of the blood group antigens and the molecular basis for their diversity. Yet despite all this sophisticated knowledge, and equally tantalising information from studies of varying susceptibility of individuals with different blood groups to a wide variety of diseases, we still don’t know why we have blood groups and in many cases have little understanding of their biological function. Since the Second World War blood transfusion medicine has changed dramatically and is likely to undergo even more dramatic developments as the new millennium evolves. Rapid progress towards the definition of subpopulations of stem cells, an increasing ability to alter the properties of cell popxi

Foreword

ulations by recombinant DNA technology and the vista, if distant, of specific organ therapy based on work on human embryonic stem cells all point to an extremely exciting future for the field. There seems little doubt that, given their expertise in handling and storing cells, blood transfusion specialists will play an increasing role in the practical applications of these new advances in cell biology. Any young person who enters the field over the next few years can be guaranteed an exciting future. The second edition of this fine book provides a comprehensive and practical account of modern blood transfusion practice. While encompassing descriptions of some of the scientific developments

which are occurring at the fringes of the speciality, it focuses mainly on the problems that are encountered daily in transfusion medicine. Because this field abuts on almost every aspect of clinical practice it should be of value to a wide range of clinicians as well as to students and practitioners of transfusion medicine. Whether by accident or intent the first edition of this book appeared exactly 100 years after Karl Landsteiner made his seminal observations. What better tribute could there have been to one of the most beautifully simple and clinically important experiments in the history of medicine, a view confirmed by the early appearance of this second edition. I wish it all the success it deserves. D.J. Weatherall Oxford, July 2004

xii

Preface to the second edition

This second edition has become necessary because of rapid changes in transfusion medicine over the last three years. The pace of change seems likely to increase with new scientific and technological developments, the challenge of ‘emerging’ pathogens, and renewed efforts to improve clinical transfusion practice. In the UK, the implications of the probable transmission of variant Creutzfeldt–Jakob disease by blood transfusion are wide-ranging across the whole transfusion chain from donor to patient. The primary aim of the second edition remains the same as the first, that is to provide a comprehensive guide to transfusion medicine. The book includes information in more depth than contained within handbooks of transfusion medicine, and is presented in a more concise and ‘userfriendly’ manner than standard reference texts. The feedback we received on the first edition from reviews and colleagues was that this objective was achieved, and that we had provided a consistent style and format throughout the book. We have strived to maintain this to provide a text that will be useful to the many clinical and scientific staff, both established practitioners and trainees, who are involved in some aspect of transfusion medicine and who require an accessible text. The book is again divided into five sections which systematically take the reader through the

principles of transfusion medicine, the use of transfusion in specific clinical areas, its practical aspects in blood centres and hospitals, the complications of transfusion and potential advances. This latter section in the first edition was particularly well received, and it is expanded in the second edition with new chapters on stem cell processing, recombinant antibodies/proteins, transfusion in the tropics, design of clinical trials in transfusion medicine, and a final chapter reviewing advances since 1995 and ‘horizon scanning’ about likely future developments up to 2010 and beyond. We are very grateful to the colleagues who have contributed to this book at a time of continuing change. Although, as with the first edition, most authors work in the blood services in the UK, contributors for the second edition include those in full-time clinical practice, and colleagues from outside the UK to provide a broader perspective. We acknowlege the contribution to Practical Transfusion Medicine of two colleagues, Cynthia Beatty and Gail Williams, who were unable to update their chapters because of new commitments. We are grateful to Janet Birchall and Simon Stanworth for providing critical comments on several chapters, and to Helen Williams for her invaluable assistance. We have again received enormous support from our publishers, particularly Maria Khan, Rebecca Huxley and Claire Bonnett. Michael Murphy Derwood Pamphilon 2004

xiii

Preface to the first edition

Blood transfusion continues to enjoy an ever increasing public profile. This has occurred in part because of the emergence of new pathogens which have posed a significant threat to the safety of the blood supply, and also due to major scientific developments. In the new millennium advances in technology have facilitated the provision of highquality blood components and a range of sophisticated diagnostic and specialist services within modern blood centres. There has been enormous progress in transfusion medicine which has developed into a specialist area of its own in the last decade. It now encompasses many important areas of medicine including haematology, immunology, transplantation science, microbiology, epidemiology, clinical practice and research and development. In this book we have aimed to provide a comprehensive guide to transfusion medicine. This includes information in more depth than contained within handbooks of transfusion medicine, but at the same time presented in a more concise and ‘user-friendly’ manner than standard reference texts. Ably assisted by many expert colleagues, we have compiled a text which should prove invaluable to haematologists in training as well as consultants in established practice. We have also aimed to provide useful information to oncologists, surgeons, anaesthetists and other clinicians, nursing staff in general and specialist units and scientific and technical staff in haematology and blood transfusion.

We have endeavoured to provide information that defines practical approaches to the problems that are encountered in transfusion medicine. To this end we have used a consistent format to make access to information easy, irrespective of whether the book is read cover to cover by haematologists updating or revising for exams, or used as a reference book by clinical or laboratory staff faced with specific problems. To facilitate this approach the book is divided into five sections which systematically take the reader through the principles of transfusion medicine, the use of transfusion in specific clinical areas, its practical aspects in blood centres and hospitals, the complications of transfusion and potential advances, some of which are already with us and some of which will continue to impact significantly on transfusion services in the future. We are grateful to the colleagues who have contributed to this book at a time of rapid development and considerable organizational change in healthcare as a whole but specifically within blood services in the UK. We are indebted to Bridget Hunt and Susan Sugden for their patience and forbearance; without their invaluable assistance in compiling the text this book would not have been possible. We have received enormous support from our publishers, particularly Andrew Robinson, who gave us considerable assistance at a time when this book was at its early conceptual stages, and Marcela Holmes whose wisdom and expertise have been invaluable in its completion. Michael Murphy Derwood Pamphilon 2001

xv

Part 1

Basic principles of transfusion

Chapter 1

Introduction Ian M. Franklin

Nearly 4 years have elapsed since the first edition of this book. Has anything occurred in the world of transfusion medicine to alter the concepts that were important then? In the past 12 months, a number of crucial events have occurred that have again acted to increase global anxiety about the safety of blood transfusion. The arrival of severe acute respiratory syndrome (SARS) and its prompt recognition as a novel coronavirus in early 2003 focused attention on the difficulties of maintaining blood safety in the face of an unknown emerging infection. In the absence of any knowledge of the epidemiology of the infection, it had to be assumed that there was the potential for SARS to be transmitted by blood. This remains an unresolved issue that will have to await a better understanding of the virus, which perhaps may be obtained in any new outbreak in 2004 or later. Anxieties over SARS were followed quickly by the expected US summer epidemic of West Nile virus (WNV), known to be a transfusion-transmitted infection, and for which precautions in the USA and Europe were urgent and needed to be robust. These included the use of nucleic acid testing for WNV genome in all donations in the USA, Canada and Mexico. In Europe, recent visitors to North America were not accepted as donors for 4 weeks after return. Most recently, after a few years in which the expected major epidemic failed to materialize, the possibility that variant Creutzfeldt–Jakob disease (vCJD) may well be a transfusion-transmitted infection in humans became more likely, following worrying results in sheep transfusion studies some years ago. A patient, one of only 48 known to be at risk through receiving a labile blood component from a donor who later developed vCJD, devel-

oped and died of vCJD in 2003, 7 years after receiving the blood. The donor was healthy at the time of donation in 1996, but became unwell and died of vCJD in 2000. In the absence of a blood test for vCJD, and with no way of confirming that the two patients had the same or different ‘strains’ of vCJD, this is not conclusive evidence for transmission. On the balance of probabilities, however, it seems likely that the transfusion recipient acquired vCJD from the blood transfusion. This event has triggered a further round of new initiatives in the UK to protect blood safety and retain confidence in the transfusion of blood. In addition to leucocyte depletion of blood components and importing both plasma for fractionation and fresh frozen plasma (FFP) for those born after 31 December 1995, it appears likely at present (January 2004) that the exclusion of donors who have received a transfusion in the UK since 1980 will be added to this list. Renewed efforts to reduce inappropriate transfusion because of concerns about the impact of this new measure on the sufficiency of the blood supply is also probable. Perhaps the one cause for optimism comes from the failure of a massive epidemic of vCJD to develop in the UK, at least to date, and most estimates of the ultimate size of the epidemic have been reduced considerably. This makes it even more important to minimize secondary cases acquired from blood transfusion. As the UK blood services prepare for additional precautions to prevent vCJD, through deferral of transfused persons as donors, fears over ‘chicken flu’ are beginning to dominate the headlines and once more pictures from Asia show citizens wearing masks as they go about their daily lives. Although this is currently topical, it may appear 3

Chapter 1

out of date later in the year and over the next 2–3 years. Other crucial events have included the relentless march of two new technologies aimed at improving blood safety. The first, nucleic acid testing (NAT) for viral pathogens, is already established, although concerns over cost–benefit analyses, at least where NAT is a second-line test to a highly effective antibody detection system, may lead to review. The second, pathogen inactivation (PI) also appeared to be heading for implementation, and one system, Intercept, had been licensed for treatment of plasma in the EU. However, after a few patients developed antibodies to aspects of the agents of the system, a delay in further trials is inevitable until the safety profile can be assessed further. Another, different, PI system has developed similar problems with neoantigen formation. Although other PI systems are continuing to be developed, all of these work by using a chemical agent to prevent nucleic acid replication, and so each must have a potential for antigenicity that will require extensive study before any such system could be introduced for large-scale use. The above events continue to make safety and supply the main priorities of blood services and this has changed little from where they were 5 or even 20 years ago. Therefore, the four key areas considered in the previous edition still appear to be as relevant now as then and are listed below. The four principal areas to be considered are: • blood safety; • the appropriate and effective use of blood and blood products; • donor recruitment and retention; and • informing patients about blood transfusion. The opinions expressed in this introduction are those of the author alone. Blood has been assumed to have mystical qualities from the early days of transfusion experiments in the seventeenth century by Lower in England and Denis in France. A number of predictable disasters caused the subject to fall into disrepute, and progress in transfusion had to wait until there was adequate understanding of blood groups to enable safe transfusions between individuals. The imperatives of the Second World War were also important in emphasizing the need for transfusion services 4

and for providing the clear logistical base from which they might be organized. The early transfusion services, certainly in the UK, were often related to military practice and modern practitioners might be forgiven for believing that the sole objectives were collection, process and supply of (at that time) bottled blood and plasma. There also seems little doubt in retrospect that there was great profligacy in the use of blood and in particular plasma. Some of this stemmed, no doubt, from inadequacies in surgical practice and an equivalent lack of understanding of blood coagulation, but the failure to collect even the most basic evidence of any benefits of blood or plasma transfusions has bedevilled the field ever since. Following a consistent increase from the 1950s, blood usage in the USA has shown a downward trend in the past decade from a peak in 1986, and demand has been decreasing for the last 3 years in the UK. The reasons for this are probably multifactorial, but include improved surgical techniques as well as concerns about blood safety. Despite this, there is evidence for disparities in blood usage between surgeons and between hospitals, for similar activities. There is also wide variation in the use of blood avoidance strategies such as autologous transfusion using cell salvage and preoperative deposit. There is, in the UK, little use of preoperative clinics to enable haemoglobin correction with iron, other haematinics or erythropoietin. In the UK, these issues will be addressed over the next few years by ‘Better Blood Transfusion’ initiatives.

Blood safety Trends in transfusion practice in the past two decades, since the identification of acquired immunodeficiency syndrome (AIDS), have been in the general direction of enhanced safety of plasma products and cellular components, as well as improved purity. With pooled fractionated plasma products there was a shift from low, then to intermediate and eventually to highly purified factor VIII, for example, which provided many benefits in safety and specificity of treatment. Together with the development of the necessary technology, these advances led to the realization that recombinant

Introduction

products, ideally free of any added human or animal proteins such as albumin, were the ultimate expression of the drive towards total safety and absolute purity. This success in improving the safety of plasma products by eliminating donor-derived material seemed to have encouraged the view that the goal of zero risk from transfusion was to be required by regulators and governments. Although there have been no specific statements to change this, a trend seems to be emerging in favour of a ‘balance of risk’ approach. In the Netherlands, the health minister has made clear that optimal, not maximal, safety is the goal. Although it is not clear what this means precisely, the inference is that some form of cost–benefit judgement must be included in the equation for achieving blood safety. The European Union (EU) Commissioner for Health and Consumer Protection, David Byrne, who has a portfolio that includes food and blood safety, stated in a speech entitled ‘Irrational Fears or Legitimate Concerns’ on 3 December 2003 that zero risk cannot be achieved. And in the UK, ministers have begun to question how much must be spent on the safety of the railways before this becomes excessive. The inference is that other areas of public life must achieve a balance between delivering an effective service without crippling costs arising from chasing absolute safety. Prior to these public statements, it appeared that the provision of blood by national blood services was almost unique in the political imperative that required total safety, at whatever cost. This obsession with reducing risks to zero led to there being a perception that there are problems with the safety of blood. Some of this came about because of later criticism of earlier decisions, in particular in the UK over delays in implementing hepatitis C virus (HCV) testing (discussed in detail in the first edition). The failure to introduce the firstgeneration test for HCV antibody led to a delay in the effective testing for this known transfusiontransmitted virus, and there is no question that some patients acquired HCV during this period. This delay was strongly criticized in the judgement in the English courts by Justice Burton, who considered that testing should have been introduced in January 1991 and not September as happened.

One obstacle to early implementation was the desire to have the whole of the UK introducing the test at the same time, so that there would be no difference in quality of component anywhere. Although in an ideal world all parts of an individual blood service would implement testing of a new agent at the same time, this seems inappropriate when a significant delay is introduced thereby. It would seem to be preferable for larger blood services to begin testing as soon as possible in some parts of the service, even if others are not yet ready. At least in this way some donations would be protected. In countries where there is no single authority managing blood services, this already happens. In the UK, leucocyte depletion of all labile blood components was implemented to prevent the then ‘theoretical’ risk that vCJD might be transmitted by blood transfusion. Leucocyte depletion was phased in as soon as it was possible operationally – there was no ‘big bang’ before which components were not leucocyte depleted and after which they were. New virus or other tests and safety measures should be managed similarly in future. The other obstacle to new tests or safety measures is the impact on supply, i.e. on donors. There is no doubt that blood donors are the essential cornerstones of a transfusion service. Nevertheless patients expect and believe that the transfusions or tissues they receive will be as safe as possible, and that those donors who may pose an additional risk to safety should not be accepted. The range of risks for which donors are deferred continues to increase, and significant numbers are turned away because of recent tattoos or body piercings, or travel to areas where there are concerns about old or emerging agents, such as WNV, malaria or Trypanosoma cruzi exposure. Testing is becoming more complex and extensive and the prospect that PI might achieve the same result as more testing, in a single manufacturing process, is most attractive. On the face of it, PI of cellular components, e.g. of platelet concentrates using the psoralen S-59 and ultraviolet A light, holds great promise by preventing virus and bacterial replication. Removing the risk of transfusion-transmitted graft-versus-host disease by preventing T-cell replication would be an added bonus. Unfortunately, the occurrence of antibodies to blood cells produced by two of these 5

Chapter 1

systems may well prove a major, if not fatal, blow to this approach for the time being. Regulation of blood services

Blood service regulation developed following a number of episodes of transfusion-transmitted infections that occurred in the 1960s and 1970s. In the USA the responsible body is the Food and Drug Administration (FDA), through its Center for Biologics and Research (CBER) division. In the UK the regulator has been the Medicines Control Agency, mainly for the production of pharmaceuticals from plasma, as cellular components were not considered to meet the requirement for a ‘product’. This latter nicety was dealt with by Justice Burton in his HCV judgement, which confirmed that cellular blood components were indeed products. The Medicines Control Agency merged with the Medical Devices Agency in 2003 to form the Medicines and Healthcare products Regulatory Agency (MHRA). For Europe, there is an overarching medicines safety body, the European Agency for the Evaluation of Medicinal Products (EMEA), but no unifying EU legislation until the EU Blood Directive entered EU law in January 2003. This will be implemented by the end of 2004 by EU member states, and requires defined standards for all aspects of the blood supply chain. For the first time in the UK there will be a legal requirement to trace blood donations to the recipient, which will take regulators into hospitals for the first time. There are still no legal requirements to consider transfusion alternatives or to implement optimal blood-use programmes. Risk management

Awareness of the importance of protecting patients from potential risk following transfusion has taken a much higher profile recently. Ten years ago there were delays in introducing tests that would clearly have impacted significantly at the 1 in 1000 or 1 in 10 000 level for HCV transmission. Now, blood services in Europe and the USA are implementing tests using nucleic acid amplification by polymerase chain reaction (PCR) where it is possible to detect events with an incidence of 6

between 1 in 300 000 and 1 in 2 000 000. These risks are now perceived to be politically and economically worth preventing. Sir Kenneth Calman, the former Chief Medical Officer of the Department of Health in the UK, addressed issues of risk in a series of articles. He provided examples of activities associated with moderate risk, such as smoking 10 cigarettes a day (1 in 200 chance of death in any one year), to infinitesimal risks, such as being struck by lightning. Schreiber and colleagues, writing on behalf of the US Retrovirus Epidemiology Donor Study, estimated the risk for transfusion-transmitted virus infections at between 1 in 63 000 for hepatitis B virus (Calman risk level, very low) and 1 in 493 000 for human immunodeficiency virus (HIV) (Calman risk level, minimal) (Table 1.1). There are no equivalent figures for the UK, although an estimate for HIV can be made from evidence of only two known transmissions of HIV since the introduction of testing in 1985. During that time some 30 million donations have been transfused. In the first year of testing for HCV RNA using PCR, only one true PCR-positive, HCV antibody-negative donation has been detected, during a period when about 3 million donations were tested. Although PCR testing for HCV RNA was initially introduced for testing plasma donations, it has been a mandatory release criterion for cellular components since 2000, in order to remove a risk of around 1 in 2 000 000 or less. The number of NAT-positive, HCV antibody-negative donations has been very small since then, and the cost of each transfusiontransmitted case avoided has been immense. Why should such minimal or even infinitesimal risks be unacceptable in blood transfusion? There is no doubt that the appalling stigmatization of individuals that occurred during the development of the AIDS epidemic in the USA and Europe has some part to play. Descriptions of transfusiontransmitted infections in the media invariably use words such as ‘tainted’ and ‘contaminated’ in relation to the blood supply. The invasion of the body by an unseen, unknown and unwelcome virus or other agent may explain some of the psychological revulsion. Commissioner Byrne alluded to this issue in his 3 December speech, and suggested that the control that individuals can exert over a risk is

Introduction Table 1.1 Descriptions of risk in relation to the risk of an individual dying (D) in any one year or developing an adverse

response (A). (From Calman 1996 with permission.) Term used

Risk range

Example

Risk estimate

High

> 1 : 100

Moderate

1 : 100–1 : 1000

Low

1 : 1000–1 : 10 000

Very low

1 : 10 000–1 : 100 000

Minimal

1 : 100 000–1 : 1 000 000

Negligible

< 1 : 1 000 000

(A) Transmission to susceptible household contacts of measles and chickenpox (A) Transmission of HIV from mother to child (Europe) (A) Gastrointestinal effects of antibiotics (D) Smoking 10 cigarettes a day (D) All natural causes, age 40 (D) All kinds of violence and poisoning (D) Influenza (D) Accident on road (D) Leukaemia (D) Playing soccer (D) Accident at home (D) Accident at work (D) Homicide (D) Accident on railway (A) Vaccination-associated polio (D) Hit by lightning (D) Release of radiation by nuclear power station

1 : 1–1 : 2 1:6 1 : 10–1 : 20 1 : 200 1 : 850 1 : 3300 1 : 5000 1 : 8000 1 : 12 000 1 : 25 000 1 : 26 000 1 : 43 000 1 : 100 000 1 : 500 000 1 : 1 000 000 1 : 10 000 000 1 : 10 000 000

crucial in the acceptance of that risk. So accepting a lift in a car enables a person to decide whether the car seems roadworthy, the driver sober and likely to be a safe choice. Similarly perhaps with issues such as home versus hospital child birth, where a significant minority of women choose home delivery despite evidence of greater risk. With a blood transfusion, or a food additive, no such choice is possible. The recipient or consumer must accept the safety of the blood or food at face value. If that acceptance is later found to have led to a transfusion-transmitted infection, then anger, compensation claims and litigation are the common responses. Unfortunately, there is no arena in which a dispassionate discussion about blood safety can be held. It would surely be helpful to discuss these important issues in a forum in which patient groups, transfusion professionals, clinical users of blood and those responsible for funding new developments could consider the issues outside the blame culture that follows a perceived transfusion problem. Involvement of ‘stakeholders’ in deciding the principles by which new, expensive and perhaps only moderately effective measures should

be introduced might be interesting and educational for all concerned. Issues to be considered when deciding whether to implement new testing or other safety measures for a transfusion-transmitted infection include the following. 1 Nature of agent being tested for, and the disease it causes. 2 Is there effective treatment? 3 How much does that treatment cost? 4 Is there perceived stigmatization or implications for subsequent lifestyle, e.g. sexually transmissible. 5 What compensation might be payable if no testing is implemented? 6 What is the potential loss of reputation to the blood service? 7 How much does the test or intervention cost? 8 How effective is the test or intervention at preventing future transmission? vCJD precautions

The publication of the results of experiments in sheep which showed that vCJD could be transmitted by whole blood transfusion suggested that it 7

Chapter 1

would be only a matter of time before vCJD was transmitted by blood transfusion between humans. Although the case reported in December 2003 remains only ‘possible’, the acquisition of vCJD by one of only 48 patients being followed who were known to be at transfusion risk is highly suggestive. The extension of the risk reduction measures already introduced (importation of plasma from countries with no or low vCJD, universal leucocyte depletion of fresh components and importation of FFP for children born after 31 December 1995) to include the deferral of all previously transfused donors is imminent. This will put pressure on supplies at a time when donor attendance seems to be falling. As was the case 4 years ago, a more appropriate move in terms of addressing concerns about safety would be a more rigorous process of thought for each and every transfusion, especially in those individuals who are likely to have a long survival after it. This would include all children and those adults who do not have life-threatening diseases, such as candidates for replacement hip surgery. The very large sums of money allocated to vCJD prevention in the UK (£70–100 million per year) might have been better spent, for example, by investing in an educational programme for hospital workers at all levels. The introduction of hospital transfusion teams to ensure that patients get the right blood and are not excessively or unnecessarily transfused would have been another approach that should have been considered.

Appropriate and effective use of blood and blood products Recent studies indicate that the most important effect on the effective use of blood within a hospital or group of hospitals seems to be its culture of transfusion. It has been known for some years that blood transfusion activity is based more on local custom and practice than on evidence. Clear differences exist in transfusion practice and blood usage between individuals and between hospitals. Although there has long been an assumption that blood must be a good thing, recent evidence suggests that even moderate transfusion practices may 8

in fact carry risks. A 1999 randomized trial of red cell transfusion thresholds in the setting of intensive care suggested that less was best, and a systematic review of albumin use in critically ill patients strongly suggested an adverse outcome in those patients who received albumin rather than crystalloids. The vested interests of those on either side of the albumin controversy demonstrated the difficulty of both collecting evidence that would be believed universally and in the acceptance by clinicians of the possibility that they may have been wrong all along. It has been well known for some time that individuals who reject blood transfusion for religious reasons, such as Jehovah’s Witnesses, can undergo open heart surgery with a reasonably high degree of safety. This in itself might suggest that for many years there has been a greatly excessive use of blood (as perioperative red cell transfusions). This is not to say that blood transfusion has not enabled new and innovative surgical procedures to be initiated. Blood remains essential for many cardiac surgery operations and for liver surgery, to cite but two, and of course many patients with malignancy could not receive chemotherapy without the use of blood components to support them. Even in situations where blood transfusion is life-saving, risks remain from errors in the transfusion process, leading to the ‘wrong blood [being] given’, issues highlighted in the UK Serious Hazards of Transfusion (SHOT) reports. In the face of an increasing body of evidence suggesting that blood transfusion carries both known and unknown risks, surely we should be seeking to eliminate unnecessary transfusion. Evidence of the clear benefits of red cell transfusion from good randomized trials is lacking, although there is now evidence that patients with cardiac decompensation tolerate anaemia badly and do benefit from transfusion to higher haemoglobin levels. It is therefore incumbent upon clinicians to think once, twice and three times before transfusing patients and serious consideration must be given to involving patients in these decisions (see below). One thing that does seem clear is that now vCJD seems likely to be transmissible through blood transfusion then there will be an interest in each and every transfusion received by a person who contracts

Introduction

vCJD (or even tests ‘positive’ for it if and when there is a test). Clinicians responsible for prescribing blood must be able to justify each transfusion. The appropriate and inappropriate indications for transfusion are covered elsewhere in this book. However, the evidence continues to accumulate that there are still hazards of blood transfusion that it is not possible to avoid, and that blood transfusion will never be zero risk. The time is overdue for a concerted effort to reduce the use of allogeneic blood to those situations where it is essential to saving or prolonging life, or the quality of life. Since the previous edition of this book, the four UK Chief Medical Officers convened a further seminar in September 2001 to consider the issue of ‘Better Blood Transfusion’. This was followed in 2002 by a further Health Service Circular (HSC) to the chief executives of hospitals in the UK, setting out an agenda for hospitals to follow. This second seminar was held partly because of a generally disappointing response to the first in 1998! Although the second HSC provides a clear toolkit for implementation of better transfusion practice, hospitals have many competing priorities, and it is still difficult to maintain blood transfusion at a high enough level of urgency for hospitals to respond in a consistent way. The cynic might be forgiven for believing that only if there is a blood shortage, sufficient to impact on surgical activity, will hospitals really tackle the issues of best transfusion practice. Reducing wastage

A discussion about the disparity between the demands for blood placed on transfusion services by clinicians and the true needs of the patients being treated is beyond the scope of this chapter. However, one good first step towards ensuring that there is always sufficient blood would be to check that no blood donation is wasted. Unfortunately, this is far from the case and figures of between 5 and 40% are quoted informally for different regions, hospitals or blood groups. The loss of potential donations begins as soon as a prospective donor arrives to offer a donation. An increasing proportion of individuals who come forward offering themselves as donors are unsuitable for reasons of low haemoglobin, lifestyle issues

known to be associated with a higher risk (e.g. transmissible infectious disease) and other temporary reasons for deferral such as body piercing, tattooing and international travel. Technical difficulties in the process of donation may also impair the percentage of units going forward for patients, such as low volume donations, long donation time and technical problems with leucocyte filtration, to give some examples. Where donors would find wastage unacceptable would be if they were aware that their donation might simply go out of date because nobody had used it or because it had been left carelessly out of a blood refrigerator. Improvements in crossmatch to transfusion ratios are continuing all the time but much more needs to be done because it is imperative that blood is not ‘tied up’ waiting for patients who are very unlikely to need it, and so unavailable to those who do. In this way so-called ‘electronic crossmatching’ holds out much hope and is already implemented safely in many parts of the world. Innate conservatism and lack of investment seem to have inhibited its more widespread acceptance. Many of these measures can be implemented if only there was a sufficient will to do so.

Donor recruitment After the end of the Second World War there was a strong sense of community, and in addition many people worked in large industrial settings with a strong sense of identity. This made blood collection easy, since workplace sessions readily recruited large numbers of willing donors. Gradually, many of the large industries have disappeared, and in their place service sector jobs more widely dispersed geographically have arisen. Competition, the changes in the place of women in society (most now work) and a perception that everyone now has less free time have provided challenges to which blood services have had to adjust. Sometimes these responses have been slow. For too long the premise seemed to be that individuals would tolerate a wait of many hours to donate, and the whole process was very centred towards the blood service collection system rather than donor requirements. Only recently has this been fully 9

Chapter 1

acknowledged as inappropriate and moves towards donation by appointment, improving the processing of donors through the session (‘donor flow’) and an increased emphasis on the professionalism of donor staff have all helped to maintain the donor base. It is essential that transfusion services continue to make it easier and more convenient for individuals to donate. There is now a more mature and active relationship developing between donors and the blood services, and this process should continue since it appears that donors are not solely motivated by general altruism – a non-specific wish to do a good thing – but are aware of specific issues. More might be done to strengthen this, perhaps by using advertising more targeted to providing information about the uses of blood and how it makes a specific difference, over and above the general exhortations such as ‘we can’t operate without you’. Is there really a reduction in altruism in the UK, as has been suggested? There may be a change, particularly in young people who perhaps appear rather more self-obsessed than previous generations. The lack of major conflicts such as wars and other common adverse circumstances, while most welcome, tends to reduce the opportunities for building community spirit. However, on reflection and reviewing some of the literature in the area, it is more likely that it is the change in society in terms of longer working hours and more commitment to careers in early adulthood causing less time to consider or attend for donation that is important. It is up to transfusion services, the healthcare industry in general and government in addition to generate and maintain interest in and awareness of the need for blood donation.

Informing patients about blood transfusion In many countries it has been a specific requirement that informed consent is obtained from each patient prior to blood or plasma transfusion. Difficulties in defining what constitutes informed consent and what information must be imparted are considerable. In the USA, a legal decision has meant that recipients must be given information about the alternatives to allogeneic blood transfu10

sion as part of the consent process. In the UK the consent issue has been considered repeatedly over the past few years and is one area where medical care is lagging well behind what is likely to be considered acceptable in the event of a legal challenge. The biggest difficulty appears to be dissecting the need to obtain informed consent and the resources required to provide staff with the time and expertise to discuss the issues. In the absence of any significant momentum to obtain consent as a matter of good practice, perhaps concerns over shortages of blood, the need to consider alternatives, and potential litigation might encourage some form of dialogue between recipient and the healthcare team. In an era of potential blood shortage, blood conservation measures might achieve importance and preadmission clinics, which would need to be a minimum of 3 weeks prior to surgery, might be one way for this to occur. Discussion of alternatives to transfusion such as correction of anaemia, perioperative salvage or predeposit donation all need time and could be combined with a formal agreement by the patient to receive allogeneic blood if that proved necessary. Certainly the current situation where most patients receive little or no pretransfusion information or advice cannot be allowed to continue for much longer without a real risk of litigation in the future. Also, the lack of information makes it impossible to discover the true opinion of individuals about to have a transfusion, or the likely interest in alternative strategies such as autologous transfusion or other blood conservation approaches, or to deliver them nationally with equity. At present, well-informed individuals in major cities probably have a chance of accessing an autologous blood programme, but certainly not the great majority of potential recipients. The challenge for the transfusion services is to convince themselves and colleagues that delivering information about transfusion really is an imperative. What else might be done in the interim? Literature for patients already exists about blood transfusion and its risks, but these do not always reach the parts of the health service that most need them, i.e. the medical and surgical wards and clinics. Perhaps, rather like package inserts for pharmaceutical products that must contain a patient infor-

Introduction

mation leaflet, a leaflet should be issued with each unit of blood, plasma or platelets and handed to the recipient. This might become tedious for blood and marrow transplant units with recipients of multiple transfusions but may be useful for the majority of patients, or their relatives, who receive blood for major surgery as a single event.

Conclusion The past two decades have seen blood transfusion services in developed nations trying desperately to minimize the risk of the next transfusion-transmitted infection, one of which seems to appear every 5 years or so. Douglas Starr, in his book Blood: an Epic History of Medicine and Commerce, spells out most forcibly the errors of omission and commission made over the years. These were more usually due to a combination of denial and naivety rather than gross negligence. Five years on, and it still makes compulsory reading for anyone working in a senior position in a blood service (see Further reading). Attempts to educate the public about risk will fail as long as blood transfusion mishaps are newsworthy, even where they occur by chance in an otherwise effectively functioning system. The only realistic way forward is to engage all participants in the blood transfusion process in active discussion. The most obvious way to begin such a dialogue would be through a pretransfusion interview that would bring physician/surgeon together with the patient to discuss blood safety, and as an obvious prerequisite would require the blood services to provide training and information for colleagues in hospitals. Such an innovation might just pave the way for a realistic debate about the wisdom of further attempts to reduce the risks of transmission of known viruses by blood transfusion to an unattainable singularity of zero risk. Much of this introduction has focused on the problems and challenges that face blood services as we enter the new millennium. That there are plenty of opportunities as well as threats is certain, and the very dependence of blood services on good manufacturing practice and good laboratory practice is opening doors for crucial collaborations in the related fields of cellular immunotherapy, gene

transfer, tissue engineering and tissue and organ banking. Exciting developments in virus inactivation and in blood cell substitutes continue to provide research opportunities at the clinical interface, and improving the education of donors and patients will provide great opportunities for those in donor and patient care services. Transfusion medicine will continue to be a little like walking through a tropical rainforest, where the known paths are clear but still require careful navigation, and new and unseen threats may still lurk around the next corner to trap the unwary. But just as the rainforest contains a huge biodiversity to keep the most jaded traveller interested, so the field of transfusion medicine can never be anything other than a fascinating and rewarding area in which to work.

Further reading Bird SM. Recipients of blood or blood products ‘at vCJD risk’. Br Med J 2004; 328: 118–19. Calman KC. Cancer: science and society and the communication of risk. Br Med J 1996; 313: 799–802. Calman KC, Royston G. Personal paper: risk language and dialects. Br Med J 1997; 313: 939–42. Chalmers I. Human albumin administration in critically ill patients. I would not want an albumin transfusion (letter, comment). Br Med J 1998; 317: 885. Cochrane IGAR. Human albumin administration in critically ill patients: systematic review of randomised controlled trials (see comments). Br Med J 1998; 317: 235–40. Goodnough LT, Shander A, Brecher MA. Lancet 2003; 361: 161–9. Hebert PC, Wells G, Blajchman MA et al. A multicentre, randomized, controlled clinical trial of transfusion requirements in critical care. Transfusion requirements in Critical Care Investigators, Canadian Critical Care Trials Group (see comments). N Engl J Med 1999; 340: 409–17. Hunter N. Scrapie and experimental BSE in sheep. Br Med Bull 2003; 66: 171–83. Llewelyn CA, Hewitt P, Knight RSG et al. Possible transmission of variant Creutzfeldt–Jakob disease by blood transfusion. Lancet 2004; 363: 417–21. McClelland B. Albumin: don’t confuse us with the facts. Rather than fulminating, seek to answer the questions raised (editorial, comment). Br Med J 1998; 317: 829–30.

11

Chapter 1 Schreiber GB, Busch MP, Kleinman SH, Korelitz JJ. The risk of transfusion-transmitted virus infections. N Engl J Med 1996; 334: 1685–90. Serious Hazards of Transfusion. Annual Report 2001–2002. www.shotuk.org Sharp Research Ltd. Altruism and Blood Donation. Qualitative Research Report. London: Research

12

Division, Central Office of Information, November 1998, Report No. SR0015. Starr D. Blood: An Epic History of Medicine and Commerce. London: Little, Brown, 1999. Wu YY, Snyder EL. Safety of the blood supply. Role of pathogen reduction. Blood Reviews 2003; 17: 111–22.

Chapter 2

Essential immunology for transfusion medicine Willem H. Ouwehand and Tim B. Wallington

The immune system is a sophisticated and multilayered defence against infection. It is based on the recognition of non-self in any potential pathogen. Since donor organs are not self, their transplantation from one individual to another is only possible if the obstacles inherent in the recipient’s immune system can be overcome. The transfer of blood components (either as therapy or during the course of pregnancy) is a form of transplantation. This discussion concentrates on the immunobiology of clinical problems that are encountered as a result. There are excellent texts which discuss human immunobiology in detail (see Further reading). The immune system has evolved distinct mechanisms for coping with extracellular pathogens, such as bacteria, based on the production of antibody and with intracellular pathogens, such as viruses, based on the activity of effector T cells. Two essential layers of defence are utilized: • innate immunity, which is primitive in evolution and not single pathogen specific (e.g. mannan-binding protein which binds microbial cell-wall saccharides and Toll-like receptors which can be directly involved in macrophage activation); and • adaptive immunity in higher animals, which adding to innate mechanisms brings specificity and memory to immune responses (e.g. antibody formation in defence against bacteria). Most problems encountered in transfusion medicine are antibody based, i.e. the humoral immune response, and this will be considered in greater detail below.

Cellular basis of the immune response The key effector cells are T cells, B cells and natural killer (NK) cells. The progenitors of T cells, B cells and NK cells are derived from the same haematopoietic stem cells (HSC) that give rise to other types of blood cell. Cells of the monocyte–macrophage series, including Langerhans’ cells and dendritic cells, process and present antigen to both T and B cells. Progenitor cells migrate from the circulation into the epithelial thymus to become T cells. There they interact with the stromal cells and their soluble products to undergo cell division, clonal selection and maturation. In addition they acquire their antigen receptor (T-cell receptor or TCR) and other surface molecules which will determine their function, CD8 on cytotoxic T cells, CD4 on helper T cells. Immature T cells initially express both CD4 and CD8 molecules, which interact respectively with major histocompatibility complex (MHC) class II or class I molecules on thymic stromal cells to influence their maturation into CD4 or CD8 T cells. Through this process self-reactive T cells are removed. Later, when they migrate to the periphery, T cells may undergo selective clonal activation triggered by antigen, which leads to proliferation and maturation. Subsequent function depends on whether the cells carry the CD4 or CD8 antigens. B-cell development is a multifocal process that is concentrated in fetal liver before bone marrow becomes the major haematopoietic organ. Progenitor cells receiving signals from local stromal cells begin to divide and begin the process that will provide an antigen receptor, in this case surface immunoglobulin (SIg). Like T cells, immature B cells are easily tolerated or killed by 13

Chapter 2

premature stimulation via their antigen receptors to prevent damage to self. After migrating from the bone marrow, B cells mature, express SIg antigen receptors, and respond to antigens together with T-cell help from CD4 cells by undergoing proliferation and plasma cell differentiation. NK cells are non-T, non-B lymphoid cells capable of killing virus-infected cells either specifically targeted by the presence of antibody on their surface (antibody-dependent cell-mediated cytotoxicity, ADCC) or through the recognition of changes in the infected cells surface that allow NK cell attack. This mechanism is greatly enhanced by the cytokine interferon (IFN)-g, illustrating the fact that these key effector cells usually act in concert in the defence against infection.

Humoral immune response Antibody

The specific effector molecule is an antibody which is secreted into the extracellular space from plasma cells. It is a unique tetramer made up of two identical heavy and two identical light chains (Fig. 2.1). These combine variability of amino acid sequence and thus variability of tertiary structure (Fab) with N terminal

VH Hinge region CH3

CH1 VL

CH2

S C terminal

CL

S

VL Fc

Fab

VH

Fig. 2.1 Basic structure of an immunoglobulin molecule.

Domains are held in shape by disulphide bonds, though only one is shown. CH1–3, constant domain of an H chain; CL, constant domain of a light chain; VH, variable domain of an H chain; VL, variable domain of a light chain.

14

a constant region (Fc), which allows the molecule to bind target antigen via Fab and trigger effector functions through the Fc portion. These molecules are more generally called immunoglobulins. They also serve as antigen receptors on B cells. Antibody effector functions

The constant regions of the heavy (H) chain of immunoglobulins are responsible for the triggering of an effector pathway. This occurs either: • by binding to appropriate Fc receptors on effector cells, such as leucocytes and mast cells; or • by activation of the complement cascade. There are five immunoglobulin isotypes based on different genes for the C domains of the H chain (Table 2.1). Immunoglobulin G (IgG) and IgA have four and two subclasses, respectively. The immunoglobulin isotypes and subtypes differ significantly in their ability to recruit effector functions (see Table 2.1). This is of clinical significance in transfusion as the ability of antibodies to bring about erythrocyte or platelet destruction varies according to their isotype and IgG subclass. Basis of antibody variability

The molecular biology of this is complex. The genes for the five heavy chains and the l and k light chains are found on separate chromosomes, at band 14q32, 22q11 and 2p11, respectively. Each chain is separately synthesized before being assembled into an antibody molecule. On chromosome 14, which carries the H-chain genes, there are three clusters: approximately 50 variable (V) region genes, which encode the first 95 amino acids of the V portion, more than 20 diversity (D) region genes and six joining (J) region genes. Together these genes encode for H-chain V regions. Like letters of the alphabet they can be joined at random into three-letter ‘words’, thus providing much variability in the receptor portion of the H chain. Similarly, 22q11 and 2p11 have two clusters of genes for the V portion of the k and l light chains, respectively, which can recombine in this way. The incredible diversity of antibody specificity, found even at the level of the germline, is the result of these events, coming together in

Essential immunology Table 2.1 Immunoglobulin classes and their functions.

Structure

Function

Isotype

Heavy chain

Light chain

Configuration*

Complement fixation†

Reaction with FcR

Placental passage

IgM IgG1 IgG2 IgG3 IgG4 IgA1 IgA2 IgD IgE

m g1 g2 g3 g4 a1 a2 d e

k, l k, l k, l k, l k, l k, l k, l k, l k, l

Pentamer Monomer Monomer Monomer Monomer Monomer Dimer in secretion Monomer Monomer

+++ +++ + +++ – – – – –

L M, N, P, L, E P, L M, N, P, L, E N, L, P — — — B, E, L

– ++ +/– ++ + – – – –

so

me

* Pentamer, five basic tetrameric units (in vitro good agglutination); dimer, two basic units; monomer, one basic unit.Two or more basic units are held together by a J chain. † Classical pathway. B, basophils/mast cells; E, eosinophils; L, lymphocytes; M, macrophages; N, neutrophils; P, platelets.

Ch

ro

mo

'Silent' area = intron (V)n

D

J

Cm

Cd Cg3 Cg1 Ca1 Cg2 Cg4 Ce Ca2

14 H The product is VHCm, i.e. an IgM heavy chain with a particular variable region

VDJCm (V)n

J

Ck

2 k

Final product IgMk or IgMl

VJCk (V)n

J

Cl

22 l Fig. 2.2 Genes encoding antibodies (see

text for explanation).

the tertiary structure of the Fab2 portion of the immunoglobulin molecule (Fig. 2.2). The variability of the receptor for antigen on T lymphocytes (TCR) is the product of similar mechanisms. Somatic mutation

Antibody function is further refined in the sequential production of immunoglobulin isotypes as the adaptive immune response matures, as follows.

VJCl

• There is switching in the immunoglobulin isotype mix to molecules that are more effective in neutralizing the wide variety of pathogen types that may be encountered; this process is controlled and driven by helper T lymphocytes. • Mutation at hotspots where the V, D and J genes join and similarly within the V portion of light chains, which refine the shape of the receptor area and thus the specificity of the antibodies produced. 15

Chapter 2

Blood cell antibodies illustrating the above principles

These mechanisms at work are well illustrated by the behaviour of commonly found blood cell antibodies. In a typical T-cell-dependent antiprotein (e.g. anti-D) immune response, the switching of the immunoglobulin isotype from IgM to IgG is associated with an increase in the affinity of the antibody. In the early phase of the response the V region is encoded by genes from the germline and antibody affinity is low. As a result, in the early phase of an anti-D response, a panreactive antibody, reactive with several blood groups, might be detected in panel studies. This reaction is most likely caused by low-affinity IgM antibodies which, although responding to RhD, are able, particularly at 4°C, to react with other blood groups because of an avidity effect. Maturation of the immune response with the selection of B and T cells with greater specificity for RhD results in improved antibody affinity and the disappearance of cross-reactivity. This is the consequence of somatic mutation of the rearranged V gene. In high-affinity anti-D antibodies, up to 20 of the 90 codons encoding the V region have changed from germline. Temperature dependency of antibody– erythrocyte interactions is used as a surrogate marker for antibody affinity. • Low-affinity antibodies generally do not bind sufficiently strongly at 37°C to be detected by agglutination but they do at lower temperatures. They are also generally of no clinical significance. • Antibodies of intermediate and high affinity do remain bound and are detected in the antiglobulin test. Pretreatment of red cells with proteases is also an effective method to reveal the presence of lowaffinity antibodies against, for example, RhD as this reduces the strength of interaction between antibody and cell required for agglutination. Other circumstances can also favour the binding of low-affinity antibodies. Some of the isotypes able to activate complement are detected in haemolysin tests if the antibody is present in excess and the activation of complement is facilitated by lowering the pH of the reaction medium. 16

Antigen in the adaptive humoral immune response

The immune response is driven by antigens which select the lymphocytes that are able to participate. Therefore selective use of V genes in antibody production against a certain antigen might be expected. Studies on the V-gene use of blood cell antibodies support this and have thrown light on certain serological anomalies. Most evidence has been acquired by studies on the molecular structure of the V domains of monoclonal antibodies against the carbohydrate antigen I and the protein antigen RhD on the red cell membrane. These studies suggest that: • there is preferential but not exclusive use of certain V genes in the generation of these specificities; • there is a significant overlap in the amino acid sequence of the V domains of cold agglutinins against the lactosylceramide I and anti-D antibodies; • pathological anti-I cold agglutinins, as observed in the majority of patients with cold haemagglutinin disease, uniquely use the VH gene segment DP63 (V-4.34); • postinfectious polyclonal anti-I antibodies seem to make use of the same VH gene segment, while cold agglutinins with other specificities do not; and • in over 50% of monoclonal IgM anti-D antibodies the VH domain is encoded by the DP63 VH gene, the same as that encountered in pathological anti-I cold agglutinins. It is attractive to speculate that RhD-specific B cells evolve from B cells with anti-I specificity. This suggests that in the germline these cells provide an SIg receptor which is best fit at that stage in the immune response for the tertiary structure presented by D. In this scenario the drift in antibody specificity from anticarbohydrate (anti-I) to antiprotein (anti-D) is best explained by minor changes in the amino acid sequence of the V domains of anti-I antibodies brought about by somatic mutation. Ultimately the low-affinity binding for I is lost. This is also influenced by the switch from IgM to IgG. In serology these structural observations are supported by the functional observations on low-affinity interactions mentioned earlier and by the fact that certain IgM

Essential immunology

monoclonal antibodies used for D typing show reactivity at 4°C with protease-treated RhDnegative red cells. The isotype and subclass of blood cell antibodies are at least partly determined by the chemical nature of the antigen which had stimulated their production. • Blood cell antibodies against carbohydrate antigens are generally IgM or IgG2 and IgG4 or a combination of these. • Antibodies against protein blood group antigens are typically of the IgG class, predominantly IgG1 and IgG3, although autoantibodies can be of the IgA type. This suggests a direct involvement of the antigen. The source of antigen might not of course be red cells if the same structure is shared, so-called crossreactivity. An increase in the titre of anti-I antibodies occurs after infection with Mycoplasma pneumoniae. Some preparations of the vaccine TAB (typhoid, paratyphoid A and paratyphoid B) stimulate anti-A and anti-B and cause not only a rise in agglutinin titre but also the development of immune characteristics. Many of the low-affinity reactions seen in red cell serology reflect part of the response to bacterial antigens, usually carbohydrate. When the affinity and concentration of such antibodies increases above certain thresholds complement-mediated haemolysis can occur, making the phenomenon of clinical significance. T-cell-independent antibody formation

As we have seen, the formation of antibodies by B cells is dependent on interaction with helper T cells. However, some antigens can stimulate certain subsets of B cells (B1 and marginal zone B cells) directly, independent of T-cell help. These cells provide a first line of defence to bacteria by producing antibody specific to bacterial polysaccharide. This route to antibody production is also important in the response to certain red cell antigens. The presence of naturally occurring IgM antibodies against A and B is an excellent example of T-cell-independent antibody formation. In the presence of an intact immune system isoagglutinins to the missing A or B antigens are always found, although there has been no exposure to red

cells carrying these antigens. The response is essentially limited to IgM because T cells are not involved and are not available to induce isotype switching, although there is evidence that some switching can occur in the absence of T cells. The repetitive carbohydrate structures (epitopes) presented by the A- and B-determining portions of the relevant cell surface glycolipids and glycoproteins are structurally the same as bacterial polysaccharide and indeed it is antigens from gastrointestinal bacteria that trigger isoagglutinin production. The isoagglutinins are present from the first months of life. IgG anti-A or anti-B antibodies can also be formed to this stimulus, usually of the IgG2 or IgG4 type. T-cell-dependent antibody formation

Unlike glycolipids and glycoproteins, the formation of antibodies against blood cell membrane proteins is always dependent on interaction with T-helper cells. The immune response to RhD is an example. • RhD is the most immunogenic red cell membrane protein antigen. RhD is a 30-kDa nonglycosylated membrane protein. • Analysis suggests that only short peptide loops, part of the molecule, are displayed on the cell surface. • Ample evidence indicates that anti-D antibodies recognize discontinuous amino acid sequences derived from several of the extracellular RhD loops. • These discontinuous residues come together in the tertiary structure of the RhD protein. Therefore isolation of RhD from the membrane disrupts the majority of B-cell epitopes and reactivity with anti-D antibodies. This is not a repetitive structure as with ABO and the B-cell response requires T-cell help. The response of T-helper cells, like B cells, is antigen specific but triggered in a totally different way. T-helper cells recognize short linear segments of amino acids derived by intracellular digestion from the RhD protein and presented to the helper T cell by the human leucocyte antigen (HLA) class II molecule, which is a cell surface molecule. This is achieved most effectively by professional antigen17

Chapter 2

presenting cells (APCs). APCs belong to a family of cells with diverse anatomical locations and of diverse ontogeny: • Langerhans’ cells in the skin; • interdigitating, follicular and germinal centre dendritic cells in lymph nodes and spleen; and • B cells and macrophages. The antibody response to a red cell antigen like RhD involves the collaboration of at least three cell types and of antigen in two forms, both intact and processed. This enables two important features of the immune response: • an efficient mechanism for tolerance to self antigens enables prevention of the failure to discriminate self from non-self that is the basis of autoimmunity; and • the maturation of the antibody response, through isotype switching, is facilitated. This is illustrated in detail by the specific example from transfusion medicine described below. Human platelet antigen-1a presentation via the HLA class II route

If a human platelet antigen (HPA-1a)-negative mother is carrying an HPA-1a-positive fetus, platelets may enter the maternal circulation and immunize her against HPA-1a. This is the end result of quite complex events and can have disastrous consequences for the fetus. Fetal HPA-1a may be ingested by maternal APCs and degraded by endosomal enzymes like cathepsin G. Short fragments of 12–15 amino acids will be produced in endosomes, which in the trans-Golgi network fuse with HLA class II-containing vesicles. The fusion results in a downwards pH shift, which results in the removal of the invariant chain from the HLA class II molecule (this chain prevents the premature loading of the cleft in the molecule that is used for presentation of the digested antigen). A plethora of peptides will be bound in the HLA class II groove, of which some will have been derived from the fetal GPIIIa. Once migrated to the surface of the APC, specific helper T cells then recognize the change in the HLA class II molecule produced by the peptides to which they are specific and T-cell proliferation will ensue. Cytokines produced by expanding clones of reactive T-helper 18

cells will drive the expansion of HPA-1a-specific B cells responding to intact HPA-1. With time the process of antigen take-up, processing and presentation will pass from the classical APCs to HPA1a-specific B cells. This helps to bring together the complex interaction of cells needed for a mature antibody response, as both the surface molecules needed for antigen presentation to helper T cells and for interaction with intact HPA-1 are present on the same cell set. HLA class II restriction of antibody response

We have seen that there must be a genetic element to an individual’s immune response in that the first encounter with antigen is dependent on the germline V-region genes, which show differences between individuals. Whether or not processed peptide from a particular alloantigen can interact with a particular HLA class II molecule to trigger a T cell is also dependent on genetic variability. In the immune response this is important to the immunogenicity of antigens in individuals. Sometimes the peptide is presented exclusively by a certain HLA class II molecule and a linkage between HLA class II type and antibody response can be observed. The HLA DRB3*0101 restricted response against HPA-1a (GPIIIa, leucine 33) is the best example of an HLA class II restricted response in humans. The very much lower immunogenicity of the antithetical antigen HPA-1b (HPA-1b, proline 33) is most likely explained by a less good fit of the peptide containing the proline at position 33 when compared with the one containing leucine at that position. Antibody-mediated blood cell destruction

Most red blood cell alloantibodies and autoantibodies of the IgG isotype bring about destruction in the extravascular compartment via the interaction of the IgG constant domain with Fcg receptors on cells of the mononuclear phagocytic system. Several receptor types are described. • FcgRI is the most important in blood cell destruction. This is a relatively high-affinity receptor found predominantly on monocytes. The consequence of adherence of IgG-coated red cells to

Essential immunology

FcgRI-positive cells is phagocytosis and lysis. This is usually extravascular and takes place in the spleen. The lysis can be demonstrated in vitro as ADCC. • FcgRII is a lower affinity receptor found on monocytes, neutrophils, eosinophils, platelets and B cells. • FcgRIII is also relatively low affinity and found on macrophages, neutrophils, eosinophils and NK cells. It is responsible for the ADCC demonstrable in vitro with NK cells. • There is also an FcgR on the placenta of a different molecular family which mediates the transfer of IgG into the fetus. The severity of red blood cell sequestration by IgG antibodies is determined by the concentration of antibody, its affinity for the antigen, antigen density and the IgG subclass. IgG2 and IgG4 antibodies are generally unable to reduce red cell survival, while IgG1 and IgG3 are. There is ample evidence in patients with warm-type autoimmune haemolytic anaemia that IgG1 and IgG3 are more effective in causing red cell destruction than IgG2 and IgG4. The level of IgG1 coating of red cells needs to exceed a threshold of approximately 1000 molecules per red cell to cause cell destruction. For a long time it has been speculated that polymorphisms in the genes of the family of Fcg receptors might be significant in causing differences of severity of blood cell destruction observed between patients with apparently similar levels of IgG coating. So far, firm evidence for such polymorphisms has been lacking, although a single amino acid polymorphism of the FcgRIIa receptor dramatically alters the affinity for human IgG2 and additional polymorphisms might have an effect on the interaction with IgG1 and IgG3.

Complement system The complement system, either working alone or in concert with antibody, is important for effective immunity to many extracellular pathogens. It also often plays an important part in immune red cell destruction and can be the reason for important systemic complications of haemolysis. Naturally occurring IgM antibodies against A and B are

often of low affinity and do not bind to red cells at 37°C. However, when ABO blood group antibodies do bind at 37°C, there will be rapid complemant-mediated destruction of incompatible red cells where there is a major A to O or B to O mismatch. This may result from a transfusion error, and remains an important cause of transfusionrelated mortality and morbidity. Blood cell antibodies that can activate complement are more effective in achieving cell destruction than non-complement-activating antibodies. In contrast to extravascular FcgR-mediated destruction, complement-mediated lysis occurs in the intravascular compartment. The ensuing release of anaphylatoxins such as C3a and C5a contributes to the acute systemic effects that occur. IgM, IgG1 and IgG3 antibodies are the most effective isotypes in binding C1q and initiating activation of the complement cascade via the classical pathway. However, they are dependent on aggregation for a sufficiently high antibody density to trigger C1q and overcome the regulators of complement activation that are present. The concentration of antibody may be too low to achieve the density necessary. The antigen topography (e.g. of RhD) can prevent the binding and activation of the C1q molecule. Complement is a complex system of plasma proteins, both part of innate immunity and vital to the effector functions of complement-fixing immunoglobulin isotypes. Central to complement’s function is the activation of C3 as this leads to the opsonization of bacteria (Fig. 2.3). C3 can be activated by three routes: the classical pathway, the alternate pathway and lectin binding. Lysis is dependent on activation, downstream from C3, of components of the membrane attack pathway. The classical complement pathway consists of: • four numbered components (C1–C4); and • two regulatory proteins (C1 inhibitor, C4binding protein). The first component (C1) comprises three subcomponents, C1q, C1r and C1s. It is the interaction between C1q and aggregated IgG or IgM bound to antigen that initiates activation of the classical complement sequence. The fixation of C1q activates C1r and C1s. C1s cleaves C4 and C2, whose 19

Chapter 2 Classical

Alternate

Clqrs

C3b MBL - MASP +

C4b

Factor D

C2b

Factor B C3 Opsonization

C3bBb

C3b4b2b Properdin

C5 Convertases C3b4b2a—classical C3bBbP—alternate

C5 C6 Lysis

Key enzymes C3 Convertases C4b2a—classical C3bBb—alternate

Mediators of inflammation

C7 C8 C9

Final lytic

active fragments C4b and C2a form the classical pathway C3 convertase. The alternative pathway to C3 activation consists of: • C3b, factor B and factor D; and • the regulatory proteins, properdin and factors H and I. Factor B binds to a cleavage fragment of C3, C3b, to form C3bB. Factor D cleaves the bound factor B to form the alternative pathway C3 convertase (C3bBb). It activates C3 in a fashion similar to the C3 convertase of the classical pathway, C4b2a. Properdin acts to stabilize this alternative pathway C3 convertase, as do carbohydrate-rich cell surfaces, by partially shielding the convertase from inhibitors. Activation via the alternative pathway would otherwise be unchecked if it were not inhibited, as it requires no specific stimulus. The lectin pathway is initiated by mannanbinding lectin. This is structurally related to C1q and binds avidly to carbohydrate on the surface of microorganisms. It activates C4 through a serine protease, which is similar to C1r and C1s with the same outcome. The attack pathway is dependent on the formation of the trimolecular complex of C4b2a3b or 20

Ba C2 Kinin C3a C5a

Fig. 2.3 The different pathways for

C567

complement activation. MBL, mannanbinding lectin; MASP, MBL-associated serine protease.

C5 convertase, which cleaves C5 to two fragments C5a and C5b. The former is a potent anaphylatoxin. C5b forms a complex with C6, C7 and C8, which facilitates the insertion of a number of C9 molecules in the membrane. The C5b-8 and the multimeric C9 molecules form the membrane attack complex (MAC), creating a lytic pore in the membrane and lysing the target cell. Cells not immediately involved in the process but close to it can also be lysed by seeded MAC, the so-called bystander lysis. Blood cells coated with C3b will bind to cells carrying receptors for C3b (CR1 or CD35). This adherence can lead to extravascular cell destruction, mainly in the liver, but if the bound C3b degrades to its inactive components iC3b and C3dg before the cell is lysed then the cell is protected from lysis. Membrane-bound molecules such as decay accelerating factor (DAF) and membrane inhibitor of reactive lysis (MIRL) protect red cells from lysis in this way. They are of clinical importance as: • DAF (CD55) and MIRL (CD59) are linked to the blood cell membrane via a glycosylphosphatidylinositol (GPI) anchor; • patients with paroxysmal nocturnal haemoglo-

Essential immunology

binuria (PNH) have an acquired mutation in the PIG-A gene in a subset of their HSC which prevents synthesis of the anchor; • progeny from the affected stem cells lack GPIlinked membrane proteins; and • the absence of DAF and MIRL from red cells increases the sensitivity for complementmediated lysis which occurs when the pH is marginally lowered during sleep, resulting in haemoglobinuria. In vitro acidification is used in Ham’s acid test to reveal the presence of a population of erythrocytes with increased sensitivity for complementmediated lysis. Flow cytometric analysis looking for the absence of GPI-linked proteins on a subset of leucocytes derived from mutated stem cells is an alternative test for the diagnosis of PNH.

Cell-mediated immunity Antigen-specific cell-mediated responses are carried out by T cells. They provide the immune system’s main defence against intracellular microorganisms and can lyse cells expressing specific antigens (i.e. cytotoxicity). In addition they release cytokines that can trigger inflammation and are responsible for delayed hypersensitivity and symptoms usually associated with infection such as fever, myalgia and fatigue. Cytotoxic T cells

Cytotoxicity is the job of cytotoxic T cells, which are distinguished by the presence of CD8 on the cell surface. This facilitates their interaction with HLA class I on the surface of cells altered by the presence of antigen, usually as the result of virus infection. Cytotoxic T cells are the main means of protection against virus infection. They are also important mediators of allograft rejection. Like the B-cell response, the cytotoxic T-cell response requires help from T-helper cells and is regulated by these cells. Delayed hypersensitivity

Delayed hypersensitivity is an example of another

crucial role for helper T cells. It is dependent on the secretion of the cytokines interleukin (IL)-1, IL-2, tumour necrosis factor (TNF) and IFN-g, the socalled proinflammatory or Th1-type cytokines. These recruit inflammatory cells, in particular macrophages, to sites of infection and arm them to kill certain bacteria, e.g. Mycobacterium tuberculosis, which normally proliferates inside cells and is resistant to killing after phagocytosis. Helper T cells also function in a Th2 manner, releasing cytokines that promote antibody formation including production of IgE, which is important in protection against parasites. The core Th2 cytokine profile is IL-4, IL-5, IL-10 and IL-13. Cell-mediated immunity in transfusion medicine

Cell-mediated immunity is of much less importance to the transfusion of blood cells than humoral immunity. It is important in the defence against blood-borne virus (discussed below). Th1type cytokines released from leucocytes in stored blood are the main cause of non-allergic febrile transfusion reactions in susceptible individuals. As we have seen, virally infected target cells are marked for recognition by cytotoxic T cells by the presence of oligopeptides derived from viral proteins in the cleft of the HLA class I molecule. This is a process analogous to that which occurs for antigen presentation on class II HLA by APCs. The cytotoxic T-cell response to this antigen is very intense. Recent studies with cytomegalovirus (CMV)-derived peptides captured in HLA class I tetramers have revealed that up to 8% of CD8+ T cells are CMV specific during CMV infection and similar results have been obtained with peptides derived from other viruses. Biologically, the antibody response can afford to lag behind this response and usually does. This is of importance to transfusion practice. Prevention of the transmission of hepatitis B, hepatitis C, human immunodeficiency virus (HIV) types 1 and 2, and CMV by blood transfusion is one of the major challenges of transfusion medicine. Counselling of donors, together with the detection of viral antigens and antibodies and the development of tests for virally derived nucleic acid, is the bedrock for the prevention of viral 21

Chapter 2

transmission. Antibody-based immunoassays are the mainstay but after a first encounter with a virus the formation of viral antibodies will require time; they are not the first line of defence or even the means to recovery from infection for many of the pathogens concerned but the basis of immunity against subsequent infection. There is a critical window, which may be several months, in which the donor carries the virus but is still antibody negative. The sequence of immunodominant oligopeptides which appear in the HLA class I molecule for all four main blood-borne viruses and several others of clinical significance has been defined and applied to treatment. Such short oligopeptides can be used to load APCs for in vitro education and proliferation of virus-specific cytotoxic T cells. These educated T cells can be clinically used for the prevention of viral infection in immunosuppressed transplant patients. • Adoptive immunotherapy for prevention of post-transplant Epstein–Barr virus-associated lymphomas by infusion of virus-specific cytotoxic T lymphocytes has been successfully applied in allogeneic bone marrow transplantation in children. • Preliminary data suggest that CMV infection in allogeneic bone marrow transplant patients can be prevented in a similar way. As blood services are adept at handling cells in vitro, they are increasingly supporting the clinical development of these novel cellular therapies, which are a direct application of the underlying immunobiology.

Antigen presentation and its clinical implications The interaction between APCs and helper T cells is complex (Fig. 2.4). We now understand which of the many interactions are prominent in turning on T-cell activity. This knowledge may provide the means of specific manipulation of immune responses. The requirement is for concurrent signalling over two independent pathways, one antigen specific via TCR and HLA/peptide and the other non-specific, the CD28/CD80 pathway being prominent. This interaction also provides control to prevent inappropriate T-cell proliferation. Initial triggering via CD28 results in T-cell proliferation and IL-2 production, which in turn induces the expression of cytotoxic T-lymphocyte antigen (CTLA)-4 on the expanding T-cell clone. CTLA-4 is a competitive inhibitor of proliferation and competes with CD28 for binding with CD80 and CD86. This molecular competition in the control of T-cell proliferation results in a balanced expansion of antigen-specific T-cell clones. The absence of initial signalling through B7 from the APC when antigen is appropriately presented leads to apoptosis of the helper T cell. This can be exploited in the modulation of immune responses through molecules which block the interaction. The following are examples which are of relevance to transfusion medicine. Clinical studies in HSC transplantation suggest that donor lymphocytes can be tolerated for Cell membrane

Cell membrane

TCR MHC class I or II APC/virus infected target cell

(ICAM-1) CD54 (LFA-3) CD58

CD4 or CD8 (LFA-1 ) CD11a/CD18 CD2

(B7.1) CD80

CTLA-4

(B7.2) CD86

CD28

CD40

CD40L

T cell

Fig. 2.4 Adhesion molecules and

signalling pathways in T-cell activation.

22

Essential immunology

incompatible HLA alloantigens on host cells by ex vivo exposure to them in the presence of a recombinant CTLA-4–IgG fusion protein which blocks B7. In platelet transfusion, alloimmunization to HLA antigens with subsequent failure to increment after further platelet transfusions is a clinical problem. This may be reduced by leucocytedepleting blood components, thus removing APCs along with other white cells. A similar effect is obtained by treatment which modifies cell membranes and particularly the B7 signal, such as ultraviolet irradiation. Conversely, immune responses can be deliberately induced by priming separated dendritic cells with immunogenic peptide derived from the relevant antigen. This approach is under investigation in cancer therapy and again blood services are involved due to their familiarity with the safe handling of such cells under conditions of good manufacturing practice. Similar molecular switches and controls influence the B-cell response to antigen. In the interaction between B cells and T-helper cells the CD40/CD154 (CD40-ligand) pathway exercises control over B-cell isotype switching and subsequent maturation of the antibody response. This pathway is also important in certain aspects of

effector T-cell function. Again there is the potential for clinical exploitation. Anti-CD40 antibodies might be powerful therapeutic reagents in kidney transplantation. HLA-incompatible transplants can be achieved successfully in monkeys if pretreated with anti-CD40 antibodies. The ability of anti-CD40 antibodies to control autoreactive B cells in treatment-refractory autoimmune thrombocytopenia looks promising. Such designer therapeutics look likely to replace less specific therapies, such as polyvalent intravenous immunoglobulin. The amount of antigen required to activate B cells is reduced if C3d is covalently bound to antigen as this leads to the concurrent signalling of CD21 (the complement receptor type 2 on B cells) as well as SIg by antigen. This is also a pathway that is open to manipulation.

Further reading Chapel H, Haeney M, Misbah S, Snowden N. Essentials of Clinical Immunology, 4th edn. Oxford: Blackwell Science, 1999. Janeway CA, Travers P, Walport M, Capra JD. Immuno Biology. The Immune System in Health and Disease, 5th edn. London: Churchill Livingstone, 2004.

23

Chapter 3

Human blood group systems Geoff Daniels

A blood group may be defined as an inherited character of the red cell surface detected by a specific alloantibody. This definition would not receive universal acceptance as cell surface antigens on platelet and leucocytes might also be considered blood groups, as might uninherited characters on red cells defined by autoantibodies or xenoantibodies. However, the definition is suitable for the purposes of this chapter. Most blood groups are organized into blood group systems. Each system represents a single gene or a cluster of two or more closely linked homologous genes. Of the 285 blood group specificities recognized by the International Society for Blood Transfusion, 245 belong to one of 29 systems (Table 3.1). All these systems represent a single gene, apart from Rh, Xg and Chido/ Rodgers, which have two closely linked homologous genes, and MNS with three genes. Most blood group antigens are proteins or glycoproteins, with the blood group specificity determined primarily by the amino acid sequence, and most of the blood group polymorphisms result from single amino acid substitutions, though there are many exceptions. The four types of red cell surface glycoproteins, based on their integration into the red cell membrane, are shown in Fig. 3.1. Some blood group antigens, including those of the ABO, P, Lewis and H systems, are carbohydrate structures on glycoproteins and glycolipids. These antigens are not produced directly by the genes controlling their polymorphisms — the genes encode transferase enzymes that catalyse the final stage in the synthesis of an oligosaccharide chain. The two most important blood group systems from the clinical point of view are ABO and Rh. 24

They also provide good models for contrasting carbohydrate- and protein-based blood group systems.

The ABO system ABO is often referred to as a histo-blood group system because, in addition to being expressed on red cells, ABO antigens are present on most tissues and in soluble form in secretions. At its most basic level, the ABO system consists of two antigens, A and B, indirectly encoded by two alleles, A and B, of the ABO gene. A third allele, O, produces neither A nor B. These three alleles combine to effect four phenotypes: A, B, AB and O (Table 3.2). Clinical significance

Two key factors make ABO the most important blood group system in transfusion medicine. Firstly, almost without exception, the blood of adults contains antibodies to those ABO antigens lacking from their red cells (see Table 3.2). In addition to anti-A and anti-B, group O individuals have an antibody, called anti-A,B, to a determinant common to A and B. Secondly, ABO antibodies are invariably IgM, though they may also have an IgG component, activate complement, and cause immediate intravascular red cell destruction, which can give rise to severe and often fatal haemolytic transfusion reactions (see Chapter 13). Major ABO incompatibility (i.e. donor red cells with an ABO antigen not possessed by the recipient) must be avoided in transfusion and, ideally, ABO matched blood (i.e. of the same ABO group) should be provided.

Human blood group systems Table 3.1 Human blood group systems.

No.

Name

Symbol

No. of antigens

Gene name(s)

Chromosome

001 002 003 004 005 006 007 008 009 010 011 012 013 014 015 016 017 018 019 020 021 022 023 024 025 026 027 028 029

ABO MNS P Rh Lutheran Kell Lewis Duffy Kidd Diego Yt Xg Scianna Dombrock Colton Landsteiner-Wiener Chido/Rodgers H Kx Gerbich Cromer Knops Indian Ok Raph John Milton Hagen I Globoside Gill

ABO MNS P1 RH LU KEL LE FY JK DI YT XG SC DO CO LW CH/RG H XK GE CROM KN IN OK RAPH JMH I GLOB GIL

4 43 1 48 20 25 6 6 3 21 2 2 5 5 3 3 9 1 1 8 12 8 2 1 1 1 1 1 1

ABO GYPA, GYPB, GYPE P1 RHD, RHCE LU KEL FUT3 FY SLC14A1 SLC4AE1 (AE1) ACHE XG, MIC2 SC DO AQP1 LW C4A, C4B FUT1 XK GYPC DAF CR1 CD44 CD147 CD151 SEMA7A GCNT2 B3GALT3 AQP3

9 4 22 1 19 7 19 1 18 17 7 X/Y 1 12 7 19 6 19 X 2 1 1 11 19 11 15 6 3 9

Single-pass Type 1 N

Type 2

Polytopic

GPI-anchored

(multi-pass)

Type 5

Type 3

C

N N C

Fig. 3.1 Diagrammatic representation

of the four types of glycoproteins of the red cell surface membrane, with examples of blood group antigens expressed on those types of glycoproteins. GPI, glycosylphosphatidylinositol.

C Glycophorins A to D, Lutheran, LW, Knops (CD35), Indian (CD44)

N Kell

N

C RhD, RhCcEe, RhAG Kidd, Diego (band 3), Colton (AQP1), Gill (AQP3), Kx

C Duffy

Yt(AChE), Dombrock Cromer (CD55) JMH (CDw108)

25

Chapter 3 Table 3.2 ABO system.

Frequency Phenotype

Genotypes

Europeans*

Africans†

Indians‡

Antibodies present

O A1 A2 B A1B A2B

O/O A1/A1, A1/O, A1/A2 A2/A2, A2/O B/B, B/O A1/B A2/B

43% 35% 10% 9% 3% 1%

51% 18% 5% 21% 2% 1%

31% 26% 3% 30% 9% 1%

Anti-A, -B, -A,B Anti-B Sometimes anti-A1 Anti-A None Sometimes anti-A1

* English people. † Donors from Kinshasa, Congo. ‡ Makar from Mumbai.

ABO antibodies seldom cause haemolytic disease of the newborn and when they do it is usually mild. The prime reasons for this are (i) IgM antibodies do not cross the placenta; (ii) IgG ABO antibodies are often IgG2, which do not activate complement or facilitate phagocytosis; and (iii) ABO antigens are present on many fetal tissues and even in body fluids, so the haemolytic potential of the antibody is greatly reduced.

membrane glycoproteins, mainly the anion transporter band 3 and the glucose transporter GLUT1, but are also on glycosphingolipids embedded in the membrane. The tetrasaccharides that represent the predominant form of A and B antigens on red cells are shown in Fig. 3.2, together with their biosynthetic precursor, the H antigen, which is abundant on group O red cells. The product of the A allele is a glycosyltransferase that catalyses the transfer of N-acetylgalactosamine (GalNAc) from

A and B subgroups

The A (and AB) phenotype can be subdivided into A1 and A2 (and A1B and A2B). In a European population, about 80% of group A individuals are A1 and 20% A2 (see Table 3.2). A1 and A2 differ quantitatively and qualitatively. A1 red cells react more strongly with anti-A than A2 cells. In addition, A2 red cells lack a component of the A antigen present on A1 cells and some individuals with the A2 or A2B phenotype produce an antibody, anti-A1, which agglutinates A1 and A1B cells but not A2 or A2B cells. Anti-A1 is seldom reactive at 37°C and generally considered clinically insignificant. There are numerous other ABO variants, involving weakened expression of A or B antigens (A3, Ax, Am, Ael, B3, Bx, Bm, Bel), but all are rare.

O

Gal

GlcNAc c

R

Gal

GlcNAc

R

Gal

GlcNAc

R

(H) Fuc

A

GalNAc Fuc

B

Gal Fuc

Biosynthesis and molecular genetics

Red cell A and B antigens are expressed predominantly on oligosaccharide structures on integral 26

Fig. 3.2 Tetrasaccharides representing A and B antigens, and their biosynthetic precursor (H), which is abundant in group O. R, remainder of molecule.

Human blood group systems

a nucleotide donor substrate, UDP-GalNAc, to the fucosylated galactose (Gal) residue of the H antigen, the acceptor substrate. The product of the B allele catalyses the transfer of Gal from UDP-Gal to the fucosylated Gal residue of the H antigen. GalNAc and Gal are the immunodominant sugars of A and B antigens, respectively. The O allele produces no transferase, so the H antigen remains unmodified. The ABO gene on chromosome 9 consists of seven exons. The A1 and B alleles differ by seven nucleotides in exons 6 and 7, which encode a total of four amino acid substitutions at positions 176, 235, 266 and 268 of their glycosyltransferase products (Fig. 3.3). It is primarily the amino acids at positions 266 and 268 that determine whether the gene product is a GalNAc-transferase (A) or Gal-transferase (B). The most common O allele (O1) has an identical sequence to A1, apart from a single nucleotide deletion in exon 6, which shifts the reading frame and introduces a translation stop codon before the region of the catalytic site, so that any protein produced would be truncated and have no enzyme activity. Another common O allele, called O1v, differs from O1 by at least nine nucleotides, but has the same single nucleotide deletion as that in O1 and so cannot produce any functional enzyme. O2, which represents about 3% of O alleles in a European population, does not have the nucleotide deletion characteristic of most O alleles and encodes a complete protein

D 261

526

exon 6 176 Arg

D 703 796 803 1059 exon7 235 266 268 Gly Leu Gly

Arg

Gly Leu Gly

Gly

Ser Met Ala

A1 A2 B O

Fig. 3.3 Diagrammatic representation of exons 6 and 7 of

the ABO gene showing the position of the nucleotide deletions (D) responsible for the common form of O (exon 6) and for A2 (exon 7), and the positions of the four nucleotide changes in exon 7 responsible for the amino acid residues that are characteristic of A- and B-transferases. Below are representations of the encoded transferases.

product, but with a charged arginine residue instead of a neutral glycine (A) or alanine (B) at position 268. This amino acid change at a vital position inactivates enzyme activity. The A2 allele has a sequence almost identical to A1, but has a single nucleotide deletion immediately before the translation stop codon. The resultant frameshift abolishes the stop codon, so the protein product has an extra 21 amino acids at its C-terminus, which reduces the efficiency of its GalNActransferase activity and might alter its acceptor substrate specificity. Biochemically related blood group systems: H, Lewis and I

H antigen is the biochemical precursor of A and B (see Fig. 3.2). It is synthesized by an a1,2fucosyltransferase, which catalyses the transfer of fucose from its donor substrate to the terminal Gal residue of its acceptor substrate. Without this fucosylation neither A nor B antigens can be made. Two genes, active in different tissues, produce a1,2-fucosyltransferases: FUT1, active in mesodermally derived tissues and responsible for H on red cells, and FUT2, active in endodermally derived tissues and responsible for H in many other tissues and in secretions. Homozygosity for inactivating mutations in FUT1 leads to an absence of H from red cells and therefore an absence of red cell A or B, regardless of ABO genotype. Such mutations are rare, as are red cell H-deficient phenotypes. In contrast, inactivating mutations in FUT2 are relatively common and about 20% of Caucasians (non-secretors) lack H, A and B from body secretions despite expressing those antigens on their red cells. Very rare individuals who have H-deficient red cells and are also H non-secretors (Bombay phenotype) produce anti-H together with anti-A and -B and create a severe transfusion problem. Antigens of the Lewis system are not produced by erythroid cells, but become incorporated into the red cell membrane from the plasma. Their corresponding antibodies are not usually active at 37°C and are not generally considered clinically significant. Lea and Leb are not the products of alleles. The Lewis gene (FUT3) product is an 27

Chapter 3 RHD Precursor of H and Lea

Gal

GlcNAc

R

Gal

GlcNAc

R

GlcNAc

R

exons

RHCE 3´

5´ 1

10

3´

10

D

1 C/c

H precursor of Leb

5´

E/e

Fuc

Gal Lea

Fuc

Gal

GlcNAc

R

Leb Fuc

N

C N

C

Fig. 3.5 Diagrammatic representation of the Rh genes,

RHD and RHCE, shown in opposite orientation as they appear on the chromosome, and of the two Rh proteins in their probable membrane conformation, with 12 membrane-spanning domains and six extracellular loops expressing D, C/c and E/e antigens.

Fuc

Fig. 3.4 Oligosaccharide structures representing Lea and Leb

expression and their biosynthetic precursors. R, remainder of molecule.

a1,3/4-fucosyltransferase that transfers fucose to the GlcNAc residue of the secreted H precursor in non-secretors to produce Lea and to secreted H in secretors to produce Leb (Fig. 3.4). Consequently, H secretors are Le(a–b+) or Le(a+b+), H nonsecretors are Le(a+b–) and individuals homozygous for FUT3 inactivating mutations (secretors or non-secretors) are Le(a–b–). I antigen represents branched N-acetyllactosamine (Galb1–4GlcNAc) structures in the complex carbohydrates that also express H, A and B antigens. The I gene (GCNT2) encodes a branching enzyme, which only becomes active during the first months of life. Consequently, red cells of neonates are I-negative. Rare individuals are homozygous for inactivating mutations in GCNT2 and never form I on their red cells. This phenotype, called adult i, is associated with production of anti-I, which is usually only active below 37°C, but may occasionally be haemolytic at body temperature.

The Rh system Rh is the most complex of the blood group 28

systems, with 49 specificities. The most important of these is D, and then C, c, E and e. Rh genes and proteins

The antigens of the Rh system are encoded by two genes, RHD and RHCE, which produce D and CcEe antigens respectively. The genes are highly homologous, each consisting of 10 exons. They are closely linked, but in opposite orientation, on chromosome 1 (Fig. 3.5). Each gene encodes a 417 amino acid polypeptide that differ by only 31–35 amino acids, according to Rh genotype. The Rh proteins are palmitoylated, but not glycosylated, and span the red cell membrane 12 times, with both termini inside the cytosol and with six external loops, the potential sites of antigenic activity (see Fig. 3.5). D antigen

The most significant Rh antigen from the clinical point of view is D. About 85% of Caucasians are D+ (Rh-positive) and 15% are D– (Rh-negative). In Africans only about 3–5% are D– and in the Far East D– is rare. The D– phenotype is usually associated with absence of the whole D protein from the red cell membrane. This explains why D is so immuno-

Human blood group systems

genic, as the D antigen comprises numerous epitopes on the external domains of the D protein. In Caucasians, the D– phenotype almost always results from homozygosity for a complete deletion of RHD. D+ individuals are either homozygous or heterozygous for the presence of RHD. In Africans, in addition to the deletion of RHD, D– often results from an inactive RHD (called RHDY) containing translation stop codons within the reading frame. Other genes containing inactivating mutations are also found in D– Africans and in D– Asians. Weak forms of D (previously known as Du) result from amino acid substitutions in the membrane-spanning or cytosolic regions of the D protein. Red cells of some D+ individuals lack some or most of the D epitopes and, if immunized by a complete D antigen, can make antibodies to the epitopes they lack. There are numerous types of these partial D antigens. They result from amino acid changes in the external loops of the D protein. Usually this is due to one or more exons of RHD being exchanged for the equivalent exons of RHCE in a process called gene conversion, but sometimes straightforward missense mutations are responsible. Anti-D

Anti-D is almost never produced in D– individuals without immunization by D+ red cells. However, D is highly immunogenic and about 85% of D– individuals will make anti-D following infusion of 200 mL or more of D+ red cells. Anti-D can cause severe immediate or delayed haemolytic transfusion reactions and D+ blood must never be transfused to a patient with anti-D. Anti-D is the most common cause of severe haemolytic disease of the fetus and newborn (HDN). The effects of HDN caused by anti-D are, at its most severe, fetal death at about week 17 of pregnancy. If the infant is born alive, the disease can result in hydrops and jaundice. If the jaundice leads to kernicterus, this usually results in infant death or permanent cerebral damage. The prevalence of HDN due to anti-D has been substantially reduced by anti-D immunoglobulin prophylaxis.

In 1970, at the beginning of the anti-D prophylaxis programme, there were 1.2 deaths per 1000 births in England and Wales due to HDN caused by antiD; by 1989 this figure had been reduced to 0.02. Prediction of fetal Rh genotype by molecular methods

Knowledge of the molecular bases for D– phenotype has made it possible to devise tests for predicting fetal D type from fetal DNA. This is a valuable tool in assessing whether the fetus of a woman with anti-D is at risk from HDN. Most methods involve polymerase chain reaction (PCR) tests that detect the presence or absence of RHD. It is important to test for more than one region of RHD, so that hybrid genes responsible for partial D antigens do not give a false result, and to test for RHDy, so that this does not give rise to a falsepositive result. Until recently the usual source of fetal DNA has been amniocytes. These are obtained by amniocentesis, which has an inherent risk of fetal loss and of fetomaternal haemorrhage. It is now possible to use sensitive PCR technology to determine fetal D type from the small quantity of free fetal DNA present in maternal plasma, as early as 12 weeks into the pregnancy. This noninvasive form of fetal D typing is now provided as a reference service in a few countries. C and c, E and e

C/c and E/e are two pairs of allelic antigens produced by RHCE. The fundamental difference between C and c is a serine–proline substitution at position 103 in the second external loop of the CcEe protein (see Fig. 3.5), although the situation is more complex than that. E and e represent a proline–alanine substitution at position 226 in the fourth external loop. Taking into account the presence and absence of D, and of the C/c and E/e polymorphisms, eight different haplotypes can be recognized. The frequencies of these haplotypes and the shorthand symbols often used to describe them are shown in Table 3.3. Anti-c is clinically the most important Rh antigen after anti-D and may cause severe HDN. On the other hand, anti-C, anti-E and anti-e rarely 29

Chapter 3 Table 3.3 Rh phenotypes and the genotypes that produce them (presented in DCE and shorthand terminology).

Phenotype

Frequency (%) Asians‡

Genotypes

0.7

56.0

2.3

1.3

3.5

+

2.1

58.9

0.2

+

-

Rare

Rare

Rare

+

-

+

34.9

13.2

8.4

-

+

+

+

11.8

18.3

2.1

+

+

-

+

+

0.2

Rare

1.1

+

+

+

+

-

0.1

Rare

0.3

+

+

+

+

+

13.4

2.1

28.1

-

+ + + + + +

+ + + + + +

+ + + + + +

+ + + + + +

Rare Rare 15.1 Rare 0.1 0.1 Rare Rare Rare

0.1 Rare 4.1 Rare 1.3 Rare Rare Rare Rare

Rare Rare 0.1 Rare 0.1 Rare Rare Rare Rare

DCe/Dce DCe/dCe DcE/DcE DcE/dcE Dce/dce Dce/Dce DCE/DCE DCE/dCE DCe/dce DCe/Dce Dce/dCe DcE/dce DcE/Dce Dce/dcE DCe/DCE DCE/dCe DCe/dCE DcE/DCE DCE/dcE DcE/dCE DCe/DcE DCe/dcE DcE/dCe DCE/dce Dce/DCE Dce/dCE dCe/dCe dcE/dcE dce/dce dCE/dCE dCe/dce dcE/dce dCe/dCE dcE/dCE dcE/dCe dCE/dce

D

C

c

E

e

Europeans*

+

+

-

-

+

18.5

+

-

+

+

-

+

-

+

-

+

+

-

+

+

+

* English donors. † Yoruba of Nigeria. ‡ Cantonese of Hong Kong.

30

Africans†

R1/R1 R1r¢ R2R2 R2r≤ Ror RoRo RzRz Rzry R1r R1Ro Ror¢ R2r R2Ro Ror≤ R1Rz Rzr¢ R1ry R2Rz Rzr≤ R2ry R1R2 R1r≤ R2r¢ Rzr RoRz Rory r¢r¢ r≤r≤ rr ryry r¢r r≤r r¢ry r≤ry r≤r¢ ryr

Human blood group systems

cause HDN and when they do the disease is generally mild, though all have caused severe disease. Other Rh antigens

Of the 49 Rh antigens, 20 are polymorphic (i.e. have a frequency between 1 and 99% in at least one major ethnic group), 21 are rare antigens and eight are very common antigens. Antibodies to many of these antigens have shown themselves to be clinically important and it is prudent to treat all Rh antibodies as being potentially clinically significant.

Other blood group systems Of the remaining blood group systems (see Table 3.1), the most important clinically are Kell, Duffy, Kidd and MNS, and are described below. Kell system

The original Kell antigen, K (KEL1), has a frequency of about 9% in Caucasians, but is rare in other ethnic groups. Its allelic antigen, k (KEL2), is common in all populations. The remainder of the Kell system consists of one triplet and three pairs of allelic antigens: Kpa, Kpb and Kpc; Jsa and Jsb; K11 and K17; K14 and K24; plus 11 highfrequency and four low-frequency antigens. All represent single amino acid substitutions in the Kell glycoprotein. Anti-K can cause severe haemolytic transfusion reactions and HDN. About 10% of K-negative patients who are given one unit of K-positive blood produce anti-K, making K the next most immunogenic antigen after D. About 0.1% of all cases of HDN are caused by anti-K; most of the mothers will have had previous blood transfusions. HDN caused by anti-K differs from Rh HDN in that anti-K appears to cause fetal anaemia by suppression of erythropoiesis, rather than immune destruction of mature fetal erythrocytes. Anti-K is a very rare antibody. It is always immune and has been incriminated in some cases of mild HDN. Most other Kell-system antibodies are rare and best detected by an antiglobulin test.

The Kell antigens are located on a large glycoprotein that crosses the cell membrane once and has a glycosylated C-terminal extracellular domain, maintained in a folded conformation by multiple disulphide bonds. The Kell glycoprotein belongs to a family of endopeptidases, which process biologically important peptides, and are able to cleave the biologically inactive peptide big endothelin-3 to produce endothelin-3, an active vasoconstrictor. Duffy system

Fya and Fyb represent a single amino acid substitution in the extracellular N-terminal domain of the Duffy glycoprotein. Their incidence in Caucasians is 66% Fya and 80% Fyb. About 70% of AfricanAmericans and close to 100% of West Africans are Fy(a–b–) (Table 3.4). They are homozygous for an Fyb allele containing a mutation in a binding site for the erythroid-specific GATA-1 transcription factor, which means that Duffy glycoprotein is not expressed in red cells, although it is present in other tissues (Table 3.5). The Duffy glycoprotein is the receptor exploited by Plasmodium vivax merozoites for penetration of erythroid cells. Consequently, the Fy(a–b–) phenotype confers resistance to P. vivax malaria. The Duffy glycoprotein (also called Duffy antigen chemokine receptor, DARC) is a red cell receptor for a variety of chemokines, including interleukin-8. Anti-Fya is not infrequent and is found in previously transfused patients who have usually already made other antibodies. It can cause haemolytic transfusion reactions, but seldom causes HDN. Anti-Fyb is very rare.

Table 3.4 Duffy system: phenotypes and genotypes.

Frequency (%) Phenotype

Genotype

Europeans

Africans

Fy(a+b-) Fy(a+b+) Fy(a-b+) Fy(a-b-)

Fy a/Fy a or Fy a/Fy Fy a/Fy b Fy b/Fy b or Fy b/Fy Fy/Fy

20 48 32 0

10 3 20 67

31

Chapter 3 Table 3.5 Nucleotide polymorphisms

Allele

GATA box sequence -64 to -69 (promoter)

Codon 42 (exon 2)

Antigen

Fya Fyb Fy

TTATCT TTATCT TTACCT

GGT (Gly) GAT (Asp) GAT (Asp)

Fya Fyb Red cells: none Other tissues: probably Fyb

in the promoter region and in exon 2 of the three common alleles of the Duffy gene.

Kidd system

MNS system

Kidd has two alleles, Jka and Jkb, which represent a single amino acid change in the Kidd glycoprotein. Jka and Jkb antigens both have frequencies of about 75% in Caucasian populations. A Kidd-null phenotype, Jk(a–b–), results from homozygosity for inactivating mutations in the Kidd gene, SLC14A1. It is very rare in most populations, but reaches an incidence of greater than 1% in Polynesians. The Kidd glycoprotein is a urea transporter in red cells and in renal endothelial cells. Anti-Jka is uncommon and anti-Jkb is very rare, but they both cause severe transfusion reactions and, to a lesser extent, HDN. Kidd antibodies have often been implicated in delayed haemolytic transfusion reactions. They are often difficult to detect serologically and tend to disappear rapidly after stimulation.

MNS, with a total of 43 antigens, is second only to Rh in complexity. These antigens are present on one or both of two red cell membrane glycoproteins, glycophorin A (GPA) and glycophorin B (GPB). They are encoded by two homologous genes, GYPA and GYPB, on chromosome 4. The M and N antigens, both with frequencies of about 75%, differ by amino acids at positions 1 and 5 of the external N-terminus of GPA. S and s have frequencies of about 55 and 90%, respectively, in a Caucasian population, and represent an amino acid substitution in GPB. About 2% of black West Africans and 1.5% of AfricanAmericans are S– s–, a phenotype virtually unknown in other ethnic groups, and most of these lack the U antigen, which is present when either S or s is expressed. The numerous MNS variants mostly result from amino acid substitutions in GPA or GPB and from the formation of hybrid GPA–GPB molecules, resulting from intergenic recombination between GYPA and GYPB. GPA and GPB are exploited as receptors by the malaria parasite Plasmodium falciparum. Anti-M and anti-N are not generally clinically significant, though anti-M is occasionally haemolytic. Anti-S, the rarer anti-s, and anti-U can cause HDN and have been implicated in haemolytic transfusion reactions.

Diego system

Diego is a large system of 21 antigens: two pairs of allelic antigens (Dia and Dib, Wra and Wrb) plus 17 antigens of very low frequency. All represent single amino acid substitutions in band 3, the red cell anion exchanger. The original Diego antigen, Dia, is very rare in Caucasians and black people, but relatively common in Mongoloid people, with frequencies varying between 1% in Japanese and 50% in some native South Americans. Anti-Dia and anti-Dib are immune and rare, but can cause HDN. Wra has a frequency of about 0.1%. Its high-frequency allelic antigen, Wrb, is dependent on an interaction of band 3 with glycophorin A for its expression. Naturally occurring anti-Wra is present in approximately 1% of blood donors. Very rarely, anti-Wra causes HDN. 32

Biological significance of blood group antigens The functions of several red cell membrane protein structures bearing blood group antigenic determinants are known, or can be deduced, from their

Human blood group systems

structure. Some are membrane transporters, facilitating the transport of biologically important molecules through the lipid bilayer: band 3 membrane glycoprotein, the Diego antigen, provides an anion exchange channel for HCO3– and Cl– ions; the Kidd glycoprotein is a urea transporter; the Colton glycoprotein is aquaporin 1, a water channel; the GIL antigen is aquaporin 3, a glycerol transporter; and the Rh protein complex might function as an ammonium transporter or a CO2 channel. The Lutheran, LW and Indian (CD44) glycoproteins are adhesion molecules, possibly serving their primary functions during erythropoiesis. The Duffy glycoprotein is a chemokine receptor and could function as a ‘sink’ or scavenger for unwanted chemokines. The Cromer and Knops antigens are markers for decay accelerating factor (CD55) and complement receptor 1 (CD35), respectively, which protect the cells from destruction by autologous complement. Some blood group glycoproteins have enzyme activity: the Yt antigen is acetylcholinesterase and the Kell antigen is an endopeptidase, though their functions on red cells are not known. The Cterminal domains of the Gerbich antigens, GPC and GPD, and the N-terminal domain of the Diego glycoprotein, band 3, are attached to components of the cytoskeleton and function to anchor it to the external membrane. The carbohydrate moieties of the membrane glycoproteins and glycolipids, especially those of the most abundant glycoproteins (band 3 and GPA), constitute the glycocalyx, an extracellular coat that protects the cell from mechanical damage and microbial attack. The structural differences between allelic red cell antigens (e.g. A and B, K and k, Fya and Fyb) are small, often being just one monosaccharide or one amino acid. The biological importance of these differences is unknown and there is little evidence to suggest that the product of one allele confers any significant advantage over the other. Some blood group antigens are exploited by pathological microorganisms as receptors for attaching and entering cells, so in some cases absence or changes in these antigens could be beneficial. It is likely that interaction between cell surface molecules and pathological microorganisms has been a major factor in the evolution of blood group polymorphism.

Summary Blood groups are inherited characters of the red cell surface detected by specific alloantibodies. Most of the authenticated 285 blood group specificities are organized into 29 blood group systems, each representing a single gene or a cluster of closely linked homologous genes. Blood group antibodies are of clinical importance in transfusion medicine as they can cause haemolytic transfusion reactions and HDN. The carbohydrate ABO antigens and protein Rh antigens are the most important from the clinical aspect. Blood group proteins serve a variety of functions, though little is known about the biological significance of blood group polymorphism.

Further reading Avent ND. Fetal genotyping. In: Hadley A, Soothill P, eds. Alloimmune Disorders of Pregnancy. Cambridge: Cambridge University Press, 2002: 121–39. Avent ND, Reid ME. The Rh blood group system: a review. Blood 2000; 95: 375–87. Chester AM, Olsson ML. The ABO blood group gene: a locus of considerable genetic diversity. Transfus Med Rev 2001; 15: 177–200. Daniels G. Functional aspects of red cell antigens. Blood Rev 1999; 13: 14–35. Daniels G. Human Blood Groups, 2nd edn. Oxford: Blackwell Science, 2002. Daniels G, Poole J, de Silva M et al. The clinical significance of blood group antibodies. Transfus Med 2002; 12: 287–95. Daniels GL, Fletcher A, Garratty G et al. Blood group terminology 2004. Vox Sang 2004, in press. Denomme GA. The structure and function of the molecules that carry human red blood cell and platelet antigens. Transfus Med Rev 2004; 18: 203–31. Henry S, Samuelsson B. ABO polymorphisms and their putative biological relationships with disease. In: King M-J, ed. Human Blood Cells: Consequences of Genetic Polymorphism and Variations. London: Imperial College Press, 2000: 1–103. Reid ME, Lomas-Francis C. The Blood Group Antigen Facts Book. San Diego: Academic Press, 1997. Watkins WM (ed.) Commemoration of the centenary of the discovery of the ABO blood group system. Transfus Med 2001; 11: 239–351.

33

Chapter 4

Human leucocyte antigens Cristina V. Navarrete

The human leucocyte antigen (HLA) system consists of a family of cell surface polymorphic molecules involved in the presentation of antigen to T cells and therefore plays a central role in the induction and regulation of immune responses. HLA molecules are also known to be involved in the pathogenesis of certain autoimmune and infectious diseases and they have an important influence on the outcome of solid organ and haemopoietic stem cell transplantation. Furthermore, HLA antigens present in blood cells are responsible for some of the serious clinical complications of blood transfusion. The genes coding for the HLA molecules are located on the short arm of chromosome 6 and span a distance of approximately 4 Mb. This region is divided into three subregions. • Class I subregion contains genes coding for the heavy chain of the classical (HLA-A, -B and -C) and non-classical (HLA-E, -F and -G) HLA class I molecules; • Class II subregion contains genes coding for the HLA class II molecules (DR, DQ and DP) and genes involved in the processing and transport of antigenic peptides. • Class III subregion lies between the other two subregions and contains genes coding for a diverse group of proteins, including complement components (C4Bf), tumour necrosis factor (TNF) and heat-shock proteins. A number of additional genes, including the major histocompatibility complex (MHC) class I chain-related gene A (MICA) and MHC class I chain-related gene B (MICB) also involved in immune responses, have been mapped between the class I and class III subregions (Fig. 4.1). In addition, the non-classical class I-like gene HFE has 34

been mapped to a locus located 4 Mb telomeric to HLA-F. Mutations in this gene have been shown to be responsible for the development of hereditary haemochromatosis. Following the development of recombinant DNA technology, it has been possible to perform a detailed analysis of the HLA region, leading to the unravelling of the genetic complexity and structure of its genes and molecules. These findings, combined with parallel studies on their function genes, have led to a better understanding of the role of the HLA system in clinical medicine.

HLA class I genes The HLA class I genes have been classified according to their structure, expression and function as classical (HLA-A, -B and -C) and non-classical (HLA-E, -F and -G). Both classical and non-classical HLA class I genes code for a heavy (a) chain, of approximately 43 kDa, non-covalently linked to a non-polymorphic light chain, the b2-microglobulin of 12 kDa, which is coded for by a gene on chromosome 15. The extracellular portion the heavy chain has three domains (a1, a2 and a3) approximately 90 amino acids long. These domains are encoded by exons 2, 3 and 4 of the class I gene, respectively. The a1 and a2 domains are the most polymorphic domains of the molecule and they form a peptide-binding groove that can accommodate antigenic peptides approximately eight to nine amino acids long. The exon/intron organization of the non-classical HLA class I genes (E, F and G) is very similar to the classical class I genes but they have a more restricted polymorphism.

Human leucocyte antigens HLA class III subregion

HLA class II subregion DMB DP

B2 A2

DOA DMA

LMP7 TAP1 TAP2 LMP2 DOB

B1 A1

DQ

MICB MICA TNF HSP70 B C

DR

B2 A2 B3 B1 A1

HLA class I subregion

B1 B2 B5 B3 B4

A

A

E

A H G F

B

Class II genes pseudogenes

Class II genes

ABC transporter genes

Proteosomelike genes

TNF

Classical class I genes

Non-classical class I genes

MICA and MICB

HSP70

Fig. 4.1 Map of the human leucocyte antigen complex. HSP, heat-shock protein; TNF, tumour necrosis factor.

α1

Peptide binding region

α2 S S

N Immunoglobulinlike region

β2-m

S

S

S

S

α3

C Transmembrane region P P P

Cytoplasmic region

HLA class A gene Fig. 4.2 HLA class I molecule. b2-m,

b2-microglobulin.

A schematic representation of the classical HLA class I gene and molecule is shown in Fig. 4.2.

HLA class II genes The class II DR, DQ and DPA and DPB genes

C

Regulatory sequences

L

α1

α2

α3

TM

CYT

3'UT

1

2

3

4

5

6 7

8

Exons

code for a heterodimer formed by two noncovalently associated a and b chains of approximately 34 and 28 kDa respectively. The expressed a and b chains consist of two extracellular domains as well as transmembrane and cytoplasmic domains. The a1/b1 and a2/b2 domains are encoded by exon 2 and exon 3 of the class II gene 35

Chapter 4

respectively. The majority of the polymorphism is located in the b1 domain of the DR molecules and in the a1 and b1 domains of the DQ and DP molecules. Similarly to the class I molecules, these domains also form a peptide-binding groove. However, in the case of the class II molecules (DR), the groove is open at both sides and it can accommodate antigenic peptides of varying size, although most of them are approximately 13–25 amino acids long. A schematic representation of the HLA class II gene and molecule is shown in Fig. 4.3.

Expression of HLA class II genes There is one non-polymorphic DRA and nine DRB genes, of which B1, B3, B4 and B5 are highly polymorphic and B2, B6 and B9 are pseudogenes. The main DR specificities (DR1–DR18) are determined by the polymorphic DRB1 gene. Furthermore, the number of DRB genes expressed in each

α1

Peptide binding region

individual varies according to the DRB1 allele expressed, e.g. HLA-DR1, -DR103, -DR10 and -DR8 alleles express the DRB1 gene only. DR15 and DR16 alleles express the DRB1 and DRB5 genes, which code for the DR51 product; HLADR17, -DR18, -DR11, -DR12, -DR13 and -DR14 alleles express the DRB1 and the DRB3 genes, which code for the DR52 specificity; and, finally, the HLA-DR4, -DR7 and -DR9 alleles express the DRB1 and the DRB4 genes coding for the DR53 product (Fig. 4.4). There are a few exceptions to this pattern of gene expression, e.g. a DRB5 gene has been found to be expressed with some DR1 alleles and non-expressed or null DRB5 and DRB4 genes have also been identified. In contrast to the DRB genes, there are two DQA and three DQB genes, of which only the DQA1 and DQB1 are expressed and both are polymorphic. Similarly, there are two DPA and two DPB genes, of which only the DPA1 and DPB1 are expressed and both are polymorphic. There are additional genes located in the class II

β1 NN S S

Immunoglobulinlike region

α2

Transmembrane region

S

S

S

Papain S cleavage sites

C

β2

C

Cytoplasmic region

Exons HLA class A gene HLA class B gene

1 L

2 α1

3 α2

5 3'UT

Regulatory sequences L 1

β1 2

β2 3 Exons

36

4 TM/CYT

TM CYT 3'UT 4 5 6 Fig. 4.3 HLA class II molecule.

Human leucocyte antigens Specifications DRB1

DRB6 Y

Dr1, DR10, DR103, DR15 DRB1 DR15, DR16, DR1

DRB9

DRA

Y

DRB6 Y

DRB5

DRB2 Y

DRB3

DRB9

DRA

Y DR51

DR17, DR18, DR11, DR12 DR13, DR14, DR1403, DR1404

DRB1

DR52 DRB9

DR8 DRB1 Fig. 4.4 Expression of HLA-DRB genes.

The classical HLA class I molecules (A, B, C) are expressed on the majority of tissues and cells, including T and B lymphocytes, granulocytes and platelets. Low levels of expression have been

DRA

Y

DR4, DR7, DR9

Distribution of HLA molecules

DRA

Y

DRB1

region which are involved in the MHC class I antigen presentation pathway. These include the low-molecular-mass polypeptide genes (LMP2 and LMP7) and the transporter associated with antigen-processing genes (TAP1 and TAP2). The LMP2 and LMP7 genes are thought to improve the capacity of the proteosomes to generate peptides of the appropriate size and specificity to associate with the class I molecules. Conversely, the TAP1 and TAP2 genes are primarily involved in the transport of the proteosome-generated peptides to the endoplasmic reticulum, where they associate with the class I molecules. In addition, the DMA and DMB genes, which code for a heterodimer involved in the loading of peptides presented by HLA class II molecules, are also located in this subregion (see Fig. 4.1). The main function of the DM molecules is to facilitate the release of the class II-associated invariant chain peptide from the peptide-binding groove of the HLA class II molecule so that it can be exchanged for the relevant antigenic peptide.

DRB9

DRB7

DRB8

Y

Y

DRB4

DRB9

DRA

Y DR53

detected in endocrine tissue, skeletal muscle and cells of the central nervous system. HLA-E and -F are also expressed on most tissues tested but HLAG shows a more restricted tissue distribution and to date the HLA-G product has only been found to be expressed on extravillous cytotrophoblasts of the placenta and mononuclear phagocytes. The HLA class II molecules are constitutively expressed on B lymphocytes, monocytes and dendritic cells but can also be detected on activated T lymphocytes and activated granulocytes. It is not clear whether they are also present on activated platelets. HLA class II expression can be also be induced on a number of cells and tissues such as fibroblasts and endothelial cells as the result of activation and/or the effect of certain inflammatory cytokines, such as interferon (IFN)-g, TNF and interleukin (IL)-10.

Genetics One of the main features of the HLA genes is their high degree of polymorphism and the strong linkage disequilibrium (LD) in which they segregate. LD is a phenomenon where the observed frequency of alleles of different loci segregating together is greater than the frequency expected by random association. The difference between the 37

Chapter 4

observed and the expected frequencies for a particular combination of alleles is called the delta value and is positive for alleles in LD with each other. Whereas some of the polymorphism and the patterns of LD are expressed with similar frequencies in all populations, others are unique to some population groups. For example, HLA-A2 is expressed at a relatively high frequency in most population groups studied so far, whereas B53 is found predominantly in black people. In addition, all HLA genes are codominantly expressed and are inherited in a mendelian fashion and the genetic region containing all HLA genes on each chromosome is termed the haplotype. Some haplotypes are also found across different ethnic groups, e.g. HLA-B44-DR7, whereas others are unique to a particular population, e.g. HLA-B42DR18 in black Africans. This characteristic is particularly relevant for the selection of HLAcompatible family donors for patients requiring solid organ or bone marrow transplantation (BMT).

Function of HLA molecules HLA molecules are directly involved in the presentation of antigenic peptides to T cells. This is a highly regulated process and requires a fine interaction between the HLA molecules, the antigenic peptide, the T-cell receptor and a number of costimulatory molecules (e.g. CD80, CD86) and adhesion molecules such ICAM-1(CD54) and LFA-3 (CD58). The HLA class I molecules are primarily but not exclusively involved in the presentation of endogenous antigenic peptides to CD8+ cytotoxic T cells. However, it has now been shown that both classical and non-classical HLA class I molecules also interact with a new family of receptors present on natural killer (NK) cells. Some of these receptors, which are polymorphic and differentially expressed, have an inhibitory role whereas others are involved in NK cell activation. The killeractivating and killer-inhibitory receptors belong to two distinct families: the immunoglobulin superfamily called killer immunglobulin receptors

38

(KIRs), and the C-type lectin superfamily CD94NKG2. The interaction between the inhibitory receptors and the relevant HLA ligand results in the prevention of NK lysis of the target cell. Thus NK cells from any given individual will be alloreactive towards cells lacking their corresponding inhibitory KIR ligands, e.g. tumour or allogeneic cells. In contrast NK cells will be tolerant to cells from individuals who express the corresponding KIR ligands. This information is currently being exploited in the clinical setting in order to promote the graft-versus-leukaemia (GVL) effect mediated by NK cells. HLA class II molecules are mostly involved in the presentation of exogenous antigenic peptides to CD4+ helper T cells. Once activated, these CD4+ cells can initiate and regulate a variety of processes leading to the maturation and differentiation of cellular (CD8+ cytotoxic T cells) and humoral effectors by the secretion of proinflammatory cytokines (IL-2, IFN-g, TNF-a) and regulatory cytokines (IL-4, IL-10 and transforming growth factor-b). The nature of the peptide presented by class I and II molecules is largely dependent on the location of the peptide within the processing machinery of the cells.

Identification of HLA gene polymorphism Traditionally, characterization of HLA polymorphisms has been carried out using serological and cellular techniques. There are, however, several limiting factors in the use of serological typing methods: • it is often difficult to obtain antibody of sufficient titre and specificity to distinguish between all described HLA types; • the antibodies available are very often of Caucasian origin, making it difficult to HLA type patients from other population groups; • patients undergoing chemotherapy normally have low white cell counts and cell antigen expression can be affected by the chemotherapy; • the requirement for viable cells, ideally 24 h after venepuncture, means that serological HLA typing results are not always reliable; and

Human leucocyte antigens

• because class II molecules are expressed on B cells and not (resting) T cells, the low number of B cells in a sample makes serological class II typing difficult. However, with the development of gene cloning and DNA sequencing it has been possible to perform a detailed analysis of these genes at the single nucleotide level. This analysis has shown the existence of shared nucleotide sequences between alleles of the same and/or different loci and the existence of certain locus-specific nucleotide sequences in both the coding (exons) and noncoding (introns) regions of the genes. The DNA sequencing of a number of alleles of various loci has also demonstrated that the majority of the polymorphism is located in regions of the a1 and a2 domain of the class I molecules and of the a1 and b1 domain of the class II molecules. These are called hypervariable regions. Based on this information, a number of techniques have been developed to characterize HLA polymorphisms. Most of the described techniques make use of the polymerase chain reaction (PCR) to amplify the specific genes or region to be analysed. These techniques include PCR-SSOP (sequence-specific oligonucleotide probing), PCRSSP (sequence-specific priming), and conformational methods including reference strand conformational analysis (RSCA) and sequencingbased typing (SBT). The number of recognized serologically defined antigens and DNA-identified HLA alleles is shown in Table 4.1. Sequence-specific oligonucleotide probing

In this technique the gene of interest is amplified using generic primers, i.e. primers designed to anneal with DNA sequences common to all alleles of the loci of interest. The amplified PCR product is then immobilized onto support (e.g. nylon) membranes and the specificity of the products analysed by reacting the membranes with labelled oligonucleotides designed to anneal with polymorphic sequences present in each allele. By scoring the probes that bind to specific regions, it is possible to assign the HLA type. A recent modification of this technique, called

Table 4.1 Number of recognized HLA antigens/alleles.

(From Marsh et al. 2002 with permission.) Alleles

Antigens

HLA class I HLA-A HLA-B HLA-C119 HLA-E HLA-F HLA-G

250 490 9 6 1 15

24 50

HLA class II HLA-DRB1 HLA-DRA1 HLA-DRB3 HLA-DRB4 HLA-DRB5 HLA-DQB1 HLA-DQA1 HLA-DPB1 HLA-DPA1

315 3 38 12 150 53 22 99 20

17 — 1 1 1 6 — 6* —

* Cellularly defined.

reverse blot, involves the addition of the PCRamplified product to labelled probes immobilized on membranes (strips) or plates. PCR-SSOP is useful when large numbers of samples need to be HLA typed, e.g. bone marrow or cord blood donors. Sequence-specific priming

This technique involves the use of sequencespecific primers in the PCR step, i.e. primers designed to anneal with DNA sequences unique to each allele and locus. The detection of the PCRamplified product is then carried out by running the product on an agarose gel. This technique allows the rapid identification of the HLA alleles in individual samples since the readout of this method is the presence or absence of the product for which specific primers were used. This technique is therefore ideally suited to HLA typing individual samples, e.g. patients requiring HLAmatched platelets. However, although this is a very rapid procedure, many PCR reactions have to be set up per sample, e.g. at least 24 reactions for low-

39

Chapter 4

resolution DR typing. Furthermore, for PCR-SSP typing the target sequence of the alleles must be known since novel unknown sequences may not always be detected. Conformational analysis methods

Conformational methods depend on the mobility of PCR products in the gels. A number of variations of this technique have been described, including single-strand conformational polymorphism analysis (SSCP), double-strand conformational polymorphism analysis (DSCA), RSCA and heteroduplex formation. In the SSCP technique, PCR-generated DNA products are denatured by heating and rapid cooling to prevent reannealing of the strands. The products are run on a polyacrylamide gel with the mobility depending upon the secondary structure of the single-stranded DNA. The major disadvantage of this approach for HLA typing is the tendency of single-stranded DNA to adopt many conformational forms under the same electrophoretic conditions, resulting in the presence of several bands from a single product. A modification of this technique that compares the mobility in polyacrylamide of duplex molecules generated by mixing PCR products is called DSCA. In this case, the mobility depends on the mismatching of the sequence and the formation of heteroduplex molecules. A further modification of this technique, the RSCA, has recently been developed and successfully applied to HLA class I and II (DP) typing. In this technique the PCR amplification is carried out on the test DNA and on a reference DNA of known sequence, using fluorescently labelled primers for the PCR of the reference DNA. The PCR products are then mixed, allowed to anneal and run in an automated sequencer. Only those duplexes containing a labelled strand are detected, i.e. the homoduplex of reference DNA and the heteroduplexes of reference and test DNA. The mobility of every known allele with the reference DNA is then established and used to compare with that of the DNA under test. Heteroduplexing

Heteroduplexing is another DNA conformational 40

technique based on the fact that mismatched DNA hybrids (heteroduplexes) migrate at a slower rate through gels (due to the presence of singlestranded loops) than fully matched DNA duplexes. Heteroduplexes are formed during the annealing stage of the PCR when the sense strand of an allele binds to the antisense strand of a different allele. The banding pattern of these products following electrophoresis through a gel can be used to identify the alleles present. This technique is particularly useful for performing DNA-based ‘crossmatching’. In this case, HLA genes from the patient and potential donor are amplified, denatured by heating to about 95°C for several minutes and then mixed together under conditions which promote reannealing. Where the donor and patient alleles are similar but not identical, then, as in RSCA, homoduplexes are formed together with chimeric or heteroduplexes where sense strands from the patient anneal to antisense strands from the donor and vice versa. These heteroduplexes would have different migration rates through gels when compared with the homoduplexes. When this mix is electrophoresed through a gel, the banding pattern observed would contain all the bands that would be observed if PCR products from the patient and donor were to be electrophoresed separately. However, in addition to these there will be extra bands that are due to the heteroduplexes. The presence of extra bands in a heteroduplex ‘crossmatch’ would therefore indicate a difference in HLA type between the patient and donor. These extra bands would not be observed if the HLA alleles of the patient and donor were identical. The sensitivity of the technique can be increased by adding DNA from a known HLA allele that is not present in the patient or donor. One advantage of typing by conformational analysis when compared with methods such as PCR-SSP and PCR-SSOP is that new mutations not previously described can be readily picked up. A disadvantage of the technique is that, like SSCP, it cannot detect some HLA class II combinations while detecting too many silent mutations. Sometimes the banding patterns observed are

Human leucocyte antigens

complex and difficult to interpret. It also suffers in that it provides no actual HLA typing information. One of the main disadvantages of all the above described techniques is that they require DNA sequence data in order to design primers and probes. Sequencing-based typing

The principle of DNA sequencing is relatively straightforward and involves the denaturation of the DNA to be analysed to provide a single-strand template. A sequencing primer is then added and the DNA extension is performed by the addition of Taq polymerase in the presence of excess nucleotides. The sequencing mixture is divided into four tubes, each of which contains specific dideoxyribonucleoside triphosphate (ddATP). When this is incorporated into the DNA synthesis, elongation is interrupted with chain-terminating inhibitors. In each reaction there is random incorporation of the chain terminators and therefore products of all sizes are generated. The products of the four reactions are then analysed by electrophoresis in parallel lanes of a polyacrylamide–urea gel and the sequence is read by combining the results of each lane using an automated DNA sequencer. The sequencing products are detected by labelling the nucleotide chain inhibitors with radioisotopes and, more recently, with fluorescent dyes. In HLA SBT, some ambiguous results can be obtained with heterozygous samples and these may need to be retested by PCR-SSP. HLA typing using the RSCA and SBT techniques permit high-resolution HLA typing, which is known to be important in the selection of HLAmatched unrelated donors. A major advantage of all DNA-based techniques is that no viable cells are required to perform detailed HLA class I and II typing. Furthermore, since all the probes and primer are synthesized to order, there is a consistency of reagents used, allowing the comparison of HLA types from different laboratories. However, although serological typing is being rapidly replaced by DNA-based typing techniques, serological reagents may still be required for antigen expression studies.

The advantages and disadvantages of the various techniques described above are given in Table 4.2.

HLA antibodies HLA-specific antibodies may be produced in any situation that exposes the host to these alloantigens, including pregnancy, transplantation, blood transfusions and planned immunizations. However, the affinity, avidity and class of the antibody produced will depend on various factors, including the route of immunization, the persistence and type of immunological challenge and the immune status of the host. Cytotoxic HLA antibodies can be identified in approximately 20% of human pregnancies. The antibodies produced are normally multispecific, high titre, high affinity and of the IgG class. Although these IgG antibodies can cross the placenta, they have not been shown to be harmful to the fetus. Conversely, the antibodies produced following transplantation seem to be largely dependent on the degree of HLA mismatch between donor and recipient and the majority of these antibodies formed are IgG, although a few IgM antibodies have also been identified. In contrast, the majority of HLA antibodies found in multitransfused patients are multispecific IgM and IgG and are mostly directed at public epitopes. The recent introduction of leucocytedepleted blood components (see Chapter 22) may lead to a reduction in alloimmunization in naive recipients but it may not be very effective in preventing alloimmunization in already sensitized recipients, i.e. women who have become immunized as a result of pregnancy. Finally, normal individuals were previously deliberately immunized with HLA-mismatched cells in order to produce potent HLA-specific reagents, but the deliberate immunization of these healthy individuals is nowadays difficult to justify ethically. However, at present, planned HLA immunization is still carried out as a form of treatment for women with a history of recurrent spontaneous abortion. These women are immunized with lymphocytes from their partners or a third party to attempt to induce an immunomodulatory 41

Chapter 4 Table 4.2 Advantages and disadvantages of DNA-based techniques.

Technique

Advantages

Disadvantages

Sequence-specific oligonucleotide probing (SSOP)

Needs only one pair of genetic primers, fewer reactions to set up Larger number of samples can be processed simultaneously Requires small amount of DNA Cheap

Different temperatures required for each probe Probes can cross-react with different alleles Large numbers of probes required to identify specificity Difficult to interpret pattern of reactions

Sequence-specific priming (SSP)

Provides rapid typing with higher resolution than SSOP All PCR amplifications are carried out at same time, temperature and conditions Fast and simple to read and interpret

Too many sets of primers are needed to fully HLA type Requires a two-stage amplification to provide HR typing

Reference strand conformational analysis (RSCA)

Easy to perform Provides higher resolution than SSOP and SSP

Requires expensive equipment Requires established data on viability value of each allele studied

Sequencing-based typing (SBT)

Provides the highest level of resolution Able to identify new alleles Does not require previous sequence data to identify new allele

Not easy to perform Requires expensive reagents and equipment Difficult to interpret Requires DNA sequence data to compare results Slower than SSOP (reverse blot), SSP and RSCA

response that results in the maintenance of the pregnancy.

Detection of HLA antibodies Over the years, a number of techniques to detect HLA antibodies have been described. These include the complement-dependent lymphocytotoxicity (LCT) test, enzyme-linked immunosorbent assay (ELISA) and flow cytometry. More recently, a new method to detect HLA antibodies, Luminex, has been described. Lymphocytotoxicity

The LCT assay, developed by Terasaki and McClelland (1964), is the most commonly used. In this technique, equal volumes of serum and cells are mixed and incubated to allow the binding of the specific antibody to the target cell. This is followed by the addition of rabbit complement and a further incubation step. Complement-fixing antibodies reacting with the HLA antigen present on 42

the cell surface leads to the activation of complement via the classical pathway and results in the disruption of the cell membrane. The lysed cells are then detected by adding ethidium bromide (EB) and acridine orange (AO) at the end of the incubation period. Live cells actively take up AO and under ultraviolet (UV) light they appear green, whereas lysed cells allow the entry of EB, which binds to DNA, and they appear red under UV light. There are a number of alternatives to EB and AO including carboxyfluorescein diacetate and EB, eosin or trypan blue. The reactions are scored by estimation of the percentage of dead cells in each well after establishing baseline values in the negative and positive controls (Table 4.3). One of the main disadvantages of the LCT assay is that it does not discriminate between HLA and non-HLA cytotoxic lymphocyte-reactive antibodies including autoantibodies, which are common in thrombocytopenic patients, particularly in post-BMT patients, in whom there is a dysfunction of the immune system. Fortunately, the majority of these lymphocytotoxic autoantibodies are IgM and can be identified by screening the

Human leucocyte antigens Table 4.3 Lymphocytotoxicity grading.

Cell death (%)

Score

Interpretation

0–10 11–20 21–50 51–80 81–100

1 2 4 6 8 0

Background cell death, negative Doubtful negative Weak positive Positive Strong positive Unreadable/invalid

serum with and without dithiothreitol (DTT). The addition of DTT to the serum results in the breakdown of the intersubunit disulphide bonds in the IgM molecule, leading to the loss of cytotoxicity due to IgM. Prolonged exposure or excess DTT can lead to the breakdown of intramolecular disulphide bonds in the IgG molecules and also inactivate complement, but this can be inhibited by the addition of cystine. In our laboratory, we tested 104 serum samples from immunologically refractory patients receiving HLA-matched platelet transfusions, 50% of the samples being positive by LCT but only 40% being positive when screened with DTT, illustrating the importance of using DTT to remove the reactivity due to IgM, non-HLA lymphocytotoxic antibodies (10%). The presence of lymphocytotoxic autoreactive antibodies in itself is not thought to be of clinical significance in solid organ transplant recipients; however, the clinical relevance of these antibodies in immunological refractoriness to random platelet transfusions is not yet clear. Another disadvantage of the LCT test is that it only detects cytotoxic HLA-specific antibodies and therefore the presence of non-cytotoxic HLAspecific antibodies may be missed. Non-cytotoxic antibodies are best detected by using an ELISA or by flow cytometry, as described below. Enzyme-linked immunosorbent assay

ELISA-based methods have often been the technique of choice for antibody detection for a number of antigen systems, particularly where there has been a requirement for testing large numbers of samples.

The basic principle of the technique is as follows. HLA antigens are purified and immobilized on a microwell plate, directly or via an antibody directed against a non-polymorphic region of the HLA antigen. HLA-specific antibodies bound to the immobilized antigen can be detected with an enzyme-linked secondary antibody which, upon addition of specific substrate, catalyses a colour change reaction that is detected in an ELISA reader (GTI QuickScreen). This ELISA test can only be used to determine the presence or absence of HLA-specific antibody. In order to detect HLA specificity, an alternative technique is used. In this case, HLA antigen is isolated from a selected panel of cell donors or cell lines derived from these donors. Antibodies directed against the non-polymorphic region of the HLA class I molecule, i.e. the a3 domain, are used to immobilize the specific HLA antigen to the microwell, ensuring that the more polymorphic a1 and a2 domains are available for antibody binding. The test is designed to cover all the major HLA specificities at least once; thus it should be possible to determine antibody specificity in those sera which show a restricted panel reactivity. Three commercial kits are now available to perform the specificity screening, SangStat.-PRASTAT, GTI QuickID and Lambda Antigen Tray LAT. One of the main advantages of the ELISA technique is that the non-cytotoxic antibodies detected are HLA specific since it relies on the binding of the antibodies to wells coated with pools of solubilized HLA antigens. It is also more sensitive and in our laboratory the ELISA test achieves a 7% increase in sensitivity over the LCT test (49% vs. 42%). The precise identification of the type, specificity and titre of antibodies is important not only for the diagnosis but also to establish the best course of treatment, since in the case of patients immunologically refractory to platelet transfusions, the majority of these patients can benefit from the provision of HLA matched or crossmatched negative platelets (see below). Luminex

This technique uses fluorochrome-dyed poly43

Chapter 4

styrene beads coated with specific antigens. The precise ratio of these fluorochromes creates 100 distinctly coloured beads, each of them coated with a different antigen. The beads are then incubated with the patient’s serum and the reaction is developed using a PE-conjugated antihuman IgG (Fc specific) antibody. Flow cytometry

The use of flow cytometric techniques was initially investigated as an alternative crossmatch technique and was shown to be more sensitive than LCT. The increased sensitivity may be attributed to a number of factors, one of which is the additional reactivity due to the detection of non-cytotoxic antibodies, some of which may be HLA specific. As with LCT screening, cells and serum are mixed and incubated to allow the binding of the antibody to the target antigen. The bound antibody is detected by using an antibody labelled with a fluorescent marker such as fluorescein isothiocyanate or R-phycoerythrin against human immunoglobulin. The flow cytometer can then be used to identify the different cell populations based on their morphology/granularity and on the fluorescence. Test

sera with median fluorescence values greater than the mean + 3SD of the negative controls are considered positive. A modification of the above procedure can be used to detect antibodies reacting with T or B cells. The main advantages of flow cytometric techniques are the increased sensitivity when compared with LCT and the detection of noncomplement fixing antibodies, allowing the early detection of sensitization. However, one of the disadvantages is that it also detects non-HLA and lymphocyte-reactive antibodies, and the clinical relevance of these antibodies is unclear. The relative advantages and disadvantages of each of these techniques are given in Table 4.4.

Clinical relevance of HLA antigens and antibodies Although the main role of the HLA molecules is to present antigens to T cells, HLA molecules can themselves be recognized as foreign by host T cells by a mechanism known as allorecognition. Two pathways of allorecognition have been identified, direct and indirect.

Table 4.4 HLA antibody screening techniques.

Advantages

Disadvantages

Lymphocytotoxicity test (LCT)

Well established Robust Requires small amount of serum Used for antibody screening and crossmatching Low cost

Viable cells required Needs separated T and B cells for class I and class II antibody screening Large and selected panel of cells required Detects non-HLA cytotoxic antibodies, e.g. autoantibodies

Enzyme-linked immunosorbent assay (ELISA)

Easy to standardize Objective readout Suitable for bulk testing Detects cytotoxic and non-cytotoxic HLA-specific antibodies More sensitive than LCT Medium cost

Not yet well established to define class I and class II specificities Large amounts of serum required Cannot be used for crossmatching

Flow cytometry

Highly sensitive Detects weak and early sensitization Detects cytotoxic and non-cytotoxic antibodies Can define class I and class II antibodies simultaneously

Not well standardized Large panel of cells required to establish antibody specificity Expensive Detects non-HLA antibodies

44

Human leucocyte antigens

In the direct allorecognition pathway, the host’s T cells recognize HLA molecules (primarily class II) expressed on donor tissues, e.g. tissue dendritic cells and endothelial cells. Indirect allorecognition involves the recognition of donor-derived HLA class I and II antigenic peptides presented by the host’s own antigen-presenting cells. Because of this mechanism, HLA antigens are therefore one of the main barriers to the success of solid organ transplantation or BMT and are responsible for the strong alloimmunization seen in patients following transplantation or blood transfusion. Solid organ transplantation

Several studies have confirmed the importance of matching for HLA-A, -B and -DR antigens. Results published by the United Network for Organ Sharing (UNOS) registry have shown that 1-year graft survival in recipients of fully HLA-A, -B and -DR matched kidneys was 94%, as opposed to the 89% and 90% survival rate observed in recipients of one HLA haplotype-matched kidney from a parent or a sibling, respectively. The differences between match grades, which are minimal at 1 year after transplant, become more apparent with increased follow-up half-life figures. Studies on the relative contribution of HLA-A, -B and -DR compatibility on the outcome of graft survival have shown that HLA-DR has the strongest effect. When grafts from cadaver donors are analysed, the 1-year graft survival rate is 88% for HLA-A, -B and -DR matched and 79% for mismatched kidneys. With the application of the PCR-based techniques, it is now possible to identify molecular differences between otherwise serologically identical HLA types of donor and recipient pairs. Correlation of these results with graft survival has shown a higher graft survival rate when recipients and donors are HLA-DR identical by serological and molecular techniques compared with when they are HLA-DR identical by serological but not molecular methods (87% vs. 69%). The use of DNA-based techniques allows the identification of ‘minor’ mismatches that were previously undetected by serological typing, particularly in the DRb1 chain.

The role of HLA antibodies in solid organ transplantation is well established. The presence of circulating HLA-specific antibodies directed against donor antigens in renal and cardiac recipients has been associated with hyperacute rejection of the graft. It is therefore important that these antibodies are detected and identified as soon as the patient is registered on the transplant waiting list to ensure that incompatible donors are not considered for crossmatching. Furthermore, recent data have also suggested that the appearance of donorspecific antibodies after transplantation may be a sign of rejection, indicating the importance of post-transplant monitoring for some groups of patients. HLA and BMT

The main risk factors affecting the survival of patients undergoing BMT are graft-versus-host disease (GVHD), leukaemia relapse, and graft rejection or graft failure. The probability of developing acute GVHD is directly related to the degree of HLA incompatibility. Although BMT between HLA-identical siblings ensures matching for the classical HLA genes, acute GVHD still develops in about 20–30% of patients transplanted with an HLA-identical sibling. This is probably due to the effect of untested HLA antigens, e.g. DP, or minor histocompatibility antigens in the activation of donor T cells. However, patients receiving grafts from HLA-matched unrelated donors have a higher risk of developing GVHD than those transplanted using an HLA-identical sibling. Original studies on the role of HLA in the outcome of BMT indicated that HLA-DR incompatibility was one of the main risk factors associated with the development of GVHD. However, more recent studies have shown that mismatches at the HLA-A alleles, and to a lesser extent HLA-C alleles, are independent risk factors for death in patients following haemopoietic stem cell transplantation from an unrelated donor. Furthermore, mismatches at the HLA-C locus, as detected by DNA sequencing, has been shown to be associated with graft failure in patients receiving an unrelated BMT. 45

Chapter 4

The use of DNA-based methods for the detection of the HLA polymorphisms mentioned above has provided a unique opportunity to improve HLA matching of patients and unrelated donors and to reduce the development of GVHD. However, it has been shown that the increased GVHD seen as a result of HLA mismatch may result in a lower relapse rate, probably due to a GVL response associated with the graft-versushost response. Furthermore, the use of T-celldepleted marrow, which has successfully decreased the incidence of GVHD, also resulted in an increased incidence of leukaemia relapse. Thus it appears that mature T cells in the marrow, which may be responsible for GVHD, may also be involved in the elimination of residual leukaemic cells. More recently, cord blood has been successfully used as a source for haemopoietic stem cells for patients requiring marrow reconstitution. Preliminary clinical data have shown reduced risk and severity of GVHD following HLA-matched and HLA-mismatched cord blood transplantation. It is possible that the immunological immaturity of cord blood mononuclear cells compared with mononuclear cells present in adult bone marrow may result in a reduced GVHD effect. The impact of this on the relapse rates is not yet clear. Conversely, the rate of graft rejection is significantly higher in recipients of an HLA-mismatched transplant than in those receiving a transplant from an HLA-identical sibling (12.3% vs. 2.0%). Graft failure, which is thought to be mediated by residual recipient T and/or NK cells reacting with major or minor histocompatibility antigens present in the donor marrow cells, has also been shown to be due to antibodies reacting with donor’s HLA antigens. Thus, rejection is particularly high in HLA-alloimmunized patients. Furthermore, studies on leukaemic patients with donor-specific antibodies and with a positive T- or B-cell cytotoxic crossmatch have a high incidence of graft failure compared with those with a negative crossmatch. Preformed cytotoxic antibodies can also increase the incidence of graft failure in patients with aplastic anaemia. However, in spite of these reports, HLA antibodies are particularly relevant in the post-BMT setting, where patients 46

receiving multiple transfusions can experience immunological refractoriness to random platelet transfusions due to the presence of HLA antibodies. These patients require transfusions of HLA-matched platelets (see Chapter 9). Blood transfusion

The clinical relevance of HLA antigens and antibodies in blood transfusion has been well documented. White cells present in transfused products express antigens which, if not identical to those present in the recipient, are able to activate T cells and lead to the production of antibodies and/or effector cells responsible for some of the serious complications of blood transfusion. On the other hand, antibodies (and sometimes cells) present in the transfused product may react directly with the relevant antigens in the recipient and in this way provoke a transfusion reaction. The immunological reactions due to the passive transfer of HLA antibodies or HLA-specific effector cells include transfusion-related acute lung injury (TRALI) and transfusion-associated graft-versus-host disease (TA-GVHD). TRALI is a relatively under-reported complication of blood transfusion, characterized by acute respiratory distress, pulmonary oedema and severe hypoxia. The development of TRALI has been associated with the transfusion of blood components containing HLA and HNA antibodies reacting with the recipient white cells, causing complement activation, accumulation of neutrophils in the lungs and oedema. TRALI cases have also been associated with the presence of white cell antibodies in recipients reacting with transfused leucocytes and/or to inter-donor antigen–antibody reactions in pooled platelets. TA-GVHD is a rare but often severe and fatal reaction associated with the transfusion of cellular blood components. It occurs primarily in immunosuppressed individuals, although it can also occur in immunocompetent recipients. In this condition, immunocompetent T lymphocytes present in blood or blood products are able to recognize HLA and/or minor histocompatibility antigens present on the recipient cells and induce a GHVD reaction similar to that seen following haemopoi-

Human leucocyte antigens

etic stem cell transplantation. Diagnosis depends on finding evidence of donor-derived cells, chromosomes or DNA in the blood and/or affected tissues of the recipient. CD4+ and CD8+ cytotoxic as well as CD4+ T-cell clones lacking direct cytotoxicity but with supernatants containing TNF-b have been isolated from the lesion of patients with TA-GVHD. Immunological reactions due to antibodies present in the recipient include non-haemolytic febrile transfusion reaction (NHFTR) and immunological refractoriness to random platelet transfusions. The occurrence of NHFTR has been commonly associated with the presence of HLA antibodies in the recipient reacting with white blood cells present in the transfused product. However more recently, and particularly in countries that have introduced universal leucodepletion, it has been observed that NHFTR can also be triggered by the direct action of cytokines such as IL-1b, TNF-a, IL-6 and/or by chemokines such as IL-8 which are found in transfused products. On the other hand, immunological refractoriness to random platelet transfusions is primarily due to HLA and, to a lesser extent, HPA and hightitre ABO alloantibodies present in the patient and destroying the transfused platelets. This results in the lack of platelet increments following the transfusion of random donor platelets. Although HLA antibodies are found in approximately 20–50% of multitransfused patients, only approximately 10–20% of them become immunologically refractory and require HLA-matched or HLAcompatible platelet transfusions. Following the introduction of universal leucodepletion, the proportion of multitransfused patients with HLA antibodies seems to have decreased to approximately 10–20% and only about 3–5% of these patients are immunologically refractory. These are previously sensitized transplanted or transfused recipients including multiparous women.

HLA and disease HLA genes are known to be associated with a variety of autoimmune, non-autoimmune and, more recently, infectious diseases. A number of

different mechanisms to explain this association have been postulated, including LD with the relevant disease susceptibility gene, the preferential presentation of the pathogenic peptide by certain HLA molecules, and molecular mimicry between certain pathogenic peptides and host-derived peptides. Among those shown to be due to linkage with the relevant HLA-related genes is hereditary haemochromatosis (HH) and among those diseases in which HLA antigens seem to be involved in the preferential presentation of peptides to antigenic T cells is neonatal alloimmune thrombocytopenia. Diseases in which the molecular mimicry mechanism has been postulated include ankylosing spondylitis and Klebsiella infection. However, the precise pathogenic mechanisms involved remain unknown. A number of diseases associated with both HLA class I and II have been described (Table 4.5).

Table 4.5 HLA-associated diseases.

HLA class I genes Birdshot chorioretinopathy: HLA-A29 Behçet’s disease: HLA-B51 Ankylosing spondylitis: HLA-B27 Psoriasis: HLA-Cw6 Malaria: HLA-B53 HLA class II genes Rheumatoid arthritis HLA-DRB1*0401 HLA-DRB1*0404 HLA-DRB1*0405 HLA-DRB1*0408 HLA-DRB1*0101/0102 HLA-DRB1*1402 HLA-DRB1*1001 Narcolepsy: HLA-DQB1*0602/DQA1*0102 Coeliac disease: HLA-DQB1*0201/DQA1*0501 Neonatal alloimmune thrombocytopenia: HLA-DRB3*0101 Malaria: HLA-DRB1*1302/DQB1*0501 Insulin-dependent diabetes mellitus: HLA-DQB1*0302/DQA1*0301 HLA-linked diseases Haemochromatosis: (HLA-A3) HFE gene C282Y, H63D and S65C 21-OH deficiency: (HLA-B47) 21-OH gene

47

Chapter 4 H63

Hereditary haemochromatosis

HH is a common genetic disorder in northern Europe, where between 1 in 200 and 1 in 400 individuals suffer from the disease, with an estimated carrier frequency of between 1 in 8 and 1 in 10. Clinical manifestations include cirrhosis of the liver, diabetes and cardiomyopathy. Detection of asymptomatic iron overload is important since removal of excess iron by phlebotomy can prevent organ damage. Previous screening methods relied on the measurement of iron saturation, confirmed with a fasting sample. Confirmation of HH was by liver biopsy. Hence a non-invasive screening method would be an advantage. Previously, a close association between HLA-A3 and HH had been described and, until recently, HLA-A3 was the only test available to aid diagnosis, although this was not very specific since the majority of HLAA3-positive individuals do not have HH. A number of mutations have been identified in the HFE gene, which is located 3 Mb telomeric of the HLA region. Clinical data indicate that at least three of these mutations (C282Y, H63D and S65C) may predispose and affect the clinical outcome of this condition. Over 90% of HH patients in the UK are homozygous for the mutation that replaces a cysteine (C) with a tyrosine (Y) at codon 282 in the HFE gene. The second and third mutations (H63D and S65C) are thought to be less important, although it may have an additive effect if inherited with the first mutation (Fig. 4.5). Recent studies on blood donors have shown that approximately 1 in 280 donors are homozygous for the mutations. A DNA-based technique to detect these three mutations simultaneously has now been developed in our laboratories and provides a simple, rapid and unambiguous definition of these mutations. Neonatal alloimmune thrombocytopenia

Neonatal alloimmune thrombocytopenia is due to fetomaternal incompatibility for human platelet antigens (see also Chapters 5 and 8). More than 80% of cases occur in women who are homozygous for the HPA-1b allele. Although the majority of cases are associated with the presence of HPA48

a1

D

a2 NH2

NH2

b2m

a3

C282

COOH

Y

COOH Fig. 4.5 HFE molecule. b2-m, b2-microglobulin.

1a antibodies, about 15% of cases are due to anti-HPA-5b. Early studies indicated that the production of HPA-1a antibodies is strongly associated with the HLA-DRB3*0101 allele. However, only approximately 35% of HPA-1a-negative, DRB3*0101positive women develop antibodies upon exposure to the antigen, suggesting that other genes or factors may be involved in the development of alloimmunization against HPA-1a. It has been shown that the amino acid substitution of leucine for proline at position 33 on the GPIIIa chain is responsible for the production of alloantibodies and, more recently, T cells responding to a peptide containing this residue have been identified in an HPA-1b/b patient with an affected child.

Further reading Brown C, Navarrete C. HLA antibody screening by LCT, LIFT and ELISA. In: Bidwell J, Navarrete C, eds. Histocompatibility Testing. London: Imperial College Press, 2000: 65–98. Campbell RD. The human major histocompatibility complex: a 4000-kb segment of the human genome replete with genes. In: Davies KE, Tilghman SM, eds.

Human leucocyte antigens Genome Analysis, Vol. 5. Regional Physical Mapping. New York: Cold Spring Harbor Laboratory Press, 1993: 1–33. Dyer PA, Claas FHJ. A future for HLA matching in clinical transplantation. Eur J Immunogenet 1997; 24: 17–28. Gruen JR, Weissman SM. Evolving views of the major histocompatibility complex. Blood 1997; 90: 4252–65. Harrison J, Navarrete C. Selection of platelet donors and provision of HLA matched platelets. In: Bidwell J, Navarrete C, eds. Histocompatibility Testing. London: Imperial College Press, 2000: 379–90. Howell M, Navarrete C. The HLA system: an update and relevance to patient–donor matching strategies in clinical transplantation. Vox Sang 1996; 71: 6–12. Lardy NM, Van Der Hjorst AR, Ten Berge IJM et al. Influence of HLA-DRB1* incompatibility on the occurrence of rejection episodes and graft survival in serologically HLA-DR-matched renal transplant combinations. Transplantation 1997; 64: 612–16. Madrigal JA, Arguello R, Scott I, Avakian H. Molecular histocompatibility typing in unrelated donor bone marrow transplantation. Blood Rev 1997; 11: 105–17.

Madrigal JA, Scott I, Arguello R, Szydlo R, Little A-M, Goldman JM. Factors influencing the outcome of bone marrow transplants using unrelated donors. Immunol Rev 1997; 157: 153–66. Marsh SG, Albert ED, Bodmer WF et al. Nomenclature for factors of the HLA system. Tissue Antigens 2002; 60: 407–64. Murac, Raguenes O, Ferec C. HFE mutations analysis in 711 haemochromatosis probands: evidence for S65C implication in mild form of hemochromatosis. Blood 1999; 93: 2502–5. Parham P, McQueen KL. Alloreactive killer cells: hindrance and help for haematopoietic transplants. Nat Rev Immunol 2003; 3: 108–21. Suthanthiran M, Strom TB. Renal transplantation. N Engl Med J 1994; 331: 365–76. Terasaki PL, McClelland JD. Microdroplet assay of human serum cytokines. Nature 2000; 204: 998–1000. Thorsby E. HLA associated diseases. Hum Immunol 1997; 53: 1–11.

49

Chapter 5

Platelet and neutrophil antigens David L. Allen, Geoffrey F. Lucas, Willem H. Ouwehand and Michael F. Murphy

Human platelet antigens As on red cells, antigens on human platelets can be categorized according to their biochemical nature. • Carbohydrate antigens on glycolipids and glycoproteins (GPs): A, B and O antigens, P, Le. • Protein antigens: human leucocyte antigen (HLA) class I A, B and C, GPIIb/IIIa, GPIb/IX/V, etc. • Haptens: quinine, quinidine; heparin; some antibiotics, e.g. penicillins and cephalosporins. These antigens can be targeted by some or all of the following types of antibodies: • autoantibodies; • alloantibodies; • isoantibodies; and • drug-dependent antibodies. Many platelet antigens are shared with other cells, e.g. ABO and HLA class I (Table 5.1). This section focuses on protein alloantigens expressed predominantly on platelets, although some of these are also present to a lesser extent on some other blood cells, e.g. human platelet antigen (HPA)-5 on activated T lymphocytes. These antigens are commonly referred to as platelet-specific alloantigens or human platelet alloantigens. Human platelet alloantigens

There are a number of well-characterized biallelic platelet alloantigen systems, and a number of rare, private or low-frequency antigens have also been described (Table 5.2). Most of these antigens were first discovered during the investigation of cases of neonatal alloimmune thrombocytopenia (NAIT). Platelet-specific alloantigens are located on platelet membrane GPs involved in haemostasis 50

through interactions with extracellular matrix proteins in the vascular endothelium and with plasma coagulation proteins. The majority of these antigens are on the GPIIb/IIIa complex, which plays a central role in platelet aggregation as a receptor for fibrinogen, fibronectin, vitronectin and von Willebrand factor. Other important GPs are GPIb/IX/V, the main receptor for von Willebrand factor, which is involved in platelet adhesion to damaged vascular endothelium; GPIa/IIa, which is involved in adhesion to collagen; and CD109, which also appears to be a collagen receptor. Congenital deficiency of these GPs results in bleeding disorders, e.g. lack of GPIIb/IIIa causes Glanzmann’s thrombasthenia and absence of GPIb/IX/V results in Bernard–Soulier syndrome. The expression of platelet alloantigens located on these GPs may be altered in these disorders and HPA typing performed by serological assays (‘phenotyping’) may give discrepant results when compared with results obtained by molecular typing (‘genotyping’). Inheritance and nomenclature

Most of the platelet alloantigen systems reported to date have been shown to be biallelic, with each allele being codominant. Historically, systems were named by the authors first reporting the system, usually using an abbreviation of the name of the patient in whom the antibody was detected. Some systems were published simultaneously by different laboratories and with different names, e.g. Zw and PlA or Zav/Br/Hc, and only later were they found to be the same polymorphism. In 1990, a working party for platelet immunology of the International Society of Blood Transfusion (ISBT)

Platelet and neutrophil antigens Table 5.1 Antigen expression on peripheral blood cells.

Antigens

Erythrocytes

Platelets

Neutrophils

B lymphocytes

T lymphocytes

Monocytes

A, B, H I Rh* K HLA class I HLA class II GPIIb/IIIa GPIa/IIa GPIb/IX/V CD109

+++ +++ +++ +++ -/(+) -

(+)/++ ++ +++ +++ +++ +++ (+)/++§

++ +++ -/+++§ (+)‡ -

+++ +++ -

+++ -/+++§ ++§ -/++§

+++ +++ +++D

* Non-glycosylated. § On activated cells. ‡ GPIIIa (b3) in association with an alternative a chain (av). D Dependent on monoclonal antibody used to assess expression.

agreed a new nomenclature for platelet polymorphisms, the HPA nomenclature. The recently founded international Platelet Nomenclature Committee has published guidelines for acceptance and naming of newly discovered platelet GP alloantigens. In the HPA nomenclature, each system is numbered consecutively (HPA-1, -2, -3 and so on) (see Table 5.2) according to its date of discovery, with the high-frequency allele in each system being designated ‘a’ and the low-frequency allele ‘b’. Newly discovered systems are only officially included when confirmed by a second party and approved by the nomenclature committee. If an antibody against only one allele has been reported, a ‘w’ (for workshop) is added after the antigen name, e.g. HPA-10bw. One possible reason why antibodies against the ‘a’ antigen have not yet been reported for many of the recently discovered systems is that the ‘b’ allele is of such low frequency that ‘bb’ homozygous individuals either do not exist or are extremely rare. In Caucasian populations, the allele frequency for the majority of HPA systems is skewed towards the ‘a’ allele and homozygosity for the ‘b’ allele is below 3%. This places significant pressures on the blood services in the management of alloimmunized patients since compatible red cells and platelets are difficult to obtain for patients with antibodies against the ‘a’ alloantigen. Allele fre-

quencies vary between populations, e.g. GPIIIaproline-33 (HPA-1b) is extremely rare or absent in the Far East, while GPIIIa-glutamine-143 (HPA4b) does not occur in Caucasians. These differences are important when investigating cases of suspected platelet alloimmunity in different ethnic groups. Until the early 1990s, platelet typing was performed by serological assays. These assays required the use of monospecific antisera, which were relatively uncommon as the majority of immunized individuals produced HLA class I antibodies in addition to the platelet-specific antibodies. The typing that could be performed therefore was limited, and many laboratories were only able to phenotype for HPA-1a. The publication of more advanced assays, such as monoclonal antibody-specific immobilization of platelet antigens (MAIPA) (Fig. 5.1) permitted more extensive phenotyping but some antisera were simply not available. With the advent of techniques such as immunoprecipitation of radioactively labelled platelet membrane proteins, MAIPA and the polymerase chain reaction (PCR), the molecular basis of the majority of clinically relevant platelet-specific alloantigen systems was elucidated (Fig. 5.1 and see Table 5.2). The molecular basis for all the HPA alloantigen systems has been determined and in all but one (HPA-14bw) the difference between the 51

Chapter 5 Table 5.2 Platelet-specific alloantigen systems.

System

Antigen

Alternative names

HPA-1

HPA-1a HPA-1b

Zwa, P1A1 Zwb, P1A2

HPA-2

HPA-2a HPA-2b

Kob Koa, Siba

HPA-3

HPA-3a HPA-3b

Baka, Leka Bakb

HPA-4

HPA-4a HPA-4b

Yukb, Pena Yuka, Penb

HPA-5

HPA-5a HPA-5b

Brb, Zavb Bra, Zava, Hca

HPA-6

HPA-6bw

HPA-7

Glycoprotein (GP)

Nucleotide change

Amino acid change

97.9 28.8

GPIIIa

T196 C196

Leucine33 Proline33

>99.9 13.2

GPIba

C524 T524

Threonine145 Methionine145

GPIIb

T2622 G2622

Isoleucine843 Serine843

>99.9 20% Maintain haemostasis by: Platelet count > 50 ¥ 109/L INR and APTT ratios < 1.5 Fibrinogen > 1.0 g/L Avoid metabolic disturbances Hypocalcaemia Hyperkalaemia Acid–base disturbances Hypothermia Treat the cause of the blood loss APTT, activated partial thromboplastin time; INR, international normalized ratio; PCV, packed cell volume.

of whole blood from ‘walk-in’ donors is now considered an unacceptable and dangerous practice due to the risks of transfusion-transmitted disease (see also Chapter 6). Practical management

• Obtain maximum venous access. One or two large-bore intravenous cannulae should be inserted and if possible a central line. • Should crystalloid or colloid be used as initial replacement therapy? This thorny argument has continued for many years and has been fuelled by the recent concerns about the use of albumin. If crystalloids are used, then larger volumes are required. • Prevent tissue hypoxia by giving adequate volume to achieve an acceptable systolic blood pressure. Initially this can be done using crystalloid or colloid until red cell transfusion is available. • Blood is required. Send a blood sample to the blood transfusion laboratory for ABO group and RhD group (can usually be performed within 5 min) and telephone the transfusion laboratory to indicate the need for blood. If possible, wait for ABO- and RhD-compatible blood. In emergency cases, use group O RhD-negative red cells until the patient’s ABO and RhD groups are known. Switch to blood of the same ABO and RhD groups as the patient as soon as possible to avoid inappropriate use of group O RhD-negative red cells as their supply may be limited. • Also send a baseline ethylenediamine tetraacetic acid (EDTA) sample for full blood count (FBC), and a citrate sample for coagulation screen and sample to biochemistry for urea and electrolytes. • Administration of blood: when a very fast rate of transfusion is required (>50 mL/kg per h in adults or >15 mL/kg per h in children), a blood warmer should be used. • Once the patient’s blood pressure has been restored, consider surgical control of the bleeding. • Haemostasis (see also Chapter 11). An early coagulation screen and platelet count or TEG will provide a guide to the use of blood components. It is important to understand that at least 1.5 blood 89

Chapter 7

volumes (i.e. 7–8 L in adults) must be transfused before the platelet count falls below 50 ¥ 109/L in an average healthy individual. There is often time to assess coagulation fully; transfusion of blood components should be given as necessary according to the results of the screening coagulation tests. • Aim to (i) keep the platelet count greater than 50 ¥ 109/L by administering platelet concentrates; and (ii) maintain PT and APTT ratios less than 1.5 times the control value by giving fresh frozen plasma (FFP). Fibrinogen in the form of cryoprecipitate can be given if fibrinogen levels are disproportionately low, in order to maintain fibrinogen concentrations greater than 1.0 g/L. Coagulation problems (see also Chapter 11)

These occur in patients with extensive bleeding because of: • loss of haemostatic factors; • consumption in clot formation; • dilution with blood components and volume expanders; • depletion of coagulation factors due to inadequate synthesis, although factor VIII deficiency is partially compensated for by increased synthesis and release as part of the stress response; • acidosis and hypothermia precipitate disseminated intravascular coagulation (DIC). DIC may occur but cannot be predicted. Massively transfused patients do not form a homogeneous group; delayed or inadequate treatment of shock is probably the common predisposing factor, while extensive tissue damage, particularly head injuries, and pre-existing hepatic and renal failure may contribute to a deterioration in haemostasis. Volume expanders may produce other haemostatic hazards, apart from dilution. Dextrans, and to a lesser extent hydroxylethyl starch, have a fibrinoplastic effect; they accelerate the action of thrombin in converting fibrinogen to fibrin, which makes clots more amenable to fibrinolysis. Both are adsorbed on platelet surfaces and on von Willebrand factor (vWF), causing decreased platelet function and an acquired von Willebrand syndrome. Gelatins produce few problems, al90

though they decrease plasma fibronectin activity, but this has little clinical significance. Consider the use of adjunct haemostatic agents. Persistent bleeding will stimulate fibrinolytic activity. Ideally D-dimers or a TEG trace could be used in this situation as a guide to fibrinolytic activity. The use of an antifibrinolytic agent such as aprotinin 500 000 kallikrein inhibitory units (KIU) intravenously or tranexamic acid 1 g intravenously may be beneficial. Other possible complications of blood transfusion

• Hypocalcaemia. Calcium gluconate (2 mL of 10% solution per unit of blood) should be given when ionized calcium or calcium concentration can be measured and shown to be low or there are clinical signs or electrocardiographic changes. • Hyperkalaemia may occur due to the high concentration (~ 40 mmol/L) in stored blood. This is usually only a problem in those with hepatic or renal disease. • Acid–base disturbances. Despite the presence of lactic acid in transfused blood, fluid resuscitation usually improves acidosis in shocked patients. Furthermore, transfused citrate produces an alkalosis once it is metabolized. • Avoid hypothermia. Warm the patient, and the blood if a fast rate of transfusion is required. Cardiopulmonary bypass

Cardiopulmonary bypass (CPB) has evolved into a reliable method for total body perfusion, maintaining an oxygenated blood supply during the time heart surgery is performed. This is normally achieved by draining venous blood under gravity from the right atrium into a reservoir, and then pumping blood through an oxygenation device back into the patient’s arterial system (Fig. 7.3). This procedure bypasses the heart and lungs and creates a bloodless surgical field. Haemorrhage is one of the most important complications of cardiac surgery since re-exploration for bleeding, which occurs after 2–6% of coronary artery bypass grafting procedures, has been associated with a case fatality rate as high as 22%. This is especially relevant now there are an increasing

Bleeding in trauma and surgery

Direction of blood flow Venous return from patient under gravity

Cannula to right atrium

Cardiotomy suction lines for return of blood from the open chest Venous reservoir and cardiotomy filters

Arterial line pressure gauge Oxygenator Cannula to aorta Heat exchanger

Fig. 7.3 Cardiopulmonary bypass

Centrifugal pump

Blood returned to patient's systemic circulation bypassing the lungs

circuit.

number of patients requiring reoperation and of patients who have received anticoagulant, antiplatelet or thrombolytic therapy prior to surgery or who undergo complex surgery such as combined heart and lung transplantation. Factors associated with the bleeding diathesis of CPB

• The extensive contact between blood and the synthetic surfaces of the circuit causes coagulation activation, which necessitates the use of heparin. Intravenous heparin is administered to the patient prior to CPB at a dose of 3 mg/kg, with repeated doses of heparin being given during CPB (approximately equivalent to heparin levels of 3 U/mL). During CPB the activated clotting time (ACT) is used to monitor anticoagulant therapy and is maintained above 350–400 s. • Thrombocytopenia and defects in platelet function are proportional to the duration of CPB, and probably related to the level of hypothermia. In addition, abnormalities of platelet function include a reduced response to aggregation stimuli

owing to discharge of a granules and loss of platelet membrane receptors such as glycoprotein (GP)Ib and GPIIb/IIIa. • Fibrinolytic activation measured by D-dimer levels has been shown to peak during CPB and decrease at the end of CPB. There is wide variation in fibrinolytic response to CPB. The increased fibrinolytic activation is mainly due to an increase in t-PA. • Haemodilution is also a consequence of extracorporeal circulation but the fall in haematocrit, platelet count and plasma proteins, including coagulation factors, is about 30% and usually not sufficient to cause bleeding. There is increasing use of minimally invasive or ‘beating’ heart surgery, which avoids the use of CPB. For example, using a piece of equipment known as an ‘Octopus’, which once placed on the heart will reduce movement in the area in which the surgeon wishes to operate, allows for surgery to proceed without stopping the heart beating. The haemostatic changes of these procedures are less severe than those associated with CPB. 91

Chapter 7

Practical management of bleeding after cardiac surgery

• Check blood loss from all the chest drains. If one is filling at a greater rate than others, then this suggests a surgical cause, so discuss with the surgeon. • Obtain a history of preoperative drugs. The use of aspirin and non-steroidal anti-inflammatory drugs is associated with increased blood loss. • If there is a steady blood loss in all the chest drains of more than 300 mL/h, arrange for further blood to be made available by calling the blood transfusion laboratory. An FBC and clotting screen could be requested. While awaiting the results of these laboratory assays, an ACT may be performed to check that heparin has been adequately reversed with protamine. If heparin has not been adequately reversed, then give the appropriate dose of protamine. • Usually after cardiac surgery there is a thrombocytopenia and platelet function defect, so the bleeding time will be prolonged and therefore of little value in differentiating the cause of bleeding. A TEG trace, if available, may help. • Discuss rationale for the use of platelet concentrates and if necessary order an adult therapeutic dose of platelets. While waiting for this to arrive, if the patient has not received an antifibrinolytic agent perioperatively consider using aprotinin 500 000 KIU. • If bleeding continues at a rate of more than 300 mL/h despite correction of any pre-existing haemostatic defects and adequate haemostatic therapy, i.e. platelet transfusion given and platelet count greater than 100 ¥ 109/L and PT and APTT ratios less than 1.5, with a fibrinogen level in the normal range (normally 2–4 g/L), then discuss reexploration of the wound with the surgeon. Hyperfibrinolysis

Bleeding may occur if there is excessive generation of plasmin secondary to the release of tissue and urokinase plasminogen activators. Plasmin is a non-specific proteolytic enzyme and will split peptides with arginyl-lysyl amino acid sequences. These include fibrinogen, factors V and VIII, and the first component of complement. 92

Some regions, especially prostatic and pelvic tissues, are rich in plasminogen activator, excessive liberation of which may occur during the following. • Pelvic and prostatic surgery, especially for carcinoma of the prostate. • Extensive surgery. • CPB. • Liver transplantation. Extremely high levels of tPA occur during the anhepatic phase of orthotopic liver transplantation, probably as a result of increased endothelial release and decreased hepatic clearance. • Iatrogenic fibrinolytic bleeding can occur through the use of exogenous fibrinolytic activators such as streptokinase or urokinase in the management of thrombosis. Useful investigations include the following. • A global test of fibrinolytic activity such as the TEG should ideally be available. • Levels of fibrin degradation products (D-dimers) are greatly increased. • PT, APTT and thrombin time are mildly prolonged due to fibrinogenolysis. • Factors V and VIII may be normal or moderately reduced, in contrast to findings in DIC. • Often it is difficult to exclude DIC, especially as the most useful fibrinolytic assays are complex, time-consuming and not widely available. Practical management

Treatment with an antifibrinolytic agent should be considered if primary fibrinolytic bleeding is suspected. Suggested management options are: • tranexamic acid up to 1 g slowly intravenously; and • aprotinin 500 000 KIU as an intravenous bolus.

Pharmacological agents to reduce bleeding These have been used in two ways, either to prevent excessive bleeding or to treat established bleeding. The majority have been used in patients having cardiac surgery with CPB. The agents used can be broadly classified into four groups: anti-

Bleeding in trauma and surgery

fibrinolytics, desmopressin, haemostatic sealants and recombinant factor VIIa.

Antifibrolytics Aprotinin

• High-dose aprotinin in cardiac surgery (2 000 000 KIU to the patient, 2 000 000 KIU into the CPB circuit and 50 000 KIU/h during CPB) reduces postoperative drainage loss by 81%, and total haemoglobin loss by 89%. It also has benefit in vascular surgery and liver transplantation. Shorter operating times were also seen. This may result from the striking reduction of oozing that is normally seen; the operative fields remain ‘bone dry’. • Aprotinin is a basic polypeptide extracted from bovine lung. It is able to inhibit certain serine proteases by binding to their active site. In low concentrations, aprotinin is a powerful inhibitor of plasmin: its molar potency in vitro is 100 and 1000 times that of tranexamic acid and e-aminocaproic acid. Its main mechanism of action is through an antiplasmin effect. In high doses, it also inhibits kallikrein. The currently licensed high-dose regimen was designed to achieve blood levels that inhibit kallikrein (about 200 KIU/mL). Kallikrein is formed during the activation of coagulation by CPB and has a central role in the activation of the inflammatory response. • Aprotinin has no effect on the fall in platelet count, but may have a minor effect on preserving platelet function by preserving platelet membrane receptors, possibly by inhibiting plasmin-mediated degradation. • Aprotinin, by inhibiting kallikrein, will prolong in vitro tests of the intrinsic system and the ACT, because kallikrein normally operates a positive feedback on the generation of factor XII. In order to allow for adequate levels of heparin, the ACT timer should be run greater than 750 s. The activator in the ACT has traditionally been celite. Recently, kaolin has been used instead in some ACT tubes, for it is less affected by aprotinin and thus ACTs can be monitored in the normal way. • Since aprotinin is a bovine protein and thus can provoke an immunological reaction, a test dose

should be given. Consideration should be given to the future need of this drug, e.g. if a patient requires repeat cardiac surgery but in the long term requires a cardiac transplant, the surgeon may wish to reserve the use of aprotinin for the transplant operation. • The risks of using a prothrombotic agent perioperatively have not been defined, especially those of increased postoperative thromboembolic disease and particularly in relation to graft patency after coronary artery bypass grafting. Until these risks are defined the use of aprotinin to prevent blood loss should be limited to its licensed indication, i.e. the prevention of blood loss in high-risk cardiac surgery. • Aprotinin can also be used in established fibrinolytic bleeding. An intravenous dose of 500 000 KIU is a good antiplasmin dose. Lysine analogues

The lysine analogues, e-aminocaproic acid and tranexamic acid, are competitive inhibitors of plasmin binding to fibrin. A continuous infusion of tranexamic acid perioperatively in open cardiac surgery reduces bleeding significantly by about one-third, although it is not as effective as aprotinin. The dose given is 10 mg/kg over 20 min preoperatively followed by 1 mg/kg for 10 h. Both can be given to treat established fibrinolysis. The recommended dose for tranexamic acid is up to 1 g by slow intravenous infusion. Desmopressin

Desmopressin acetate (DDAVP) is a synthetic vasopressin analogue that is relatively devoid of vasoconstrictor activity. It increases the plasma concentrations and activity of vWF, probably by inducing the release of vWF from Weibel–Palade bodies in the endothelium. Plasma levels of vWF increase from two to five times from the baseline within an hour. It also improves platelet function. It thus leads to shortening of the bleeding time in patients with von Willebrand’s disease, platelet function defects and uraemia. Trials of DDAVP 0.3 mg/kg given after CPB to reduce bleeding had variable results; overall it was 93

Chapter 7

not beneficial. It may well be useful in patients with platelet function defects preoperatively, but this is not proven as yet. However, DDAVP may yet have a place in reducing bleeding, for it may be useful in reducing blood loss in patients with functional platelet disorders, notably those patients who have received aspirin preoperatively. Adverse effects include flushing and an antidiuretic effect. Haemostatic sealants

Haemostatic sealants mimic the final part of the coagulation cascade. • Fibrin sealants provide a source of thrombin and fibrinogen that when mixed together in the presence of calcium form a clot. They can be administered by a ‘gun’ which produces mixing of the reagents. Fibrin glue is best suited to secure haemostasis in patches and suture lines, for in situations where there are high blood flow rates then it is in danger of being washed off before it has ‘set’. A systematic review has shown that they do reduce allogeneic blood use and reduce bleeding but generally the trials were small and of poor methodological quality. • The initial source of thrombin was of bovine origin, which led to the development of a bleeding diathesis postoperatively. This is due to the formation of antibodies to bovine thrombin, which cross-react with human factor V, leading to acquired factor V deficiency in the recipients. Currently human thrombin is used in most fibrin sealants. • Methods of preparing autologous fibrin glue have been developed. They are currently made by units with large transfusion centres and commercial companies also exist to manufacture components. • Other haemostatic sealants, such as Floseal (Baxter), use bovine thrombin and gelatin. These are reconstituted and mixed together. The mixture is applied to the tissue surface, the gelatin expands to physically restrict the flow of blood and then thrombin converts endogenous fibrinogen to fibrin. The structural integrity of the gelatin–fibrin matrix enables it to remain in place at the tissue surface. The clot is resorbed within 6–8 weeks, 94

consistent with the time frame of normal wound healing. Recombinant activated factor VII

Recombinant activated factor VIIa (rVIIa) was initially used to treat haemophiliacs with inhibitors. The mechanism of action of factor VIIa in this setting is not fully understood but it does bypass the need for factor VIII and IX by generating thrombin and thus fibrin via direct activation of factor X. It is unclear whether tissue factor is necessary for this. The use of rVIIa has been explored ‘off licence’ for uncontrolled bleeding in a number of clinical scenarios. A group in Israel first described it as being efficacious in trauma patients who continued to bleed despite conventional component therapy. Subsequently, it has been used to reverse over-anticoagulation with warfarin, uncontrolled bleeding in hepatic dysfunction, orthotopic liver transplantation, cardiac surgery and in patients with thrombocytopenia and platelet dysfunction. There is a paucity of double-blind randomized trials to assess efficacy and safety in these multiple potential applications, and a number of such trials are currently in progress. The major concern about safety is the theoretical risk of thrombosis: rVIIa will bind to exposed tissue factor and initiate local thrombosis. Thus in patients with atheroma, tissue factor in plaques may be exposed to blood at the time of plaque rupture, while in DIC tissue factor may be exposed on monocytes. Accumulating safety data suggest that the risk of thrombosis with rVIIa is low, but this may relate to its low use in those with atheromatous disease. A randomized double-blind trial explored the use of rVIIa in preventing perioperative bleeding in 36 patients undergoing retropubic prostatectomy randomized to receive a perioperative dose of 20 or 40 mg/kg rVIIa or placebo. Median blood loss was significantly reduced in those receiving 40 mg/kg and none of this group required transfusion. There were no adverse events in the study group. This study suggests that rVIIa requires further study to assess its efficacy and safety in preventing perioperative bleeding and reducing the use of allogeneic blood.

Bleeding in trauma and surgery

Organization of transfusion for patients with trauma and for major accidents (see also Chapter 6) In a major accident, large numbers of people may be injured within a short space of time. The rescue and management of patients requires a coordinated approach from the rescue services and the hospital designated for admission of the casualties. A ‘major accident procedure’ is a necessity within every hospital. It should be tested periodically by holding a ‘major accident exercise’. The following must be incorporated into the procedure. • The telephone numbers of those who ‘need to know’ should be held by the hospital switchboard and called. • Suspend the issue of blood for non-emergency cases. • Increase the stocks of blood components to a predefined level by arranging deliveries from the nearest transfusion centre and maintain the necessary stocks of blood and blood products throughout the emergency. The blood transfusion laboratory must have a telephone line independent of the hospital switchboard because the main hospital switchboard may be inundated with telephone calls. • The risk of errors in the identification of patients and blood samples can be high in an emergency. Special care must be taken in the identification of casualties, and in labelling blood samples. In accident and emergency every attempt to maintain good clinical practice should be made. The minimum patient identification details for the request form and blood sample are the hospital number of the patient (or accident and emergency or major incident number) and their gender. • A telephone call from accident and emergency to the blood transfusion laboratory to inform them of estimates of potential future blood requirements is essential. • To avoid errors, the practice of issuing blood in a major disaster should not be changed from the routine practice of providing blood for urgent requests. • When the recipient’s blood group is not known, group O RhD-negative blood should be given to

girls and women of reproductive age, unless there is life-threatening bleeding and group O RhDnegative blood is not available. Group O RhDpositive blood can be given to males with unknown blood groups. • Blood components such as FFP and platelet concentrates need to be available for those who are receiving massive transfusion. • Dealing with requests to donate blood: following a major accident, there may be a number of telephone calls from the public, offering to donate blood. These potential donors should be given the telephone number of the Blood Service National Call Centre so that they can attend one of the routine blood donor clinics.

Conclusions There have been some exciting advances in monitoring and managing bleeding in surgical and trauma patients in the last few years. Despite this, many more adequately powered, prospective studies are required to investigate the utility and safety of these approaches in different surgical settings.

Further reading Carless PA, Anthony DM, Henry DA. Systematic review of the use of fibrin sealant to minimize perioperative allogeneic blood transfusion. Br J Surg 2000; 89: 695–705. Carson JL, Poses RM, Spence RK, Bonavita G. Severity of anaemia and operative mortality and morbidity. Lancet 1996; 348: 1055–60. Clark RAF. Fibrin sealant in wound repair: a systematic survey of the literature. Exp Opin Invest Drugs 2000; 9: 2371–91. Cochrane Injuries Group Albumin Reviewers. Human albumin administration in critically ill patients: systematic review of randomised controlled trials. Br Med J 1998; 317: 235–40. [This article received an enormous critical response, and the reader is referred to one of the reviewers of the original article whose response is typical of the criticism: Soni N. Human albumin administration in critically ill patients. Validity of review methods must be assessed. Br Med J 1998; 317: 883–4.]

95

Chapter 7 Friederich PW, Henny CP Messelink EJ et al. The effect of recombinant activated factor VII on perioperative blood loss in patients undergoing retropubic prostatectomy: a double-blind placebo-controlled randomised trial. Lancet 2003; 361: 201–5. Ghorashian S, Hunt BJ. Off license use of recombinant activated factor VII. Blood Rev 2004; 18: 245–59. Hébert PC, Wells G, Blajchman MA et al. A multicenter, randomized, controlled clinical trial of transfusion requirements in critical care. N Engl J Med 1999; 340: 409–17. Horrow JC, Hlavecek J, Strong MD et al. Prophylactic tranexamic acid decreases bleeding after cardiac operations. J Thorac Cardiovasc Surg 1990; 99: 70–4. Martinowicz U, Kenet G, Segal E et al. Recombinant activated factor VII for adjunctive haemorrhage control in adults. J Trauma 2001; 51: 431–9.

96

Oz MC, Delos M, Cosgrove DM et al. Controlled clinical trial of a novel hemostatic agent in cardiac surgery. Ann Thorac Surg 2000; 69: 137–82. Porte RJ, Leebeek. Pharmacological strategies to decrease transfusion requirement in patients undergoing surgery. Drugs 2002; 62: 2193–211. Sanguis Study Group. Use of blood products for elective surgery in 43 European hospitals. Transfus Med 1994; 4: 251–68. Segal H, Hunt B. Aprotinin: pharmacological reduction of perioperative bleeding. Lancet 2000; 355: 1289–90. Shore-Lesserson L, Manspeizer HE, DePerio M, Francis S, Vela-Cantos F, Ergin MA. Thromboelastography-guided transfusion algorithm reduces transfusion in complex cardiac surgery. Anesth Analg 1999; 88: 312–19.

Chapter 8

Prenatal and childhood transfusions Irene Roberts

Obstetrics One of the most important aspects of obstetric transfusion medicine is the prevention, recognition and treatment of haemolytic disease of the newborn (HDN), which causes at least 50 neonatal deaths per year in the UK. This is considered in detail in this chapter. Other topics covered include: • aspects of maternal platelet and white cell disorders relevant to transfusion; • maternal haemorrhagic disorders, including major obstetric haemorrhage; and • transfusion requirements during pregnancy of patients with major haemoglobinopathies. Antenatal red cell antibody testing

The three factors essential in the pathogenesis of HDN are: • maternal red cell alloantibodies that cross the placenta; • fetal red blood cells that express antigens against which the antibodies are directed; and • antibodies which are able to mediate red cell destruction. Clinically relevant alloantibodies are almost always IgG and are reactive at 37°C. Women develop these antibodies as a result of previous transfusions, previous pregnancies or both. Identification of such antibodies is the main goal of antenatal screening.

• Identify RhD-negative women who require antiD prophylaxis (around 16% women are RhD negative). • Ensure swift provision of compatible blood for obstetric emergencies. • Identify the fetus requiring treatment in utero or in the neonatal period. • Identify additional red cell alloantibodies. • Identify new red cell antibodies induced by intrauterine transfusion. Red cell serology at the booking visit

At the booking visit, which should take place before week 16 of pregnancy, all women should have their ABO and RhD group determined and should be screened for red cell alloantibodies. If red cell antibodies are detected at the booking visit and/or if there is a history of HDN, the antibodies should be identified, quantified and monitored as outlined below. It is particularly important to monitor women with anti-D, anti-c and anti-K, since these antibodies may be associated with severe HDN. If no red cell alloantibodies are detected at booking, all pregnant women should be retested at 28–36 weeks of gestation. Further testing of women without detectable antibodies is unnecessary since immunization later in pregnancy is unlikely to result in antibody levels sufficient to cause HDN requiring treatment. Partial/weak D

Objectives of red cell antibody testing in pregnancy

• Identify the pregnancy at risk of fetal or neonatal HDN.

Du (weak D) individuals are regarded as RhD positive and do not form immune anti-D. Individuals with known partial D status, e.g. DVI, are more likely to make anti-D; therefore it is important that 97

Chapter 8

reagents for RhD grouping do not detect DIV (so that these individuals group as RhD negative). ABO antibodies

There is no need to test for ABO immune antibodies in antenatal samples as their presence is not predictive of HDN and such antibodies very rarely cause significant haemolysis in utero. Samples at delivery

At the time of delivery a maternal and a cord blood sample should be collected from all pregnancies in RhD-negative women. Women with no previously detected anti-D should have prophylactic anti-D administered if the infant is RhD positive. A Kleihauer test should also be carried out on all such women to assess the requirement for additional anti-D. A direct antiglobulin test (DAT) should be performed on the cord blood: a positive DAT is a good predictor of HDN. It is important to note that in women who have had prophylactic anti-D during pregnancy the anti-D remains detectable in serum for up to 12 weeks and may cause confusing serological results in the mother and a positive DAT in the baby in the absence of HDN. In the case of women with other clinically significant red cell alloantibodies (see below), a DAT should be carried out on cord blood; if the DAT is positive, a red cell eluate may help identify the red cell antibody. Infants born to mothers with clinically significant antibodies should be monitored for 48–72 h for the presence of haemolysis. Clinically relevant red cell alloantibodies

The main antibodies implicated in HDN include the following. • Rh group: D, c, C, e, E, Ce and Cw. • Kell group: K1, K2 and Kpa. • Duffy group: Fya. • Kidd group: Jka. The antibodies most commonly implicated in severe to moderate HDN are anti-D, anti-c and anti-Kell. Anti-D is the commonest cause of HDN (approximately 40% of cases in the UK). This is because anti-D is highly immunogenic and the high 98

proportion of women who are RhD negative (16%). Most anti-D antibodies are IgG1 or IgG1 plus IgG3. The presence of IgG3 alone, which has 100 times the destructive ability of IgG1, is uncommon and rarely associated with HDN in utero, but can cause severe postnatal manifestations of HDN. Anti-c is found most commonly in women with the R1R1 genotype (CDe/CDe), which occurs in 20% of pregnant women. Such women also have the propensity to make anti-E. HDN due to anti-E is both less common and less severe. However, anti-E and anti-c in combination cause more severe HDN than either antibody alone. Note that in such cases only the anti-E is detectable in eluates from cord blood red cells. Anti-K1 is the most common red cell alloantibody outside the ABO and Rh system. K1 is the principal antigen of the Kell blood group system and is highly immunogenic; 5% of K1-negative individuals will produce anti-K1 if transfused with K1-positive blood. K1 has around twice the potency of c and E and 20 times the potency of Fya. Anti-K1 often causes severe HDN; the haemolytic anaemia is compounded by suppression of erythropoiesis due to anti-K1 inhibiting the growth of erythroid progenitor cells. Anti-K titres can be an unreliable predictor of the severity of HDN. Therefore it is important to identify the fetuses at risk of HDN by determining the fetal Kell genotype in all mothers with anti-K1 whose partners are heterozygous for K1 (since only 50% of such fetuses will be K1 positive). Moderate to severe HDN may also be caused by anti-K2 (anti-cellano) and anti-Kpa. A number of other red cell alloantibodies have also been reported to cause HDN of variable severity, e.g. anti-U. These initially present with a positive indirect antiglobulin test (IAT) in maternal serum; therefore all women with a positive IAT should have further investigation to try to identify any clinically relevant red cell alloantibodies. Red cell alloantibodies not implicated in HDN

These include the following: • anti-Lea and anti-Leb; • anti-Lua; • anti-P; • anti-N;

Prenatal and childhood transfusions

• anti-Xga; • anti-Gerbish. Management of pregnant women with red cell alloantibodies Anti-D

• Women with anti-D should have their anti-D titres monitored monthly until 28 weeks of gestation and then every 2 weeks. • All samples should be checked in parallel with the previous sample. • An increase in anti-D by 50% or more compared with the previous sample is a significant increase irrespective of gestation. It is important to note that these are guidelines. Titres of anti-D do not always correlate closely with the development of HDN. Therefore all women with affected pregnancies should be referred early to specialist fetal medicine units for fetal assessment by ultrasound, amniocentesis or fetal blood sampling as indicated (for management of the fetus and neonate, see pp. 107–9). Anti-c

• Women with anti-c should have their anti-c titres monitored monthly until 28 weeks of gestation and then every 2 weeks. • All samples should be checked in parallel with the previous sample. • An increase in anti-c by 50% or more compared with the previous sample is a significant increase irrespective of gestation. • Anti-c titres of greater than 10 IU/mL are associated with a moderate risk of HDN and may require intrauterine transfusion (IUT). • All women with anti-c should be referred to a specialist fetal medicine unit early in pregnancy. Anti-Kell

• Women with anti-K1 should have their anti-K titre monitored monthly until 28 weeks of gestation and then every 2 weeks. • Anti-K titres may not accurately reflect the degree of fetal anaemia.

• Fetal K typing by chorionic villous sampling (CVS), amniocentesis or fetal blood sampling should be performed where the father is heterozygous for K1. • Fetal growth, fetal anaemia and the presence of hydrops should be monitored by serial ultrasound and Doppler and anaemia confirmed by fetal blood sampling as indicated. • Amniocentesis is not a good indicator of the severity of fetal anaemia since anaemia due to antiK results from a combination of haemolysis and red cell hypoplasia. • All women with anti-K should be referred to a specialist fetal medicine unit early in pregnancy. Other red cell alloantibodies

• If the antibody is likely to cause problems with provision of blood to cover an obstetric emergency, it is important to inform the obstetrician in charge of the case and the transfusion laboratory in the hospital and efforts should be made to ensure that appropriate blood products can be supplied. • Any babies born to mothers with an IATreacting antibody must be assessed at birth for evidence of HDN. Blood transfusion support for mother and baby Mother

• Red cell components of the same ABO and RhD group must be selected. • Group O blood may be used, provided it is plasma depleted and does not contain high-titre agglutinins. • Note that in pregnancy, immunization following a transfusion is most likely to occur in the third trimester. • Samples used for pretransfusion testing should ideally be taken immediately before transfusion and must never be more than 7–10 days old (Table 8.1). Fetus and neonate: crossmatching and general considerations

Management of the fetus and neonate at risk of 99

Chapter 8 Table 8.1 Pretransfusion testing of maternal samples.

Timing of last transfusion

Timing of pretransfusion sample

3–14 days before 14–28 days before 28 days to 3 months

24 h before transfusion 72 h before transfusion 1 week before transfusion

HDN is discussed in detail below. The general principles of the blood products used are summarized here. 1 Prior to the first transfusion, samples should be obtained from the mother for ABO, RhD grouping and antibody screening and from the fetus/neonate for ABO, RhD and DAT (plus an antibody screen if no maternal sample is available). 2 In the fetus/neonate the ABO group is determined on the cells only (as reverse grouping can detect passive maternal antibodies). 3 Red cells which are ABO compatible with maternal and neonatal plasma, RhD negative (or RhD identical with neonate) should be used. (NB If exchange or ‘top-up’ transfusion is required for HDN due to ABO incompatibility, group O red cells with low titre anti-A and B or group O red cells suspended in AB plasma should be used.) 4 Group O blood is acceptable; units with hightitre anti-A/anti-B must be excluded. 5 If the mother’s blood group is unknown, blood for the fetus/neonate should be crossmatched against the baby’s serum. 6 If no atypical antibodies are present in the maternal (or infant) sample, and if the DAT of the infant is negative, crossmatching is not necessary for the first 4 months of postnatal life. 7 If the antibody screen or DAT is positive, full serological investigation and compatibility testing are necessary. 8 An electronic crossmatch is not advisable unless an appropriate algorithm has been created, as ABO-identical adult blood transfused to an infant with maternal anti-A or anti-B may haemolyse even if the pretransfusion DAT is negative, due to stronger ABO antigen expression on adult cells. 9 Red cells (and platelets if given) should be cytomegalovirus (CMV) negative and leucocyte depleted. 100

10 Note that alloantibody formation is rare in the fetus and neonate and is usually associated with massive transfusion or with the use of fresh or whole blood. 11 Gamma-irradiation of cellular blood components to reduce the risk of transfusion-associated graft-versus-host disease (TA-GVHD) is recommended for: (a) all IUT; (b) all transfusions to neonates previously transfused in utero; (c) exchange transfusions as long as gammairradiation would not result in a delay in transfusion; (d) all transfusions from a family member; (e) all neonates with known inherited immune deficiencies (e.g. severe combined immunodeficiency). These precautions are due to the immaturity of the fetal and neonatal immune system which may lead to a reduced ability to reject transfused allogeneic lymphocytes, immune tolerance and the persistence of donor lymphocytes for up to 6–8 weeks after exchange transfusion. HDN: guidelines for prevention

The introduction of anti-D prophylaxis for recently delivered RhD-negative women in the UK in 1969 led to a reduction in new immunizations against anti-D from 17% of pregnancies to 1.5%. Every year in the UK 80 000 RhD-negative women have a RhD-positive infant and despite national guidelines sensitization still occurs, largely due to non-compliance with the guidelines. A dose of anti-D of 125 IU (25 mg) suppresses immunization by 1 mL of RhD-positive red cells (i.e. 2 mL of whole blood). (Note that in the UK the dose of anti-D is given in IU, whereas in other countries it is expressed in milligrams.) While anti-D is extremely effective as prophylaxis, it cannot reverse immunization once it has occurred and has no effect on the development of non-D antibodies. Indications for anti-D immunoglobulin (Table 8.2)

Anti-D should be given to all RhD-negative

Prenatal and childhood transfusions

women without anti-D antibodies after the following sensitizing events: • abortion (see below); • CVS; • ectopic pregnancy; • amniocentesis; • external cephalic version; • abdominal trauma; • antepartum haemorrhage; • premature labour; • pre-eclampsia; and • intrauterine death (associated with chronic fetomaternal haemorrhage). Anti-D should be administered following all abortions after 12 weeks, both spontaneous and induced, and following abortion at any gestation following surgical or medical treatment, including the use of abortifacients. Anti-D should also be administered in cases of threatened abortion if there is any bleeding after 12 weeks of gestation. Current UK guidelines also recommend the administration of anti-D immunoglobulin as antenatal prophylaxis since fetomaternal bleeding can happen at any gestation. Dose and schedule of administration of anti-D during the antenatal period

Therapeutic anti-D immunoglobulin to prevent the development of immune anti-D after sensitizing events should be given within 72 h of the sensitizing event; however, anti-D may still be

Table 8.2 Antenatal and postnatal prophylaxis with anti-D.

Indications for anti-D

Dose and schedule of administration

Sensitizing event 4 mL occur in 0.8% and of >15 mL in 0.3% of deliveries) so that an additional dose of anti-D (125 IU/mL blood loss) can be given. • The standard dose of anti-D in the USA and some European countries is higher (1500 IU). It takes 48 h following an intramuscular dose of anti-D to reach a good level and 72 h for clearance of sensitized red cells. If clearance of the RhDpositive cells is not complete, further anti-D must be given until RhD-positive cells can no longer be detected in maternal blood. Kleihauer test

This simple and inexpensive test is used to detect whether there has been a fetomaternal haemorrhage and the size of that haemorrhage. The principle and method for the test is as follows. 101

Chapter 8

• Fetal haemoglobin (HbF)-containing fetal red cells resist acid elution and therefore stain dark pink in comparison with HbA-containing cells, which appear as unstained ‘ghost’ cells (Plate 8.1, shown in colour between pp. 304 and 305). • To quantitate fetomaternal haemorrhage, the numbers of pink-staining HbF-containing cells in each single low-power field are counted; using this method a count of 200 HbF-containing cells or less in 50 low-power fields is equivalent to a fetomaternal haemorrhage of 4 mL or less. • Samples of maternal blood for the Kleihauer test must be taken within 2 h of administration of antiD to avoid a falsely low estimate of the size of the fetomaternal bleed. • Maternal hereditary persistence of fetal haemoglobin may cause a false-positive Kleihauer test due to maternal HbF-containing red cells.

Anti-D immunoglobulin not indicated

Anti-D immunoglobulin is not indicated in the following circumstances: • patients who are already sensitized; • those classified as weak D (e.g. Du); • if the infant is RhD negative; • for women not capable of child-bearing (following transfusion of RhD-positive blood); • for complete abortions before 12 weeks of gestation if there has been no surgical treatment. Preparation of anti-D immunoglobulin

Anti-D is a polyclonal antibody prepared by plasmapheresis of hyperimmunized donors, 95% of whom are women who have been sensitized during pregnancy. Anti-D is now prepared using US donor plasma because of concerns about transmission of variant Creutzfeldt–Jakob disease.

Large fetomaternal bleeds

Larger bleeds (>4 mL) may be measured using flow cytometry as well as the Kleihauer test. Larger fetomaternal bleeds are associated with: • amniocentesis; • abdominal trauma; • antepartum haemorrhage; • stillbirth; • twin pregnancy; • manual removal of the placenta. For any fetomaternal haemorrhage >4 mL an appropriate supplementary dose of anti-D must be given immediately AND a repeat test for fetal cells and free anti-D should be carried out on the mother 48 hours after the initial anti-D injection. A further appropriate dose of IgG anti-D should be given to the mother: • if fetal cells are no longer present but there is no residual free anti-D detectable (to make sure there is sufficient anti-D to eliminate small numbers of fetal cells below the limits of detection); • if fetal cells are still present but there is no detectable anti-D. Note that if fetal cells are still present after 48 hours but there is still detectable anti-D, a repeat test for fetal cells and free anti-D should be carried out after a further 48 hours to determine whether more anti-D IgG should be given to the mother. 102

Platelet and white cell disorders in pregnant women Differential diagnosis of thrombocytopenia in pregnancy

The most common causes of maternal thrombocytopenia are the following. • Gestational. • Pregnancy induced: pre-eclampsia, eclampsia, HELLP (haemolysis with elevation of liver enzymes and low platelets) syndrome. • Immune: immune thrombocytopenia (ITP) and systemic lupus erythematosus (SLE). • Virus associated, e.g. human immunodeficiency virus (HIV). Less common causes of maternal thrombocytopenia include: • antiphospholipid syndrome; • thrombotic thrombocytopenic purpura; • disseminated intravascular coagulation (DIC); • type IIb von Willebrand’s disease; • congenital bone marrow failure (e.g. Fanconi’s anaemia); • heparin-induced thrombocytopenia; • folate/B12 deficiency; • myelodysplasia/acute leukaemia. It may be difficult to distinguish between gestational, pregnancy-induced and immune thrombo-

Prenatal and childhood transfusions

cytopenia in pregnancy. ITP is more likely if the platelet count was subnormal prior to or during the first trimester of pregnancy. Further investigation depends on careful evaluation of the blood film and marrow smear, which may reveal characteristic changes (e.g. acute leukaemia). The disorders of particular relevance to transfusion medicine are ITP, HELLP and type IIb von Willebrand’s disease.

moderate not severe. Fulminant pre-eclampsia precipitating early delivery may be associated with DIC and require treatment with platelet transfusion and fresh frozen plasma (FFP) (with or without cryoprecipitate). Thrombocytopenia in HELLP syndrome is more often severe and platelet transfusion may be indicated, particularly at delivery, which is usually by urgent Caesarean section, and postpartum.

Management of maternal ITP

ITP usually presents in an otherwise well mother with or without a previous history of ITP or, less commonly, SLE. In those without a previous history, the diagnosis may be difficult and platelet antibody studies are of limited value since they may be positive even in the absence of ITP. The management of the mother with active ITP and who is thrombocytopenic should be as conservative as possible. The most common approach to therapy is with intravenous immunoglobulin (0.3–0.5 mg/kg daily) for 3–5 days. However, prednisolone (1 mg/kg) can also be used. The indications for treatment of maternal ITP are: • platelets less than 20 ¥ 109/L in the first, second or early third trimester; • aim to have platelets above 80 ¥ 109/L in the late third trimester; • avoid epidural or spinal anaesthesia if the platelet count is less than 80 ¥ 109/L; • platelet transfusion is very rarely indicated; • splenectomy should be postponed until after delivery if possible; • fetal blood sampling and elective Caesarean section for maternal ITP are unnecessary since significant fetal thrombocytopenia is uncommon (12%) and intracranial haemorrhage is even less common (1%). The fetal platelet count cannot be predicted from maternal platelet counts nor from platelet serology. The most important factor predicting the presence and severity of fetal thrombocytopenia is a history of maternal ITP prior to pregnancy: in this higher-risk group, 10–30% of babies will have significant thrombocytopenia ( 240/unit

6.4–7.4 6.4–7.4

< 5 ¥ 106/unit < 5 ¥ 106/unit

265

Chapter 23

Frozen and washed red cells

This process is used only for red cells from donors with rare phenotypes or from occasional patients with multiple red cell alloantibodies for whom provision of compatible donor blood is extremely difficult. Red cells may be stored for years in the vapour phase of liquid nitrogen at –80°C provided a cryoprotectant (usually glycerol) is added prior to freezing. Such red cells must be washed several times in saline after thawing, which renders the final product leucocyte depleted. The same considerations regarding shelf-life apply to this product as to red cells washed without prior freezing. Pathogen inactivation in red cells

Red cells present a particular challenge for pathogen inactivation. Photochemical methods suitable for platelets and plasma cannot be applied to red cells because of the high degree of light absorption by haemoglobin. A number of compounds that do not depend on light activation are in development or in clinical trial, but none is yet available for routine use.

Platelet concentrates Sources of platelets for transfusion

Platelets may be produced either from wholeblood donations or by apheresis, in which platelets with or without plasma are collected and the red cells returned to the donor. Specifications for platelet yield and residual leucocyte count are similar for the two methods (see Table 23.3). Apart from exposing the patient to fewer donors and the possibility of HLA/human platelet antigen (HPA) matching with the patient, apheresis platelets are not intrinsically of higher quality. Platelet production from whole blood

Platelet production from whole blood may be carried out either from pooled buffy coats generated by BAT processing (see Fig. 23.3) or from platelet-rich plasma (PRP) as an intermediate step. 266

Buffy coat-derived platelets have long been favoured in Europe, and are now standard in the UK, while the PRP method is standard in North America. Leucocyte depletion by filtration may be routinely incorporated into either process. An adult therapeutic dose of platelets (2.5–3 ¥ 1011) can be manufactured from buffy coats, or by the PRP method, from four to six whole-blood donations, whereas the same dose can be harvested from an apheresis donor in an hour. Platelet production by apheresis

Systems are now available to collect one, two or even three adult doses of platelets during one collection procedure. This may require some preselection of donors according to platelet count and haematocrit. In the USA, donors were for a time administered thrombopoietin to boost platelet yields. This is no longer permitted, and has never been practised in the UK. Platelets are collected into citrate supplemented with adenosine and dextrose. Leucocyte depletion is performed either on the apheresis machine by physical separation, or by filtration soon after collection. Platelet storage

• After production and resting, platelet concentrates are stored in plasma in incubators set at 20–24°C for up to 5 days. Platelet concentrates should never be placed in the refrigerator. • Platelets must be agitated during storage, either in the horizontal plane or by tumbling rotation. Storage packs for platelet concentrates are manufactured from plastic designed to maximize gaseous exchange. • During storage, platelets undergo a fall in pH due to accumulation of lactate, express increased surface expression of activation markers such as Pselectin (CD62p), and change shape from discoid to round. Many different laboratory assays have been advocated to monitor development of this socalled ‘platelet storage lesion’ but few have been demonstrated to correlate with in vivo survival (Table 23.4). pH remains the only quantitative

Production and storage of components Table 23.4 International Society of Blood Transfusion BEST

(Biomedical Excellence for Safer Transfusion) proposals for platelet quality assurance. Routine quality assurance pH at day 5 Platelet count White cell count Volume of concentrate Swirling Evaluation of new component As routine plus: Surrogate for viability Morphology score Hypotonic shock Shape change ATP Platelet activation Surface P-selectin b-Thromboglobulin release Platelet lysis Supernate lactate dehydrogenase In vitro function Aggregation to pairs of agonists Metabolic activity PO2 and PCO2 Lactate production

change which must be monitored routinely, and must be between 6.4 and 7.4 at outdate. Visual inspection to look for the ‘swirling’ effect of discoid platelets has been recommended, but this is highly subjective and changes only when the platelets have been grossly damaged. • Bacterial contamination of platelets has been demonstrated with a frequency as high as 1 in 1000 platelet pools. Now that viral risks are so low, this is the commonest transfusion-transmitted infection. The clinical effects range from mild fever and rigors to acute septicaemia with hypotension and death, and several cases each year are reported to the Serious Hazards of Transfusion (SHOT) scheme. Platelets showing turbidity or unusual colour should be returned to the blood bank for culture. Because the major source of bacteria is the skin of the donors, improved skin cleansing and diversion of the first 20–30 mL of the donation into a side pouch which can be used for testing are

estimated to reduce the bacterial risk by as much as 70%. Bacterial screening of platelets is under consideration in a number of countries, including the UK, since semi-automated methods suitable for mass screening are now available. Because bacterial culture delays issue of the platelets by 1–2 days, there is increased interest in extending shelf-life to 7 days, which was permitted before the bacterial risks became apparent. With prestorage leucocyte depletion and the improved gas exchange offered by modern storage packs, platelets stored for 7 days in plasma maintain their pH well, and several countries are performing additional studies to assess their functionality. If this proves satisfactory, it might be possible to return to a 7-day shelflife for bacterially screened platelets. ‘Washed’ platelets and platelet additive solution

For patients with severe anaphylactic-type reactions, which are usually due to plasma proteins, it is possible to prepare platelets to be virtually plasma-free using one of a number of new platelet additive solutions. This product is sometimes referred to as ‘washed’ platelets, although washing is unnecessary and may lead to platelet activation. These solutions differ from red cell additive solutions in that they contain some or all of potassium, acetate, citrate, phosphate, gluconate and chloride. Platelets in 100% additive solution have only a 24-h shelf-life, but a number of countries have begun to produce platelets in a mixture of 30% plasma and 70% platelet additive solution. This strategy makes more plasma available for fractionation and allows a normal 5-day shelf-life. Data to day 7 and beyond are limited, although one new solution promises storage to day 9 and beyond. These solutions have great potential, but require careful validation, which may need to include volunteer and patient studies. Pathogen reduction in platelets

A group of compounds called psoralens have been developed for their virus and bacterial killing properties, and a system for pathogen inactivation of platelet concentrates using a second-generation psoralen called amotosalen (S-59) has been 267

Chapter 23

licensed in Europe. Psoralens form adducts with DNA and RNA; when activated by exposure to ultraviolet light (UV)A, binding becomes irreversible and nucleic acid replication is blocked. Amotosalen/UVA treatment results in a high degree of killing of the major transfusiontransmitted pathogens human immunodeficiency virus (HIV), hepatitis C virus (HCV) and hepatitis B virus (HBV), including intracellular pathogens such as CMV and HTLV. However, there is no effect on prions, which lack nucleic acid. Additional properties of S-59 photoinactivation include killing of Gram-positive and Gram-negative bacteria, inactivation of the antigen-presenting cells important in HLA alloimmunization, and inhibition of the donor T-cell proliferation that characterizes transfusion-associated graft-versus-host disease (TA-GVHD). Thus there could potentially be multiple benefits from this approach, and in clinical trials of S59-photoinactivated platelets it was considered unnecessary to perform gammairradiation for prevention of TA-GVHD. Randomized clinical trials, albeit on small numbers of patients, have shown that S59-treated platelets, whether prepared by apheresis or from pooled buffy coats, are effective in preventing haemorrhage in thrombocytopenic patients with haematological malignancies. However, platelet increments and intertransfusion intervals were less favourable than in control patients, raising the possibility that increased numbers of platelet units might be required to support such patients. This question can only be answered by further large-scale clinical studies. Other systems for pathogen reduction in platelets are also in development.

Fresh frozen plasma Definition and specification

FFP is the plasma from a single donation, usually 250–300 mL, which has been frozen soon after collection (usually within 8 h) without pooling to a core temperature of less than –30°C. To minimize virus risk, FFP is not manufactured from first-time donors in the UK. FFP can also be derived from apheresis collections, in 300-mL or 600-mL volumes. It is used primarily as a source of multiple 268

coagulation factors in situations such as massive transfusion, disseminated intravascular coagulation and liver disease (see Chapter 11). • The permitted shelf-life depends on storage temperature, e.g. less than –30°C for 12 months; 24month storage is now permitted in Europe. • At least 75% of units processed must contain more than 0.7 IU/mL of factor VIII. Although most FFP is prescribed for patients with normal or elevated factor VIII levels, it is selected for quality monitoring purposes as it is labile and hence sensitive to exposure to adverse conditions. • A standard unit of FFP may contain over 107 leucocytes/mL. There are no specific indications for leucocyte-depleted FFP but if universal leucocyte depletion is implemented, a filtration step is required. This can be done as whole blood prior to separation or by use of a specific plasma filter. • FFP is thawed (in a protective overwrap to prevent bacterial contamination) in a water bath, either with or without agitation; purpose-designed microwave ovens are also available. It has been traditional to recommend that, once thawed, FFP should be maintained at room temperature and transfused as soon as possible. However, more recent data suggest that post-thaw storage at 4°C for up to 24 h results in an acceptable product. Virus inactivation

Two virus-inactivated FFP preparations are now available in the UK, methylene blue (MB) treated and solvent–detergent (SD) treated, with other systems in development. Both methods offer good virus protection but are associated with loss of clotting factors. The key features of MB FFP and SD FFP compared with untreated FFP are shown in Table 23.5. MB is a phenothiazine dye that, when exposed to white light, generates reactive oxygen species which damage nucleic acids, preventing viral replication. Treatment is applied to single unpooled units of plasma, and requires prior removal of white blood cells by filtration or freeze–thawing. MB is contained in or added to the integral pack system, mixed with the plasma, then placed on a light box for activation. MB is generally removed using an adsorption filter prior to transfusion to

Table 23.5 Comparison of standard fresh frozen plasma (FFP) with methylene blue (MB)-treated FFP and solvent–detergent

(SD)-treated FFP. Standard FFP

MB FFP

SD FFP

UK donors, all previously virus tested. Single-unit format

US volunteer donors, all male. Single-unit format

Non-UK donors; pools of up to 380 L (600–1500 ABO identical donations)

Donation testing Serology Genomic

HIV, HBV, HCV, HTLV HCV

HIV, HBV, HCV, HTLV HCV, HIV

HIV, HBV, HCV, HTLV HAV, HCV, B19, HIV, HBV

Virus risk HIV-1 and HIV-2 HCV HBV HAV Parvovirus B19

1 in 10 million 1 in 50 million 1 in 1.2 million Rare event Rare event

No proven cases reported to date for HIV, HBV, HCV (1 possible HCV transmission)

No reported transmissions to date of HIV, HBV, HCV in SD FFP or SD-treated plasma products None reported Batch withdrawals due to possible B19 content. Seroconversion in patients no greater than with untreated FFP

Product characteristics Volume Coagulation factor content

180–300 mL + 50 mL paediatric size Variable between units; 75% units, factor VIII > 0.7 IU/mL

235–305 mL + 50 mL paediatric size Variable between units; 75% units, factor VIII > 0.5 IU/mL; all other factors > 0.5 IU/mL; no reduction ATIII, protein C, protein S. No coagulation factor/complement activation May become available

Not available

Source

No greater than for standard FFP. None reported to date

200 mL; no paediatric size Constant within batch. All factors > 0.5 IU/mL

Cryoprecipitate/ cryosupernatant

Available

Residual additives

None

< 0.3 mmol/L MB. No toxicity seen or predicted at this level, even in premature neonates

< 2 mg/mL TNBP; < 5 mg/mL Triton-X 100. Residual levels not toxic

Allergic reactions

May be reduced by leucocyte depletion 1% 0.1%

Reactions attributable to cells would be expected to be reduced No data No data

Probably less frequent than FFP

As for standard FFP

Pooling reduces all of these risks High-titre anti-A,B not a problem since donations pooled Only 1 possible TRALI case reported

Mild Severe Adverse reactions due to antibody Red cell

Tested for high-titre anti-A,B

Not tested for high-titre anti-A,B

TRALI Thrombocytopenia

> 20 cases/year (SHOT) Very rare

None reported to date

Cellular content

Leucocyte depleted

Leucocyte depleted

No intact cells or fragments; no need to RhD match

Product licence

Not required

Medical device; CE marked

Licensed, batched product

As for FFP

As for FFP

> 1 million units in Europe

3 million units in Europe

Indications Usage to date

300 000 units/year in UK

ATIII, antithrombin III; B19, parvovirus B19; HAV, hepatitis A virus; HBV, hepatitis B virus; HCV, hepatitis C virus; HIV, human immunodeficiency virus; HTLV, human T-cell leukaemia/lymphoma virus; SHOT, Serious Hazards of Transfusion; TNBP, tri(N-butyl)phosphate; TRALI, transfusion-related acute lung injury.

269

Chapter 23

the patient, leaving residual MB concentrations of less than 0.3 mmol/L. At these concentrations, no toxicity has been demonstrated or is predicted. Glucose 6-phosphate dehydrogenase deficiency is not a contraindication to use of this product. There is approximately 30% loss of activity of factors VIII and XI; of particular note is the effect of MB on fibrinogen, with up to 30% loss of activity. It is, however, unclear as yet whether these changes require increased volumes of the product to be prescribed. SD treatment can be applied only to pools of several hundred ABO-identical units; as the treatment destroys the lipid envelope of red cells, no RhD matching is required. Exposure to SD destroys the lipid envelope of HIV, HBV and HCV, and no such transmissions have been reported. Non-lipid-coated viruses such as parvovirus B19 and hepatitis A are not specifically inactivated but their titre may be reduced in downstream processing. In addition, plasma pools with high genomic titres of these viruses are rejected, and pools contain specified levels of viral antibodies, which may be at least partially protective. No increase in clinical cases of hepatitis A virus or B19 in SD FFP recipients is evident. Concerns have been expressed in the USA that SD FFP might be associated with increased thrombotic risk in certain clinical situations, attributed to loss of proteins C and S during SD treatment. However, these complications have not been prominent in recipients of SD FFP manufactured by the European method, which results in greater preservation of these proteins. A variant of SD FFP in which ABO groups are mixed is in development. By neutralization of anti-A and anti-B by soluble A and B substance, it is intended to produce a ‘universal FFP’ which could be given to patients of any ABO group. vCJD and FFP

In animal studies, plasma was found to contain infective prion. Therefore, a precautionary measure for the UK was announced in 2002 whereby FFP for children born on or after 1 January 1996 would receive FFP imported from

270

the USA, where no cases of BSE or vCJD have been reported. This is now being implemented, with plasma coming from volunteer donors extensively tested for transfusion-transmitted viruses, including West Nile virus. Plasma is pathogen inactivated by the MB process on arrival in the UK.

Cryoprecipitate and cryosupernatant Cryoprecipitate

Cryoprecipitate is manufactured from single units of FFP by rapid freezing to less than –30°C then slow thawing overnight at 4°C. This precipitates out the so-called cryoproteins, namely factor VIII, fibrinogen, fibronectin and factor XIII. By removing most of the supernatant plasma (‘cryosupernatant’), a component providing a high concentration of these clotting factors is obtained. Current UK guidelines specify that 75% of cryoprecipitate units must contain over 70 IU of factor VIII and 140 mg of fibrinogen, with a 24-month storage period. Although originally developed for factor VIII deficiency (haemophilia A), most cryoprecipitate is now prescribed to treat congenital or acquired hypofibrinogenaemia, usually in the context of liver disease, disseminated intravascular coagulation or massive transfusion. An adult dose of 10–12 packs is generally indicated once the fibrinogen level falls to less than 0.5–1.0 g/L. No licensed preparations of fibrinogen concentrate are yet available in the UK. Cryosupernatant

This is the portion of FFP remaining after separation of cryoprecipitate. It is generally discarded, although contains sufficient fibrinogen to be used as start material for production of fibrinogen concentrate. Cryosupernatant has been used succesfully as replacement fluid in plasma exchange procedures for thrombotic thrombocytopenic purpura. It may have advantages over FFP, possibly because it lacks the highest molecular weight multimers of von Willebrand factor.

Production and storage of components

Virus inactivation of cryoprecipitate and cryosupernatant

Production of cryoprecipitate from SD FFP has been performed experimentally. Such cryoprecipitate contains insufficient von Willebrand factor to treat patients with von Willebrand’s disease. Fibrinogen levels are reduced but acceptable. Manufacture of cryoprecipitate and cryosupernatant from MB plasma is under assessment, but is challenging because of the fibrinogen loss in the start plasma. However, potentially altered recovery in the cryoprecipitation process and the inhibitory effect of MB on fibrin polymerization mean that it may be difficult to achieve adequate concentrations of functional fibrinogen in the MB-treated cryoprecipitate.

Granulocytes for transfusion The use of transfused granulocytes is now uncommon. They are sometimes used for severely neutropenic patients (granulocyte count < 0.5 ¥ 109/L) with focal bacterial or fungal infection refractory to antimicrobial therapy, but there are difficulties in obtaining sufficient functional cells from donors and administering them frequently enough to the patient. Production options have been to use 10 or 12 buffy coats from random donors, or to harvest apheresis granulocytes from family members using a red cell sedimenting agent such as hydroxyethyl starch. Animal studies suggest that greater than 1 ¥ 1010 granulocytes once or twice daily are required to treat an adult, but apheresis can produce no more than 0.5 ¥ 1010 per dose. Therefore unstimulated apheresis granulocytes are directed towards children, in whom an adequate dose can be achieved, with adult patients receiving buffy coat-derived granulocytes. There has been renewed interest in the use of granulocytes in studies of granulocyte colonystimulating factor (G-CSF)-mobilized granulocytes collected by apheresis. Administration to the donor of a single subcutaneous injection of 10 mg/kg of G-CSF plus oral dexamethasone 8 mg 12–24 h prior to apheresis raises the peripheral

leucocyte count to more than 25 ¥ 109/L. This, coupled with Pentaspan sedimentation, allows collection of a therapeutic dose of granulocytes of 5–20 ¥ 108/kg body weight of the recipient. This can result in a measurable rise in the peripheral granulocyte count in the patient, and recovery of migrated cells from saliva. Clinical trials of such granulocytes are ongoing. At present, use of GCSF for granulocyte collection is not permittted in volunteer donors unrelated to the patient. All granulocyte preparations should be released for issue as soon as possible after collection, which may mean that certain time-consuming screening assays such as HCV genome testing cannot be done prior to release. They must be gammairradiated to prevent TA-GVHD and should be administered to the patient without delay. A short period of storage is unavoidable, which should be at 22°C without agitation. Because of red cell contamination, a red cell crossmatch should be performed.

Components for intrauterine transfusion, and for neonates and infants General requirements (Table 23.6)

• In the UK, such components are not manufactured from ‘first-time’ donors. FFP has not been shown to transmit CMV so provision of CMV antibody-negative FFP is not critical. • For components other than those in additive solution, they must be free of clinically significant red cell antibodies, including high titre-anti-A and anti-B. • Blood from haemoglobin sickle heterozygous donors should not be used. • Although components are leucocyte depleted at source, they should still be administered through a 170–200 mm filter to remove any microaggregates formed during storage. • Gamma-irradiation is required for intrauterine and exchange transfusions. ‘Top-up’ red cell transfusions need be irradiated only if there has been a previous intrauterine transfusion (IUT) or if the component is prepared from a family member. Family donations are not encouraged except in

271

Chapter 23 Table 23.6 UK specifications for red cells for intrauterine transfusion (IUT), exchange/large-volume transfusions and ‘top-up’

transfusions for neonates.

Virology-negative donation in previous 2 years Free of high-titre anti-A, anti-B Use of additive solution permitted Leucocyte depleted Gamma-irradiation Haemoglobin S negative Cytomegalovirus seronegative Shelf-life

IUT

Exchange transfusion

Top-up transfusion

Yes Yes No Yes Yes Yes Yes 24 h after irradiation and < 5 days total

Yes Yes No Yes Yes Yes Yes 24 h after irradiation and < 5 days total

Yes No Yes Yes No Yes Yes 35 days

rare cases of fetomaternal alloimmunization where the infant’s requirements cannot be met from donor blood.

should apply to other large-volume transfusions in neonates, such as for cardiac surgery or extracorporeal membrane oxygenation.

Intrauterine transfusions

Top-up transfusions for neonates

Red cells are given in utero to treat severe fetal anaemia due to haemolytic disease of the newborn or parvovirus B19 infection. Red cells for IUT are prepared from blood less than 5 days old to a haematocrit of more than 0.7–0.9 and are gammairradiated. They should be administered within 24 h of irradiation, and always by the end of day 5. In cases of fetomaternal alloimmunization to platelets, weekly transfusions of selected platelets (usually HPA-1a negative) are given in utero via the umbilical vessels. Production of hyperconcentrated platelets by apheresis of a genotyped panel of platelet donors is now possible, yielding a platelet ‘hyperconcentrate’ of more than 120 ¥ 109 platelets in 60 mL of plasma.

Premature neonates are among the most heavily transfused patients in any hospital. Most red cell transfusions are given to replace repeated samples taken for laboratory testing. As each infant may require multiple small transfusions, adult packs are split into four to eight ‘paedipacks’ of 30– 60 mL, which are allocated to one infant for the duration of transfusion dependence. Such a strategy reduces donor exposure considerably. For these small-volume transfusions, red cells in additive solution may be used, up to the normal 35-day shelf-life. Gamma-irradiation of these is not required in the UK, unless there has been a previous IUT or the blood comes from a family member. Studies of erythropoietin in premature infants have not convincingly shown a reduction in transfusion requirements.

Exchange/large-volume transfusion of neonates

This is undertaken to treat hyperbilirubinaemia due to either haemolytic disease or prematurity. Either whole blood or partially packed red cells with a haematocrit of 0.6 may be used. Red cells in additive solution are not recommended by some paediatricians, because of concerns regarding the adverse effects of mannitol, but some countries use this component for exchange transfusion without apparent problems. The same considerations 272

Platelet concentrates and FFP

These are most simply prepared from apheresis donations. Multiple aliquots can be allocated to the same infant if required. An alternative strategy for platelets is to prepare a platelet concentrate from a single buffy coat or from a unit of whole blood using the PRP method. These components are generally used for sick babies with multiple

Production and storage of components

coagulation defects. Platelets from a panel of HPA1a- and HPA-5b-negative donors are available ‘off the shelf’ for rapid availability for suspected cases of neonatal alloimmune thrombocytopenia (see Chapter 5).

Further reading British Committee for Standards in Haematology. Guidelines on the clinical use of leucocyte-depleted blood components. Transfus Med 1998; 8: 59–71. British Committee for Standards in Haematology. Guidelines for the use of platelet transfusions Br J Haematol 2003; 122: 10–23. British Committee for Standards in Haematology. Transfusion guidelines for neonates and older children. Br J Haematol 2004; 124: 433–53. British Committee for Standards in haematology. Guidelines for the use of fresh-frozen plasma. Br J Haematol 2004; 126: 11–28. Council of Europe. Guide to the Preparation, Use and Quality Assurance of Blood Components, 9th edn. Strasbourg: Council of Europe Publishing, 2003.

Dumont LJ, Dzik WH, Rebulla P, Brandwein H, and members of the BEST Working Party of the ISBT. Practical guidelines for process validation and process control of white cell-reduced blood components: report of the Biomedical Excellence for Safer Transfusion (BEST) Working Party of the International Society of Blood Transfusion (ISBT). Transfusion 1996; 36: 11–20. McClelland DBL, ed. Handbook of Transfusion Medicine, 3rd edn. London: The Stationery Office, 2001. Pamphilon DH. Viral inactivation of fresh frozen plasma. Br J Haematol 2000; 109: 680–93. Pamphilon DH, Rider J, Barbara JAJ, Williamson LM. Prevention of transfusion-transmitted cytomegalovirus infection. Transfus Med 1999; 9: 115–23. United Kingdom Blood Transfusion Services/National Institute for Biological Standards and Control. Guidelines for the Blood Transfusion Services in the United Kingdom, 6th edn. The Stationery Office: London, 2002. Williamson LM. How should the safety and efficacy of platelet transfusions be assured? Blood Rev 1998; 12: 203–14.

273

Chapter 24

Medicolegal aspects Patricia E. Hewitt

The provision of blood and blood products from donation to transfusion must be based on sound ethical principles and quality guidelines, and ideally should be controlled within a robust regulatory framework based on statute.

Ethical principles The International Society of Blood Transfusion (ISBT) some years ago instituted a code of ethics setting out the guiding principles for blood donation and transfusion. Following revision, this code of ethics has been adopted by the World Health Organization (WHO). It has also been used to support ethical standards in the drafting of the European Blood Directive. It is recommended that all blood transfusion provision is in accordance with the principles included in this code. The main provisions are listed below. • There should be no coercion to donate blood. • Both donors and recipients must be adequately informed. • Confidentiality must be maintained. • Adequate standards should be enforced. • Clinical need must be the determinant of transfusion therapy.

Quality guidelines Uniformity and process control can be achieved by compliance with detailed quality guidelines. The Guidelines for the Blood Transfusion Services in the United Kingdom and the American Association of Blood Banks Technical Manual are two examples of such documents. Although such 274

guidelines do not possess legal status, they can be interpreted as the definitive statement of minimum professional standards, and set out the requirements to be met for good manufacturing practice. Deliberate non-compliance would be regarded very seriously. Unavoidable non-compliance should be carefully documented and should include a clear explanation of the reasons for noncompliance.

Regulatory framework in the UK The regulatory framework in the UK is briefly described. Similar arrangements will be in place in all developed countries. Medicines Act 1968

The Medicines Act provides the framework for the regulation and control of all dealings with medicinal products. It is clear that cellular and fractionated blood products are included within the terms of the Act. Fractionated products (e.g. albumin, coagulation factor concentrates, intravenous immunoglobulin preparations) are individually licensed. The provision of labile blood components (red cells, platelets, fresh frozen plasma) is enabled by means of an organizational licence awarded to individual blood centres by the Medicines and Healthcare Products Regulatory Agency (MHRA) following appropriate inspection and demonstration of compliance with the standards of good manufacturing practice. This Manufacturers Special Licence is valid for 5 years and individual facilities are reinspected and revalidated on a 2-year cycle. Failure to comply with the licensing

Medicolegal aspects

regulations can lead to the imposition of a fine or closure of an organization, and could proceed to a claim of negligence.

although the means of detecting the defect (the availability of a hepatitis C test) was not necessarily available during the whole period.

Consumer Protection Act 1987

NHS Act 1999

The Consumer Protection Act creates a strict liability action against manufacturers and suppliers when physical injury or property damage is caused by a defective product. The Consumer Protection Act 1987 was enacted as a result of a European Community Directive in 1985 and clearly includes within its terms the provision of all blood and blood products. Its premise is the principle of product liability, that is that there is no need to prove that a negligent action has taken place, merely that the end product is defective and has caused harm. In section 3 (1) of the Act a ‘defect’ is defined as follows ‘There is a defect in the product . . . if the safety of the product is not such as persons generally are entitled to expect . . . ’. Blood providers can be held liable under the terms of the Act as producers, suppliers or keepers. The liability therefore extends from the blood centre producing the product to the hospital blood bank which stores and issues products. There are possible defences within the terms of the Act, such as the ‘state of the art defence’. In essence this means that if a product is found to be defective based on current knowledge, that information cannot be used to prove that the same product was defective sometime previously when the current knowledge was not available (Table 24.1). This defence was not held to apply in the case of the hepatitis C litigation in England, since the defect (the transmission of hepatitis C) was apparent at the time of the claimants’ transfusions (1988–91),

This Act modernizes the NHS in England, Wales and Scotland. Raising standards in the quality of NHS care is at its heart. A statutory duty of quality is now placed on all NHS providers, monitored by means of the Commission for Healthcare Audit and Inspection (CHAI). Additionally, the National Institute of Clinical Excellence (NICE) ensures minimum standards of healthcare are developed throughout the UK. It is accepted that the spur to set up these statutory provisions had been inequality in care. All aspects of healthcare including blood transfusion will no doubt in time be scrutinized; the emphasis is likely to be on clinical aspects of transfusion therapy.

Table 24.1 Relevant UK statutes.

Act

Terms

Medicines Act 1968

Regulates the donation, testing, processing and issuing of blood Encompasses all aspects of provision of blood and blood products from donation to hospital blood bank Concentrates on clinical use of blood

Consumer Protection Act 1987

NHS Act 1999

Duty of care Putting aside the ethical principles, quality guidelines and regulatory framework described above, there remains the clear duty of care which must be at the heart of the provision of blood transfusion. This duty must be according to an accepted standard. At present the standard is determined according to the Bolam principle: ‘The test is the standard of the ordinary skilled man exercising and professing to have that special skill.’ This defines the standard as that of a responsible body of doctors skilled in the same specialty. The standard of care can be supported by the application of professional guidelines, although currently the latter have no legal standing. It must be remembered that inexperience is not a defence and a junior doctor in a specialty is deemed to be equally liable with a senior doctor in the same specialty. The duty of care of blood services, according to the defined standard, is both to the blood donor and to the recipient patient. Duty to the donor

The two general principles that underpin blood 275

Chapter 24

donation are that there should be no harm to the health of the donor, and no risk to the health of the recipient patient. The duty to the donor would therefore include compliance with strict medical selection procedures. The donor needs to be informed about the screening tests performed on the donation and should provide a written consent to human immunodeficiency virus (HIV) testing. Information should be provided on any situation that could potentially pose risk to the donor, for example the administration of growth factors prior to stem cell donation or the use of general anaesthetic during bone marrow donation. In these instances a donor would need to consent formally to the procedures. The blood services in the UK make leaflets available at all routine blood donation sessions, to inform the prospective blood donor of relevant issues. Additionally, the blood service has a duty to maintain the confidentiality of a donor, particularly in the event of a recipient patient being harmed by blood obtained from a single donor. The duty of care to the donor also extends to the clinician prescribing the blood, to ensure appropriate use, particularly as that donation is provided on a voluntary basis with no expectation of monetary gain. Duty to the recipient patient

In the UK the standard of care for patients receiving blood transfusion is likely to be addressed under the legislation referred to above. It is suggested that, as a minimum, this standard of care should include the provision of adequate information to the recipient patient and ensuring appropriate clinical use of individual blood components.

Consent to transfusion Any patient being asked to consent to a medical treatment or investigation has the right to be informed of the aims, benefits and risks of the treatment, and to be given details of any alternatives. Without such information, consent is not valid. The standard NHS ‘Patient agreement to investigation or treatment’ form includes a section completed by the health professional who has the 276

discussion with the patient. This section documents that an explanation has been given to the patient about the proposed investigation or treatment, including the possibility of extra procedures which may be found necessary, and blood transfusion is specifically mentioned at this point. The patient signs a general consent to the procedure/investigation described, embracing the possibility of additional procedures, but has the opportunity to list any procedures for which he or she witholds consent without further discussion. The patient must have the capacity to consent. It is presumed that an adult will have that capacity. No doctor should force a competent adult to accept any treatment even if that adult’s decision appears to be irrational. An adult could be incapacitated and therefore unable to give consent because of loss of consciousness or mental retardation. In general no other person can give consent on behalf of an incapacitated adult. (In some countries, e.g. Scotland, the power of parens patriae applies, where another adult can take responsibility as a parent for an incapacitated individual.) Prior wishes may be taken into consideration where the adult has previously been competent and the treatment is regarded as noncontroversial. Treatment may be given to an adult incapable of consenting if the treatment is urgent and in the patient’s best interests. In an elective situation, however, it would be best to seek a ruling from a court of law. The General Medical Council has published a comprehensive guide to consent and this is recommended for more detailed reading (Table 24.2). In the case of children, the Family Law Reform Act 1969 makes it lawful for a minor to consent to, or refuse, treatment when he or she reaches the age of 16 years. In the case of a child below 16 years of age, the parents usually give consent, although such children are able to give valid consent in their own right if they are capable of understanding Table 24.2 Informed consent must include these elements.

Capacity to understand Should be based on adequate information Should be obtained without coercion

Medicolegal aspects

clearly the nature of the proposed treatment. Here the difficulty could be where parents could have specific religious beliefs that prevent them from consenting to blood transfusion for their child. In this instance, a doctor can decide to provide a treatment, including blood transfusion, in the child’s best interests. To fulfil these criteria, the treatment must be carried out in order to save life or to ensure improvement of or to prevent deterioration in the physical or mental health of the child. This would form the basis of a doctor’s individual decision during an emergency situation, but in the case of a planned blood transfusion it would be appropriate to seek a ruling from a court of law. In these circumstances it is recommended that the doctor’s medical defence body is consulted for advice on how to proceed. For consent to be informed and valid, it must be based on adequate information. Various attempts have been made to define what should be included as adequate information and it is legally acceptable that the explanation need not include all the potential adverse consequences if the risk of them occurring is small or immaterial. Minor insignificant reactions to transfusion occur relatively commonly, whereas the risk of complications with serious or fatal long-term consequences, e.g. transmission of HIV, is extremely low. However, there is heightened public awareness of such low risk, and it would therefore be appropriate for these events to be included in a preliminary explanation. Again, the standard that applies in the UK regarding the provision of information is the standard of a responsible body of skilled doctors, the Bolam principle. In the USA, however, a different rule applies, the standard being judged not according to the information that a reasonable doctor would think relevant to impart, but rather according to that which a prudent patient would think relevant to receive, a situation which is likely to develop within the UK within the next few years. There must be no coercion in obtaining consent. A competent adult is able to accept or refuse treatment even if that decision could lead to harm or indeed death. Should an individual doctor decide to treat an adult without consent, then that doctor should be prepared to explain and justify the decision.

Jehovah’s Witnesses Jehovah’s Witnesses, because of their religious beliefs, will never accept normal blood transfusion therapy, although in appropriate circumstances could find cell salvage in continuous circulation acceptable. Many Jehovah’s Witnesses carry an Advance Medical Directive which states the individual’s views and requirements to be followed in the event that the individual is unconscious or otherwise unable to express his or her views. What should be done in this instance? Where the situation is one relating to a competent adult, as long as it is clear there is no coercion, the decision to refuse treatment must be respected, even if it would lead to harm or indeed death of the patient. In an emergency situation, if the patient is unconscious, and therefore incapacitated, then prior previously held beliefs must be taken into account and blood transfusion should not be prescribed if those beliefs made it clear that it was unacceptable. In the situation of a child, where the parents’ religious beliefs could prevent the child from being given a necessary blood transfusion, it would be advisable to seek a proper legal ruling, which would usually mean the child becoming a ward of court and therefore decisions on the treatment being taken by the court.

Patient recourse Despite compliance with standards and appropriate care, things do and will go wrong. In some countries, for example New Zealand and the Scandinavian countries (Sweden, Norway, Finland and Denmark), compensation for medical accidents is provided under a ‘no fault’ system. However, in most countries there is a need to prove liability. In these circumstances, liability will rest either with the individual doctor or with the health employer if vicarious liability applies (this is the current position in the UK for all NHS work). There have been examples of ‘no fault’ compensation awarded in the UK in specific circumstances, for example the Vaccine Damage Payment Scheme, which provides payment of a fixed lump sum where serious mental or physical damage has been caused by the 277

Chapter 24

administration of specified vaccines. In relation to UK blood transfusion, payments were provided to recipients infected by HIV through transfusion both before and after the introduction of mandatory screening of the blood supply in the UK. These were ex gratia payments to those affected, with the government emphasizing that they should not be regarded as an admission of liability or as compensation, but as a response to a particular and tragic situation. Requests for similar treatment for individuals infected with hepatitis C (and hepatitis B) were, until recently, refused. However, an announcement in August 2003 indicated that patients who had been infected with hepatitis C through treatment with blood products would receive payments in a similar manner. At the time of writing, it has been stressed that similar treatment will not apply to recipients infected with, for example, hepatitis B (or any other agent) through blood transfusion, although such cases are small in number and it becomes difficult to explain the different treatment of patients who have suffered apparently similar unfortunate and unexpected adverse effects through their treatment with blood transfusion. Furthermore, the announcement of ex gratia payments for hepatitis C infection followed the successful claims under the Consumer Protection Act (see below) and has led to a number of questions which remain unresolved. For example, can recipients receive payments twice over and how will the level of payments be decided in the hepatitis C scheme? A patient who has suffered harm can bring an action either in medical negligence or under product liability. If brought in negligence, there would be a need to prove a breach in the duty of care, and that the breach directly caused harm to the patient. If brought under product liability, negligence need not be present; a defective product must have directly caused the harm (Table 24.3). An example of an action that could be brought under medical negligence would be that of the transfusion of a unit of red cells to the wrong recipient patient because of failure to check patient identification. Here there would be a clear breach of the duty of care, that is in checking the patient identification against the red cell unit, and it would also be simple to demonstrate that harm, in the 278

Table 24.3 Comparison of medical negligence versus

product liability. Medical negligence

Product liability

Duty of care Breach of the duty Harm caused directly by the breach

Defective product Harm caused directly by the defect

form of a haemolytic transfusion reaction, had occurred as a direct result of the breach. The recipient patient would be able to seek damages for the injury and compensation for any consequent financial loss. Cases of product liability in relation to blood transfusion in Europe are few. The most notable case was that of a number of recipients (114) in England and Wales who brought a claim under the Consumer Protection Act which was heard in 2000–01. In his judgement, Burton found that the Blood Service was liable for the damage because the product (i.e. the blood) did not provide the safety that the consumer (patient) was ‘generally entitled to expect’. The claimants were awarded damages on a provisional basis according to the damage (extent of hepatitis C disease) present at the time of the action. Provisional damages allow for the claimants to return with a future claim should their medical condition deteriorate. This case has attracted much attention within Europe and will surely be used as a precedent in future claims. Other case law within Europe is scarce, although some European countries (e.g. France and the Scandinavian countries) provide for ‘no fault’ compensation in relation to infection acquired through medical treatment. The judgement in the hepatitis C litigation was not appealed. Future similar claims are likely to be made and will be settled, unless the judgement is contested.

Note added in proof The European Blood Safety Directives (2002/93 and 2004/33) will be transposed into UK law as the Blood Safety and Quality Regulations 2005, due to come into force on 8th February 2005.

Medicolegal aspects

The new regulations impose safety and quality requirements on human blood collection, testing, processing, and storage. The requirements apply to blood transfusion service in England, Scotland, Wales, and Northern Ireland. Many of the provisions of the regulations also apply to hospital blood banks. A further Direcive on haemovigilance and traceability, and on quality systems is being discussed by the European Commission’s Expert Group on Blood. It is likely that this further Directive will not be finalized until 2005 and will be then transposed into UK law by separate amending legislation. The regulations will replace some of those currently covered under the Medicines Act (inspection, licensing and accreditation). It is intended that these activities will cease to be regulated under the Medicines Act, which will therefore need to be amended. At the time of writing, the new regulations are subject to public consultation. They will have wide-reaching implications for both Blood Services and hospital blood transfusion laboratoris.

Further reading

American Association of Blood Banks. Technical Manual, 14th edn. Bethesda, MA: AABB, 2002. Bolam v Friern Barnet Hospital Management Committee [1957] 2 All ER 118. Braithwaite M, Beresford N. Law for Doctors: Principles and Practicalities. London: Royal Society of Medicine Press, 2002. The Consumer Protection Act 1987. In: Halsbury’s Statute of England. HMSO: London. F v West Berkshire Health Authority [1989] 2 All ER 545. General Medical Council. Seeking Patients’ Consent: the Ethical Considerations. London: GMC, 1999. Gillick v West Norfolk and Wisbech Area Health Authority [1984] 3 All ER 402. Goldberg R. Paying for bad blood. Strict product liability after the hepatitis C litigation. Med Law Rev 2002; 10: 165–200. Grubb A, Pearl DA. Blood Testing, Aids and DNA Profiling. Bristol: Family Law (Jordan and Sons Ltd), 1990. Guidelines for the Blood Transfusion Services in the United Kingdom, 6th edn. The Stationery Office: Norwich, 2002. The Medicines Act 1968. In: Halsbury’s Statute of England. HMSO: London. NHS Conferation. The coming year in Parliament for health. Briefing 1998; issue 24. Warden J. HIV infected haemophiliacs: 90 million more. Br Med J 1989; 299: 1358. Williams FG. Consent for transfusion. Br Med J 1997; 315: 380–1.

A and others v National Blood Authority [2002] 3 All ER 289.

279

Chapter 25

Blood transfusion in hospitals Sue Knowles and Geoff Poole

Transfusion medicine in a hospital setting is focused on ensuring that a patient receives the correct blood component support which is clinically indicated, in a safe, timely and cost-efficient manner. Specialists in all branches of medicine and surgery are involved in prescribing blood components and the transfusion process itself involves multiple steps and the cooperative action of several groups of staff. Haemovigilance reports from around the world have confirmed that mistransfusion, i.e. giving the patient the incorrect unit of blood, resulting in an ABO-incompatible transfusion, is the most frequent cause of mortality and morbidity resulting from blood transfusion. Mistransfusion is the result of human error that can occur at all steps in the transfusion process, due to failures to comply with clerical or technical procedures and often compounded by systems that are either poorly constructed or not understood. Multiple errors can be made in the transfusion process, some of which can be detected during an effective bedside check at the time of administering blood. It has been observed that as many as 1 in 19 000 red cells are given erroneously and 1 in 33 000 will involve ABO-incompatible units. Estimates of mortality from mistransfusion range from 1 in 600 000 units to 1 in 1.8 million. While there has been enormous progress made in reducing the risks of transfusion-transmitted infections, there is no evidence from serial annual haemovigilance reports that the safety of blood transfusion in a hospital setting is improving. Although there are numerous guidelines and directives available, and an understanding of measures which can be taken to improve transfusion safety, in practice an effective transfusion quality assurance programme is required to minimize the risk of 280

mistransfusion and to avoid the other patient risks and wastages associated with the transfusion process (Table 25.1). This in turn requires committed leadership and adequate resources.

Key features of a quality assurance system for transfusion practice A quality assurance system can be described as the sum of the activities planned and performed to provide confidence that all processes and their elements that influence the quality of transfusion practice are working as expected. In the UK, the Department of Health emphasized this requirement in a health service circular in 2002 entitled ‘Better Blood Transfusion’ (HSC 2002/009). The key features for assuring safe and effective transfusion therapy can be summarized as follows. • Hospital or Trust executive/board: to provide commitment and appropriate resources for the quality system. • Hospital transfusion committee (HTC): to provide multidisciplinary ownership, leadership, and review of the quality system. • Hospital transfusion team (HTT), consisting of the specialist practitioner(s) of transfusion (hospital transfusion safety officer), the lead consultant for blood transfusion and the blood bank manager: to deliver the HTC’s objectives of improving transfusion medicine practice and to implement and monitor the processes of delivering patient care. • Staff involved in the transfusion process: to deliver and/or receive continuous education in transfusion medicine, training in specific procedures and assessment of their competencies.

Blood transfusion in hospitals Table 25.1 Errors in the transfusion process and some potential outcomes.

Problem

Outcome

Unnecessary prescriptions

Patient subjected to unnecessary risk Wastage of blood components

Failure to prescribe specialist components

Risk of, for example, transfusion-associated graft versus host disease

Failure to keep blood in a controlled environment

Wastage of blood components

Incorrect interval between sampling for pretransfusion testing and transfusion

Potential for acute and delayed haemolytic transfusion reactions

Sample for pretransfusion testing taken from incorrect patient. Transposition of samples or other errors in the laboratory. Incorrect unit of blood collected for and/or administered to the patient

Mistransfusion and potential for an ABO- or RhD-incompatible transfusion

Insensitive techniques in pre-transfusion testing

Potential for acute and delayed haemolytic transfusion reactions

Poor laboratory stock control

Wastage of units Inappropriate use of group O and national shortages of that group

Delays in provision of blood components in an emergency

Patient morbidity from hypoxia or coagulopathy

• Policies and guidelines: documented principles and recommendations which guide all activities. • Standard operating procedures/integrated care pathways: documented work instructions and steps which should be followed. • Equipment/material: planned maintenance and calibration of equipment (e.g. blood bank refrigerators, blood warmers and centrifuges), validation of reagents and techniques used in pretransfusion testing. • System review: (a) performance monitoring, e.g. utilization statistics, wastage, results of laboratory external quality assessment; (b) audits, i.e. compliance with guidelines or procedures; and (c) incident and error investigation, analysis and reporting.

Healthcare Organizations. In the UK, the Department of Health initially issued a health services circular ‘Better Blood Transfusion’ (HSC 1998/224) in December 1998, requiring all hospitals/Trusts to have an HTC in place by March 1999. Since then, HSC 2002/009 has expanded the requirements to encompass the key features of a quality infrastructure. To be effective, the HTC requires a dedicated HTT and adequate resources, including IT and clerical support to facilitate data retrieval and audit. Given the importance of the HTC with respect to clinical governance, it should report to the hospital/Trust management board, through options which include the clinical governance committee, the risk management committee or the clinical executive. Terms of reference

Hospital Transfusion Committee The HTC is the focal point for overseeing transfusion practice. In the USA, an HTC has been a requirement since 1972 for hospital accreditation by the Joint Commission on Accreditation of

The HTC is charged with the review of: • clinical transfusion practice; • performance of the hospital transfusion service; • performance of the local blood centre as a provider; and • legal implications of transfusion practice. 281

Chapter 25

Functions Clinical transfusion practice

• To ensure that local policies and procedures are in place, based upon national guidelines and regulations. These should include the following: (a) the administration of blood and blood components and the management of the transfused patient, including the collection of samples for compatibility testing and the management of adverse events related to blood transfusion; (b) a maximum surgical blood ordering schedule (MSBOS); (c) the appropriate use of blood components and blood products; (d) indications for specialist blood components (gamma-irradiated, CMV seronegative and phenotyped); (e) management of massive transfusion; (f) use of autologous blood; (g) use of pharmacological agents to reduce blood usage during surgery. • To ensure that all relevant groups of staff are trained in accordance with local policies, and their competencies are assessed. • To commission audits of compliance with these policies and procedures. • To organize continuing education in transfusion medicine for all members of the hospital staff involved in prescribing or administering blood. • To ensure that investigations are conducted of near misses and adverse events, which in turn will focus the need for further education or amendments to existing procedures. • To ensure that transfusion incidents are reported to the national haemovigilance scheme, i.e. Serious Hazards of Transfusion (SHOT) in the UK. • To review performance regularly, e.g. wastage rates, inappropriate blood component usage, utilization of blood components by directorate, user or surgical procedure. Monitoring the performance of the hospital transfusion service

• To review the operational effectiveness of the service, e.g. response times for emergency requests, elective work undertaken on call, crossmatch to 282

transfusion ratios for individual surgical procedures and users. • To review quality assurance measures including performance in external quality assessment schemes and the outcome of accreditation inspections and other external audits. Monitoring the performance of the local blood centre as provider

• To review the adequacy and timely provision of blood and blood components. • To review the adequacy of diagnostic reference services and consultant advisory services. Legal implications of transfusion practice

• To ensure that patients undergoing transfusion are provided with access to information in relation to the risks and benefits of transfusion. • To ensure that an audit trail of documentation exists to trace the ultimate fate of all blood components received. • To ensure that all relevant aspects of product liability and health and safety are adequately addressed. Composition

The Chair should be appointed by the Trust chief executive and should have a good understanding and experience of transfusion medicine practice. Ideally, the Chair should not be the consultant responsible for the hospital transfusion laboratory, who could be perceived to have a vested interest. The following membership is suggested: • representatives of major speciality users of blood in all directorates; • lead consultant haematologist for blood transfusion; • hospital blood bank manager; • specialist practitioner(s) of transfusion; • senior nursing officer; • representative from clinical risk management; • representative of junior medical staff; • representative of Trust management; • local blood centre consultant (ex officio);

Blood transfusion in hospitals

• other representatives may be co-opted as required, e.g. from medical records, portering staff, clinical audit, training or pharmacy.

Administration of blood and blood components and the management of the transfused patient This process involves several steps: • counselling the patient of the need for a blood transfusion, given that alternative approaches (autologous transfusions and/or erythropoietin) are insufficient or inappropriate for their circumstances; • the prescription of blood and blood components; • requests for blood and blood components; • sampling for pretransfusion compatibility testing; • collection and delivery of blood and blood components from transfusion issue refrigerator to clinical care area; • administration of blood and blood components; and • monitoring of transfused patients. Errors can occur at each of these steps and data provided in the UK SHOT report for 2002–03 shows that 20% of errors are made at the time of

prescription, sampling and request, 29% arise in the hospital blood bank and 48% are made at the time of collecting and administering the component. The Trust should have written procedures to cover all these steps, to which the relevant staff are trained and assessed. The responsibilities, actions, documents and potential errors are provided in Tables 25.2–25.7. The blood transfusion compatibility form, the prescription chart and the nursing observations related to the transfusion should be kept in the patient’s medical case notes as a permanent record. Since correct identification of both patient and blood unit are critical control processes, computerized systems have been developed for the blood administration process, which includes bedside verification of the match between patient identification and blood unit identification.

Technologies to reduce errors in administering blood Additional manual systems of patient identification

Sets of red labels with the same unique number can be allocated to a transfusion episode. A label can be incorporated into an additional patient wristband at the time of phlebotomy, affixed to the

Table 25.2 Prescription of blood and blood components.

Responsibility

Action

Document

Potential errors

Medical officer

Ensure patient is aware of the need for a blood transfusion and has read and understood information related to the risks and benefits of transfusion Prescribe component, any specialist requirements, quantity, and duration of transfusion

Patient information leaflet Hospital consent form

Failure to take account of patient’s religious beliefs or other views

Prescription chart

Unnecessary prescription, in the failure to follow hospital guidelines or as a result of an error in baseline blood count or coagulation screen. Lack of awareness of or failure to prescribe specialist components

Document rationale for transfusion

Patient case notes

Related hospital procedures and documents Guidelines for the use of blood and blood components, including specialist components. Practice guidelines/procedures for individual diseases/treatments.

283

Chapter 25 Table 25.3 Requests for blood and blood components.

Responsibility

Action

Documentation

Potential errors

Doctor or registered nurse

State full patient identification, location, diagnosis, details of type and quantity of component and time required. State previous obstetric and transfusion history when requesting red cells

Written/electronic request form or laboratory telephone log in an emergency

Incomplete or incorrect patient information leading to failure to recognize historical laboratory record, recording requirement for specialist components or phenotyped units Failure to request specialist components Failure to comply with hospital MSBOS

Hospital blood transfusion laboratory staff

Review historical record and whether a further sample for pretransfusion testing required

Previous laboratory record

Patient identification error in transcribing telephone request Failure to locate/heed information contained in historical record Failure to request a new sample in a recently transfused patient and potential for overlooking newly developed red cell antibodies

Related hospital procedures and documents Timing of pretransfusion sampling with respect to previous transfusion. Maximum surgical blood ordering schedules (MSBOS).

Table 25.4 Sampling for pretransfusion compatibility testing.

Responsibility

Action

Medical, nursing or phlebotomy staff, authorized and trained

Direct questioning of patient to provide surname, first name and date of birth when judged capable. Check that details given match those on patient wristband and on request form Take blood sample and immediately label at bedside with the required patient information

Hospital blood transfusion laboratory staff

To determine that sample labelling meets requirements for pretransfusion testing. If unacceptable, to inform requester of the need for another sample

Documentation

Potential errors Patient misidentification as a result of failing to positively identify patient or as a result of wristband missing or with incomplete information

Sample labelled and signed

Patient misidentification as a result of: • Prelabelled sample tube with another patient’s ID • Labelled away from the bedside with another patient’s ID • Addressograph label affixed from incorrect patient

If unacceptable, to document reasons and log

Potential to issue inappropriate unit if inadequately labelled samples accepted Failure to provide blood in a timely manner if clinicians unaware of the need for another sample

Related hospital procedures and documents Trust sample labelling policy. Trust policy for allocation and maintenance of unique patient identifiers and for resiting wristbands in theatre or intensive care.

284

Blood transfusion in hospitals Table 25.5 Collection and delivery of blood and blood components from transfusion issue refrigerator to clinical care area.

Responsibility

Action

Documentation

Potential errors

Staff authorized and trained

Take documentation bearing patient identification to the issue refrigerator Check that unit removed and accompanying blood transfusion compatibility form bear the identical patient identification details Record time and sign that correct unit has been collected

Prescription chart or a completed collection slip

Incorrect unit collected if no documentation bearing patient ID Incorrect unit removed

Lack of audit trail from failure to sign out unit from issue refrigerator

Related hospital procedures and documents Trust blood collection policy.

Table 25.6 Administration of blood and blood components.

Responsibility

Action

Documentation

Potential errors

Doctor or registered nurse

At the bedside, direct questioning of patient to provide surname, first name and date of birth when judged capable. Check that this identity is identical with documents

Prescription chart Case notes Compatibility form Compatibility label Patient’s wristband

Unit transfused to wrong patient if unit checked away from bedside or no verification of patient identity

Check that blood group is compatible

Case notes Compatibility form Compatibility label Base label on blood pack

Incorrect ABO/D group transfused if failure to detect laboratory grouping or labelling error

Check that special requirements are fulfilled

Prescription chart Blood pack

Inappropriate component transfused if failure to note laboratory issuing error

Check that unit of blood has of not passed its expiry date, and it is intact with no evidence visual discoloration Document date and time of commencement of unit and sign

Failure to note transfusion of timeexpired component Failure to note unit potentially contaminated with bacteria Compatibility form and/or prescription chart

Related hospital procedures and documents Trust blood administration policy.

request form, sample tube and into the current medical notes, and the unique number can also be printed onto the compatibility labels and compatibility report form. At the time of administration, the additional unique number provides a supplementary means of cross-checking.

Barcoded patient administration systems

The use of hand-held computers and portable barcode label and wristband printers provide the means for improving patient identity and patient safety. At the stage of sample collection, the 285

Chapter 25 Table 25.7 Monitoring of transfused patients.

Responsibility

Action

Documentation

Potential errors

Staff authorized and trained

Measure temperature, pulse and blood pressure before the start of each unit Explain to the patient possible adverse effects to be reported and keep patient under close visual observation in first 15–20 min of each unit Measure temperature and pulse 15 min after start of each unit Measure temperature, pulse and blood pressure at the end of each unit

Observation chart, recording date and time

In absence baseline observations, cannot detect any change giving a warning of transfusion reaction Patient not aware of symptoms to be reported that can provide first warning of a transfusion reaction In absence early observation, potential to miss a serious transfusion reaction In absence of timed final observation, cannot know whether any subsequent changes in patient’s condition are temporarily related to an ongoing transfusion

Observation chart, recording time Observation chart, recording time

Related hospital procedures and documents Trust policy on monitoring transfused patients. Management of transfusion reactions.

phlebotomist’s and patient’s identity can be scanned and a barcoded label generated at the bedside to attach to the tube. In the laboratory, the allocated unit is labelled to incorporate the patient’s unique identification barcode and the unit number. At the time of administration, the member of staff is prompted by the hand-held computer to scan their own identification barcode, the barcoded patient wristband, the compatibility label and the unit number. The computer confirms whether the unit is correct for the patient and also provides prompts to check for special requirements, pretransfusion observations and the expiry date of the pack.

Influencing clinical practice There are several potential factors that influence transfusion medicine practice and decision-making: • physician knowledge; • physician perception based upon clinical experience; • peer pressure and feedback; • effectiveness of the hospital governance framework; 286

• educational prompts at the time of the decision; • financial pressures or incentives; and • public and political perceptions and the fear of litigation. Improving transfusion medicine practice within a hospital community requires a planned consistent approach which is endorsed and implemented through clinical governance. Guidelines, algorithms and protocols

Guidelines are defined as systematically developed statements to assist practitioner and patient decisions about appropriate healthcare for specific clinical circumstances. Controlled data are unavailable to assess the impact of professional guidelines, but most would agree that nationally derived documents rarely lead to change unless there is a local implementation and dissemination strategy, which requires time and resources. Developing a local strategy to implement the guidelines is a useful opportunity to gain ownership, in that it can provide educational opportunities in examining the evidence basis, and identify dissension and other local barriers to its implementation, e.g. staff resources, laboratory turnaround times.

Blood transfusion in hospitals

Local groups should adopt the recommendations of pre-existing guidelines but customize them for local use. This may involve separating a guideline into several sections or incorporating some of its recommendations into other local protocols for specific conditions, e.g. a fresh frozen plasma (FFP) guideline incorporated into protocol for the management of disseminated intravascular coagulation (DIC), massive haemorrhage and obstetric haemorrhage. Experience in other medical fields has also demonstrated that embedding the recommendations of a guideline into documents in use at the time of the clinical consultation/decision can significantly improve compliance. Examples of this approach could include: • listing the indications for specialist blood components on the blood transfusion request forms or electronic request screens; • listing the nursing actions to be taken in the event of a transfusion reaction on a specific transfusion observation chart; and • detailing the checks to be made prior to administering blood on the compatibility form. Intraoperative algorithms for the use of platelet concentrates and FFP to correct microvascular bleeding during and after cardiac bypass surgery have also proved to be successful in reducing inappropriate use of these components, when combined with near-patient testing and the rapid availability of results to feed into the decision tree. The local documents should be disseminated alongside training events for all involved staff. A list of relevant guidelines is included at the end of this chapter. Audit

The audit cycle consists of defining the area to be studied and comparing observed practice with a standard. Analysis of the findings should lead to recommendations for improved practice, which may include a revision of the content and clarity of the standard (Fig. 25.1). The audit process has been criticized since it has been said to consume considerable resources and result, at best, in only a transient change in clinical behaviour. However, audits frequently fail because

What are we aiming to do? Standard and subsequent review

Have we made things better? Further comparison

Doing something to make things better Corrective action

Are we doing it? Audit project; comparison of activity with standard

Why are we not doing it? Analysis of causes of non-compliance

Fig. 25.1 The audit cycle.

the cycle is not completed, i.e. when the results and educational messages are not disseminated and discussed by those whose practice could be improved, when no analysis is undertaken of the corrective actions which should be introduced to improve practice, and no resources are provided to implement the actions identified. Audits can also be made more effective when they are conducted repeatedly or on an ongoing basis. Audits can be conducted retrospectively or prospectively. If adopting a prospective approach in monitoring the appropriateness of the requests for blood components, this can be considered to be intrusive and potentially delay the delivery of patient care, but does prevent unnecessary transfusion. Immediate retrospective audit of component utilization does not prevent unnecessary transfusion but, if conducted in a timely and individual fashion, can provide effective educational feedback and a sustained change to a physician’s prescribing habit. Ongoing computerized prospective or retrospective auditing, using agreed algorithms, is the only realistic way of monitoring and providing individual clinician feedback on the inappropriate use of blood components. Regular audit cycles have been shown to 287

Chapter 25

improve compliance with the bedside procedures for administering blood. Regular audits of an MSBOS are essential if the schedule is to be kept in line with changing practice. Audits that involve several healthcare organizations are particularly effective since peer pressure is applied in comparing practice. Participating hospitals’ results can be made anonymous for all other participants. Suitable audits would include the percentage of inappropriate use of blood components, of patients with correct wristband identification and of identification checks made at the time of administering blood. Surveys

Many activities which fall under an ‘audit’ banner are not comparing practice with a standard but are monitoring or surveying practice. These activities, many of which can be quantified, often lead to the development of quality indicators or performance indicators. Trend analysis, or comparison of one organization with another, or one blood user with another, are powerful means of exerting peer pressure and influencing practice (benchmarking). Performance indicators or benchmarking can be applied, for example, to: • percentage of patient wristband errors; • percentage of mislabelled samples; • hospital blood wastage; • percentage of group O usage; • number of units crossmatched to number of units transfused (C : T) ratio; • red cells used per surgical procedure (for each surgical team); • percentage of primary arthroplasties requiring allogeneic transfusion; • percentage of patients receiving platelets after coronary artery bypass grafting; • percentage of patients with refractory anaemia having received a trial of treatment with erythropoietin. National schemes

A number of national schemes in the UK set out to monitor or assure transfusion practice. Such

288

schemes can be used to influence policy and educate users within hospitals. They are generally made anonymous to encourage full participation. National schemes include the following. • The SHOT scheme was launched in November 1996. It is a voluntary system for collecting data on serious adverse events in the transfusion of blood and blood components. It produces an annual report of its findings and recommendations. • External Quality Assurance (EQA) schemes, for example the National External Quality Assurance (NEQAS) scheme in Blood Transfusion Laboratory Practice (BTLP), provide ‘clinical’ material to laboratories on a regular basis. Laboratory results are returned to the scheme organizers for analysis and collated reports are disseminated to users. • The Blood Stocks Management Scheme is a joint venture between hospital laboratories and the National Blood Service. It collates and publishes through its website details of blood stock inventory and wastage and allows participants to compare their practice with that of comparably sized hospitals. • The Royal College of Physicians and the National Blood Service in England have established a national comparative audit initiative, which has concentrated so far on benchmarking aspects of safe transfusion practice, and will soon encompass usage of blood in certain common medical and surgical conditions. Public and political perceptions and fear of litigation

The knowledge that human immunodeficiency virus (HIV) could be transmitted by blood transfusion in 1982 led to a decline in the use of allogeneic red cells in the USA, from 12.2 million units in 1986 to 11.4 million units in 1997. This decline is even more significant if the growth and ageing of the population in the USA during this period are taken into account. Over the same period, autologous donations increased by a factor of more than 30. Individual physicians were sued in the USA if their patient contracted HIV through the blood supply and their transfusion was not clinically indicated. Concern about variant Creutzfeldt–Jakob

Blood transfusion in hospitals

disease (vCJD) being transmissible through blood led in 1998 to the Department of Health in the UK requiring that all hospitals/trusts should have HTCs, implement good transfusion practice and that they should have explored the feasibility of autologous blood transfusion. Universal leucocyte depletion of blood, when introduced in the UK in 1999 as a preventive measure for vCJD, led to a significant increase in the price of red cells, and encouraged a more judicious approach to transfusion and the use of transfusion alternatives. As a consequence, total red cell usage in the UK reached a plateau between 2000 and 2001 and has subsequently fallen by about 1% per year, despite the increase in surgical procedures performed over this period. Local investigation and feedback following critical incidents and ‘near misses’

The UK SHOT scheme defines a ‘near miss’ as any error which, if undetected, could result in the determination of a wrong blood group, or the issue, collection or administration of an incorrect, inappropriate or unsuitable component but which was recognized before transfusion took place. ‘Near-miss’ errors are common, and if systematically analysed and collated provide the opportunity to understand the potential weaknesses in the process of blood transfusion. Corrective action can then be taken to minimize the occurrence of a critical incident. Identified weaknesses include staff misconceptions or ignorance, defective or risky protocols or processes. Sample errors, most importantly those where the tube is labelled with the intended patient’s details but is subsequently found to contain blood from another patient, are the commonest detectable errors. These inevitably arise as a result of a failure to systematically and positively identify the patient at the bedside. Corrective action should involve counselling and educating the individual concerned who failed to comply with the correct procedure. However any investigation will also uncover compounding latent factors contributing to the event, which need to be collectively understood and addressed, for example:

• the practice of not positively identifying patients, since healthcare workers perceive this as denoting an inadequate knowledge of the patients under their care; • reduced junior doctors’ hours and shift patterns of all those involved in direct patient management leading to unfamiliarity with patients; and • patient ‘hot bedding’ in the UK, which frequently leads to preoperative patients having to be sampled for pretransfusion testing before case notes are made available on the wards or wristbands are applied. Sadly, exposure to avoidable patient morbidity or fatality is often the trigger for affecting the local medical and nursing community’s perception of the risks of blood transfusion and for instigating a change in procedures. Education and continuing professional development

Education of all individuals in the transfusion process is difficult to achieve unless it is made an integral part of a hospital/Trust documented mandatory training programme and the process is subject to external inspection. Even then, it requires a dedicated resource and a flexible and pragmatic approach to accommodate shift patterns, staff shortages and agency staff. Education is nevertheless an essential component of every strategy to gain clinician compliance with clinical procedures and guidelines and to modify practice as a result of audit, surveys or investigations into errors. Educational interventions have been found to be more successful when they are interactive, focused on a specific objective and directed at groups of individuals with reflections on their own practice. Continuing professional development schemes exist for the various professional groups involved in healthcare. The schemes vary but all are intended to encourage knowledge acquisition. In a typical scheme, such as the one introduced by the British Blood Transfusion Society (BBTS) in April 2001, members keep a portfolio in which accredited activity in educational, professional and vocational areas is recorded.

289

Chapter 25

Maximum surgical blood ordering schedule

Table 25.8 Maximum surgical blood order schedule

(general surgery).

This is a table of elective surgical procedures that lists the number of units of blood routinely crossmatched for each. The number of units allocated takes into account the likelihood of the need for transfusion and the response time for receiving blood following an immediate spin crossmatch or electronic issue. An MSBOS reduces the workload of unnecessary crossmatching and issuing of blood, and can improve stock management and wastage. The successful implementation of an MSBOS depends upon all parties agreeing to the tariff, the education of junior staff, the confidence of senior staff that there is a robust system for accessing blood promptly when there is unexpected blood loss and the ability to override the tariff when there are reasons to indicate that greater blood loss will occur. A tariff is constructed by: • analysing each surgical procedure in terms of the C : T ratio; • managing procedures with a C : T ratio greater than 2, i.e. a low probability of transfusion, with a group and screen, and issuing blood only when there is a need for transfusion; and • allocating an agreed number of units for procedures with a C : T ratio of less than 2. In recipients with red cell alloantibodies, consideration should be given to the time taken to acquire and crossmatch antigen-negative units. An overall C : T ratio of 1.5 for elective surgery is achievable when the laboratory is centrally issuing blood. However, lower ratios would be possible with remote electronic issue in theatre suites. An example of an MSBOS is provided in Table 25.8.

Pretransfusion compatibility testing This testing comprises: • determination of the ABO and D group of the recipient; • a screen for red cell alloantibodies reactive at 37°C in the plasma of the recipient; • a check for previous records or duplicate records, and comparison of current with historical 290

Operation

Units crossmatched or group and screen (G & S)

Adrenalectomy Colectomy Cholecystectomy Gastrostomy, ileostomy, colostomy Gastrectomy (partial) Liver biopsy Mastectomy Oesophagectomy Pancreatectomy Parathyroidectomy Partial hepatectomy Splenectomy Thyroidectomy Vagotomy Bile duct stricture repair

3 2 G&S G&S G&S G&S G&S 4 4 G&S 6 2 G&S G&S 3

findings (these elements comprise a group and screen); • identification of the specificity of any alloantibody detected in the antibody screen; • selection of blood of an appropriate blood group or extended phenotype; • a serological or electronic crossmatch; • labelling of the blood with the recipient’s identifying information. Detection of red cell antigen–antibody reactions

The phenotyping (‘blood grouping’) of red cells and the detection of red cell alloantibodies depend upon interpretations of serological interactions between red cell antigens and antibodies. Various serological methods and test systems are available to demonstrate these interactions, and these must be optimized in order to obtain the appropriate sensitivity and specificity for their intended clinical use. Failure to follow the instructions provided by reagent manufacturers can lead to incorrect conclusions. Test methods have been developed to allow the detection of antibodies of different isotypes. Antibodies that have specificities for red cell antigens

Blood transfusion in hospitals

are usually IgG or IgM. IgM antibodies are pentameric molecules that can cross-link between antigens on adjacent cells, thus causing direct agglutination of red cells. Conversely, IgG antibodies are monomeric, and although they are divalent, the distance between the Fab regions on a single IgG molecule is in general insufficient to allow direct agglutination to take place, as there are stronger intercellular repulsive forces between red cells at these distances. Methods such as the indirect antiglobulin test (IAT) (which uses a secondary antibody; see Fig. 25.2) or the enzyme method (which uses proteolytic enzymes such as papain to cleave negatively charged, hydrophilic residues from the red cell membrane) must therefore be used to detect most IgG red cell antibodies. Test systems for the detection of serological reactions can be classified into three broad categories. Liquid-phase systems

Liquid-phase systems rely on the visualization of haemagglutination reactions in individual glass/plastic tubes or 96-well microplates. The presence or absence of agglutinated red cells distin-

guishes positive and negative reactions, allowing the grading of reaction avidity according to the strength of haemagglutination. These liquid-phase systems are more commonly used for blood grouping than for antibody screening, as IgM blood grouping reagents can be used in simple, rapid, direct agglutination methods. IAT methods using red cells suspended in a low ionic strength solution remain the gold standard for the detection of clinically significant red cell alloantibodies, although these methods require meticulous attention to procedure, in particular during the washing phase to remove unbound IgG. Column-agglutination systems

The introduction of column-agglutination systems during the last decade has resulted in a very significant change to routine laboratory practice in the UK. One of these systems, first described by Lapierre in 1990, uses a plastic card containing six channels, each of which contains a mixture of Sephadex and Sephacryl gels. This gel mixture is formulated to allow the passage of unagglutinated cells but not of agglutinated red cells. Positive reactions are therefore distinguished by agglutinates at

Red cells incubated with serum containing IgG antibody

Unbound antibody washed away. Anti-human globulin added to precipitate cells

Fig. 25.2 Indirect antiglobulin test.

291

Chapter 25

or near the top of the gel column and negative reactions appear as buttons of red cells at the bottom (Fig. 25.3). A similar column-agglutination method involving a glass microbead density barrier in place of a gel is also available. Reagent (IgM) antibody can be incorporated into the gel or bead columns, allowing phenotyping to be undertaken simply by the addition of test cells to the top of the column. Similarly, the IAT can be performed in columns containing antiglobulin reagent to which plasma and reagent red cells are added. Because plasma proteins are less dense than the gel, a washing phase is not needed. This property, and the relative stability of the reaction end point, give column agglutination methods a degree of simplicity and reliability not achieved by other methods. Solid-phase sytems

These systems are based on 96-well microplates and provide another alternative to the tube or column IAT. Although differing in detail, all these methods achieve a positive reaction end point that is characterized by a monolayer of red cells across the surface of the microplate well. A discrete button of red cells at the bottom of the well indicates a negative reaction (Fig. 25.4). Solid-phase systems suffer from the disadvantage that they require carefully standardized centrifugation and

washing steps; however, unlike the situation with liquid-phase test systems, fully automated equipment allows these steps to be performed safely and consistently without operator intervention. Reduction of error in pretransfusion compatibility testing

An analysis of cases in the UK during 1996–2002 where blood components had been incorrectly transfused showed that laboratory errors were implicated in 28% of incidents. Many of these laboratory errors were due to the transposition of samples or to simple human error in the setting up or interpretation of tests. Provided that the correct laboratory identifier (as a barcode) is placed on the patient’s sample, these errors can be avoided by using a fully automated system that is interfaced to the blood transfusion laboratory computer. The basic features of a fully automated (‘walk away’) system should include: • trays or carousels to stack samples; • automated liquid handling and other robotic operations; • devices to ensure that positive sample identification is maintained; • clot sensor and liquid level alarms; • an optical device to record reaction patterns; and • a comprehensive system management software

Fig. 25.3 Column-agglutination

technology. Positive results are seen in the first and last columns.

292

Blood transfusion in hospitals

Fig. 25.4 Solid-phase technology.

that interprets reaction patterns and flags discrepant results. Automated systems can utilize solid-phase/microplate or column-agglutination technology. Where full automation is not available, or cannot be used for the whole process (as is currently the case with antibody identification), steps must be taken to minimize the occurrence of error or minimize its impact. A high standard of training, participation in internal and external quality assessment schemes, and strict adherence to validated documented procedures are among the measures that reduce the occurrence of human error. The most critically important procedure in the blood transfusion laboratory, the determination of the ABO group, should be performed by two people (except in urgent situations) if there is no record of a grouping result from a previous sample. Similarly, the determination of a D group should be performed in duplicate, in the absence of full automation. ABO and D grouping

• The patient’s red cells should be tested against monoclonal anti-A and anti-B grouping reagents. • The patient’s serum/plasma should be tested against A1 and B reagent red cells, except in neonates. • The expected reaction patterns in ABO grouping are illustrated in Table 25.9.

Table 25.9 ABO grouping patterns.

Group

Anti-A

Anti-B

A1 cells

B cells

O cells

A B O AB

+++ +++

+++ +++

+++ +++ -

+++ +++ -

-

• The patient’s red cells should be tested with an IgM monoclonal anti-D reagent, which does not detect the ‘partial D’ group, DVI. • ABO and D groups must be repeated when a discrepancy (anomaly) is found. Repeats should be performed using a fresh suspension of washed cells. An autocontrol should be included. Antibody screening

• The IAT performed at 37°C is the best method available for the detection of red cell antibodies of clinical importance. It is simple (especially when using a column-agglutination system), has an appropriate level of sensitivity and has a high degree of specificity. • The patient’s serum/plasma should be tested against two or more ‘screening cells’ using the IAT. • The reagent red cells used for screening should between them express antigens reactive with all clinically significant antibodies; ideally the 293

Chapter 25

phenotypes R1R1 or R1wR1 and R2R2 should be represented in the screening cell set. It is recommended that the following phenotypes should also be represented: Jk(a+b–), Jk(a–b+), S+s–, S–s+, Fy(a+b–), Fy(a–b+). • Antibody screening is the most reliable and sensitive method of detecting a clinically significant antibody, since stronger reactions may be obtained with cells having homozygous expression of the antigen and the red cells are preserved in a medium to minimize loss of antigens during the storage period. • Antibody screening performed in advance of the requirement for transfusion also provides the laboratory with time to identify the specificity of the antibody and, when clinically significant, to select antigen-negative units for crossmatching. Antibody identification

• When an antibody has been detected in the screening test, the specificity should be determined by testing the patient’s serum/plasma against a panel of reagent red cells of known phenotypes. • In addition to the IAT, other methods (e.g. using enzyme-treated red cells) may be helpful, particularly when a mixture of antibodies is present. • The specificity of the antibody can be determined when the serum/plasma is reactive with at least two examples of red cells bearing the antigen and non-reactive with at least two examples of red cells lacking the antigen. • When one antibody specificity has been determined, it is essential that additional clinically significant antibodies are not overlooked. Multiple antibodies can only be confirmed by choosing red cells that are antigen negative for the recognized specificity but positive for other antigens to which clinically significant antibodies may arise. Autoantibodies

These may be suspected when the patient’s serum/plasma reacts with all cells used in the reverse ABO group or with all cells in the antibody identification panel including the patient’s own red cells. Not all autoantibodies give rise to haemolysis. 294

Serological investigations should focus on obtaining the correct ABO and RhD group of the patient and on excluding the presence of underlying alloantibodies. Cold-type autoimmune haemolytic anaemia

• The red cells should be washed at 37°C for performing the direct antiglobulin test (DAT), which will usually be strongly positive due to coating with C3d. • Underlying alloantibodies can be excluded by using cells and serum separately warmed to 37°C. Warm-type autoimmune haemolytic anaemia

• The red cells will usually have a positive DAT due to coating with IgG with or without complement. Rarely, the red cells may be coated with IgA or IgM and IgG. • Underlying alloantibodies may be detected following the removal of autoantibodies from the patient’s serum. This may be achieved by either absorbing the serum with the patient’s own red cells (e.g. using a combination of papain and dithiothreitol ‘ZZAP’ to elute the autoantibody and enzyme treat the red cells) or, if the patient has been recently transfused, with red cells of similar phenotype (if already known) or with two or more examples of red cells of known phenotypes. Selection of red cells for transfusion

• Red cells of the same ABO and D group as the patient should be selected, except in a lifethreatening situation before the patient has been grouped. In this case, group O should be used; if the patient is a premenopausal female, group O, D negative should be used. Group-specific units should be provided as soon as the patient’s group is known. • The selection of blood for patients with red cell alloantibodies is summarized in Table 25.10. However, in life-threatening situations, the immediate need for red cell transfusion may necessitate the use of incompatible units. • Units for fetal or neonatal exchange transfu-

Blood transfusion in hospitals Table 25.10 Recommendations for the selection of blood for patients with red cell alloantibodies.

Typical examples

Procedure

Antibodies that could be considered clinically significant

Anti-D, -C, -c, -E, -e Anti-K, -k Anti-Jka, -Jkb Anti-S, -s, -U Anti-Fya, -Fyb

Select ABO-compatible, antigen-negative blood for serological crossmatching

Antibodies directed against antigens with an incidence of < 10%, and where the antibody is often not clinically significant

Anti-Cw Anti-Kpa Anti-Lua Anti-Wra (anti-Di3)

Select ABO-compatible blood for serological crossmatching

Antibodies primarily reactive below 37°C, and never or only very rarely clinically significant

Anti-A1 Anti-N Anti-P1 Anti-Lea, -Leb, -Lea+b Anti-HI (in A1 and A1B patients)

Select ABO-compatible blood for serological crossmatching, performed strictly at 37°C

Antibodies sometimes reactive at 37°C and clinically significant

Anti-M

If reactive at 37°C, select ABO-compatible, antigen- negative blood for serological crossmatching If unreactive at 37°C, select ABO-compatible blood for serological crossmatching, performed strictly at 37°C

Other antibodies active by IAT at 37°C

Many specificities

Seek advice from blood centre

IAT, indirect antiglobulin test.

sions should be selected to be compatible with the maternal serum/plasma. • Premenopausal females should ideally receive K-negative red cells, and R1R1 units if they are c-negative. • Patients with a lifelong dependency on red cell support should receive red cells matched for Rh antigens and K (see Chapter 9). • Recipients of ABO- and D-incompatible allogeneic haemopoietic stem cell grafts will need to be transfused with red cells of the donor group in the case of a minor ABO mismatch, or group O in the case of a combined ABO mismatch. D-negative red cells should also be selected for D-positive recipients of a D-negative donation (see Chapter 9). Serological crossmatch

Crossmatching techniques have been simplified in recent years, and only the immediate spin crossmatch and the IAT crossmatch remain in common use.

The IAT crossmatch can be abolished when antibody screening is performed with screening cells that share apparent homozygous expression of common antigens capable of stimulating clinically significant antibodies, and the patient’s serum/plasma has never been found to contain clinically significant antibodies. An exception to this rule should be made for patients who have received an ABO-incompatible solid organ transplant and who may develop an IgG anti-A or antiB produced by passenger lymphocytes. Several retrospective and prospective studies have shown that there is negligible risk in omitting the IAT crossmatch. Although up to 0.2% IAT crossmatches may reveal an unpredicted incompatibility, few of these transfusions result in haemolysis. Antibodies directed against low-frequency antigens may be missed but the majority of these are naturally occurring and do not cause patient morbidity. If the IAT crossmatch is omitted, there must be some check included to detect ABO incompati295

Chapter 25

bility. The immediate spin crossmatch (i.e. agglutination in saline following a 2–5 min incubation) is a serological check that can be used. However, this technique is fallible when the patient has low levels of anti-A or anti-B and, unless ethylenediamine tetra-acetic acid (EDTA) saline is used, falsenegative results may also arise as a result of steric hindering of agglutination by C1. False-positive crossmatches arising from rouleaux or cold agglutinins not detected in the antibody screen have the potential to delay the issuing of compatible units. The limitations of the immediate spin crossmatch have heralded the acceptance of electronic issue as an alternative method of preventing the release of ABO-incompatible units of blood. Electronic issue

Electronic issue should only be used for detecting ABO incompatibility between the donor unit and the patient sample that was submitted for pretransfusion testing. There are several essential requirements for adopting this approach, which are common to the various professional standards. • The computer contains logic to prevent the assignment and release of ABO-incompatible blood. • No clinically significant antibodies are detected in the recipient’s serum and there is no record of previous detection of such antibodies. • There are concordant results of at least two determinations of the recipient’s ABO and D groups on file, one of which is from a current sample. • Critical elements of the system (application software, readers and interfaces) have been validated on-site and there are mechanisms to verify the correct entry of data prior to release of blood, such as the use of barcode identifiers to enter information when it cannot be automatically transferred. Fully automated blood grouping and antibody screening, although not a requirement in current UK guidelines, is strongly recommended. Electronic issue has been routinely practised in Sweden for over 10 years, during which time one error has been noted due to an incorrectly labelled unit of blood, which was supplied by a small noncomputerized blood centre. 296

Electronic issue has several potential advantages: • reduced technical workload; • rapid availability of blood; • improved blood stock management and reduced wastage; • less handling of biohazardous material; • elimination of unwanted false-positive results in the immediate spin crossmatch; and • ability to issue blood electronically at remote sites, using trained non-laboratory staff. This last characteristic has allowed the development of networked electronic blood release systems. When the patient details are entered, the system checks that the criteria for electronic issue are fulfilled and lists the compatible units available in the remote site blood refrigerator. The barcodes of the unit selected are scanned into the computer and, if ABO and D compatible, a compatibility label is printed that is attached to the unit and rescanned. This step generates a second label, or compatibility form, which is signed by the clinical staff at the time of the transfusion.

Summary The transfusion process is unique for several reasons: • it links one sector of the community (the donors) with another (the patient) in an altruistic, potentially life-saving activity; • it links many grades of staff across a healthcare organization; and • for many patients, there is still no substitute for human-derived blood components. Prescribers of blood components have a moral obligation to the donors to ensure that the donations are used appropriately. Prescribers of blood components also have a duty of care to their patients to ensure that the benefits of the transfusion outweigh the risks. There are many different members of staff involved in the hospital transfusion process, which provides too many opportunities for human error to prevail. Investment in a quality infrastructure, computerization and automation is essential to prevent errors in this process.

Blood transfusion in hospitals

Further reading Audet AM, Greenfield S, Field M. Medical practice guidelines: current activities and future directions. Ann Intern Med 1990; 113: 709–14. Department of Health. Better Blood Transfusion. HSC 2002/009. London: HMSO, 2002. Dzik WH, Corwin H, Goodnough LT et al. Patient safety and blood transfusion: new solutions. Transfus Med Rev 2003; 17: 169–80. Eisenstaedt RS. Modifying physicians’ transfusion practice. Transfus Med Rev 1997; 11: 27–37. Handbook of Transfusion Medicine, 3rd edn. London: The Stationary Office, 2001 (www.transfusionguidelines.org). Heddle NM, O’Hoski P, Singer J, McBride JA, Ali MA, Kelton JG. A prospective study to determine the safety of omitting the antiglobulin crossmatch from pretransfusion testing. Br J Haematol 1992; 81: 579–84. Jensen NJ, Crosson JT. An automated system for bedside verification of the match between patient identification and blood unit identification. Transfusion 1996; 36: 216–21. Judd WJ. Requirements for the electronic crossmatch. Vox Sang 1998; 74 (Suppl. 2): 409–17. Mollison PL, Engelfriet CP, Contreras M. Detection of the

reaction between red cell antigens and antibodies. In: Blood Transfusion in Clinical Medicine. Oxford: Blackwell Science, 1997: 242–76. Research Unit of the Royal College of Physicians. Audit Measures for Good Practice in Blood Transfusion Medicine. London: RCP Publications, 1995. Serious Hazards of Transfusion. Annual Report 2002–2003. Manchester: SHOT Office, 2004 (www.shotuk.org). Turner CL, Casbard AC, Murphy MF. Barcode technology: its role in increasing the safety of blood transfusion. Transfusion 2003; 43: 1200–9.

Guidelines British Committee for Standards in Haematology. Guidelines for compatibility procedures in blood transfusion laboratories. Transfus Med 2004; 14: 59–73. British Committee for Standards in Haematology. The administration of blood and blood components and the management of transfused patients. Transfus Med 1999; 9: 227–38. British Committee for Standards in Haematology. Guidelines for blood bank computing. Transfus Med 2000; 10: 307–14.

297

Chapter 26

Autologous transfusion Dafydd Thomas

Autologous transfusion is the collection of the patient’s own blood for reinfusion. This can be by prior collection either in the weeks before or immediately prior to surgery. Alternatively the blood loss during or after surgery can be collected and reinfused. These methods can be summarised as: • intraoperative cell salvage (ICS); • postoperative cell salvage (PCS), which may be washed or unwashed on reinfusion; • predeposit autologous donation (PAD), taken and stored in the weeks before surgery; • acute normovolaemic haemodilution (ANH), taken just before surgery and returned following the operation.

Reasons to consider autologous transfusion Clinical transfusion practice includes an attempt to reduce the risks involved in blood transfusion. This may be achieved in part with extensive testing of the blood to decrease the risk of transfusiontransmitted infection. Improvements in the safety of blood over the last 20 years (Fig. 26.1) has ensured there continues to be considerable demand for allogeneic blood. Blood conservation strategies should be adopted to minimize the use of allogeneic blood by withholding transfusion until strictly clinically necessary, and employing techniques such as autologous transfusion. In some situations autologous transfusion is definitely indicated, such as patients with very rare blood groups or complex red cell antibodies for whom it is difficult to find compatible blood. Autologous transfusion should be also used instead of, or to supplement, allogeneic blood in situations where it 298

has been shown to be effective and safe. It has been suggested that more than 20% of surgical demand can be met by autologous transfusion. There are certain procedures that can be undertaken with virtually no allogeneic blood use, allowing conservation of supplies for areas of medicine where there are currently few alternatives to allogeneic blood, such as haematological oncology. The provision of comprehensive testing of donor blood and other safety measures have led to a significant increase in cost. The escalating costs, and fears over a continued adequate supply of donor blood have led to a greater interest in autologous transfusion. Considerable experience has been gained in the various methods of autologous transfusion elsewhere in the world particularly in North America. The autologous transfusion procedures which seem to offer the most cost-efficient and clinically effective method in the current situation are ICS and PCS.

Blood conservation strategies Avoidance of unnecessary blood transfusion includes optimizing the preoperative haemoglobin (Hb) concentration, minimizing blood loss, and being clear about when and how much to transfuse (see Chaper 6). Autologous transfusion options are at their most effective when combined with these other strategies (Table 26.1). Before surgery: optimizing Hb and haemostasis

This involves assessing a patient in advance of surgery and taking steps to reduce the patient’s requirements for transfusion. If a patient is

Autologous transfusion

Risk of infection per unit transfusion

1/100

HCV HBV HIV

1/1000

1/2000

1/10 000 1/58 000 to 1/149 000

1/100 000

1/872 000 to 1/1.7 ¥ 106

Nucleic acid testing for HCV/HIV

Testing for p24 antigen

Screening for HCV-antibody

Screening for HIV-antibody Surrogate screening for non-A, non-B hepatitis

of blood over the last 20 years. HBV, hepatitis B virus; HCV, hepatitis C virus; HIV, human immunodeficiency virus. (From Goodnough et al. 2003 with permission.)

1/1.4 ¥ 106 to 1/2.4 ¥ 104

19 8 19 3 8 19 4 8 19 5 8 19 6 8 19 7 8 19 8 8 19 9 9 19 0 9 19 1 9 19 2 9 19 3 9 19 4 9 19 5 9 19 6 9 19 7 9 19 8 9 20 9 0 20 0 01

Fig. 26.1 Improvements in the safety

Donor screening criteria changed

1/1 000 000

Years

Table 26.1 An approach to blood conservation and the reduction of risk associated with blood transfusion in patients having

elective surgery. Some principles apply to other situations where blood transfusion is being considered. • • • • • •

Check the blood count well in advance of surgery and correct any treatable anaemia Ask about any drugs the patient is taking and consider if any should be stopped Check antibody status and blood group so group-specific blood can be used in an emergency Consider whether patient has a hereditary or acquired bleeding tendency and investigate/treat as appropriate During surgery consider ways to reduce bleeding, e.g. the use of lasers or the use of aprotinin in cardiac surgery Postoperatively, consider whether blood transfusion is clinically indicated (transfusion trigger) and, if it is, consider how many units are required to achieve the desired Hb (transfusion target) • If operation would normally require blood transfusion, consider the option of autologous blood transfusion.The options are listed below Technique

Situations in which it might be considered

Intraoperative cell salvage

Any patient (generally without infection or malignancy in the operative field) with blood loss > 0.5 L; especially suitable for massive blood loss Patients with postoperative blood loss from a clean site Suitably fit patients expected to lose a moderate amount of blood (1–2.5 L) Suitably fit patients expected to lose more than 20% blood volume (~ 2 L)

Postoperative cell salvage Predeposit autologous transfusion Acute normovolaemic haemodilution

anaemic, it is important to consider the reasons and address the underlying cause, e.g. iron deficiency may be due to a gastrointestinal malignancy. If a patient with iron deficiency anaemia is started on iron, the Hb can be expected to rise by about 1 g/dL per week. It is therefore important to check the blood count sufficiently far in advance of surgery to allow time for treatment to be given if required. Improved preparation of patients for elective surgery may ensure that correctable anaemia is treated prior to surgery. Patients presenting with a normal or near-normal Hb will

require transfusion at a later stage or may even avoid blood transfusion altogether. Recent evidence suggests that the newer preparations of intravenous iron can be given with fewer adverse reactions than were associated with these preparations in the past. Intravenous iron is almost immediately available for red cell production. Research is beining undertaken to determine whether administration of intravenous iron as late as the preoperative day can improve red cell production in response to surgical anaemia and thus decrease the use of donor blood. 299

Chapter 26

It is also important to consider if there are any patient factors which might cause excessive blood loss during surgery and which can be corrected in advance. The following list outlines some of the preoperative measures that may need attention. • Drugs: if patients are on aspirin or clopidogrel, it is usually advisable to stop them 7 days before surgery (except if the patient is at high risk of suffering a myocardial infarction). Patients on warfarin can often stop a few days before surgery. If it is imperative to continue anticoagulation, the patient may be changed to heparin to cover the surgical period. • Inherited disorders of coagulation: it is important to take a bleeding history when the patient is being seen prior to surgery. Screening tests and specific treatment may be required (see Chapter 9). • Bleeding diathesis: important to consider in patients with renal or liver disease. Agents such as desmopressin (DDAVP) or tranexamic acid may decrease the bleeding tendency temporarily (see Chapter 7). Expert advice from a haematologist should be sought. During surgery: reduction in blood loss

With advancing surgical and anaesthetic techniques, blood loss in many operations has fallen significantly. Use of lasers and microsurgical techniques have had a huge impact on blood usage. • Maintenance of normothermia allows optimum coagulation perioperatively and has been shown to decrease blood loss. • Cardiac surgery: patients on bypass are at particular risk of bleeding. Aprotinin, an inhibitor of fibrinolysis, has been shown to reduce blood loss, particularly in high-risk operations, and does not appear to increase the reinfarction rate (see Chapter 7). During/after surgery: when to transfuse

No blood transfusion is without risks, but equally the administration of blood may be life-saving. In making the decision to transfuse the balance of risks must be considered for each individual patient. Factors influencing the decision to transfuse include: 300

• Hb concentration; • the patient’s life expectancy, i.e. age/prognosis (many of the adverse effects of transfusion are delayed, e.g. transfusion-transmitted infections often have a long latent period before the patient becomes symptomatic); and • clinical judgement about the patient’s ability to tolerate anaemia including the presence of other factors such as cardiac and respiratory disease and sepsis.

Transfusion triggers Data from patients who refuse blood on religious grounds or who live in parts of the world where blood is scarce or dangerous have helped our understanding of the effects of different levels of Hb. In otherwise healthy patients the following transfusion triggers for stable anaemia might be considered: • 10 g/dL: transfusion rarely required. A study of patients in intensive care showed that less severely ill patients (Acute Physiology and Chronic Health Evaluation II score £20) and patients under 55 years actually had a survival advantage if the Hb was maintained between 7 and 9 g/dL rather than between 10 and 12 g/dL. For patients with clinically significant cardiac disease the mortality was similar in both groups. For otherwise fit patients with a previously normal Hb who are actively bleeding the following guidelines might be applied. • Blood loss 40% blood volume: transfusion indicated. Note that blood volume is about 70 mL/kg in adults, and that 20% of blood volume is approximately 1 L. For patients with a short life expectancy or those

Autologous transfusion

with chronic anaemia and impaired red cell production, the main trigger for transfusion should be the patient’s symptoms.

Transfusion targets In addition to considering when to transfuse, a target Hb should be established for each clinical scenario using the best data available (see above and Chapter 6), and it is also important to consider how many units to give. When a patient is actively bleeding, replacement of red cells should be guided by an estimate of blood loss. A guide to how many units are required to achieve the target is shown in Table 26.2. Oneunit transfusions have previously been discouraged. However, Table 26.2 shows that it might be reasonable to give a single unit to a small elderly woman who is symptomatic with an Hb of 7 g/dL to bring it up to just under 9 g/dL. The transfusion of blood just because it has been made available for the patient should be avoided. If blood is not used, it can be returned to the blood bank and used for another patient.

tion of a single unit of red cells may be enough to raise a patient’s Hb to avoid donor blood transfusion. After much debate, red cell salvage now seems to offer the most cost-effective method of autologous transfusion. Future issues in blood supply and demand combined with the discovery of other blood-borne diseases may change this view. A full description of all the autologous methods has therefore been retained in this chapter. All autologous blood must be clearly labelled and be distinct from allogeneic blood. An example of an autologous blood label is shown in Fig. 26.2; they are more easily identified if printed a different colour to the ones used for allogeneic blood. Perioperative cell salvage Principle

During surgical operations when blood loss is expected, blood can be collected, processed and then returned to the patient. This can be done SWANSEA NHS TRUST AUTOLOGOUS TRANSFUSION HOSPITAL NO.

Techniques for providing autologous blood The measures for avoidance of blood transfusion described above, when combined with some form of autologous technique, can maintain the Hb at a clinically acceptable level, sufficient to avoid the transfusion of allogeneic blood. Even the collec-

Table 26.2 Guide to number of units required to achieve the

NAME

DOB OPERATOR

SIGNATURE

DATE

TIME

EXPIRY DATE/TIME

‘target’ haemoglobin (Hb).

Amount of Hb in 1 unit of red cells Example: volume bled 450 mL ¥ average Hb concentration 13 g/dL = 58 g/unit

HOSPITAL NO.

Weight (kg) 43 57

71

3L 1.9

5L 1.2

AFFIX IN TRANSFUSION RECORD DOB

NAME LD. CHECKED BY

Blood volume (70 mL/kg in adults) Increase in Hb after one unit transfusion (g/dL)

4L 1.6

ANH

WITNESS PHE DONATION

WASHED

DATE UNWASHED

Fig. 26.2 Autologous colour-coded labels.

301

Chapter 26

either intraoperatively or postoperatively depending on the type of operation. This process can be cost-effective even when even small volumes of blood (i.e. more than 500 mL) are collected. The amount salvaged not only decreases the use of allogeneic blood but in many instances completely removes the need for allogeneic blood transfusion. Intraoperative cell salvage

ICS involves the collection and reinfusion of red cells lost during surgery. This may be performed as follows. • Single-unit reinfusion devices (contain anticoagulant, unless defibrinated blood is being collected): these are simple and cheap for low-volume losses. • Continuous reinfusion of unprocessed blood: may be used in conjunction with cardiac bypass but is not of proven benefit and may be associated with risk of haemolysis. • Reinfusion of processed blood (discussed below).

There are a number of machines available which wash red cells by centrifugation and resuspend them in saline (an example is shown in Plate 26.1). Blood is aspirated from the wound site and mixed with anticoagulant (Fig. 26.3) before it enters the reservoir of the machine (Plate 26.2) (Plates 26.1 and 26.2, shown in colour between pp. 304 and 305). The cycle can be either run automatically or controlled manually. In general about 75% of red cells can be recovered for reinfusion back into the patient. The machines can deliver the equivalent of 10 units of blood per hour. Advantages of ICS • Considerable reduction in allogeneic blood usage in cases where blood loss is large (>2 L). Suitable operations might include open heart surgery, liver transplantation and ruptured ectopic pregnancy. • Available to all patients having appropriate surgery regardless of medical fitness. • In some situations of uncontrolled blood loss it may be life-saving.

Anticoagulant

Blood from site of operation

Reservoir Reinfusion bag

Wash solution

Pump

Waste bag

Centrifugal bowl

302

Fig. 26.3 Complete cell saver set-up.

Autologous transfusion

• Unlike other techniques, ICS can be used selectively in cases where the actual, rather than the predicted, blood loss is high. • Blood can be collected in the reservoir and the decision to use the machine and harness can be deferred until it is clear that the blood loss is sufficient to warrant processing. • Unlike PAD and ANH, in which blood is stored outside the body, cell salvage may be accepted by Jehovah’s Witnesses, provided the collected blood remains in continuity with the patient via tubing which is connected to the patient’s intravenous cannula and hence the patient’s circulation. Disadvantages/risks of ICS • The reinfusion of haemolysed blood is unlikely, providing the wash process is undertaken correctly. Currently available machines operate an automatic washing process, and a sensor monitors the effluent from the wash cycle, continuing the wash process until the liquid being discarded is completely clear, suggesting removal of all free Hb and fragmented red cells. • There have been no recent deaths associated with air embolism as a result of improved design and greater awareness of such potential problems. Deaths were associated with air embolism in the earlier machines and collected blood should not be used with pressurized reinfusion. • It does not recover all the blood lost so allogeneic blood is often required if major haemorrhage has occurred. • It requires a capital outlay and trained operators, so ICS can only be used in hospitals with suitable operations to become cost-effective. As the cost of allogeneic red cells continues to rise with the introduction of safety measures such as universal leucocyte depletion of blood and increasingly sensitive and expensive microbiology testing, cell salvage has become cost-effective for many hospitals. • It is very important to follow agreed standard operating procedures and to document all stages of the process to avoid mistakes being made. Indications Operations where expected blood loss is likely to be in excess of 500 mL. Even when blood loss is

unpredictable, the collection of operative blood loss may be worthwhile. Providing this blood is anticoagulated, it can be processed and reinfused if sufficient volumes are collected. The processing kits are separately packaged so only the collection jar is wasted if small volumes are salvaged with a decision not to wash. It is cost-effective providing 1 unit of packed red cells is produced and used. Relative contraindications There are a number of situations where the use of cell salvage has been discouraged. However, in the presence of massive haemorrhage, ICS may be considered as the possibility of hypovolaemic shock and perhaps death becomes part of the risk–benefit equation. • Potential for aspiration of malignant cells: although leucocyte filters may remove the majority of malignant cells, and small numbers may not be clinically significant compared with the numbers that enter the circulation during surgery, some would advocate the use of gamma-irradiation in this setting but this is logistically difficult to arrange. • Presence of infection: although the balance of risk depends on the clinical urgency for salvaged blood, antibiotics may be added to the anticoagulant solution used and given parenterally to the patient to minimize any bacteraemia that occurs. • Presence of ascitic or amniotic fluid in the operative field, which may cause embolism/disseminated intravascular coagulation. Filtering of the salvaged blood via a leucocyte filter removes all lamellar bodies and decreases the risk of amniotic fluid embolism. • Sickle cell disease: cells may sickle in the low oxygen tension of the machine. • Where topical clotting agents have been used, e.g. fibrin glue, or iodine has been used to wash out the abdomen: in practice these contaminants haemolyse red cells. Even if these agents are collected they are washed out during the centrifugal process because of their low molecular weight relative to the red cells, and therefore not reinfused. Postoperative cell salvage

Postoperative recovery of blood involves the 303

Chapter 26

collection of blood from surgical drains followed by reinfusion with or without processing. The blood recovered is dilute, partially haemolysed and defibrinated and may contain high levels of cytokines. If the collected wound drainage blood is simply reinfused, some centres limit the quantity reinfused. Other centres recommend that all blood is washed and resuspended in saline. This can either be done with the apheresis machines used in the main theatre suite or with the newer and more compact processing machines that wash collected blood by the patient’s bedside (Plate 26.3, shown in colour between pp. 304 and 305). Its most common application is the collection of blood after the tourniquet is removed at the end of knee surgery. In this setting it may help avoid the need for allogeneic blood in the majority of patients undergoing such procedures. Acute normovolaemic haemodilution Principle

Whole blood is taken into standard blood bags containing citrate anticoagulant, either just before or during induction of anaesthesia. The volume removed is simultaneously replaced with crystalloid or colloid so that the patient remains normovolaemic. The amount of blood withdrawn depends on the target haematocrit and can be calculated using standard formulas. By this process the Hb that remains is diluted so that if the patient then bleeds a given volume, the total quantity of Hb lost is not as great as if the concentration had been higher. At the end of the procedure, or before if necessary, the blood removed at the start is returned to the patient. ANH is only really beneficial if significant haemodilution is achieved and the blood loss is large, over 20% of the total blood volume. Safety

In the hands of experienced anaesthetists, ANH has been shown to be safe when performed on patients who are fit and closely monitored. It requires the establishment of, and adherence to,

304

standard operating procedures. As microbiology testing is not performed, the units should be treated as high risk and labelled ‘untested blood for autologous use only’. Units should remain with the patient and not be put in a blood refrigerator where they could be used for the wrong person. Any unused autologous blood should be disposed of as hazardous waste. All stages of the process should be carefully documented. As for all autologous transfusions, any serious events should be reported to the hospital transfusion committee and to the Serious Hazards of Transfusion (SHOT) scheme. Efficacy

Studies of ANH have generally been small or retrospective and no conclusive reduction in the use of donor blood has been demonstrated (see Chapter 6). Although it seems to be less effective at reducing the need for red cell transfusion, studies have not looked at other potential benefits. This blood may supply coagulation factors and platelets which have not been diluted within the intravascular compartment during surgery or been exposed to the stress response. At the UK conference ‘Autologous transfusion: 3 years on. What is new? What has happened?’ in November 1998, it was concluded that ‘randomised controlled trials are required before this technique can be widely recommended’. The few studies published on the topic since that meeting have not provided strong support for the technique. Indications

Guidelines from the British Committee for Standards in Haematology (BCSH) for perioperative haemodilution suggest than ANH can be considered when: • the likely surgical blood loss exceeds 20% of the patient’s blood volume; • preoperative Hb is more than 11 g/dL; • patients do not have severe myocardial disease ( patients >45 years should be assessed with caution).

(a)

(b)

Mix thoroughly and incubate at 22°C for 30 min

Add 50 mL FITC labelled anti-IgG or IgM

1000

64

1:

(i)

(ii)

10 1

1

Hist Region

Count

2

B % +ve 1.9 C x Ch 100

90.6 10 000 190 10 000

1

10 100 log F1-1 (FITC)

Mean X

Mean Y

PK PK PK Pos X Pos Y Cnt

27.3

7.33

21

3.29 0.262

64 48

32

6.6

B

HPCV FPCV

10 1.15 51.7 4344 0.000 106.2

64

Anti-HPA -1a (NAIT)

2:A

48

C

32 16 0 0.1

1 10 100 1000 log F1-1 (FITC)

1000

133

2.3 0.10

C

0 0.1

B

0 0.1

1000

AB serum

2:A

16

C

32

Count

Count

%

A autoa

48

(iv)

10 100 ss log

1

64

2:A

16

0.1 0.1

(iii)

Stop reaction Read in flow cytometer or wash x 2 in PBS/BSA and read microscopically

48 Count

FS log

100

Wash x 4 in PBS/BSA

Count

(c)

Mix thoroughly and incubate at 37°C for 30 min

Add 50 mL serum

50 mL platelet suspension in PBS

(v)

Plate 5.1 Indirect platelet immunofluorescence test (PIFT).

(a) Outline of assay. (b) Results of microscopic analysis of PIFT showing a strongly positive reaction. (c) Results of flow cytometric analysis of PIFT: (i) the platelet population is identified from forward/side scatter characteristics; (ii) the gated population is analysed for fluorescence intensity;

B

Anti-HPA -1a (PTP)

2:A C

32 B

16 0 0.1

1 10 100 1000 log F1-1 (FITC) (vi)

1

10

1

10

log F1-1 (FITC)

(iii) results may be expressed as percentage positive cells within a region (Region B, ‘% + ve’) or as mean channel fluorescence (Region C, ‘x Ch’); (iv)–(vi) plots of fluorescence intensity versus number of events for (iv) a negative sample, (v) a sample containing a weak anti-HPA1a and (vi) a potent anti-HPA-1a.

1

Plate 8.1 Acid elution technique (Kleihauer test) for

haemoglobin F containing cells; the blood specimen was taken from a postpartum woman and shows that a fetomaternal haemorrhage had occurred. A single stained fetal cell is seen against a background of ghosts of maternal cells. From Bain (1995).

Plate 20.2 Section through the brain of a patient with variant CJD with immunohistochemical staining for PrP demonstrating abnormal accumulation of PrPSc throughout the brain. (Reproduced with the permission of Professor James Ironside.)

2

Plate 20.1 Section through the brain of a patient with variant CJD demonstrating spongiform degeneration of neuronal tissue and a florid amyloid plaque (centre). (Reproduced with the permission of Professor James Ironside.)

Plate 20.3 Section through the lymphoid tissue of a patient with variant CJD with immunohistochemical staining for PrP demonstrating abnormal accumulation of PrPSc in follicular dendritic cells. (Reproduced with the permission of Professor James Ironside.)

Plate 26.1 A centrifuge bowl within an apheresis machine

Plate 26.2 Collection reservoir which may be used either

showing the dense red cell layer towards the outside of the bowl and separation from the buffy layer and plasma.

operatively or postoperatively to collect the spilt blood or wound drainage.

Plate 26.3 Equipment is now available that can wash

salvaged blood in the ward environment.

3

R1

R2

a)

Region Statistics (a and b) (Gate: No Gate) Region Events % Gated R1 27491 96.46 R2 27674 97.10

b) Quadrant Statistics (c) (Gate: G2) Quad Events % Gated UL 10 0.04 UR 434 1.57 LL 561 2.03 LR 26669 96.37

c)

d)

Plate 32.1 Haemopoietic stem cell (CD34 cell) +

enumeration in a mobilized apheresis sample using International Society for Hematotherapy and Graft Engineering guidelines. Fluorochrome-conjugated antiCD45 antibodies (fluorescein isothiocyanate, FITC) and anti-CD34 antibodies (phycoerythrin, PE) are used in this flow cytometric assay. Plot 1 (a) shows the initial gating of all cells with exclusion of visual debris (platelets, cell fragments), or R1. Plot 2 (b) sets CD45-gating (R2) to include only leucocytes for further analysis. Plot 3 (c) shows gating on cells from R2 with addition of anti-CD34-PE. The

4

Quadrant Statistics (d) (Gate: G2) Quad Events % Gated UL 354 1.28 UR 65 0.23 LL 15101 54.57 LR 12154 43.92 arrow indicates cells of interest in the upper right (UR) quadrant (CD45+/CD34+ cells = 1.57% of population). This plot serves as quality control in support of the final plot. Plot 4 (d) showing side scatter versus CD34 positivity with gating on R2 is the final plot, and CD34+ cell enumeration is reported as 1.28%; arrow at upper left (UL) quadrant points towards cells exhibiting low side scatter and CD34 positivity. (Courtesy of Dr W. Jaszcz and M. KraftWeisjahn, Fairview-University Medical Center and University of Minnesota.)

Autologous transfusion

Hypervolaemic haemodilution

A variation of ANH, which is not strictly autologous transfusion, is known as hypervolaemic haemodilution. Rather than removing blood preoperatively and replacing it with fluid, the fluid is given without removing any blood. This serves to dilute the Hb concentration with a reduction in the total amount of Hb lost, as described above. At the end of the procedure the patient is made normovolaemic either as a result of the operative blood loss or by diuresis. This technique is not widely practised. It is not without risks and its efficacy remains to be established. Preoperative autologous donation Principle

The principle behind PAD is that units of blood are collected, usually at weekly intervals, in the 4–5 weeks before surgery, during which time the patient will make up the blood lost. Advantages

This technique has been shown to reduce exposure to allogeneic blood: in one series of 116 adolescent patients undergoing spinal surgery, it was reduced from 60 to 11%. Disadvantages Increased risk of transfusion While PAD reduces the number of allogeneic units transfused, it may in fact increase the likelihood of being transfused, with the inherent risk this poses. This is because by taking several units of blood preoperatively, the preoperative Hb is often lower than it would otherwise have been. Autologous blood may also be perceived as safer and therefore be given more readily, but any transfusion carries risks. Predeposited autologous blood is just as likely as allogeneic to be associated with administrative errors, fluid overload or bacterial sepsis. Allergic or febrile reactions have also been reported. The effect of non-leucocyte-depleted blood whether allogeneic or autologous may still

suffer the storage lesion and result in immunomodulation in the recipient. Risk of donation The other additional source of morbidity/mortality is in the donation process itself. One North American study looked at more than 4 million whole-blood donations over a 10-month period from July 1993 and found that the risk of hospital admission following PAD was 1 in 16 783 (more than 10 times the rate for homologous donors). The reason was most commonly a severe vasovagal reaction but angina or trauma due to the venepuncture also occurred. Wastage One of the biggest problems associated with PAD is predicting how much blood an individual patient will lose during surgery and hence what the blood requirements will be. Even if autologous blood is only collected in those who would normally be crossmatched according to a maximum surgical blood-ordering schedule (MSBOS) (see Chapter 25), 30–50% will be wasted. This means that for more than one-third of patients, the resources used and the time, travelling and discomfort involved will have been wasted. Expense With the high wastage rates, PAD is an expensive option. In the USA, Medicare and some private insurers will not reimburse the costs. Looking at cost-effectiveness models from the USA, the cost of PAD per quality-adjusted life-year ranges from $230 000 in a total hip replacement to over $23 million in a transurethral resection of the prostate. These compare with less than $50 000 for most commonly accepted medical interventions. The advantages and disadvantages of PAD are summarized in Table 26.3. Practicalities Patient selection In order to minimize the risks of donation and reduce the wastage of autologous blood, guidelines have been drawn up for the selection of

305

Chapter 26 Table 26.3 Advantages and disadvantages of predeposit

autologous blood donation. Advantages Reduced exposure to allogeneic blood reducing the risks of: Transfusion-transmitted infection Transfusion-associated graft-versus-host disease Immunization to red cell/platelet and HLA antigens Provides compatible blood for patients with complex red cell antibodies or antibodies to common red cell antigens Supplements the blood supply but uses more resources and staff time to collect per unit transfused than homologous blood Disadvantages Reduces preoperative Hb, thus increasing the risk of receiving a blood transfusion Does not abolish the risks of: Bacterial contamination ABO-mismatched blood being given as a result of administrative errors Some febrile reactions Fluid overload Is associated with risks due to the donation of several units of blood 30–50% of units are wasted, exposing many patients to unnecessary morbidity and the healthcare system to unnecessary expense

• In children under 8 years old or weighing less than 25 kg, PAD is technically difficult and rarely justified; 8–16 year olds can be considered if the child is willing and the parents or guardian gives consent. The guidelines recommend that donations should be collected in a hospital in close collaboration with a paediatrician. • Exercise caution with patients under 50 kg: may need small-volume packs (250-mL packs are available) to avoid taking more than 12% of blood volume. Other requirements • The patient must have suitable venous access. • The operation would normally require blood to be crossmatched (according to MSBOS). • The operation date is fixed. • There is sufficient time to donate the required number of units: the last unit should be taken 7–10 days before the planned surgery. The minimum safe interval before surgery is 72 h, which allows the blood volume to be replenished. The process

suitable patients (see the BCSH ‘Guidelines for autologous transfusion’, part 1). In summary the medical exclusion criteria for preoperative donation are as follows. • Active bacterial infection. • Positive results for human immunodeficiency virus (HIV) or hepatitis C virus (HCV). • Epilepsy. • Prolonged or frequent angina, left main stem disease, significant aortic stenosis or cyanotic heart disease. Other patients with cardiac disease can be considered subject to assessment by a cardiologist. • Caution with patients on b-blockers or angiotensin-converting enzyme inhibitors: consider isovolaemic replacement with crystalloids. • Uncontrolled hypertension. • Pregnancy, especially if impaired placental flow or intrauterine growth retardation, due to possible harm to the fetus. • Patients with a previous history of prolonged faint after blood donation. • Hb below 11 g/dL at the start or below 10 g/dL for subsequent donations; 306

As outlined above, the use of PAD should be confined to suitable patients in whom the risk of taking autologous blood is smaller than the risk associated with the use of allogeneic blood. The request may come from patient or clinician. In either case the patient should be given a full explanation of the risks and benefits (preferably with an information leaflet). The referring physician should be satisfied that the patient is medically fit and meets the criteria for PAD. The discussion should include the following. • The risks and benefits of PAD. • The requirement for microbiology testing for HIV, HBV and HCV (to reduce the risk if the unit was inadvertently transfused to another). • The possibility that the patient will not need a transfusion and the units will be wasted. In the UK, unlike some other countries, units are not ‘crossed over’ for other patients’ use since many autologous donors would not meet the required criteria. • The possibility that more blood will be required than has been collected so the patient may need additional, allogeneic blood.

Autologous transfusion

• The possibility that some or all of the units collected may not be usable, e.g. if the bag had a leak or the patient develops an infection soon after donation. If a patient wants to proceed, appointments are made, usually at weekly intervals. The first appointment should be less than 5 weeks before the operation date as units can only be stored for a maximum of 35 days. The last one should be at least 72 h (preferably a week) before the planned surgery to allow time for plasma and the majority of the Hb to be replenished. It has been shown that oral iron improves the yield, even in iron-replete patients, so patients are advised to take iron. Patients who are iron deficient should start this in advance of the first appointment. Erythropoietin is licensed for use in PAD but is expensive and may be associated with an increased risk of thrombosis, and is therefore not generally recommended. The use of intravenous iron is likely to increase due to the improved safety profile of the iron sucrose preparations. Whether the units are collected in a hospital or a transfusion centre, it is vital that strict standard operating procedures are drawn up and followed to ensure that any risk of administrative error is minimized. Once the units are collected they should be clearly labelled with the patient’s details and the label signed by the patient. The units are then transferred to the hospital blood bank where they should be kept securely segregated from the stocks of donor blood. Prior to surgery a further sample from the patient should be sent and the autologous units crossmatched against it as an added safety check. This sample is also required in case additional allogeneic blood is needed. It is vital that there is good communication between all the staff involved. These include the referring clinician, the transfusion centre, the hospital blood bank, junior staff requesting the blood and the doctor who prescribes it. One of the most common errors in autologous transfusion is to give allogeneic blood instead of the predeposited units, often because it was not realized that autologous units had been collected.

Blood substitutes The use of blood substitutes is discussed in Chapter 30.

Summary The place of the various forms of autologous transfusion should be considered as part of a strategy for minimizing the risk associated with transfusion for patients having surgery, where blood loss is expected. • All transfusions, whether allogeneic or autologous blood, carry a risk. • Planning and appropriate treatment in advance of or during surgery may reduce transfusion requirements. • Before transfusing a patient always consider whether it is necessary and if so how many units are required. • ICS and PCS are considered the most costeffective methods of autologous transfusion.

Further reading Birkmeyer JD, Goodnough LT, Aubuchon JP, Noordsij PG, Litenberg B. The cost effectiveness of preoperative autologous blood donation for total hip and knee replacement. Transfusion 1993; 33: 544–51. British Committee for Standards in Haematology. Guidelines for implementation of a maximum surgical blood order schedule. Clin Lab Haematol 1990; 12: 321–7. British Committee for Standards in Haematology. Guidelines for autologous transfusion. 1. Preoperative autologous donation. Transfus Med 1993; 3: 307–17. British Committee for Standards in Haematology. Guidelines for autologous transfusion. 2. Perioperative haemodilution and cell salvage. Br J Anaesth 1997; 78: 768–71. Calman KC. Cancer: science and society and the communication of risk. Br Med J 1996; 313: 799–802. Consensus statement. Autologous Transfusion: 3 years on. What is new? What has happened? Transfus Med 1999; 9: 285–6. Goodnough LT, Brecher ME, Kanter MH, AuBuchon JP. Transfusion medicine. Part 1. N Engl J Med 1999; 340: 438–47.

307

Chapter 26 Goodnough LT, Brecher ME, Kanter MH, AuBuchon JP. Transfusion medicine. Part 2. N Engl J Med 1999; 340: 525–33. Goodnough LT, Shander A, Brecher ME. Transfusion medicine: looking to the future. Lancet 2003; 361: 161–9. Hébert PC, Wells G, Blajchman MA et al. A multicentre, randomised controlled clinical trial of transfusion requirements in critical care. N Engl J Med 1999; 340: 409–17. Newman JH, Bowers M, Murphy J. The clinical advantages of autologous transfusion. A randomised controlled study after knee replacement. J Bone Joint Surg 1997; 79B: 630–2.

308

Popovsky MA, Whitaker B, Arnold NL. Severe outcomes of allogeneic and autologous blood donation: frequency and characterization. Transfusion 1995; 35: 734–7. Royal College of Physicians of Edinburgh. Consensus Conference on Autologous Transfusion. Transfusion 1996; 36: 625–67. Thomas D, Wareham K, Cohen D, Hutchings H. Autologous blood transfusion in total knee replacement surgery. Br J Anaesth 2001: 86: 669–73. Vanderlinde ES, Heal JM, Blumberg N. Clinical review: autologous transfusion. Br Med J 2002; 321: 772–5.

Chapter 27

Tissue banking Deirdre Fehily and Ruth M. Warwick

The banking of cord blood, as described in Chapter 28, has been a natural progression for the UK blood services, where much of the expertise and infrastructure required already existed for the banking of blood. Less obvious, perhaps, is the parallel that exists between the requirements for safe blood banking and the safe banking of other human tissues such as bone, skin and heart valves, which are used for orthopaedic procedures such as revision hip surgery, burns and heart valve replacement respectively. Although the tissues themselves may be very different from blood, the principles applied to blood and cord blood banking are equally relevant to the banking of these tissues. In recognition of the UK blood services’ expertise and infrastructure, the National Blood Service in England has greatly increased its involvement in tissue banking over the last 10 years. Blood centres in other European countries, notably France and Spain, also contribute significantly to the provision of tissue banking services. There are some features of tissue banking, however, that make it distinctly different from blood or cord blood banking, and these must be addressed by any organization undertaking tissue banking. • In many cases, tissue donors are necessarily deceased. Donor selection, consent, testing and counselling issues all present different challenges in this context. • Processing is by necessity very ‘open’, and therefore the requirements for processing facilities are much more stringent than for blood. Methods of processing also vary greatly and tissues can be supplied for transplant in either a viable or non-viable state. • The medical director of a blood service tissue

bank is normally a haematologist, who will not be expert in the clinical applications of tissue transplantation in the surgical specialities of orthopaedics, plastic surgery, and cardiac surgery for example. Consequently, the clinical dialogue between the tissue bank and the user hospital is different to that which exists for blood and may rely on the existence of a medical advisory committee with clinical experts from the relevant disciplines. While tissue banking within the blood services has grown significantly, the service is also provided by many hospital departments and particularly in the USA by independent organizations some of which operate for profit. Where cadaveric tissues are banked, there is always some degree of collaboration with organ transplantation programmes and in some cases organ and tissue donation is coordinated in a fully integrated way, though this is unusual in the UK. Hospital-based tissue banks usually concentrate on the banking of a single tissue and are directed by a surgeon who is also a user of that tissue. This latter type of arrangement gives the benefit of very good clinical feedback and motivation to maximize collection but has the disadvantage of a lack of expertise in donor selection and good manufacturing practice (GMP). Guidance is available from a number of sources on the quality standards that should be applied in tissue banking. Table 27.1 includes references to the most important of these documents. Legally binding regulation is in place in the USA, where the Food and Drug Administration (FDA) has published a number of relevant rules (also listed in Table 27.1). In the European Union, legally binding regulation will soon be in place. A new Directive, adopted in 2004, will require all 309

Chapter 27 Table 27.1 Pertinent laws and guidance in the field of tissue

Consent

banking. The Human Tissue Act 1961 (UK) Human Organ Transplants Act 1989 (UK) Anatomy Act 1984 (UK) Coroners Act 1988 (UK) Human Tissue Ethical and Legal Issues published by the Nuffield Council on Bioethics 1995 Committee on Microbiological Safety of Blood and Tissues for Transplantation, Department of Health: guidance on the microbiological safety of human tissues and organs used in transplantation (NHS Executive,August 2000 A Code of Practice for the Diagnosis of Brain Stem Death (including Guidelines for the Identification and Management of Potential Organ and Tissue Donors) (UK Department of Health, March 1998) Rules and Guidance for Pharmaceutical Manufacturers (EEC Orange Guide) (London, HMSO, 1997) Standards for tissue banking, British Association of Tissue Banks (Transfus Med 1996; 6: 155–8) Standards for tissue banking,American Association of Tissue Banks (1998) ‘Guidelines for tissue banking’ (section 4, added in 1999) in Guidelines for the Blood Transfusion Services in the United Kingdom (3rd edn., 1996) FDA regulations governing human tissue intended for transplantation. Title 21 Code of Federal Regulations Part 1270 FDA Guidance for the Industry on Validation at www.fda.gov/cver/guidelines.htm The Belgian Tissue Banking Law 1988 The Spanish Tissue Banking Law 1996 The French Tissue Banking Law 1994 Ethical Aspects of Human Tissue Banking. Opinion of the European Group on Ethics in Science and New Technologies to the European Commission (21 July 1998) A Code of Practice for Tissue Banks Providing Tissues of Human Origin for Therapeutic Purposes (UK Department of Health, 2001) Guide to safety and quality assurance for organs, tissues and cells (Council of Europe, 2002)

member states to have inspection and accreditation systems in place by June 2006 which ensure that all banks providing these services comply with an agreed set of standards. This chapter aims to identify the key considerations for a blood centre embarking on the banking of human tissues.

310

Patients undergoing surgery for joint replacement or heart transplant can donate bone and heart valves respectively. In the former cases, written consent for all the relevant aspects of tissue donation and testing should be obtained in advance of tissue retrieval and be witnessed, and the process of obtaining consent for donation should be entirely separate from the process of obtaining the patient’s consent for surgery. The legal requirements for obtaining permission for the retrieval of tissues after death vary from country to country. However, even where ‘opting out’ or ‘presumed consent’ systems are operated, it is considered best practice to confirm that relatives do not object to the donation. In the UK the Human Tissue Act 1961 makes it clear that the requirement in these cases is to establish a ‘lack of objection’ from the deceased’s family, rather than ‘consent’, and, once established, the donation can proceed even in the absence of a donor card. The UK legal framework for consent to donation is currently under review. Consent should cover the following areas: • the intended clinical use of the donation; • virological testing at the time of donation (and, for living donors, again at least 6 months later) including specific mention of human immunodeficiency virus (HIV); • review of medical records held elsewhere, if necessary; • receiving counselling in the event of positive virology markers in living donors or cadaveric donors, if they are deemed to have clinical significance for surviving relatives; and • the use of the tissue for research and development, if it proves unsuitable for clinical use.

Donor selection Living donors

Where patients can be interviewed face to face, the selection process can be very similar to blood donation but should also include a review of the patient’s hospital notes.

Tissue banking

Cadaveric donors

The primary source of donor selection information for cadaveric donors is the interview with the donor’s relative(s). The interviewee should be the person who knew the potential donor best, even if they are not the next of kin. This should be established at the beginning of the interview. For cadaveric donors, additional information should be sought from the family doctor as an added security measure required in the absence of a face-toface interview with the donor. Additionally, where a postmortem has been performed, the results should be reviewed as part of the donor selection process. Interviews with donors or their families should include enquiries as detailed in Table 27.2.

Donor testing Testing of cadaveric and living donors closely follows the testing undertaken for blood donation. Living donors

In many countries, including the UK, there is a requirement to quarantine living tissue donations and to obtain a further blood sample from the donor at least 180 days following their initial donation. Virology testing at donation and after quarantine should be as for blood donors. Cadaveric tissue donors

The quality and nature of blood samples removed from deceased donors varies considerably due to autolysis and haemolysis and there is documented observation of a high rate of false positivity in antibody assays and a significant rate of inhibition in nucleic acid technology (NAT) assays. It is important to standardize, as much as possible, the site and method of sample collection and to minimize the time period between death and blood sampling. Consideration should also be given to the validity of the sample that is tested. A sample that is taken close to an intravenous administration or central line may be diluted even if the donor has

not received a significant amount of fluids. Wherever possible, antemortem blood samples should be taken, as long as they can be reliably identified. Recent work suggests that newer DNA extraction techniques can be adapted and applied to remove the inhibition caused by cadaveric samples in NAT assays. It appears that NAT may therefore provide the most secure method for mandatory marker testing in cadaveric donors if applied in the appropriate way. An additional consideration with cadaver donors is whether transfusion of blood or other fluids in the antemortem period, particularly where the donor has also lost blood, may have resulted in a plasma dilution effect which might render virology tests unreliable. It is essential therefore to record all fluids administered in the 48 h prior to death. An estimation of any plasma dilution effect can be calculated. It is generally accepted that a blood sample that is more than 50% dilute should not be considered valid. An algorithm can be applied in the calculation of plasma dilution for cadaveric donors (Table 27.3). In the context of hepatitis B surface antigen (HBsAg) testing, high rates of non-specific reactivity are recognized in viral screening tests undertaken on cadaver blood samples. American Association of Tissue Bank guidance requires the discard of donations associated with such samples. In the UK, where confirmatory tests clearly indicate the absence of infection, tissues derived from cadaveric donors whose blood samples are repeatably reactive in an HBsAg screening test may be utilized, according to the Guidelines for the Blood Transfusion Services in the UK. Clarification of the significance of the repeatably reactive tests can be undertaken using the following approach. • Samples must lack anti-hepatitis B core antigen (HBc) using a validated assay for blood donors or, when available, an assay validated in the context of cadaveric testing. • The sample must not demonstrate evidence of neutralization of HBsAg with a specific antibody. • Testing must be undertaken in a designated diagnostic laboratory with proven expertise in the context of blood donor testing and cadaveric donor testing.

311

Chapter 27 Table 27.2 Enquiries to be made of tissue donors or their families.

Enquiry category

Information sought and outcome

Notes, examples and special circumstances

General past medical history

Malignancy specifically excludes

Exceptions are cured in situ cancer of the cervix and basal cell carcinoma. Primary brain tumours should be excluded unless benign nature is confirmed histologically. This is due to the risk that a solitary metastasis might be mistaken for a primary. Diagnostic procedures during life may increase the chance of extracranial metastases by breaching the blood–brain barrier For example sarcoidosis, Crohn’s disease and ulcerative colitis have some features in common with some infectious disorders Parkinson’s disease and multiple sclerosis are common diseases in this category Examples include systemic bacterial infections such as fulminant pneumonia (although small foci of infection associated with ventilation, even if currently being treated, may not exclude the donor), septicaemia, acute myocarditis and active tuberculosis Examples include rheumatoid arthritis, systemic lupus erythematosus and polyarteritis nodosa.These may affect the quality of the tissues. May be treated with drugs that may affect the quality of the tissues. May be treated with immunosuppressants that may affect validity of test results.Are of unknown aetiology, possibly with an infectious trigger

Diseases of unknown aetiology are reason to exclude Diseases of neurodegenerative aetiology specifically exclude Diseases of known infectious origin usually exclude

Multisystem autoimmune diseases exclude donation

Medication

Enquiries primarily aimed at identifying underlying disease that may make the donor ineligible

Long-term steroid therapy can affect the quality of skin and bone and immunosuppression may render antibody-based tests invalid

Hepatitis and HIV transmission risk

Risks due to acupuncture, tattooing, ear or body piercing, etc. in the previous year. Receipt of an organ or tissue transplant

Living donors with these risks can be accepted although the retest should be timed to ensure that it is conducted at least 1 year after the risk event. Consideration should be given to adding an anti-HBc test at the recall. Corneal recipients excluded from donation

CJD

Enquiries should elicit any family history of CJD and any brain or spinal surgery before 1992. Hormone treatment for infertility or growth before 1985

Details and dates of brain or spinal surgery should be recorded and further investigations made with the hospital concerned

Travel (malaria and Chagas’ disease)

The rules for history-taking, acceptance and malarial antibody testing of blood donors should be applied equally to tissue donors

It is not clear whether any risk of malaria transmission remains in non-viable tissues. Cornea banks do not exclude donations on the basis of malarial risk

Tissue-specific medical history

Depending on the tissues to be donated, enquiries should be made to exclude donors on the basis of a medical history which may imply that the quality of the specific tissue is compromised

For example, previous eye surgery in eye donors, or previous hip surgery in femoral head donors

Recent history

Enquiries should establish circumstances surrounding the death, including whether a hospital or coroner’s postmortem is to be performed

CJD, Creutzfeldt–Jakob disease.

312

Tissue banking Table 27.3 Algorithm for the

calculation of plasma dilution.

Interval prior to sampling

Volume infused (mL)

Per cent retained

Volume retained (mL)

Crystalloid infused > 24 h 2–24 h 1–2 h 1 ¥ 109/L by day 21, platelets > 20 ¥ 109/L by day 30, unrelated donors need > 3.7 ¥ 108 nucleated cells/kg recipient body weight for sturdy graft and good outcome

Doses of > 4 ¥ 106 CD34+ cells/kg recipient body weight will give reliable prompt engraftment slightly faster than bone marrow Neutrophils > 1 ¥ 109/L by day 17, platelets > 20 ¥ 109/L by day 21

Engraftment is slow and uncertain, and at least 28 days may be needed for both neutrophil and platelet recovery Uncertain if one cord blood donation can engraft an adult > 75 kg reliably

Graft-versus-host disease (GVHD) Acute Standard-risk cases will get acute GVHD in about 65%, with about 25% getting severe disease Greater risk in unrelated donor transplants

Appears to be no different from bone marrow in sibling transplants

Since cord blood transplants mainly used in children, incidence is low, either because children known to have low incidence of GVHD or naive T cells in cord blood

Chronic

Risk is increased to 50–60%

Not known for sure as yet

Some 20–30% of patients will develop some form of chronic GVHD In about 10% this will be severe and debilitating

G-CSF, granulocyte colony-stimulating factor.

Indications for haemopoietic stem cell transplants

The detailed indications for intensive therapy requiring haemopoietic stem cell rescue or support change over time. Current indications as considered appropriate by the European Group for Blood and Marrow Transplantation (EBMT) are regularly updated and provide a broad consensus of what the transplant community in Europe considers reasonable. The transplant data have 372

been analysed from the major transplanting countries contributing to the EBMT database. This has enabled a hierarchy of indications to be prepared based on actual transplant activity. That is, it should reflect what transplant physicians do and not merely what they say they would do. Naturally, the indications agreed by transplanters are likely to be at the limits of the procedure. Some indications may be greeted with less

Stem cell transplantation and immunotherapy

enthusiasm by more conservative physicians. Also, as trial data emerge, enthusiasm and indications may change. Recent studies looking at the role of high-dose chemotherapy and autologous BMT in breast cancer have not shown a significant benefit for BMT. Although not conclusive, it is clear that further trials are needed in breast cancer before the role of intensive therapy with BMT can be clarified, if indeed there is any role. Most clinicians would agree that the following groups are highly appropriate for BMT. Allogeneic BMT

This is indicated in the following conditions. • CML in first chronic phase with a matched sibling or unrelated donor. The introduction of the specific tyrosine kinase/bcr/abl inhibitor imatinib (Glivec) has provided improved responses for many patients with CML, but is still not believed to be a curative agent. • Acute leukaemias in second complete remission. • Very severe aplastic anaemia in children and young adults. • Some high-risk acute leukaemias in first complete remission. Relative indications for allogeneic BMT or indications appropriate for clinical trial would be: • acute leukaemias in first complete remission in adults; • multiple myeloma; and • myelodysplastic syndromes in children or young adults. Autologous transplantation has less clear-cut indications, most remaining appropriate for clinical trials that compare BMT with less intensive approaches. Clear-cut indications based on randomized clinical trial data include: • high-grade non-Hodgkin’s lymphoma in second complete remission or in first relapse with responsive disease; and • multiple myeloma. Patients with Hodgkin’s disease who have relapsed after at least two different treatments, e.g. radiotherapy and chemotherapy, have improved disease-free survival after autograft but overall survival may not be so different since some patients can be ‘salvaged’ by transplant

after further relapses following conventional chemotherapy. Complications of transplantation

Patients who are being considered for any form of blood or marrow transplant must be given full information about the procedure prior to their agreement to proceed. Although procedures such as allogeneic BMT for CML and severe aplastic anaemia have transformed the prognosis such that the majority of patients with these diagnoses will survive, all transplant procedures carry major risks of mortality, morbidity and later ‘collateral damage’. Some of these risks will remain lifelong. The chronology of the major complications of allogeneic blood or marrow transplant is shown in Fig. 33.1. Regimen-related toxicity

This refers to the immediate toxic effects of the radiotherapy or chemotherapy used for the transplant. With the exception of some current experimental procedures, all transplants have used intensive therapy that will ablate (kill off) the patient’s own bone marrow. Other organs are also damaged, especially the gut, with severe mucositis a major problem. Less commonly, liver, heart, lungs and kidneys may suffer transient or even permanent damage. Careful pretransplant assessment of each patient is essential. The use of RIC transplants is intended to minimize this toxicity and enable patients to receive transplants who might be unfit for full chemotherapy and radiotherapy conditioning. Rejection

Rejection is an immune-mediated event in which the pretransplant conditioning and immunosuppression are insufficient to prevent recognition by the recipient of HLA on the donor cells. It only occurs in allogeneic transplants, although graft failure can occur in autografts. Graft failure is due to inadequate numbers of stem cells in the transplant and/or pre-existing damage to the marrow microenvironment. 373

Chapter 33

Before transplant After transplant Alloimmunization GVHD

Chronic GVHD

Acute GVHD

Pneumocystis Bacteria Infection

Fungi Viruses

High risk: continuing while GVHD or poor graft function

CMV Zoster Infertility Central line problems: thrombosis and infection

Central line removed

Day + 50

Day + 100

Risk factors for rejection include the following. • HLA incompatibility between patient and donor, especially in unrelated donor transplants. • Low numbers of stem cells infused (< 3.7 ¥ 108 nucleated cells per kilogram body weight in unrelated BMT). • Prior sensitization of the patient to HLA or other marrow cell antigens. This is particularly problematic in patients with severe aplastic anaemia and can be prevented by using leucocytedepleted blood components from presentation onwards. The kinetics of engraftment of bone marrow and PBSC is shown in Fig. 33.2.

Fig. 33.1 Relationship of the major

complications of BMT to the time before and/or after transplant. Such a diagram can give only a broad view of the most common periods of relatively uncomplicated transplants. Those using an unrelated or other alternative donor will have greater risks of graft failure, graft-versus-host disease (GVHD) and thus continuing likelihood of infections. CMV, cytomegalovirus; HSV, herpes simplex virus; Zoster, varicella-zoster virus.

HSV

after the transplant. Acute GVHD is characterized by involvement of the following. • Skin: from an erythematous sunburn-like rash to a blistering exfoliative erythroderma. • Liver: typically the bile ducts are attacked and an obstructive jaundice-type picture develops. Milder forms may lead to elevated transaminases and cause considerable difficulties with diagnosis. Liver biopsy is advised. • Gastrointestinal tract: classically a profuse watery diarrhoea develops, bloody in the most severe cases. Upper gastrointestinal upset is not uncommon, with nausea and sickness. Rectal biopsy or upper gastrointestinal endoscopy is required for diagnosis.

Graft-versus-host disease

GVHD is caused by immune-competent T lymphocytes in the transplant recognizing antigens in the patient as foreign. This then leads to cytokine release, which increases and perpetuates the response. Despite immunosuppression of the patient to prevent GVHD, e.g. using cyclosporin plus methotrexate or corticosteroids, more than half of patients receiving allogeneic transplants will develop acute GVHD in the first 100 days 374

Relapse

Despite the intensive preparation for transplant, a significant proportion of patients will suffer recurrent disease after transplant (Fig. 33.3). Risk factors for acute leukaemia recurrence include the folowing. • Patient not in remission at time of transplant, especially if disease is resistant to conventional therapy.

Stem cell transplantation and immunotherapy (a)

PBSC infusion

120

Red cell transfusions

Hb (g/L)

100 80 60 40 20 0 300

myeloid leukaemia who received a peripheral blood stem cell (PBSC) allogeneic transplant from a sibling using reduced-intensity conditioning. The blood count falls following the chemotherapy (fludarabine and melphalan) and recovers quickly once engraftment begins. The patient had minimal acute graft-versus-host disease and the marrow graft has remained robust. Molecular evidence of chronic myeloid leukaemia was detected some 9 months after the transplant and was treated with low-dose donor lymphocyte infusion. The patient is alive and disease-free 6 months later. Hb, haemoglobin; WBC, white blood cells.

200 150 100 50 0 12 10

WBC (¥109/L)

Fig. 33.2 (a) A patient with chronic

Platelets (¥109/L)

Platelet transfusions 250

8 6 4 2 0 –20

–10

• Patient is beyond first complete remission, i.e. has already relapsed once after chemotherapy even if now in remission. • Autologous marrow or PBSC transplant (no GVL). • No GVHD after allogeneic transplant. Although GVL can be separated from GVHD in some situations, the presence of GVHD, especially chronic GVHD persisting more than 100 days after transplant, is associated with a potent GVL effect. However, overall survival may still be worse due to the toxic effects of GVHD.

0

10 20 30 Days before and after transplant

40

50

60

Infectious complications

The immune system of the transplant recipient must be suppressed to allow the graft to be accepted, and antitumour therapy such as total body irradiation (TBI) ensures that the patient has minimal immune function at the time of the transplant. Even RIC transplants have this risk, because they utilize intensive immunosuppression in order to ensure that the transplant is not rejected. The agents used (fludarabine and anti-Tcell or pan-lymphoid monoclonal antibodies) provide prolonged and profound immune deficiency. Haemopoietic recovery takes at least 2–3 375

Chapter 33

(b) Bone marrow infusion 125

Hb (g/L)

Red cell infusions

100

75 –10

0

10

20

30

40

50

60

Day

1000 Platelet infusions

WBC (x 109/L)

100

WBC (x 109/L)

Platelets (x 109/L)

10

1

0.1 –10

0

10

20

30 Day

376

40

50

60

Fig. 33.2 (b) A patient with severe aplastic anaemia secondary to hepatitis. Blood counts are low before the transplant, reflecting the aplasia and are relatively slow to recover. Severe acute graft-versus-host disease occurred but was controlled eventually with corticosteroids and antilymphocyte globulin. The patient is alive and well 4 years later. Hb, haemoglobin; WBC, white blood cells.

Stem cell transplantation and immunotherapy

Probability %

100

Fig. 33.3 Probability of survival after

allogeneic transplants for chronic myeloid leukaemia in chronic phase by donor type and disease duration, 1994–99.

80

HLA-identical sibling, 20 years) (£ 20) CR2 or other remission (> 20 years) (£ 20) Not in remission(> 20 years) (£ 20) CR1 (> 20 years) (£ 20) CR2 or other remission (> 20 years) (£ 20) Not in remission(> 20 years) (£ 20) CR1 CR2 or other remission Not in remission

CP1 (first chronic phase) Acc. phase Blast phase CP1 (first chronic phase) Acc. phase Blast phase

Disease stage

1912 524 1067 202 252 387 729 265 177

741 455 312 778 363 163 176 233 171 652 191 153 95 59 39

3528 750 249 1240 460 174

Number of transplants

59 ± 2 45 ± 5 23 ± 3 36 ± 8 25 ± 7 10 ± 4 51 ± 5 41 ± 8 22 ± 8

49 ± 5 61 ± 6 32 ± 7 52 ± 4 15 ± 5 27 ± 8 37 ± 11 48 ± 8 25 ± 8 38 ± 4 13 ± 6 19 ± 8 51 ± 14 15 ± 15 9±9

64 ± 2 45 ± 5 30 ± 7 50 ± 3 28 ± 5 20 ± 7

Survival at 3 years post transplant (%)

56 ± 3 36 ± 7 18 ± 4 36 ± 8 25 ± 7 5±4 39 ± 8 34 ± 9 20 ± 8

44 ± 5 58 ± 7 29 ± 7 46 ± 6 13 ± 5 24 ± 9 22 ± 20 48 ± 8 25 ± 8 33 ± 6 NA 17 ± 8 46 ± 15 NA NA

60 ± 3 33 ± 12 26 ± 8 45 ± 4 25 ± 6 20 ± 7

Survival at 5 years post transplant (%)

Stem cell transplantation and immunotherapy

379

380

Non-Hodgkin’s lymphoma (NHL) Follicular NHL

All ages; 90% > 20 years. All ages; 77% > 20 years

Hodgkin’s disease

Severe aplastic anaemia

Allogeneic sibling

Myelodysplasia (all ages; 85% > 20 years)

Allogeneic identical sibling

Autologous

Allogeneic sibling Other related or unrelated

Autologous

Other related or unrelated

Allogeneic sibling

Allogeneic unrelated donor

Transplant type

Disease

Table 33.2 Continued

CR1 CR2 CR3 or greater Relapse no. 1 Relapse beyond no. 1 Primary treatment failure Relapse no. 1 Primary treatment failure

CR1 CR2 CR3 or greater Relapse no. 1 Relapse beyond no. 1 Primary treatment failure Advanced disease assumed Advanced disease assumed

Age £ 20 years Age > 20 years. Age £ 20 years Age > 20 years.

Refractory anaemia (RA) RA with excess blasts (RAEB) RAEB in transformation (RAEBT) RA RAEB RAEBT

Disease stage

118 189 45 476 223 252 108 100

106 375 96 912 347 364 172 34

679 636 244 118

229 296 279 87 116 124

Number of transplants

84 ± 9 71 ± 9 67 ± 17 70 ± 5 60 ± 8 64 ± 7 69 ± 10 62 ± 11

87 ± 9 76 ± 6 72 ± 10 65 ± 4 58 ± 7 53 ± 7 33 ± 9 34 ± 17

78 ± 3 68 ± 4 52 ± 7 32 ± 10

53 ± 8 39 ± 7 41 ± 7 31 ± 13 34 ± 9 27 ± 9

Survival at 3 years post transplant (%)

79 ± 13 71 ± 9 51 ± 24 59 ± 11 54 ± 11 42 ± 20 59 ± 20 62 ± 11

87 ± 9 64 ± 10 67 ± 12 54 ± 7 43 ± 12 48 ± 8 25 ± 12 34 ± 17

77 ± 4 66 ± 4 50 ± 8 27 ± 13

50 ± 9 27 ± 11 40 ± 8 27 ± 13 34 ± 9 15 ± 14

Survival at 5 years post transplant (%)

Chapter 33

Allogeneic sibling Other related or unrelated

Fanconi’s anaemia (all ages; 98% < 30 years)

Diagnosis to transplant £ 18 months Diagnosis to transplant > 18 months Diagnosis to transplant £ 18 months Diagnosis to transplant > 18 months Diagnosis to transplant £ 18 months Diagnosis to transplant > 18 months

CR1 Relapse no. 1 Primary treatment failure CR1 Relapse no. 1 Primary treatment failure

CR1 CR2 CR3 or greater Relapse no. 1 Relapse beyond no. #1 Primary treatment failure Relapse no. 1 Primary treatment failure

175 111

526 31

2690 835 564 211 52 33

53 30 24 49 26 27

259 435 67 1027 305 597 94 99

76 ± 8 28 ± 10

81 ± 3 67 ± 18

60 ± 3 48 ± 5 46 ± 5 41 ± 8 32 ± 14 11 ± 10

65 ± 16 35 ± 22 42 ± 22 (30 months) 55 ± 16 14 ± 14 49 ± 20

66 ± 7 53 ± 6 38 ± 15 44 ± 4 31 ± 6 49 ± 5 33 ± 12 22 ± 10

76 ± 8 NA

81 ± 3 NA

40 ± 7 25 ± 9 42 ± 5 29 ± 10 26 ± 16 NA

65 ± 16 35 ± 22 NA 55 ± 16 NA 49 ± 20

53 ± 11 45 ± 8 32 ± 16 37 ± 5 30 ± 8 41 ± 7 33 ± 12 22 ± 10

* Results from the International Bone Marrow Transplant Registry (IBMTR) (allogeneic) and the American Autologous BMT Registry (ABMTR) (autologous) for adult patients over 20 years old (except where stated). The data presented here were obtained from the Statistical Center of the International Bone Marrow Transplant Registry and the Autologous Blood and Marrow Transplant Registry. The analysis has not been reviewed or approved by the Advisory Committees of the IBMTR/ABMTR. Transplants took place between 1994 and 1999 and reflect all those registered with the IBMTR and the ABMTR. Figures are given for survival at 3 and 5 years, where available. This is not disease-free survival, although most patients with acute leukaemia, myelodysplasia, aplastic anaemia and chronic myeloid leukaemia are likely to be disease-free, if alive.The reliability of the data is clearly greatest where the greatest numbers of transplants have been performed.These results do not reflect the emerging trend for reduced-intensity conditioned transplants. CR, complete remission (CR1, first complete remission); CP, chronic phase of chronic myeloid leukaemia.

Allogeneic sibling Other related or unrelated

Other related or unrelated

Allogeneic sibling

Autologous

Allogeneic identical sibling

Autologous

Allogeneic identical sibling

Autologous

Thalassaemia (all ages; 93% < 20 years)

Multiple myeloma

Lymphoblastic NHL

Diffuse large-cell NHL

Stem cell transplantation and immunotherapy

381

Chapter 33 Table 33.3 Classification of indications

Allogeneic haemopoietic stem cell transplant

Autologous haemopoietic stem cell transplant

CV < 50; very high degree of consensus

AML CR1 AML other than CR1 ALL CR1 ALL other than CR1 CML CP1 CML other than CP1 Myelodysplasia Non-Hodgkin’s lymphoma

Multiple myeloma Hodgkin’s disease Non-Hodgkin’s lymphoma

CV 50–80; some variation in practice between BMT units/nations

Multiple myeloma Chronic lymphocytic leukaemia

AML CR1 AML other than CR1

CV > 80; little consensus as to evidence in support of indication. Clinical trials highly appropriate in these conditions

Hodgkin’s disease

ALL CR1 ALL other than CR1 CML CP1 CML other than CP1 Myelodysplasia Chronic lymphocytic leukaemia

Degree of consensus

for blood and marrow transplant according to BMT unit practice in Europe in 2001.

All transplant procedures are arduous, even though mortality has fallen over the past years. In addition, the use of allogeneic donors causes major problems with immune reconstitution such that few patients over 60 years old would be considered for sibling transplants, or for unrelated procedures over 50 years of age. Reduced-intensity conditioned transplants are increasing these age thresholds but no mature data are available for long-term outcome. For autografting some groups have extended the limit to 75 years and the author has experience up to 68 years. Fitness of the patient and the likelihood of benefit are the most important considerations. ALL, acute lymphoblastic leukaemia; AML, acute myeloid leukaemia; CML, chronic myeloid leukaemia; CR, complete remission (CR1, first complete remission); CP, chronic phase of CML; CV, coefficient of variance.

• 35% LFS in second or subsequent complete remission (relapse risk 45%); and • 26% LFS for patients in relapse at the time of allogeneic BMT (relapse risk 57%). Results for other common indications are shown in Table 33.3. Registry data are of great importance but cannot replace careful assessment of individual patients. Such individual factors may serve to improve or worsen the risks for a particular case, e.g. coexistent disease or toxic effects of prior chemotherapy. Also, registries report only mature data, usually with a minimum of 3 years follow-up. Since RIC transplants are a recent development, experience with these is not yet 382

reflected in registry data, and primary research publications and meeting reports must form the basis of current outcomes. Regulatory aspects of haemopoietic stem cell transplantation

Concern over the appropriate indications for transplants and an increased awareness that the support services needed for BMT require careful quality assurance has led to an increasingly stringent regulatory environment. Some of these support services include: • stem cell cryopreservation;

Stem cell transplantation and immunotherapy

• stem cell manipulation, e.g. CD34 selection, stem cell expansion; and • molecular and cellular diagnostic facilities. The number of teams performing transplants in Europe increased from eight in 1973 to almost 100 in 1983. Over 340 teams reported data to the EBMT group in 1995, and 479 in 2001, when 16 555 transplants were performed. Some of the less complex transplants are no longer the domain of specialist units but can be performed safely and appropriately in smaller centres. However, safety of patients remains the first concern and this has led to the current regulatory environment, most of which remains managed by professional groups and is essentially voluntary. The European Union is likely to develop its own guidance in the near future, which will be statutory. At present the regulatory structures in Europe and the UK, and their status, are as follows. • Joint Accreditation Committee of EBMT and International Society for Cellular Therapy (ISCT) Europe (JACIE). Voluntary/professional guidelines. Accreditation in the UK will be administered by the British Society of Blood and Marrow Transplantation (BSBMT). Implementation not yet complete. • European Directive 85/374/EEC (1985). Statutory in relation to product liability in general. • UK Department of Health guidance notes on the collection, storage and infusion of bone marrow and stem cells (1997): advisory. • Medicines and Healthcare Products Regulatory Agency (MHRA). Accreditation of UK establishments involved in tissue banking, including haemopoietic stem cells. In legal terms a voluntary process, but the intention is that this will be an essential requirement in the UK. • British Committee for Standards in Haematology (BCSH). Professional guidelines on stem cell collecting, processing and storage as well as separate guidance on the appropriate clinical facilities for the care of patients with severe bone marrow failure. • European Union Directive on Tissues. Currently being drafted, this will include stem cell regulation. It is likely to enter EU law in 2004, and to come into force in individual member states in 2006.

It is clear that if clinicians wish to ensure that they are operating a safe service they will need to show adherence to clear quality systems including: • attention to the quality of raw materials, premises, equipment, reagents and storage; • each product, e.g. CD34-selected stem cells, must have an appropriately defined specification set; • safety and efficacy of each technical procedure must be validated and documented; • standard operating procedures (SOPs) should be written for each procedure; • evidence of the competence, training and proficiency of each staff member must be available; • internal and external quality assurance systems must be in place; • quality audits should be performed by inspectors external to the BMT service; • each stem cell donation or product must be uniquely identifiable, and the harvest procedure, time, date and nature of processing documented; • records must be kept for at least 11 years or until the patient/donor has reached 21 years of age, whichever is the longer. Although these quality measures may often appear pedantic at best or even restrictive of innovation at worst, they are essential if, over time, the confidence of donors, patients and the public in the safe conduct of haemopoietic stem cell transplants is to be secured and maintained.

Immunotherapy The idea that the immune system might be exploited to prevent recurrence of malignant disease or even treat it in the first place has been present for a long time. Initially, studies involved crude preparations of bacteria to provide immune stimuli. Landmarks include: • pyogenic bacteria (erysipelas) used by William Coley in New York in 1893 to stimulate antitumour responses; • use of bacille Calmette–Guérin (BCG) as an anticancer ‘vaccine’ by Holmgren in Sweden (1935); • Mathe and colleagues showed a beneficial effect of BCG against ALL in 1969, although 383

Chapter 33

Resting lymphocyte

1

Activated lymphocyte

2 3 4 Tumour cell

Apoptosis Fig. 33.4 Interaction between tumour cells and

lymphocytes. The tumour cell expresses antigens essential for recognition by T lymphocytes. Adhesion molecules (1) maintain cell-to-cell contact while major histocompatibility complex (MHC) structures present peptide antigens that may be tumour specific (2). If costimulatory molecules such as B7 (3) are present as well, the T lymphocyte will become

subsequent randomized trials failed to show any benefit; and • South-west Oncology Group in the USA showed clear benefit of local BCG in bladder cancer in 1980. It is only in the past 5–10 years, however, that essential knowledge of the details of the immune response, and how it may be enhanced or abrogated, has emerged. In the 1980s the identification first of IL-2 and its occasional effects against malignant melanoma, despite frequent major toxicity, showed that there might be a role for more specific immune targeting. Adding lymphokine (IL-2)-activated killer cells or isolating tumourinfiltrating lymphocytes on the basis that they may have a specific effect against tumours again showed anecdotal promise but has not entered the mainstream of therapy. IL-2 does have a clear, albeit inconsistent, effect against renal cell cancers, although the mechanisms for this and the other non-specific cellular therapies are not known. More recently, a greatly improved understanding of the activation and stimulation of cells in 384

activated and immunity will develop. In the absence of B7, however, the T lymphocyte will either die by apoptosis or become anergic (inactive but alive) and there will be tolerance towards the antigen. Tumour cells may produce soluble cytokines or express other molecules (4) that can either block activation or induce apoptosis.

the cellular immune response has triggered much research. This, combined with an explosive increase in knowledge of antigen-presenting cells such as dendritic cells, means that protocols being developed now have the benefit of immunological logic rather than clinical empiricism. It is now understood that immunity against foreign antigens or cells is mediated, at least partially, by a combination of surface antigens having to be all in place on both the target tumour cell and the effector cells of the immune system. In addition the most potent effects are seen when foreign antigens have been processed by follicular dendritic cells. The necessary antigens are shown in Fig. 33.4. The combination of T-cell receptors, adhesion molecules and the costimulatory molecules B7.1 (CD80) and B7.2 (CD86) are all required for effective signalling to trigger a cytotoxic T-cell response. Firstly, two areas that have established the role of cellular immunotherapy will be considered. These include GVL and the use of cytotoxic T cells to treat post-BMT lymphomas. Secondly,

Stem cell transplantation and immunotherapy

the future potential of immunotherapy will be summarized. Graft versus leukaemia

During the 1980s there was a vogue for BMT procedures in which the T lymphocytes were removed in order to prevent GVHD. It was recognized in the late 1980s that these transplants were associated, particularly in CML, with a very high incidence of relapsed disease, especially if the patient developed no GVHD at all. Although a relationship between GVHD and reduced relapse had been noted earlier, this was the first occasion on which there was clear evidence of a GVL effect being mediated via GVHD. This is caused by alloreactive T lymphocytes in a BMT identifying foreign antigens in the host. Subsequent to these data from Tcell depleted transplants, it became apparent that there was a difference in relapse rate with other forms of GVHD prophylaxis. For example, with cyclosporin and methotrexate there was a higher incidence of relapse after transplant than with either cyclosporin or methotrexate alone, presumably because of the reduced GVHD seen with the dual therapy. Overall survival was not impaired, however, since GVHD remains the most important adverse event predicting survival after allogeneic BMT.

further clinical, cytogenetic and molecular remission by infusions of donor lymphocytes collected from the original transplant donor. However, this treatment was complicated by the following. • Severe GVHD. This could be avoided by giving lower and graded doses of donor T lymphocytes. In the majority of patients 1 ¥ 107/kg will induce complete molecular and haematological remission without a significant risk of GVHD or graft failure. • Failure of the graft leading to marrow aplasia. Prevented by monitoring the patient more carefully so that when donor lymphocyte infusions (DLI) are given early in relapse, before a significant proportion of the transplant has been rejected and replaced by leukaemia, they can be successful without the risk of bone marrow failure. Thus, by using DLI early in relapse and giving a specific dose of cells, it is possible to separate GVL from a graft-versus-host response. In addition to CML, DLI has been tried in numerous diseases with varying success. In multiple myeloma there is good evidence of reinduction of remission, although the number of cases remains small. Also it appears to be necessary to use rather more lymphocytes, e.g. 1 ¥ 108/kg. A scheme for immunotherapy in myeloma is shown in Fig. 33.5. In other diseases, such as acute leukaemias and myelodysplasia, the impact of DLI has been less great.

Post-BMT monitoring

With the advent of PCR it was possible to monitor leukaemic clones using sensitive molecular techniques. It became apparent that the loss of the molecular marker of malignancy in CML and some cases of ALL occurred gradually after transplant and not immediately. This suggested that there was some continued immune surveillance in addition to the initial impact of the radiotherapy and highdose chemotherapy used to prepare patients for BMT. Donor lymphocyte infusions

In the early 1990s it was shown that those patients with CML who suffered recurrence of their leukaemia after transplant could be induced into a

Cytotoxic T-cell therapy

Another form of potent GVL or antilymphoma is that demonstrated by Brenner and his colleagues in treating Epstein–Barr virus (EBV)-related lymphoproliferative disorder (LPD) in children who had received unrelated donor BMT. Intensive immunosuppression was given to ensure the graft was not rejected. This led to increased risk of reactivation of EBV, which triggered LPD. If untreated this progressed into an aggressive lymphoma. EBV causes glandular fever as well as Burkitt’s lymphoma in other patient groups. Although in the very early stages EBV LPD can be treated with antiviral therapy (aciclovir), once established it is likely to progress and respond poorly to conventional antilymphoma chemother385

Chapter 33

Dendritic cells

VEGF Muc-1

Plasma cells

Fas L TGF-b

Lymphocytes

TGF-b

IL-1b, IL-6

Fibronectin, collagen

TGF-b

Stromal cells Fig. 33.5 Immunosuppression in multiple myeloma. In

myeloma, a number of strategies have been identified that enable the plasma cells to prevent their killing by cytotoxic T lymphocytes. They produce transforming growth factor (TGF)-b, which stimulates stromal cells to produce more matrix materials (fibronectin, collagen) to provide a microenvironment conducive to plasma cells. The plasma cells employ three strategies to deflect or kill T lymphocytes.

apy. However, by isolating T cells and exposing them to EBV in vitro it was possible to generate clonal cytotoxic T cells that would recognize EBV antigens. These T cells were grown in the laboratory and then infused into the patients. In the main this treatment has been highly effective in both the prophylactic and therapeutic management of EBV LPD. These data show that by presenting tumourrelated antigens, it is possible to generate sufficient antitumour effect to induce remissions. It is of interest that the one disease that has been treated effectively with immunotherapy is a virus-driven malignancy. In the past few years more progress has been made in developing strategies for using cytotoxic T cells against viruses in the BMT setting than for antitumour indications. This reflects the fact that viruses possess foreign antigens not pos386

(1) Muc-1 antigen expression may cause apoptosis of T cells. (2) Fas L (Fas ligand) binding to Fas on T cells will also lead to apoptosis. (3) TGF-b secretion prevents T lymphocytes responding appropriately to interleukin (IL)-2 and they fail to increase in numbers sufficiently to kill plasma cells. Vascular epithelial growth factor (VEGF) may prevent ingress of dendritic cells and so prevent antigen presentation to T cells.

sessed by the human patient, and the relative lack of progress in antitumour immunotherapy over the past 5 years, beyond basic DLI approaches, suggests that this is a complex area difficult to overcome. Future of immunotherapy

Approaches to immunotherapy may be either passive or active. Passive immunotherapy

This involves the use of monoclonal antibodies. Studies in the 1980s using mouse monoclonal antibodies produced some encouraging results in Bcell malignancies but immunity to mouse protein prevented prolonged effects. Currently, antibodies

Stem cell transplantation and immunotherapy

engineered to be mainly human in origin have overcome this.

conjoined immunoglobulin idiotype component ensures specificity against the B-cell malignancy.

CD20 Antibodies to CD20 present on B lymphocytes have proven of value in the management of lowgrade B-cell malignancies.

Tumour cell vaccination This may involve: • apoptotic cells; • fresh tumour cells irradiated; or • apoptotic and/or fresh tumour cells fed to dendritic cells to process tumour antigens.

CD40 Such antibodies hold out great promise for treating more aggressive tumours. In mice, not only can lymphomas expressing CD40 be successfully treated by antibodies against CD40, but such immunotherapy appears to induce immunity such that the mice are resistant to subsequent challenge by tumour cell injections. Active immunotherapy

Tumour cell antigens, i.e. the proteins, may be used as the target for vaccination. More effect seems to be obtained with DNA vaccines, in which the DNA encoding the target gene or protein is used as the vaccine. Cancer vaccines have been reviewed in detail (see Further reading). Examples of antigens to target using immunotherapy approaches are: • virus antigens, e.g. EBV LPD; • mutated proto-oncogenes, e.g. p53; • mucins, e.g. Muc-1 in multiple myeloma and breast cancer; • idiotypic proteins, e.g. for B cells, have been used to immunize a sibling donor against the myeloma paraprotein of the recipient; • oncofetal antigens such as a-fetoprotein; and • products of chromosomal translocations such as t14;18 and bcr/abl.

Dendritic cell therapy Dendritic cells are essential for processing antigen to the immune system and many malignant conditions suppress the normal immune system and prevent antitumour effects. By taking dendritic cells out of the body and exposing them to tumour antigens (so-called pulsing), dendritic cells carrying specific antitumour antigens may be reinfused into the patient and stimulate T-cell responses in vivo. Cytotoxic T-cell therapy Such cells can be isolated and developed as performed by Brenner’s group for EBV LPD. One major problem is that tumour cells have developed specific strategies for avoiding control by the immune system of the host. Thus an understanding of how cancer cells avoid destruction is needed before detailed immunotherapy protocols can be developed. For example, in multiple myeloma it is known that antigens such as Muc-1 and Fas, and secretion of the cytokine transforming growth factor-b can adversely influence T cells to enable the tumour cells to proliferate. A current possible schema for this is shown in Fig. 33.6.

Conclusion DNA vaccines These show great promise since they can be more easily engineered than peptide antigens for vaccination and composite antigens can be created. An idiotype motif has been constructed together with fragment C of tetanus toxoid to immunize patients against their B-cell malignancy. The tetanus toxoid acts as a ‘danger signal’ or recall antigen, since most individuals have been immunized previously against tetanus toxoid, and the

At present there are still no clearly developed indications or proven strategies for cellular immunotherapy other than for generating GVL against CML (see Fig. 33.6) (and possibly myeloma) and for the prevention and treatment of EBV LPD. The last 5 years have been disappointing in failing to deliver any of the above approaches other than in single-centre studies or small-scale pilot trials. The exploitation of our 387

Chapter 33

Diagnosis

Disease control first chronic phase (CPI)

Collect HSC T lymphocytes Dendritic cells

Further cytoreductive therapy (? autograft)

Preparation of bcr/abl directed Cytotoxic T cells Dendritic cells pulsated with bcr/abl DNA vaccine

Minimal residual disease (molecular evidence only)

Complete remission Bone marrow and molecular Cure Molecular relapse (a)

Diagnosis

Collect HSC T lymphocytes Dendritic cells

Possible future routine use to prevent infection and maintain remission

Disease control first chronic phase (CPI)

HLA-compatible donor (sibling or unrelated)

Allogeneic transplant

Haemopoietic stem cells

Complete remission Bone marrow and molecular

Peripheral blood lymphocytes

Complete remission Bone marrow and molecular

Post-transplant virus infections CMV EBV +/– lymphoma (b)

388

Molecular relapse

Fig. 33.6 Future immunotherapy

strategies in chronic myeloid leukaemia (CML). Imatinib has transformed the initial treatment of CML, but may not be curative. From experience with allogeneic BMT it seems likely that some form of graftversus-leukaemia effect will be needed to maintain molecular and clinical remission. Ideally, this might involve autologous cells (dendritic cells and T lymphocytes) with no risk of graftversus-host disease. Patients with an allogeneic donor may still benefit from a transplant, particularly if immunotherapy enables the use of reduced-intensity conditioning. CMV, cytomegalovirus; EBV, Epstein–Barr virus; HSC, haemopoietic stem cells.

Stem cell transplantation and immunotherapy

increased knowledge of the immune system has not been easy. At the moment, the routine role of immunotherapy in cancer seems further away than a few years ago. However, its huge potential means that further effort and trials are clearly justified. The most likely application appears to be in a more refined approach to the use of DLI after transplant, and the advent of RIC transplants may permit the extension of haemopoietic stem cell transplants to more non-malignant diseases in which immune modulation may have a role. Meanwhile, what is the long-term future for BMT generally? These transplants can save life in many patients with incurable leukaemias and lymphomas. Patients who survive the first 3 years are likely to enjoy long-term survival, although life expectancy does not return to normal. RIC transplants will extend the benefits to more patients who might have been unfit to undergo the rigours of a myeloablative procedure. Whether the extension of RIC transplants to a wider range of malignant diseases is wise will become apparent in the next few years. Certainly, GVL effects are most potent in the context of minimal residual disease in slow-growing malignancies, such as CML and myeloma, in which there also happen to be specific tumour antigens (bcr/abl products and paraprotein). DLI-induced GVL struggles to compete with rapidly emerging relapses of acute leukaemias. Additional approaches are needed to deal with those patients whose primary disease is poorly responsive to current chemoradiotherapy.

Dazzi F, Goldman JM. Adoptive immunotherapy following allogeneic bone marrow transplantation. Annu Rev Med 1998; 49: 329–40. Goldman JM, Schmitz N, Niethammer D, Gratwohl A. Allogeneic and autologous transplantation for haematological diseases, solid tumours and immune disorders: current practice in Europe in 1998. Accreditation Sub-Committee of the European Group for Blood and Marrow Transplantation. Bone Marrow Transplant 1998; 21: 1–7. Gratwohl A, Baldomero H, Passweg J, Frassoni F, Neuderweisser D, Schmidz, N, Urbano-Ispizua A. Hematopoietic stem cell transplantation for hematological malignancies in Europe. Leukaemia 2003; 17: 941–59. Greten TF, Jaffee EM. Cancer vaccines. J Clin Oncol 1999; 17: 1047–60. Kolb HJ, Holler E. Hematopoietic transplantation: state of the art. Stem Cells 1997; 15 (Suppl. 1): 151–7; discussion 158. Laupacis A, Fergusson D. Erythropoietin to minimize perioperative blood transfusion: a systematic review of randomized trials. The International Study of Perioperative Transfusion (ISPOT) Investigators. Transfus Med 1998; 8: 309–17. Slavin S. Immunotherapy of cancer with alloreactive lymphocytes. Lancet Oncol 2001; 2: 491–8. Socie G, Stone JV, Wingard JR, Weisdorf D, HersleeDowney PJ, et al. Long-term survival and late deaths after allogeneic bone marrow transplantation. N Engl J Med 1999; 341: 14–21. Stovek J, Joseph A, Espine G, Dawson MA, Douek MC, Sullivan KM et al. Immunity of patients surviving 20 to 30 years after allogeneic or syngeneic bone marrow transplantation. Blood 2001; 98: 3505–12. Thomas ED. Does BMT confer a normal life span? N Engl J Med 1999; 341: 50–1.

Further reading

Web resources

Barker JN Weisdorf DJ, Defor TE, Blazar BR, Miller JS, Wagner JE. Rapid and complete donor chimerism in adult recipients of unrelated donor umbilical cord blood transplantation after reduced-intensity conditioning. Blood 2003; 102: 1915–19. Barrett AJ, van Rhee F. Graft-vs.-leukaemia. Baillières Clin Haematol 1997; 10: 337–55.

http://www.ibmtr.org International Bone Marrow Transplant Registry (IBMTR) and the Autologous Blood and Marrow Transplant Registry (ABMTR). Website of the International Allogeneic database and the Autologous database for North and South America. http://www.ebmt.org/ European Group for Blood and Marrow Transplantation (EBMT).

389

Chapter 34

Gene therapy Colin G. Steward and Marina Cavazzana-Calvo

The ability to cure disease by genetic manipulation is one of the most fascinating concepts in modern medicine. Although widely discussed in the early 1970s, ‘gene therapy’ underwent an explosion of popularity in the early 1990s. This was fuelled by improved methods of blood cell purification/ expansion, the development of more sophisticated vectors (the vehicles used to transport genes into cells) and the birth of the biotechnology industry. By 2003, more than 640 gene therapy protocols had been sanctioned, and more than 3500 patients recruited. This chapter concentrates on basic principles, methods of gene delivery and the major clinical benefits reported so far in monogenic diseases of the immune system. However, a considerable section is also devoted to potential uses in cancer therapy, since this has come to be the major focus of clinical trials and financial investment.

Basic concepts and definitions Somatic gene therapy

Research is concentrated on genetic manipulation of somatic cells, since manipulation of gametes is unethical and is prohibited in all countries. Transfection and transduction

The process of introducing extraneous genetic material into a cell is termed ‘transfection’ and may be performed either in the laboratory (in vitro) or by direct injection into tissue or blood (in vivo). The gene being transferred is termed the ‘transgene’. Once it has been introduced by a viral vector, the cell is said to be ‘transduced’. Genes may be trans390

ferred into cells using either chemical methods (transfection) or a wide variety of physical or viral vectors, summarized in Tables 34.1 and 34.2. Methods of gene transfer

Vectors are the vehicles used to carry genes into cells and are usually modified viruses. Although no perfect vectors exist, an ideal vector system for clinical purposes needs to have the following characteristics. • Highly efficient, transducing a large proportion of target cells • Result in stable integration so that the therapy is long-lasting. This can only be achieved by vectors, e.g. retroviruses and adeno-associated viruses (AAV), which insert their genetic material permanently into that of the host, so that it is replicated in all cell progeny. • Either transfect specific organs or cell types (targeted therapy) or transfect cells indiscriminately but ensure that the gene product is only expressed in cells requiring a therapeutic effect (tissuespecific or disease-specific expression). This necessitates the identification of physiological gene control mechanisms such as the locus control region (LCR) that promotes globin gene expression in red blood cells exclusively. In vitro techniques allow manipulation of target cells and include the following. • Cytokine stimulation of haemopoietic progenitor cells, in order to stimulate cell proliferation and improve retroviral transduction rate. • Selection of successfully transduced cells: the commonest methods involve introducing either a neomycin resistance (neoR) gene (which imparts resistance to successfully transduced cells when

Gene therapy Table 34.1 Major viral vectors used in gene therapy experiments.

Virus type

Advantages

Disadvantages

Retrovirus

Permanent integration

Small capacity for gene inserts (6–7 kb) Low titre (106–107 virus particles/mL), necessitating large volumes of stock Potentially pathogenic and carcinogenic Can only infect cycling cells

Adenovirus

Capacity for large gene inserts (7–36 kb) Can be prepared at high titre (1011–1012/mL) Infect non-dividing cells

Immunogenic, causing inflammatory reaction Lost on cell division because not permanently inserted into DNA

Adeno-associated virus

Insert preferentially on chromosome 19, a ‘safe’ area of DNA Non-pathogenic Can be prepared at high titre (106–1012 virus particles/mL)

May be poor for infecting HSC Low infection efficiency Small capacity for gene inserts (2–4.5 kb) Requires helper adenovirus for infection

Herpes virus

High potential as vectors for CNS-directed gene therapy Capacity for large gene inserts (10–100 kb)

Potentially pathogenic

Lentivirus

Permanent integration Able to divide non-dividing cells

Major concerns about safety

CNS, central nervous system; HSC, haemopoietic stem cell.

Table 34.2 Physical agents used in gene therapy experiments.

Category

Advantages

Disadvantages

Naked DNA

Directly targeted injection (e.g. to muscle via ‘gene gun’)

Many tissues inaccessible Low transfection efficiency Expression lost if cells divide

Liposomal DNA

Non-pathogenic Capable of transferring large genes

Low transfection efficiency Expression lost if cells divide

DNA–protein conjugates

Tissue targeting possible

May be degraded in circulation Low transfection efficiency Expression lost if cells divide

Oligonucleotides

Cheap and easily prepared Non-toxic Can target specific genes

Transient effect Non-specific binding Poor efficiency May be degraded in circulation

Ribozymes

Potential for highly specific RNA cleavage/repair Catalytic: each particle can perform multiple reactions

Developmental: may be problems of delivery, degradation, expression or specificity

391

Chapter 34

grown in the presence of a neomycin analogue called G418) or a gene for a fluorescence-activated cell sorter (FACS)-selectable marker (e.g. CD24 or green fluorescent protein). Selected cells may then be further expanded. In vivo techniques tried have included: • naked DNA coated onto gold particles and fired into tissue via a ‘gene gun’ (an inefficient method of gene transfer); • injected/inhaled adenoviral or AAV vectors; • protein–DNA conjugates, i.e. DNA coupled to proteins for which specific cellular receptors exist, in order to allow tissue targeting. Viral vectors

The major characteristics of the more frequently used viral vector systems are summarized in Table 34.1. Oncoretroviruses

The most widely used retroviral vectors have been those based on the Moloney murine leukaemia virus (MLV). The core structure of a retrovirus is shown in Fig. 34.1. These allow permanent integration of therapeutic genes and infect a wide variety of cell types. Depending on the envelope type of the virus it may only be able to infect cells of the species that it was originally isolated from or identified in (termed ‘ecotropic’) or to more broadly infect mammalian cells (‘amphotropic’).

LTR

gag

pol

env

LTR

Packaging signal (psi)

Fig. 34.1 The core structure of a retrovirus: the genome

comprises gag, pol and env genes (encoding core proteins, reverse transcriptase and viral envelope proteins, respectively) sandwiched between two long terminal repeat (LTR) elements. These genes can synthesize an empty virion but require a packaging signal (psi) to insert a copy of the viral genome, thus creating a virus capable of selfreplication.

392

The MLV-derived vector can only integrate successfully into the genome of cells which pass through mitosis. They can also only accommodate small genes (maximum size of incorporated DNA is 8 kb). Larger genes can still be utilized in their complementary DNA (cDNA) form, where the non-coding sequences (introns) have been spliced out, leaving only the coding segments (exons). This explains the use of cDNA rather than whole genomic DNA in most protocols. Problems are that (i) intronic sequences may contain control elements critical for transcriptional regulation and (ii) the limited capacity precludes the inclusion of large stretches of upstream and downstream DNA which may also contain regulatory sequences. Therapeutic viral particles (retroviruses containing a therapeutic gene) are synthesized in modified murine fibroblast cells, termed a packaging cell line. In the simplest form of packaging cell line (‘first generation’), the cells contain viral genome from which the packaging signal has been deleted. Instead this signal is ligated to a plasmid containing a therapeutic gene and transfected into packaging cells. A successfully transfected packaging cell will then release into overlying supernatant culture fluid viral particles containing the therapeutic gene, packaging signal and long terminal repeat (LTR) regions (which contain retroviral promoter elements) but no other replication-competent retroviral genes. These particles are capable of one round only of infection, thereby inserting the therapeutic gene randomly into host DNA. Retroviral vectors appear prone to silencing in humans at a much greater rate than in animals. This results in rapid loss of expression, although transgene can be shown to persist by genetic analysis such as polymerase chain reaction (PCR). Processes which may explain silencing include the following. • Methylation of promoter sites. • Position effects imparted by chromosomal sequences at their integration site: these may be reduced by introducing chromatin insulator elements (e.g. cHS4). • Immune responses against vector or transgene sequences: silencing occurs more slowly when immunosuppressant drugs are given.

Gene therapy

Lentiviruses

The most significant hope for widespread effective application of gene therapy resides with this group of vectors derived from human or animal immunodeficiency viruses. Lentiviral vectors have the following characteristics. • They can infect both dividing and non-dividing cells as their ‘preintegration complex’ (viral ‘shell’) can get through an intact nuclear membrane. Cell cycling is therefore not needed for infection of cells, allowing the possibility of in vivo gene therapy and avoiding differentiation or other deleterious effects on target cells by in vitro culture. • They show wide infectivity of non-dividing or terminally differentiated cells, including neurones, haemopoietic stem cells (HSC), muscle and liver cells. • They are associated with a low incidence of immune responses against vector. • Target specificity is provided by protruding membrane proteins in the lipid coat of the virus; these include gp120 which provides the CD4 T-cell specificity of human immunodeficiency virus (HIV). In the future it may be possible to engineer vectors specific for target cell type. • The cell specificity of HIV can be greatly broadened by exchanging the gene encoding gp120 for genes encoding other glycoproteins, e.g. G glycoproteins from vesicular stomatitis virus. This is called ‘pseudotyping’ the vector. As one example, a vector pseudotyped with glycoprotein from RD114 feline endogenous virus is relatively resistant to inactivation induced by human complement and has shown augmented transduction of human primary blood lymphocytes and CD34+ cells. • Lentiviral vectors can be designed as ‘selfinactivating’ vectors by making deletions in the LTR of the virus. These inactivate the LTR promoter and eliminate the production of vector RNA, with transcription being driven from an exogenous viral or cellular promoter that is inserted into the lentiviral vector. There is interest in two other classes of lentiviral vectors. • Non-human lentiviral vectors: examples are vectors derived from feline immunodeficiency virus (present in 2–20% of cats but never known

to have infected humans) and equine infectious anaemia virus. These may be less likely to elicit immune responses but have so far not been well developed for human applications. • Spumaviruses (foamy viruses): these are poorly characterized due to lack of association with human diseases, but can often be isolated from primate bone marrow cultures. They appear to have a larger packaging capacity, can infect more cell types and are better at transducing nondividing cells than MLV vectors. Adenoviruses

The major attractions of adenoviruses are as follows: • a high propensity to infect respiratory epithelium, neurones and hepatic cells, all major targets for gene therapy; • the ability to infect both dividing and quiescent cells; • the ability to be prepared in high titre, reducing the volume of stock solutions and glassware needed for transfection procedures; • large capacity, carrying up to 36 kb of transgene. A major drawback is that DNA is not integrated but held in the form of ‘episomes’, which are lost when the cell divides. This necessitates repeated treatments with a highly immunogenic virus. Adeno-associated virus

Major attractions of this vector are: • the propensity to integrate at a specific site on chromosome 19, i.e. 19q13 (an area lacking known tumour-suppressor genes); • the ability to transduce non-dividing cells; • the fact that it is naturally replication deficient (requiring coinfection with a helper virus such as adenovirus for productive replication); • they are less immunogenic than standard adenoviral vectors (although humoral responses may occur to capsid proteins). These features make AAV vectors a popular choice for in vivo gene therapy experiments. There are promising animal studies to suggest that infection of liver or muscle could act as a source of 393

Chapter 34

coagulation factors but so far no beneficial results in humans. However, it is not clear that human CD34 cells have appropriate surface receptors for AAV infection and enthusiasm is tempered by the small capacity (< 5 kb) of these vectors. Historically, the production of these vectors has relied on coinfection of helper cell lines with adenoviruses, requiring subsequent separation from the more immunogenic adenovirus. This has been overcome by cotransfection of producer lines with adenovirus helper genes. The problems of capacity and poor tropism for human haemopoietic cells may be improved by creation of vectors with chimeric capsids containing adenovirus type 35 fibres. Herpes viral vectors

The particular attractions of herpes-based vectors are: • their predilection for infecting nervous tissue; and • large capacity for inserted transgene (30–40 kb). Non-viral vectors (see Table 34.2) Cationic liposomes

• A plasmid carrying a therapeutic gene is linked by charge to the liposome surface. • Attractive for gene therapy in that there is widespread experience with liposomal drug formulations, they can be given repeatedly and the process is non-infectious, reducing regulatory problems. • Subject to degradation in vivo and gene expression is only transient. • Gene transfer efficiency is low, although it may be possible to improve this in the future, for example by coating the liposomes with polyethylene glycol (PEG) to slow clearance from the blood, and by conjugating them with antibodies to receptors on specific cells or viral cell fusion proteins. • Liposomes are typically 100 nm in size, but nanoliposomes (25 nm) coated with positively charged peptides are small enough to enter pores in the nuclear membrane.

394

DNA–protein conjugates

DNA is coupled via a linker arm to a protein recognized by a cell surface receptor. A typical example is transferrin–polylysine–DNA, where the transferrin will bind to receptors on hepatocytes. These vectors share many of the problems of liposomal vectors. Oligonucleotides and antisense approaches

Approaches that can neutralize the process of gene expression are of major interest in the gene therapy of cancer and HIV. Examples include the following. • Oligonucleotides: DNA fragments, typically 15–25 bases in length, of complementary sequence to either the normal, sense mRNA transcript (antisense therapy) or specific genomic DNA sequences (‘antigene’ therapy). Problems include nonspecificity of target binding, the requirement for large doses of injected oligonucleotides in order to attain adequate intracellular concentrations and degradation of oligonucleotides in the circulation (although this can be reduced by chemical modification). • Ribozymes: these are short segments of RNA which form complementary base pairing with mRNA and cleave the target in a highly specific manner. They then dissociate and repeat the process on other mRNA strands in a catalytic manner. Ribozyme technology is still in the early phase of development, but may even allow targeted repair of specific DNA mutations. • Small interfering RNA molecules: small (21–23 nucleotides) double-stranded RNA molecules were first identified in petunias but have become recognized as a powerful potential tool for gene therapy by post-transcriptional gene silencing in humans. One application may be in combating HIV infection. Targeted gene repair

Although homologous recombination can be performed in vitro where cells can be cloned and expanded, this is not possible on a therapeutic scale at present. However, there is growing interest

Gene therapy

in the use of ribozymes and chimeric DNA/RNA oligonucleotides to repair specific mutations in DNA.

Transducing haemopoietic stem cells Gene therapy using HSC has been greatly hampered by the: • inability to identify primitive stem cells with precision; • predominant dormant state of these cells (oncoretroviral transduction requires cells to proceed through the S phase of the cell cycle); • low expression of retroviral receptors on their cell surface. These problems have forced scientists to use cytokine cocktails to stimulate cell proliferation. These appear to differentiate progenitor cells, reducing the duration of the therapeutic effect, and succeed typically in transducing less than 1% of true primate/human stem cells (although this has been sufficient to correct several human immunodeficiency diseases). Scientists are now engineering vectors to contain proteins which bind specifically to haemopoietic cells.

Conditioning therapy In rare instances, spontaneous genetic mutation may revert the mutated gene in a cell to a normal genotype. If this confers a genetic advantage to that cell over diseased counterparts, the reverted cell will proliferate more successfully and produce progressive disease reversion. This phenomenon has been observed in several forms of severe combined immunodeficiency (SCID), Wiskott–Aldrich syndrome and Fanconi’s anemia and provides compelling evidence that gene therapy should be effective, even if only a minority of cells are transduced successfully. Similar evidence comes from bone marrow transplantation (BMT) where patients often engraft with a mixture of donor and recipient cells (‘mixed chimerism’) but eventually become 100% donor chimerae (e.g. thalassaemia, Fanconi’s

anaemia, SCID). This implies competitive advantage for normal (donor) cells over diseased (recipient) cells. After gene therapy such competitive advantage would be expected to favour progressive engraftment/expansion of gene-modified cells and avoid the necessity for chemoradiotherapy. However, this advantage typically appears to operate at the level of differentiated cells rather than the stem cell. In the absence of a selective advantage for the transduced cells, this can be achieved artificially by using conditioning therapy such as low-dose cyclophosmamide or busulfan.

Risks of gene therapy Immune responses

A recurring theme in the literature on gene therapy is the problem of immune responses hindering therapy. These may take two forms: Responses to the therapeutic protein

It is well recognized that tolerance to self proteins is established by 14 weeks in utero. Any new protein expressed after this time is perceived as foreign and cells expressing it become the target for an immune response. Transgene products can induce both humoral and cell-mediated responses, and the subsequent loss of transduced cells is well documented. Responses to the vector

Immune responses against viral vectors have also proved to be a major problem. These are a major problem in protocols using adenoviruses, especially where repeated administrations are required. In one trial of inhaled gene therapy for cystic fibrosis using the cystic fibrosis transmembrane conductance regulator gene (CFTR) this resulted in severe pneumonitis. The risks were most tragically demonstrated in 1999 by the death of an 18year-old American teenager, Jesse Gelsinger, who developed multiorgan failure after an adenoviral vector injection into the hepatic artery for ornithine decarboxylase deficiency. 395

Chapter 34

(a) Packaging cell Promoter

gag

Promoter

env

pol

Empty retrovirus particles Fig. 34.2 (a) ‘Third-generation’

(b) 1 Packaging cell transfected with retroviral plasmid DNA containing therapeutic gene

2 Retroviral transcript produced containing therapeutic gene

LTR

Therapeutic gene

LTR

PSI Promoter

gag

Promoter

env

pol

3 Packaged into empty retroviral particle

4 Therapeutic retroviruses released into supernatant

Risks associated with retroviral vectors Risk during production

In first-generation packaging cell lines, a single genetic recombination event will result in formation of replication-competent wild-type ‘helper’ virus, i.e. MLV. In an early gene therapy trial in rhesus monkeys, helper virus contamination resulted in some animals developing T-cell nonHodgkin’s lymphoma. This risk can be reduced by further splitting the retroviral genome on the chromosomes of packaging cells, a process exemplified in Fig. 34.2. Gene therapy regulatory authorities insist on scrupulous testing for helper virus in retroviral supernatant before clinical use. 396

packaging cell line for production of retroviral vectors: a mouse fibroblast cell line has been transfected with separate plasmids containing gag/pol and env genes, respectively, with 3¢LTR removed. Multiple recombination events are now required to produce live retrovirus containing gag, pol and env genes and an LTR. (b) The therapeutic gene, together with a packaging signal, is introduced on a separate plasmid and viral particles containing this gene are then produced and released into the supernatant. Although thought to be an unlikely event, formation of retroviruses capable of self-replication has been documented from such cells and meticulous testing for ‘helper’ virus therefore remains essential.

Risks during vector production and following administration

In vivo recombination events between retroviral vectors and endogenous retroviral sequences already present in human DNA could result in the formation of replication-competent retroviruses. In the case of lentiviral vectors based on HIV this could result in devastating new diseases, especially since a new virus would probably lack the CD4+ Tcell specificity of HIV. Self-inactivating vectors with modifications of the lentiviral LTR may overcome this risk. Random incorporation of therapeutic genes could disrupt potentially dangerous genes such as tumour-suppressor genes or activate protooncogenes. This has occurred in 2 of 14 children

Gene therapy

3 years after they received gene therapy for Xlinked SCID. In both cases the vector had inserted within or upstream of the LMO2 proto-oncogene, resulting in a T-cell lymphoproliferative disease that required chemotherapy for control. Selfinactivating lentiviral vectors are thought to reduce the risk of insertional mutagenesis.

Gene therapeutic approaches to cancer The past decade has seen rapid advances in the understanding of tumour immunology, particularly of mechanisms of immune evasion and the (relatively) specific antigens expressed by some tumours. Tumour antigen profiling is likely to advance at an even more rapid pace as DNA array profiling becomes more routine. Coupled with greater recognition of tumour-suppressor genes and transforming oncogenes, this has led to major interest in genetic therapeutic approaches to cancer. As a result, 80% of experimental protocols are directed to cancer rather than to the cure of monogenic diseases. It seems likely that combinations of immunological and genetic approaches with conventional therapy will be the next major step in improving survival. The particular role of these newer therapies may lie in eliminating the last vestiges of cancer, which have been reduced to a state of ‘minimal residual disease’ by conventional therapy, and in preventing relapse. Clinical application of many of these techniques is hampered by the fact that mouse models often falsely predict responses which do not then occur in humans, and because of our inability to target all of the cells of a tumour. Transducing all cells within a cancer may not be necessary because of a ‘bystander effect’, the death of untransduced cells surrounding a tumour cell that is dying as a consequence of genetic manipulation. This may result from intercellular transfer of harmful cytokines or prodrugs (see below) or due to immunological reactions, although the mechanisms are not well understood.

Gene marking

The integration of foreign ‘marker’ DNA sequences into cells in vitro allows their fate to be followed after reinfusion by PCR amplification of gene sequences or FACS detection of expressed proteins. This has allowed demonstration of the source of relapse when this follows autologous bone marrow transplantation for acute and chronic myeloid leukaemia and tracking of Epstein–Barr virus (EBV)-specific cytotoxic lymphocytes given to abrogate post-transplant lymphoproliferative disease (EBV-LPD). Suicide genes

This term is used to describe genes that convert non-toxic prodrugs into cytotoxic agents. The most widely employed has been thymidine kinase from herpes simplex virus (HSV-Tk). This phosphorylates the antiviral drug ganciclovir, producing a metabolite that interferes with DNA synthesis on subsequent cell division and leads to cell death. For example, HSV-Tk transduction of T cells administered following allogeneic BMT (for control of EBV-LPD or leukaemic relapse) has enabled reversal of acute graft-versus-host disease by administration of ganciclovir in some patients. However, there are various disadvantages to this approach: • the prolonged transduction processes required currently may cause loss of function in the target lymphocytes; and • transfected cells may be lost or silenced due to immune responses directed against HSV-Tk. There are many other candidate suicide gene/prodrug combinations, including cytosine deaminase (which converts 5-fluorocytosine to 5-fluorouracil) and P450-2B1 (which converts cyclophosphamide to the active metabolite 4hydroperoxycyclophosphamide). It is essential to target the suicide gene only to those cells that are to be killed. For this it is possible to take advantage of the enhanced mitotic rate of tumour cells or, in brain tumours, the fact that normal neurones do not divide. An alternative technique of tumour targeting is to put the suicide gene under the control of a promoter that is only 397

Chapter 34

active in tumour cells, e.g carcinoembryonic antigen in colorectal carcinomas. This is termed ‘gene-directed enzyme prodrug therapy’. Increasing the drug resistance of normal haemopoietic progenitors

It is possible to confer resistance to cytotoxic drugs by transducing them with genes such as MDR-1, O6-alkyl DNA transferase, dihydrofolate reductase or aldehyde dehydrogenase. If normal bone marrow cells are transfected with these genes and returned to the patient before intensive chemotherapy, this could potentially shorten the subsequent period of pancytopenia or allow dose intensification. The principal problem with this elegant approach is that the threshold of major toxicity for other organs is close to that of bone marrow, limiting the potential for dose intensification. Immunotherapeutic approaches

Tumours appear to impair immunological responsiveness by generating tumour antigen-specific tolerance and by inducing global immunosuppression. This may be explained in part by factors such as: • downregulation of expression of tumour antigens, major histocompatibility complex (MHC) class I or costimulatory molecules or progressive selection of tumour cells expressing low amounts of these proteins; • downregulation of the z-chain of the T-cell receptor (TCR) CD3 complex (a key protein in Tcell signalling) on tumour infiltrating lymphocytes; • production of immunosuppressive cytokines, such as transforming growth factor-b and interleukin (IL)-10; • induction of lymphocyte apoptosis by tumour cells using the Fas/Fas ligand (Fas L) pathway. Yet many tumours express potential novel protein targets (e.g. immunoglobulin and TCR idiotypes, fusion proteins secondary to chromosomal translocation) and tumour antigens at high concentration. This has encouraged a number of developments based either on: • preparing tumour-specific cytotoxic T cells ex 398

vivo (e.g. for trials in controlling disseminated melanoma or EBV-LPD); or • enhancing the immunogenicity of tumour cells, thus creating a ‘tumour vaccine’. Tumour vaccines

Tumour vaccines have been created by transducing tumour cells with genes encoding proteins involved in antigen presentation or in eliciting inflammatory reactions. After return to the animal these sometimes stimulate immunological antitumour responses that eradicate the injected cells and may even cause the death of metastatic non-transduced tumours or confer resistance to subsequent tumour injections. Major interest is currently focusing on the identification, characterization, expansion and reinjection of dendritic cells pulsed with tumour antigens or peptides. In humans, efforts have principally been concentrated on those tumours which show occasional spontaneous regression or which express fetal antigens not normally expressed after birth. These include melanoma, renal cell carcinoma and neuroblastoma. The commonest approach has been to transduce tumour cells with cytokine genes, such as those encoding IL-2 or granulocyte– macrophage colony-stimulating factor (GM-CSF), thereby making them the target of an inflammatory response. Some encouraging responses have been seen. Tumour infiltrating lymphocytes

There is much evidence that the lymphocytes contained within tumours are either themselves suppressed or actively suppressive of immune responses. Examples include: • an increased frequency of a regulatory phenotype (Treg: CD4+CD25+); • high production of suppressive cytokines (e.g. IL-4 and IL-10); and • reduced TCR z expression. These characteristics can be partially reversed by culture and transfection with proinflammatory cytokines such as IL-2. Homing and tumour responses to gene-modified cells have been largely disappointing.

Gene therapy

Manipulation of tumour-suppressor genes and oncogenes

Some of the earliest interest centred on supplementing tumours with mutated tumoursuppressor genes with normal (wild-type) gene copies, e.g. injection of wild-type p53 directly into bronchial carcinomas or gliomas, or into the peritoneal cavity of women with disseminated ovarian carcinoma. A major limitation of this approach is the necessity to genetically modify a large proportion of the cells in tumour nodules, which often contain large necrotic elements and have a compromised blood supply. Typically, effects are only seen close to the injection site. Where dominant oncogenes (e.g. myc, ras) are implicated, research is concentrated on antisense technologies (see above). Modifying angiogenesis

Tumours above 2 mm in diameter are critically dependent (for adequate supply of oxygen and nutrients) on the development of new blood vessels from an existing vascular network. This process, called angiogenesis, depends on a balance of cytokines responsible for stimulating and suppressing the growth of blood vessels. Angiogenesis presents an attractive target for interfering with tumour growth and spread, either by: • developing drugs that suppress endothelial cell responses to tumour-derived growth factors; or • directly inhibiting the pro-angiogenic activities of other cell types in tumours. Candidates include the angiogenesis inhibitors, endostatin and linomide, and a 50-kDa proteolytic fragment of fibrinogen.

Progress in gene therapy for monogenic haematological disorders Gene therapy was initially largely targeted towards inherited single-gene disorders, especially those governed by small, unregulated and easily transferable genes. This has resulted in detailed, highly publicized experiments concerning compar-

atively rare disorders, for example SCID in its Xlinked form (SCID-X1) or secondary to adenosine deaminase (ADA) deficiency (ADA-SCID), and cystic fibrosis. However, there is now encouraging animal work on commoner disorders involving large genes with complex control mechanisms (e.g. haemoglobinopathies). It should be remembered that truly successful gene therapy of such conditions depends on the following caveats. • It is only possible for recessively inherited diseases. In dominant diseases the abnormal proteins encoded are essentially toxic. Successful approaches to dominant diseases will therefore require widespread blockade of RNA production, a much more difficult task than gene supplementation. • It needs permanent gene integration for lasting cure without repeated rounds of therapy. • It needs transgene expression without eliciting immune responses. Some workers have proposed continuing immunosuppression after gene therapy in order to reduce this risk. However, this carries long-term risks and is not viable. • It may be difficult in diseases where a high proportion of cells must be modified, e.g. congenital erythropoietic porphyria, where a majority of red cell precursors would need to be modified to prevent the excess production of toxic porphyrins (there is no evidence of selective advantage for normal red cell precursors over diseased cells). Important considerations will be the number and type of cells that need to be transduced for success. Although the difficulty of transducing HSC has already been discussed, some haematological diseases may be curable by transfecting liver or muscle cells, or by the use of implantable reservoirs of gene-corrected cells. If this improved factor VIII in a haemophiliac by 5–10%, this could dramatically improve lifestyle. Unfortunately, no study has managed to show so far that this threshold is reachable. Immunodeficiency diseases

Immunodeficiency diseases represent ideal targets for gene therapy for the following reasons: • they are monogenic; 399

Chapter 34

• responsible genes have been identified in a number of conditions; • they are very severe diseases, being lethal within the first year in absence of any treatment; • symptoms can be largely alleviated or abolished by BMT from matched siblings without the use of conditioning chemotherapy; • full chimerism is not required for cure; • normal donor cells (especially T lymphocytes) carry competitive advantage over diseased cells; • immune response to vectors and therapeutic genes are less likely. ADA-SCID

ADA is responsible for the detoxification of metabolites in the purine salvage pathway. Deficiency causes accumulation of deoxyATP, which inhibits cell division and causes death by apoptosis. This results in early block of lymphoid differentiation and decreased survival of mature T cells and early lymphoid precursors, resulting in SCID. The most severe symptoms of this disease can be ameliorated by use of infused enzyme (modified with PEG in order to enhance its persistence in the circulation). This allows patients lacking an appropriate donor for BMT to be kept in reasonable health while attempting gene therapy for long-term disease correction. The first administration of human gene therapy was to a 4-year-old girl with ADA-SCID on 14 September 1990 by W. French Anderson and colleagues at the National Institutes of Health, Bethesda, USA. Since that time, five clinical trials have been conducted in this condition. All cells administered have been transduced in vitro by MLV retroviral vectors. In the early studies T cells were transduced but in more recent studies CD34+ HSC from marrow or cord blood have been used. Major findings are as follows. • Results were generally poor in children maintained on PEG-ADA following gene therapy. This is probably because accumulation of toxic metabolites may offer a selective advantage to cells that produce adequate vector-derived ADA and this advantage is lost in the presence of a therapeutic level of PEG-ADA.

400

• Transduced T cells have lasted for more than 12 years even after in vitro activation to induce their cycling. Therefore, at least in these cells, the MLV promoter can drive transgene expression for long periods without gene silencing. • The best results were seen in children who received transduced bone marrow HSC and who were given mild myeloablation with busulphan before receiving modified cells. Sustained HSC engraftment was followed by differentiation into multiple lineages, increased lymphocyte counts, improved immune function and reduction in toxic metabolites. • However, even immunodeficient patients were able to mount immune responses to both the retroviral envelope and the fetal calf serum used in transfection procedures. SCID-X1

• Accounts for 50–60% of all cases of SCID. • Results from mutation of the gene encoding the common g-chain subunit of the haemopoietic cytokine receptor family (IL-2, IL-4, IL-7, IL-9, IL15 and IL-21). • Between 1999 and 2002, 14 children received HSC retroviral gene therapy in Paris or London. • Thirteen patients developed substantial or complete correction of T- and B-cell function (although intriguingly reconstitution of natural killer cell function has mostly been poor in the long term). These patients are alive and well without any additional treatment. • Two children developed uncontrolled Tlymphoproliferation due to insertional mutagenesis inducing hyperexpression of the LMO2 protooncogene. They have been treated using acute lymphoblastic leukemia chemotherapy protocols and are in complete remission. Chronic granulomatous disease

Chronic granulomatous disease (CGD) results from mutations in any of four genes encoding essential subunits of respiratory burst NADPH oxidase, the enzyme complex required for the production of reactive oxygen intermediates in phago-

Gene therapy

cytes. The absence of oxidants results in recurrent bacterial and fungal infections and development of inflammatory granulomas. The potential for successful gene therapy in CGD is highlighted by the finding that female carriers with only 5–10% NBT-positive cells (due to highly skewed X-inactivation) are usually healthy and the knowledge that post-BMT patients with mixed chimerism and as little as 30% donor neutrophils remain well. Mouse models of CGD transplanted with bone marrow cells transduced with gp91phox (the gene defective in X-linked CGD) after lethal irradiation show neutrophil expression at 10% and superoxide production at approximately one-third that of wild-type neutrophils. Human phase I clinical studies in CGD patients have yet to produce clinically beneficial numbers of corrected neutrophils for significant periods. Current trials are focusing on techniques that allow higher levels of gene transfer efficiency (50–80%) into CD34+ haemopoietic progenitor cells and use of conditioning chemotherapy. Haemoglobinopathies

b-Thalassaemia and sickle cell diseases should be amenable to gene therapy if normal b-globin or anti-sickling genes, respectively, could be expressed at sufficiently high levels in the red blood cell lineage. An attractive approach for sickle cell disease is introduction of a g-globin gene, since g-globin is a much stronger inhibitor of HbS polymerization than b-globin. Unfortunately, progress in developing gene therapy for these disorders has been slow due to: • the large size of the a- and b-globin genes; • poor transduction efficiency of HSC by oncoretroviral vectors; • complex mechanisms which control globin gene expression, notably the existence of a specific LCR; • instability of b-globin/LCR retroviral vectors partly caused by unwanted splicing of retroviral RNA before incorporation into virions; • gene expression from retroviral vectors is position dependent, being influenced by elements in surrounding chromatin.

Progress has come with the following observations. • The use of lentiviral vectors incorporating RNA export elements that allow incorporation of more genetic material and larger LCR segments; these approaches have allowed dramatic improvement in mouse b-thalassemia and SCID models. • Identification of one amino acid residue from the g-globin gene that appears to be responsible for most of the anti-sickling properties, and incorporation of this in a b-globin gene variant. In mice this has produced erythroid-specific accumulation of anti-sickling protein in most red blood cells. Haemophilia

Haemophilia A and B make attractive candidates for gene therapy because of the dramatic improvements in quality of life which result from relatively small increments in the levels of clotting factors (to 5% or above). Clinical effect would not be dependent on transducing HSC and might be achieved by targeting either liver or muscle. Progress so far in these diseases includes the following observations. • Promising demonstrations of effect in dog models of haemophilia A and B using portal venous or multiple intramuscular injections and AAV vectors. Factor levels up to 2–4% of normal have been seen for up to 14 months after the gene therapy procedure. • Evidence of effective gene transfer also using high-capacity adenoviral or lentiviral vectors. • Phase I and II studies in humans have showed only limited efficacy (with plasma levels less than 1% of normal) but no significant toxicities.

Gene therapy for HIV infection Potential techniques of gene therapy for HIV infection include the following. • Use of antisense constructs or ribozymes which can negate HIV gene expression. In vitro these are capable of substantial reduction, though not cessation, of HIV replication. • Infection of CD4+ cells with HIV-based vectors

401

Chapter 34

containing HSV-Tk under the control of the HIV LTR promoter. If these cells subsequently become infected with wild-type HIV, HSV-Tk is upregulated and renders the cell sensitive to killing by ganciclovir. • Expression of small inactivating RNA species from vectors. This could interfere with production of proteins involved in HIV-1 infection (e.g. CCR5 coreceptor) and so reduce the rate of lymphocyte infection.

Vascular diseases In occlusive vascular disease stimulation of angiogenesis is being investigated widely. • The safety and tolerability of therapeutic angiogenesis by gene transfer has been demonstrated in phase I clinical trials. • Evidence of efficacy from early phase II studies of angiogenic gene therapy for ischaemic myocardial and limb disease is limited. • An alternative strategy to the use of transgenes encoding angiogenic growth factors is therapy based on transcription factors such as hypoxia inducible factor-1a, which can regulate the expression of multiple angiogenic genes. • Large, randomized, placebo-controlled phase II and III trials will be required to ascertain the value of therapeutic angiogenesis for ischaemic cardiovascular disease.

Future directions There are two major areas of challenge if gene therapy is to become a routine component of haematological medicine. Firstly, scientists need to strive yet further to produce efficient nonimmunogenic vectors capable of carrying large pieces of genetic material into primitive HSC. Ideally these vectors should be sufficiently safe for direct injection into the bloodstream or organs.

402

This will depend critically on a more thorough understanding of the mechanisms of gene integration and control. Secondly, regulators have the unenviable task of deciding what constitutes acceptable risk in the USA for HIV patients.

Further reading Brenner S, Malech HL. Current developments in the design of onco-retrovirus and lentivirus vector systems for hematopoietic cell gene therapy. Biochim Biophys Acta 2003; 1640: 1–24. Ciceri F, Bordignon C. Suicide-gene-transduced donor Tcells for controlled graft-versus-host disease and graftversus-tumor. Int J Hematol 2002; 76: 305–9. Hacein-Bey-Abina S, Le Deist F, Carlier F et al. Sustained correction of X-linked severe combined immunodeficiency by ex vivo gene therapy. N Engl J Med 2002; 346: 1185–93. Hacein-Bey-Abina S, Von Kalle C, Schmidt M et al. LMO2associated clonal T cell proliferation in two patients after gene therapy for SCID-X1. Science 2003; 302: 415–19. Howe S, Thrasher AJ. Gene therapy for inherited immunodeficiencies. Curr Hematol Rep 2003; 2: 328–34. Humrich J, Jenne L. Viral vectors for dendritic cell-based immunotherapy. Curr Top Microbiol Immunol 2003; 276: 241–59. Long MB, Jones JP III, Sullenger BA, Byun J. Ribozymemediated revision of RNA and DNA. J Clin Invest 2003; 112: 312–18. Persons DA, Nienhuis AW. Gene therapy for the haemoglobin disorders. Curr Hematol Rep 2003; 2: 348–55. Richter J, Karlsson S. Clinical gene therapy in hematology: past and future. Int J Hematol 2001; 73: 162–9. St George JA. Gene therapy progress and prospects: adenoviral vectors. Gene Ther 2003; 10: 1135–41. Thomas CE, Ehrhardt A, Kay MA. Progress and problems with the use of viral vectors for gene therapy. Nat Rev Genet 2003; 4: 346–58. VandenDriessche T, Collen D, Chuah MK. Gene therapy for the haemophilias. J Thromb Haemost 2003; 1: 1550–8. Wall NR, Shi Y. Small RNA: can RNA interference be exploited for therapy? Lancet 2003; 362: 1401–3.

Chapter 35

Recombinant antibodies and other proteins Marion Scott

Introduction Many potentially useful human proteins for therapeutic, diagnostic and research use are expressed in the body at very low concentrations, and it is difficult, if not impossible, to isolate them by conventional biochemical methods. Other proteins, such as antibodies of a particular specificity, are difficult to purify from a complex mixture of very similar proteins. However, once the gene encoding a protein has been cloned and sequenced, it becomes possible to express the protein at high concentrations, using virally derived expression vectors designed to produce full-length proteins at high levels in various different in vitro culture ‘host’ cell systems. Some blood proteins, such as the clotting factors to treat haemophilia, have been efficiently purified by fractionation of pooled human plasma, but have been shown to have the potential of transmitting diseases, such as human immunodeficiency virus (HIV) and hepatitis C virus (HCV). The cloning and expression of these proteins has led to the availability of recombinant clotting factors for the treatment of haemophilia, with reduced risk of infection. As the recombinant clotting factors are grown in vitro, there is also the advantage of an unlimited supply of constant guaranteed product. Similar drivers have led researchers to try to develop recombinant replacements for specific immunoglobulins currently fractionated from high-titre blood donations, such as anti-D and anti-varicella-zoster virus (VZV). Some concern has been expressed about the safety of such recombinant products, as they may potentially contain viruses or other infectious agents arising from the host cells used to express the protein, or the culture medium components

used to grow the host cells. Increasing awareness of the risks from pooled polyclonal blood products have been heightened by the variant Creutzfeldt– Jakob disease (vCJD) crisis in the UK. Are the potential risks from such biotechnology products any worse than the risks from blood products derived from pooled human plasma? Apart from cloning and expressing such naturally occurring proteins, it is possible using recombinant DNA technology to produce modified forms of the proteins that do not occur naturally, that might have desired therapeutic effects or diagnostic advantages. Indeed, the technology has been used to produce totally novel recombinant proteins for particular applications.

General methods for recombinant protein expression Choice of the host cell system to use for recombinant protein expression relies on several factors. Bacterial expression systems, such as Escherichia coli are the cheapest, simplest and most effective, but cannot be used for many types of human proteins that require eukaryotic posttranslational modifications for biological activity, e.g. glycosylation. Most of the enzymes used in recombinant DNA technology, e.g. many restriction enzymes, DNA polymerases, DNA ligases, polynucleotide kinase and reverse transcriptases, are now produced as recombinant proteins themselves. These enzymes are readily produced in E. coli. The ready availability of these recombinant proteins has paved the way for the routine use of recombinant DNA technology in every biological discipline. In addition, 403

Chapter 35 Table 35.1 Production systems for recombinant mammalian proteins.

System

Cost

Production time scale

Scale-up capacity

Product quality

Glycosylation

Contamination risks

Bacteria Yeast Plants

Low Medium Low

Short Medium Long

High High High

Low Medium High

None Incorrect Some differences

Insect cells Mammalian cells Transgenic animals

Medium High High

Medium Long Very long

High Low Low

Medium Very high Very high

Incorrect Correct Correct

Endotoxins Low risk Low risk, but environmental concerns Low risk Animal viruses Animal viruses

some simple eukaryotic proteins, such as protein hormones that are medically important, e.g. insulin, growth hormone, granulocyte colonystimulating factor (G-CSF), can be produced in bacteria and in sufficient quantities for use as therapeutic agents. Toxicity to the host organism can be a problem if a gene is overexpressed at too high a level, so it is important to control levels of expression. Strong promotors, which can be easily regulated, are used to overcome the problems of toxicity and yet still maintain high levels of expression and protein production. Two promoter systems are in common use for expressing proteins in E. coli, the lac complex and the T7 late promoter system. Prokaryotes lack the enzymes that catalyse many of the post-translational modifications found on eukaryotic proteins. Proteins produced in prokaryotes may not be folded properly and/or can be insoluble, forming inclusion bodies. It can be very difficult to resolubilize proteins found in inclusion bodies and restore biological activity. Research is currently underway to produce genetically modified strains of yeast (Pichia pastoris) that have human glycosylation pathways. Insect cells have also been used for recombinant protein expression, using baculoviral vectors. Again, there are issues about the correct folding and glycosylation of mammalian proteins in these systems, and attempts have been made to produce insect cells that have been transfected with mammalian glycosylation enzymes. Other workers have expressed recombinant proteins in plants and plant cell cultures, but this again 404

raises issues about correct folding and glycosylation of proteins, and introduces further concerns about environmental containment of genetically modified crops. Transgenic animals have also been produced, with targeted production of recombinant proteins in milk. A comparison of different production systems for recombinant proteins is shown in Table 35.1. For many types of human proteins, expression in a mammalian system is the best option, as this is the approach most likely to yield soluble, biologically active proteins, although it is considerably more expensive than expression in E. coli, yeast or insect cells. Cell lines commonly used are NS0 (mouse myeloma), CHO (chinese hamster ovary) and COS-7 (African green monkey fibroblast). A number of techniques have been developed for rapid one-stage purification of recombinant proteins. Epitope tags are short amino acid sequences for which commercial monoclonal antibodies are available, and can be placed anywhere within the protein where it will not disrupt the protein’s function. It is also common to create fusion proteins, i.e. to create a single open reading frame that encodes a well-characterized protein such as glutathione S-transferase (GST) together with the sequence of the protein of interest. The most popular tag systems for the purification of recombinant proteins in mammalian expression systems are GST, His6-tag, myc and FLAG. When the tag protein is produced, the protein of interest is produced as well, as one fusion or chimeric protein. Fusion proteins are useful because they enable rapid purification by affinity chromatography, and

Recombinant antibodies Restriction site Promoter

Tag

2 Restriction endonuclease, ligate with insert

Gene

Antibiotic resistance gene 1 Expression vector 3 Express fusion protein Fig. 35.1 Production of recombinant

fusion proteins.

the fused tag can be removed after purification using a specific protease. Some workers have used IgG Fc domain tags, and used protein-A and protein-G affinity matrices. Plasmids (expression vectors) used for expression commonly contain a viral promoter sequence, an antibiotic resistance gene, a fusion tag sequence and a restriction endonuclease site for insertion of the coding sequence of interest (Fig. 35.1). cDNA coding for the protein sequence of interest is normally derived by reverse trancriptase polymerase chain reaction (RT-PCR) from cells expressing the protein, using sequence-specific primers to amplify the region required. This cDNA is then inserted into the expression vector, and used to transfect a mammalian cell line. Growth in medium containing the antibiotic to which the vector codes resistance results in selection of transfected cells only. Production of the fusion protein can then be detected using antibodies to the fusion tag sequence, and the fusion protein purified and characterized. Some expression vectors do not insert into the host cell nuclear material and give rise to transient expression. Other vectors insert into the host cell DNA and give rise to stable expression. Integration into the host genome is random, such that it is worth screening different clones of infected cells, as some may show higher levels of expression than others. Different clones also vary in their stability of secretion of the protein, such

protein

Tag

that several high-producing clones should be kept in culture over a period of several months and secretion monitored. Once a high-producing stable clone has been isolated, master and working cell banks of the clone should be frozen in liquid nitrogen, and each batch of the recombinant protein grown from one vial of the working cell bank. This process ensures that every batch of the recombinant protein produced is identical.

Recombinant antibodies Limitations of rodent monoclonal antibodies

Conventional monoclonal antibody technology uses immunization of mice or rats with antigen to yield hyperimmunized spleen cells, which are then fused with non-secreting myeloma cell lines to yield hybridoma cell lines that can be grown in vitro to produce monoclonal antibodies. Effectively, the fusion process inserts the DNA from the spleen cells into the myeloma cells. While many such conventional monoclonal antibodies were very successfully developed into diagnostic reagents (such as the high-avidity anti-A and antiB now used routinely worldwide for blood grouping), it was not possible to produce antibodies of certain specificities in rodents, and attempts to use rodent monoclonal antibodies in humans as therapeutics rapidly ran into problems, as the recipients developed a strong human anti-rodent response, 405

Chapter 35 Table 35.2 Success rates for

1980–82 1983–85 1986–88 1989–91 1992–94 1995–97 1998–2000 All Murine Chimeric Humanized

Total monoclonal antibodies

Discontinued

Approved

Completed (%)

Success (%)

2 9 33 34 41 33 34 186 49 23 59

1 8 29 29 23 12 2 104 34 13 15

1 0 2 2 5 0 0 10 1 4 5

100 89 94 91 68 36 6 61 71 74 34

50 0 6 6 18 0 0 9 3 24 25

monoclonal antibodies entering clinical trials.

which rapidly cleared the antibodies from the body.

pipeline (see Tables 35.2 and 35.3 and Figs 35.2 and 35.3).

Humanizing rodent monoclonals

Human recombinant antibodies

The early promise of monoclonal antibodies as therapeutics was not realized, and many became disillusioned with the concept of the ‘magic bullet’. The success rate of rodent monoclonal antibodies that entered clinical trials was only 9% over the 20 years from 1980 to 2000, and only 10 were approved as products out of 186 entering trials (Table 35.2). However, when the possibility of making recombinant antibodies became available in 1986, things rapidly changed. Using recombinant DNA technology, it was possible to replace the mouse constant domains of antibodies with corresponding human domains, and express these chimeric recombinant immunoglobulin molecules in myeloma cell lines. There was far less human anti-mouse response to such antibodies, such that 4 of 23 chimeric antibodies entering clinical trials have now been approved as products (Tables 35.2 and 35.3; Figs 35.2 and 35.3). Further engineering work also allowed the replacement of the framework regions of the mouse variable domains with human framework regions, resulting in virtually fully humanized antibodies. Such antibodies are proving to be very successful, with five already approved as products and many others in the

Human circulating B cells can be selected from immune individuals and transformed into cell lines that can be grown in culture by transformation with Epstein–Barr virus (EBV). The cDNA coding for the antibodies can then be derived from these cells using RT-PCR, ligated into expression vectors and expressed in a suitable mammalian host cell line. Alternatively, phage display technology can be used. Bacteriophage that infect E. coli are modified such that they carry the cDNA encoding for antibody variable domains, while at the same time they express the antibody protein on their surface. This permits in vitro selection of antibodies of the required specificity, and then expansion in E. coli. RT-PCR is used to amplify all the heavy and light chain variable domains in a buffy coat sample. PCR is then used to assemble these randomly into VH and VL pairs, by inclusion of DNA encoding for a flexible linker chain between the heavy and light chain domains. A ‘tag’ sequence is also included, normally c-myc, to aid detection and purification. These linked heavy and light chain domains are known as single-chain Fv (scFv) (Figs 35.4 and 35.5). The scFv constructs are then

406

Recombinant antibodies Table 35.3 Therapeutic monoclonal

antibodies approved by the Food and Drug Administration, USA.

Name

Specificity

Trade name

Company

Type

Approval

Muronomab Abciximab Rituximab Daclizumab Basiliximab Palivizumab Infliximab Trastuzumab Gemtuzumab Alemtuzumab

CD3 GPIIb/IIIa CD20 CD25 CD25 RSV TNF-a HER-2 CD33 CD52

Orthoclone ReoPro Rituxan Zenapax Simulect Synagis Remicade Hercetin Mylotarg Campath

Ortho Centocor Genentech Hoffman La Roche Novartis MedImmune Centocor Genentech Wyeth-Ayerst Milennium/ILEX

Murine Chimeric Chimeric Humanized Chimeric Humanized Chimeric Humanized Humanized Humanized

1986 1994 1997 1997 1998 1998 1998 1998 2000 2001

RSV, respiratory syncytial virus;TNF, tumour necrosis factor.

12 10 8

Murine Chimeric Humanized

6 4 2

00 20

98 19

96 19

94 19

90

88

86

92 19

19

19

19

84 19

82 19

19

Fig. 35.2 Number of monoclonal

80

0

antibodies entering clinical trials.

ligated into a phage display vector. The scFv domain is ligated into the vector next to regions that code for the PIII phage coat protein. The recombinant phage then express the scFv protein alongside their PIII coat protein at the tip of the phage (Fig. 35.6). Phage libraries can be panned against antigens, and those phage selected that are displaying scFv bind to the antigen. Selected phage are eluted from the antigen, expanded by culture in E. coli, and then repanned against antigen. Selected human scFv can then be removed from the phage vector and can be ligated to cloned human IgG constant domains to express full-length human recombinant antibody molecules. One large advantage of this approach is that antibodies can be derived from phage display libraries made

from non-immunized individuals, and that normally restricted antibodies (e.g. anti-self) can be derived. Human recombinant monoclonal anti-D

Despite the overall success in the production of rodent monoclonal antibodies to human ABO blood group antigens, no such monoclonal antibodies have been produced to the Rh antigens. Analyses of immunoglobulin gene usage in human monoclonal anti-D have shown that it is very restricted: it is possible that the rodent strains used lacked the appropriate immunoglobulin genes. It may also be that Rh antigen processing and presentation in rodents is different to that in humans. 407

Chapter 35

PCR RNA

PCR assembly V H VH VH VL VH

VL

VH

VL

VH

VL

VL 108 B lymphocytes

VL V-gene repertoires

VL VL

Clone108 clones scFv phage library

scFv-Gene repertoires

Fig. 35.4 Generation of scFv phage libraries. PCR,

polymerase chain reaction Mouse

Chimeric

Human

Humanized

Fig. 35.3 Chimeric/humanized antibodies.

It is also possible that the D antigen is very similar to a rodent antigen, such that rodents do not respond because the lymphocytes recognizing the common structure have been deleted from their repertoire. Various different approaches have been developed to produce human monoclonal antibodies specific for RhD. Early work used the immortalization of human B cells by infection with EBV. However, although it is relatively easy to establish lymphoblastoid cell lines producing specific antibody in this way, antibody production is commonly lost on expansion of the culture. Similarly, specific antibody production 408

is often lost during cloning. Repeated selection for cells producing antibody using antigen (by rosetting with RhD-positive red cells) is required to produce stable cell lines. Improvements in the stability of human cell lines have been achieved by back-crossing human anti-D-secreting EBV lines to a mouse–human heterohybridoma line or to a mouse myeloma line. Use of these approaches has enabled the production of a large number of blood group-specific human anti-Rh monoclonal antibodies. More recently, various molecular biology techniques have been used to produce recombinant human antibodies with Rh specificity. Single-chain human Fv specific for Rh antigens have been produced by panning phage display libraries of scFv derived from non-immune donors. Fab fragments with RhD specificity have been produced by panning Fab libraries from a hyperimmune donor, and the Fab fragments have been converted to fulllength immunoglobulin molecules by cloning the variable regions into expression vectors containing genomic DNA encoding the immunoglobulin constant regions. The resultant IgG1 constructs have been successfully expressed in CHO cells. DNA coding for anti-D in lymphoblastoid cell lines and heterohybridomas has been isolated, modified and expressed in rodent myeloma cell lines. Another group has expressed anti-D DNA in a baculovirus–insect cell expression system. Candidate monoclonal anti-Ds for immunoprophylaxis are selected, firstly, on their ability to bind to the RhD antigen via the Fv part of the molecule

Recombinant antibodies myc tag

link VH

VL

30 kDa

VL

Ag

NH2

Fig. 35.5 Structure of scFv.

myc

p lll coat protein

VH

VL

p vlll coat protein DNA

Fig. 35.6 scFv displayed on phage surface.

and, secondly, on their ability to interact with Fc receptors via the Fc part of the molecule to bring about immunomodulation. The exact mechanism of immunosuppression by anti-D is not known, but it is clear that it involves interaction of anti-D with Fc receptors. To be effective, prophylactic antibody must be capable of not only binding to

COOH

VH

the RhD antigen on the red cells via its Fv regions, but also interacting with the effector cells of the immune system via its Fc region. Selection of recombinant monoclonal anti-D for therapeutic use therefore depends not only on the antigen specificity and avidity of the monoclonal antibody but also its functional activity in interacting with effector cells. The exact mechanism of immunosuppression by anti-D is not known, but it is clear that it involves interaction of anti-D with Fc receptors. To suppress immunization, IgG-coated red blood cells need to be rapidly cleared from the maternal circulation and localized in the spleen. It has been suggested that D antigen-specific B cells in the spleen are then deactivated by the simultaneous binding of the Fc region of the anti-D to FcgRIIb together with binding of the B-cell receptor to the D antigen. Interactions of anti-D with FcgRI, FcgRIIb and FcgRIIIa may thus all be required for effective immunosuppression. IgG monoclonal anti-D antibodies have been evaluated in various in vitro systems to test how effective the antibodies are at interacting with immune system effector cells. Each assay tests efficacy at binding to different Fc receptors. Rosette formation of sensitized cells with monocytes and phagocytes, adherence of sensitized cells to monocyte monolayers, and chemiluminescent measurements of the oxidative burst caused when monocytes react with sensitized red cells are all in vitro measures of interaction with FcgRI. Measure409

Chapter 35

ments of antibody-dependent cellular cytotoxicity (ADCC) by radiolabelled chromium release from natural killer cells measures interaction with FcgRIIIa. It is not clear at present how well performance in these various in vitro assays will predict in vivo efficacy. Some of the functional activities may be dependent on glycosylation of the antibodies, and therefore dependent upon the host cell line used for expression of recombinant antibodies. For example, expression of an anti-D in NSO cells produced a recombinant antibody with significantly lower ADCC activity than the same antibody expressed in CHO cells. Compliance with regulatory requirements

Regulatory bodies in the USA and Europe have guidelines for the manufacture and production of recombinant antibodies for therapeutic use. These have been harmonized under the auspices of the International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH). Most of the guidelines are aimed at ensuring the microbiological safety of biotechnology products derived from cell lines of human or animal origin. The risk of viral contamination is a feature common to all biotechnology products produced from cell lines. Such contamination could have serious clinical consequences and can arise from the contamination of the cell sources themselves or from adventitious introduction of virus during production. To date, however, biotechnology products derived from cell lines have not been implicated in the transmission of viruses. Nevertheless, it is expected that the safety of these products with regard to viral contamination can be reasonably assured only by the application of a virus-testing programme and assessment of virus removal and inactivation achieved by the manufacturing process. Three principal, complementary approaches have evolved to control potential viral contamination of biotechnology products. 1 Selecting and testing cell lines and other raw materials, including media components, for the absence of undesirable viruses which may be infectious and/or pathogenic for humans.

410

2 Assessing the capacity of the production processes to clear infectious viruses. 3 Testing the product at appropriate steps for the absence of contaminating infectious viruses. The current regulations state that ‘where a known human pathogen is identified, the product may be acceptable only under exceptional circumstances’. This statement may mean that production of monoclonal anti-D directly from EBV-transformed human cells may not be acceptable to the regulatory authorities. However, technology is now readily available to express antibody DNA from EBV-transformed cells in other cell types, and the guidelines further state that Cell lines such as Chinese hamster ovary (CHO), C127, baby hamster kidney (BHK) and murine myeloma cell lines have frequently been used as substrates for drug production with no reported safety problems related to viral contamination of the products. For these cell lines in which the endogenous particles have been extensively characterized and clearance has been demonstrated, it is not usually necessary to assay for the presence of the non-infectious particles in purified bulk product. From the guidelines it would appear that expression of anti-D monoclonal antibodies in CHO cells would be the easiest and safest method of production. Clinical trials

In the first clinical trial of monoclonal anti-D, the IgG1 antibody used (UCHD4) did not promote phagocytosis by purified monocytes and did not induce accelerated clearance of Dpositive cells. In a later trial, two monoclonal anti-D antibodies, BRAD-3 and BRAD-5, were selected because of their high activity in in vitro functional assays, high avidity and specificity for the immunodominant epitope region of the RhD antigen. Initial studies in D-negative male volunteers showed expected half-lives and pharmacokinetics after injection. Further studies on the antibodies administered with 51 Cr-labelled D-positive red cells demonstrated

Recombinant antibodies

accelerated red cell clearance in all subjects and provided preliminary evidence for protection from immunization. It is clear from these clinical trials to date that recombinant anti-D has the potential to replace polyclonal prophylactic anti-D. There is a case for universal antenatal prophylaxis if sufficient supplies of anti-D are available. How quickly recombinant anti-D becomes available will be largely determined by commercial investment and regulatory procedures. There have been concerns about the safety of biotechnology products in general, and trying to develop a biotechnology product for routine administration to healthy, young, pregnant women has been seen as a high commercial risk, that some are not prepared to take. Cost of litigation, if something went wrong, would be massive. As recombinant antibody products become more established, with a good safety record, such concerns will hopefully diminish.

model. We are currently engineering human chimeric versions of these mouse antibodies to progress this work into clinical trials for the potential treatment of vCJD. Anti-HCV recombinant antibodies

Two mouse monoclonal antibodies that beween them cover virtually all strains of HCV have been produced by conventional monoclonal technology, by immunizing mice with synthetic peptides corresponding to known neutralization sites on the virus. The antibodies neutralize the virus in vitro. Human chimeric versions of these antibodies are now being constructed before proceeding to clinical trials. Anti-HPA-1a recombinant antibodies

In a similar fashion to anti-D, recombinant antiVZV antibodies have been produced by DNA technology. Human cell lines were derived by EBV transformation of selected B cells from immunized individuals. cDNA coding for the antibodies has been expressed in CHO cells. Antibodies have been shown to neutralize VZV in vitro, and studies are currently underway in a guinea-pig model of the disease prior to clinical trials.

scFv specific for the human platelet antigen HPA1a have been derived from a phage display library prepared from an individual with antibodies to HPA-1a. The scFv has been ligated to scFv specific for the RhD antigen on red blood cells, and the novel bispecific recombinant antibody can be used in a mixed passive haemagglutination test for the HPA-1a antigen on platelets. The scFv has also been expressed as a full-length human IgG antibody by ligation to the constant domains of human IgG1, and this antibody has been used either fluorescently labelled or enzyme labelled in other diagnostic tests for the HPA-1a antigen on platelets.

Anti-prion recombinant antibodies

‘Null’ recombinant antibodies

A range of monoclonal antibodies to human recombinant prion proteins has been produced by immunizing prion knockout mice. Selected antibodies have been developed into a diagnostic test for bovine spongiform encephalopathy using homogenized bovine brain post-mortem. Work is currently underway to try to increase the sensitivity of the assay to make it suitable for screening human blood for vCJD. It has been shown that these mouse monoclonal antibodies can prevent the spread of vCJD prion disease in a mouse

Using site-directed mutagenesis, the Fc domains of human IgG antibodies have been mutated to have as little biological function as possible. Recombinant anti-D and anti-HPA-1a antibodies have been produced with this ‘null’ Fc region. In vitro studies have shown that these ‘null’ antibodies can effectively compete with clinically significant antibodies, and prevent them causing immune destruction of red cells and platelets respectively. A clinical trial in male volunteers showed that the ‘null’ anti-D protected D-positive red cells

Human recombinant anti-varicella-zoster immunoglobulin

411

Chapter 35

from clearance by anti-D in vivo. The aim is to see if the ‘null’ anti-HPA-1a antibody can be administered to HPA-1a-negative pregnant women who are carrying HPA-1a-positive fetuses and prevent neonatal alloimmune thrombocytopenia by crossing the placenta and competing with maternal anti-HPA-1a that can cause destruction of fetal platelets. Secretory IgG

Intravenous immunoglobulin is a pooled blood production used to treat patients with primary immune deficiency throughout their lives. Whereas the product does protect from fatal infection, many patients still suffer from recurrent infections of the mucosal surfaces, presumably because intravenous immunoglobulin does not provide primary immune defence at these surfaces. Experiments have shown that antibodies to the polymeric immunoglobulin receptor (normally responsible for the transport of polymeric IgA and IgM to the mucosal surfaces) can trigger transport across epithelial cells. scFv specific to the receptor have been produced by panning phage display libraries and also shown to trigger transport. Bispecific Fv constructs are being produced which have specificity for the receptor and the Fc domain of IgG. In vitro studies have shown that such a bispecific recombinant antibody molecule can promote transport of IgG across epithelial cells in culture. A mouse model is being developed to study the efficacy of this approach in vivo. Recombinant phenotyping reagents

Many monoclonal human IgG antibodies have been produced to blood group antigens, but these require use in enzyme or antiglobulin techniques that are not suited to high-throughput automated blood grouping machines. By cloning the variable regions of these antibodies, it is possible to ligate them to the constant domains of human IgM antibodies, and express hybrid recombinant molecules in myeloma cells. These antibodies are highly potent because they are hexameric rather than pentameric structures, and they combine the high affinity of the affinity-matured IgG antibodies with 412

the polymeric structure of primitive hexameric IgM. They are very potent direct agglutinins that can readily be used in automated blood grouping machines.

Recombinant antigens Detection and identification of clinically significant blood group, platelet and granulocyte antibodies currently relies on the availability of high-quality antibody screening and identification cells that cover all clinically significant antigens, and carry them in combinations such that the specificity of antibodies can be deduced. The quality of these panels of cells is critical, and their variability has been shown in UK National External Quality Assessment Scheme exercises to be the main cause of error in the detection and identification of antibodies. Quantitation of antibodies during pregnancy is carried out using titration or autoanalyser technology, both of which show high levels of variation, such that it is difficult to set levels at which clinical action is required. Some studies have used enzyme-linked immunosorbent assay techniques to try to also determine the subclass of the antibody, in case that might aid clinical judgement, but this is not widely used. Most of the relevant antigens have been sequenced and cloned. Some have been inserted in expression vectors and expressed in the membranes of in vitro cultured cells, e.g. expression of the Rh protein in K562 human erythroleukaemia cells. For some antigens it is possible to amplify just the extracellular domain of the protein that carries the antigen, ligate this to a fusion partner protein or tag (such as FLAG), and express a soluble recombinant protein that carries antigenic activity. This has been demonstrated for the Kell, Lutheran, Duffy, MNSs and Cartwright red cell antigens and HPA-1a and HPA-1b platelet antigens. More work is underway to produce further antigens. The target of this work is to be able to produce microarrays of recombinant antigens which could then be used for high-throughput antibody screening, identification, subclass determination and quantitation of antibodies in transfusion recipients

Recombinant antibodies

and pregnant women. Such recombinant antigen microarrays have already been developed and used to study autoantibodies.

Recombinant cytokines Erythropoietin (EPO) and thrombopoietin (TPO) have both been produced as recombinant proteins and licensed for therapeutic use for the treatment of anaemias and thrombocytopenias respectively. Improvements have been made to recombinant EPO by engineering different glycosylation of the recombinant protein. Darbepoetin-alfa has two amino acid substitutions engineered into the native molecule, which result in a hyperglycosylated molecule. It has a prolonged half-life in plasma and increased biological activity. Similarly, recombinant TPO is available as a molecule with identical amino acid sequence to naturally occurring TPO, or as a pegylated recombinant megakaryocyte growth and development factor, which is a nonglycosylated molecule consisting of the first 163 amino acids of TPO and is coupled to polyethylene glycol. Several studies have reported red cell aplasia due to autoantibodies made to the recombinant erythropoietins administered, and thrombocytopenia due to autoantibodies induced by administration of recombinant TPO. Such development of neutralizing antibodies to endogenous cytokines after administration of recombinant growth factors occurs rarely. Mechanisms for the immune responses are not currently understood, nor is it clear whether there is a higher incidence of antibodies to the modified recombinant growth factors compared with the ‘native’ recombinant factors. The availability of recombinant cytokines and growth factors has advanced research on understanding the processes of erythropoiesis and thrombopoiesis. It is now possible to isolate stem cells from cord blood and direct their differentiation in vitro to produce mature red cells or megakaryocytes. This introduces the possibility of engineering and growing ‘designer’ blood cells in vitro for multitransfused patients for whom it is difficult to resource compatible blood.

Recombinant clotting factors Recombinant clotting factors have been successfully used for the treatment of haemophilia for several years. Recombinant protein technology has virtually eliminated transmissible disease risk from these products, such that recombinant products are the products of choice for haemophiliacs. Other non-infectious complications, including inhibitor formation, remain a concern. There is no evidence to date that the recombinant clotting factors induce a higher level of inhibitor formation than fractionated plasma products.

Recombinant haemoglobin Recombinant protein technology is also being used to engineer haemoglobin variants to act as red cell stroma-free oxygen-carrying solutions. The only recombinant haemoglobin subjected to clinical trials is expressed in E. coli and is a modified human haemoglobin tetramer cross-linked with a glycine bridge between the a-subunits.

Conclusions Recombinant protein technology has rapidly advanced over the last 20 years, and we are now starting to see the routine use of recombinant proteins in transfusion medicine. Recombinant proteins will probably totally replace clotting factors and specific immunoglobulins that are currently produced from fractionated pooled plasma. However, it is unlikely that recombinant products will replace intravenous immunoglobulin or albumin. Intravenous immunoglobulin works because of its broad specificity: it would be very difficult/impossible to mimic this successfully with a recombinant product. However, a recombinant product may be able to improve the efficacy of intravenous immunoglobulin by targeting part of it to the secretions to provide primary immune defence. Albumin could be produced as a recombinant protein, but this is unlikely to be economically viable compared with the ease of production 413

Chapter 35

from plasma. Only evidence of disease transmission by plasma-derived albumin could drive the production of recombinant albumin. Further specific recombinant immunoglobulins are being produced that are not currently available as blood products (anti-HCV and anti-vCJD) and the efficacy of these needs to be investigated in clinical trials. Haemopoietic growth factors have also been produced by recombinant technology and used to treat anaemia and thrombocytopenia. Blood group antigens are now available as recombinant molecules, such that we may no longer need to use red cells, platelets and granulocytes for antibody screening, identification and quantitation. Various recombinant haemoglobin molecules have been produced and these are currently undergoing clinical trials. They may be useful as blood replacements in particular surgical situations but are unlikely to replace most applications of blood transfusion. The most likely application of recombinant protein technology that may one day replace donated blood for transfusion is the use of recombinant growth factors to grow ‘designer’ blood cells from stem cells in vitro.

Summary Any protein that has a known DNA sequence can be expressed as a recombinant protein. cDNA is inserted into a virally derived expression vector and transfected into host cells. Recombinant proteins can be expressed in bacteria, yeasts, plants, insect cells, mammalian cells or in transgenic animals. Novel proteins that do not occur in nature can be engineered. Recombinant antibodies

414

have been produced and may replace specific immunoglobulins currently derived from fractionated plasma. Novel bispecific and null antibody molecules have the potential to form novel therapeutics. Recombinant antigens, growth factors, clotting factors and haemoglobins have also been produced.

Further reading Bretthauer RK. Genetic engineering of Pichia pastoris to humanize N-glycosylation of proteins. Trends Biotechnol 2003; 21: 459–62. Hesse F, Wagner R. Developments and improvements in the manufacturing of human therapeutics with mammalian cell cultures. Trends Biotechnol 2000; 18: 173–80. Kato T, Miyasaki H. Therapeutically induced autoantibodies in patients treated with recombinant hematopoietic growth factors: a brief summary. Curr Pharm Des 2003; 9: 1129–32. Ma JK-C, Drake PMW, Christou P. The production of recombinant pharmaceutical proteins in plants. Nat Rev Genet 2003; 4: 794–805. Macdougall IC. Erythropoietin and kidney failure. Curr Hematol Rep 2003; 2: 459–64. Robinson WH, DiGennaro C, Hueber W et al. Autoantigen microarrays for multiplex characterisation of autoantibody responses. Nat Med 2002; 8: 295–301. Schlesinger KW, Ragni MV. Safety of the new generation recombinant factor concentrates. Expert Opin Drug Saf 2002; 1: 213–23. Scott ML. Monoclonal anti-D for immunoprophylaxis. Vox Sang 2001; 81: 213–18. Watkins NA, Ouwehand WH. Introduction to antibody engineering and phage display. Vox Sang 2000; 78: 72–9. Winslow RM. Alternative oxygen therapeutics: products, status of clinical trials and future prospects. Curr Hematol Rep 2003; 2: 503–10.

Chapter 36

Blood transfusion in a global context David Roberts, Jean-Pierre Allain, Alan Kitchen, Stephen Field and Imelda Bates

Introduction 17% of the world’s population has access to 60% of the global blood supply Inequality in the provision of ‘safe blood’ around the world mirrors the unequal distribution of almost all other resources crucial for effective health services or indeed for health itself. Unfortunately, in many countries, providing safe blood is made more difficult by lack of donors and the high frequency of transfusion-transmissible infections. At the same time, the problems posed by the poor supply of safe blood are compounded by the frequent need for urgent life-saving transfusions in childbirth and in children with malaria. The purpose of this chapter is not to guide those developing transfusion services in less affluent countries but to inform a wider audience of the problems faced in the development of effective transfusion services in these countries. A secondary aim is to stimulate some debate and analysis of the problems faced by transfusion services globally. Finally, a short chapter must be selective and our choice of topics and examples and their solutions reflects our own experience in SouthEast Asia and tropical Africa south of the Sahara (or sub-Saharan Africa). Blood safety

An unsafe blood supply is costly in both human and economic terms. Transfusion of infected blood not only causes direct morbidity and mortality in the recipients, but also has an economic and emotional impact on their families and communities and undermines confidence in modern healthcare.

Those who become infected through blood transfusion are infectious to others and contribute a significant secondary wave of iatrogenic infections. Investment in safe supplies of blood is therefore a cost-effective investment for every country, even those with few resources. The World Health Organization (WHO) has identified four key objectives of all strategies used to ensure that blood is safe for transfusion. • Establish a coordinated national blood transfusion service that can provide adequate and timely supplies of safe blood for all patients in need. • Collect blood only from voluntary nonremunerated blood donors from low-risk populations and use stringent donor selection procedures. • Screen all blood for transfusion-transmissible infections and have standardized procedures in place for grouping and compatibility testing. • Reduce unnecessary transfusions through the appropriate clinical use of blood, including the use of intravenous replacement fluids and other simple alternatives to transfusion, wherever possible. WHO also emphasizes that effective quality assurance should be in place for all aspects of the transfusion process, from donor recruitment and selection, through infection screening, blood grouping and blood storage to administration to the patients and clinical monitoring for adverse reactions. It is axiomatic that transfusion medicine is a distinct and multidisciplinary sector of the health service and should be incorporated into all national health plans. It is therefore the responsibility of governments to develop policies and legislation that will facilitate the development of a national transfusion service and ensure that the 415

Chapter 36

blood transfusion process and its associated quality assurance programmes are of a high standard. WHO has simply and succinctly recommended the structure of national blood transfusion services. It suggests that at the national level the transfusion service should have a medical director, an advisory committee and clear national transfusion policies and strategies with the appropriate statutory instruments to ensure the national coordination and standardization of blood testing, processing and distribution. However, less than 70 out of the 191 member states meet WHO’s recommendations for a national blood programme. In Africa, in 2002 WHO estimated that among the 46 member states in the African continent, only 14 had a national blood policy and just six had a policy to specifically encourage and develop a system of voluntary non-remunerated donation. It is worthwhile reflecting on why the development of national transfusion services has been delayed. One reason may be that the emphasis on primary healthcare over the last 25 years has diffused interest in hospital-based curative medicine. A second reason may have been the high cost of blood transfusion in relation to disposable income and healthcare budgets. The average annual income in sub-Saharan Africa is in the range of $400–1000 and a unit of blood costing $10–20 is an expensive commodity in relation to the annual per-capita budget for healthcare in these countries. Nevertheless, blood transfusion for severe malarial anaemia and severe haemorrhage can be lifesaving and here the cost of transfusion is well within the generally accepted cost–benefit range for health interventions in poorer countries of $1 per year of life saved. However, it is also true that many of these countries cannot afford sustainable basic healthcare systems, let alone develop a national transfusion service, which supplies blood according to the standards established by affluent countries, where a unit of blood now costs in the region of $200. To prepare enough safe blood in a sustainable fashion, African countries need to develop their own ways to produce affordable safe blood. Uncritical adoption of external advice and models

416

may lead to unsustainable and inappropriate solutions. What then are the transfusion services in Africa and what are the consequences for the supply of safe blood? Organization of transfusion services in sub-Saharan Africa

African countries have developed a variety of systems to try to achieve a sustainable safe blood supply. These vary from large, modern, national blood centres to locally organized donor programmes for isolated district healthcare facilities. A minority of countries have invested significant resources in transfusion services, often with financial support and advisers from European governments, USAID or non-governmental organizations (NGOs), including Red Cross, Red Crescent, Family Health International and the Safe Blood for Africa Foundation. In these countries there has been a commitment to establishing centralized systems based on a model of sustainability and effectiveness, similar to that used in wealthy nations. These centres typically collect over 10 000 units a year, use automated equipment and produce some components. Blood donor recruitment and screening and processing of donated blood are carried out in specifically designed premises away from the hospitals where blood is transfused. The cost of a unit of blood produced by these centres is of the order of $40 per bag. However, as we have seen, many countries in sub-Saharan Africa do not operate a centralized transfusion service. Each hospital recruits donors for its own patients and processes the blood for transfusion itself. These hospitals often handle less than 1000 units a year and experience difficulties in standardization, quality assurance and in maintaining supplies of high-quality reagents. Recruiting voluntary donors from the community is complex and expensive and depends on regular education programs, venesection teams, vehicles and cold storage. Because of these difficulties, the majority of donors in poorer countries are ‘replacement’, not volunteer, donors, although many local initiatives exist to create small local panels of tested volunteers who may be called

Global blood transfusion

upon to donate at short notice. In some regional schemes, anti-human immunodeficiency virus (HIV), hepatitis B surface antigen (HBsAg) and anti-hepatitis C virus (HCV) are screened before donation with high-performance rapid tests so those blood bags are not wasted. Furthermore deferred donors can be identified, informed and counselled, consequently decreasing the prevalence of viral markers in repeat volunteer donors. In 2002, in Africa as a whole, WHO estimated that over 60% of blood originated from replacement/family donors. In sub-Saharan Africa the proportion of blood derived from replacement donors is certainly higher. These replacement donors should be family members, but are too often professional donors paid directly by the family, who are asked to provide blood for their relatives in the hospital. Cultural taboos and lack of education about donating blood (e.g. ‘men will become impotent if they donate blood’; ‘HIV can be caught from the blood bag needle’) makes relatives reluctant to donate so they may choose to purchase blood from ‘professional’ donors. It is worth noting that similar problems faced widespread acceptance of blood donation when Percy Oliver and Geoffrey Keynes began to establish the first blood banks of volunteer donors in London over 70 years ago. Nevertheless, the local systems allow many patients to survive serious illness and are often maintained by dedicated staff in difficult circumstances. However, even with the best input from local staff these district services experience problems of supply, safety and cost of blood. Supply of blood

Patients in poorer countries usually present late in the course of their disease and the delays and lack of stored blood inherent in the replacement donor system mean that patients may die before a blood transfusion can be organized. By the time a donor has been found, screened and venesected, and the blood transfused into the patient, several hours or even days can elapse. A survey of the blood supplied by a dedicated district service in East Africa showed that the average delay in sourcing blood

for children with severe malarial anaemia was 6 h. Anecdotal evidence suggests that in some areas, and occasionally in many areas, blood may not be available at all. Finally, locally based services make it difficult to separate blood, even into simple fractions such as red cells and plasma, in order to provide specific components if needed. Safety of blood

Local blood transfusion systems encounter many problems, almost always centred around lack of funding, including inconsistency and high prices of screening tests, breakdown of the cold chain mostly related to frequent power cuts, and poorly trained staff. Blood frequently has to be collected in small hospital-based units often with no dedicated staff or specifically allocated budget. The WHO review estimated in 2000 that 25% of the blood in sub-Saharan Africa was untested for antiHIV and that blood transfusion was the origin of 5–10% of new HIV infections. HBsAg was screened in 50% of donors or donations and only 19% were screened for anti-HCV. Most countries in sub-Saharan Africa do not screen for human Tcell leukaemia virus (HTLV) since the prevalence is low (